From tim_one at email.msn.com  Wed Jul 28 22:27:57 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 28 Jul 1999 22:27:57 -0400
Subject: 1.5.2 for: else:
In-Reply-To: <379F6E37.777322BF@compaq.com>
Message-ID: <000d01bed969$f89ec220$71a22299@tim>

[/F, on the sparkling consistency of for/else, while/else, if/else]
> what else did you expect from Guido?

[Greg Ewing]
> I would have expected Guido to shriek in horror
> and refuse to even consider such a semantically
> ambiguous construct.
>
> It seems quite out of character for both him
> and Python.

Indeed, if for/else hadn't been in Python from the start, anyone suggesting
it would get flamed without mercy!  I happen to like it, but it's Guido-like
only in its inscrutablitily at first sight <wink>.   The only Pythonism less
Pythonic is that e.g.

    4 < "3"

silently returns true.

although-that-would-be-even-less-pythonic-if-it-returned-false-ly y'rs  -
tim





From loredo at spacenet.tn.cornell.edu  Fri Jul  2 16:17:40 1999
From: loredo at spacenet.tn.cornell.edu (Tom Loredo)
Date: Fri, 02 Jul 1999 16:17:40 -0400
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773CA67.7057F797@quasar.ipa.nw.ru> <3773D55B.8C966174@spacenet.tn.cornell.edu> <slrn7naiqb.8ll.zeitlin@seth.lpthe.jussieu.fr> <7li9t6$j2k$1@srv38.cas.org>
Message-ID: <377D1E64.C0DC11D0@spacenet.tn.cornell.edu>

lvirden at cas.org wrote:
> 
> According to Vadim Zeitlin <zeitlin at dptmaths.ens-cachan.fr>:
> :
> : The sad truth is that macs are not very common nowadays. I would like to help
> :develop wxMac, but I don't have one (nor at home nor anywhere near me in the
> 
> I wonder why various industry rags have proclaimed during the past
> 10 months about Apple's G3 machines outselling other vendor
> machines.  ...

Make me a third.  About a million iMacs have been sold in less than
a year, and although the Mac market share is only around 10% right
now (that's a huge number of Macs, by the way!), the fraction
of all personal computers *in use* that are Macs is actually about 20%.
It is simply way off the mark to believe that Macs are not very common
nowadays.  I spend most of my time at a Unix box, and work on all
three major platforms regularly (Unix, Mac, Win32), so please don't
dismiss this as just another Mac-only user mouthing off.  It's a great
platform, and I simply wouldn't want to see python GUI development
ignoring it.

-Tom Loredo



From Andrew.Cooke at p98.f112.n480.z2.fidonet.org  Thu Jul  1 08:16:00 1999
From: Andrew.Cooke at p98.f112.n480.z2.fidonet.org (Andrew Cooke)
Date: Thu, 01 Jul 1999 13:16:00 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <0000160b@bossar.com.pl>

From: Andrew Cooke <andrew at andrewcooke.free-online.co.uk>

In article
<64F6457BE0585BB3.1CF2F67BC8A680BD.A19A9FEE23BC9E58 at lp.airnews.net>,
  mikael at pobox.com (Mikael Lyngvig) wrote:
> On 26 Jun 1999 21:37:12 GMT, zeitlin at seth.lpthe.jussieu.fr (Vadim
> Zeitlin) wrote:
>
> > While I don't think wxPython can replace TkInter as a de-facto GUI
standard
> >for Python I still think that it has several advantages compared to
TkInter.
> >Instead of saying how much better, prettier, faster (this one has
really made
> >me smile, BTW: I don't think wxWin is currently very fast, if it has
been
> >cited as its advantage it must really mean that TkInter is extremely
slow.
>
> Well, TkInter is extremely slow, but not because of the Python part.
> Even the Tk widget tour, included with Tk, is so slow that you almost
> think "Please, draw the next character - you can do it." whenever a
> window with some text needs to be redrawn.
>
> In contrast, the wxPython demo is so fast that you probably wouldn't
> realize it is implemented in Python, unless you're used to do GUI
> programming yourself.
>
> One important point is that I probably experience Tk(Inter) as slow
> because I'm used to Win32 desktops with very fast graphics hardware
> and no X/Windows style client/server architecture.  I imagine that
> many X/Windows users experience Tk(Inter) as fast enough.

I'm confused.  I'm sitting at a moderate NT box (266MHz, 96Mb) - my
work machine, and running the Tk demos.  There is no problem at all -
they are instantaneous.

Is this problem to do with memory?  Or is it a W95 thing?  If so,
is it due to thread scheduling?

Andrew



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From htrd90 at zepler.org  Fri Jul  9 17:22:25 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Fri, 09 Jul 1999 21:22:25 GMT
Subject: win32 python as a dll?
References: <1280583297-81012085@hypernet.com> <37863B4F.FB1780F3@enews.nrl.navy.mil>
Message-ID: <3787648e.2225295@news.freeserve.net>

Scott Wolford <wolford at enews.nrl.navy.mil> wrote:

>I guess I didn't pose my question properly. Under Unix, can you not link with a
>debug foo.so at link-time, then at run-time load an optimized version, as long
>as the code used to generate the two is roughly the same? It seems to work for
>me. Why can't you do the same thing under Windows?

msvcrt.dll and msvcrtd.dll are not even roughly the same. There are
differences in the data structures behind malloc, fopen, etc. They are
given different names to prevent them getting mixed up.

If optimisation level was the only difference then you would not need
separate .lib files.



Toby Dickenson



From catlee at my-deja.com  Tue Jul  6 15:47:13 1999
From: catlee at my-deja.com (catlee at my-deja.com)
Date: Tue, 06 Jul 1999 19:47:13 GMT
Subject: Getting contents of Text / Entry widgets
Message-ID: <7ltmfu$nbt$1@nnrp1.deja.com>

Hi,

I'm writing a little utility that uses a Tkinter Text widget (and an
Entry widget).  I want to perform an action for every key that is
entered.  So I do something like:

def callback(event):
   print event.widget.get()

widget=Entry()
widget.bind("<Key>",callback)

The problem with this is that the output generated by callback is
always one character behind.  I assume this is because when the event
is generated, the character you've typed hasn't been entered into the
buffer yet.  I could modify callback to look something like:
def callback(event):
   print event.widget.get() + event.char

But this doesn't handle backspace and delete characters well.

Is there a nice way of retrieving the contents of a Text / Entry widget
every time the widget's text buffer is changed (as opposed to whenever
a key is pressed)?

Thanks,
Chris


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From emile at fenx.com  Mon Jul 26 10:04:04 1999
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 26 Jul 1999 07:04:04 -0700
Subject: How to use a variable form main script in a module.
References: <7nhmt5$su0@enews3.newsguy.com>
Message-ID: <sYZm3.4214$VL2.443976@news.direcpc.com>

Hi Peter,

You need to fully specify the module variable name in your main program.
For example, if an imported module defines x as an unbound (quasi)class
variable:

>>> x = 3
>>> import testx
>>> x
3
>>> testx.x
15

HTH
--

Emile van Sebille
emile at fenx.com
-------------------


Peter Torstenson <p.t at iname.com> wrote in message
news:7nhmt5$su0 at enews3.newsguy.com...
> This is probaly a silly question from a newbie.
> In my main script I put a value for to a variable called "x".
> Now I want to use the "x" variable in a module.
> The module is imported into the main script. How to do this?
>
> I've been trying to find out how to do it in tutorials an the
reference
> manual but I can?t find anyting written about this. Perhaps because
it's a
> to silly question.
>
> /Peter Torstenson
>
> Rio de Janeiro
>
>





From C.Kotsokalis at ece.ntua.gr  Tue Jul 13 17:49:07 1999
From: C.Kotsokalis at ece.ntua.gr (Constantinos A. Kotsokalis)
Date: 13 Jul 1999 21:49:07 GMT
Subject: problem with SocketServer -- address already in use
References: <931899928snz@vision25.demon.co.uk>
Message-ID: <slrn7ond1f.jk.C.Kotsokalis@ppp032.dialup.ntua.gr>

On Tue, 13 Jul 99 21:05:28 GMT, Phil Hunt wrote:
>I'm trying to use SocketServer.py to write a TCP server. I ran
>my program with ``python serv.py'', which went into an infinite
>loop (which is what it should do). I exited my program with ^Z,
_____________________________________________________________^^
>and tried to start it up again. I got an error message:
>
>socket.error: (98, 'Address already in use')

ctrl-Z suspends your server, so the port is still occupied.
Use ctrl-C to interrupt it and wait, or use setsockopt as Gordon
suggests and ctrl-C. In any case, you can't use that port again
with a suspened server.

  --C

-- 
Constantinos A. Kotsokalis || C.Kotsokalis at ece.ntua.gr
National Technical University of Athens - GREECE
Electrical and Computer Engineering Department
"Bus error -- driver executed."



From billtut at microsoft.com  Thu Jul 22 21:46:04 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Fri, 23 Jul 1999 01:46:04 GMT
Subject: Non-ODBC MS Sql Connect
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F98@RED-MSG-50>

You can always use ADO from Python on NT.

Quick example:
import win32com.client
import win32com.client.makepy
import win32com.client.genpy
import win32com.client.gencache
adolib = win32com.client.gencache.EnsureModule(
	'{00000200-0000-0010-8000-00AA006D2EA4}', # TypeLib GUID for ADO
	win32con.LANG_NEUTRAL, # LCID
	2, # Major version #
	0, # Minor version #
	win32com.client.genpy.GeneratorProgress()) # Silent progress
indicator.
# ALL loaded typelibrary constants...
k = win32com.client.constants
cxnSQLServer = adolib.Connection()
cxnSQLServer.CursorLocation = k.adUseNone
cxnSQLServer.Provider = "SQLOLEDB"
cxnSQLServer.Open("Server=MachineName;Database=master;UID=user;PWD=apple")

rs = adolib.Recordset()
rs.Open("select * from sysobjects", cxnSQLServer, k.adOpenForwardOnly, 
	k.adLockReadOnly)
# etc.....


Bill




> -----Original Message-----
> From: Roberts, Robert J [mailto:robert_j_roberts at rl.gov]
> Sent: Thursday, July 22, 1999 5:33 PM
> To: python-list at cwi.nl
> Subject: Non-ODBC MS Sql Connect
> 
> 
> How to do I make a non-odbc connection to a MS SQL Server?
> 
> Am running Python Windows NT/IIS.
> 
> Thanks,
> 
> 



From fdrake at cnri.reston.va.us  Wed Jul 28 12:12:42 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Wed, 28 Jul 1999 16:12:42 GMT
Subject: API question -- decref after adding to dictionary?
In-Reply-To: <rpu849$1$37nspbi$7gc@news.supernews.com>
References: <rps64v$0$37nspbi$7gc@news.supernews.com>
	<379EBC8B.E36F020B@lemburg.com>
	<rpu849$1$37nspbi$7gc@news.supernews.com>
Message-ID: <14239.11258.615882.396388@weyr.cnri.reston.va.us>

Rich Salz writes:
 > The API documentation doesn't seem to say what the functions do to
 > object references.  Am I missing it, or do I really have to read the

Rich,
  This is a matter of completeness; the API docs just aren't "there"
yet in this regard, and the time to devote to adding the information
hasn't been available.
  In the meanwhile, I'd certainly be glad to receive patches which
added the information to the API reference.
  I do hope that this isn't a deciding factor as to whether you use
Python.  The discussions of reference counting in the API and
Extending & Embedding manuals do try to explain the approaches taken
in Python and reduce the pain significantly.  I've written a number of 
C extensions myself, including one very early in my experience with
Python, and haven't had a lot of difficulty with it.  (Note: I did
*not* say I haven't had any difficulties!)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From pepper at xcs.tut.fi  Tue Jul 27 08:27:00 1999
From: pepper at xcs.tut.fi (Petteri)
Date: Tue, 27 Jul 1999 12:27:00 +0000
Subject: upload file with cgi
References: <7mro7g$kr5@newsops.execpc.com> <40kk3.351$ui4.121238@news.shore.net>
Message-ID: <379DA594.BB2FD011@xcs.tut.fi>

Hi,

This must be a FAQ question, but I'm wondering where the file disappears
from FieldStorage? My own cgi-upload script used to work but now what
happens is that the FieldStorage is empty. Is this just some silly mistake
or does it have something to do with netscape?

<FORM ACTION="/cgi-bin/upload.py" method=POST enctype=multipart/form-data>
<p>Upload a file!</p>
<SELECT NAME=dest>
<OPTION value=koulutus selected>koulutus
<OPTION VALUE=tiedotus>tiedotus
</SELECT>
<INPUT TYPE="file" name="userfile" SIZE="30">
<INPUT TYPE=SUBMIT value=SEND!>
</FORM>


An excerpt from the script:

        form = cgi.FieldStorage()
        print form
        sys.stdout.flush()
        fileitem = form['userfile']
        if fileitem.file and form.has_key('dest'):
            Upload(fileitem, form['dest'].value)

and what 'print form' prints if I send the form with a file selected:
FieldStorage(None, None, [])

and, of course, an exception is raised
Traceback (innermost last): File "/usr/lib/cgi-bin/upload.py", line 53, in
Main fileitem = form['userfile'] File"/usr/lib/python1.5/cgi.py", line 907,
in __getitem__ raise KeyError, key KeyError: userfile

I'm confused. Please help.

Sami




From billtut at microsoft.com  Wed Jul 21 10:36:23 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Wed, 21 Jul 1999 14:36:23 GMT
Subject: VC++ 6.0 larger exe files explained
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F88@RED-MSG-50>


> From: Gordon McMillan [mailto:gmcm at hypernet.com]
> 
> 
> Robin Becker wrote:
> > <rant>
> > As part of the Wintel push to take over the world it seems as though
> > VC++ 6.0 will now only generate files with sections 0x1000 (4k) or
> > longer. Under 5.0 and before the linker would allow much smaller exe
> > file sections. So python.exe which used to be 5k becomes 20k.
> > Needless to say this should also apply to dll's etc and appears to
> > offer no obvious advantage. Bill has never been other than cavalier
> > with our space time resources. </rant>
> 
> Ay yi yi. The whole PE header is full of convolutions so the loader 
> can take a packed disk image and align the sections on the fly.
> 
> I guess this "feature" designed to "optimize" load time. Or maybe MS 
> has been investing in drive manufacturers...
> 

I believe this is indeed the case. (optimizing load time, and not the drive
manufacturer bit...) I vaguely recall this having something to do with Win98
and FAT32, and it not making much difference for NTFS for other reasons.
Unfortunately I can't seem to find the exact details atm. Blech.

Bill



From bh at cellware.de  Fri Jul 16 06:38:45 1999
From: bh at cellware.de (bernard)
Date: Fri, 16 Jul 1999 10:38:45 +0000
Subject: Accessing an object name ?
Message-ID: <378F0BB5.9AD878D@cellware.de>

Is there any way of mapping from an object id to an object name
so that given the object declaration:

obj1 = myClass()

I can, within an instance method, access the object name 'obj1' ?

bernard.
-- 
 _________________________________________________________________
                                               __
                                             _|__|
 Bernard Hayes                     cellware |_|__
                                              |__| broadband

 Email  bh at cellware.de                          Cellware Broadband
 Phone  (+49 30) 670 08-114                     Rudower Chaussee 5
 Fax    (+49 30) 670 08-108                     12489 Berlin
 WWW    http://www.cellware.de                  Germany
 _________________________________________________________________



From conrad at vanir.srl.caltech.edu  Sat Jul 10 16:08:37 1999
From: conrad at vanir.srl.caltech.edu (Conrad Steenberg)
Date: 10 Jul 1999 20:08:37 GMT
Subject: Printing returned object in embedded app
References: <378694FC.D857F68A@srl.caltech.edu> <6wxh3.1472$6M6.450919@news.shore.net> <slrn7oep57.a57.conrad@vanir.srl.caltech.edu>
Message-ID: <7m9tnq$6qg$69@rks1.urz.tu-dresden.de>

On 10 Jul 1999 18:24:33 GMT, Conrad Steenberg <conrad at vanir.srl.caltech.edu>
wrote:
>On Sat, 10 Jul 1999 01:40:18 GMT, Michael P. Reilly <arcege at shore.net> wrote:
>>
>>Try:
>>  PyObject_Print(object, stdout, Py_PRINT_RAW);
>
>Thanks! 
>
>Conrad
>-- 
Hmmm... Sorry to reply to my own message, but it seems like PyRun_String 
returns only the Py_None object! 

How can I get the result of a calculation, say
obj=PyRun_String ("1+1", start, globals, locals);
which should be a numeric object? 

The behaviour I'm trying to get is the same as running python on a terminal:
when an instruction have a printable result, I want to print it.

Cheers

Conrad

-- 



From Haimo.Zobernig at cern.ch  Thu Jul  8 10:28:14 1999
From: Haimo.Zobernig at cern.ch (Haimo G. Zobernig)
Date: Thu, 08 Jul 1999 16:28:14 +0200
Subject: os.kill on NT
Message-ID: <3784B57E.C0A57A@cern.ch>

Some time ago (maybe not all that long) someone posted
a small function implementing os.kill on Windows NT, using the PythonWin
extensions. I didn't save it then and now I can't seem to find that post
with either dejanews or egroups. Since I'm not very familiar with NT
I'm having trouble writing it myself. 
Could somebody repost or email it?

Many thanks!

  - Haimo

-- 
Haimo G. Zobernig                                  tel   +41 22 7677361
Univ. of Wisconsin                                 fax   +41 22 7678370
c/o CERN Bld. 32 R-A02
1211 Geneve 23, Switzerland                email Haimo.Zobernig at cern.ch



From skip at mojam.com  Sun Jul 25 13:22:03 1999
From: skip at mojam.com (Skip Montanaro)
Date: Sun, 25 Jul 1999 12:22:03 -0500 (CDT)
Subject: Access dictionary in order?
In-Reply-To: <roy-2507991300370001@mc-as01-p61.med.nyu.edu>
References: <roy-2507991300370001@mc-as01-p61.med.nyu.edu>
Message-ID: <14235.18248.924848.52652@dolphin.mojam.com>

    Roy> I've got a dictionary of items which I will sometimes want random access
    Roy> to, but sometimes I'll want to be able to step through the items in a
    Roy> fixed order as well.  Any good way to do that?

You demonstrated the best way.  There is no need to save the key list unless
the dictionary is very large though.  Just grab 'em and sort 'em when you
need:

    keylist = dict.keys()
    keylist.sort()
    for key in keylist: do_stuff(dict[key])

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From claird at starbase.neosoft.com  Wed Jul 21 10:37:56 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Jul 1999 14:37:56 GMT
Subject: Distributed application architectures (was: [Q] In-Browser technology)
References: <1279557217-15543582@hypernet.com>
Message-ID: <8F18075464DB9982.F6A8F284AFA328EB.A0A519CF82F3C805@lp.airnews.net>

In article <1279557217-15543582 at hypernet.com>,
Gordon McMillan <gmcm at hypernet.com> wrote:
			.
			.
			.
> Despite all the hype, the only things you can do reasonably portably 
>in browsers (NS & IE) is vanilla JScript / JavaScript and vanilla 
>Java applets. Even there you will almost certainly have to detect 
Not that that's to be sneered at,
in general.  Vanilla JavaScript
and vanilla Java applets can do
more than most programmers will
touch in their whole lifetimes.
On the other hand, yes, there's a
LOT that they leave out.  'Depends
on what ones needs are--the maxim
Gordon and I invoke in all situa-
tions, anyway.
>which browser you're in, and code accordingly. You may even have to 
>detect which version of which browser. And things get even worse as 
>soon as you need to deal with the browser's security. And it's very 
>difficult to do anything useful without running up against a security 
>rule.
I confirm Gordon's melancholy
cautions.
>
> If you have long term relationships with your clients (so you can 
>ask them to install something) and if you only have to deal with a 
>limited (the smaller the better) number of browser / versions, you 
>can do some neat things. (But, as Cameron points out, why bother with 
>a 15 meg browser when they could just install an app - HTML forms are 
>easy to write, but don't make particularly good GUIs.)
It varies.  HTML forms are *perfect*
for some situations, and risibly in-
adequate for others.  The anti-anti-
... backlash against HTMLization has
now ascended to an uncountable height,
as near as I can tell.  My personal
bias is to assume vanilla Webification
is enough, and get customers to tell
me for what more they truly are willing
to pay.
>
> This whole noise about browsers-as-thin-clients <choke> was started 
>by the Java folks. Well, at this point, according to the same folks, 
>applets are "out" and server-side is "in". And they're right - Java 
>makes much more sense on the server than on the client. You can 
>blame it on the browsers, but the JVM spec pushes all the hard 
>work into the JVM implementation, so it should be no surprise that 
>the promise was hollow. 
Thin clients--ohboy.  I've been sketching
a whole essay I want to write on that
subject.  Whatever else we say, though,
it's really, really valuable to be able
to assume a Navigator or IE is sitting
"out there", on a client's desktop, pro-
foundly flawed though they are.
>
>On the server, Zope makes even more sense.
Isn't *that* the truth!  (That's Hoosier
for, "I agree with you heartily.")
>
> I found Juice very intriguing when I looked at around 5 yrs ago. 
>It's not encouraging that they're still using the same wire-frame 
>demo they used back then.
I appreciate your understatement.  No,
that is far from encouraging.
			.
			.
			.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From roy at popmail.med.nyu.edu  Thu Jul  8 12:39:50 1999
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Thu, 08 Jul 1999 12:39:50 -0400
Subject: Two Pythons talking to each other?
References: <199907072203.AAA09601@axil.hvision.nl> <378465E9.7762BE9A@fnmail.com> <14212.51977.601241.772478@buffalo.fnal.gov> <3784CD0C.24437EFC@fnmail.com>
Message-ID: <roy-0807991239500001@qwerky.med.nyu.edu>

Per Kistler <kistler at fnmail.com> wrote:
> But does NT and Mac have loopback IP addresses?

The Mac certainly does.

I was able to ping localhost using WhatRoute, and when I start up Telnet's
built-in ftp server, I am able to connect to it by telnetting to
"localhost.med.nyu.edu 21".  Oddly enough, Fetch was unable to connect to
localhost, giving me an "Error -1", and my telnet process crashed a little
while later, but the existance of buggy software should not be taken as
evidence that Macs do not have a loopback address :-)

Much as I think evil about NT and all things Microsoft, I would be
surprised (astonished, actually) if NT didn't handle loopback.

-- 
Roy Smith <roy at popmail.med.nyu.edu>
New York University School of Medicine




From claird at starbase.neosoft.com  Tue Jul 20 19:41:52 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Jul 1999 23:41:52 GMT
Subject: Python suitability questions
References: <0CCE781E44F2C872.F528377A62B41DCA.73CAA63AD6E79314@lp.airnews.net>
Message-ID: <AFBA2663BF7BCC33.531B12EB7F7128A6.EB5C4953D3050C61@lp.airnews.net>

In article <0CCE781E44F2C872.F528377A62B41DCA.73CAA63AD6E79314 at lp.airnews.net>,
Jason Maskell <jmaskell at saskpublishers.sk.ca> wrote:
>Well, after a day or two of reading lots of newsgroup posts and
>documentation, I'm still not quite certain if Python is suitable for the
>project I have in mind. I'm rather reluctant to jump in and start fiddling
>with it without having some idea whether it's what I'm looking for or not. A
It's what you're looking for.  You're at a stage
in evaluation where it'll be cheaper for you to
jump in than to continue to poke at Python from a
distance.  This is one of the charms of Python,
as you'll learn; it invites experimentation, and
is friendly, so frequently the best answer to how
things work is to experience them for yourself.
			.
			.
			.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From wlfraed at ix.netcom.com  Sun Jul  4 17:13:31 1999
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Sun, 04 Jul 1999 21:13:31 GMT
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.i <m3btdvhrlt.fsf@localhost.localdomain> <d2Of3.1233$U5.237262@ptah.visi.com>
Message-ID: <377fc323.69785805@nntp.ix.netcom.com>

On Sun, 04 Jul 1999 18:50:49 GMT, thornley at visi.com (David Thornley)
declaimed the following in comp.lang.python:


> "Thinking Forth" is an excellent book to read.  So is Stroustrup's
> "Design and Evolution of C++".  I haven't seen a similar book about
> the Lisp way of thinking, not counting books specifically on learning
> Lisp and doing it well.  Then again, I find Common Lisp revolutionary
> enough sometimes.
>
	Unfortunately, it predates much of what is now considered to be
LISP.

	"Anatomy of LISP", John Allen, (1978 McGraw-Hill Computer
Science Series).

	While using LISP, it was started as a data structures type
course...

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <



From arnold at dstc.edu.au  Thu Jul 15 17:29:56 1999
From: arnold at dstc.edu.au (David Arnold)
Date: Thu, 15 Jul 1999 21:29:56 GMT
Subject: Calling superclass method with variable args? 
In-Reply-To: Your message of "Thu, 15 Jul 1999 11:35:10 MST."
             <378E29DD.1613913E@pixar.com> 
References: <378E29DD.1613913E@pixar.com>
Message-ID: <199907152129.HAA02590@piglet.dstc.edu.au>

-->"Steve" == Steve Johnson <slj at pixar.com> writes:

  Steve> But what if the method takes unnamed positional and keyword
  Steve> arguments. Then, how do you pass these along to the
  Steve> superclass's version of the method such that it gets the
  Steve> arguments in the same form that the subclass's method did?

so close!  the trick is the built-in function "apply(func, args [,
keywords])" has the optional keywords dictionary as a third parameter.

<http://www.python.org/doc/current/lib/built-in-funcs.html>


so,

    class Job(Task):
        def __init__(self,a,b,c, *moreargs, **morekeyargs):
            apply(Task.__init__, (self,a,b,c) + moreargs, morekeyargs)
            <whatever>
    
    aJob = Job(x,y,z,m1,m2,m3,m4,m5,foo="bar")


should work just fine,



d







From tismer at appliedbiometrics.com  Wed Jul 21 15:08:12 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Wed, 21 Jul 1999 19:08:12 GMT
Subject: [Tutor] I want to learn to hack
References: <Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org>
Message-ID: <37961A9C.60201FBC@appliedbiometrics.com>


David Ascher wrote:
> 
> On Wed, 21 Jul 1999, Gerrit Holl wrote:
> 
> > On Tue, Jul 20, 1999 at 02:31:55AM -0400, Tim Peters wrote:
> > > > I want to learn to hack
> > > So, little Roger, you want to learn how to hack?  You came to the right
> > > place!  First walk down the hall and pick up a copy of Python from Guido.
> > > He'll have you cracking bank accounts in no time <wink>.
> > >
> > > the-psa-is-the-richest-organization-on-earth-ly y'rs  - tim
> > >
> >
> > Well, I think Roger meant hack, not crack. Have a look at Eric S Raymonds
> > "How to become an hacker FAQ" at http://www.tuxedo.org/~esr/ (and some
> > further).
> >
> > Hacking is good,
> > cracking is bad.
> >
> > I also want to learn to hack, I'm learning it!
> > I *don't* want to learn to crack!!
> 
> Gerrit -- just to fill you in -- Roger was presumably either joking or
> testing a spam filter -- he is a colleague of Guido's, and knows all about
> hacking.  Whether he knows about cracking or not is not something he's
> likely to disclose in public. =)
> 
> (In other words, Tim was kidding, Roger was kidding, I'm not kidding!).

No, you're just emailing to the wrong list :-)))

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From gmcm at hypernet.com  Tue Jul 27 19:48:01 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 27 Jul 1999 18:48:01 -0500
Subject: Problem with signals & exceptions...
In-Reply-To: <2BB05D8530326D0C.420888E8906528AF.C6BF325D12E9ED90@lp.airnews.net>
Message-ID: <1279007017-12004369@hypernet.com>

Eric Lee Green writes:

> I'm having trouble catching my thrown exceptions under FreeBSD 3.2.
> I am using the Python 1.5.2 that came with FreeBSD 3.2 (well, on
> disk 1 of the <n>-disk set). The following program works properly
> under Red Hat Linux version 6.0 running a self-compiled Python 1.5.2
> (in place of the 1.5.1 which comes with Red Hat 6.0, which does not
> properly handle signals during socket i/o calls). How it works: it
> listens to port 6664. When it gets a connection, it then sets an
> alarm handler and alarm and then tries to read from the socket. If
> it does not get data within 10 seconds, it then times out, at which
> point the alarm handler throws an exception and aborts the socket
> read.  

Why not avoid the pain of signals (which have enough gotchas to 
destroy portability) and use a select with a 10 sec timeout. If 
there's nothing in the readable list when it returns, bail out.

- Gordon



From neelk at brick.cswv.com  Wed Jul  7 22:37:33 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 7 Jul 1999 21:37:33 -0500
Subject: Arguments of builtin functios?
Message-ID: <7m12td$13n$1@brick.cswv.com>

I'm working on a tiny little module that makes it easy to dynamically
examine out the call signature of functions and methods in Python.
This seems like something that would be very useful when trying out
code snippets in an interactive environment.

For native Python functions and methods, I've managed to figure this
out by by looking at the code object associated with the function, but
I'm at a loss to figure out how to do the same thing for builtin C
functions and method objects.

Any hints?


Neel




From holger at phoenix-edv.netzservice.de  Fri Jul  2 05:31:28 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Fri, 2 Jul 1999 09:31:28 GMT
Subject: Why must path be in args in os.execv?
References: <377A2590.D95D72FE@phoenix-edv.netzservice.de> <slrn7nn2bb.81d.jepler@craie.inetnebr.com>
Message-ID: <377C86F0.A95BE6B4@phoenix-edv.netzservice.de>

Thanks, Jeff.

That's a good answer.

Ciao,
Holger

Jeff Epler schrieb:
> 
> Holger,
> 
> On Unix, read the manpage for execve and friends.  You will see that the
> first argument is the path to the executable, and the subsequent arguments
> describe the arguments (including argv[0]) and optionally the environment.
> By convention, argv[0] is the basename of the executable when the
> executable is on the PATH, and the full path otherwise.  (Another unix
> convention is that when the first letter of argv[0] is "-", the shell is a
> login shell)
> 
> DOS doesn't really have execve.  It's possible that the imitation is less
> perfect than it could be.  Instead, it may do the equivalent of entering
> the string
>         string.join(argv, " ")
> in the shell, i.e., ignoring the path argument.
> 
> This looks like a shortcoming in the DOS (windows 9x, NT) implementation of
> exec*, in a way that makes it work differently from the Unix it attempts to
> imitate.  This wouldn't surprise me, since I recently discovered that NT's
> spawnl takes a unix-like array of arguments (const char *argv[]) but when
> any element contains a space, it is split into two arguments in the child
> process.  So, I had to break some code that had worked fine on Unix to
> generate argv, by quoting each argument (something that would *not* work on
> Unix) ..
> 
> Bad enough that Windows wants to copy Unix, they have to consistently get
> it subtly wrong too.
> 
> Jeff



From Richard.Jones at fulcrum.com.au  Tue Jul 27 19:35:54 1999
From: Richard.Jones at fulcrum.com.au (Richard Jones)
Date: Tue, 27 Jul 1999 23:35:54 GMT
Subject: Qt 
In-Reply-To: Message from Kikutani Makoto of 1999-Jul-27 13:49:2,
            <7nkdce$t1c$1@oak.prod.itd.earthlink.net> 
References: <7nkdce$t1c$1@oak.prod.itd.earthlink.net>
Message-ID: <199907272335.JAA06521@envy.fulcrum.com.au>

[Kikutani Makoto]
> I fetched it and did make install.
> But I got the following error at trying examples.
> 
> mako% python aclock.py                /usr/local/src/Lang/pyKDEsrc-0.8/exampl
> es
> Traceback (innermost last):
>   File "aclock.py", line 4, in ?
>     from kde import *
>   File "/usr/lib/python1.5/kde.py", line 7, in ?
>     import libkdec
> ImportError: /usr/lib/libkimgio.so.2: undefined symbol: png_set_strip_16
> 
> What's wrong ?
> My Linux Box is Debian potato, and kdelibs2g is version 1.1.1-19990523.
> Python is 1.5.2 (python-gtk 0.6.1-1)

   Not sure - do you have the PNG library version that Qt requires? The example 
works for me under redhat 6.0 with qt 1.44 and png 1.0.3 (not sure what the 
actual requirements of Qt are though).


       Richard





From billtut at microsoft.com  Thu Jul 15 12:33:24 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Thu, 15 Jul 1999 16:33:24 GMT
Subject: making a Windows NT service out of a python program
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F50@RED-MSG-50>

> From: Christian Tismer [mailto:tismer at appliedbiometrics.com]

> Bill Tutt wrote:
> > 
> > > From: Christian Tismer [mailto:tismer at appliedbiometrics.com]
> > >
> > >
> > > This is all great stuff if you have understood it already.
> > > For me, it is still unclear what makes up a service, and
> > > what I need at the minimum to turn a Python program into
> > > an NT service.
> > >
> > 
> > Not knowing ANYTHING about Medusa... I'll provide another 
> example service
> > right here, and try and explain things as I go...
> 
> That's fine. Medusa is just a Python program. I just need
> to run any Pprogram as a service. :-)
> 

Medusa uses non-blocking sockets, so check out get_request in the
TCPServerService class.

In particular after accepting the new connection in Medusa, you don't need
to issue the .setblocking(1) call. You should still call WSAEventSelct()
just in case someone else down the chain would like to turn the socket into
a blocking socket again.

Bill



From fire at netlimit.com  Wed Jul 21 18:02:39 1999
From: fire at netlimit.com (Joacim Löwgren)
Date: Wed, 21 Jul 1999 22:02:39 GMT
Subject: cin
Message-ID: <3srl3.1302$Gj8.191985152@newsa.telia.net>

Hi,

I'm new to Python and I'm wondering what the C++ "cin" (or C
"scanf()")equivalent in Python is. Doesn't seem there's a quick function for
this.





From pierrette at p98.f112.n480.z2.fidonet.org  Sun Jul  4 16:54:50 1999
From: pierrette at p98.f112.n480.z2.fidonet.org (pierrette)
Date: Sun, 04 Jul 1999 21:54:50 +0100
Subject: Upgrade or not ??
Message-ID: <000017c7@bossar.com.pl>

From: pierrette <pserrand at worldnet.fr>

Thanks, exactly what i needed.
Is there a way to find out what events a widget support ?
I browsed documentation on python.org but didn't find anything on FocusIn en
FocusOut. I there somewhere documentation on that ??

Pierrette

Greg McFarlane wrote:

> Depending on exactly what you want to do, it may be possible in
> Tkinter.  You can use something like this to find out when a widget
> gets and loses focus:
>
>     import Tkinter
>
>     def gotFocus(event):
>         print 'Focus in'
>     def lostFocus(event):
>         print 'Focus out'
>
>     root = Tkinter.Tk()
>
>     e = Tkinter.Entry()
>     e.bind('<FocusIn>', gotFocus)
>     e.bind('<FocusOut>', lostFocus)
>     e.pack()
>
>     root.mainloop()
>
> Also, you can use this to set the focus to a widget.
>
>     e.focus_set()
>
> On 3 Jul, pierrette wrote:
> > First a general question :
> > Actually we're working with Python 1.5.1 and we're satisfied with it.
> > Should we upgrade to 1.5.2 and if yes WHY?
> >
> > No a Tkinter guestion :
> > We developped a prototype of an application with python an tkinter, now
> > we have to
> > decide if we're gonna write the real app in python or something else !
> > We now allready that in principle we could do it with python, the thing
> > we're missing
> > in tk is a 'on get focus' ,a 'on lost focus' event and a 'set focus to'
> > command
> > Pe: on lost focus of inputfield 1:
> >             if inputfield1 = A:
> >                 set focus to field2
> >             elif inputfield1 = B:
> >                 set focus to field 3
> >
> > If those things are not available in tkinter what alternatives do we
> > have ?
> > - Changing the GUI (would be a shame but feasable), knowing that is has
> > to be multi-platform WIN-LINUX? If so to what ?
> > - Not write in python ? In this case we lose the work done for the
> > prototype
> >
> > Thanks for comments and remarks
> >
> > Pierrette Serrand
> > pserrand at worldnet.fr
> >
> >
> >
> >
> >
>
> --
> Greg McFarlane:    INMS Telstra Australia (gregm at nms.otc.telstra.com.au)
> Today's forecast:  Sunny, with occasional cloudy periods and a chance
>                    of precipitation in some areas.



From posselt at daimi.au.dk  Tue Jul  6 11:28:49 1999
From: posselt at daimi.au.dk (Peter Posselt Vestergaard)
Date: Tue, 06 Jul 1999 17:28:49 +0200
Subject: problem with lists
Message-ID: <378220B1.9E0C272@daimi.au.dk>

It seems like I've completely misunderstood something or at least done
something wrong. There's noway I can see why these two scripts are
outputting nothing or more correctly only the 4 first lines which cannot
be seen in a browser. Anybody knows why?
----script 1------------------------------------------------

#!/usr/local/bin/python
import cgi

print "Content-type: text/html"
print

print "<HTML><HEAD>"
print "<TITLE> ?ndring af data </TITLE>"
print '</HEAD><BODY BGCOLOR="#FFFFFF"><FONT FACE="Helvetica">'

nydata.insert(0,'hej')
print nydata[0]
print 'strange'

print '</BODY></HTML>'

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

----script 2---------------------------------------------------

#!/usr/local/bin/python
import cgi

print "Content-type: text/html"
print

print "<HTML><HEAD>"
print "<TITLE> ?ndring af data </TITLE>"
print '</HEAD><BODY BGCOLOR="#FFFFFF"><FONT FACE="Helvetica">'

nydata[0]='weird'
print nydata[0]
print 'strange'

print '</BODY></HTML>'

------------------------------------------------------------------
This can be found at
http://dovs-www.daimi.au.dk/cgi-posselt/kurtilmelding/test.py


best regards
Peter Posselt Vestergaard



From bwarsaw at cnri.reston.va.us  Mon Jul 19 11:36:19 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Mon, 19 Jul 1999 11:36:19 -0400 (EDT)
Subject: Python and Boehm-Demers GC, I have code.
References: <o+zLoAAp4ak3Ewjk@jessikat.demon.co.uk>
	<000101bed15c$c5685ca0$e7a22299@tim>
Message-ID: <14227.17907.298990.975953@anthem.cnri.reston.va.us>

>>>>> "TP" == Tim Peters <tim_one at email.msn.com> writes:

    TP> It's Cyclops.py, and I just uploaded it to python.org's ftp
    TP> contrib site.  It should show up in the System directory later
    TP> this week (send a complaint every 5 minutes to
    TP> ftpmaster at python.org until it does <wink>).

    TP> that-barry-gets-too-much-sleep-ly y'rs - tim

Tim was already
Pushing it with the bogus
Haiku that he sent

-Barry



From sreekantk at freenet.co.uk  Thu Jul  8 21:52:12 1999
From: sreekantk at freenet.co.uk (Sreekant Kodela)
Date: Fri, 09 Jul 1999 02:52:12 +0100
Subject: Help with strings please
Message-ID: <378555CC.8B65F809@freenet.co.uk>

Dear all

I seem to be in trouble here. I am trying to make some cgi scripts for
my web site. I am using the postgresql as the backend db and a form
based address book and news forum etc. The problem is everything is
working ok as long as noone uses ' and " in the text. If they do, bingo
, postgres refuses to accept. I tried using string library [mainly
string.join(string.split('awkward'"'st'ring')). It works sometimes and
doesn't some times.

Any better and efficient way to deal with this situation !

Any suggestions and ideas will be greatly appreciated.

Thanks
sreekant





From heather at thalamus.wustl.edu  Mon Jul 19 12:26:00 1999
From: heather at thalamus.wustl.edu (Heather A. Drury)
Date: 19 Jul 1999 11:26:00 -0500
Subject: binding keys to slider in Tkinter?
Message-ID: <7mvjio$dh8@thalamus.wustl.edu>

Hi,

I'm trying to bind the arrow keys to allow a user to increment
(or decrement) a counter using the Scale widget and Tkinter.
I seem to be unable to bind anything to the Scale widget.
I'm sure I'm missing something.

Here's a code blurb:

        self.slider = Scale(self, from_=0, to=100,
                            orient=HORIZONTAL,
                            length="3i",
                            variable = self.myval,
                            label="happy slider",
                            command=self.print_value)
        self.slider.bind ("<KeyPress-f>", self.doit)

How come I can't seem to do this? Also, anyone know how to bind
the right and left arrow keys (is it <LeftArrowKey> or something
like that?)?

TIA,

Heather



From emile at fenx.com  Thu Jul 22 19:50:27 1999
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 22 Jul 1999 16:50:27 -0700
Subject: Fw: ANNOUNCE: Mailing list for Australian Python Users
References: <007201bed42e$c4316c30$f29b12c2@secret.pythonware.com> <wkogh56j2o.fsf@ifi.uio.no> <7n78nm$lic$1@news.accu.uu.nl> <379798BC.33FC0CA9@equi4.com>
Message-ID: <r9Ol3.3999$VL2.425132@news.direcpc.com>


--

Emile van Sebille
emile at fenx.com
-------------------


Jean-Claude Wippler <jcw at equi4.com> wrote in message
news:379798BC.33FC0CA9 at equi4.com...
> Martijn Faassen wrote:
> >
> > Lars Marius Garshol <larsga at ifi.uio.no> wrote:
> >
> > > * Fredrik Lundh
> > > |
> > > | python-se, n?gon?  Link?ping?  :-)
> >
> > > Tjah. Si Oslo, s? kan vi bli en del. :-)
> >
> > Utrecht is veel beter denk ik, Python-NL en zo..
>
> Doen.  Ik kom op de fiets.
>
> > An-large-percentage-of-pythoneers-speaks-Dutch-anyway-ly yours,
>
> De meerderheid, toch?
>
> -- Jean-Claude

Niet op de brommer?  (sp?)





From smalleys at gte.net  Fri Jul 23 00:44:02 1999
From: smalleys at gte.net (Sue)
Date: Fri, 23 Jul 1999 04:44:02 GMT
Subject: Merging two lists as a dictionary
References: <19990720211946.A12202@pooh.frostnet.net> <19990721153539.S6170@xs4all.nl> <19990721084614.C15502@pooh.frostnet.net>
Message-ID: <3797F338.E560D59A@gte.net>

I think there's a C function in mxTools, as well.
Sue.



From petravic at fnal.gov  Thu Jul  8 13:01:48 1999
From: petravic at fnal.gov (petravick@FNAL.GOV)
Date: Thu, 8 Jul 1999 17:01:48 GMT
Subject: Two Pythons talking to each other?
In-Reply-To: <14212.51977.601241.772478@buffalo.fnal.gov>
References: <14212.51977.601241.772478@buffalo.fnal.gov>
Message-ID: <Pine.SGI.3.95.990708120053.9140Q-100000@hppc.fnal.gov>

AF_UNIX is likely to me more effcient if that matters.

AF_INET will run about the whole protocol stack, AF_UNIX
is usually implemented to be more like a pipe...

-- DOn

On Thu, 8 Jul 1999, Charles G Waldman wrote:

> Date: Thu, 08 Jul 1999 11:00:09 -0500 (CDT)
> From: Charles G Waldman <cgw at fnal.gov>
> To: Per Kistler <kistler at fnmail.com>
> Cc: python-list at cwi.nl
> Subject: Re: Two Pythons talking to each other?
> Newsgroups: comp.lang.python
> 
> Per Kistler writes:
>  > Hi Hans
>  > 
>  > One could use sockets with AF_INET, if it runs on another
>  > machine, and sockets with AF_UNIX, if it's on the the same
>  > host.
> 
> There's no particular reason to use AF_UNIX, even if the two processes
> are running on the same node.  AF_INET with a loopback IP address
> works fine, and is portable (not all platforms support unix-domain
> sockets).
> 




From postmail at mail3.intellect.com.tw  Sat Jul  3 12:50:12 1999
From: postmail at mail3.intellect.com.tw (postmail at mail3.intellect.com.tw)
Date: Sat, 03 Jul 1999 16:50:12 GMT
Subject: Save your printing money
Message-ID: <98572UT3.7M33741Y@mail3.intellect.com.tw>

A.R. Tech Communication (CartridgeOutlet) is specialized
in remanufactured toner cartridge & replacement inkjet
cartridges

Our mission is to save your money in printing . All 
remanufactured toner cartridges & replacement inkjet
cartridges we offer are under lifetime warranty and 60 days
cash back guarantee.

Laser Printer Cartridge                          USD$
======================================================
HP92274 (PX).....................................52.20
HP92275 (LX).....................................47.95 
HP92285 (LX).....................................54.95 
HP92291 (NX).....................................61.95 
HP92295 (SX).....................................42.95 
HP92298 (EX).....................................59.95 
HP3900 (BX)......................................71.95  
HP3903 (VX)......................................66.45 
HP3906 (AX)......................................54.10  
HP3909 (WX)......................................94.95 
A-30       ......................................56.95  
E-31       ......................................71.20 
FX-1       ......................................57.95 
FX-2       ......................................68.95 
FX-3       ......................................64.95 
IBM 4019/28/29...................................79.95 
IBM 4039/49.....................................101.60 

Laser Printer Cartridge w/Micr Toner             USD$
======================================================
HP92274 (PX) Micr Toner..........................77.95
HP92275 (LX) Micr Toner..........................74.95 
HP92291 (NX) Micr Toner..........................97.95 
HP92295 (SX) Micr Toner..........................81.95 
HP92298 (EX) Micr Toner..........................87.95 
HP3900 (BX) Micr Toner..........................101.95  
HP3903 (VX) Micr Toner...........................93.95  
HP3906 (AX) Micr Toner...........................93.95  
HP3909 (WX) Micr Toner..........................111.95 
HP4000A Micr Toner..............................164.95 
HP4000X Micr Toner..............................169.95 

InkJet Printer compatiable Cartridge              USD$
======================================================
SO20025 STYLUS 800, 1000 black....................9.69
SO20034 STYLUS PRO XL black.......................9.69 
SO20036 STYLUS PRO XL C/Y/M......................18.75 
SO20047 STYLUS 200, 820, COLOR II black...........9.69 
SO20049 STYLUS 200, 820, COLOR II C/Y/M..........15.49 
SO20066 STYLUS PRO XL C/Y/M......................18.49 
SO20089 STYLUS 400, 600 C/Y/M....................16.97 
SO20093 STYLUS 400, 600 black.....................9.99 
SO20108 STYLUS 800, 1520 black....................9.99 
SO20110 STYLUS PHOTO /PM 700C F.  CC/MM/YY.......17.29 
SO20118 STYLUS COLOR 3000 black..................22.99  
SO20122 STYLUS COLOR 3000 yellow.................22.99 
SO20126 STYLUS COLOR 3000 magenta................22.99 
SO20130 STYLUS COLOR 3000 cyan...................22.99 
SO20138 STYLUS COLOR 300 B/Y/M/C.................14.99  
201B BJC 600 SERIES black.........................6.29 
201C BJC 600 SERIES cyan..........................6.29  
201M BJC 600 SERIES magenta.......................6.29  
201Y BJC 600 SERIES yellow........................6.29  
201HCB BJC 600E SERIES black......................6.99 
21B BJC 4000 SERIES black.........................9.49 
21CLR BJC 4000 SERIES Y/M/C......................11.49 
642B BJC 300 SERIES black.........................8.99 

For more information please visit our website at
http://www.cartridgeoutlet.com

Per Section 301, Paragraph (a)(2)(C) of S. 1618, further transmissions to you
by the sender of this email may be stopped at no cost to you by sending a
reply to this email address with the word "remove" in the subject line.. 
 



From Mike.C..Fletcher at p98.f112.n480.z2.fidonet.org  Fri Jul  2 14:25:26 1999
From: Mike.C..Fletcher at p98.f112.n480.z2.fidonet.org (Mike C. Fletcher)
Date: Fri, 02 Jul 1999 19:25:26 +0100
Subject: FOX-USERS: Re: Python GUIs: Abandoning TkInter and welcoming wxPyth
Message-ID: <0000168a@bossar.com.pl>

From: "Mike C. Fletcher" <mfletch at vrtelecom.com>

I should clarify, when I'm speaking about "rich" text editing widgets, I'm
talking about colouring/styling, font selection, selection management, and
preferably some sort of layout engine.  Rich text as in RTF format, or HTML
can both easily be built on top of a generic "rich" text widget.

My personal ideal would be to allow a cascading style sheet-style layout
system.

Looking at the code for FXEditor, it seems to be very tightly focused on
source-code editing, but subclassing it to allow for a colouriser which also
changes fonts looks possible (it uses a low-level device context to draw
graphic fonts).  It is not a good base for a full layout engine, however, as
it rather lacks in abstraction, doesn't provide for formatting blocks, and
is generally a straight-forward-no-looking-back text-placement engine.
Possibly adapting the Mozzilla layout engine (Gecko?) might be a better
option.

Enjoy yourselves,
Mike


Lyle Johnson <jlj at cfdrc.com> wrote in message
news:377CFD82.8EA084A2 at cfdrc.com...
...
> I think any kind of rich text (i.e. RTF-compliant) or HTML widget is much
> farther off for FOX.
...




From jepler at inetnebr.com  Sun Jul 25 09:17:33 1999
From: jepler at inetnebr.com (Jeff Epler)
Date: Sun, 25 Jul 1999 13:17:33 GMT
Subject: simplebeginnerquestion(**args)
References: <379afff7.131877652@news.mindspring.com>
Message-ID: <slrn7pm3ue.sj2.jepler@craie.inetnebr.com>

Brian,

When you're going to use "named" arguments, such as
	myfunc(a=1, b=2, c=3, d=4)
then it's appropriate to use
	def myfunc(**kw): pass
but when you're using unnamed arguments, such as
	myfunc(1, 2, 3, 4)
you want to use
	def myfunc(*args): pass

This makes sense, because for unnamed arguments the position is important
but there's no sensible key to give them if they were placed in a
dictionary---so instead, you get them as a tuple with "*args".  But if the
arguments are named, obviously they need to go into a dictionary.

Jeff
-- 
\/ http://www.infidels.org/                      Jeff Epler jepler at inetnebr.com
"Pay no attention to the man behind the curtain."
-- Karl, as he stepped behind the computer to reboot it, during a FAT



From greg.ewing at compaq.com  Thu Jul 15 23:50:45 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 16 Jul 1999 15:50:45 +1200
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mlu8n$6ug$1@brick.cswv.com>
Message-ID: <378EAC15.20269642@compaq.com>

Neel Krishnaswami wrote:
> 
> What's the difference, exactly? AFAICT you need to save the execution
> state when suspending both coroutines and generators, but I might be
> missing something obvious....

The crucial difference is that a generator is never resumed
after its caller has returned. This means that the generator's
state can be pushed onto the same stack as the caller's.

A coroutine, on the other hand, can outlive the context in
which it was created, and therefore needs a stack all of
its own.

Another way to think about it is that a generator call is
equivalent to an ordinary call where one of the parameters
is a procedure. For example, where in "generator python"
you might write

  def even_numbers_up_to(n):
    for i in range(n):
      if i % 2 == 0:
        yield(i)

  for e in even_numbers_up_to(42):
    print e, "is an even number!"

you could do the same thing in ordinary python as

  def for_each_even_number_up_to(n, do_it):
    for i in range(n):
      if i % 2 == 0:
        do_it(i)

  def print_it(e):
    print e, "is an even number!"

  for_each_even_number_up_to(42, print_it)

which clearly can be executed quite happily using 
a single stack.

If Python had something akin to Smalltalk code blocks,
generators wouldn't be needed. It would be nifty to
be able to write something like

  for_each_even_number_up_to(42) -> (e):
    print e, "is an even number!"

Greg



From phd at emerald.netskate.ru  Tue Jul 27 08:15:44 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Tue, 27 Jul 1999 16:15:44 +0400 (MSD)
Subject: 1.5.2 for: else:
Message-ID: <Pine.LNX.4.04.9907271612140.9849-100000@emerald.netskate.ru>

Hello!

   The following program:

----------
for a in ['a', 12]:
   print a
else:
   print "Empty!"
----------

   prints under 1.5.2 on Linux and Solaris:

-----
a
12
Empty!
-----

   Is it a bug, or is it just me, who do not understand the feature? I
expected just "a" and "12"...

   This:

----------
for a in []:
   print a
else:
   print "Empty!"
----------

   prints:

-----
Empty!
-----

   and I think it is correct.

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From Alex.Maranda at p98.f112.n480.z2.fidonet.org  Thu Jul  1 09:12:56 1999
From: Alex.Maranda at p98.f112.n480.z2.fidonet.org (Alex Maranda)
Date: Thu, 01 Jul 1999 14:12:56 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001606@bossar.com.pl>

From: Alex Maranda <amaranda at nospam.com>

Mikael Lyngvig wrote:
 
> In contrast, the wxPython demo is so fast that you probably wouldn't
> realize it is implemented in Python, unless you're used to do GUI
> programming yourself.
I'll second that.

> The wxWindows performance problems must be more of the "this could be
> done better" than "this has to be done better" nature.
I'll second that.

> A very strong plus for wxWindows is that the Python API and the native
> C++ API are so close that it actually makes sense to use Python in the
> design and prototyping phase and then rewrite the Python code to C++,
> if you want to, when you've reached a stable design.
This is the way to go indeed. Of course only the performance sensitive
bits - there's no point in rewriting everything. Then put freeze to work
on the remaining python code, and here you go. I've actually done that
with the (sizable) wxPython demo on Linux (linking everything statically
- libpython, wxPython wrappers, wxGTK, GTK, GLIB, even stdc++; in the
end the only dynamic dependencies of the binary were libc, libdl,
libX11). I got a 4.5 Meg binary, from which 0.5 Meg was the demo
bytecode. 4 Meg of 'runtime' it's not much nowadays, and boy it's fast.
I'm not advocating this for casual use, I just explored packaging
possibilities for the average U*ix platform. The Windows version is
smaller 'cause GDI and the other libs are always there :)

Cheers,
-- 
Alex Maranda         mailto: amaranda at spider dot com
Spider Software Ltd. Tel: +44 (0)131 4757036
Edinburgh, UK        http://members.xoom.com/Alex_Maranda
STREAMS based communications protocols for embedded systems



From parkw at better.net  Thu Jul  8 06:33:15 1999
From: parkw at better.net (William Park)
Date: Thu, 8 Jul 1999 06:33:15 -0400
Subject: I am v. stupid, please help
In-Reply-To: <000015aa@bossar.com.pl>; from Newbie on Wed, Jun 30, 1999 at 12:31:49AM +0100
References: <000015aa@bossar.com.pl>
Message-ID: <19990708063315.A2111@better.net>

On Wed, Jun 30, 1999 at 12:31:49AM +0100, Newbie wrote:
> From: "Newbie" <8499 at greenhead.ac.uk>
> 
> 
> I have just downloaded Python 1.5 and I have 
> had a brief look at the helpfiles and some 
> source code and my stupid question is this :
> How do you define a string and then use it 
> for input, e.g,
> 
> name = input('What is your name : ')
> 
> This works for integers but not for strings,
> 
> Help Me Please!!
> 
> --Posted from EarthWeb Discussions. http://discussions.earthweb.com
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list

Since 'eval' will simply execute your expression just as you would
type it interactively, try putting quotations around your input:

>>> input("==> ")
==> 123
123
>>> input("==> ")
==> 'sss'
'sss'
	
William Park




From jsight at mindspring.com  Wed Jul  7 00:55:26 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Wed, 07 Jul 1999 00:55:26 -0400
Subject: Advertising Management in Python?
Message-ID: <3782DDBE.E666F6D4@mindspring.com>

Is there a Python library for easily handling banner ads along with
statistics and analysis?  This would be very nice, as most of the ones
in Perl are truly annoying.  :-)  Ok, so I don't like Perl.  :-)

Thanks in advance,

-- 
---------------
Jesse D. Sightler
http://www3.pair.com/jsight/

"Do not use a hatchet to remove a fly from your friend's forehead." 
      - Chinese Proverb



From rvollmert at gmx.net  Sat Jul 17 10:53:26 1999
From: rvollmert at gmx.net (Robert Vollmert)
Date: Sat, 17 Jul 1999 16:53:26 +0200
Subject: at / os.popen problems
Message-ID: <19990717165326.D599@krikkit.vollmert.home>

Hello,

I found a problem with os.popen in an at job. Closing the pipe after
calling pipe.read() produces an IOError for some commands.

popen.py (attached) produces the following output when called from at
as:
popen.py '/bin/ls /' '{ /bin/ls / ; }' 'echo hello' '{ echo hello ; }'

error while closing { /bin/ls / ; }: IOError: [Errno 10] No child processes
error while closing echo hello: IOError: [Errno 10] No child processes
error while closing { echo hello ; }: IOError: [Errno 10] No child processes
/bin/ls / status: 0
/bin/ls / output:
[snipped correct ls output]

{ /bin/ls / ; } status: -1
{ /bin/ls / ; } output:
[snipped correct ls output]

echo hello status: -1
echo hello output:
hello

{ echo hello ; } status: -1
{ echo hello ; } output:
hello

This works correctly from cron and in the shell. Also, when read()
is left out, close() works as normal.

I'm using a self-compiled version 1.5.2 python on Debian/GNU Linux 2.1
with kernel 2.2.10. at is version 3.1.8.

Thanks, 
Robert.

popen.py:
#! /usr/bin/env python
import sys, os, string
def main():
    for command in sys.argv[1:]:
        try:
            pipe = os.popen('%s' % command, 'r')
        except Exception, inst:
            sys.stderr.write('error while opening %s: %s\n' %
                             (command, inst.__class__.__name__, inst))
        try:
            output = pipe.read()
        except Exception, inst:
            sys.stderr.write('error while reading %s: %s\n' %
                             (command, inst.__class__.__name__, inst))
        try:
            status = pipe.close() or 0
        except Exception, inst:
            sys.stderr.write('error while closing %s: %s: %s\n' %
                             (command, inst.__class__.__name__, inst))
            status = -1
        print command, "status:", status
        print command, "output:"
        print output
if __name__ == '__main__':
    sys.exit(main())

-- 
Robert Vollmert                                      rvollmert at gmx.net



From clgonsal at kami.com  Fri Jul  9 07:16:22 1999
From: clgonsal at kami.com (C. Laurence Gonsalves)
Date: Fri, 09 Jul 1999 11:16:22 +0000
Subject: VIM and [Python] block jumping
References: <3781347D.CB2FBAB3@starvision.com> <Pine.SOL.4.05.9907051548470.13052-100000@saturn.radsci.uci.edu> <37815B80.B4B51F94@starvision.com> <3785D753.DB2FF1ED@kami.com>
Message-ID: <3785DA06.974567F3@kami.com>

Of course, right after I posted that, realized that there's a problem.
The blockMotion function doesn't account for lines that are
continuations of other lines. That means that triple-quoted string
literals, lines with more {[('s that }])'s, and lines joined with \ can
screw it up. Are there any other situations? I'll try to post a
corrected version later. Some of those problems are pretty nasty though
(triple-quoted strings in particular).

-- 
  C. Laurence Gonsalves              "Any sufficiently advanced
  clgonsal at kami.com                   technology is indistinguishable
  http://www.cryogen.com/clgonsal/    from magic" -- Arthur C. Clarke



From dyoung at pobox.com  Mon Jul 19 23:12:32 1999
From: dyoung at pobox.com (David Young)
Date: 20 Jul 1999 03:12:32 GMT
Subject: python to perl conversion?
Message-ID: <slrn7p7q8j.n4.dyoung@baltazar.baltazar.org>

Is there a program that converts Python source to Perl source?  There is
a program that a client wants written in Perl so that their Perl-only
programmers (can you imagine?) can hack on it later. I haven't the time or
inclination to learn Perl, and this project fairly cries out for Python,
but I cannot justify the time expense of translating from Python to Perl
by hand.

Dave

-- 
``I always thought that anybody who told me I couldn't live in the
past was trying to get me to forget something that if I remembered
it would get them in serious trouble.'' --Utah Philips



From wolford at enews.nrl.navy.mil  Thu Jul  8 11:01:31 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Thu, 08 Jul 1999 11:01:31 -0400
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com>
Message-ID: <3784BD4B.7938B7BA@enews.nrl.navy.mil>

Here's the smallest case that gives me problems:
#include "Python.h"
#include <stdio.h>
int main(int argc, char **argv)
{
//  return Py_Main(argc, argv);
  FILE *file = fopen(argv[0], "rt");
  PyRun_AnyFile(file, argv[0]);
}

I just modified python.c to run the script like I do in my application. I
get the following error:
Fatal Python error: PyThreadState_Get: no current thread

The Py_Main line works fine. Any takers?

Thanks in advance,
Scott

Gordon McMillan wrote:

> Scott Wolford writes:
>
> > Now I'm a little confused.
>
> I was trying to figure out how to tell you that politely...

Thanks for doing so.

> > Python does seem to be a dll (?); I found
> > python15.dll in winnt/system32 directory.
>
> Correct. That's 99.999+% of python. python.exe does nothing but load
> it and call Py_Main.

I knew that; brain fart. I've written my own makefile for python and my
application works fine when I build it myself.

> > But I can't find
> > python15_d.dll anywhere. Was that supposed to be installed?
>
> Nope, you have to build it.
>
> > I think
> > python15.dll bombs because it was compiled as optimized using the M$
> > compiler and that optimizer has problems (at least w/ O2; I've been
> > pretty successfull w/ O1).
>
> Probably not. It's been pretty thoroughly tested at this point.

I agree, but in the *normal* fashion.

> > So I want to use python15_d but I can't
> > find the dll.
> >
> > Any help would be appreciated,
>
> Probably the most common mistake of Windows embedders is to link to
> the wrong C runtime lib. Python uses Multithreaded DLL.

Nope, I use msvcrt and msvcrtd. You say Multithreaded *DLL* so that's what
you mean (and not libcmt.lib), right?

> But you might have done something more creative. Telling us how and
> when it bombs complete with messages and as small a code sample as
> possible will get you more and better answers.

see top of post.

> - Gordon




From markh at angband.org  Fri Jul 16 08:09:18 1999
From: markh at angband.org (Mark Howson)
Date: Fri, 16 Jul 1999 13:09:18 +0100
Subject: Expand tuple into argument list?
References: <87aesy4w3f.fsf@solano.in.renlabs.com>
Message-ID: <378F20EE.5F0567AA@angband.org>

Steven Work wrote:
 
> I'm new to Python.  I want a dispatch function -- takes a function
> name and other arguments, calls the named function with the remaining
> arguments.  This works for limited length argument lists:
> 
> def dispatch(fnname, *rest):
>     fn = dispatch.func_globals[fnname]

How about:

      return( apply(fn, rest))

Mark



From gerrit.holl at pobox.com  Thu Jul  8 16:41:10 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 8 Jul 1999 20:41:10 GMT
Subject: Why are index() and count() only for mutable sequences?
In-Reply-To: <3784F778.6A6B4568@lemburg.com>; from M.-A. Lemburg on Thu, Jul 08, 1999 at 09:09:44PM +0200
References: <004701bec95f$186faa80$a0ee6dd1@BobAlexDesk> <3784F778.6A6B4568@lemburg.com>
Message-ID: <7m9tlp$6qg$7@rks1.urz.tu-dresden.de>

On Thu, Jul 08, 1999 at 09:09:44PM +0200, M.-A. Lemburg wrote:
> Date: 	Thu, 08 Jul 1999 21:09:44 +0200
> From: "M.-A. Lemburg" <mal at lemburg.com>
> To: Bob Alexander <bobalex at ix.netcom.com>
> Cc: python-list at python.org
> Subject: Re: Why are index() and count() only for mutable sequences?
> 
> Bob Alexander wrote:
> > 
> > There is no modification of the sequence performed by these functions
> > -- it seems as though they could be available for all sequences. It
> > would certainly be useful for tuples as well as lists (strings, too,
> > for that matter, even though we have the more general string.find()).
> 
> You could try count(), index(), forall() and exists() from mxTools:
> 
> 	http://starship.skyport.net/~lemburg/mxTools.html
> 
> Albeit, they work with conditions (functions that return 1/0) to
> implement the testing process.
> 

Is this an anwser??
He asked: "why"...

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From greg.ewing at compaq.com  Tue Jul 13 19:50:45 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 14 Jul 1999 11:50:45 +1200
Subject: console show in Python?
References: <378626CF.9495FAB0@mry.rbd.com>
Message-ID: <378BD0D5.A05CDD21@compaq.com>

Scott Shillcock wrote:
> 
> Is there something similar to the Tcl/Tk 'console show' command in
> Python.

Use IDLE - it has a console window that does
what you want.

Note that you don't need to call mainloop()
at all when you're using IDLE, because IDLE
itself is implemented using Tkinter and is
therefore running a mainloop() all the time.

Greg



From larsga at ifi.uio.no  Sun Jul  4 06:23:03 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 04 Jul 1999 12:23:03 +0200
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> 	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> 	<375FDCA7.62FF@mailserver.hursley.ibm.com> 	<7josm1$bgc$1@Starbase.NeoSoft.COM> 	<37628EBF.207F7A8F@iname.com> 	<37682DD3.43923243@compaq.com> 	<7ka2be$qja$1@nntp6.u.washington.edu> 	<37776FFA.2D85@mailserver.hursley.ibm.com> 	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it> 	<876748mcxg.fsf@home.ivm.de> 	<37788BB9.ABD@mailserver.hursley.ibm.com> 	<lw67471fyi.fsf@copernico.parades.rm.cnr.it> 	<3779ED0F.2C67@mailserver.hursley.ibm.com> 	<nkjzp1hizvb.fsf@tfeb.org> 	<377A3383.D453468@iname.com> 	<377aa228.31380423@news.jpl.nasa.gov> 	<slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> 	<wkzp1g2t2e.fsf@ifi.uio.no> 	<ey3r9mskvja.fsf@lostwithiel.tfeb.org> 	<x27lojoums.fsf@todday.aiai.ed.ac.uk> 	<wmcn1xenodh.fsf@polytechnique.fr> <14206.52213.119553.809177@lrz.de>
Message-ID: <wk3dz4u320.fsf@ifi.uio.no>

* Jeff Dalton
|
| Well *Lisp* dates back to the late 50s.  Scheme may be from 75, but
| Lisp is a different matter.  And lambda calculus is, what, 1949?

* Jerome Kalifa
| 
| No, as a mathematical theory, it was developed before the war.

* Eugene Leitl
| 
| Church, A., The Calculi of Lambda Conversion, Princeton University
| Press, Princeton, N.J., 1941

According to David Harel's Algorithmics, the correct reference is
actually S.C. Kleene "A Theory of Positive Integers in Formal Logic",
Amer. J. Math. 57 (1935), pp. 153-173,219-244.

And there seems to be general agreement that the lambda calculus dates
back to the 30s. 

--Lars M.



From algaba at my-deja.com  Mon Jul 26 14:13:38 1999
From: algaba at my-deja.com (algaba at my-deja.com)
Date: Mon, 26 Jul 1999 18:13:38 GMT
Subject: createfilehandler, how to use it with Tk?
References: <7nht1s$grn$1@nnrp1.deja.com> <012c01bed77a$d562b370$f29b12c2@secret.pythonware.com>
Message-ID: <7ni8ga$p0k$1@nnrp1.deja.com>

In article <012c01bed77a$d562b370$f29b12c2 at secret.pythonware.com>,
  "Fredrik Lundh" <fredrik at pythonware.com> wrote:
> algaba at my-deja.com wrote:
> > I have to communicate a tk front end with another modules.
> > And I'm trying:

> forget about createfilehandler. use asyncore/asynchat instead:
> http://www.python.org/doc/current/lib/module-asyncore.html
>
> to use them under Tk, you can use a
> simple polling scheme:
>
>             asyncore.poll(0.05)
>             self.master.after(100, self.poll)
>         except:
>             pass # error in network callback
>
> ...
>
asyncore is a new module ? I can't find it in my python 1.5.1

Anyway, thanks.
And... what's the difference between this and a simple
loop with time.sleep(0.001), ... a bit of more freedom?
Thanks


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From bwarsaw at cnri.reston.va.us  Wed Jul  7 11:17:58 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Wed, 7 Jul 1999 11:17:58 -0400 (EDT)
Subject: ++Python - Emacs functions to expand shorthand
References: <7ltrua$pkr$1@nnrp1.deja.com>
Message-ID: <14211.28582.924601.108525@anthem.cnri.reston.va.us>

>>>>> "PL" == Preston Landers <mithy at my-deja.com> writes:

    PL> py-shorthand-line-expand -- bound to C-z -- works only on the
    PL> current line.

    PL> py-shorthand-buffer-expand -- bound to C-c C-z -- works on the
    PL> whole buffer.

    PL> These keys expand shorthand expressions.  A couple of
    PL> examples:

    | foo += bar  --->  foo = foo + bar
    | ++foo.bar   --->  foo.bar = foo.bar + 1

Here's a thought:

Why not make `+', `-', and `=' electric and do the expansion
automatically when you type those characters?  You'd have to use

    foo.bar++  instead of ++foo.bar

but that doesn't seem too bad.  I think those expansions would be
pretty easy to do, and if they work out well and a lot of people like
them, I might consider adding them to python-mode.

-Barry



From mwh21 at cam.ac.uk  Wed Jul 14 13:28:17 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Jul 1999 18:28:17 +0100
Subject: OS/2 Python: editing in interactive use
References: <378C55A9.411BE9AB@tu-clausthal.de>
Message-ID: <m3673nkuod.fsf@atrus.jesus.cam.ac.uk>

Stefan Schwarzer <Stefan.Schwarzer at tu-clausthal.de> writes:

> Hi there :)
> 
> It seems to me that the OS/2 port of python lacks the
> "sophisticated" features for input editing (history,
> C-A, C-E, etc.)

Those features are provided by readline. I doubt readline works on
OS/2, though you could have a look. A search on google turned up a
reference to a Python readline module for win32, though the site seems
to be off net, at least temporarily.
 
> If I have overlooked a possibility to activate it, please
> give me a hint. If not, how difficult would it be to extend
> Python/2 appropriately? I would appreciate this very much. :)

Can you use idle? Most Python IDE's have some commandline editing
features.

> Regards
>  Stefan

HTH
Michael

-- 
Oh, very funny. Very sar-cah-stic.         http://www.ntk.net
http://www.ntk.net/doh/options.html - ho ho



From Alan.Daniels at p98.f112.n480.z2.fidonet.org  Sat Jul  3 00:51:47 1999
From: Alan.Daniels at p98.f112.n480.z2.fidonet.org (Alan Daniels)
Date: Sat, 03 Jul 1999 05:51:47 +0100
Subject: Python suitable for a game engine?
Message-ID: <000016a2@bossar.com.pl>

From: daniels at mindspring.com (Alan Daniels)

On Fri, 02 Jul 1999 11:48:36 +0100, the infinitely wise Alex Maranda
(amaranda at nospam.com) spoke forth to us, saying...

[snip...]
>Alan Daniels wrote:
>> I have the (VERY!) preliminary code at:
>>     http://www.mindspring.com/~daniels/engine/index.html
>> if you're interested, although there isn't much to look at.
>I'm downloading it now. You can use negative version numbers;
>SmallEiffel is now at version -0.78. They'll stop when they'll reach 0.0
>:-)

In that case, I'm going to restart with negative numbers.
I was just going to call it the "Zebra" engine, but that name is
taken. So, I'm going to rename it to the "Zeroth" engine. Seems
appropriate, considering the state its in. :=)

-- 
=======================
Alan Daniels
daniels at mindspring.com
daniels at cc.gatech.edu



From m.faassen at vet.uu.nl  Mon Jul 19 04:05:54 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 19 Jul 1999 08:05:54 GMT
Subject: J, you can't pass your own test.
References: <378E3481.EE31C97@uswest.net> <19990715175320.B20155@toast.internal> <7mnj5k$id9$1@nnrp1.deja.com> <19990716173936.F20935@toast.internal>
Message-ID: <7mum92$15a0$1@news.accu.uu.nl>

jam <jam at newimage.com> wrote:
> myself *did* take the time to respond to such a post says a lot for the
> community as a whole, which is far more important than my own political
> standing in the group at the moment.

Cool, I didn't know I *had* a political standing. So what's my political
standing? What's my ranking on the Tim Peters scale (where the timbot is
100)?

Trying-to-lighten-the-mood-ly yours,

Martijn



From vdkoijk at linux01.prc.tno.nl  Mon Jul  5 03:34:04 1999
From: vdkoijk at linux01.prc.tno.nl (John van der Koijk)
Date: 05 Jul 1999 09:34:04 +0200
Subject: popen2 on Windows
Message-ID: <877lof7doz.fsf@linux01.prc.tno.nl>

I hope this is not a FAQ:

First, let me set the stage:

I'm trying to implement a control loop using python and a DDE
application under Windows 95. The DDE link freezes as soon as I start
a thread in the python program, in order to run a Tkinter GUI.

(((
I've taken care of the Tkinter mainloop vs. threading problem using

# Ernst-Udo Wallenborn <wallenborn at phys.chem.ethz.ch> says:
def mainloop(self):
	dummy = 0
        while 1:
		Frame.update(self)
		time.sleep(0.02)
#		if self.quitting == 1:
#			sys.exit(0)

root.mainloop = mainloop
root.mainloop(root)

which is puzzling because `quitting' does not seem to be known here.
)))

After coping with my initial frustration, I've decided to make an
unthreaded stub program talking to the DDE client, using popen2 to do
the IPC with the GUI and controller program. Sadly, this approach also
does not seem to work, because using popen2 gives me:

[....] os.fork()
AttributeError: fork

This is amazing, since the manual for 1.5.2 says:

8.13 popen2 -- Subprocesses with accessible standard I/O streams

   Availability: Unix, Windows.

I'm using Python 1.5.2 (#0). Could anyone give me a clue what could be
wrong?

Kind regards,
-- 
John van der Koijk.
--
TNO Institute of Industrial Technology

PO Box 5073, 2600 GB, Delft, The Netherlands
Phone +31 15 2608833, Fax +31 15 2608846 



From wtbridgman at radix.net  Fri Jul 23 06:46:24 1999
From: wtbridgman at radix.net (T. Bridgman)
Date: Fri, 23 Jul 1999 6:46:24 -0400
Subject: Python Mode for Alpha ?
References: <lawter-2107990058210001@client-151-197-119-26.bellatlantic.net> <14229.57251.996871.502133@anthem.cnri.reston.va.us> <14229.58766.234709.815774@weyr.cnri.reston.va.us> <o9vhbevwgh.fsf@panix3.panix.com>
Message-ID: <01HW.B3BDC040002912FD0CB6CED0@news1.radix.net>

I know that I've never been able to get it working on Alpha 7.x

Tom

On Wed, 21 Jul 1999 11:44:46 -0400, Josef Sachs wrote
(in message <o9vhbevwgh.fsf at panix3.panix.com>):
> 
>>>>>> On Wed, 21 Jul 1999 15:21:50 GMT, Fred L Drake said: 
> 
>> No, he's asking about an editing mode for Alpha, a Mac editor.  I remember 
>> having one when I used a Macintosh, but that was almost 5 years ago.  I 
>> have no idea where to get such a beast these days. 
> 
> <URL: ftp://ftp.python.org/pub/python/mac/old/AlphaPythonMode.sit.hqx>? Did 
> that stop working with some recent Alpha release? 




From jsight at mindspring.com  Tue Jul 20 18:13:44 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Tue, 20 Jul 1999 18:13:44 -0400
Subject: python to perl conversion?
References: <slrn7p7q8j.n4.dyoung@baltazar.baltazar.org> <7n1t4q$ik$1@nnrp03.primenet.com>
Message-ID: <3794F498.AD6FF3B9@mindspring.com>

Well, COBOL does to, but that doesn't mean I'd write any of it.  :-)

John Jensen wrote:
> 
> David Young <dyoung at pobox.com> wrote:
> : Is there a program that converts Python source to Perl source?  There is
> : a program that a client wants written in Perl so that their Perl-only
> : programmers (can you imagine?) can hack on it later. I haven't the time or
> : inclination to learn Perl, and this project fairly cries out for Python,
> : but I cannot justify the time expense of translating from Python to Perl
> : by hand.
> 
> I think I lean towards Python for beauty, but Perl does have its role in
> commerce:
> 
>         http://www.robojob.com/lang.html
> 
> It might be good to know both, to use the one you prefer on projects you
> control, but to use the other when the client specifies.
> 
> John

-- 
---------------
Jesse D. Sightler
http://www.biddin.com/pyPit/



From kikutani at sprintmail.com  Tue Jul 27 01:32:48 1999
From: kikutani at sprintmail.com (Kikutani Makoto)
Date: 27 Jul 1999 05:32:48 GMT
Subject: Qt
Message-ID: <7njga0$361$1@ash.prod.itd.earthlink.net>

There are kdebinding and qtbinding in:
    ftp://ftp.kde.org/pub/kde/devel/

I guess qtbinding is obsolete and kdebinding should be used.
Am I right ?

I'm afraid that kdebinding is not updated recently, either.

-- 
Kikutani, Makoto  kikutani at sprintmail.com




From billtut at microsoft.com  Fri Jul 30 18:54:57 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Fri, 30 Jul 1999 22:54:57 GMT
Subject: Need help with TCP/IP client access from Windows
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90FF4@RED-MSG-50>


> From: Eric S. Raymond [mailto:esr at thyrsus.com]
> 
> 
> Gordon McMillan <gmcm at hypernet.com>:
> > > OK.  Will giving the filename ".\\intriguer" work for this?  And,
> > > given that the filename is going through the C library, 
> will leaving
> > > it as "./intriguer" work?
> > 

Well, as long as your current working directory is what you want it to be,
sure it'll work, otherwise you'll need to pull the standard "where did I get
run from" nonsense.

Bill



From Fred.L..Drake at p98.f112.n480.z2.fidonet.org  Thu Jul  1 16:33:14 1999
From: Fred.L..Drake at p98.f112.n480.z2.fidonet.org (Fred L. Drake)
Date: Thu, 01 Jul 1999 21:33:14 +0100
Subject: print color strings?
Message-ID: <00001637@bossar.com.pl>

From: "Fred L. Drake" <fdrake at cnri.reston.va.us>


Gerrit Holl writes:
 > is there an easy way to print color strings in python?
 > 
 > A function like printcolor('brightyellow', 'red', 'spam'), that finds
 > out on what kind of terminal you are and prints it to stdout?

Gerrit,
  If you're doing this on Unix, take a look at the ncurses library;
there's a Python interface to that.  (I think Oliver Andrich did it;
searching DejaNews should find it fairly quickly since this keeps
coming up.)
  The standard curses module will be documented in the next release of 
the Library Reference.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From Donn.Cave at p98.f112.n480.z2.fidonet.org  Thu Jul  1 17:55:55 1999
From: Donn.Cave at p98.f112.n480.z2.fidonet.org (Donn Cave)
Date: Thu, 01 Jul 1999 22:55:55 +0100
Subject: print color strings?
Message-ID: <00001640@bossar.com.pl>

From: donn at u.washington.edu (Donn Cave)

"Fred L. Drake" <fdrake at cnri.reston.va.us> writes:
|Gerrit Holl writes:
|> is there an easy way to print color strings in python?
|> 
|> A function like printcolor('brightyellow', 'red', 'spam'), that finds
|> out on what kind of terminal you are and prints it to stdout?
|
| Gerrit,
|   If you're doing this on Unix, take a look at the ncurses library;
| there's a Python interface to that.  (I think Oliver Andrich did it;
| searching DejaNews should find it fairly quickly since this keeps
| coming up.)
|   The standard curses module will be documented in the next release of 
| the Library Reference.

And if that doesn't work, probably your termcap or terminfo definitions
lack the necessary information about color escape sequences.  That's a
widespread problem.

The color escape sequences that have worked for me on a couple of terminal
types have been of the form <ESC>[32m, where 32 is a number from 30 to 37.
I use only 31 to 36.  30 and 37 are more or less black and white, but the
details vary, so I switch from color to non-color with <ESC>[m instead of
trying to sort out the details of black and white vs. background and
foreground.

Of course these are terminal specific escape sequences.  They're evil and
you must not use them, especially if you have a choice.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu





From mal at lemburg.com  Thu Jul  8 15:09:44 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 08 Jul 1999 21:09:44 +0200
Subject: Why are index() and count() only for mutable sequences?
References: <004701bec95f$186faa80$a0ee6dd1@BobAlexDesk>
Message-ID: <3784F778.6A6B4568@lemburg.com>

Bob Alexander wrote:
> 
> There is no modification of the sequence performed by these functions
> -- it seems as though they could be available for all sequences. It
> would certainly be useful for tuples as well as lists (strings, too,
> for that matter, even though we have the more general string.find()).

You could try count(), index(), forall() and exists() from mxTools:

	http://starship.skyport.net/~lemburg/mxTools.html

Albeit, they work with conditions (functions that return 1/0) to
implement the testing process.

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   176 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From nascheme at ucalgary.ca  Sat Jul 17 00:01:11 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 17 Jul 1999 04:01:11 GMT
Subject: Python and Boehm-Demers GC, I have code.
References: <000901becffa$b0f25e00$91a22299@tim>
Message-ID: <7mov67$bfg@ds2.acs.ucalgary.ca>

Tim Peters <tim_one at email.msn.com> wrote:
>[Neil Schemenauer]
>> I have added the Boehm-Demers garbage collector to Python and it
>> works well!
>> ...
>
>That's good, but I'm not clear on what this means.  The patch didn't seem to
>do more than replace malloc/calloc/realloc/free with the BDW versions; in
>particular, Py_USE_GC_FREE was #define'd, and you ran tests without
>significant cyclic structures, so it doesn't *appear* that anything more
>happened here than that a different implementation of the malloc family got
>plugged in.  Did the "collection" phase of BDW find any trash at all?

It doesn't unless you create some.  The patch creates a version
of Python that uses GC only to clean up reference cycles.

>... it would be great if BDW could be invoked
>"just sometimes" to reclaim the cycles RC can't catch.

That is what is happening now.

>semi-encouraging-ly y'rs  - tim

Thanks.


    Neil


BTW, _tkinter is not working perfectly yet.  I think lambdas
passed as callbacks are getting collected.



From brian.pedersen at mail.danbbs.dk  Mon Jul 26 17:40:18 1999
From: brian.pedersen at mail.danbbs.dk (Brian Pedersen)
Date: Mon, 26 Jul 1999 21:40:18 GMT
Subject: Python and Red Hat Linux 6
References: <379CD547.22A10A50@postoffice.pacbell.net>
Message-ID: <379CD5C2.5BA096A2@mail.danbbs.dk>

Rob wrote:
> 
> When programming in Perl or Python, I am using a header like the
> following on an executable file to cause the shell (bash or csh) to run
> the scripts as Perl/Python files:
> #!/usr/bin/perl   for Perl   and #!/usr/bin/python   for Python. I've
> checked and the compilers are located in the appropriate directories. I
> keep getting this when I run scripts:
> bash: <script name>: command not found. Does Linux support the '#!'
> notation in the file header, or do I always have to run the scripts by
> explicitly typing in 'perl' or 'python' before the script name?
> ~Rob~

Try 

!#/usr/bin/env python 

instead (don't ask me why, i read it in a FAQ somewhere :-)

--                                                             
Brian Pedersen, M.Sc.E.E.                         /\               
DSP Software and Algorithm Development       .~. /  \ .~.
URL: http://www.danbbs.dk/~kibria/brian/        ~    ~
NOTE: New e-mail address !



From skip at mojam.com  Tue Jul 27 18:45:44 1999
From: skip at mojam.com (Skip Montanaro)
Date: Tue, 27 Jul 1999 17:45:44 -0500 (CDT)
Subject: Python Performance
In-Reply-To: <379e2890.6069849@news.erols.com>
References: <3794C642.AB6A3E5D@freenet.co.uk>
	<7n5f0c$snu$1@ffx2nh5.news.uu.net>
	<al81zdxkktf.fsf@sirppi.helsinki.fi>
	<p51zdvn4c7.fsf@bidra241.bbn.hp.com>
	<14236.32616.980578.536614@dolphin.mojam.com>
	<379D89BC.DA91BA8C@hp.com>
	<14237.60407.130426.461846@dolphin.mojam.com>
	<14238.335.458728.140169@lrz.de>
	<14238.1372.334823.378192@dolphin.mojam.com>
	<379e2890.6069849@news.erols.com>
Message-ID: <14238.12851.809608.195391@dolphin.mojam.com>

    Robert> Just to address this for-instance, you can compile Python
    Robert> extensions for Win32 systems using the free egcs compiler.  See
    Robert> <URL: http://starship.python.net/crew/kernr/mingw32/Notes.html>
    Robert> for more information.

That's all well and good.  This is why I mentioned "technical depth".
Someone who's done very little programming, discovers Python and is merrily
whacking things into shape, then finds they need to recode a hotspot or two
in C is quite probably not going to have the mental model built up in their
brain to allow them to connect egcs, Python, extension modules, compilation,
dll building, (yadda, yadda, yadda) into a coherent whole.  It's also
unlikely they'll find Robert's excellent notes about egcs & Python on
Windows.

I am a Unix weenie with an MS in computer science.  I found my brief brushes
with the Windows C development environment terrible affairs ("where's make?"
"where's the debug flag?" "where's the *&!@+ debugger?" - to the point that
I simply installed cygwin and tried to make things look as much like Unix as
possible).  Most Windows weenies with CS degrees would probably find the
sort of Unix environment I happily muck around in on a daily basis equally
bothersome.  I think those of us who program for a living (or at least have
a good reason to invest a lot of time in writing software) all tend to
forget how much time and effort we've invested (over many years) learning
the systems we use and take them pretty much for granted.  I don't think
many of the people who write to help at python.org or the python-tutor list are
prepared to expend the effort necessary to learn the magic incantations -
not based on many of the questions I read there anyway.

Q-E-D-ly y'rs

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From jcw at equi4.com  Wed Jul 21 07:20:02 1999
From: jcw at equi4.com (Jean-Claude Wippler)
Date: Wed, 21 Jul 1999 13:20:02 +0200
Subject: Tcl/Python: Who was first?
References: <932552685.620225@news.adtranzsig.de>
Message-ID: <3795ACD5.FCFF67A1@equi4.com>

Dirk-Ulrich Heise wrote:
> 
> I just saw a message called
> 
> README: Tcl-URL! - the weekly guide to Tcl resources
> 
> on comp.lang.tcl.announce. (Reading it for the first time in my life)
> Did Python steal this (good) idea from Tcl or was it the other way round?

No and no.

This was first started for Tcl by yours truly last year, and quickly
caught on with the help of Cameron Laird, Mark Roseman, and Matt Newman.
 The archive for Tcl-URL! is stored and searchable on the "Tcl'ers Wiki"
(which is clearly another of those grand ideas by yours truly, now that
I'm into self-glorifying mode anyway):
	http://purl.oclc.org/thecliff/tcl/url.html
	http://purl.oclc.org/thecliff/tcl/wiki/

Tcl-URL! is also tracked and stored on the Scriptics site and on Linux
Weekly News, and a few others I believe.  If you scroll to the bottom of
the first URL page just mentioned, there is a little blurb about the
background of this.

Cameron and I, both multi-lingual and scripting enthousiasts by choice,
considered this simple idea to be easily graftable onto other scripting
communities.  So we started a similar setup for Python, but somehow it
never reached a stable weekly state.

Which is a big, big, pity if you ask me.  All it would take is for one
of the leading Pythoneers to take the Python-URL! stick for a month or
so, and I'm sure we'd be having Python-URL! back on track in no time. 
Soooo... GvR, TP, DA, MH, CT, GM, anyone want to bring this thing back
on track?  I'm sure many others would be more than happy to follow your
lead and take things from there later on.

If you're interested, please contact Cameron or me.  We can take care of
all the details and the mechanics, and I'd be happy to commit to do just
that for a long time, but the real issue here is that volunteers need to
be willing to come up with a list of 5 to 10 one-liners and URLs they
wish to highlight each Monday.

Tcl-URL! works well IMO, because it gives an interesting peek into what
key players choose and find important, and because it is gradually
building up a stable set of links to main topics, announcements, and
discussions.  It has become part of the Tcl folklore by now, I guess,
now that John Ousterhout has mentioned it on several occasions.

-- Jean-Claude



From tim_one at email.msn.com  Sat Jul 17 00:34:34 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Jul 1999 00:34:34 -0400
Subject: I envy newbies
In-Reply-To: <1279991447-21085630@hypernet.com>
Message-ID: <000101bed00d$ab59e4a0$d2a22299@tim>

[Gordon McMillan]
> ...
> I think the Tcl install does modify your path.

It may under NT (can't remember), but definitely does not under W95.

> But Python will go looking for it in the default places. Which is,
> of course, C:\Program Files\...

Which should really be taken from the registry, at key

    SOFTWARE\Microsoft\Windows\CurrentVersion\ProgramFilesDir

I think Guido has been thoroughly Americanized <wink>.

> [I think it's hilarious that MS was so enamored of their "advance" in
> being able to put spaces in paths that they made the "proper" place
> to install everything a demo of this "feature"...
> Which demos the fact that spaces in path names are a pain in the 
> ass].

yup!-and-they're-regretting-it-all-the-way-to-the-bank-ly y'rs  - tim





From tim_one at email.msn.com  Thu Jul 15 22:15:20 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 15 Jul 1999 22:15:20 -0400
Subject: Stackless & String-processing
In-Reply-To: <7mlu8n$6ug$1@brick.cswv.com>
Message-ID: <000c01becf31$0da1b080$c7a22299@tim>

[Tim]
> BTW, generators are much easier to implement than coroutines -- ...

[Neel]
> What's the difference, exactly? AFAICT you need to save the execution
> state when suspending both coroutines and generators, but I might be
> missing something obvious....

Just that a generator (in the Icon sense) is a semi-coroutine:  it always
suspends directly to its caller, at the point it was called.  A world of
difference follows from that:

1) From the caller's point of view, it's an ordinary call.  So on *that* end
of it, generators are no harder than vanilla procedure calls.

2) From the generator's point of view, "the execution state" is a single
frame (its own).  A coroutine can transfer to any other at any time, with
intervening calls piling up an arbitrary number of frames in the state.

3) From a recursive interpreter's point of view, the implementation language
(C, in Python's case) can't get any of its own frames stuck between caller
and callee during suspension of a generator (the generator *returns* to its
caller; it doesn't *resume* it; this asymmetry is the heart of it).

It's #3 (dealing with an intertwined C stack) that requires
platform-specific hackery to make coexps work in Icon, while the lack of C
stack pollution is why generators can work without platform-specific cruft.

Continuations and threads are harder still to implement (efficiently) -- the
less restrictive the rules, the less the implementation can exploit to make
its life easier.

a-phenomenon-you-may-observe-again-some-day<wink>-ly y'rs  - tim





From Gordon.McMillan at p98.f112.n480.z2.fidonet.org  Fri Jul  2 06:22:44 1999
From: Gordon.McMillan at p98.f112.n480.z2.fidonet.org (Gordon McMillan)
Date: Fri, 02 Jul 1999 11:22:44 +0100
Subject: Debugging embedded pgm under WinNT
Message-ID: <0000168f@bossar.com.pl>

From: "Gordon McMillan" <gmcm at hypernet.com>

Kevin Rodgers wrote:

> I have Python 1.5.2 embedded in an application on WinNT (4.0, SP4)
> using Visual C++ 6.0 SP2.  When I compile the project in Debug mode,
> even though it links correctly to the python15_d.lib file, it won't
> run because it complains about "can't find python15_d.dll".  So I
> got the source, compiled the python15_d.dll, put it in
> WINNT/system32 right next to python15.dll, ran my program . . . and
> now it couldn't find "multiarray" (my embedded Python code uses the
> Numeric extensions).  I verified that sys.path was correct and
> included the Numeric directories. Do I have to build the Numeric
> extensions under Debug mode? 

Yup. A pain in the butt, ain't it? 

Everything has to be running the same c runtime lib. However, you can 
compile python15.dll and your stuff in debug mode but using the 
release c runtime (Multithreaded DLL). This at least lets you debug 
your stuff and step into python15.dll. Other stuff will load, but 
won't have any symbols.  As I recall I had to tweak quite a few 
things in the build to get that to work, too.

- Gordon




From gmcm at hypernet.com  Tue Jul  6 10:43:46 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 6 Jul 1999 09:43:46 -0500
Subject: Newbie question: embedding python in a c++ main
In-Reply-To: <3781F81F.D8730AEE@platonix.com>
Message-ID: <1280853816-64731835@hypernet.com>

Yishai Beeri writes:

> I am trying to embed the python interpreter in a c++ program. All it
> does is a simple main, which initializes python, initializes several
> python modules (extensions written in c++), and run a python
> program.
> 
> When I read the python program into a char* and use
> PyRun_SimpleString, it seems to work fine. However, when I try to
> use PyRun_SimpleFile, after opening the file, I get a crash
> somewhere inside the PyParser functions.

The most likely problem is using a different C runtime lib in your 
program than is used by python. Fits the symptoms (works until you 
try to pass a FILE around), though without knowing the specific error 
this is just a guess.

python15.dll is linked to the multithreaded DLL C runtime.

- Gordon



From amk at mira.erols.com  Fri Jul  9 19:34:12 1999
From: amk at mira.erols.com (A.M. Kuchling)
Date: 9 Jul 1999 23:34:12 GMT
Subject: PyApache and Apache 1.3.6 anyone ???
References: <19990708221333.B16851@hoffleit.org> <3785836D.3BE3EBC7@cs.ucsd.edu> <19990709112714.A2615@hoffleit.org>
Message-ID: <7m9tn7$6qg$49@rks1.urz.tu-dresden.de>

On Fri, 9 Jul 1999 11:27:14 +0200, 
	Gregor Hoffleit <flight at mathi.uni-heidelberg.de> wrote:
>On Thu, Jul 08, 1999 at 10:06:53PM -0700, Pavel Velikho wrote:
>Nope, there's no entry at all in the error log for all the silently
>dying starting attempts. I don't know how to debug apache with gdb,
>therefore I'm comletely lost where to look for hints about the
>failure.

This is probably because Apache is segfaulting for some reason,
something that's become very familiar since I started working on
mod_pcgi.  To debug an Apache binary with GDB, run "gdb httpd", and
then run Apache with the -X command-line argument.  -X disables
forking, so everything is handled in a single process; you may also
want to specify "-f <configfile".  Then you can set breakpoints, get
stack traces at segfaults, and all that sort of thing.  

	Running Apache under GDB is really slow, though; an
application such as Zope that makes lots of little HTTP requests will
run glacially slowly.  That shouldn't be a problem for getting
PyApache working.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
It is a rotten world / Artful politicians are its bane / Its saving grace is
the / Artlessness of the young / And the wonders of the sky.
    -- Epitaph, Ross Bay Cemetery, Victoria



From mwh21 at cam.ac.uk  Sun Jul 18 08:55:04 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 18 Jul 1999 13:55:04 +0100
Subject: scoping assertion
References: <37839491.3F7C1F6@mikemccandless.com> 		 <3783C878.B8C3790B@bioreason.com> <19990708211651.A17658@optiplex.palga.uucp> <7cQP0BAYKRh3Ew6K@jessikat.demon.co.uk> <378DB7F5.1A48506A@unisys.co.jp>
Message-ID: <m3aesu9kyf.fsf@atrus.jesus.cam.ac.uk>

Isao AKIYAMA <Isao.Akiyama at unisys.co.jp> writes:

> > I use
> > try:
> >         raise None
> > except:
> >         ....
> 
> Wow!
> 
> >>> import sys
> >>> try: raise None
> .. except: print sys.exc_info()
> ..
> (<class exceptions.TypeError at 89a4e0>, <exceptions.TypeError instance
> at 8cc17
> 0>, <traceback object at 8cc150>)
> >>>
> 
> 'raise None' actually raises a TypeError ;-)

Well, it would. You can only raise strings, classes or instances as
exceptions:

Python 1.5.2+ (#2, Jul 14 1999, 00:03:34)  [GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> raise 4
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: exceptions must be strings, classes, or instances
>>> 

So Python gets the thing being raised (None), checks it's type, finds
it wanting - and raises an exception. Ho hum. I cant work out whether
that is a) trivial b) deep c) working, but by lucky accident or d)
silly or e) this post is a complete waste of time. Or all of the
above.

Cheers,
Michael
 
> Isao AKIYAMA
> 

-- 
Oh, very funny. Very sar-cah-stic.         http://www.ntk.net
http://www.ntk.net/doh/options.html - ho ho



From wang at rjka.com  Wed Jul 28 13:03:17 1999
From: wang at rjka.com (Huayin Wang)
Date: Wed, 28 Jul 1999 12:03:17 -0500
Subject: A Zope build question: where is sha.py
Message-ID: <199907281703.MAA05703@ws89.rjka.com>

Fredrik Lundh wrote:

> (some hints, though: are you using Python 1.5.2?
> have you configured Python to include the sha
> module -- it's on by default, but you maybe you're
> using old configuration files...).

I am amazed about your hint.  IT IS the old "Modules/Setup" that throw me off.

thank you very much! 
my thanks goes to Skip Montanaro too!




From reggie at merfinllc.com  Wed Jul  7 20:08:10 1999
From: reggie at merfinllc.com (Reggie Dugard)
Date: Wed, 7 Jul 1999 17:08:10 -0700
Subject: Pmw activate method
Message-ID: <7m0q5b$6ba@dfw-ixnews11.ix.netcom.com>

Please forgive me if this is a stupid question, because I'm pretty new to
Python and really new to Pmw.

I'm writing an GUI application using Pmw on NT 4.0 SP5 and, during
debugging, I call it repeatedly during a given python interactive session.
As part of the application, I use some dialogs which I display using the
activate method.  My problem is, once I've activated a dialog during a
"run", any subsequent "runs" in the same python session will generate the
following background error if I attempt to activate a dialog:

Error: 1
TclError Exception in Tk callback
  Function: <method test.dialog of test instance at 898fe0> (type: <type
'instance method'>)
  Args: ()
Traceback (innermost last):
  File "...\Python\Lib\Pmw\Pmw_0_8\lib\PmwBase.py", line 1349, in __call__
    return apply(self.func, args)
  File "test2.py", line 18, in dialog
    dialog.activate()
  File "...\Python\Lib\Pmw\Pmw_0_8\lib\PmwBase.py", line 846, in activate
    showbusycursor()
  File "...\Python\Lib\Pmw\Pmw_0_8\lib\PmwBase.py", line 1033, in
showbusycursor
    if window.state() != 'withdrawn':
  File "...\Python\Lib\lib-tk\Tkinter.py", line 861, in wm_state
    return self.tk.call('wm', 'state', self._w)
TclError: can't invoke "wm" command:  application has been destroyed

I have found that the root window does not seem to get cleared from the
Pmw._toplevelBusyInfo dictionary and if I manually delete it, the error goes
away, but I'm sure there is a better way of fixing this problem.

A typical run looks like:

H:\python>python
Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import test2
>>> a=test2.test()  <-- Deletion necessary if dialog was activated
Deleting root
>>> a=test2.test()  <--- No deletion if I don't activate the dialog during
the run
>>>


Here is my test2 code to give you a better idea of what I'm doing.  Any
advice would be appreciated.
import Tkinter, Pmw

class test:
    def __init__(self):
        root = Tkinter.Tk()
        Pmw.initialise(root)
        button = Tkinter.Button(text = 'Dialog', command = self.dialog)
        button.pack(pady = 20)
        exit = Tkinter.Button(text = 'Exit', command = root.destroy)
        exit.pack(pady = 20)
        root.mainloop()
        if hasattr(Pmw,'_toplevelBusyInfo') and
Pmw._toplevelBusyInfo.has_key(root):
            print 'Deleting root'
            del Pmw._toplevelBusyInfo[root]

    def dialog(self):
        dialog = Pmw.MessageDialog(title = 'Warning', message_text =
'Testing')
        dialog.activate()

--
Reggie Dugard

Merfin, LLC                             Voice:  925-937-4560
1460 Maria Lane, Suite 420              FAX:    925-937-4530
Walnut Creek, CA 94596                  Mailto:reggie at merfinllc.com





From bucodi at worldnet.fr  Mon Jul 19 09:09:23 1999
From: bucodi at worldnet.fr (pserrand)
Date: Mon, 19 Jul 1999 15:09:23 +0200
Subject: designing question
References: <1279735663-4809742@hypernet.com>
Message-ID: <37932383.A9AA3B3B@worldnet.fr>

Gordon McMillan wrote:

> Pierrette writes:
>
> > I have two questions:
> > 1. The database will hold abouth 50.000 lines in its biggest table,
> > can gadfly handle this without problems, what abouth acces time ??
>
> Gadfly is in-memory, and (to support transactions) has 2 copies of
> the db. If the machines have the memory, access time will be fine.

With the cost of memory now this should be no problem :) All machines we
have or that we deliver to clients has at least 128 MB

>
>
> > 2. The application will have abouth 50 different inputscreens wich
> > acces all to different tables. I will create a class for every
> > window wich inheret from a general 'input-screen'-class. I wonder if
> > a should place the db-access code (insert-update-delete) in every
> > specific window or if i should  create a big db class wich has all
> > possible db-acces in it ?
>
> Completely philosophical question. For flexibility, I favor the GUI
> making no assumptions about how the data is stored. I would have the
> GUI making calls like "app.addthingie(thingie)", where the app has an
> "abstract" UI.

I'm surprised that this is just a philosophical question :) I tought
there would be some
practical reasons to do it one or another way (p.e. maintenance of the
application)
The reason i posted this question is that i have very little experience
with OOP. (actually i learned it with python <wink>) So i'm looking for
some experience here

>
>
> Of course, it almost nevers works out that cleanly in practice, but
> it's certainly a worthy goal.
>
> Now if each screen maps exactly to a table, this is probably
> overkill. But if your data is normalized, this makes a lousy UI -
> users usually want to know all kinds of silly things, like whether
> there's a "next" record, or the owning customer's name (not the
> foreign key in this row, which is just a number) that violate
> normalization.
>
> - Gordon

Thanks for your quick response Gordon

Pierrette




From da at ski.org  Fri Jul  2 15:30:55 1999
From: da at ski.org (David Ascher)
Date: Fri, 2 Jul 1999 12:30:55 -0700 (Pacific Daylight Time)
Subject: Debugging embedded pgm under WinNT
In-Reply-To: <1281178906-45167517@hypernet.com>
Message-ID: <Pine.WNT.4.04.9907021229500.310-100000@rigoletto.ski.org>

On Fri, 2 Jul 1999, Gordon McMillan wrote:

> > WINNT/system32 right next to python15.dll, ran my program . . . and
> > now it couldn't find "multiarray" (my embedded Python code uses the
> > Numeric extensions).  I verified that sys.path was correct and
> > included the Numeric directories. Do I have to build the Numeric
> > extensions under Debug mode? 

FWIW, I have a version of compile.py (the thing which builds Numeric)
which, when debugged, will allow building multiple versions automatically,
both Release and Debug (thanks to MHammond).

--david




From gmcm at hypernet.com  Tue Jul 13 10:12:29 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 13 Jul 1999 09:12:29 -0500
Subject: Running 2 gui libs simultaneously
In-Reply-To: <7mf2rn$dj9$1@news1.saix.net>
Message-ID: <1280251163-5463910@hypernet.com>

Riaan Booysen writes:
> 
> How difficult would it be to run 2 window libraries simultaneously
> (Tcl/Tk / wxPython) in one Python program. The problem is that both
> need an event processing loop ( lib.mainloop() ). Are threads my
> only solution?

You don't have to use Tk's mainloop. You can call update() in a loop. 
Perhaps wxWindows has something similar, or you can call Tk's update 
from within some idletask in wxWindows.

Not something I'm likely to try...

- Gordon



From pvelikho at cs.ucsd.edu  Fri Jul  9 01:06:53 1999
From: pvelikho at cs.ucsd.edu (Pavel Velikho)
Date: Thu, 08 Jul 1999 22:06:53 -0700
Subject: PyApache and Apache 1.3.6 anyone ???
References: <19990708221333.B16851@hoffleit.org>
Message-ID: <7m9tlo$6qg$5@rks1.urz.tu-dresden.de>

Gregor Hoffleit wrote:

> Is anybody successfully running PyApache on an Apache 1.3.6 ?
> Preferably on a Linux system ? When I build PyApache and load it from
> httpd.conf, /usr/sbin/apache just doesn't start up. It doesn't give
> any error or what, it just starts up and quits immediately.

>
> Any experiences ?

Did you check the error log? If it is a configuration error it will
usually keep quiet about
it and just write it in the log file.
Hope this helps

>
> Thanks in advance,
>
>         Gregor

Pavel




From kernr at mail.ncifcrf.gov  Fri Jul  2 18:30:22 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Fri, 02 Jul 1999 22:30:22 GMT
Subject: Python GUIs: Portability or not?
References: <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net> <377B711A.E8B65E0B@cfdrc.com> <377BC620.F5553FF5@easystreet.com> <b%Pe3.4033$lr3.164497@news2.jacksonville.net> <14204.4556.177837.851763@lrz.de> <Pine.LNX.4.10.9907012309060.6263-100000@us2.mayo.edu> <5DEB4D7CBF56EC8D.37F29301ECE96FC0.D3C7C2F2D9068D65@lp.airnews.net> <7li101$7fg$1@nnrp1.deja.com>
Message-ID: <377d3cd9.5405704@news.erols.com>

On Fri, 02 Jul 1999 09:32:48 GMT, Georg Mischler
<schorsch at schorsch.com> wrote:

>Mikael Lyngvig wrote:

[snip]

>There are other X servers around, but those who blend into the
>host GUI (instead of creating their own X desktop within a window)
>all seem to be commercial.
>
>Now *this* would be a task for an open software project... ;)

http://siddiqi.webjump.com/
It doesn't work yet, but it *is* an open source project.  Not sure if
they are aiming for the "blending," though.

>Have fun!
>
>-schorsch
>
>--
>Georg Mischler  --  simulation developper  --  schorsch at schorsch.com
>+schorsch.com+  --  lighting design tools  --  http://www.schorsch.com/

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From Gaetan_Corneau at baan.com  Wed Jul 14 14:16:20 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 14 Jul 1999 18:16:20 GMT
Subject: Redirect stdout to file
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF47C@ex-quebec-u1.baan.com>

Frederik,

I tried it and tought it didn't work, but it seems it was my fault. I did it
this way:

from sys import stdout
stdout = open("out.txt", "w")

I changed to:

import sys
sys.stdout = open("out.txt", "w")

and it worked. Do you know why?

Thanks a lot for your help!
______________________________________________________
   Gaetan Corneau
   Software Developer (System integration Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"

-----Original Message-----
From: Fredrik Lundh [mailto:fredrik at pythonware.com]
Sent: Wednesday, July 14, 1999 2:04 PM
To: Gaetan Corneau
Cc: Python Mailing List
Subject: Re: Redirect stdout to file


Gaetan Corneau <Gaetan_Corneau at baan.com> wrote:
> I'm in Barneveld Holland right now (hey, maybe I'll go visit the famous
> Stichting Mathematisch Centrum ;) and I would like to redirect output from
> the print statement to a text file. I think a saw a message about that
some
> time ago, but I can't find it right now.

try this:

import sys
sys.stdout = open("my.log", "w")
print "this ends up in the logg file"

also see: http://www.python.org/doc/FAQ.html#5.8

</F>



From ivnowa at hvision.nl  Tue Jul 27 18:09:29 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Tue, 27 Jul 1999 22:09:29 GMT
Subject: 1.5.2 for: else:
In-Reply-To: <Pine.LNX.4.04.9907271632450.9849-100000@emerald.netskate.ru>
References: <Pine.LNX.4.04.9907271632450.9849-100000@emerald.netskate.ru>
Message-ID: <199907272208.AAA06264@axil.hvision.nl>

On 27 Jul 99, Oleg Broytmann wrote:

> > Loop statements may have an else clause; it is executed when the loop
> > terminates through exhaustion of the list (with for) or when the
> > condition becomes false (with while), but not when the loop is
> > terminated by a break statement. [..]
> [skip]
> > Else clauses get executed on normal loop exit, even if the loop was
> > 'empty'. It does not get executed when you 'break' out of a loop.
> 
>    Aha, thanks. I expected "else" to be executed when "for" not executed
>    on
> empty list. Wrong assumption.

This one has bitten me too in the past. One of the few non-intuitive 
parts of Python, methinks. :(

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
Python Snippets: http://www.hvision.nl/~ivnowa/snippets/
The Purple Kookaburra Forum: http://www.delphi.com/kookaburra/



From news at dorb.com  Fri Jul  2 13:58:39 1999
From: news at dorb.com (Darrell)
Date: Fri, 2 Jul 1999 13:58:39 -0400
Subject: C shared object files
References: <199907021717.KAA03837@ogl.llnl.gov>
Message-ID: <w77f3.4656$U6.26648@newsr2.twcny.rr.com>

Check the docs under "Extending and Embedding" and search those key words at
dejanews.com.
Also check out some examples of how other people have done it, which will
show up in your search.
--
--Darrell


Martin Casado <casado at ogl.llnl.gov> wrote in message
news:199907021717.KAA03837 at ogl.llnl.gov...
>
>   Python Guru's,
>
>    I am trying to get python to dynamically load a C object module, and
> first off I can't figure out how to compile myCollProggy.c into a
> shareable object file, and secondly I was curios if there is a way to
check
> at runtime if a module is on the system or not.
>
>   I have made arrangements with the proper authorities and there is a
veritable
> wealth of Karma points waiting for anyone who can help me :-)  Hehe thanks
> very much in advance.
>
> -Martin
>
>





From tim_one at email.msn.com  Fri Jul 23 03:23:21 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 23 Jul 1999 03:23:21 -0400
Subject: Cyclops 0.9.4
In-Reply-To: <37978EBC.41BE8CE4@compaq.com>
Message-ID: <000101bed4dc$3e51e280$392d2399@tim>

[Tim]
> You want to know about objects still living that you don't expect
> to be living.  No system can answer that for you!  They can't know
> what you expect <wink>.

[Greg Ewing]
> I'm quite happy to go through the list of living objects
> and decide for myself which ones should be dead.

Nobody can sift thru hundreds of thousands of live objects by eye, which
happens routinely in large long-running apps.

> What I meant was, what I really want to know is how those objects
> got reached.

Well, 100,000 objects can often be reached in 500,000 ways via one direct
link -- & it gets messier the longer the path length you consider.  You have
to reduce the set of objects that are potentially interesting; Cyclops
requires you to identify them in advance; I'm not sure how it could be
easier to identify them later.

> If there are cycles involved, it wouldn't hurt to be told about
> them, but they're not the most important thing.

Cycles do have special status in CPython, though.  I find them the most
fruitful thing to look at first.  The more of them I see, though, I agree
they're not the whole banana they've been made out to be.

>> If what you want is a list of all reachable objects,

> No, I'd like a list of *all* objects, reachable or not.

All objects ever allocated, or all objects whose refcounts haven't yet
fallen to 0?

The former implies no object memory is ever recycled, which would make it
impractical for the large long-running apps most in *need* of help.

The latter requires a way to forget objects whose refcounts have reached 0,
and so likely requires a doubly-linked list of objects (so that dead
non-cyclic objects can unlink themselves efficiently when they die).  Since
"even ints are boxed" in Python, that's a major memory hit.

A short while back Guido pitched a compromise, maintaining a list of all (&
only) allocated dicts (whether dicts whose refcount had reached zero would
be purged from this list wasn't resolved).  Since each class and instance
object has a dict, this probably allows getting at 99% of problem cases.
Although descriptor-based slicing of large array objects seems to account
for all major leaks reported by others on c.l.py this month <wink/sigh>.

> Cyclops could then do what it does now, but for the isolated islands
> of garbage as well.

Isolated garbage in CPython necessarily contains cycles, and all isolated
garbage (cyclic or not) is necessarily reachable *from* cycles, so Cyclops
has *a* handle on that now.  It doesn't need more information so much as it
needs ways to present less to the user.  Let's say you have everything you
want, and even more:  for every object ever allocated, Cyclops2 can give you
a timestamped history of every change ever made to it, every path from which
it could be reached at any given time, and similarly for every object
reachable from it.

Then what?  Beats me -- every app seems different, and each requires a lot
of thought to whittle the last trillion machine cycles down to the seven
that matter.  More info doesn't help without a clear plan for exploiting it.

>> All of those are also needed if Python is ever to move toward builtin
>> portable mark-&-sweep (optional or not).  The advantage over "walking an
>> (explicit) list" is no overhead (time or space) unless & until
>> it's used.

> If there were mark & sweep there wouldn't be unreachable
> garbage, so in that case you are right -- there would be
> no point in keeping a list. I'm only suggesting it as an
> interim measure.

No, what I described is essential for the M half of portable M&S, but on top
of that a list is essential for the S half.  Without a list of allocated
objects, Python *can't find* the "isolated islands of garbage" in order to
clean them up -- it could mark all reachable objects, but couldn't sweep the
unreachable ones.

You should take that as good news:  if Python is ever to *reach* builtin
portable M&S, the list you want will be there in one form or another.  But
in my repeated experience here, unreachable garbage is not the problem --
it's reachable stuff you expected *would* be garbage but isn't; and a list
isn't needed to find that.

>> then a Cyclops-like thingy could tell you not only that something
>> unexpected is still alive, but also from where it can be reached

> I don't understand why Cyclops couldn't be made to do
> that now. It reached those objects somehow -- all it needs
> to do is remember how!

I think you should try using it once <wink>.  You have to explicitly
register the set of objects you're curious about, and in practice *that's*
(i.e., your registration) how it "reached them".  "Yes, object X is
reachable in 12 ways, and here are the ways I know about:  (1) you
registered it.  (2) umm ... well, no other way I can find -- register more
stuff, chase more types, filter less away, and try again".

There was once another pile of code to trace out all ways of reaching each
registered object from all other registered objects, but that didn't turn up
anything but cycles (which it finds more cheaply via other means), plus
hundreds of thousands of useless paths ("OK, A.x can be reached from A"), so
I tossed it.  I think I'll put it back in, though -- one of these days it's
bound to stumble into something worth finding even if by accident <0.6
wink>.

> I don't mean to sound ungrateful for Cyclops, by the way --
> it's a great idea.

Greg, it's much more than an idea:  it's code you can use <smile>.

> All it needs is a little bit more help from the Python core to make
> it even greater...

The core help I think it could best exploit was covered in the preceding msg
(namely, the "M" half of M&S machinery).  That's not enough to find
unreachable islands unless an element of the island was registered, but
those appear to be the tail of the leaky dog.

canine-metaphors-in-service-of-a-barfworthy-problem-ly y'rs  - tim





From arcege at shore.net  Wed Jul  7 11:57:12 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Wed, 07 Jul 1999 15:57:12 GMT
Subject: Py_BuildValue, PyArg_ParseTuple
References: <37835CB3.2B876BE3@ncsa.uiuc.edu>
Message-ID: <sNKg3.1098$6M6.349834@news.shore.net>

Randy Heiland <heiland at ncsa.uiuc.edu> wrote:
: Here's a very simple program where I want to get back out
: (PyArg_ParseTuple) what I put in (Py_BuildValue) - however, I can't seem
: to do that.  Can someone PLEASE enlighten me?

: thanks,
: --Randy

: /* Py_BuildValue, PyArg_ParseTuple  usage */

: #include "Python.h"

: main(argc, argv)
:     int argc;
:     char **argv;
: {
:   unsigned char ba[3] = {2,4,8};  /* byte array */
:   PyObject *pobj;
:   char *cstr;
:   int len;

:   Py_Initialize();

:   printf("Py_BuildValue of %d %d %d\n",ba[0],ba[1],ba[2]);
:   pobj = Py_BuildValue("s#", ba,3);
:   if (pobj == (PyObject *)NULL)
:      printf("pobj is NULL\n");

:   printf("\nPyArg_ParseTuple(pobj,...)\n" );
:   PyArg_ParseTuple(pobj,"s#", &cstr,&len);
:   printf("len=%d\n",len);
:   printf("cstr=%s\n",cstr);

:   printf("cstr[0]=%c\n",cstr[0]);
:   printf("cstr[1]=%c\n",cstr[1]);
:   printf("cstr[2]=%c\n",cstr[2]);

:   printf("(int)cstr[0]=%d\n",(int)cstr[0]);
:   printf("(int)cstr[1]=%d\n",(int)cstr[1]);
:   printf("(int)cstr[2]=%d\n",(int)cstr[2]);
: }


PyArg_ParseTuple expects to be given a Python tuple object, not a
string.  The compliment to Py_BuildValue is PyArg_Parse.  Use that
instead of PyArg_ParseTuple and it should work. :)

  -Arcege




From bwarsaw at cnri.reston.va.us  Fri Jul 23 11:03:56 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Fri, 23 Jul 1999 11:03:56 -0400 (EDT)
Subject: [Tutor] I want to learn to hack
References: <19990721154610.B3663@optiplex.palga.uucp>
	<Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org>
	<14231.11726.916541.492153@nobot.cnri.reston.va.us>
	<slrn7pfq6v.f9h.bwinton@tor.dhs.org>
Message-ID: <14232.33884.244828.631178@anthem.cnri.reston.va.us>

>>>>> "BW" == Blake Winton <bwinton at tor.dhs.org> writes:

    BW> This was the first one I've seen in all my many years as a
    BW> reader of the tutor list...  Are you planning on forwarding
    BW> them to the tutor list, and wanted to test the waters, or did
    BW> you just want some ideas on how to handle them when they came
    BW> your way?

When I see such messages on webmaster at python.org I basically ignore
them.  The replybot on that addr gives all the information that
someone with an inkling of a clue needs to figure out where to go next 
(I hope -- if anybody thinks otherwise, feel free to suggest
improvements!).

I'm usually pretty good about hitting `d' on messages to webmaster
that aren't dealing strictly with the operation or content of the
site, but every once in a while I'll forward a message on to
python-help.  I usually don't forward to tutor.

-Barry



From ajung at sz-sb.de  Wed Jul 14 10:01:45 1999
From: ajung at sz-sb.de (Andreas Jung)
Date: Wed, 14 Jul 1999 14:01:45 GMT
Subject: How to trace Python/sys.settrace() ?
Message-ID: <19990714160145.A21443@sz-sb.de>

In some situations it is neccessary to be able to do a complete
trace of Python script (similiar to truss/strace) - that means one
want to know all called functions (function names) and functions parameters. 

sys.settrace() seems to be a candidate for building such a tracing functionality.
However I am not sure if this is the right way. I'd like to ask the Python gurus
and Python hackers here  about a possible solution. Maybe there is already some usable
code available.

Thanks in advance,
Andreas

-- 
                                _\\|//_
                               (' O-O ')
------------------------------ooO-(_)-Ooo--------------------------------------
   Andreas Jung, Saarbr?cker Zeitung Verlag und Druckerei GmbH
   Saarbr?cker Daten-Innovations-Center
   Gutenbergstr. 11-23, D-66103 Saarbr?cken, Germany
   Phone: +49-(0)681-502-1563, Fax: +49-(0)681-502-1509
   Email: ajung at sz-sb.de (PGP key available)
-------------------------------------------------------------------------------



From fredrik at pythonware.com  Tue Jul 13 18:38:10 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 13 Jul 1999 22:38:10 GMT
Subject: problem with SocketServer -- address already in use
References: <931899928snz@vision25.demon.co.uk>
Message-ID: <004201becd80$633c8090$f29b12c2@secret.pythonware.com>

Phil Hunt <philh at vision25.demon.co.uk> wrote:
> I'm trying to use SocketServer.py to write a TCP server. I ran
> my program with ``python serv.py'', which went into an infinite
> loop (which is what it should do). I exited my program with ^Z,
> and tried to start it up again. I got an error message:
> 
> socket.error: (98, 'Address already in use')
> 
> which I presume is because something thinks the old invokation of
> my program is still using port 7070. 
> 
> How do I clear the port, before I start up my server?

use setsockopt to enable local address reuse; see

    http://www.deja.com/getdoc.xp?AN=484041505&fmt=text

for an example.

</F>




From tbryan at arlut.utexas.edu  Fri Jul  9 09:45:11 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Fri, 09 Jul 1999 08:45:11 -0500
Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
References: <tot2m7.su.ln@final.math.uni-bonn.de> <19990708223856.A29441@optiplex.palga.uucp>
Message-ID: <7m9tm9$6qg$22@rks1.urz.tu-dresden.de>

Gerrit Holl wrote:

> I used HTMLgen and I don't really understand why to use it. If you don't
> know HTML, you can't use it: you need to understand tags like <B>, <I>, <U>
> and such. I strongly recommend to have a look at HTML:

If you know HTML, then you can use HTMLgen to help you maintain a 
consistent appearance from page to page on a web site.  In particular, 
many users find a consistent use of navigation aides (such as next page,
previous page, and home) to be very helpful.  It is also useful for 
helping multiple people working together to produce pages with a 
consistent appearance by using the same document base classes.
It also provides a nice abstraction for programatically generated
documents (for example, from CGI).  That way, you don't have to 
output raw HTML with print statements from your program.  Again, if
several CGI scripts should produce pages with a similar appearance, 
you can factor out the common parts into a class.  Then each program 
only needs to instantiate the class and fill in the extra information.

For a site of a dozen pages, HTMLgen is definitely unnecessary.
For a site of millions of pages, I think that HTMLgen would prove
insufficient.  For a site of a few hundred pages, I would probably
consider HTMLGen, especially if python scripts will be generating 
those pages (either as CGI scripts or as a cron job).  

As a bonus, it seems that it would be simple to use the HTMLgen
document classes to generate multiple formats.  I have been thinking
about writing output methods for LaTeX output. :)

-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From greg.ewing at compaq.com  Wed Jul 21 22:09:16 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 22 Jul 1999 14:09:16 +1200
Subject: Cyclops 0.9.4
References: <001501bed315$8888e740$d0a22299@tim>
Message-ID: <37967D4C.9D9434ED@compaq.com>

Tim Peters wrote:
> 
> 2. Unless you're fighting a bug in the interpreter's reference-counting code
> (very rare), those groups of unreachable objects necessarily contain cycles.
> That's why Cyclops focuses on cycles.

I'm not sure that focusing on cycles is the right thing to
do, then. The cycles themselves aren't necessarily wrong;
what's wrong is the code that was meant to break them is
buggy. But you don't find that out until you leak memory.

I think the Python core needs an enhancement to allow
walking a list of all allocated objects. Cyclops in
conjuction with that would be very useful. Maybe a compile
time or run time option, since it would slow down 
allocation/deallocation a little bit.

> That's the mode it's being used in with IDLE now:  things *are* still
> reachable that nobody expects "should be" reachable.  What Cyclops can't do
> (yet <wink>) is tell you from *where* they can be reached.  All it can say
> for now is that they don't appear to be in cycles -- they're still reachable
> "for some other" reason.

So, again, cycles aren't really what you want to know about!

Greg



From m.faassen at vet.uu.nl  Mon Jul 26 04:19:00 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 26 Jul 1999 08:19:00 GMT
Subject: Fw: ANNOUNCE: Mailing list for Australian Python Users
References: <007201bed42e$c4316c30$f29b12c2@secret.pythonware.com> <7n78nm$lic$1@news.accu.uu.nl> <379798BC.33FC0CA9@equi4.com> <19990723124741.A29309@optiplex.palga.uucp> <1E4AC88B7E04EAB0.8140D4DF3103F8D3.0124F5E4F4834F96@lp.airnews.net>
Message-ID: <7nh5lk$e46$1@newshost.accu.uu.nl>

Cameron Laird <claird at starbase.neosoft.com> wrote:

> In article <19990723124741.A29309 at optiplex.palga.uucp>,
> Gerrit Holl  <gerrit.holl at pobox.com> wrote:
>>But I don't see anyone using python in nl.comp.programmeren...

> Dutch is more popular among Pythoneers than Python is
> among the programming Dutch?!

This may be possible. But we're working on it -- it's just a matter
of time now. :)

I should take a look at nl.comp.programmeren sometime.

Regards,

Martijn




From nick at NOSPAMvideosystem.co.uk  Fri Jul  2 05:39:35 1999
From: nick at NOSPAMvideosystem.co.uk (Nick)
Date: Fri, 2 Jul 1999 10:39:35 +0100
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain>	<377B7E17.5BD7@nospam.com> <377bbf9b@news.xtml.co.uk> <377BD789.C4A1D191@appliedbiometrics.com>
Message-ID: <377c870b@news.xtml.co.uk>

The language doesn't necessarily need to support it. You could implement the
threaded system yourself in Lua, you then have more control over what goes
on. Proper threads are relatively slow in Windoze (I believe) and would be a
nightmare to debug? If Lua wasn't fast enough you could write a scheduler
with priority assessment in C/C++ and attached Lua script callbacks to this.
Personally I wouldn't dream of useing proper threads for game logic/AI I
would write a basic scheduler, it wouldn't be difficult. Grim Fandango uses
Lua as its scripting language and I think they did something along these
lines. I think Balders Gate is another game that is rumoured to use it.

N


Christian Tismer <tismer at appliedbiometrics.com> wrote in message
news:377BD789.C4A1D191 at appliedbiometrics.com...
>
>
>Nick wrote:
>>
>> Don't want to put the cat amongst the pigeons but Lua
>> (http://www.tecgraf.puc-rio.br/lua/), is a language specifically designed
to
>> be embedded and it runs about 50-100% faster than Python and is a lot
more
>> compact (30k or something). It has a lot of similarities syntactically as
>> far as I can see.
>>
>>  If I was to use an embedded scripting language in a game I would use Lua
>> not Python. If I was to write tools for the game I would use Python.
>
>I had a quick look, but not the time to check it completely.
>Does it have coroutines or small threads to do very
>quick context switches? This is something which
>Python will be able to do very fast, soon. You can have
>thousands of current contexts and switch them in no time.
>This can simplify the implementation of games in a scale,
>that the speed penalty might be compensated by that.
>Well, we have to try...
>
>ciao - chris
>
>--
>Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
>Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
>Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
>10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
>PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
>     we're tired of banana software - shipped green, ripens at home





From nick at NOSPAMvideosystem.co.uk  Tue Jul 13 04:03:33 1999
From: nick at NOSPAMvideosystem.co.uk (Nick Trout)
Date: Tue, 13 Jul 1999 09:03:33 +0100
Subject: idle
References: <000c01becce4$b1799e60$31a02299@tim>
Message-ID: <378af10a@news.xtml.co.uk>

I wish I could find the time to fix as well! Maybe I'll delve into CVS one
day.

Where-did-this-ly-y'rs-thing originate!-ly y'rs,
Nick.


Tim Peters <tim_one at email.msn.com> wrote in message
news:000c01becce4$b1799e60$31a02299 at tim...
>[Nick Trout]
>> I can't get breakpoints to work in the version I have, 0.4. I'm running
it
>> under Windows.
>>
>> Is this just me?!
>
>I doubt it, Nick.  I don't use debuggers much (& haven't used IDLE's at
>all), but you're not the first to mention this.  I installed a recent CVS
>version of IDLE for my coworkers a couple weeks ago, and they complain
>breakpoints (under Windows) can't be set under that either.  So my guess is
>that it's been broken since at least IDLE 0.4.
>
>Luckily, you have the source code <wink>.
>
>not-something-i'm-likely-to-fix-anyway-ly y'rs  - tim
>
>
>
>





From Darrell at p98.f112.n480.z2.fidonet.org  Fri Jul  2 06:03:30 1999
From: Darrell at p98.f112.n480.z2.fidonet.org (Darrell)
Date: Fri, 02 Jul 1999 11:03:30 +0100
Subject: Tuples & lists
Message-ID: <00001686@bossar.com.pl>

From: "Darrell" <news at dorb.com>

Did an experiment on NT and lists and tuples took the same amount of space.
Humm ? Not what I expected.

--
--Darrell
Terry Reedy <tjreedy at udel.edu> wrote in message
news:7lh375$5ea$1 at news.udel.edu...
> In article <377bbde1 at news.xtml.co.uk>, nick at NOSPAMvideosystem.co.uk
says...
>
> >Why both having tuples and lists? As far as I can see there're the same,
but
> >tuples are read only..? Have I missed something in the tutorial?
>
> There is probably something in the the FAQ.  BAsically, tuples are stored
> more compactly and can be used as dictionary keys (because they ARE
> readonly).
>
> TJR
>




From wolford at enews.nrl.navy.mil  Fri Jul  9 14:11:28 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Fri, 09 Jul 1999 14:11:28 -0400
Subject: win32 python as a dll?
References: <1280583297-81012085@hypernet.com>
Message-ID: <37863B4F.FB1780F3@enews.nrl.navy.mil>

Gordon McMillan wrote:

>  > How about:
> > python.lib/dll and msvcrt.lib/dll
> > python.lib/dll and msvcrtd.lib/dll
> > python_d.lib/dll and msvcrt.lib/dll
> > python_d.lib/dll and msvcrtd.lib/dll
>
> Debug python linked to non-debug msvcrt makes sense - means you can
> load non-debug extension modules (see below).

Exactly my point; hence the four combos (or two if you don't want to give out
python15_d.dll).

> > > It is also worth noting that Gordon pointed this out to you before this
> > > particular thread, and you still had the problem.  Im not sure that
> > > documenting it would be effective in all cases.
> >
> > He said to use the Multithreaded DLL. Unfortunately, msvcrt.dll and
> > msvcrtd.dll are *both* the Multithreaded DLL; one is debug the other
> > is not. It never occured to me (well, it finally did) that it
> > mattered whether msvcrt.dll or msvcrtd.dll were used. I assumed that
> > the semantics of the two DLLs were essentially the same, just the
> > implementation differed slightly. I'm pretty sure (though I haven't
> > tested it, maybe you'll know) that you can mix and match optimized
> > and debug code pretty easily under Unix.
>
> To beat a dead horse into an unrecognizable pulp:
>
> That's the whole point - they are *different* dlls. So a file open in
> one will not be found in the other. Kaboom. Nothing to do with
> semantics - it's just where buffers, files and other statics /
> globals live.

I guess I didn't pose my question properly. Under Unix, can you not link with a
debug foo.so at link-time, then at run-time load an optimized version, as long
as the code used to generate the two is roughly the same? It seems to work for
me. Why can't you do the same thing under Windows?

> You can mix debug and optimized in Windows, too - you'll just end up
> stepping thru unintelligable assembler when you hit non-debug code.

That's not what I meant and I know that.

> [And for the masochistically curious - the reason the runtime lib is
> specified to the MSVC compiler and not the linker is that the
> compiler generates slightly different code depending on whether the c
> runtime is in a static lib or in a dll.]

I still don't see why you have to jump through these hoops just to get simple
stuff working on Windows.




From anamax at earthlink.net  Thu Jul  1 12:19:07 1999
From: anamax at earthlink.net (Andy Freeman)
Date: Thu, 01 Jul 1999 16:19:07 GMT
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <lwoghyexy5.fsf@copernico.parades.rm.cnr.it> <377B454C.6F59@mailserver.hursley.ibm.com>
Message-ID: <7lg4dd$hs5$1@nnrp1.deja.com>

In article <377B454C.6F59 at mailserver.hursley.ibm.com>,
  Paul Duffin <pduffin at mailserver.hursley.ibm.com> wrote:

> > Dynamic binding is another beast altogether, which Common Lisp and
> > Scheme (statically scoped languages) allow in through a back door.
>
> If they are statically scoped how does the following work.

All of the following book references are to Steele's Common
Lisp the Language, 2nd edition.

See Chapter 3 (esp pg 45, which explicitly states that the
default scope is static), Chapter 1.2.3 (discussion of expansion),
and Chapter 8 (esp 8.1).  Chapter 3 has a nice description
of what scope, extent, static, and dynamic mean.

> Look at the following code.
>
> 	(let ((a 2))
> 	  (let ((b 3))
> 	    (+ a b)))
>
> When you expand the macro let you get something like.
>
> 	((lambda (a)
> 	   ((lambda (b)
> 	      (+ a b)) 3)) 2)

So far, so good.  However, it doesn't really go further
because lambda and application are primitives.

> Which is equivalent (apart from side effects) of.
>
> 	(defun inner (b)
> 	  (+ a b))
>
> 	(defun outer (a)
> 	  (inner 3))
>
> 	(outer 2)

No, it isn't.  That's not what lambda means, and it isn't how
macros work, and so on.  Lambda "makes" an UNNAMED function IN
a given lexical context - the above is an invalid transformation.

One can use lambda to create CL variables with dynamic scope (that
is indefinite scope and dynamic extent), but nested lambdas don't
work as above.

It isn't clear that nested lambdas EVER worked that way, even in
dynamically scoped lisps.

> So what is the process which associates the use of "a" inside inner
> with the formal argument of outer.

mu.  Since there is no "inner" or "outer", the question makes
no sense.

> Lisp macros manipulate S-expressions but the macro definitions
themselves
> are interpreted differently to Lisp expressions.

Macro definitions just define source to source transformations
that happen "before" before evaluation.

> A simple macro version of setq would convert from
> 	(setq symbol '(list))
> to
> 	(set 'symbol '(list))

Setq can not be defined in terms of set because set can't work for
statically scoped vars, which is the default in CL.

> Lisp does not really need macros, they are really just a way of
> creating short cuts.

I don't need * or even + either (if I have +1 and -1), but .....
(In fact, I can even get by with an "if" which always evaluates
all of its arguments.)

> I would say that Tcl (without macro system) can do anything that Lisp
> (with macro system) can do.

Since both are Turing machines....

-andy


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From markusk at bidra241.bbn.hp.com  Wed Jul 28 05:32:48 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 28 Jul 1999 11:32:48 +0200
Subject: make or similar in Python
References: <378390DD.58CA6BD5@nevex.com> <7nlj9o$mj7$1@ssauraac-i-1.production.compuserve.com>
Message-ID: <p5emhtm85b.fsf@bidra241.bbn.hp.com>

"T. C. Mits" <71351.356 at compuserve.com> writes:

> Howdy:
>     Funny, I'm searching for the exact same thing.
> 
>     My idea is to use XML as a repository for the dependencies and other
> required configuration properties.  A Python script will query the db and
> generate dynamic makefiles.   A pain is that this has to work on NT and Unix
> and perhaps even by web based via Servelets.
> 
> 
> Greg Wilson <gvwilson at nevex.com> wrote in message
> news:378390DD.58CA6BD5 at nevex.com...
> > Hi, folks.  Has anyone ever written a dependency management/update tool
> > (like 'make') in Python?
> > I'm not looking for something that reads makefiles, and happens to be
> > written in Python; I'm looking
> > for a Python module that users import, then provide data to, in order to
> > manage dependencies.  (The
> > answer I got from the Perl crowd was 'cool idea, but no'; I'm hoping
> > I'll have more luck here.)
> > Thanks,
> > Greg (who'd really, really like to override some of the things that
> > GNU make does)
> >
> >

Ask sam at mitre.org for PyMake. 


Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From greg.ewing at compaq.com  Thu Jul  1 17:14:27 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 02 Jul 1999 09:14:27 +1200
Subject: total idiot question: +=, .=, etc...
References: <000001bebffc$12060200$da9e2299@tim> <slrn7ngj6b.g9.behrends@allegro.cse.msu.edu> <m3so7buvr6.fsf@greebo.nodomain.de>
Message-ID: <377BDA33.144EB96D@compaq.com>

Bernhard Herzog wrote:
> 
>         .last = .lines[-1]
>         .lines = .lines[:-1]

Hey, I've just had an idea: allow an *empty*
name as an identifier. Then we could write:

  class C:
    def foo(,x,y):
      .z = .z + x * y

and kill two birds with one syntax change!

Greg



From philh at vision25.demon.co.uk  Tue Jul 13 17:05:28 1999
From: philh at vision25.demon.co.uk (Phil Hunt)
Date: Tue, 13 Jul 99 21:05:28 GMT
Subject: problem with SocketServer -- address already in use
Message-ID: <931899928snz@vision25.demon.co.uk>

I'm trying to use SocketServer.py to write a TCP server. I ran
my program with ``python serv.py'', which went into an infinite
loop (which is what it should do). I exited my program with ^Z,
and tried to start it up again. I got an error message:

socket.error: (98, 'Address already in use')

which I presume is because something thinks the old invokation of
my program is still using port 7070. 

How do I clear the port, before I start up my server?

-- 
Phil Hunt....philh at vision25.demon.co.uk




From r.skowronek at fz-juelich.de  Thu Jul  1 02:43:45 1999
From: r.skowronek at fz-juelich.de (r.skowronek at fz-juelich.de)
Date: 1 Jul 1999 06:43:45 GMT
Subject: Decent Tutorial
References: <377925EF.1CBB640F@starvision.com>  <011a01bec2d3$c215a900$f29b12c2@secret.pythonware.com> <3779e9a5@discussions>
Message-ID: <7lf2n1$s5r@zam201.zam.kfa-juelich.de>

Newbie <8499 at greenhead.ac.uk> wrote:

: Where can I find a decent tutorial which is 
: for beginners to Python, preferably including 
: some documented source code and avaliable for 
: download.

Have a look at http://www.spindoczine.com/folder5

There is an (almost?) complete book which I as a newbee to python found very
helpful. It is in MS-Word 97 format though.

Rolf

-- 
r.skowronek at fz-juelich.de



From Jean-Claude.Wippler at p98.f112.n480.z2.fidonet.org  Fri Jul  2 15:50:47 1999
From: Jean-Claude.Wippler at p98.f112.n480.z2.fidonet.org (Jean-Claude Wippler)
Date: Fri, 02 Jul 1999 20:50:47 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001679@bossar.com.pl>

From: Jean-Claude Wippler <jcw at equi4.com>

Mikael Lyngvig wrote:
> 
> On 2 Jul 1999 12:04:54 GMT, lvirden at cas.org wrote:
> 
> >I wonder why various industry rags have proclaimed during the past
> >10 months about Apple's G3 machines outselling other vendor
> >machines.  ...
> 
> That makes two of us.

Sales != installed base.
Also: Sales of one vendor vs. combined sales of N.

-- Jean-Claude, proving that PB G3's are fine for posting (even off-topic)



From hank at prosysplus.com  Thu Jul  1 16:19:25 1999
From: hank at prosysplus.com (Hank Fay)
Date: Thu, 1 Jul 1999 16:19:25 -0400
Subject: wddx wrapper?
Message-ID: <%0Qe3.4034$lr3.164499@news2.jacksonville.net>

Does anyone know a reference to a Python wrapper for wddx (www.wddx.org)?
It looks like a neat way to pass data around (being able to transport
recordsets between differnet languages, in particular), and I'd like to be
able to have Python doing some of the heavy lifting on the server side.

thanks,

Hank Fay

--
http://www.prosysplus.com





From godzilla at netmeg.net  Fri Jul 23 14:53:28 1999
From: godzilla at netmeg.net (Les Schaffer)
Date: Fri, 23 Jul 1999 18:53:28 GMT
Subject: Python interface to IBM Data Explorer?? (DX/OpenDX)
References: <3798841F.B9ED6E95@brown.edu>
Message-ID: <m2u2qv9ozh.fsf@netmeg.net>

[follow-ups set to comp.lang.python]

>>>>> ">" == Daniel Faken <Daniel_Faken at brown.edu> writes:

    >>   Has anyone developed a python interface (via SWIG or
    >> something else) for the Data Explorer visualization system?

i have been following the user and development email lists for DX for
a month or so, and havent heard anything like this yet.

the way DX gets built is there are all these library built first and
then major interface components get linked with these libs.

whether that means that these static libraries have well-enough
defined interfaces that we could SWIG 'em in, thats another. 

i'd be interested in pursuing this further, if you wanna look at this
together. 

les

-- 
____        Les Schaffer              ___| --->> Engineering R&D <<---
Theoretical & Applied Mechanics          |  Designspring, Inc. 
Center for Radiophysics & Space Research |  http://www.designspring.com/
Cornell Univ.  schaffer at tam.cornell.edu  |  les at designspring.com



From fredrik at pythonware.com  Wed Jul 14 14:45:16 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 14 Jul 1999 18:45:16 GMT
Subject: Redirect stdout to file
References: <816010E2456BD111A48700805FBBE2EEFDF47C@ex-quebec-u1.baan.com>
Message-ID: <000701bece29$04821ed0$f29b12c2@secret.pythonware.com>

Gaetan Corneau <Gaetan_Corneau at baan.com> wrote:
> I tried it and tought it didn't work, but it seems it was my fault. I did it
> this way:
> 
> from sys import stdout
> stdout = open("out.txt", "w")

to make things a bit clearer, let's rewrite
this slightly.  the above is roughly equivalent
to:

import sys
stdout = sys.stdout
del sys
stdout = open("out.txt", "w")

now, keep in mind that all python variables
are just named pointers.  so:

stdout = sys.stdout

creates a variable called "stdout" in *your*
module, which points to the same thing as
"sys.stdout" currently points to...

stdout = open("out.txt", "w")

replaces the existing variable called "std-
out" with a new one, which points to the
object returned from "open".

"stdout" in your  module, that is.  the "sys"
module isn't affected, and "print" still uses
whatever "sys.stdout" happens to point to...

in contrast,

sys.stdout = open("out.txt", "w")

modifies the "stdout" variable in the sys-
module, which is exactly what you wanted.

or in other words,

    "You Cannot Mutate Simple Variables
    By Assigning To Them"

    (the only way to modify something in place
    is to use a method.  this includes sugared
    methods like __setattr__, __setitem__ etc,
    of course...)

also see:
http://www.pythonware.com/people/fredrik/fyi/fyi06.htm
(seems like you violated the "You Know
Exactly What You're Doing" rule here ;-)

</F>




From Robert.Kern at p98.f112.n480.z2.fidonet.org  Fri Jul  2 18:11:38 1999
From: Robert.Kern at p98.f112.n480.z2.fidonet.org (Robert Kern)
Date: Fri, 02 Jul 1999 23:11:38 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001694@bossar.com.pl>

From: kernr at mail.ncifcrf.gov (Robert Kern)

On Thu, 1 Jul 1999 23:16:00 -0500, Travis Oliphant <olipt at mayo.edu>
wrote:

>
>This has been a very interesting discussion on GUI's.
>
>It struck me that all this talk of a cross-platform GUI might be
>preventing a lot of very good GUI's from being used.
>
>So, I looked in to PyGtk and PyGnome (with its Gnome Canvas) and I'm
>impressed.  Right now it fits what I want to do very well and it works on
>systems with an X server (aren't X servers available for MAC's and
>Windows?)  

Not for free.  Well, Microimage's X server, MI/X, is freely avialable
for Windows and Mac, but it's only approximately X11R5 compliant
(http://www.microimages.com/freestuf/mix/mix.htm).  PyGTK and PyGnome
are unlikely to work with it satisfactorily.  Heck, tkinter doesn't
work satisfactorily with it.

Starnet does have an X11R6 server for Windows for which they give away
a free demo that has a 2 hour time limit per run, although it can be
restarted again.  (http://www.starnet.com/demo.htm)

There is a project in progress to port XFree86 to Windows with the
Cygwin tools.  Everything compiles now; we just need to wait a little
before it runs.  :-)

GTK+ for Windows (http://www.gimp.org/~tml/gimp/win32/) still hasn't
had enough of the libraries ported to compile PyGTK.  This could
change soon, though.

>So, I'm happy with it. If you're in the same situation you
>might like it too.  I tend to agree that there is plenty of room for more
>than one GUI.
>
>Travis Oliphant

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From arcege at shore.net  Thu Jul 15 17:24:48 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 15 Jul 1999 21:24:48 GMT
Subject: Calling superclass method with variable args?
References: <378E29DD.1613913E@pixar.com>
Message-ID: <Aksj3.132$ui4.39372@news.shore.net>

Steve Johnson <slj at pixar.com> wrote:
: Hi,

: I'm a long-time C++ programmer, but I'm really new to Python, so please
: forgive me if this is a dumb question with a simple or obvious answer.

: A fundamental idea in OOP is the ability to have a subclass augment the
: behavior of a superclass's method by redefine that method, and then
: calling the superclass's version of the method before doing anything
: else.  The key to this idea, and to polymorphism in general, is that the
: two methods have the same argument signature so that the superclass's
: method gets passed the same set of arguments from the caller whether the
: subclass overrides the method or not.

: If these methods have a static argument list, this is an easy thing to
: do in Python. Here's an example:

: class Task:
:     def __init__(self,a,b,c):
:         <whatever>

: class Job(Task):
:     def __init__(self,a,b,c):
:         Task.__init__(self,a,b,c)
:         <whatever>

: aJob = Job(x,y,z);

: But what if the method takes unnamed positional and keyword arguments.
: Then, how do you pass these along to the superclass's version of the
: method such that it gets the arguments in the same form that the
: subclass's method did?  Here's the code in question:

: class Task:
:     def __init__(self,a,b,c,*moreargs,**morekeyargs):
:         <whatever>

: class Job(Task):
:     def __init__(self,a,b,c, *moreargs, **morekeyargs):
:         Task.__init__(self,a,b,c,?????)
:         <whatever>

: aJob = Job(x,y,z,m1,m2,m3,m4,m5);

: My question really is...what do I replace the ????? (or that whole line)
: with to get this fundamental OOP behavior?

: This doesn't work:

:         Task.__init__(self,a,b,c,moreargs,morekeyargs)

: because now Task.__init__ will see exactly 5 arguments, instead of some
: arbitrary number.  This almost works:

:         apply(Task.__init__,(self,a,b,c) + moreargs)

: but the extra keyword arguments don't get passed along. So, I took the
: idea one step further and tried to introduce the keyword arguments back
: into the parameter list that gets passed to apply:

: def reconstruct(attribs):
:     r = ()
:     for key in attribs.keys():
:         r = r + (key + "=" + attribs[key]",)
:         print r
:     return r

: class Task:
:     def __init__(self,a,b,c,*moreargs,**morekeyargs):
:         <whatever>

: class Job(Task):
:     def __init__(self,a,b,c, *moreargs, **morekeyargs):
:         apply(Task.__init__, (self,a,b,c) + moreargs +
: reconstruct(morekeyargs))
:         <whatever>

: aJob = Job(x,y,z,m1,m2,m3,m4,m5);

: but Task.__init__() doesn't recognize the keyword parameters as such,
: instead it thinks they are string arguments of the form "foo=bar".

: Can anyone tell me how to get the classic OOP behavior I'm looking for
: here?

You need to look at apply more, it takes up to three arguments:
  apply(function, args[, keywords])

So you could have:
  def __init__(self, a, b, c, *moreargs, **morekeyargs):
    apply(Task.__init__, (self, a, b, c) + moreargs, morekeyargs)

The URL for the documentation of apply() is:
 http://www.python.org/doc/current/lib/built-in-funcs.html#l2h-100

  -Arcege




From markus_kohler at hp.com  Wed Jul 28 03:24:29 1999
From: markus_kohler at hp.com (Markus Kohler)
Date: Wed, 28 Jul 1999 09:24:29 +0200
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk>
		<7n5f0c$snu$1@ffx2nh5.news.uu.net>
		<al81zdxkktf.fsf@sirppi.helsinki.fi>
		<p51zdvn4c7.fsf@bidra241.bbn.hp.com>
		<14236.32616.980578.536614@dolphin.mojam.com>
		<379D89BC.DA91BA8C@hp.com>
		<14237.60407.130426.461846@dolphin.mojam.com> <14238.335.458728.140169@lrz.de>
Message-ID: <379EB02D.3890926@hp.com>

Eugene Leitl wrote:
> 
> Skip Montanaro writes:
> 
>  > True.  I was thinking about the situation I'm in (small developer, a few
>  > machines to maintain/upgrade/replace).
> 
> What's wrong with recoding the hot spots in C? Python is primarily a
> rapid prototyping/scripting/glue language, it was never meant as a
> sole implementation language for applications.

It's not wrong, but it takes more time. 
The less I need to use C the better should be my producitvity. 

Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From mgushee at havenrock.com  Tue Jul 20 23:56:00 1999
From: mgushee at havenrock.com (Matt Gushee)
Date: 21 Jul 1999 12:56:00 +0900
Subject: binding keys to slider in Tkinter?
References: <7mvjio$dh8@thalamus.wustl.edu> <KAJIYAMA.99Jul21061013@pseudo.grad.sccs.chukyo-u.ac.jp>
Message-ID: <m3vhbe4pwv.fsf@havenrock.com>

kajiyama at grad.sccs.chukyo-u.ac.jp (Tamito Kajiyama) writes:

> heather at thalamus.wustl.edu (Heather A. Drury) writes:
> | 
> | I'm trying to bind the arrow keys to allow a user to increment
> | (or decrement) a counter using the Scale widget and Tkinter.
> | I seem to be unable to bind anything to the Scale widget.
> | I'm sure I'm missing something.
> 
> For a widget to get key press/release events, the widget first needs to
> get the input focus.  You can pass around the input focus by pressing
> the tab key.  The Scale widget has the key bindings that you want
> already.
> 
> If you want to control the Scale widget using arrow keys even when
> another widget has the input focus, you have to let the widget to pass
> the relevant events to the Scale widget.

Haven't tried this in a while, but IIRC if you create a binding for
the arrow keys on the toplevel window, that binding is accessible from 
all child windows of that toplevel.

Matt Gushee
Portland, Maine, USA
mgushee at havenrock.com



From dwolff567 at my-deja.com  Thu Jul 29 16:33:18 1999
From: dwolff567 at my-deja.com (dwolff567 at my-deja.com)
Date: Thu, 29 Jul 1999 20:33:18 GMT
Subject: Animating gifs in Tkinter
Message-ID: <7nqdqd$5ci$1@nnrp1.deja.com>


Hello,

Has anyone had success with animating gif files using tkinter?
I have searched around and haven't found any example code for this.
My program runs a simulation in C which outputs a gif file every
once in a while.  The C program then calls a python method which
updates the gif file in a canvas widget.  I have had lots of trouble
with this for some strange reason.  The program will freeze at random
times, but only when
the picture is being updated.  Here is the code which updates the
image:

 def update(self,time):
        if(self.image == self.image1):
            self.image2 = PhotoImage(file="plot.gif")
            self.image = self.image2
        else:
            self.image1 = PhotoImage(file="plot.gif")
            self.image = self.image1
        self.simCanvas.delete(self.curImage)
        self.curImage =  self.simCanvas.create_image(0,0,anchor=NW,
                         image=self.image)

        self.simCanvas.update_idletasks()

I had to use two seperate PhotoImage instances or it wouldn't work at
all.

If anyone can be of any help, I would greatly appreciate it.  I would
also appreciate any example code you could send me.

Thanks,

  Dave


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From alex at somewhere.round.here  Tue Jul  6 19:17:39 1999
From: alex at somewhere.round.here (Alex)
Date: 06 Jul 1999 19:17:39 -0400
Subject: optimization
References: <XFMail.990705181259.nathan@islanddata.com>
Message-ID: <etdbtdp74h8.fsf@m2-225-12.mit.edu>

How does this look?
Alex.

import re, string

class dictionary_parser:
    def __init__ (self, file):
        self.file = file
        self.dict = {}
        self.parse_file ()
    def parse_file (self):
        lines = self.file.readlines ()
        multiline_delimiters = filter (re.compile ('\.').match, lines)
        for delimiter_index in range (0,len (multiline_delimiters),2):
            delimiter = multiline_delimiters [delimiter_index]
            assert delimiter==multiline_delimiters[delimiter_index+1]
            key = delimiter[1:]
            assert not self.dict.has_key (key)
            start = lines.index (delimiter)
            prefix = lines [:start]
            self.parse_single_line_entries (prefix)
            lines = lines [start+1:]
            end = lines.index (delimiter)
            # Check there's no single-line key in the multi-line value
            for line in lines [:end]: assert not re.search (':', line)
            self.dict [key] = string.join (lines[:end-1], '')
        self.parse_single_line_entries (lines [end+1:])
    def parse_single_line_entries (self, lines):
        single_line_pairs = filter (re.compile (':').search, lines)
        for pair in single_line_pairs:
            (key, value) = string.split (pair, ':', 1)
            assert not self.dict.has_key (key)
            self.dict[key] = value

if __name__ == '__main__':
    class file:
        contents = ['Key: one line value\n',
                    '.KEY\n',
                    'multiline value\n',
                    'multiline value\n',
                    'multiline value\n',
                    'multiline value\n',
                    '.KEY\n',
                    'Other Key: one line value\n']
        def readlines (self):
            return file.contents
    print dictionary_parser (file ()).dict



From larsga at ifi.uio.no  Thu Jul 22 07:05:19 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 22 Jul 1999 13:05:19 +0200
Subject: Fw: ANNOUNCE: Mailing list for Australian Python Users
References: <007201bed42e$c4316c30$f29b12c2@secret.pythonware.com>
Message-ID: <wkogh56j2o.fsf@ifi.uio.no>

* Fredrik Lundh
| 
| python-se, n?gon?  Link?ping?  :-)

Tjah. Si Oslo, s? kan vi bli en del. :-)

--Lars M.



From sachs at panix.com  Wed Jul 21 11:44:46 1999
From: sachs at panix.com (Josef Sachs)
Date: 21 Jul 1999 11:44:46 -0400
Subject: Python Mode for Alpha ?
References: <lawter-2107990058210001@client-151-197-119-26.bellatlantic.net>
	<14229.57251.996871.502133@anthem.cnri.reston.va.us>
	<14229.58766.234709.815774@weyr.cnri.reston.va.us>
Message-ID: <o9vhbevwgh.fsf@panix3.panix.com>

>>>>> On Wed, 21 Jul 1999 15:21:50 GMT, Fred L Drake said:

> No, he's asking about an editing mode for Alpha, a Mac editor.  I
> remember having one when I used a Macintosh, but that was almost 5
> years ago.  I have no idea where to get such a beast these days.

<URL: ftp://ftp.python.org/pub/python/mac/old/AlphaPythonMode.sit.hqx>?
Did that stop working with some recent Alpha release?



From arcege at shore.net  Mon Jul 12 12:34:30 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Mon, 12 Jul 1999 16:34:30 GMT
Subject: Printing returned object in embedded app
References: <378694FC.D857F68A@srl.caltech.edu> <6wxh3.1472$6M6.450919@news.shore.net> <slrn7oep57.a57.conrad@vanir.srl.caltech.edu> <7m9tnq$6qg$69@rks1.urz.tu-dresden.de>
Message-ID: <qOoi3.1613$6M6.533266@news.shore.net>

Conrad Steenberg <conrad at vanir.srl.caltech.edu> wrote:
: On 10 Jul 1999 18:24:33 GMT, Conrad Steenberg <conrad at vanir.srl.caltech.edu>
: wrote:
:>On Sat, 10 Jul 1999 01:40:18 GMT, Michael P. Reilly <arcege at shore.net> wrote:
:>>
:>>Try:
:>>  PyObject_Print(object, stdout, Py_PRINT_RAW);
:>
:>Thanks! 
:>
:>Conrad
:>-- 
: Hmmm... Sorry to reply to my own message, but it seems like PyRun_String 
: returns only the Py_None object! 

: How can I get the result of a calculation, say
: obj=PyRun_String ("1+1", start, globals, locals);
: which should be a numeric object? 

: The behaviour I'm trying to get is the same as running python on a terminal:
: when an instruction have a printable result, I want to print it.

You should be able to wrap the code you want in something like:

  char buffer[256];
  char expression[] = "1+1";
  sprintf(buffer, "result = (%s); print result", expression);
  PyRun_String(buffer, start, globals, locals);

This would only work for expressions, but then.. you can't print the
results of statements anyways. ;)

  -Arcege




From pduffin at mailserver.hursley.ibm.com  Thu Jul  1 04:37:43 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Thu, 01 Jul 1999 09:37:43 +0100
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <lwoghyexy5.fsf@copernico.parades.rm.cnr.it>
Message-ID: <377B28D7.773C@mailserver.hursley.ibm.com>

Lets leave these good Python people alone and move to comp.lang.lisp.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From fredrik at pythonware.com  Mon Jul 12 04:27:58 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 12 Jul 1999 08:27:58 GMT
Subject: python, tkInter -> tcl/tk problems
References: <7m9tid$2sl$1@rzsun02.rrz.uni-hamburg.de>
Message-ID: <007401becc40$7627ddd0$f29b12c2@secret.pythonware.com>

martin.lemburg at bigfoot.de wrote:

>TclError: Can't find a usable init.tcl in the following directories:

>E:/programming/Python/tk8.0/library/tk.tcl: bad event type or keysym "MouseWheel"

http://www.pythonware.com/people/fredrik/fyi/fyi02.htm

</F>




From gmcm at hypernet.com  Mon Jul 19 09:23:37 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 19 Jul 1999 08:23:37 -0500
Subject: [Q] Type checking...
In-Reply-To: <7mv000$sun$1@feed.teaser.fr>
Message-ID: <1279735660-4809931@hypernet.com>

Olivier Deckmyn asks:
> 
> Abstract : [Q]How to check the type of an instance before(after?)
> sending it to a method...

The builtin function "isinstance".

If class A has method frobnicate(), your code will be much more 
general if you do
 class B:
   def DoSomething(self, thing):
     try:
       thing.frobnicate()
     except:
       ...

This lets your users implement the A interface without having to 
actually be an A.


- Gordon



From hyoon at bigfoot.com  Fri Jul 16 10:07:22 1999
From: hyoon at bigfoot.com (Hoon Yoon)
Date: Fri, 16 Jul 1999 10:07:22 -0400
Subject: Case mismatch on NT
References: <378CCB4A.BDAD6FC7@bigfoot.com> <5loghdcyr2.fsf@eric.cnri.reston.va.us>
Message-ID: <378F3C9A.872455B0@bigfoot.com>

Guido,

  The problem here is that both Dialog and dialog comes with pretty
standard Python distribution for tk and mfc. I had similar prb with COM
and callDll some time ago.
  If I use the first solution, then I am kinda lost because I don't know
what those standard distribution call each other when and where.
  If I use the second way, then I cannot use either portion of tk (and
Dialog.py in Tk is pretty important) or mfc (I am not sure how dialog.py
is important, but..).
  My current thought is that I use sys.path with insert to control the
precedence during run time, but it's not an amicable situation.
  Best solution is I guess standard packages to avoid name confusion
like this. Or what I proposed earlier about searching through the whole
PYTHONPATH before rasing case mismatch err.
  Thanks for the reply,

Hoon,

Guido van Rossum wrote:
> 
> Hoon Yoon <hyoon at bigfoot.com> writes:
> 
> >  I realize that this has been discussed before. And you might think the
> > solution is  os.environ["PYTHONCASEOK"] = "hotdog" but try this one.
> >  There is dialog.py sitting in pythonwin/.../mfc directory and Dialog.py
> > in lib-tk directory. If I have both directory in my Pythonpath, I cannot
> > call the one sitting at the second directory with out encountering the
> > dreaded "Case Mismatch" error.
> >  I feel that "Case Mismatch" should only occur after all directories has
> > been searched to find the 'right' case import. Not after finding the
> > first non-matching one. Is there a way to set environ to do what I want?
> >  If not, can anyone tell me how to get around this prob or tell me if
> > what I am proposing makes sense or not?
> 
> This is intentional -- it prevents people from *relying* on the case
> matching to make their program work.  (The case sensitivity is there
> to make your program fail if you typed "import String" when you meant
> "import string".)
> 
> Two workarounds: (1) make a copy of the second one with a different
> name, and use that; (2) rearrange your sys.path so that the one you
> want to use comes first.  Neither might work if you have some library
> module that uses one and some other library module that uses the other...
> 
> --Guido van Rossum (home page: http://www.python.org/~guido/)

-- 
*****************************************************************************
S. Hoon Yoon (Quant) Merrill Lynch Equity Trading, 
yelled at yahoo.com hoon at bigfoot.com(w)
"Miracle is always only few standard deviations away, but so is
catastrophe."
* Expressed opinions are often my own, but NOT my employer's.
"I feel like a fugitive from the law of averages."    Mauldin
*****************************************************************************
-------------- next part --------------
A non-text attachment was scrubbed...
Name: vcard.vcf
Type: text/x-vcard
Size: 202 bytes
Desc: Card for Hoon Yoon
URL: <http://mail.python.org/pipermail/python-list/attachments/19990716/367f92af/attachment.vcf>

From MHammond at skippinet.com.au  Fri Jul  2 23:34:38 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Sat, 3 Jul 1999 13:34:38 +1000
Subject: Waiting for an external process
References: <377D9E05.E4F91D2F@globalserve.net>
Message-ID: <7lk08q$sgt$1@m2.c2.telstra-mm.net.au>

Juan M. Gonzalez wrote in message <377D9E05.E4F91D2F at globalserve.net>...
>Is there a better way to determine when an external process has finished
>than using
>win32process.CreateProcess / GetExitCodeProcess ?

Depends what you mean by "better".  Using win32process, it is possible to
have an event set when the process terminates, which can avoid polling.  Im
not sure if this meets your definition of "better" tho - certainly not
"simpler".

Mark.






From shoh at lgsoft.com  Fri Jul 30 10:26:08 1999
From: shoh at lgsoft.com (Sang Hun Oh)
Date: Fri, 30 Jul 99 14:26:08 GMT
Subject: Looking for Paos and Chautauqua
Message-ID: <mt2.0-24150-933344768@news.informatik.uni-bonn.de>

Hi!

I am looking for Paos (Python active object server) and Chautauqua
developed by Carlos Maltzahn and David Vollmar under the auspice of Skip
Ellis
at the University of Colorado at Boulder. They are written by Python.

I failed to download them from the 'ftp.cs.colorado.edu' site.
Please send them to me, if you have a copy.

Thanks in advance.

Sang Hun Oh
shoh at lgsoft.com



From Hans-Christian.Holm at nrk.no  Thu Jul  8 11:00:19 1999
From: Hans-Christian.Holm at nrk.no (Hans-Christian Holm)
Date: Thu, 8 Jul 1999 17:00:19 +0200
Subject: Assigning to ASP Session dictionary
Message-ID: <7m2ee6$5of$1@elle.eunet.no>

Has anyone found out how? I can access session variables by code like

a = Session('a')

but

Session('a') = 12

does not work! I've tried many other ways, like using Session.Contents etc.,
but I get nothing but errors.

Hans-Christian Holm
NRK interaktiv






From Gareth.McCaughan at pobox.com  Thu Jul  1 12:51:28 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 01 Jul 1999 17:51:28 +0100
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.i
	<lwoghyexy5.fsf@copernico.parades.rm.cnr.it>
	<7ldt5p$srh$1@dragon.flightlab.com>
	<lwhfnosvuz.fsf@copernico.parades.rm.cnr.it>
	<377B922A.4A7B@mailserver.hursley.ibm.com>
Message-ID: <86n1xgwbxr.fsf@g.pet.cam.ac.uk>

Paul Duffin wrote:

> I was referring to the fact that CL macros 'complicate' the parsing /
> evaluation rules of Lisp which otherwise would be very simple and
> consistent.
> 
> I personally like Lisp a lot and take advantage of lots of short cuts
> that macros provide but I think that there is a big learning curve to
> climb if you want to write one yourself (which is why I referred to it
> as another language). Tcl on the other hand allows me to use the same
> commands that I use to write normal programs to extend the language by
> adding new programming constructs.

So does Lisp. If you want (foobar x y) to be the same as (setq y x)
you say

    (defmacro foobar (x y)
      `(setq ,y ,x))

which is hardly terribly hairy. If you want (n-times n (wibble spong))
to be the same as n repetitions of (wibble spong) you say

    (defmacro n-times (n form)
      `(progn . ,(loop repeat n collect form)))

If you want (with-foo (var arg1 arg2) blah blah blah) to be the same
as (let ((var (foo arg1 arg2))) blah), you say

    (defmacro with-foo ((var . args) &body forms)
      `(let ((,var (foo . ,args))) . ,forms))

This is all the same Lisp language as you write functions in.
The backquote form is most commonly used when writing macros,
but it's useful for other things too. For instance, I hacked up
a symbolic differentiation program recently, and it's full of
things like `(+ (* ,(d f x) ,g) (* ,f ,(d g x))). Writing
macros is certainly a slightly different *skill* from writing
ordinary code, but it's not remotely a different *language*.

Incidentally, how would you implement equivalents of those
things in Tcl?

Of course a macro isn't the *same* as a function. But then,
if you want something the same as a function, you can use a
function :-). Perhaps what you want is a function that doesn't
evaluate its arguments. Some old versions of Lisp had things
called FEXPRs, which were exactly that. But, after experience
of using them, it turned out that they weren't the Right Thing,
and something much more like the present macro system was born.
Kent Pitman (who posts regularly here) wrote an article about
why FEXPRs aren't good enough, back around the time when the
transition to macros happened. It's available on the web
somewhere.

If you don't like having to construct code explicitly (though
I find the backquote syntax makes this a non-issue), you might
prefer the Scheme macro system. It's rather interesting.

-- 
Gareth McCaughan            Dept. of Pure Mathematics & Math. Statistics,
Gareth.McCaughan at pobox.com  Cambridge University, England.



From amk at mira.erols.com  Sun Jul  4 14:19:20 1999
From: amk at mira.erols.com (A.M. Kuchling)
Date: 4 Jul 1999 18:19:20 GMT
Subject: wddx wrapper?
References: <%0Qe3.4034$lr3.164499@news2.jacksonville.net>
Message-ID: <slrn7nv9qd.178.amk@207-172-57-156.s156.tnt2.ann.va.dialup.rcn.com>

On Thu, 1 Jul 1999 16:19:25 -0400, Hank Fay <hank at prosysplus.com> wrote:
>Does anyone know a reference to a Python wrapper for wddx (www.wddx.org)?

There's a WDDX module in the current XML-SIG's CVS tree (not in the
released code).  It's in xml.marshal.wddx, and the interface is much
like the marshal module:

>>> from xml.marshal import wddx
>>> print wddx.dumps( ('tuple', 1, 3.5, wddx.TRUE) )
<?xml version="1.0"?><!DOCTYPE wddxPacket SYSTEM "wddx_0090.dtd">
<wddxPacket version="0.9">
  <array length="4">
    <string>tuple</string>
    <number>1</number>
    <number>3.5</number>
    <boolean value="true"/>
  </array>
</wddxPacket>

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
If there's another world, he lives in bliss; / If there is none, he made the
best of this.
    -- Robert Burns, "Epitaph on William Muir"



From bmcd at es.co.nz  Sun Jul 11 00:46:49 1999
From: bmcd at es.co.nz (Ben Caradoc-Davies)
Date: 11 Jul 1999 04:46:49 GMT
Subject: Dreams.
References: <slrn7ofkrt.55r.sdossey@dolphin.openprojects.net>
Message-ID: <slrn7og8m8.17k.bmcd@ripley.sulaco.net>

Hmmm. I'm not sure what your background is, Scott, but you may not be aware of
some previous work --- the term you may be looking for is "continuation". The
use of continuations is the explicit aim of a number paradigms, particularly in
the functional language world and "artificial intelligence". Continuations can
be used to implement "backtracking", where, for example, a program explores
one branch of execution, reaches a dead end, and restores it's previous state
by falling back on a saved continuation.

A continuation can be described as a function representing the work yet to be
done.

Have you read "The Structure and Interpretation of Computer Programs" by
Abelson and Sussman? A real eye-opener, especially for someone with a physics
Fortran/C/Matlab background. Changed my world-view. I have heard this book
often recommended as one of the best. It is (used to be?) required reading for
MIT comp. sci. and e.e. undergraduates. The second edition came out recently.
The book uses the Scheme language
  http://www.swiss.ai.mit.edu/scheme-home.html

Check out Ken Dickey's intro to scheme
  ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/intro.txt
which has some examples of how to rewrite a function in a continuation passing
style.

Also look in on comp.lang.scheme, comp.lang.lisp and comp.lang.functional. Read
the functional programming faq.

One of the basic ideas behind functional (a.k.a. applicative) programming is
avoiding the use of assignment to get around the problems of saving state
which, as you have correctly deduced, causes a lot of problems. There has been
a large amount of work done on these problems since the seventies. Look for
resources on "closures," "continuations," "functional programming," and
"backtracking."

ObPython:
The funny thing is, coming from Scheme, I found that Guido had put the best of
Scheme into Python. Python's major advantages are clear syntax and semantics, a
well-defined foreign function interface, and (consequently) a large library and
numerous extensions.

"Alms for a poor *ex*-schemer!"
"Did you say *ex*-schemer?"
"Yes. One minute I was a schemer with a livelihood. Next, along comes Guido.
  *Cures* me! No so much as a by-your-leave!"

-- 
Ben Caradoc-Davies <bmcd at es.co.nz>



From Andrew.Dalke at p98.f112.n480.z2.fidonet.org  Thu Jul  1 01:06:50 1999
From: Andrew.Dalke at p98.f112.n480.z2.fidonet.org (Andrew Dalke)
Date: Thu, 01 Jul 1999 06:06:50 +0100
Subject: 1.5.1 on IRIX
Message-ID: <00001626@bossar.com.pl>

From: Andrew Dalke <dalke at bioreason.com>

rceci at my-deja.com said:
> The libs originally where built/installed  -> /usr/local/lib/python1.5
> re-installed as ${somedir}/lib/irix32, and set PYTHONPATH to
> ${somedir}/lib/irix32.

> Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]

> Any ideas how to "fix" or suppress them?


Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]

The default module search path uses <prefix>/python1.5.


Actually, it's a little more complicated than that. The default
module search path uses <prefix>/python1.5 so you'll need to
move
  ${somedir}/lib/irix32
to
  ${somedir}/lib/irix32/python1.5

then setenv PYTHONHOME ${somedir}/lib/irix32 .

						Andrew
						dalke at acm.org



From loril at mmrd.com  Thu Jul  8 16:09:28 1999
From: loril at mmrd.com (Lori Ludick)
Date: Thu, 08 Jul 1999 16:09:28 -0400
Subject: setting up shortcuts from python
Message-ID: <37850578.EEEB73FE@mmrd.com>

Hi all,

I need to set up shortcuts from inside of my python script.
I have looked through some of the documentation so far and
have not even found a direction to go in.

I assume I need to find the magic windows api that
will let me create a shortcut and set up all the nice shortcut
informaton it needs...but I haven't had any luck.

Anyone try this before?


Thanks,
Lori





From lvirden at cas.org  Fri Jul  2 08:26:13 1999
From: lvirden at cas.org (lvirden at cas.org)
Date: 2 Jul 1999 12:26:13 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <00f701bec212$f6a83050$f29b12c2@secret.pythonware.com>
Message-ID: <7lib55$jnj$1@srv38.cas.org>

According to Fredrik Lundh <fredrik at pythonware.com>:
:What native API?  As of 8.0, Tk still emulates all widgets
:except the scrollbar, and use a somewhat ineffective
:X emulation layer for graphics.  AFAIK, they haven't
:changed that in 8.1...

I suspect that Tk features such as these are the native 'api's to which
the original poster was referring.  The following are a few of
the entries from Tk 8.x changes file.


8/22/96 (new feature) There are four new commands for opening common
dialog boxes: tk_chooseColor, tk_getOpenFile, tk_getSaveFile and
tk_messageBox. Native dialog boxes are used wherever available. (IL)

11/7/96 (feature change) Under Windows, buttons and scrollbars now
have native look and feel.  This affects the default class bindings
and the way the some configuration options are interpreted.  Refer to
the widget manual pages for more details. (SS)
*** POTENTIAL INCOMPATIBILITY ***

12/17/96 (feature change) Native Windows labels do not get a focus-ring
border.  (CS)

7/30/97 (feature change) As a result of native menus, you can no
longer drag through a frame of menubuttons on Macintosh and Windows
and have the menus pop down. You can still click on individual
menubuttons and their menus will pop down.  Applications needing to
present a menubar should consider using the new "-menu" configuration
of the toplevel widget to set up menubar which behaves correctly on
Macintosh, Windows and X Windows. (SRP)

10/21/97 (bug fix) Tk did not reset the result after native dialog
modal loops on Window so background events could perturb the dialog
result. (SS)

-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.



From tim_one at email.msn.com  Thu Jul  8 23:24:06 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 8 Jul 1999 23:24:06 -0400
Subject: Regular expression (re) anomaly with group names and alternatives
In-Reply-To: <AF223B12F1FDD2119253006097C4FCEC0AD159@cats.pdri.xerox.com>
Message-ID: <000301bec9ba$808432c0$c49e2299@tim>

[Bob Alexander]
> It would be nice (IMO) if group names could be duplicated in different
> alternatives (sections separated by the | operator) of a regular
> expression, with the name taking on the value in the alternative that
> matches (if any).  However, that doesn't work -- is that a bug or by
> design?

I believe it's a bug, but for a different reason <wink>:  the re compiler
should raise an exception when a group name is repeated.  Group names are
simply symbolic names for numbered groups:

>>> p = re.compile("(?P<a>a1)(?P<b>b)(?P<a>a3)")
>>> p.groupindex
{'b': 2, 'a': 3}
>>>

"a" started life as a symbolic name for group 1, but the re compiler
silently changed it to a symbolic name for group 3 (forgetting about group
1) when it saw the 2nd occurrence of the name.  Since this is surprising (&,
I believe, also an accident), it should complain.

Note that, in general, and as illustrated by the specific pattern above,
there's no general sense to be made of a repeated name.  If that pattern
matched, would you want group "a" to refer to string "a1" because that
matched "first"?  To "a3" because that matched "last"?  To their catenation?
To a list of matching substrings?

When there's not a *clear* meaning for something, Python generally declines
to make one up.

not-that-there's-a-clear-meaning-for-any-regexp<wink>-ly y'rs  - tim





From greg.ewing at compaq.com  Tue Jul  6 00:34:58 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 06 Jul 1999 16:34:58 +1200
Subject: Getting list of data sources from mxODBC?
Message-ID: <37818772.1C27503E@compaq.com>

Is there a way to get a list of available
data sources from mxODBC? I can't find any
mention of this in the supplied docs.

Thanks for any help,
Greg Ewing
greg.ewing at compaq.com



From JamesL at Lugoj.Com  Thu Jul 15 23:49:48 1999
From: JamesL at Lugoj.Com (James Logajan)
Date: Thu, 15 Jul 1999 20:49:48 -0700
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net> <19990715175320.B20155@toast.internal>
Message-ID: <378EABDC.EE9A3A46@Lugoj.Com>

jam wrote:
> 
> On Thu, Jul 15, 1999 at 02:20:33PM -0500, mary jeanette wrote:
> >
[Plea for help from Mary elided.]
> your post appears to be begging for an argument, so I shall refuse to give
> you one.
[Jam's response mostly elided.]

> as a rule of thumb: if your post can't pass the "you asshole" test, don't
> bother hitting the send button lest you humiliate yourself in front of
> a huge audience-- here's a hint: the post you sent out *did not*, in my
> opinion, pass the 'you asshole' test. the python community (as with others
> on the internet today) is driven (mostly) by mutual respect for one another
> and *not* greed. you simply need to ask for help instead of demanding it,
> and you will be given the answers you seek.

I never knew there was a "you asshole" test. Probably only an asshole would
invent such a test. Any person with an ounce of empathy would recognize that
the post by Mary Jeanette was obviously born of frustration, so a civil
request for more details about the problem encountered would have sufficed.
I can't figure out what got into "jam" to post such a vicious and uncalled
for attack.

P.S. If Mary Jeannette can give a few more specific details, post them or
even e-mail them direct to me and I or others will try to help.



From bobalex at ix.netcom.com  Sat Jul 17 11:52:04 1999
From: bobalex at ix.netcom.com (Bob Alexander)
Date: Sat, 17 Jul 1999 08:52:04 -0700
Subject: I envy newbies
References: <000001bed00d$a9a70480$d2a22299@tim> <3790884A.A58864B@callware.com>
Message-ID: <011601bed06c$53893960$2cee6dd1@BobAlexDesk>

Even the Windows 98 installer modifies the PATH. But I suppose one could
argue that it's the operating system, and the operating system owns the
PATH, so it can do whatever it darn well pleases!

However, although well-intentioned, it botched up my AUTOEXEC.BAT. I had to
figure out what it really *meant* to do and fix it by hand.

The problem with automatic modification of AUTOEXEC is that it's a script,
and if you've done anything more than having a simple, single PATH
statement, the odds of successful modification diminish.

It's the custom for Windows installers to be intrusive: new unwanted icons
on the desktop, new little icons in that little area by the Start button,
and on and on...

-- Bob




From " HPeter.Friedrichs" at alliedsignal.com  Thu Jul  1 12:30:31 1999
From: " HPeter.Friedrichs" at alliedsignal.com (H. P. Friedrichs)
Date: Thu, 01 Jul 1999 09:30:31 -0700
Subject: Help: Can you explain this error? (contains code snip)
References: <377AAD25.96947EBA@alliedsignal.com> <slrn7nlf4u.5ng.jepler@craie.inetnebr.com>
Message-ID: <377B97A7.473BBB36@alliedsignal.com>

Thanks!

Pete

Jeff Epler wrote:

> On Wed, 30 Jun 1999 16:49:58 -0700, H. P. Friedrichs
>  <@alliedsignal.com> wrote:
> >Hello:
> >
> >This is week 2 for me in python...What fun!
> >
> >I have stumbled across an interesting problem. When the code below is
> >run under Win95 (python 1.5.2),it works without problems. If the same
> >code is run under Linux (python 1.4) I get the error:
>
> Make things simpler:
>
> >>> int("123")
>
> this works in Python 1.5.x and fails in 1.4.  I believe this feature is
> present in all 1.5.x and no standard 1.4 versions.
>
> Use
>
> >>> string.atoi("123")
> to work on both versions.
>
> Jeff

--

-----------------------------------------------
| The opinions expressed here are my own,     |
| and not necessarily those of my employer.   |
-----------------------------------------------





From gmcm at hypernet.com  Thu Jul 22 23:45:58 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 22 Jul 1999 22:45:58 -0500
Subject: Dreaded Newbie Question
In-Reply-To: <K4tK+KAA26l3EwvK@epstein.co.uk>
Message-ID: <1279424740-23511939@hypernet.com>

Dani Epstein wrote:
> 
> BTW, I'm running PythonWin 1.5.1

Everybody seems to have missed this. To find a lot of the os 
stuff under 1.5.1, you had to look in the docs under the the 
particular implementation for your platform: ntpath, posixpath etc.

This has been greatly improved in the recent docs.

- Gordon



From robin at jessikat.demon.co.uk  Sun Jul  4 09:34:25 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sun, 4 Jul 1999 14:34:25 +0100
Subject: ordered threading
In-Reply-To: <1281033720-53905059@hypernet.com>
References: <Y2tnuCAVprf3EwNj@jessikat.demon.co.uk>
 <1281033720-53905059@hypernet.com>
Message-ID: <bSKHzCAhL2f3EweJ@jessikat.demon.co.uk>

In message <1281033720-53905059 at hypernet.com>, Gordon McMillan
<gmcm at hypernet.com> writes
>Robin Becker writes:
>
>> In article <000a01bec583$30332b40$769e2299 at tim>, Tim Peters
>> <tim_one at email.msn.com> writes
>
>> >[Robin Becker]
>> >> I have an unbounded number of worker threads which occasionally require
>> >> access to a unique server thread. I can easily mutex access to the
>> >> server thread, but is there any obvious/easy way to ensure that the
>> >> blocked workers wake up in the right order?
>> >
>> >No threading system on earth allows specifying the order in which blocked
>> >threads will wake up, although some systems (not Python's) allow you to
>> >*influence* the order via assorted "priority" gimmicks.
>
>> the workers are HTTP requests in separate threads from Zope I guess.
>> I want to serialise access to the time/space hog. I guess it could
>> run uniquely in the worker threads ie the locked resource would be
>> access to the hog code. I'm not sure that would let me guarantee
>> unlocking as there are ways for the worker threads to suicide.
>
>If that's what you want, you don't really need to order the 
>unblocking. Just give each thread an input queue protected by a 
>Condition. Worker threads put requests on the Hog thread's input 
>queue (including a ref to the return queue). If you find that some 
>worker threads seem to get better access to the Hog's queue than 
>others, a time.sleep(0.001) will probably fix it - since it will 
>pretty much force a context switch.
>
>
>- Gordon
thanks; I knew already that this could work. I didn't know how to force
equalisation.
-- 
Robin Becker



From amackay at starvision.com  Mon Jul  5 18:41:01 1999
From: amackay at starvision.com (Angus MacKay)
Date: Mon, 05 Jul 1999 15:41:01 -0700
Subject: VIM and block jumping
Message-ID: <3781347D.CB2FBAB3@starvision.com>

this is a question for all of the VIMers out there.

is there any way to get [# ]# style logical python block jumping
to work for python? (to make up for the lack of {} in python)

cheers, Angus.



From matomira at iname.com  Thu Jul  1 08:04:11 1999
From: matomira at iname.com (Fernando Mato Mira)
Date: Thu, 01 Jul 1999 14:04:11 +0200
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <lwoghyexy5.fsf@copernico.parades.rm.cnr.it> <377B454C.6F59@mailserver.hursley.ibm.com>
Message-ID: <377B593B.F7F7504B@iname.com>

OK. Let's dispel Lisp mith #4 and then
let's get this out of comp.lang.python
[economics are different than other groups,
 BEWARE of Netscape's scrollbar!]

Paul Duffin wrote:

> > > I am surprised at this because Tcl's [upvar] is simply an explicit form
> > > of dynamic binding which I seem to remember is what Lisp uses. (That
>
> Mistake, I think I meant dynamic scoping.

>         (defun inner (b)
>           (+ a b))
>
>         (defun outer (a)
>           (inner 3))
>
>         (outer 2)

>
> A simple macro version of setq would convert from
>         (setq symbol '(list))
> to
>         (set 'symbol '(list))

All 3 the same instance of the first misconception.
That's OLD, BROKEN Lisp!!






From ddainese55x at x89dsi.unive.it  Tue Jul 27 17:00:01 1999
From: ddainese55x at x89dsi.unive.it (Diego Dainese)
Date: 27 Jul 1999 21:00:01 GMT
Subject: Access dictionary in order?
References: <roy-2507991300370001@mc-as01-p61.med.nyu.edu> <14235.18248.924848.52652@dolphin.mojam.com> <slrn7po8t3.6k.diego@blanka2.blankanet.it> <Wy_m3.4215$VL2.444209@news.direcpc.com>
Message-ID: <slrn7ps7eh.9r.diego@blanka2.blankanet.it>

On Mon, 26 Jul 1999 07:45:29 -0700, Emile van Sebille wrote:
[...]
> > But how much time requires the keylist generation?
> 
> I tried this test and found that the results varied on the number of
> items.  For example,
[snip]

You are right, my question was a bit naive...

> > That is, the keylist is made traversing all the dictionary nodes or is
> > cached inside it?
> 
> I can't help you with this one...

AFAICT (from you test program) the nodes are linked in a list and the
key-sequence is generated traversing this list. Moreover the sequence
is not cached:

  import time
  d = {}
  r = ('spam','eggs','eggs','bacon','eggs','and','spam')
  k = []
  for i in xrange(1000000):
    d[i] = i, 2*i, r
    k.append(i)
  
  test1start = time.clock()
  k = d.keys()
  test1end = test2start = time.clock()
  k = d.keys()
  test2end = time.clock()
  
  print 'test1: ',test1end - test1start
  print 'test2: ',test2end - test2start
  
The last 2 prints give always the same results (about 0.8 seconds on
my Linux box).

> > --
> > Sorry for my bad English!
> 
> Not bad at all.

Thanks :)

+++

Thanks for the reply (and to Skip Montanaro too), bye

-- 
Diego Dainese
--
To reply remove the numbers and the `x' from my address
--
Sorry for my bad English!



From gerrit.holl at pobox.com  Tue Jul 20 15:48:35 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Tue, 20 Jul 1999 19:48:35 GMT
Subject: python + bash + new user
In-Reply-To: <3793C729.55DE42D0@netcard.com>; from Joyce Stack on Tue, Jul 20, 1999 at 12:42:42AM +0000
References: <3793BE52.43E023BB@netcard.com> <7n0f3d$1d7@dfw-ixnews10.ix.netcom.com> <3793C729.55DE42D0@netcard.com>
Message-ID: <19990720214835.B30258@optiplex.palga.uucp>

On Tue, Jul 20, 1999 at 12:42:42AM +0000, Joyce Stack wrote:
> Sorry about being vague...i meant the Bash prompt...i have gotten other simple
> programs to work...changed mode and executed the programs ...but i dont know how to
> run the program and pass a variable at the same time...
> 

$ python -c "print 'Hello, I\'m called from the bash prompt... how nice\!'"

regards,
Gerrit.



From amk at mira.erols.com  Wed Jul  7 18:59:12 1999
From: amk at mira.erols.com (A.M. Kuchling)
Date: 7 Jul 1999 22:59:12 GMT
Subject: python core dump (SIGBUS) on Solaris
References: <3F8BA119B48AD11197470008C7A493CB98B443@HQOBGEX1> <37830BB2.94C4613D@uk.sun.com> <37837154.7A6889A@ix.netcom.com>
Message-ID: <slrn7o7nb9.eo.amk@207-172-111-58.s58.tnt1.ann.va.dialup.rcn.com>

On Wed, 07 Jul 1999 11:25:08 -0400, 
	Stephen J. Turner <sjturner at ix.netcom.com> wrote:
>> >>> C.load("webJudgeUser=garyp;")
...
>There's a bug in cPickle that causes heap corruption if loads is passed
>a string (as opposed to a pickled object) starting with 'g', 'h' or
>'j'.  Try applying the attached patch to the Python 1.5.2

	Wait a sec... why is it automatically trying to unpickle the
value of the cookie, when the value obviously wasn't pickled in the
first place.  While the cPickle core-dumping bug should definitely be
fixed, cookie.py also shouldn't try to unpickle things by default.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
If Abaton's the unearthly reflection of all that we are, Britain's in far
worse shape than I thought.
    -- John Constantine in HELLBLAZER #93



From ashfordv at usa.net  Fri Jul  2 05:35:36 1999
From: ashfordv at usa.net (cyberescue)
Date: Fri, 2 Jul 1999 04:35:36 -0500
Subject: newbie
Message-ID: <7li14e$mts$1@Nntp1.mcs.net>

I am new to programming. Where can I get a copy of Python?

mailto: ashfordv at usa.net







From gmcm at hypernet.com  Mon Jul 19 09:23:37 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 19 Jul 1999 08:23:37 -0500
Subject: Killing a thread
In-Reply-To: <7mu6a5$h61$1@nnrp1.deja.com>
Message-ID: <1279735662-4809835@hypernet.com>

Chris (catlee at my-deja.com) writes
> 
> I'm writing a progrma where some user code is executed in a
> restricted environment running in a separate thread.  Now, if the
> user is malicious, s/he can just put 'while 1: 1+1'.  I'd like to be
> able to recover from this.  Is there any way I can kill the thread
> that is stuck in the infinite loop?

Python uses OS threads. An OS thread is a "lightweight" process. Much 
of the savings in weight is because the OS does not clean up after 
the thread (closing files, releasing locks...). So the general advice 
is that threads should never be killed from outside.

- Gordon



From swdevpatrick at my-deja.com  Tue Jul  6 14:52:43 1999
From: swdevpatrick at my-deja.com (swdev)
Date: Tue, 06 Jul 1999 18:52:43 GMT
Subject: Developer Wanted, Portland, Oregon - e-commerce
Message-ID: <7ltj9m$lvv$1@nnrp1.deja.com>

Hello All!

Sorry to make such a blanket post (and, though I've been subscribed to
c.l.p for a couple of years now, this is one my first).  I'm really
hoping, though, to get a response from this group.

We're looking for a developer to join our team at Able Solutions, makers
of AbleCommerce (check both on the web).  We're located just north of
Portland, Oregon on Vancouver, Washington (not BC :)

Here's the text of the ad we're running:

----------------------
Software Developer Wanted

In this role you will be encouraged to contribute creatively
in the development of cutting edge applications for our
customers. Responsibilities include design and production of modules
and customizations for our award winning e-commerce products.

Requirements: candidates will have, above all, a bright mind and a
desire to work in a team environment. In addition, the candidate should
have significant, experience in as many of the following skills as
possible:

Web Development: HTML, IIS, scripting (ColdFusion, PHP, Perl, ASP,
JavaScript). Awareness of DHTML and XML would be helpful.

Database Technologies: SQL and query design. Experience with schema
design and database administration is a plus.

Object Oriented Development: Any experience with an object language
(Python, Java, C++, Smalltalk) or object design technologies (UML,
Booch) will prepare you for our future.
----------------------

To clarify, for the python community, the developers coming aboard will
join a web-based application effort with several targets.  Initially
this will mean implementation in ColdFusion (talking via SQL and taking
advantage of JavaScript where appropriate, yadayada).  The next phase
will be more cutting edge, javaservlets, JPython, XML/XSL and
Apache/Coccoon.

We have needs at several levels, from experienced team lead, to an entry
level sufficiently experienced not to require much handholding.  If
you're in the Portland/Vancouver area, PLEASE APPLY!

To Join Our Team email your resume and salary requirements to:
        info at ablecommerce.com
or mail to:
        Able Solutions Corporation 11700 NE 95th Street, Suite 100
        Vancouver, WA 98682 ATTN: Personnel Manager

Sincerly-hoping-for-a-fellow-python-er-ly yrs!
--p
Patrick Curtain, Husband & Father ( i also write software )
VP of Emerging Technologies, Able Solutions, Inc.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From zaks at prioris.im.pw.edu.pl  Sun Jul  4 13:37:33 1999
From: zaks at prioris.im.pw.edu.pl (Slawek Zak)
Date: Sun, 04 Jul 1999 17:37:33 GMT
Subject: Out of Office AutoReply: Daily Digest: comp.lang.python 1/2
References: <6935EFF15210D311896D00A0C9DAF28F0470D4@exchange.delphi.se>
Message-ID: <87g134uxsg.fsf@medievalist.org>

** David.sparrman at delphi.se wrote:
David> Jag har f?r n?rvarande semster ?ter p? kontoret m?ndag den 25
David> juli
No kidding :)

-- 
* Suavek Zak (Systems Administrator)
* email: zaks at im.pw.edu.pl   voice: +48 (0) 22 674 66 79
* PGP v2.6: 2048/9A7CBF71,   finger://zaks at prioris.im.pw.edu.pl



From bob at bluestarscrap.com  Fri Jul 23 11:12:36 1999
From: bob at bluestarscrap.com (Robert Zurer)
Date: Fri, 23 Jul 1999 08:12:36 -0700
Subject: Extreme Newbie Question
Message-ID: <MPG.1202263db22c797c989681@news.globix.net>

It's probably ridiculously obvious but it's my first hour working with 
Python...

In interactive mode how do you save the script that you've entered line 
by line to a file?

Bob Zurer



From python-list at teleo.net  Tue Jul 13 18:05:47 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Tue, 13 Jul 1999 15:05:47 -0700
Subject: XML-RPC
References: <931899884snz@vision25.demon.co.uk>
Message-ID: <9907131509510Q.08196@quadra.teleo.net>

On Tue, 13 Jul 1999, Phil Hunt wrote:
> Does anyone know if an XML-RPC server exists for Python?

Fredrik Lundh of PythonWare wrote a very nice XML-RPC library for
Python. (See http://www.pythonware.com/products/xmlrpc/ to get a copy.)

Zope now incorporates it in the core package, due to additional work
from Eric Kidd of Frontier and Jim Fulton of Digital Creations.

http://www.zope.org



From tismer at appliedbiometrics.com  Sun Jul 18 13:30:41 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Sun, 18 Jul 1999 17:30:41 GMT
Subject: How fast can we multiply?
Message-ID: <37920F41.185E012B@appliedbiometrics.com>

Hi all,

Lately, an old issue about long numbers and bitcounts
showed up again. I argued something about "if at all, then
build an extension with fast multiply".

I want to retract this. Python's C code needs a couple of
helper functions. But fast multiplication itself can be
easily done with plain vanilla Python.

Please try the attached script and enjoy.
The Python multiplication is twice as fast as the internal,
at about 100000 bits.

After this break, back to continuous stacklessness :-) chris
______________________________________________________________________
"""
fastlong.py

Fast long integers. 

Christian Tismer
990718

>From time to time I have argued on the main list that Python's
long integers should have a faster multiplication. It is a well
known fact that by some factorization, integer multiplication's
complexity can be brought from O(n**2) down to O(n**1.5) .

The method is sometimes called Karatsuba algorithm. But studying
chapter 12.7 of (1), it turns out to be nothing else but a special
case of preconditioning polynomials.

I think to remember that the GMP library uses this algorithm,
but I didn't look it up and tried my own straightforward way.

The reason why I don't aim to modify the C implementation any
longer is: It will not be reasonably faster in C. The real
working horse is still the basic internal long multiplication
(sometimes called "highschool algorithm") which is fast for small
long integers. The only effect of a C port would be to move the
break_even point to something smaller.

My proposal to get something like this into the C code:
For very long numbers to multiply, the C version could
be enabled to call back into the Python implementation.

A function to obtain the bit length of a long would
be most helpful. Some kind of split function as well.
Currently, with a "clean" nbits implementation, the
break even point appears to be 50000 bits.
With a marshal-based inaccurate hack, break even is 
around 10000 bits.

For a bit length of 100000, my algorithm outperforms
the built in long numbers by a factor of 2.


(1) Aho/Hopcroft/Ullmann: 
    "The Design and Analysis of Computer Algorithms"

"""

class fastlong:
    """a class around long numbers which can multiply faster than
long"""
    break_even = 50000 # clean nbits
    break_even = 10000 # hacked nbits
    def __init__(self, val):
        if isinstance(val, fastlong):
            self.val = val.val
        else:
            self.val = long(val)
        
    def __repr__(self):
        return repr(self.val)
        
    def __add__(self, other):
        return self.val + fastlong(other).val
        
    def __sub__(self, other):
        return self.val - fastlong(other).val

    def __div__(self, other):
        # not optimized yet
        return self.val / fastlong(other).val
        
    def __pos__(self):
        return self
        
    def __neg__(self):
        return fastlong(-self.val)
        
    def __lshift__(self, bits):
        return self.val << bits
 
    def __rshift__(self, bits):
        return self.val >> bits
 
    def __mul__(self, other):
        """Karatsuba alogrithm
        (a+bX)(c+dX)=(X^2+X)(bd)+(X)(a-b)(d-c)+(X+1)(ac)
        """
        val1 = self.val
        val2 = fastlong(other).val
        if not val1 and val2:
            return 0
        bitcount = max(nbits(val1), nbits(val2))
        if bitcount < self.break_even:
            return val1 * val2
        shift = bitcount / 2
        (hi1, lo1) = split_num(val1, shift)
        (hi2, lo2) = split_num(val2, shift)
        sum1 = lo1-hi1
        sum2 = hi2-lo2
        mul1 = hi1 * hi2
        mul2 = lo1 * lo2
        mul3 = sum1 * sum2
        res = mul2 + ((mul1 + mul2 + mul3) << shift) + (mul1 <<
shift+shift)
        return res
        
def split_num(longval, shift):
    scale = 1L << (shift)
    mask = scale - 1
    lo = longval & mask
    hi = longval >> shift
    return fastlong(hi), fastlong(lo)
        
def nbits(x):
    """clean, accurate, not so fast"""
    if x <= 2:
        if x < 0:
            return nbits(-x)+1
        return x
    sum = 1
    while x:
        shift = 1
        x = x2 = x >> 1
        while x2:
            x = x2
            sum = sum + shift
            shift = shift + shift
            x2 = x >> shift
    return sum

import marshal, struct

def nbits(x):
    """not clean, rather fast, inaccurate"""
    s=marshal.dumps(x)
    return abs(struct.unpack("i", s[1:5])[0])*15 -7

#---------------------------------------------

def test():
        import sys
        be = fastlong.break_even
        for i in range(be/2, be*16, be/4):
            arg = (1l << i+1) -1
            fastarg = fastlong(arg)
            tim1, res1 = timing(lambda x:x*x, arg)
            tim2, res2 = timing(lambda x:x*x, fastarg)
            if res1 <> res2:
                raise ValueError, "different results fo i=" % i
            print i, tim1, tim2
            sys.stdout.flush()

def timing(func, args, n=1, **keywords) :
	import time
	time=time.time
	appl=apply
	if type(args) != type(()) : args=(args,)
	rep=range(n)
	before=time()
	for i in rep: res=appl(func, args, keywords)
	return round(time()-before,4), res
	
# the end ----------------------------------------------------

______________________________________________________________________
-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From sjturner at ix.netcom.com  Fri Jul 23 09:08:53 1999
From: sjturner at ix.netcom.com (Stephen J. Turner)
Date: Fri, 23 Jul 1999 09:08:53 -0400
Subject: Problem with informixdb and python1.5.2
References: <000016a3@bossar.com.pl> <378AEC72.8D12D0A7@aethos.co.uk>
Message-ID: <37986965.F8931610@ix.netcom.com>

Dominic Binks wrote:
> 
> Ken Liu wrote:
>>
>> Python 1.5.2 (#8, Jul  1 1999, 20:22:25)  [GCC 2.7.2.3] on linux2
>> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>> >>> import informixdb
>> >>> con = informixdb.informixdb('acme_online_demo')
>> Traceback (innermost last):
>>   File "<stdin>", line 1, in ?
>>   File "informixdb.py", line 47, in __init__
>>     self.conn = _informixdb.informixdb(logon)
>> InformixdbError: Error -922 performing LOGON: Cannot get name of current
>> working directory.
>> >>>
> 
> Can informix read your current working directory?  I've certainly seen
> this cause problems with dbaccess, so I don't see why the ESQL/C
> libraries should be much different.

As it happens, Ken's problem turned out to be not with the informixdb
module nor the Informix ESQL/C runtime, but with the Python interpreter
-- or more precisely, the C runtime library(ies) with which Python was
linked.  All calls to the getcwd C function were failing with an errno
of ENOENT, as seen below in the exception raised by the os.getcwd Python
function:

>>> import os
>>> os.getcwd()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
OSError: [Errno 2] No such file or directory
>>>

Using ldd to check the list of shared objects loaded for the python
executable revealed the problem's root cause.  Ken explained in an
email:
> thanks, this allowed me to track down the problem.  In order to get readline
> support, I had to link against libtermcap, which linked against libc.so.5,
> so my python was linked against both libc.so.6 and libc.so.5, and chaos
> ensued.  So now I changed it to link against libncurses and everything is
> working great.

Don't you just love a story with a happy ending? :-)

--
Stephen J. Turner <sjturner at ix.netcom.com>



From conversy at lri.fr  Mon Jul 26 12:57:46 1999
From: conversy at lri.fr (Stephane Conversy)
Date: Mon, 26 Jul 1999 18:57:46 +0200
Subject: Tkinter bug ?
References: <379C8F64.E48BF26@lri.fr> <001301bed786$e88a7490$f29b12c2@secret.pythonware.com>
Message-ID: <379C938A.27F0EF@lri.fr>

FAQ:

Well, the Tk button widget keeps a reference to the internal
  photoimage object, but Tkinter does not.  So when the last
  Python reference goes away, Tkinter tells Tk to release the
  photoimage.  But since the image is in use by a widget, Tk
  doesn't destroy it.  Not completely.  It just blanks the image,
  making it completely transparent...

ok, but does it have something to do with the fact that in one case
I do a from file1 import * and in the other import file1 ?

I'm a beginner in Python, and don't know exaclty how modules
are implemented...

    stef


--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr






From ron at library.ucsf.edu  Fri Jul 30 13:47:20 1999
From: ron at library.ucsf.edu (Ron Klatchko)
Date: Fri, 30 Jul 1999 10:47:20 -0700
Subject: Is bsddb.pyd thread safe in 1.5.2 (win95)
References: <7npimc$l9b@romeo.logica.co.uk> <Z37o3.41724$jl.29736754@newscontent-01.sprint.ca> <7nrpm0$hcq@romeo.logica.co.uk>
Message-ID: <37A1E528.FE2E982F@library.ucsf.edu>

Ian Clarke wrote:
> Is there any way to find out which version of dbm the 1.5.2 bsddb.pyd file
> is based on?

Python's bsddb module uses the DB 1.* interface.  You can compile it
against the DB 2.* by changing the include from:
  #include <db.h>
to:
  #include <db_185.h>

which implements a compatibility interface.  I've been looking at the
internals of the bsddb module to have it detect which version it's using
and to handle things appropriately but don't have anything production
quality yet.

moo
----------------------------------------------------------------------
          Ron Klatchko - Manager, Advanced Technology Group           
           UCSF Library and Center for Knowledge Management           
                        ron at library.ucsf.edu



From landauer at apple.com  Fri Jul 30 13:45:45 1999
From: landauer at apple.com (Doug Landauer)
Date: Fri, 30 Jul 1999 10:45:45 -0700
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <slrn7psisn.da.wtanksle@dolphin.openprojects.net> <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com> <slrn7pum9t.2ro.wtanksle@dolphin.openprojects.net> <001501bed935$561b8210$f29b12c2@secret.pythonware.com> <slrn7q1c10.5jp.wtanksle@dolphin.openprojects.net> <00ad01beda66$d6bdb9b0$f29b12c2@secret.pythonware.com>
Message-ID: <landauer-3007991045450001@landauer.apple.com>

> for an "if" statement, "else" is a path taken if the
>    expression evaluates to false.

In my mind, for an "if" statement, "else" is the path taken
when the BODY is not executed at all.

> for a "while" statement, "else" is a path taken if the
>    expression evaluates to false.  or in other words,
>    when the loop terminates by natural causes.

And therefore, the "else:" on either while or for loops, as
defined in Python, is completely counterintuitive.

> not that complicated, was it?

It's not complicated, it's just something that is easily
open to multiple interpretations, depending on how one views
the normal, common usage of the word.  As such, it is one of the
very few things in Python that (for some of us) is not clear from
context -- it just doesn't look obvious what the behavior should
be.  So, in the spirit of python, we don't use it.  Unfortunately,
others do.  I hope no one ever uses it without a comment explaining
what it means.

I know it ain't gonna change soon; but one can hope that perhaps
Python 2.0 could choose a better keyword for it -- one that doesn't
come with preconceived notions that the current definition violates.
-- 
    Doug Landauer        landauer at apple.com (work)
                         landauer at scruznet.com (not-work)



From bece at bece.net  Mon Jul  5 06:23:26 1999
From: bece at bece.net (BeCe)
Date: Mon, 05 Jul 1999 10:23:26 GMT
Subject: Python, ASP - Namespace (again)
References: <O18e3.449$U%.3201@news1.online.no> <7lbj2t$2es$1@m2.c2.telstra-mm.net.au> <377b4c67.12387312@news.mch.sni.de>
Message-ID: <3780868f.1123180109@news.telepac.pt>

On Thu, 01 Jul 1999 11:18:23 GMT, nikolai.kirsebom at siemens.no (Nikolai
Kirsebom) wrote:

>The next time I checked the code written in my first message, it
>worked ok !!!!
>
>As I indicated in the message, I have one ASP file and two PY files.
>Updating the ASP file (saving it) followed by a refresh in the browser
>results in updated browser page content.  However, updating any of the
>PY files does not result in any changes in the browser output.  Could
>anyone answer when the PY modules are loaded.  I thought the PY
>modules automatically are reloaded if the datestamp of the *.py file
>is newer than the *.pyc file.

IIS does the same thing with include files, i.e. only when the .asp
timestamp changes it is reprocessed, the same thing happens with
python. If the code is inside the .asp it works fine, if the code is
outside (i.e. in a .py) you have to touch the .asp so it gets
reprocessed.

>I'm using the statement:
>
>from <module> import *
>
>How (if possible) would I reload explicitly such a module ?

Touching the .asp.




From dwelton at linuxcare.com  Fri Jul 30 21:31:28 1999
From: dwelton at linuxcare.com (David N. Welton)
Date: 30 Jul 1999 18:31:28 -0700
Subject: Import a lot of modules, or, import 'foo'
References: <874silabe7.fsf@linuxcare.com> <1%ro3.4357$VL2.467111@news.direcpc.com>
Message-ID: <871zdpa9lb.fsf@linuxcare.com>

"Emile van Sebille" <emile at fenx.com> writes:

> Hi David,
> 
> This seems to work for me:
> 
> >>> for t in tests:
>  exec ('import %s' % t)

Eewwwww, that's kind of twisted, but I guess if it works:-)

Thanks,
-- 
David N. Welton, Web Engineer, Linuxcare, Inc.
415.354.4878 x241 tel, 415.701.7457 fax
dwelton at linuxcare.com, http://www.linuxcare.com/
Linuxcare. At the center of Linux.



From jbooth at ux4.cso.uiuc.edu  Sun Jul 18 23:06:19 1999
From: jbooth at ux4.cso.uiuc.edu (jonathan allen booth)
Date: 19 Jul 1999 03:06:19 GMT
Subject: Restricted Execution
Message-ID: <7mu4nb$9gd$1@vixen.cso.uiuc.edu>

I was looking at embedding Python where it will get user-input code,
which may be malicious.

Obviously I'm going to be using rexec to run it, to take care of part of
that problem, but that still leaves the problem of non-terminating code
open -- someone passing in a while 1: with no break or other exit
mechanism.

Since rexec can't help here (according to the docs), I perused the code
and came up with two options:

1. It looks like profiling support may be able to simulate this, written
   in python code (hacked profiling), but at a preformance penalty.

2. Of course instruction count limiting can always be hacked into the
   interpreter. Perusing the source, I saw a recursion depth check (which I
   don't think is settable dynamically in code -- it'd be nice if it was)
   in ceval.c which could have a sister instruction count check added
   alongside it (but that only makes sense if it would be dynamically
   settable, so you could turn off the limit when you're running known-safe
   code).

I'd prefer 1 even if it has a preformance penalty, if it isn't likely
that the changes in 2 would become part of the main distro (don't want
to have to tell people to get python, apply this patch, compile it and
so fourth).

Is there already a solution out there I've overlooked? Any comments on
either of the above solutions for this problem (I want to avoid having
to fork a whole new process or use threads if I can)?

Thanks,

-- 
Jonathan A. Booth				kamikaze at N0$PAMimsa.edu
http://www.N0$PAMimsa.edu/~kamikaze		jbooth at N0$PAMuiuc.edu
Remove NOSPAM from hostnames. Unsolicited commercial e-mail: access denied



From tjreedy at udel.edu  Thu Jul  1 21:04:37 1999
From: tjreedy at udel.edu (Terry Reedy)
Date: 2 Jul 1999 01:04:37 GMT
Subject: Tuples & lists
References: <377bbde1@news.xtml.co.uk>
Message-ID: <7lh375$5ea$1@news.udel.edu>

In article <377bbde1 at news.xtml.co.uk>, nick at NOSPAMvideosystem.co.uk says...

>Why both having tuples and lists? As far as I can see there're the same, but
>tuples are read only..? Have I missed something in the tutorial?

There is probably something in the the FAQ.  BAsically, tuples are stored 
more compactly and can be used as dictionary keys (because they ARE 
readonly). 

TJR




From skip at mojam.com  Fri Jul 23 19:26:33 1999
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 23 Jul 1999 18:26:33 -0500 (CDT)
Subject: Xemacs Python Mode.
In-Reply-To: <3798E523.7AF3DE05@jps.net>
References: <199907231949.PAA07568@python.org>
	<3798CF69.1FC04C23@jps.net>
	<3798DE2D.C008286D@aw.sgi.com>
	<3798E523.7AF3DE05@jps.net>
Message-ID: <14232.63872.409563.299007@227.chicago-26-27rs.il.dial-access.att.net>

    Chris> This gets me close enough to be functional, but isn't quite
    Chris> there. The problem is that if there are spaces in there, it won't
    Chris> reset them to tabs, which is what I would like.

Sounds like a one-shot pass over the code with sed, tr, indent or Emacs'
tabify command should do the trick.  I'm not used to TAB==SPC*4 mode, so I
won't hazard a guess at the appropriate settings, but a little
experimentation should converge on a solution fairly quickly.

In fact, maybe Emacs' tabify command will just do the right thing if you
have tab-width set properly.  Try C-h f tabify RET.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From robin at jessikat.demon.co.uk  Thu Jul 22 06:34:21 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 22 Jul 1999 11:34:21 +0100
Subject: VC++ 6.0 larger exe files explained
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC15@molach.origin.ea.com>
 <379695A7.EB1E7A6F@compaq.com>
Message-ID: <4ypvvfAtOvl3Ew5y@jessikat.demon.co.uk>

In article <379695A7.EB1E7A6F at compaq.com>, Greg Ewing
<greg.ewing at compaq.com> writes
>"Stidolph, David" wrote:
>> 
>> This might (stress this is a pure guess) have something to do with virtual
>> memory and its size being 4k pages.
>
I agree that pure code and data sections might be more desirably a
multiple of the page size; but most sections aren't actually pure. They
must be manipulated after loading to do fixups etc etc. Also some are
nothing to do with the traditional executable model eg I see the list
below (woe betide the 'hello world' with all of these) and this also
applies to the most trivial dll. The existing loader handles the old
non-aligned exe's quite happily. Python.exe would fit in 2x4k now it
needs 4x4k so the disk time loss is probably worse than the saving in
cpu. I have thousands of dlls on my machine; if the each wasted 10k I
would be upset.

Sections 

.text Machine code instructions.

.data Initialized data.

.rdata Read only data. OLE GUIDs are stored here, among other things.

.rsrc Resources. Produced by the resource compiler, and placed into RES
files. Linker copies it to the executable.

.reloc Base relocations. Produced by the linker. Not found in OBJs.

.edata The exported function table. Created by the linker and placed in
an EXP file. Linker copies it to the executable.

.idata Imported function table in an executable file.

.idata$XXX Portions of an imported function table. The librarian creates
these sections in an import library. The linker combines them into the
final .idata section in the executable.

.CRT Tables of initialization and shutdown pointers in the executable
that are used by the Microsoft C++ runtime library.

.CRT$XXX Initialization and shutdown pointers in OBJs, prior to the
linker combining them in the executable.

.bss Uninitialized data.

.drectve OBJ file section containing linker directives. Not copied to
executable.

.debug$XXX COFF symbol table information in an OBJ file.


>That sounds like a pretty good guess. If Windows VM is anything
>like Unix, it'll want to keep the code & data in separate pages
>so it can share the code between processes.
>
>In the case of extremely small programs (i.e. ones whose code
>& init data together fit in one page), it might be better to
>forget about trying to share the code and just load a separate
>copy into each process.
>
>Mind you, Microsoft programmers have probably long since
>forgotten that there's any such thing as an executable smaller
>than about 5MB, so it's understandable that they wouldn't
>have thought of this...
>
>Greg

-- 
Robin Becker



From m.faassen at vet.uu.nl  Tue Jul  6 06:14:11 1999
From: m.faassen at vet.uu.nl (m.faassen at vet.uu.nl)
Date: 6 Jul 1999 10:14:11 GMT
Subject: ANN: Minotaur 0.1 - bridging Tcl, Python, and Perl
References: <37812D1F.F18721A4@equi4.com>
Message-ID: <7lsktj$t7u$1@news.accu.uu.nl>

Jean-Claude Wippler <jcw at equi4.com> wrote:
[Minotaur description, involving at least 4 languages (forth, python, perl,
tcl, and probably C too) snipped]

Wow, amazing.. I'm not sure if I'll *use* it a lot, but this is a wild bit of
uberhacking indeed!

My compliments.

Regards,

Martijn




From Alex.Maranda at p98.f112.n480.z2.fidonet.org  Thu Jul  1 12:41:27 1999
From: Alex.Maranda at p98.f112.n480.z2.fidonet.org (Alex Maranda)
Date: Thu, 01 Jul 1999 17:41:27 +0100
Subject: Python suitable for a game engine?
Message-ID: <00001614@bossar.com.pl>

From: Alex Maranda <amaranda at nospam.com>

> their leader is killed, etc. The events would be things such as
> "create", "monster sees player", "clock tick", etc. All the functions
> in the script would be callbacks written in C++, to manipulate the
> monster to do things such as "patrol area", "play an animation",
> "sleep for X seconds", etc. In other words, the scripts would be glue,
> not the main thread of execution.
you use confusing terms; I understand what you're saying, but you
shouldn't use the word 'thread' for anything else than its OS meaning.
You could say the 'the bulk of processing'. You might very well end up
using real Python threads; don't forget in this case to release/reaquire
the global interpreter lock in your C++ callbacks, if you make blocking
syscalls (unlikely in a game).

> 
> I'm been using Python for a couple of years now, and it is my favorite
> scripting language. But, I don't have much experience with embedding
> it in a another program, and I don't know enough about the speed
> issues. So, my questions are...
> 
> - Will Python be fast enough? Even with possibly 100 or 200 active
>   items on a level?
Unlikely. You won't know until you try. To my knowledge UnrealScript is
eventually translated to C++.

> - Especially when all of the scripts will be live, and usually in the
>   middle of a callback?
> - Should I use the interpreter in the library, or should I take an
>   approach of crunching the byte-codes myself?
??

Cheers,
-- 
Alex Maranda         mailto: amaranda at spider dot com
Spider Software Ltd. Tel: +44 (0)131 4757036
Edinburgh, UK        http://members.xoom.com/Alex_Maranda
STREAMS based communications protocols for embedded systems



From larsga at ifi.uio.no  Mon Jul 26 10:31:29 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 26 Jul 1999 16:31:29 +0200
Subject: How to use a variable form main script in a module.
References: <7nhmt5$su0@enews3.newsguy.com>
Message-ID: <wkhfmrscse.fsf@ifi.uio.no>

* Peter Torstenson
|
| This is probaly a silly question from a newbie.
| In my main script I put a value for to a variable called "x".
| Now I want to use the "x" variable in a module.
| The module is imported into the main script. How to do this?

>From the way you describe it this sounds like something you shouldn't
do. Communication between different pieces of code through globabl
variables is in general frowned upon, and with good reason. It's good
for some things, but for those the variables tend to reside in a
module, and not in the global namespace.

Feel free to explain the reason why you want this if you want a
verdict on whether it's a good or bad idea.

If you still want to do this you can use the 'global' keyword, like
so:

def myfunc(blah):
  global var
  return var+blah

--Lars M.



From tbryan at arlut.utexas.edu  Tue Jul  6 13:13:02 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Tue, 06 Jul 1999 12:13:02 -0500
Subject: changing some lines in a file
References: <1280853815-64731922@hypernet.com> <37821035.57FCDE9E@daimi.au.dk>
Message-ID: <3782391E.95791628@arlut.utexas.edu>

Peter Posselt Vestergaard wrote:
> 
> Hi again, and thanks for your answer, but it doesn't seem to help.
> What I meant by stopping was that I have tried to put some prints
> different places(se below), and it won't print anything out after the
> problematic line.

It's *very* difficult to troubleshoot programs online without a lot
of information from you. Generally, we're not going to solve it, 
we're just going to ask you so many questions that you'll eventually
find a bug in your code. :)

Does the program run to completion?
  (If so, could you provide a sample input, output, and 
   values in Form?  If not, does it exit with an exception, 
   or does it just run forever?)
Does it print something before the problematic line 
 and not after it?  Or do neither print statements work?
Are you sure that Form has the data you think it has?
  (That is, have you dumped Form before the for loop?)
Is found ever being set to 1?
This program looks like it's dealing with input from a
 web form or an interactive session?  Is that true?
 Which one?

> db.close

By the way, you missed one of Gordon's remarks.
This line (and the db.close at the end) should be 
db.close()
 
---Tom



From arcege at shore.net  Thu Jul  1 11:37:34 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 01 Jul 1999 15:37:34 GMT
Subject: ioctl in python
References: <377A71D7.470AF9DA@starvision.com> <377A7374.C6919FD3@starvision.com> <377AA64A.B4C63805@bigpond.NO.SPAM.com> <377AB10E.B0B85931@starvision.com>
Message-ID: <2XLe3.518$6M6.157342@news.shore.net>

Angus MacKay <amackay at starvision.com> wrote:
: hun?

: why is this expecting a 4 byte buffer? my C version uses an int*
: and it works no problem.

: as for 80 being a input I think you have it backwards (for linux-i386 at least):
: 0x80046601  EXT2_IOC_GETFLAGS      int *
: 0x40046602  EXT2_IOC_SETFLAGS      const int *
: 0x80047601  EXT2_IOC_GETVERSION         int *
: 0x40047602  EXT2_IOC_SETVERSION         const int *

: cheers, Angus.

The four bytes is the sizeof(int *), but also (according to some linux
headers I was looking at) it is expecting a pointer to a long, not an
int.  A long is a 4 byte value (not to mention that an int is also a
long on most systems).

But, the ioctl call takes some 4 byte value, it could be cast from a
char, short, pointer or whatever; it is up to the called request to
interpret the value (int *, in this case) given.  In your original
(Python) program, you gave ioctl() a fairly short character string
(three bytes if I remember correctly), which would not be the size of a
long.

  -Arcege




From emile at fenx.com  Fri Jul  2 08:29:25 1999
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 2 Jul 1999 05:29:25 -0700
Subject: Alternatives to pickle/class nesting
Message-ID: <Ch2f3.2457$VL2.321674@news.direcpc.com>

Good morning,

Unlike Gerrit's problem with seemingly large amounts of space being
consumed for storage, I cpickled my object and was impressed that only
81 bytes were used.  Of course, the entire object didn't get pickled.
In browsing dejanews, I found a post by Guido saying that nested class
structures, being rare, are not supported by pickle.  Being new to this,
I didn't anticipate the pending pickle problem, and now need to either
restructure the classes into one pickle-friendly object with the same
functionality, or look at alternatives for swizzling these objects?

Essentially, my class instances consist of a few variables, and two
dictionaries of string keys and UserList values modified for write-only
stack-like behavior.  If I do restructure the classes, by retaining the
same/similar structure without external class instances in the
structure, is it likely pickle will deal with it properly?

Can __getinitargs__, __getstate__ and __setstate__ be used to correct
this?

I have a debug display routine that dumps the contents of these
instances.  Any way to use this routine to create a local __dump__ /
__load__ pair.  Dumping I think I get, but are there traps in the
loading?

Any suggestions?

--

Emile van Sebille
emile at fenx.com
-------------------







From da at ski.org  Wed Jul 21 14:35:22 1999
From: da at ski.org (David Ascher)
Date: Wed, 21 Jul 1999 11:35:22 -0700 (Pacific Daylight Time)
Subject: [Tutor] I want to learn to hack
In-Reply-To: <19990721154610.B3663@optiplex.palga.uucp>
Message-ID: <Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org>

On Wed, 21 Jul 1999, Gerrit Holl wrote:

> On Tue, Jul 20, 1999 at 02:31:55AM -0400, Tim Peters wrote:
> > > I want to learn to hack
> > So, little Roger, you want to learn how to hack?  You came to the right
> > place!  First walk down the hall and pick up a copy of Python from Guido.
> > He'll have you cracking bank accounts in no time <wink>.
> > 
> > the-psa-is-the-richest-organization-on-earth-ly y'rs  - tim
> > 
> 
> Well, I think Roger meant hack, not crack. Have a look at Eric S Raymonds
> "How to become an hacker FAQ" at http://www.tuxedo.org/~esr/ (and some
> further).
> 
> Hacking is good,
> cracking is bad.
> 
> I also want to learn to hack, I'm learning it!
> I *don't* want to learn to crack!!

Gerrit -- just to fill you in -- Roger was presumably either joking or
testing a spam filter -- he is a colleague of Guido's, and knows all about
hacking.  Whether he knows about cracking or not is not something he's
likely to disclose in public. =)

(In other words, Tim was kidding, Roger was kidding, I'm not kidding!).

--david




From aahz at netcom.com  Wed Jul 28 18:36:59 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 28 Jul 1999 22:36:59 GMT
Subject: a little motivation, anyone?
References: <19990728052704.10608.00002260@ng-fq1.aol.com> <MBJn3.859$Yk3.6771@newsr1.twcny.rr.com> <GTJn3.79756$eF3.24937@news.rdc1.nj.home.com>
Message-ID: <7no0mb$47k@dfw-ixnews16.ix.netcom.com>

In article <GTJn3.79756$eF3.24937 at news.rdc1.nj.home.com>,
Brett g Porter <BgPorter at NOacmSPAM.org> wrote:
>
>Hint #4: Read lots of books about writing code. Some of my faves:
>'The Practice of Programming', Kernighan & Pike
>'Code Complete', Steve McConnell
>'The Zen of Code Optimization', Michael Abrash

_The Mythical Man-Month_, Frederick Brooks
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From fredrik at pythonware.com  Fri Jul 23 08:42:57 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 23 Jul 1999 12:42:57 GMT
Subject: Q: anyone running Python on OS X server?
References: <004401bed44f$1a5c6400$f29b12c2@secret.pythonware.com> <37979524.C9F4F6BC@freenet.co.uk>
Message-ID: <008c01bed508$e4e85a80$f29b12c2@secret.pythonware.com>

sreekant kodela <sreekantk at freenet.co.uk> wrote:
> Do you mean Linux / Xfree ?
> If that is what you asked then answer is Yes.

no, I meant "OS X," as in "Mac OS X Server":
http://www.apple.com/macosx/server/

turns out that the answer is "yes, works right
out of the box!".  (which is pretty cool, IMO).

not sure if it's runs X, though ;-)

</F>




From poinot at onera.fr  Mon Jul 19 08:38:07 1999
From: poinot at onera.fr (Marc POINOT)
Date: Mon, 19 Jul 1999 14:38:07 +0200
Subject: Python/Gadfly and SQL wildcards
References: <1279735665-4809662@hypernet.com>
Message-ID: <37931C2F.726B1D55@onera.fr>

Gordon McMillan wrote:

> Colm Rafferty writes:
>
> > Does anyone know how to implement the wildcard in gadfly SQL?
> > e.g.
> > Select * from table where fieldname = 'A*'
> > or
> > Select * from table where fieldname like 'A*'
> >
> > * does not work!
> > Though it does in other SQL languages
>
> In ANSI SQL, that would be "fieldname like 'A%'"
>
> Last I used Gadfly, it didn't support 'like', so you're stuck with
>
> ..fieldname >= 'A' and fieldname < 'B'
>
> or the same using "between".
>

The GadFly doc says the "like" statement will not be
implemented. It says you have to use the Python strings
or RE facilities to do that.

Marcvs [alias The Gadfly doc is clear about what is SQL compilant or not
]







From dan at detached.demon.co.uk  Thu Jul 15 19:20:34 1999
From: dan at detached.demon.co.uk (Daniel Barlow)
Date: 16 Jul 1999 00:20:34 +0100
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <377B454C.6F59@mailserver.hursley.ibm.com> <7liguf$58o$1@Starbase.NeoSoft.COM> <m0JTEFA0DTf3EwLV@ugglan.demon.co.uk> <slrn7nss0n.5u0.wtanksle@dolphin.openprojects.net> <378077E1.5E65@mailserver.hursley.ibm.com> <slrn7o58l7.4vr.wtanksle@dolphin.openprojects.net> <3783305C.33FA@mailserver.hursley.ibm.com> <slrn7o7d7o.6la.wtanksle@dolphin.openprojects.net> <3785B997.2001@mailserver.hursley.ibm.com>     <w6btdmw4mm.fsf@wallace.nextel.no> <3789ABAB.1E71@mailserver.hursley.ibm.com> <slrn7okdnn.ep3.wtanksle@dolphin.openprojects.net> <n7vi3.21348$AU3.605286@news2.giganews.com> <378CCAD2.97D4FEA7@research.att.com> <dkbj3.37476$AU3.735796@news2.giganews.com>
Message-ID: <m3n1wxo5z1.fsf@detached.demon.co.uk>

cbbrowne at news.hex.net (Christopher Browne) writes:
> You mean I *forgot* to indicate that the type of each object needs to
> be attributed with its unique punctuation?

It would help if you were also to overload the + operator to do string
concatenation.  Oh, and silently coerce things to strings.

-dan



From commando at netvigator.com  Sun Jul 18 08:12:26 1999
From: commando at netvigator.com (commando at netvigator.com)
Date: Sun, 18 Jul 1999 20:12:26 +0800
Subject: Please visit
References: <7msg8t$or7$1@imsp009a.netvigator.com>
Message-ID: <7msgal$l90$1@imsp009a.netvigator.com>

Wrong Post, Sorry!!
<commando at netvigator.com> wrote in message
news:7msg8t$or7$1 at imsp009a.netvigator.com...
> http://www.digitalworld.com.hk/front_c.htm
>
>





From posselt at daimi.au.dk  Tue Jul  6 10:18:29 1999
From: posselt at daimi.au.dk (Peter Posselt Vestergaard)
Date: Tue, 06 Jul 1999 16:18:29 +0200
Subject: changing some lines in a file
References: <1280853815-64731922@hypernet.com>
Message-ID: <37821035.57FCDE9E@daimi.au.dk>

Hi again, and thanks for your answer, but it doesn't seem to help.
What I meant by stopping was that I have tried to put some prints
different places(se below), and it won't print anything out after the
problematic line. 
Best regards
Peter Vestergaard

db = open("db.dat","r")                                          
data = db.readlines()
db.close

found=0
for i in range(len(data)):
    line = data[i]
    if found==1:
        if line[0:2]=='fi':
            print data[i]
            data[i] = 'fi=' + Form["fi"].value
            print data[i]
        elif line[0:2]=='la':
            line='la='+Form["la"].value
        elif line[0:2]=='em':
            line='em='+Form["em"].value
        elif line[0:2]=='ho':
            line='ho='+Form["ho"].value
        elif line[0:2]=='st':
            line='st='+Form["st"].value
            found=2
    if line[0:9]==('id='+Form["id"].value):
        found=1
if found==0:
    print '<H1>Fejl i ?rskortnummeret!</H1>'
    stop 
#udskift databasen med den ?ndrede version
db = open("db.dat","w")
db.writelines(data)
db.close


Gordon McMillan wrote:
> 
> Peter Posselt Vestergaard writes:
> 
> > Hi, I'm quite a newbie in Python-programming, so I hope you'll
> > forgive me asking about some pretty things. :-) My problem is that I
> > have a file in which i want to find some lines and replace them with
> > some other data. I've tried a lot of things but nothing seems to
> > work.
> 
> Unfortunately, you haven't said what "doesn't work" means.
> 
> > Below is my best try, but it seems that the script stops when
> > it is about to execute the line:
> 
> Stops how? With a traceback?
> 
> > line=('fi='+Form["fi"].value)
> >
> > I'll be thankfull for any kind of help.
> > Best regards
> >
> > Peter Vestergaard
> >
> > ------------------------------------------------------------------
> >
> >
> > db = open("db.dat","r+")
> 
> Why are you opening for update?
> 
> > data = db.readlines()
> > db.close
> 
> MIssing something?
> 
> >
> > found=0
> > for line in data:
> >     print found
> >     if found==1:
> >         if line[0:2]=='fi':
> >             line=('fi='+Form["fi"].value)
> 
> Ah. Well you'll never see any results, because you're rebinding the
> variable "line" to a new string. When you return to the "for", you
> lose all your hard work.
> 
> You can either modify the data list by using
> 
> for i in range(len(data)):
>   line = data[i]
>   .....
>      data[i] = 'fi=' + Form[whatever]....
> 
> Or create a new list and append the modified lines to it.
> 
> (Updating a file in place usually means lots of seek()s and tells()
> and is basically never worthwhile on text files.)
> 
> >         elif line[0:2]=='la':
> >             line='la='+Form["la"].value
> >         elif line[0:2]=='em':
> >             line='em='+Form["em"].value
> >         elif line[0:2]=='ho':
> >             line='ho='+Form["ho"].value
> >         elif line[0:2]=='st':
> >             line='st='+Form["st"].value
> >             found=2
> >     if line[0:9]==('id='+Form["id"].value):
> >         found=1
> > if found==0:
> >     print '<H1>Fejl i ?rskortnummeret!</H1>'
> >     stop
> >
> > db = open("db.dat","w")
> > db.writelines(data)
> > db.close
> >
> > --
> > http://www.python.org/mailman/listinfo/python-list
> 
> - Gordon

-- 
Med venlig hilsen
Peter Posselt Vestergaard



From nick.bower at ssec.wisc.edu  Sun Jul 18 15:25:00 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sun, 18 Jul 1999 14:25:00 -0500
Subject: what is python?
References: <7mt81n$s6e$1@bgtnsc03.worldnet.att.net>
Message-ID: <37922A0C.D82170BB@ssec.wisc.edu>

> the fastest way is to put it to work right away.  Therefore, I guess I'm
> trying to figure out what python excels in and then I'll allocate a
> problem-domain (in my own work) to apply it to -- hopefully this will be the
> most pragmatic approach.

I think these are the kind of questions outlined in any FAQ.  The Python
website (http://www.python.org) has exceptional help for (but not
restricted to) new programmers, including a link to an FAQ.  Perhaps the
reading the sections "1.1. What is Python?", "1.17. What is Python good
for?", Who uses Python? etc. should help you rethink your question.

cheers, nick



From bhoel at server.python.net  Thu Jul  8 13:41:59 1999
From: bhoel at server.python.net (Berthold =?iso-8859-1?Q?H=F6llmann?=)
Date: Thu, 08 Jul 1999 19:41:59 +0200
Subject: Trouble with HTMLGen, HTMLtest
References: <e3UDcTMy#GA.342@nih2naae.prod2.compuserve.com> <37845270.BD3E08DD@starship.python.net> <#dvnA1Uy#GA.131@nih2naaf.prod2.compuserve.com>
Message-ID: <3784E2E7.DD86D47B@starship.python.net>

Dave Borgelt wrote:
> 
> Sorry to be so slow here but how would I be 'in' that directory.  I added it
> to my PythonPath so that the import works and the traceback shows that the
> system knows about looking in that directory
> (c:\programming\python\HTMLGen\filename.py).
> 
> What else would I need to do to get this to work?  To be 'in' that
> directory?
> 
> Dave Borgelt
> 
> Berthold H?llmann <bhoel at server.python.net> wrote in message
> news:37845270.BD3E08DD at starship.python.net...
> > Dave Borgelt wrote:
> > >
> > > I am having a problem with HTMLGen that is driving me nuts.  It has to
> do
> > > with running the test code in HTMLtest.  (HTMLtest.test()).
> > >
> > > I expect the solution is simple but I can't see it.  The path to the
> HTMLGen
> > > files is in my PythonPath (hence the import works) but as you can see
> the
> > > test fails.
> > >
> > > -----------------------------
> > >
> > > >>> import HTMLtest
> > > >>> HTMLtest.test()
> > > Traceback (innermost last):
> > >   File "<interactive input>", line 1, in ?
> > >   File "c:\programming\python\HTMLGen\HTMLtest.py", line 26, in test
> > >     'overview.html')      # manual top
> > >   File "c:\programming\python\HTMLGen\HTMLtest.py", line 35, in overview
> > >     doc = SeriesDocument('HTMLgen.rc')
> > >   File "c:\programming\python\HTMLGen\HTMLgen.py", line 256, in __init__
> > >     if resource: execfile(mpath(resource), self.__dict__)
> > > IOError: [Errno 2] No such file or directory
> > > >>>
> > >
> > > ------------------------------
> > >
> > > Any help will be appreciated.
> > >
> > > Dave Borgelt
> >
> > This only works in the directory where HTMLgen.rc (and probably other
> > files) lie. This is usually the directory generated when unpacking
> > HTMLge.
> >
> > Greetings
> >
> > Berthold

try:

import os
os.chdir("c:\programming\python\HTMLGen\")
import HTMLtest
HTMLtest.test()

Good luck

Berthold
-- 
bhoel at starship.python.net / http://starship.python.net/crew/bhoel/
        It is unlawful to use this email address for unsolicited ads
        (USC Title 47 Sec.227). I will assess a US$500 charge for
        reviewing and deleting each unsolicited ad.



From graham at sloth.math.uga.edu  Sat Jul 17 12:09:19 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 17 Jul 1999 16:09:19 GMT
Subject: Stackless & String-processing
References: <000801becff3$3fe92880$91a22299@tim>
Message-ID: <7mq9rf$r27$1@cronkite.cc.uga.edu>

Tim Peters (tim_one at email.msn.com) wrote:
: A nice paper on Clean parsing combinators can be downloaded as item 5 
: > 
: Interestingly enough, the next paper on the list ("An Interpreter") declines
: to use  combinators for its parsing task, instead building a conventional
: "by hand" recursive descent parser.

Yes it's kind of strange that they don't use the combinators to build
the interpreter. One of Peyton-Jones' papers has a good (but small) 
example of making an interpreter with combinators (it's probably the
first paper of his on monads). 

Tim Peters (tim_one at email.msn.com) wrote:
: another-two-years-and-they'll-discover-lex/yacc<wink>-ly y'rs  - tim

I'll take that as a BIG <wink>. Once you use these combinators you will
never go back to lex/yacc (unless you want LALR(n) grammars).

graham
-- 
            The girls is all salty, the boys is all sweet, 
                      the food ain't too shabby 
                    an' they piss in the streets
    In France, way down in France, way on down, way on down in France



From AAndersen at ACM.Org  Thu Jul  1 07:05:28 1999
From: AAndersen at ACM.Org (Anders Andersen)
Date: 01 Jul 1999 13:05:28 +0200
Subject: Reflection in Python?
References: <7ldglr$ita$1@nnrp1.deja.com>
Message-ID: <nxlnd0y6iv.fsf@raillkatt.itek.norut.no>

> I've just started programming in Python, and I could really use some
> help figuring out something.  I'd like to implement Kent Beck's unit
> tester in Python (someone posted a very nice alternative a few weeks
> ago, but Junit's approach works better for my project's needs).  How
> do I do the equivalent of what Java calls reflection?

I am not familiar with Kent Beck's unit tester, but I did write a note
a while ago about Python and reflection.  This work has now matured a
little bit more, and is a part of my Dr. thesis that will be submitted
soon.  Anyway, the note is available at:

  http://www.itek.norut.no/~anders/abstract/andersen98a.html

> Thanks, Anders Schneiderman 

Anders,

-- 
Anders Andersen - <mailto:aa at computer.org> - <http://w3.to/andersen/>



From stidolph at origin.ea.com  Fri Jul 30 13:46:44 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Fri, 30 Jul 1999 17:46:44 GMT
Subject: Recursive constructors?
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC70@molach.origin.ea.com>

Of course it is ok to call recursively.  self.bar will contain a different
instance of class foo.  Never had to write this, but shouldn't be any kind
of problem.  Make sure you have initialized the class first of course.

-----Original Message-----
From: roy at endeavor.med.nyu.edu [mailto:roy at endeavor.med.nyu.edu]
Sent: Friday, July 30, 1999 12:18 PM
To: python-list at cwi.nl
Subject: Recursive constructors?


Is it legal for a constructor function to call itself recursively?  I.e.:

class foo:
   def __init__(self, x):
      if (whatever):
         self.bar = foo(y)

-- 
Roy Smith <roy at popmail.med.nyu.edu>
New York University School of Medicine




From tim_one at email.msn.com  Mon Jul 19 22:28:14 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 19 Jul 1999 22:28:14 -0400
Subject: Cyclops 0.9.4
Message-ID: <000c01bed257$84f52a80$139e2299@tim>

    http://www.python.org/ftp/python/contrib/System/Cyclops.py

# Module Cyclops version 0.9.4.
# Released to the public domain 18-Jul-1999,
# by Tim Peters (tim_one at email.msn.com).

Cyclops.py implements a CycleTracker class that provides major help in
finding and analyzing runtime cycles in Python programs.

Strong points:

+ Much faster than previous similar modules.  Searching for cycles
  takes time linear in the number of objects reachable from the
  registered objects plus the number of arcs connecting them, and all
  computation needed to display cycles is delayed until it's really
  needed (if ever). Tens of thousands of objects and hundreds of
  thousands of arcs can be analyzed in less than a minute on my creaky
  old P5-166 system.

+ Many kinds of optional output reports, from a simple listing of
  objects found in cycles, to a partitioning of cyclic objects into
  maximal strongly-connected components.

+ Easy to add new types to the set of objects CycleTracker knows how
  to "chase":  pass appropriate functions to a CycleTracker instance's
  chase_type() method.  It's done this way instead of via subclassing
  for speed.

+ An optional cycle-filter callback can be registered to ignore expected
  cycles (typically those created by the Python implementation itself).

Weak points:

+ The problems this module helps to address are inherently difficult, and
  CycleTracker doesn't make them any easier to understand or to fix -- it
  only helps identify what and where they are.  It's a tool, not a
  solution.

+ All the optional abilities make for a steep learning curve.

+ Registering objects can distort the behavior of the program under
  investigation, by keeping objects alive that would otherwise have died.
  See the module docstring for discussion and workarounds.

<P><A HREF="http://www.python.org/ftp/python/contrib/System/Cyclops.py">
Cyclops.py 0.9.4</A> - tool for tracking cyclic structures in Python
programs.  (18-Jul-99)





From MHammond at skippinet.com.au  Fri Jul  9 08:40:47 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 9 Jul 1999 22:40:47 +1000
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com> <37850F5B.F80CB0A7@enews.nrl.navy.mil> <7m3ajc$s3q$1@m2.c2.telstra-mm.net.au> <37859ef4.11675718@news.ricochet.net>
Message-ID: <7m9tm2$6qg$16@rks1.urz.tu-dresden.de>

phil wrote in message <37859ef4.11675718 at news.ricochet.net>...
>On Fri, 9 Jul 1999 09:03:03 +1000, "Mark Hammond"
><MHammond at skippinet.com.au> wrote:
>
>available but not the _d.dl. Couldn't there be two downloads, one with
>the optimised/non-debug and one without? I rather like the

There could of course - just a matter of "who"?  Guido does a fanstasic job
of putting together the Python binary set for Windows, even though Windows
is not his primary platform.

It is just asking too much for the existing people to do this for nothing.
This is the reason I am adding it as a registered user benefit - it is
simply too much of a burden to go _this_ far for free!

Mark.






From aahz at netcom.com  Mon Jul 19 17:59:05 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 19 Jul 1999 21:59:05 GMT
Subject: SMTP problem
References: <hrMk3.1158$Gc6.192154624@newsb.telia.net>
Message-ID: <7n0739$npi@dfw-ixnews5.ix.netcom.com>

In article <hrMk3.1158$Gc6.192154624 at newsb.telia.net>,
Joacim L?wgren <fire at netlimit.com> wrote:
>
>I copied this straight from the Python SMTP example, however it doesn't seem
>it's working. I'm not quite sure what ^D does and if it is different on
>Unix/Windows as I think this is the problem with the script.

Yes, on Windoze you want ^Z instead.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From fredrik at pythonware.com  Mon Jul 26 11:23:32 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 15:23:32 GMT
Subject: createfilehandler, how to use it with Tk?
References: <7nht1s$grn$1@nnrp1.deja.com>
Message-ID: <012c01bed77a$d562b370$f29b12c2@secret.pythonware.com>

algaba at my-deja.com wrote:
> I have to communicate a tk front end with another modules.
> And I'm trying:
> I've been browsing  fnorb example of use of createfilehandler
> but it's as complex as hell.
> has anyone a simple,straightforward example of it?

forget about createfilehandler. use asyncore/asynchat instead:
http://www.python.org/doc/current/lib/module-asyncore.html

to use them under Tk, you can use a
simple polling scheme:

import asyncore

class myApp:
    def __init__(self, master):
        self.master = master

        ...

        self.poll()

        # issue a HTTP request (see handbook example)
        http_client("www.python.org", "/")

    def poll(self):
        # keep the server up and running
        try:
            asyncore.poll(0.05)
            self.master.after(100, self.poll)
        except:
            pass # error in network callback

...

</F>




From Nikolas.Kauer at p98.f112.n480.z2.fidonet.org  Fri Jul  2 01:53:43 1999
From: Nikolas.Kauer at p98.f112.n480.z2.fidonet.org (Nikolas Kauer)
Date: Fri, 02 Jul 1999 06:53:43 +0100
Subject: execfile(), import and namespaces
Message-ID: <00001678@bossar.com.pl>

From: "Nikolas Kauer" <nkauer at students.wisc.edu>

We use a modification of "Listing One" of Andrew Kuchling's "A CGI Framework
in Python"
article in Web Techniques (February 1998) for our dynamic Web site.  We just
made our
Web site public and wanted to switch from files with Python source (under
/htdocs) to compiled files with Python object code (i.e. python -c 'import
file.py' -> file.pyc) as suggested in CGI FAQ 1.8, mainly to make it harder
to (download and) re-use our code.  The object files would be loaded as
modules with 'import'.

Problem: In the following code snipet environment variables are passed to
the
executed Python file via the corresponding global namespace argument of the
execfile() function; whereas the 'import' statement does not allow for such
arguments.

namespace = {
             'headers': headers,
             'webvars': webvars,
             'environ': os.environ
            }

[ more code ]

try:
    execfile(filename, namespace)
except:
    # The page's code raised some sort of exception.

Does anyone know of an elegant way to pass namespace objects to modules in a
similar
way execfile() does?

Thanks,
Nikolas

PS I'm not on the mailing list, please reply directly.






From mstenber at cc.Helsinki.FI  Tue Jul 13 13:42:43 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 13 Jul 1999 20:42:43 +0300
Subject: PIL+Tkinter query (I must be doing something _really_ stupid)
Message-ID: <al8673oe99o.fsf@myntti.helsinki.fi>

Ok, even with the topic, this is a warning: total clueness incoming ;)
>From Markus's adventures in PIL+Tkinter-land..

	from Tkinter import *
	tk = Tk()
	Label(tk, text='foo').pack()
	tk.mainloop()

This works. Ok. (me, a Tcl/Tk guru, snicker)

Then, we add:
	import ImageTK
and change Label to have
	Label(tk, image=ImageTk.PhotoImage(image)).pack()

The image is fine; image.save() produces the required output; it's RGB
image with some map data, to be precise. HOWEVER: The label has correct
size (size of the image) but no image data whatsoever. The same problem
applies to all approaches I've tried (using Canvas and putting it on one,
etc). 

What am I doing incredibly stupidly now? (It _must_ be something trivial,
but..). The s/w used is PIL-1.0b1 with Python 1.5.2.

-Markus

-- 
"...very few phenomena can pull someone out of Deep Hack Mode, with two
noted exceptions: being struck by lightning, or worse, your *computer*
being struck by lightning."
	- Matt Welsh



From wlfraed at ix.netcom.com  Wed Jul 21 23:27:14 1999
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Thu, 22 Jul 1999 03:27:14 GMT
Subject: gadfly RDBMS ???
References: <Pine.LNX.4.04.9907211629470.12211-100000@emerald.netskate.ru> <3795C21F.A0F13648@onera.fr> <3795EA7F.4A452490@arcus.tel.hr>
Message-ID: <37968aa6.3558767@nntp.ix.netcom.com>

On Wed, 21 Jul 1999 17:42:55 +0200, Darko Cvagic
<darko.cvagic at arcus.tel.hr> declaimed the following in comp.lang.python:

> 
> 
> Thanks, I think that I'm exchange terms like RELATION and LINK
> So, Gadfly IS RELATIONAL, but I can't make something like:
> 	CREATE LINK order_customer
> 	FOREIGN KEY order(cno)
> 	REFERENCES customer
> 	ON DELETE RESTRICT
> 
> I can't do it in gadfly, am I wright?  
> What is name for construction like this? 
> I was thought that it is RELATIONAL , I was wrong!

	That sounds like a definition of referential integrity
constraints... I've not looked at gadfly -- perhaps it doesn't actually
maintain such internally, but only via SQL select statements, which
would require one to code the integrity tests themselves...


	It is confusing in a way -- based upon some readings of my old
DB textbooks, "relation" in "relational" doesn't really apply to
linkages between tables. The "relation" is that the key value /in/ the
table uniquely identifies the "related" data values, and that each
"record" (tuple in the texts) of a table is identical in format to every
other in that table.

	Relational algebra and calculus are methods of selecting data
from multiple tables to produce a result -- but again, these do not
enforce referential integrity. In truth, "relational" is a view of the
data, which may be stored in any format -- even in hierarchical or
network format.

	I think I begin to ramble...
 
--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <



From gmcm at hypernet.com  Wed Jul  7 09:38:04 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 7 Jul 1999 08:38:04 -0500
Subject: Multiple program defined types -- access from C?
In-Reply-To: <m3908tp99t.fsf@froydnj.laptop.rose-hulman.edu>
Message-ID: <1280771407-69691738@hypernet.com>

Nathan Froyd wrote:
> I'm writing a program that needs a lot of program-defined types that
> are also exposed as python classes to the user. The problem is, I
> also need to access the innards of a lot of these classes from C.
> Are there any elegant ways to do this? It doesn't seem as though
> Python likes for class functions to be used from outside the C file
> they're defined in.

Python asks you to declare almost everything as file static to reduce 
the chance of name clashes when loaded dynamically in Unix (or maybe 
I should say non-Windows, where this is not a problem).

The "proper" way to make a C API available to other modules is to use 
a CObject (see Extending / Embedding section 1.12) to pass a void * 
around. If you're not trying to provide a C API to other extensions, 
you can probably get away with losing the "static", as long as you're 
careful about how you name your functions.

- Gordon



From chad at vision.arc.nasa.gov  Fri Jul 23 03:22:36 1999
From: chad at vision.arc.nasa.gov (Chad Netzer)
Date: Fri, 23 Jul 1999 00:22:36 -0700
Subject: Q: anyone running Python on OS X server?
References: <004401bed44f$1a5c6400$f29b12c2@secret.pythonware.com> <37979524.C9F4F6BC@freenet.co.uk>
Message-ID: <3798183C.D216CCEF@vision.arc.nasa.gov>

sreekant kodela wrote:

> Do you mean Linux / Xfree ?
> If that is what you asked then answer is Yes.

I suspect he meant Apple's MacOS X server (where `X'
means `10' in this case), not `X server' as in the X Windows
system...

Chad






From kistler at fnmail.com  Thu Jul  8 04:36:50 1999
From: kistler at fnmail.com (Per Kistler)
Date: Thu, 08 Jul 1999 10:36:50 +0200
Subject: PIL anim gif
References: <378364C4.E18D7C1E@fnmail.com> <008e01bec919$2e919210$f29b12c2@secret.pythonware.com>
Message-ID: <37846322.ED3ED756@fnmail.com>

Hi Fredrik

Thanks a lot!!!!!!!!!!!

Cheers, Per.

Fredrik Lundh wrote:
> 
> Per Kistler <kistler at fnmail.com> wrote:
> > How to make an animated gif from several Image
> > objects in PIL? According to the docu, it seems,
> > that it can only anim gifs which are such already.
> 
> see the Scripts/gifmaker.py sample script.
> 
> </F>

-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net
------------------------------------------------------------



From tim_one at email.msn.com  Wed Jul 21 00:57:32 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 21 Jul 1999 00:57:32 -0400
Subject: Stackless & String-processing
In-Reply-To: <7n27r8$mm4$1@cronkite.cc.uga.edu>
Message-ID: <000701bed335$8a812740$a72d2399@tim>

[Fernando Pereira]
> ...
> But from my readings (those mentioned in the thread and others) as well
> as your comments, the whole point of monadic combinators for parsing is
> to interweave the grammatical structure with other computations,

That was at least part of the point, the rest being the perceived advantages
of using the same notation (language) for both parsing and semantic
processing.  Based on your work with weighted automata, I'd guess you'd be
the last person on earth to dismiss the advantages of a unifying framework
<wink>.

> thus making the underlying grammatical structure less accessible to
> optimization.

Whereas your approaches have factored things in such a way as to make them
more accesible to optimization.

If it isn't clear yet <wink>, efficiency isn't driving Graham here.  He's
found a cool new approach, and is having fun with it.

Different strokes.  Most of my parsing life has been concerned with
artificial languages, and there I've got a different complaint about
intermixing syntax with semantics:  the grammar is interesting for its own
sake, and it's a positive burden at times to clutter it up with semantic
annotations!

Case in point:  Python's grammar is formally defined by the file
Grammar/Grammar in the source distribution.  That's the input to a parser
generator.  At 60 non-noise lines, it's quite possible to digest the whole
thing, and minor changes in syntax can be effected without touching anything
else (how minor? exactly what doesn't require changing anything else
<wink>).

This is a joy as it is; even extending it to a conventional attributed
grammar would get in the way of the purposes to which the grammar file is
*usually* put.

I haven't tried using parser combinators, but in reading the papers the
question I had over and over is whether they scale well in practice, whether
measured by speed or maintainability (they eventually answered the "speed"
one: no).  It didn't bode well that each new example seemed to require
creating a new combinator, and then you've got the implementation of *that*
to worry about on top of the target grammar and the semantics.  It seemed to
carry the flavor of programming in an assembly language for "a real parsing
engine" (as Icon often feels to me in contrast to SNOBOL4 as grammars get
large).  "You can have any sequencing you want, but you have to build it
yourself."

> As I suggested above, it would be interesting to define classes of
> monadic programs that still allow (an appropriate extension of) those
> optimizations. But AFAIK that has not been done.

It would.  Not for me, though <wink>.

lex/yacc-looking-better-all-the-time-ly y'rs  - tim





From aaron_watters at my-deja.com  Wed Jul 21 11:33:18 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Wed, 21 Jul 1999 15:33:18 GMT
Subject: gadfly RDBMS ???
References: <Pine.LNX.4.04.9907211629470.12211-100000@emerald.netskate.ru> <3795C21F.A0F13648@onera.fr>
Message-ID: <7n4p7u$v89$1@nnrp1.deja.com>

In article <3795C21F.A0F13648 at onera.fr>,
  poinot at onera.fr wrote:

> SQL is a query langage, OSQL is an extension to OODBMS (Object
oriented).
>
> The "relationnal" model is an old model of data [1] but it as a
theorical
> background
> with set theory, and it is widely spread in the industry.

Yes and the primary difference between an RDBMS and an OODBMS is
that if something doesn't work in an RDBMS it's a bug, whereas
if something doesn't work in an OODBMS it's a feature, in fact
an indispensible and highly useful feature of great importance ;).

[note smiley]

  -- Aaron Watters http://www.chordate.com

====

http://starship.skyport.net/crew/aaron_watters/bplustree/term.cgi?
2393368


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From wtanksle at dolphin.openprojects.net  Thu Jul 29 22:40:42 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Fri, 30 Jul 1999 02:40:42 GMT
Subject: 1.5.2 for: else:
References: <1278928389-16733806@hypernet.com>
Message-ID: <slrn7q2459.6nk.wtanksle@dolphin.openprojects.net>

On Wed, 28 Jul 1999 16:38:24 -0500, Gordon McMillan wrote:
>William Tanksley wrote:
>> On Tue, 27 Jul 1999 21:43:00 -0500, Gordon McMillan wrote:

>[snip Billy's silly error and his entirely insufficient embarassment 
><wink>]

:)

>> However, you're wrong about the best way to detect a 2 in the list
>> using this style.  To steal the same example:

>> try:
>>  for x in [1,2,3]:
>>   print x
>>   if x == 2: raise neverMind
>>  print "modern else clause here"
>> except neverMind: print "found two"
>> else:
>>  print "another place to put the modern else clause"

>> Of course, by demonstrating both ways to do it, I'm making the code
>> more complex; but you can see that this solution solves the problem
>> quite elegantly.

>WIth an extra class, and 3 or 4 more lines of code, not to mention 
>treading on the toes of all who think exceptions should only be used 
>for exceptional conditions (or maybe it _is_ exceptional when 
>something you wrote works <1e3 wink>).

You don't need the extra class -- you can raise a string if somebody's
linecounting you.  But who really cares?  The result is pretty darn clear.
There are no common misconceptions to fight when you're writing or reading
it (I count three people who have posted to say nothing more than "gee,
that wasn't what I expected!" about the current behavior).  This code just
works!

And as for exceptions -- anyone who uses break together with else-loop but
shuns exceptions is in serious need of psychological help.  That's WAY
tweaked.  Exceptions are a wonderful tool; break is a useful hack.

Oh, and if you REALLY like the current behavior of else-loop, you can
still have it if else-loop were gone -- like this:

try:
 for x in list:
   if iffers(x): raise "no!"
   yadda(x)
except "no!": pass
else:
 twiddle()

That's two extra lines, and one level of nesting -- for the _exact_ same
functionality, only much more likely to be comprehended.

>> The place to use MY definition of 'else', OTOH, is when you're
>> building a structure during the iteration, but the result of an
>> empty loop is different than the initial value the variable has to
>> have when going through the loop.

>> Here you go.  This example might convert a Python list into a Lisp
>> tree-list -- but to imitate Lisp (okay, it's a toy problem, I'm
>> cheating) an empty list produces NIL ~= None.

>> var = Node(None,None)
>> for x in pyList:
>>   var.car = x
>>   var.cdr = Node()
>> else:
>>   var = None

>Not a Lisper, but isn't what you're doing better done thusly:

No no no!  Don't be silly -- I was giving an example of how my idea of
else should work.  I don't have time to design a system which NEEDS to be
done this way.

but-I-do-have-time-to-whine-about-something-that-isn't-going-to-change-ly
yr's,

-- 
-William "Billy" Tanksley



From vysteric at my-deja.com  Fri Jul 30 15:13:26 1999
From: vysteric at my-deja.com (vysteric at my-deja.com)
Date: Fri, 30 Jul 1999 19:13:26 GMT
Subject: embedding
Message-ID: <7nstgb$st3$1@nnrp1.deja.com>

I've been reading through my docs for hours, and I can't seem to find an
answer to this.. How do you create an object in the main namespace and
how do you get the value of an object in the main namespace? Without
using PyRun*

thanks

Patrick


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From ivanlan at callware.com  Mon Jul 19 15:35:17 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 19 Jul 1999 19:35:17 GMT
Subject: How fast can we multiply?
References: <000201bed181$a5bf0be0$4ba02299@tim> <37936AE4.810F2152@appliedbiometrics.com>
Message-ID: <37937DF5.519CF1AA@callware.com>

Hi All--

Christian Tismer wrote:
> 
> Tim Peters wrote:
>

[bobbit]

> At the same time, the reacton of the list was a little
> disappointing to me. There seems to be near to zero interest.
> Is this the hot summer? Holidays? Bad title of my post?
> 

Oh, bad title, most certainly!  ;-)  *I'm* interested!  I don't
understand most of what you guys are talking about, but I can see that
if I could afford to spend the time studying it I would be rewarded.

On topic, though, doesn't Knuth include in his "How Fast can We
Multiply?" section a discussion of using modular arithmetic to speed up
multiplies by a pretty large factor?  Now, maybe I've misunderstood the
discussion so far so badly that I missed your use of it, but I don't
recall seeing anything about it.

> Then I also had a look into NumPy and its matrix multiplication.
> I stunned. Again no attempt to optimize big O. It could be done
> with Python, the algorithms are known, the fruit is waiting
> to be picked.
> 
> Challenge: I claim Python could be a competitor to Mathematica,
> if only some people were interested.
> 

I would be interested--just not competent!  But a Python version of
Mathematica could indeed be a killer app.  (Or is that a killer rabbit?)

<guess-who-flunked-high-school-algebra>-ly y'rs,
Ivan 
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From kernr at mail.ncifcrf.gov  Wed Jul  7 18:58:50 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Wed, 07 Jul 1999 22:58:50 GMT
Subject: make or similar in Python
References: <378390DD.58CA6BD5@nevex.com>
Message-ID: <3783db76.118974654@news.erols.com>

On Wed, 07 Jul 1999 13:39:42 -0400, Greg Wilson <gvwilson at nevex.com>
wrote:

>Hi, folks.  Has anyone ever written a dependency management/update tool
>(like 'make') in Python?
>I'm not looking for something that reads makefiles, and happens to be
>written in Python; I'm looking
>for a Python module that users import, then provide data to, in order to
>manage dependencies.  (The
>answer I got from the Perl crowd was 'cool idea, but no'; I'm hoping
>I'll have more luck here.)
>Thanks,
>Greg (who'd really, really like to override some of the things that
>GNU make does)

Check out
ftp://ftp.python.org/pub/python/contrib/System/pymake-0.1-sb.tar.gz
You will probably have to hack it up a bit to do exactly what you
want, but it's probably a really good start.

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From michael.jungmann at online.de  Fri Jul 16 16:19:47 1999
From: michael.jungmann at online.de (meikel)
Date: Fri, 16 Jul 1999 22:19:47 +0200
Subject: Calling C code from Python
Message-ID: <378F93E3.C4F37BED@online.de>

Hi,

I would like to call a C-function from Python. The function is not
located within a DLL but in a running application. That would probably
mean that I would have to embed Python into the C-application, make
the function available to Python (in a magic way?) and then execute a
script (which has to be a text file on disk) which can call the
mentioned C-function. How can I do something like that?

What is this all about? I have to provide a textual script which can
be called from an running C-application. In order to do the job the
script has to know certain values from the application. Some of the
information can not be computed before the script is called but the
script has to call functions within the application.

I hope my question is not too dumb and someone can help me (especially
for the 'magic' part :o).


Greetings,

Michael & Udo



From ivnowa at hvision.nl  Thu Jul  8 01:36:41 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Thu, 8 Jul 1999 05:36:41 GMT
Subject: Error Trapping In Python
In-Reply-To: <7m0pgt$cij$1@eskinews.eskimo.com>
References: <7m0pgt$cij$1@eskinews.eskimo.com>
Message-ID: <199907080535.HAA19715@axil.hvision.nl>

On 7 Jul 99, Jonathon wrote:

> 
>  The script I am writing expects the user to input an
>  integer.   Being on the safe side, it reads the raw
>  data as a string.   However, to convert that string
>  to an integer, is perplexing me.
> 
>  string.atoi.(read_data) will give a value error if
>  read_data contains something other than a number.
> 
>  int ( read_data ) gives a Value Error, if non numerical
>  data is included.  <<  Or even if 10.0 is written instead
>  of 10  >>

You can trap such errors with exceptions... A simple example:

try:
    i = int(mystring)
except ValueError:
    print 'Cannot be converted'

There are others ways, of course. For instance, you could scan that 
string and filter out any unwanted characters (other than 0-9, 
probably). Or in the case of 10.0, you could try to convert to a 
float and if that succeeds, convert to an int, which causes any 
decimals to be chopped off.

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
Python Snippets: http://www.hvision.nl/~ivnowa/snippets/
The Purple Kookaburra Forum: http://www.delphi.com/kookaburra/



From pmagwene at midway.uchicago.edu  Fri Jul 23 12:58:35 1999
From: pmagwene at midway.uchicago.edu (Paul Magwene)
Date: Fri, 23 Jul 1999 16:58:35 GMT
Subject: how to access ports ?
References: <Pine.LNX.4.04.9907231714040.30978-100000@emerald.netskate.ru>
Message-ID: <37989F3B.726AE879@midway.uchicago.edu>

Oleg Broytmann wrote:
> 
> On 23 Jul 1999 r.skowronek at fz-juelich.de wrote:
> > Well, how do I access arbitrary ports?
> > I managed to read and write to serial ports easy, but I would like to access
> > an mca card at addresses 0x220 - 0x31f.
> > I checked the docs but did not see any reference to hardware access - or did
> > I not look in the right places?
> 
>    Hardware access is too hardware/software/OS dependent, generic langauage
> like Python cannot handle it in portable way.
>    If you can program in C, and if you agree to learn "Python Extending and
> Embedding" - write extension module for your particular task in C.
> 


That said, if you're using Windows check out the following qoute from
http://www.python.org/windows/

"(Many people find the Serial-IO module for Python under windows to be
very useful -
you can find it in the System utilities contrib dir, as sio-151.zip, or
the older
siomodule.zip for Python 1.4.)"

You may have to write your own extension for other platforms (though I
would be surprised if no one else has tried to "scratch this itch"
before).

--Paul



From amaranda at nospam.com  Fri Jul  2 06:28:38 1999
From: amaranda at nospam.com (Alex Maranda)
Date: Fri, 02 Jul 1999 11:28:38 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <377a6bb0.10820509@news.oh.verio.com> <377BDEF7.C34A6088@compaq.com> <maedel-0107991843320001@209-122-226-23.s277.tnt1.nyw.ny.dialup.rcn.com>
Message-ID: <377C9456.62FE@nospam.com>

Giambattista Vico wrote:
> 
> I'd like to hear what people have to say in this thread if we consider
> freezing or otherwise creating a standalone app, even if only for Unix and
> Win. Has anyone found a Python tool or combination of tools that solves
> this problem?
> 
> --tom
Read my previous post on this thread.

Cheers,
-- 
Alex Maranda         mailto: amaranda at spider dot com
Spider Software Ltd. Tel: +44 (0)131 4757036
Edinburgh, UK        http://members.xoom.com/Alex_Maranda
STREAMS based communications protocols for embedded systems



From ivnowa at hvision.nl  Wed Jul  7 08:59:27 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Wed, 7 Jul 1999 12:59:27 GMT
Subject: interaction between modules... globals?
In-Reply-To: <19990707100532.A951@optiplex.palga.uucp>
References: <19990707100532.A951@optiplex.palga.uucp>
Message-ID: <199907071258.OAA25940@axil.hvision.nl>

On 7 Jul 99, Gerrit Holl wrote:

> Hello,
> 
> I have a program which consists of several little modules, but I encounter a
> problem with that. I have one module which returns a class instance, with
> messages as members, either Dutch or English. Every module needs that instance,
> but I don't want to pass it as an argument to every little function, the
> modules should find out theirselves. But I _also_ don't want to make that
> instance 5 times.
> 
> So I thougt: I need a sort of global between the modules. Is this possible?

Why not use it like this:

[module mymodule.py]
class MyClass:
    # blah blah
myinstance = MyClass()

Now we have the module with your much-used instance. This can be used 
by other modules as follows:

[module a.py]
import mymodule

mymodule.myinstance.do_something(x,y)

But maybe you tried this approach already?

Hans Nowak (ivnowa at hvision.nl)
Homepage: http://fly.to/zephyrfalcon



From tim_one at email.msn.com  Sat Jul 10 23:50:42 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 10 Jul 1999 23:50:42 -0400
Subject: Another reason assignment expressions would be nice (?)
In-Reply-To: <3787c452.1487712890@news.slip.net>
Message-ID: <000301becb50$8c2045a0$479e2299@tim>

[alan at littleford.net]
> ...
> Now here is the rub, I might want two lines generated with a related
> value. Lets assume SomeFunction() throws a random number, so I can't
> just call it again, so what I would like to do is this:
>
> 	a = foo()
> 	a.setTemplate("'A test line %d' %
> 		(self.i = self.SomeFunction()"))
> 	a.setTemplate("'Same value as above %d' % self.i")
> 	...
> 	a.run(100)
>
> But of course I can't do this because '=' is a statement in the first
> template.
>
> [and skipping insane alternatives]
>
> ...
> All in all it seems I have to jump through hoops because i = 3
> doesn't 'return' the value 3.
>
> Or am I missing something frightfully easy here ..

Writing code to generate code with magical side-effect instance attr binding
via magical side-effect embedded assignments ... oh ya, sign me up for two
of those <wink>.

How about

    a = foo()
    a.setAttr('i', 'self.SomeFunction()')
    a.setTemplate("'A test line %d' % self.i")
    a.setTemplate("'Same value as above %d' % self.i")

?  Then (a) the code is easier to write; and, (b) when it all falls apart
you can at least find your mutations in an editor via searching for
"setAttr", or even turn on a debug flag causing setAttr to log what it's
doing.

different-operations-different-methods-ly y'rs  - tim





From 801d at my-deja.com  Sat Jul  3 16:12:29 1999
From: 801d at my-deja.com (Scott McClellan)
Date: Sat, 03 Jul 1999 20:12:29 GMT
Subject: calldll & PyArg_ParseTuple strings
Message-ID: <7llqr8$crd$1@nnrp1.deja.com>

Hello,

I'm attempting to use Sam Rushing's calldll module to call a C function
in a dll, with the following prototype:

long __declspec(dllexport)func(long *x, long *y, long *z)

I expect the function to give certain values back in x, y, and z, but so
far have been unable to make that happen. Let's say that I have three
long integers in Python that I initialize as follows:

x=y=z=0L

My question is, in the following call from Python,

call_foreign_function(pFunc, 'lll', 'l', (<?>,<?>,<?>))

what should the tuple (<?>,<?>,<?>) contain?  If I pass (x,y,z),
the Python interpreter (Idle in my case) crashes, and this is
understandable, since it results in the contents of x, y, and z being
used as addresses to write to.  I've tried (id(x),id(y),id(z)), since
id() is supposed to return the address of a Python object - Python
didn't crash as before, but nothing was copied into x, y, and z, either.

Can anyone point me in the right direction?

Thanks,
Scott McClellan


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From hochberg at psn.net  Fri Jul 23 10:35:29 1999
From: hochberg at psn.net (Tim Hochberg)
Date: Fri, 23 Jul 1999 08:35:29 -0600
Subject: Jpython Question
References: <Pine.GSO.4.05.9907191525580.20433-100000@allegro> <wkr9m3846k.fsf@ifi.uio.no> <Pine.GSO.4.05.9907231035200.14333-100000@allegro>
Message-ID: <L10m3.2062$J3.58215@news15.ispnews.com>

Ian Kezsbom wrote in message ...
>Just out of curiosity...lets say I've written a Java program, with many
>classes and I want to run that program through JPython and have access to
>all its classes...how would I go about doing this...


Its almost too easy to describe. I really recomend you just try it out.
However, let's say you have a java class Triangle in the default java
package. After compiling Triangle with your favorite java compiler (e.g.,
javac) you can import the class directly.

# For example
import Triangle
triangle = Triangle()
print  = triangle.area()

More likely, you'd have your java code in a package. If you're java class
triangle is in the package shapes you'd do something like:

from shapes import Triangle
# ....

Note that the java class files must be on your classpath. If the class files
are the current directory, this works automatically. Otherwise you're going
to have to modify the classpath in the jpython launcher (called either
"jpython" or "jpython.bat").

-tim







From miller at nialas.npl.uiuc.edu  Tue Jul  6 16:22:04 1999
From: miller at nialas.npl.uiuc.edu (M.A.Miller)
Date: 06 Jul 1999 15:22:04 -0500
Subject: total idiot question: +=, .=, etc...
References: <000015b4@bossar.com.pl>
Message-ID: <76u2rha5qr.fsf@four.npl.uiuc.edu>

>>>>> "Reimer" == Reimer Behrends <Reimer.Behrends at p98.f112.n480.z2.fidonet.org> writes:

    > And personally, I'd rather live with the inconvenience of
    > the explicit self rather than having Python's design
    > obfuscated by some strange syntax; syntax highlighting can
    > go a long way towards making the current state of affairs
    > more manageable.

Does anyone have a tweak for python-mode.el that will highlight
occurrences of self and self.?

Mike



From wtanksle at dolphin.openprojects.net  Fri Jul 30 21:37:58 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Sat, 31 Jul 1999 01:37:58 GMT
Subject: Include statements with python scripting
References: <7nsu3t$kui$1@nw001t.infi.net> <slrn7q44ck.8kv.wtanksle@dolphin.openprojects.net> <7nt4pr$qin$1@nw001t.infi.net>
Message-ID: <slrn7q4krm.95d.wtanksle@dolphin.openprojects.net>

On Fri, 30 Jul 1999 17:18:45 -0400, Sean Robertson wrote:
>William Tanksley <wtanksle at dolphin.openprojects.net> wrote in message

>> On Fri, 30 Jul 1999 15:24:38 -0400, Sean Robertson wrote:
>> >How do you do an include statement in python?  I know that in Cold Fusion
>it

>> ><html><head><title><!--#var name--> Search Results</title></head>

>> This isn't Python -- I'm sure you're trying to ask about Zope.  Remarkable

>> I wasn't aware of a #include in Zope; I've been using #var.  Generally
>> speaking, it doesn't make any sense to include an entire HTML page inside
>> another, so I think you might possibly be confused.

>I was told that the program that uses this template uses python, but am
>checking out zope nonetheless.

The program that reads that template is Zope, and uses Python.
Nonetheless, that template isn't written in Python :).

>The reason I need to include and external
>html fragment is that I want the page to have two seperate interfaces
>depending on which part of the site the user is coming from.  If they are
>searching CareerConnection, the mainclass variable will be Employers and the
>results page needs to fit the interface of the CareerConnection site,
>otherwise, they should get the old default interface.  I was using
>"#include" only as a placeholder until I figured out how to do it in python.
>If you know how, please let me know.

It depends on your environment, and the Zope people will be able to help
you a lot more, but my Zope website uses things like:

<!--#var page_header-->

to handle that kind of thing.  The correct page header is automatically
inserted there depending upon where you are -- no need for ifs.  Look up
an example of 'acquisition' in the Zope docs.

Once you learn it (and it's easier than it looks), you'll never look back.

-- 
-William "Billy" Tanksley



From spamkiller at epstein.co.uk  Fri Jul 23 08:04:03 1999
From: spamkiller at epstein.co.uk (Dani Epstein)
Date: Fri, 23 Jul 1999 13:04:03 +0100
Subject: Dreaded Newbie Question
References: <K4tK+KAA26l3EwvK@epstein.co.uk>
 <000701bed4bb$20388360$2c2d2399@tim>
Message-ID: <S3i8NRAzoFm3EwEN@epstein.co.uk>

In article <000701bed4bb$20388360$2c2d2399 at tim>, Tim Peters
<tim_one at email.msn.com> writes
>First download 1.5.2 (1.5.1 is more than a year old, and in ways that count
><wink>).

Shall do.

>&-keep-a-printout-of-the-library-manual-next-to-the-toilet-ly y'rs  - tim

My throne room is already overflowing with a variety of excellent
documentation that I print out on 40 gsm paper so that it can be reused
once read. Um.

What I have found in the docs at hand (oh dear) is that there is a
definite unix bias to them, which I suppose is understandable, but makes
life terribly complicated for us Windoze fellows (and fellowettes of
course).

I presume this has changed in the latest docs.

Thanks to all for your replies. This adventure looks like its going to
mean good company as well as fun!
-- 

Dani Epstein




From tismer at appliedbiometrics.com  Thu Jul 15 11:23:58 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Thu, 15 Jul 1999 15:23:58 GMT
Subject: making a Windows NT service out of a python program
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F4B@RED-MSG-50>
Message-ID: <378DFD0E.99BE1E99@appliedbiometrics.com>


Bill Tutt wrote:
> 
> > From: Christian Tismer [mailto:tismer at appliedbiometrics.com]
> >
> >
> > This is all great stuff if you have understood it already.
> > For me, it is still unclear what makes up a service, and
> > what I need at the minimum to turn a Python program into
> > an NT service.
> >
> 
> Not knowing ANYTHING about Medusa... I'll provide another example service
> right here, and try and explain things as I go...

That's fine. Medusa is just a Python program. I just need
to run any Pprogram as a service. :-)

[great list of great dubious functions]

> Ok, so those are the pieces, but Bill, how does all of this nonsense fit
> together? Good question. :)

I couldn't have said it better. The big "fine. And now?".

[and a full example follows, ready to tweak for my needs]

> I hope this helps,
> Bill

Bill, you are great. Not only you explained all about horses, but
you also gave me a basic course on riding.

thanks a lot - chris   ...hopp hopp *ouch* hopp hopp *yeah* hopp...

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From robin at jessikat.demon.co.uk  Fri Jul 30 14:16:48 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 30 Jul 1999 19:16:48 +0100
Subject: Python (Win) Database Advice needed
References: <7ns78n$mnu@romeo.logica.co.uk>
Message-ID: <ATrxHHAQweo3EwEd@jessikat.demon.co.uk>

In article <7ns78n$mnu at romeo.logica.co.uk>, Ian Clarke
<clarkeia at logica.com> writes
>Hi, I am planning to write a python based web-database retrieval system.  I
>am not very familiar with databases in general (although I have used SQL)
>and was looking for some advice.  Basically this system will run on an NT
>server and allow selected users to view and change data over the web.  I
>need a mechanism for storing the data so that:
>1) It can be accessed by several HTTP threads running at the same time
>without risk of corruption
>2) It can be migrated to other database formats such as Microsoft Access.
>3) It is Open Source
>4) It is simple to deploy
>
>Thanks for any advice,
>
>Ian.
>
>
check out www.zope.org; they're exactly on this road if not quite there
at the moment.
-- 
Robin Becker



From chad at vision.arc.nasa.gov  Sat Jul 31 00:23:48 1999
From: chad at vision.arc.nasa.gov (Chad Netzer)
Date: Fri, 30 Jul 1999 21:23:48 -0700
Subject: Import a lot of modules, or, import 'foo'
References: <874silabe7.fsf@linuxcare.com> <1%ro3.4357$VL2.467111@news.direcpc.com> <871zdpa9lb.fsf@linuxcare.com>
Message-ID: <37A27A54.A737AE62@vision.arc.nasa.gov>

"David N. Welton" wrote:

> "Emile van Sebille" <emile at fenx.com> writes:
>
> > This seems to work for me:
> >
> > >>> for t in tests:
> >  exec ('import %s' % t)
>
> Eewwwww, that's kind of twisted, but I guess if it works:-)
>

I do a similar thing occasionally, and the big problem is that it makes
the Traceback system
work poorly, so that tracking down where unexpected exceptions have
occurred during
debugging can be a pain (especially Tkinter or Pmw generated exceptions,
etc.).

But seriously, you should use the __import__ command instead. ie.:

    string = __import__("string")

which is similar (equivalent at the module level?) to:

    import string

To import names from within the module (ala. "from spam_module import
eggs"):

    globals()[name] = getattr(__import__(module_name), name)

Chad





From Sim.&.Golda.Zacks at p98.f112.n480.z2.fidonet.org  Fri Jul  2 12:16:13 1999
From: Sim.&.Golda.Zacks at p98.f112.n480.z2.fidonet.org (Sim & Golda Zacks)
Date: Fri, 02 Jul 1999 17:16:13 +0100
Subject: Python supporting web hosting?
Message-ID: <00001699@bossar.com.pl>

From: Sim & Golda Zacks <simngolda at voyager.net>

I am looking for a web hoster that supports Python 1.5 and also has a
database with a python interface (MySQL, for example). If you know of
one with affordable pricing please let me know. Currently I'm with an
ISP with Python 1.4 and they say they'll upgrade but not for some time,
and they have MySQL but only with the PERL interface installed, and they
will think about the Python one. And I'm getting sick of a lot of other
things. I just need a good hosting site, that preferably has anonymous
reseller accounts.

Thanks
Sim



From fredrik at pythonware.com  Fri Jul  9 05:59:57 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 9 Jul 1999 09:59:57 GMT
Subject: why is there no put in urllib ?
References: <Pine.LNX.4.05.9907082250040.11616-100000@ns.ac-copy.net>
Message-ID: <7m9tlv$6qg$12@rks1.urz.tu-dresden.de>

Joachim Schmitz <perry at ac-copy.net> wrote:
> may be this is a total dump question. But it would allow us Zopistas to
> upload our dtml pages with a simple python script to the Zope-server.

so what are you waiting for?  you have the
sources, don't you?  ;-)

(also see the XML-RPC firewall thread)

on the other hand, httplib supports PUT just
fine.  as long as there's no firewall in your way,
it's not much harder to use than urllib.

</F>




From sekter at mail.matav.hu  Sat Jul 31 19:38:03 1999
From: sekter at mail.matav.hu (Arpad Kiss)
Date: Sun, 01 Aug 1999 01:38:03 +0200
Subject: os.name
References: <1278785134-25350540@hypernet.com> <V5ho3.4340$VL2.463886@news.direcpc.com> <001e01beda8e$2111d2e0$f29b12c2@secret.pythonware.com> <Ggto3.4361$VL2.467372@news.direcpc.com> <7nuac5$3rs$1@chief.death.de> <RECo3.4365$VL2.468864@news.direcpc.com>
Message-ID: <37A388DB.BFEB9A80@mail.matav.hu>

Emile van Sebille wrote:
> 
> It sure would.  Any idea how you would get the result back into python
> without having the dos command windows flash up on the screen?
On Windows NT I use win32pipe:
i,o,e=win32pipe.popen3('ver')
ret=o.readlines()
> 
> but-the-you-wouldn't-have-the-fun-of-parsing-the-file-ly yr's
> 
> --
> 
> Emile van Sebille
> emile at fenx.com
> -------------------
> 
> Felix Puetsch <puetsch at gmx.de> wrote in message
> news:7nuac5$3rs$1 at chief.death.de...
> > emile at fenx.com (Emile van Sebille) wrote in
> > <Ggto3.4361$VL2.467372 at news.direcpc.com>:
> >
> > >By the way, in windows at a command prompt:
> > >
> > >95,98:    net config
> > >nt:       net config [server|workstation]
> > >
> > >gives a line that starts : Software version
> >
> > Wouldn't "ver" be easier?
> >
> > --
> > Felix P?tsch
> 
> --
Arpad Kiss



From greg.ewing at compaq.com  Thu Jul 22 00:12:45 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 22 Jul 1999 16:12:45 +1200
Subject: IDLE mini-hack & ?
References: <199907211257.HAA14822@cro.ots.ac.cr>
Message-ID: <37969A3D.EEDC201@compaq.com>

Isidor wrote:
> 
> ##def date_time_stamp(self):
> def date_time_stamp():

I think the first version of that line was better.
If it's a method of a class, it'll need a self
argument. That would account for the "No arguments
expected" exception - it's being called with one
argument (self) but it's not expecting any.

By the way, does anyone else keep getting annoyed
by the fact that when that sort of error occurs
(i.e. a callback from Tkinter with the wrong
number of args) the traceback gives you NO
useful filename/linenumber?

Could the traceback printing thingy perhaps be
enhanced to go just a little bit further down the
chain and report the filename/linenumber of the
function being called, not just the line it was
being called from? Is the necessary information
present in the exception & traceback?

Greg



From johanw at easics.be  Fri Jul  2 11:45:13 1999
From: johanw at easics.be (Johan Wouters)
Date: Fri, 02 Jul 1999 17:45:13 +0200
Subject: Why does shelve make such large files?
References: <19990701211450.A6827@optiplex.palga.uucp> <377BF01D.16F8AA33@cup.hp.com> <19990702123152.B9975@optiplex.palga.uucp> <377CC17C.4487A36A@easics.be> <377CCF95.DECB8CF5@equi4.com>
Message-ID: <377CDE89.EDEF4BB5@easics.be>

<snip>
> Ahem, with all due respect to everyone... this is humbug.
> 

Why so?

> Pickle serializes your data in one sweep to/from disk.  It's compact.

So? I don't remember me having any trouble with that statement. As a matter
of fact this is just what I stated.


> But the last claim is horrendously misleading, I'm afraid: data storage
> is not a space/time tradeoff.  It's about throughput (I/O bottlenecks)
> and overhead in managing the supported data and indexing schemes.  There
> are order-of-magnitude performance differences in how several solutions
> work, because of this.

The general idea was some typical engineering tradeoff. Larger database
environments *might* give you better performance for some criteria like
searching or combining data. Say you have an amount of data you want to 
search. You could keep it as a serial stream (minimal space) or you could
use some clever hash method and fixed size records to gain speed. This
way you can easily bypass the overhead of accessing the whole database
(and the slow disks and, ...) by simply calculating where the data should
be. On the other hand, this structuring of the data will incur some extra
space. It looks as if this is just the space/time tradeoff I was talking about ...


> 
> Shudder.  The notion that a large database package, or a large datafile,
> is faster, is so far from reality that it has to be corrected, even in
> this Python-oriented newsgroup.  My apologies for the S/N ratio drop.

I was talking about the datafile, not the package! Also I never stated
that bigger files WILL give you better performance.

Maybe the whole story was a little simplistic, but with all respect: how
does your contribution add to the S?

So Gerrit, I hope you still got some insight from all this!

Kind regards,
Johan Wouters



From gkr at is.in-berlin.de  Tue Jul  6 04:50:31 1999
From: gkr at is.in-berlin.de (Gernot Krause)
Date: Tue, 6 Jul 1999 10:50:31 +0200
Subject: Tkinter._test() problem Tcl/Tk8.1 incompatibility python-1.5.2
References: <377F38AD.C6CDBA6D@fnmail.com> <37814F74.DE625933@tag.csiro.au>
Message-ID: <7lsf18$r0f$1@black.news.nacamar.net>

>> -----cut
>> >>python
>> Python 1.5.2 (#4, Jul  4 1999, 12:25:56)  [GCC egcs-2.91.66 19990314
>> (egcs-1.1.2  on linux2
>> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>> >>> import Tkinter
>> >>> Tkinter._test()
>> Traceback (innermost last):
>>   File "<stdin>", line 1, in ?
>>   File "/per/lib/python1.5/lib-tk/Tkinter.py", line 1947, in _test
>>     root = Tk()
>>   File "/per/lib/python1.5/lib-tk/Tkinter.py", line 886, in __init__
>>     self.tk = _tkinter.create(screenName, baseName, className)
>> TclError: Can't find a usable init.tcl in the following directories:
>>
>> This probably means that Tcl wasn't installed properly.
>> ----cut
>>
>> It seems, that python looks in some directories, which it does not
>> have. What do I have to do?


Hi,


if tk not installed in standard path (see --prefix in source code if you
have it compiled), than you must set the environment variable
TK_LIBRARY. The value contains the path, on which installed the tk files.
You can set also TK_PATH. For tcl you can set  TCL_LIBRARY,
TCL_PATH and TCL_PROGRAMS for your environment.
Please read the documentation in /usr/doc/packages/tk*
or  /usr/doc/packages/tcl*.

Gernot Krause







From da at ski.org  Wed Jul 14 14:01:46 1999
From: da at ski.org (David Ascher)
Date: Wed, 14 Jul 1999 11:01:46 -0700 (Pacific Daylight Time)
Subject: Proposed Birds of a Feather Sessions at O'Reilly Conference
Message-ID: <Pine.WNT.4.04.9907141046441.212-100000@rigoletto.ski.org>

I'd like to propose a few BOF sessions for the O'Reilly Conference
(conferences.oreilly.com), and ask that folks who think they might be
interested in attending these contact me so that we can get a feel for
which BOF clashes are most important to avoid.  If you're interested
enough that you'd consider being the 'sponsor' for a BOF, let me know as
well. (suggestions for different BOFs also heartily encouraged -- these
are off the top of my head.)

Finally, if you are the author of a popular Python package or tool and
you'll be at the conference, could you email me?  (those that are on the
program need not do that =).

Tentative BOFs:

- Python and Java: JPython and friends.

- Distributed Python & CORBA: Fnorb, ILU, XML-RPC, DCOM etc.

- How to promote Python within your company.

- Python in Science and Engineering.

- Planning for Python 2.0: "yell-at-guido time"

- Using Python in system administration ("A snake among the camels")

- Python and the web.

- Python and education / Python for the people / 
    Teaching Python /  Learning Python

- How to get rich quick by writing Python books.

- How to get rich quick by selling Python software.

- How to get rich quick, period.

--david ascher

PS: The deadline for 'early registration' at the O'Reilly conference is
    very soon! (July 15).  Register now and save money and get a t-shirt.






From dalke at bioreason.com  Thu Jul  8 21:41:49 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Thu, 08 Jul 1999 19:41:49 -0600
Subject: scoping assertion
References: <37839491.3F7C1F6@mikemccandless.com>
	 <3783C878.B8C3790B@bioreason.com> <19990708211651.A17658@optiplex.palga.uucp> <7cQP0BAYKRh3Ew6K@jessikat.demon.co.uk>
Message-ID: <3785535D.7CB0679@bioreason.com>

Robin Becker <robin at jessikat.demon.co.uk> said:
> I use
> try:
>         raise None
> except:
>         ....

Just hope you don't get a ^C inside the try-block because it
will be ignored.  (I know, the odds are low, but still present.)

How about

try:
	raise ""
except "":
	....


It's exactly the same number of keystrokes, if you count
pressing the shift key as one of them :)

						Andrew
						dalke at acm.org



From robin at jessikat.demon.co.uk  Wed Jul 21 06:20:32 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Wed, 21 Jul 1999 11:20:32 +0100
Subject: [Q] In-Browser technology
References: <932337314.611.103@news.remarQ.com>
 <BE2245BA1DCDAE31.8804258C1F60F9A7.2EB4BCD57DBA05D1@lp.airnews.net>
 <932522335.684.43@news.remarQ.com> <7n42qo$6ug$2@news.accu.uu.nl>
Message-ID: <qtU7cFAw7Zl3Ew7o@jessikat.demon.co.uk>

In article <7n42qo$6ug$2 at news.accu.uu.nl>, Martijn Faassen
<m.faassen at vet.uu.nl> writes
>[this reply only posted to comp.lang.python -- hope Vladik sees it :)]
>
>In comp.lang.python Vladik <reason at shadow.net> wrote:
>> OK, I will try to clarify my question.
>> With the warning that I may not be speaking
>> authoritatively about the web technologies yet:
>> I am just starting this research.
>
>Having followed this thread a little, I still am not sure how much 
>in browser technology the system you're describing actually _needs_. You
>can do a lot with dynamically generated HTML from the server, and some
>JavaScript.
>
>Client side action in browsers is hard to do with Python (and Perl and
>Eiffel), though I'd wish. Although I must mention jpython here, a version
>of Python that runs on top of a Java virtual machine, and can therefore
>be used in 'Java' applets (www.jpython.org).
>
>For the server side, where you'd have your business rules and database
>functions presumably, I recommend you take a look at Zope. Of course I'm 
>biased to Python, and Zope uses Python, but it's really very
>very nice. Check it out at: http://www.zope.org
>
>Browser technology is hard to do with Python (and Perl and Eiffel) -- 
>browsers don't tend to support these languages. 
>
>Regards,
>
>Martijn
>
python has with Mark Hammond's assistance got to the stage where it
works with wsh so it can also appear in ie html. It's a real pity that
no-one got around to making the equivalent of a plugin for python, but
from using the tcl plugin I know that this is tough to get right and
there are a whole load of other issues eq who manages the events,
security etc etc.

I think I detect some confusion about the actual requirements. Pure java
applets live on the server as apparently does the juice applet these
have to be downloaded to the Java/Juice/Tcl plugins.

The scriptlets JScript, VBScript, PyScript and for all I know
JuiceScript live in the HTML and also have to be downloaded (along with
the HTML). The browser farms them off to the relevant interpreter.

The ASP is done purely on the server; Zope is also a dynamic server in
this sense.


########### sample of a python script in html
<HTML>
<BODY>

<SCRIPT>
b="Hello"
</SCRIPT>

<SCRIPT LANGUAGE="Python">
import win32traceutil
print "Hello"
a="Hi there"
print "Location is", document.location
document.write("Hello", " from version ", 2, " of the Python AXScript
Engine","<P>")
document.writeln("Hello again")

</SCRIPT>

<P>The caption on the first button is set by the Window Load code.
Clicking
that button changes the text in the first edit box.

<P>The second button changes its own text when clicked.

<P>The fourth button calls a global function, defined in the global
'script' scope,
rather than the 'MyForm' scope.

<FORM NAME="MyForm" METHOD="GET">
   <SCRIPT LANGUAGE="Python">
print "Hello from in the form"
   </SCRIPT>
   <INPUT NAME="Button1" TYPE="Button" OnClick="MyForm.Text1.value='Hi'"
LANGUAGE="Python">
   <INPUT TYPE="TEXT" SIZE=25 NAME="Text1">
   <INPUT NAME="Button2" TYPE="Button" VALUE="Click for 'Hi'"
OnClick="a='Howdy'; MyForm.Button2.value='Hi'" LANGUAGE="Python">
   <INPUT NAME="Button3" TYPE="Button" VALUE="Click for URL" OnClick="My
Form.Text2.value=document.location" LANGUAGE="Python">
   <INPUT TYPE="TEXT" SIZE=25 NAME="Text2">
   <INPUT NAME="Button4" TYPE="Button" VALUE="Call global fn"
OnClick="foo1()" LANGUAGE="Python">
   <INPUT_NAME="Button5" TYPE="Button" VALUE="Script for... Test">
<script for="Button5" event="onClick" language="Python">
print "HelloThere";
window.alert("Hello")
def ATest():
        print "Hello from ATEst"

ATest()
</script>
   <INPUT NAME="Button6" TYPE="Button" VALUE="Set Other"
OnClick="Form2.Text1.Value='Hi from other'" LANGUAGE="Python">

</FORM><BR>
<P>
And here is a second form
<P>
<FORM NAME="Form2" METHOD="GET">
   <INPUT NAME="Button1" TYPE="Button" OnClick="Form2.Text1.Value='Hi'"
LANGUAGE="Python">
   <INPUT NAME="Button2" TYPE="Button" VALUE="Set Other" OnClick="MyForm
.Text1.Value='Hi from other'" LANGUAGE="Python">
   <INPUT TYPE="TEXT" SIZE=25 NAME="Text1">
   <INPUT NAME="ButRExec" TYPE="Button" VALUE="RExec fail"
OnClick="import win32api;win32api.MessageBox(0,'Oops')"
LANGUAGE="Python">
   <INPUT NAME="ButRExec2" TYPE="Button" VALUE="RExec fail 2"
OnClick="import sys,win32traceutil;print sys.modules;from
win32com.client import dynamic;import win32com.client.dynamic,
pythoncom, win32com.client;o=win32com.client.Dispatch('Word.Application'
)" LANGUAGE="Python">
   <INPUT NAME="ButVB" TYPE="Button" VALUE="VBScript Button"
OnClick='alert("Hi from VBScript")'>
</FORM><BR>

<SCRIPT LANGUAGE="Python">
x=13

def foo1():
   y = 14
   for name, item in globals().items():
        print name, `item`
   alert ("Hello from AXCode")
   print "Y is ",y

def Window_OnLoad():
        print "X is", x
        print "a is", a
        print "MyForm is", `MyForm`
        print "MyForm.Button1 is", `MyForm.Button1`
        MyForm.Button1.Value = "Python Rules!"
        Form2.Button1.value = "Form2!"
        MyForm.Text1.value = document.location
</SCRIPT>

</BODY>
</HTML>
-- 
Robin Becker



From jodywinston at my-deja.com  Thu Jul 22 10:47:42 1999
From: jodywinston at my-deja.com (jodywinston at my-deja.com)
Date: Thu, 22 Jul 1999 14:47:42 GMT
Subject: Is anybody aware of a tree GUI widget for tkinter or Pmw?
References: <3796E213.9F326412@eng.cam.ac.uk>
Message-ID: <7n7aua$t4o$1@nnrp1.deja.com>

In article <3796E213.9F326412 at eng.cam.ac.uk>,
  "D. Lopez-De-Ipina" <dl231 at eng.cam.ac.uk> wrote:
> Hi,
>
> Could anybody please tell me if there exits any TREE control widget
for
> tkinter or Pmw? I know that Pmw provides a rather rich set of
> 'megawidgets' but the tree control has not been implemented yet as far
> as I know.
>
> Thanks a lot,
>
> Diego
>
>

The PMW Contrib widgets (doughellmann at mindspring.com) has a tree control
widget.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From greg.ewing at compaq.com  Thu Jul 22 20:02:55 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 23 Jul 1999 12:02:55 +1200
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have code.)
References: <001601bed410$ec2ba0c0$642d2399@tim>
Message-ID: <3797B12F.AE6BF826@compaq.com>

Tim Peters wrote:
> 
> I think it worked out great!  I was surprised.

I wouldn't be quite so enthusiastic. I would say that
one can learn to live with the inconvenience, and
even almost forget that it's there most of the time.
But I still feel annoyed whenever I bump up against
it.

> and maintaining state via class instances will always be more
> Pythonic.

I don't think you can assert that until a Python
with lexical scoping has been in use for a while
and we see how people prefer to write things.
My bet is that a lot of things we write now like

  b = Button(..., command = lambda self=self: self.blat(42))

would get written as

  b = Button(..., command = lambda: self.blat(42))

and we would all wonder why we didn't storm
Guido's palace years ago and force him to release
his lexical scoping code at gunpoint...

Greg



From catlee at my-deja.com  Fri Jul  9 17:44:02 1999
From: catlee at my-deja.com (catlee at my-deja.com)
Date: Fri, 09 Jul 1999 21:44:02 GMT
Subject: Python re module
Message-ID: <7m5qev$lue$1@nnrp1.deja.com>

Try the following:

import re
re.search("((.)\\1+)","a")

Now, I know this isn't proper syntax (you shouldn't reference a group
inside itself), but on my machine python hangs.

Is this a bug, or a feature? =)

Cheers,
Chris


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From reason at shadow.net  Sat Jul 17 22:37:39 1999
From: reason at shadow.net (Vladik)
Date: Sat, 17 Jul 1999 22:37:39 -0400
Subject: [Q] In-Browser technology
Message-ID: <932337314.611.103@news.remarQ.com>

Hello,
Sorry for cross-posting, but
my question is really to the people on the news groups
who know about Eiffel, Python and Perl.

Basically,
We are to develop a web client (a program that
can be ran witting a web browser) for our otherwise n-tier
cross-platform system, our research department gave the
recommendation and actually a prototype of the client
in MS ASP (active server pages) and, I think but not
sure, blended with Visual Basic. Talk about bias...
But this means that it can only run in MS IE and not
in Netscape, also it limits the number of platforms not only
for the client on which web browser runs, but also for
the web server (MS platforms only, because web server with
active pages can only run there)
There is some talk about Java, but our web client
will also be ran via Serial links (not T1 connections)
on relatively inexpensive systems -- therefore Java is
something to look into but it may be slow.
So I am set on a quest to find something
that
a) can be executed within IE or Netscape,
b) be faster and overall more resource friendly then JavaScript technology
c) supports OO programming
d) does not depend on the platform on which web server is running.( if the
web server is running NT or UNIX (sun, linux)

I have found one thing that almost satisfies the requirements:
it is based on Oberon-2 programming language and so far sounds
VERY promising.
  The reason why I am still looking is because
there are more books and internet support for Python,Perl and Eiffel
then for Oberon (this will be a factor when presenting the proposal
to the management).  The other problem is that there is plugin
for this technology available for NT and Mac platforms (for both
IE and Netscape) and is not available for UNIX platforms (which means that a
web browser can not be ran on UNIX). There is source code,
though.
If interested: http://caesar.ics.uci.edu/juice/

So, finally, my question is is there a technology
I am looking for available for either Perl or Eiffel or Python

Since I am cross-posting, I would appreciate a reply
by mail as well.

Thanks in advance,
Vladislav






From cees at pcraster.nl  Tue Jul  6 10:50:18 1999
From: cees at pcraster.nl (Cees Wesseling)
Date: Tue, 06 Jul 1999 16:50:18 +0200
Subject: xml-0.5.1 and comments
Message-ID: <378217AA.2ED7E731@pcraster.nl>

Hi,

I am trying to manipulate some xml documents using the xml-0.5.1
package.
My goal is to write scripts that transform some xml using either SAX or
DOM
where apropriate.
With the DOM Walker, I am already succesfull in doing the most things I
want,
great tools!  But the DOM walker seems not to pass comment's through.
What I want is an exact copy of the input xml minus the transformations
I want
to made. But in Walker overriding doComment does not trigger any
comments.

Also wading to some of the SAX code, I started wondering if I can do
this with
SAX.
Does anyone has some sample code, for DOM or SAX that does nothing else
then
replicating the xml input exactly, is that possible?

Thanks in advance,
Cees Wesseling





From parkw at better.net  Wed Jul 21 20:10:07 1999
From: parkw at better.net (William Park)
Date: Wed, 21 Jul 1999 20:10:07 -0400
Subject: Repost: Re: Scientific Plotting?
In-Reply-To: <7n5fhn$9e5$1@nnrp1.deja.com>; from mhuster@hotmail.com on Wed, Jul 21, 1999 at 09:54:05PM +0000
References: <379392E7.EB0F29E7@ssec.wisc.edu> <7n5fhn$9e5$1@nnrp1.deja.com>
Message-ID: <19990721201007.B1050@better.net>

On Wed, Jul 21, 1999 at 09:54:05PM +0000, mhuster at hotmail.com wrote:
> Original reply:
> 
> I am also trying to find an adequate *object* based plotting package. I
> am writing some signal processing software and want to code plotting
> into the classes I am defining.
> 
> Some requirements:
> 1) Postscript output (for inclusion in high quality Word & Powerpoint
> documents) GIF, JPG, etc just do not have the resolution. They default
> to 72 dpi rendering. You can overcome this on Linux using Image Magick's
> convert -dens 300. The resulting files are large.

If you need to print it, then you should consider PiCTeX which is a
TeX macro package.

> 
> 2) Object oriented

I guess you could write python module.  From my experience, plotting
is impossible to standardize because of varying personal taste and
need (ie. scaling, labeling, title, fonts, lines, ...)

> 3) Images

??

> 
> Our current assessment:
> 
> GIST: not enough flexibility in configuring the plots. It has the
> Postscript we need, though. We have it coded in now for some look at the
> data. But it is not the long term solution.
> 
> Plplot: uses stroke fonts, even in postscript output. We cannot get a
> uniform look with matlab produced plots. Bummer. Otherwise it looked
> good. I am afraid when blown up the fonts will look even worse. For
> example. To get a bold font you have to play with pen width and font
> size. Isn't "Helevetica-Bold" easier to specify? Also, it is at a
> developmental (& FORTRAN) deadend. No chance to change things.
> 
> Gnuplot: Runs gnuplot through a pipe. All of the data (upto many MB)
> also has to go through the pipe. Unsat. I want to plot efficiently.
> 
> plotlib: not enough high level wrappers. I do not want to have to
> calculate default axis size, where to put tick marks, etc.
> 
> Pgplot: looks like plplot.
> 
> Piddle & graphite: looks good, but not mature enough. We are keeping an
> eye on it. Probably will try it.
> 
> PyMat: I want to get away from needing Matlab licenses
> 
> I am kind of bummed. If we wrap intelligent python wrappers around
> something, we will post it.
> 
> Michael Huster
> mhuster at hotmail.com
> 
> 
> 
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From greg.ewing at compaq.com  Wed Jul 21 23:53:11 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 22 Jul 1999 15:53:11 +1200
Subject: VC++ 6.0 larger exe files explained
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC15@molach.origin.ea.com>
Message-ID: <379695A7.EB1E7A6F@compaq.com>

"Stidolph, David" wrote:
> 
> This might (stress this is a pure guess) have something to do with virtual
> memory and its size being 4k pages.

That sounds like a pretty good guess. If Windows VM is anything
like Unix, it'll want to keep the code & data in separate pages
so it can share the code between processes.

In the case of extremely small programs (i.e. ones whose code
& init data together fit in one page), it might be better to
forget about trying to share the code and just load a separate
copy into each process.

Mind you, Microsoft programmers have probably long since
forgotten that there's any such thing as an executable smaller
than about 5MB, so it's understandable that they wouldn't
have thought of this...

Greg



From mwh21 at cam.ac.uk  Sat Jul 10 19:17:32 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 11 Jul 1999 00:17:32 +0100
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu>
Message-ID: <7m9to1$6qg$76@rks1.urz.tu-dresden.de>

Nick Bower <nick.bower at ssec.wisc.edu> writes:

> I feel really dumb asking this question, but how do I change the current
> working directory?  For browsing the file system I mean.
> 
> I've found exec commands, but these would surely be device dependent and
> there doesn't seem to be anything obvious in the os or os.path
> documentation.  i looked through the tutorials too.
> 
> thanks in advance, nick

os.chdir?

as in

Python 1.5.2 (#1, May 11 1999, 11:42:00)  [GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> os.chdir('/')
>>> os.getcwd()
'/'
>>> print os.chdir.__doc__
chdir(path) -> None
Change the current working directory to the specified path.
>>> os.chdir('~')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
OSError: [Errno 2] No such file or directory: '~'
>>> os.chdir(os.path.expand'~')
os.path.expanduser  os.path.expandvars
>>> os.chdir(os.path.expanduser('~'))
>>> os.getcwd()
'/home/mwh21'
>>> 

HTH
Michael



From tismer at appliedbiometrics.com  Wed Jul 21 07:43:33 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Wed, 21 Jul 1999 11:43:33 GMT
Subject: How fast can we multiply?
References: <001401bed345$d3934200$a72d2399@tim>
Message-ID: <3795B265.195C4FAE@appliedbiometrics.com>


Tim Peters wrote:
> 
> [Tim]
> >> Andrew Kuchling's wrapping gmp for Python is the most useful thing
> >> that's ever been done in this area.
> 
> [Christian Tismer]

[wants to rebuild the world using nuts & bolts]

> I play with & investigate bigint algorithms in Python, & I enjoy it.  But
> serious work in this area requires serious speed, and GMP doesn't compromise
> on that -- from slaved-over assembler to low-level mutable packed storage
> limbs, it's doing the right stuff for its task.  Factors of 5 count here
> <wink>.

Agreed. Still, if we resolve to low level functions which deal
with an optimized number of bits at a time (some reasonable cutoff),
then the Python overhead to operate intelligently with them
is (hopefully) small enough to be justified.

> You do "long1 & long2" in Python, and it's handled by an all-purpose bitop
> function that redispatches on the operation in each iteration of the inner
> loop.  It's "like that" everywhere:  the bigint code was written for
> compactness (longobject.c is only twice the size of intobject.c!), and
> no-hassle porting to the feeblest machine on earth.  The compromises have
> consequences (some good!  that code has been rock solid since the first
> release, while it can take a week to track down all the bugfix patches to
> the last official GMP release).

Yes. I think this simplicity is a real advantage.

The Python generalism to do the general &-code with
typechecks and everything on every inner loop appears
to be expensive for small objects only.
If I can write algorithms in a way that they operate
only above some "cutoff" level (say 5000 bit ints),
this is again not expensive.

...
[matmult]
> OTOH, Winograd's refinement of the matmult procedure requires 15 floating
> adds, where straight 2x2 matmult requires only 4; float + isn't much cheaper
> than float * on many modern processors.  So the win here isn't nearly as
> pure as for intmult, and the reduction is only from O(n**3) to O(n**2.81)
> operations anyway.  There are reasons few vendors offer this form of matmult
> in their libraries <0.5 wink>.

15 adds, sure. Would be expensive for a 2x2 matrix of reals.
But since we recusrively apply this to 2x2 matrices of submatrices,
the complexity of multiplication quickly voidens the real-processors
optimization again, and we win quickly.

Thinking of Jeffrey Chang's 6000x6000 matrices, what can we expect?
I didn't try, but let's assume a cutoff at about 94x94 matrices,
which appears fairly high for me.
Jeffrey can than think of the multiplication of 64x64 matrices,
where the scalars are the 94x94 matrices of above.
Cost for addition of two 94x94 matrices vanishes, compared to
multiplication.
So Jeffrey might expect to go from 64 ** 3 == 262144
down to 64 ** ( log(7)/log(2) ) == 117649.0 which gives him
a speedup factor of 2.23 .

If we are more lucky, and the cutoff enables us to 47*47 matrices,
the speedup will be 2.55 . And for 24*24 matrices, we end up at
a speedup of 2.91 .

May we expect a win for 12x12 matrices, already?
It would give us a great speed gain of 3.33, but we have to
guess the cutoff.

Reading Knuth about that, the necessary scaling for stability
implies that this techniques pay off for n >= 40, so if
we are lucky, the 47*47 case above could be implemented,
hopefully giving a factor of between 2and 2.5 .


One concern I have, and which stops me to dig seriously into
this is matrix stability. Since we are dealing with floats in
the low level, the result may be impacted by more rounding
errors, since more additions are involved.
At least this has to be studied carefully before replacing
one algorithm by another. I don't know what kind of
preconditioning is necessary here, but Knuth said it is.

Not so with fastlong, this is a perfect replacement.

I have not further investigated if Fourier transform
can be applied, and if matrix multiplication can be
expressed in terms of convolutions and Fourier transforms.
Multiplication of very long integers *can* be done this way.

[and about Jurjen's exact math, we'll wait for volunteers]

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From chris at frostnet.advicom.net  Tue Jul 13 13:29:01 1999
From: chris at frostnet.advicom.net (Chris Frost)
Date: Tue, 13 Jul 1999 17:29:01 GMT
Subject: MySQL Module
Message-ID: <19990713122901.A2538@pooh.frostnet.net>

Are than any modules for connecting to a MySQL sever? I didn't see
anything in the contributed section of python.org. (If there aren't, what
other databases do people think are better, why?)

thanks!
-- 
Chris Frost  |  <http://www.frostnet.advicom.net/chris/>
-------------+------------------------------------------
Public PGP Key:
   Email chris at frostnet.advicom.net with the subject "retrieve pgp key"
   or visit <http://www.frostnet.advicom.net/chris/about/pgp_key.phtml>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 256 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19990713/1eaad15a/attachment.sig>

From clarkeia at logica.com  Thu Jul 29 11:44:21 1999
From: clarkeia at logica.com (Ian Clarke)
Date: Thu, 29 Jul 1999 16:44:21 +0100
Subject: Wrong "From:" address!
References: <7npimc$l9b@romeo.logica.co.uk>
Message-ID: <7npssc$p7k@romeo.logica.co.uk>

Oops, Andrew Dolan didn't send this posting, I did.  Please send any replies
to clarkeia at logica.com.

Ian.





From kas at maps.magnetic-ink.dk  Fri Jul 16 06:03:45 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Fri, 16 Jul 1999 12:03:45 +0200
Subject: Python 1.5.2 + Linux 2.0.36 == Bad Combo?
Message-ID: <slrn7ou0ru.3sp.kas@titan.magnetic-ink.dk>

  G'day mates,

I'm running Python 1.5.2 on a Linux 2.0.36 kernel (RedHat 5.1 distro partly
upgraded to v5.2) and have recently experienced malfunctioning and/or non-
working python applications -- specifically:

  *  PageCast
     -- This one causes a Kernel Oops, though not a serious one.
     The author, Preston Landers, have managed to invoke a similar
     Kernel Oops using the same python/kernel combination.
  
  *  PyGCS
     -- The program exits silently as soon as the second is loggin in.
     The author, Jeff Blaine, hasn't been able to reproduce the behaviour.
     I have asked Preston Landers to try out the program on his set-up
     but have yet to hear the results
  
  *  DejaGrabber
     -- The program connects without problem but seems to only retrieve
     headers, not the full articles.
     Haven't contacted the author yet.

I would like to hear if anybody else have experienced similar problems using
the same Python / Linux kernel combination and, if so, whether is has helped
shifting to another kernel version or downgrading python.


Cheers,

  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From tim_one at email.msn.com  Wed Jul 21 23:29:12 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 21 Jul 1999 23:29:12 -0400
Subject: Stackless & String-processing
In-Reply-To: <7n4nrl$sfe$1@cronkite.cc.uga.edu>
Message-ID: <001101bed3f2$5e8874a0$642d2399@tim>

[Tim]
> ...
> If it isn't clear yet <wink>, efficiency isn't driving Graham
> here.  He's found a cool new approach, and is having fun with it.

[Graham Matthews]
> No that's not the case.

OK, my mistake:  you didn't find it, it's not cool, it's not new, and/or
you're not having fun <wink>.

> Efficiency drives all my programs *when it's a necessity*.

Ack, I'm speaking English here.  If efficiency isn't your *primary* concern,
then efficiency isn't "driving" you.  Dictionary.  Look up.  Efficiency is
Fernando's primary concern wrt parsing approaches -- it drives him.  That's
how you two differ.  It's important to note that efficiency really is a
necessity in his applications (have you read any of his papers?  no claim
makes sense out of context).

> My point is that using combinators:
>
> a) is a very nice way of structuring a parser since it is very amenable
>    to change and robust under change.

Could well be.

> b) doesn't have to cost anything in efficiency since you can freely mix
>    combinators with other parsing techniques.

Like programming in Python doesn't have to cost anything in efficiency since
you can freely mix Python with other languages?  It's a true enough
argument, but useless in practice unless 5% of the code accounts for 95% of
the runtime.  Which it often *does* in general-purpose Python apps.  For
parsing, it's not clear what kind of sequencing you could want that isn't
already covered by the "other parsing techniques" that are presumably doing
95% of the (presumed to be) efficiency-critical computation.  Have something
concrete in mind?

> c) is a general programming technique useful in a vast range of programs
>    outside of just parsing. For example I have used it for type checking
>    and writing a GUI. The literature has a lot of other examples of this
>    approach.

I didn't see anyone argue against that.  This thread started with string
processing, and everyone else has stuck to that.

> Moreover the point of combinators is not to "interweave the grammatical
> structure with other computations". Combinators do do that, and I think
> it's a great benefit.

Expect that depends on the app:  sometimes a benefit, sometimes not.  I've
used systems of both kinds, and neither is a pure win.

> But the point is to:
>
> a) clearly separate sequencing code from other code. In a combinator
>  	 program the precise control flow of the program is clear.

In a way deeper than it is in a functional program not using monads?  IOW,
is the clarity of "control flow" due to monads specifically or just to
functional semantics in general?  AFAIK, "combinator" hasn't sprouted a new
meaning beyond its old "higher-order curried function" one.

> b) allow programs to program their own sequencing combinators (rather
>    than providing some fixed set that the language designer hopes will
>    be enough.

WRT parsing specifically, what kinds of desirable sequencing do you find
lacking in other approaches?  Any parsing framework caters more-than-less
directly to juxtaposition, choice, optionality, repetition and (at least
conceptual) backtracking.  I understand you can model all those with
combinators.  What I haven't seen is an example of something clearly useful
(wrt parsing specifically) *beyond* those.

> ...
> I have been using combinators for some time now and never noticed the
> "grammar" being cluttered with all these "semantic annotations".

We often see most clearly that which we hate <wink>.  I saw 'em all over the
place in the papers.

> What annotations one requires can easily be separated from the grammar
> either by sensible layout, or by the design of clever combinators that
> hide semantic annotations away.

I mentioned indirectly that what drives me is variously "speed or
maintainability" in parsing systems.  "clever" is at odds with the latter --
unless all the cleverness comes built in to the system once & for all.

> ...
> Combinators are very robust under change. I have a parser written using
> combinators and I often tweak the grammar, without requiring large
> changes to other code or the parser itself.

That's encouraging!

> Have you ever written a combinator based parser or are you just
> "theorising"?

You ask that immediately before quoting my

> : I haven't tried using parser combinators, but in reading the papers ...

?  Prophetic <wink>.

> Ah there's the rub -- you haven't written a combinator based parser so you
> your comments are just what you *think* might happen if you did.

Oh yes!  Parsing with combinators is a Porsche.  I've driven all kinds of
cars for thirty years, in various terrains and in all kinds of weather; even
built a few engines from metal I smolt myself <wink>, that other people
relied on.  I used functional languages extensively too, although very
lightly this decade.  So when I read the Porsche owner's manual it's not
exactly blind imagination driving my concerns.  It's certainly no substitute
for a test drive, though.  I installed Clean with the aim of doing that, but
had too much fun poking at other aspects of the system to get around to it.

> I don't know where you got the idea that speed is a problem!

*Can* be a problem.  Fernando covered that exhaustively, and you actually
haven't disgreed with his points for all your words to the contrary <wink>.
I also have much other experience crafting elegant functional solutions that
had to be tossed in practice for the same kinds of "can't automatically
optimize a huge nest of clever higher-order applications" reasons.

> For some kinds of parsing I wouldn't use combinators rather I would use
> automaton techniques. But for most kinds of parsing they are plenty
> fast enough.

"Most" depends on what you're doing.  For most things I do for fun, even
SNOBOL4 was plenty fast enough (and I don't doubt for an instant that even a
casually tossed-together combinator-based program is faster than that!).
Half the things I do for work deal with artificial languages that are
specifically designed for one-token lookahead techniques, because they're
designed for peak speed over megabytes of input; the other half is parsing
huge ambiguous grammars on cheap PCs via dynamic programming techniques, and
there we *claw* to save bytes and milliseconds, no holds barred.

Different strokes.  I believe I'd enjoy combinator approaches for my own
fun; I still doubt they'd be practical for my work (but I'd rather try that
than continue arguing about it <0.9 wink>).

> ...
> Each new example in these papers requires a different combinator because
? each new example illustrates some new sequencing idea -- eg.
determinisation,
> coninuation, ...
> The papers are illustrating the generality of the techniques, and hence
> you expect to see a lot of different combinators.

Yes, and I confess to over-panicking at what I indeed should have expected.

> In practice however (something you wouldn't know about right?) you don't
> use all these combinators.

Which casts some doubt on the great utility of being able to create new
ones.  Parsing is an old problem.  If only a handful of combinators are of
actual use in parsing problems, freeze the semantics, package them in a
library, and implement a pleasant parser generator on top of them <wink>.

> I have written a parser for a new programming language and I used 10
> combinators (5 of which I re-used in the type checker!).
> Yes you do have to build your own combinators (that's the point of the
> approach!), but once done you stick them in a library and re-use them.

That's fine for a one-man show.  "Maintainability" in my life cuts across
programmers, and it's Not Good for each person to reinvent their own basic
machinery -- even to have the ability to do so.  One reason I like Python is
that it *doesn't* let the jerk down the hall invent their own control
structures <0.5 wink>.

I will give parser combinators a try; they do sound interesting to me.  More
interesting would be to hear about that "new programming language"!

i-believe-you-can-parse-it<wink>-ly y'rs  - tim





From ivanlan at callware.com  Mon Jul 19 18:11:34 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 19 Jul 1999 22:11:34 GMT
Subject: SMTP problem
References: <hrMk3.1158$Gc6.192154624@newsb.telia.net>
Message-ID: <3793A296.7A0FAD18@callware.com>

Hi All--

"Joacim L?wgren" wrote:
> 
> Hi,
> 
> I copied this straight from the Python SMTP example, however it doesn't seem
> it's working. I'm not quite sure what ^D does and if it is different on
> Unix/Windows as I think this is the problem with the script.
> 

1)  It appears that you're on Windows, so you would use ^Z (control-z)
instead of ^D.  In either case, the function of the control character is
to indicate the end-of-file, which tells the SMTP script you're done
typing your message and that it should be sent.

2)  Since you're on Windows, and windows doesn't have any native SMTP
processing abilities, you'll need to replace 'localhost' in the example
program with the name of your mail server.  If netlimit.com is your
host, then it is possible that you'll have a mail server at
'mail.netlimit.com' -- but I would check with your IS people to make
sure, as you need a mail account on the server in order to be able to
send and receive mail.

3)  If I'm mistaken and you really are on UNIX, then you need to make
sure that the system you're running this script on has an SMTP
server/sendmail daemon running.  If not, the script won't work, since
it's hard-coded to your localhost.

<windows-don't-need-no-stinkin-sendmail>-ly y'rs,
Ivan 
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From aahz at netcom.com  Fri Jul 30 11:06:56 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 30 Jul 1999 15:06:56 GMT
Subject: Need help with TCP/IP client access from Windows
References: <slrn7q1ej3.18d.esr@snark.thyrsus.com> <1278839420-22085362@hypernet.com> <19990730005211.T7149@thyrsus.com>
Message-ID: <7nsf2g$ifp@dfw-ixnews11.ix.netcom.com>

In article <19990730005211.T7149 at thyrsus.com>,
Eric S. Raymond <esr at thyrsus.com> wrote:
>Gordon McMillan <gmcm at hypernet.com>:
>>
>> Probably better just to have a default file, and let the user 
>> override that if need be. Which may well never happen.
>
>OK, but where to put it?  I don't know what the conventions are.

I'd stick it in the directory your script is in.  That's pretty standard
for Windoze.  In fact, I'd make it "<base script name>.ini".
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From amaranda at nospam.com  Thu Jul  1 10:41:27 1999
From: amaranda at nospam.com (Alex Maranda)
Date: Thu, 01 Jul 1999 15:41:27 +0100
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain>
Message-ID: <377B7E17.5BD7@nospam.com>

> their leader is killed, etc. The events would be things such as
> "create", "monster sees player", "clock tick", etc. All the functions
> in the script would be callbacks written in C++, to manipulate the
> monster to do things such as "patrol area", "play an animation",
> "sleep for X seconds", etc. In other words, the scripts would be glue,
> not the main thread of execution.
you use confusing terms; I understand what you're saying, but you
shouldn't use the word 'thread' for anything else than its OS meaning.
You could say the 'the bulk of processing'. You might very well end up
using real Python threads; don't forget in this case to release/reaquire
the global interpreter lock in your C++ callbacks, if you make blocking
syscalls (unlikely in a game).

> 
> I'm been using Python for a couple of years now, and it is my favorite
> scripting language. But, I don't have much experience with embedding
> it in a another program, and I don't know enough about the speed
> issues. So, my questions are...
> 
> - Will Python be fast enough? Even with possibly 100 or 200 active
>   items on a level?
Unlikely. You won't know until you try. To my knowledge UnrealScript is
eventually translated to C++.

> - Especially when all of the scripts will be live, and usually in the
>   middle of a callback?
> - Should I use the interpreter in the library, or should I take an
>   approach of crunching the byte-codes myself?
??

Cheers,
-- 
Alex Maranda         mailto: amaranda at spider dot com
Spider Software Ltd. Tel: +44 (0)131 4757036
Edinburgh, UK        http://members.xoom.com/Alex_Maranda
STREAMS based communications protocols for embedded systems



From gmcm at hypernet.com  Mon Jul 19 12:42:13 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 19 Jul 1999 11:42:13 -0500
Subject: Problems with PyRun_SimpleFile() 
In-Reply-To: <379336D5.799D83E0@gmx.net>
Message-ID: <1279723778-5524667@hypernet.com>

Markus Ewald writes:

> Excuse this dumb question, 

Not dumb, just very frequent.

> but how can I use PyRun_SimpleFile() ?
> I'm using VC6 on Win98, where I've declared some simple functions
> for testing.
> 
> Well -- everything works fine when I use PyRun_SimpleString().
> 
> If I put the same lines I gave PyRun_SimpleString() in a File and
> then use PyRun_SimpleFile() to execute it, I get an access
> violation.

Make sure you have linked your stuff against the exact same run time 
as python15.dll. In other words, choose "Multithreaded DLL". NOT 
"Debug Mulitthreaded DLL" or any other choice.

- Gordon



From tim_one at email.msn.com  Mon Jul 26 22:13:32 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 26 Jul 1999 22:13:32 -0400
Subject: Newbie question
In-Reply-To: <379d0ee6@news.gci.net>
Message-ID: <000c01bed7d5$a0a92c80$eea22299@tim>

[Arnaldo]
> I'm just trying out some examples from the book Learning Python and I run
> into a exercise that is not working for me.
> It's on page 95 , exercise 4b.
> This is the answer the book has
> ===========================================
> l = [1, 2, 4, 8, 16, 32, 64]
> x = 5
> i = 0
> while i < len(l):
>     if 2 ** x == l[i]:
>         print 'at index', i
>         break
>         i = i + 1
> else:
>      print x , 'not found'
> ============================================
>
>  When I run this , python hangs. When I do Ctrl-c , this is what I get
> Traceback (inntermost last):
>     File "powera.py, line 5, in ?
>         if 2 ** x  == l[i]:
>
> What am I doing wrong?

What do you *think* could be going wrong?  The program appears to be running
forever.  How could that happen?  Even if 2**x==l[i] is never true, the loop
is supposed to end as soon as "i < len(l)" is false.  So how can "i <
len(l)" always be true?

Bad hint:  The code above is not the answer the book gives -- although it's
close.

an-hour-of-thinking-will-save-10-seconds-of-reading<wink>-ly y'rs  - tim





From dalke at bioreason.com  Sun Jul 11 02:47:39 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Sun, 11 Jul 1999 00:47:39 -0600
Subject: code coverage
Message-ID: <37883E0B.D956F029@bioreason.com>

Hello,

  After the discussion a couple weeks back about code coverage
utilites for Python, I decided to look into them because I want
to ensure that our regression tests really do execute all (or at
least most) of the code in our packages.

  The two mentioned in the thread were:
 trace.py - http://www.musi-cal.com/~skip/python/trace.py
      by Skip Montanaro

 pycover - ftp://ftp.python.org/pub/python/contrib/All/pycover-0.1.tar.gz
      by Andrew Csillag


  Neither of these fit my needs, so I ended up highly modifying
trace.py.  For those interested, the new version is at
   ftp://starship.python.net/pub/crew/dalke/trace.py

I would appreciate people taking a look at it and testing it out;
It's in a pretty complete state, but needs review and suggestions
for improvements.


  The major features I added are:

     a more extensive command-line interface using getopt and
reporting a full --help message

     support for packages.  trace.py would use only basename of
the __file__ when generating a coverage log.  Thus, if there were
two submodules with the same name, one would overwrite the other.
Now the results are saved to the file named __name__ + '.covered'

     able to exclude coverage on a module based on its __file__
and/or on its __name__ (so you can exclude all the system modules
using sys.prefix).  pycover, for example, has built-in exclusion
of /usr/local, which isn't always appropriate for our installs.

     a big problem I had with both of the coverage programs was
they didn't do a very good job at telling me what executable lines
were not covered.  For example, one of my test scripts is:

=====
import daylight.Smiles

def main():
        mol = daylight.Smiles.smilin("COON")
        if len(mol.atoms) > 9:
                print "Hello" #pragma: NO COVER
                print "there"
        else:
                print "Small"
        print mol.cansmiles()

if __name__ == "__main__":
        main()
=====

when run through pycover, the un-executed lines start with a '!'

======
> import daylight.Smiles
  
> def main():
>       mol = daylight.Smiles.smilin("COON")
>       if len(mol.atoms) > 9:
!               print "Hello" #pragma: NO COVER
!               print "there"
!       else:
>               print "Small"
>       print mol.cansmiles()
  
> if __name__ == "__main__":
>       main()
======

As you see, the "else:" is marked as unexecuted, and it always will
be even if both branches of the if statment are taken, because else
generates no SET_LINENO instruction.  The basic problem is in
determining if a line should have been executed or not, and the
programs do it by looking for non-empty, non-comment ("""detection
of multiline quotes is also a problem
for these two programs""")

Now, nearly exery executable line, when compiled for the PVM, contains
code listing the line number.  So I wrote some code (based on dis.py)
to disassemble a module's code objects, look for the SET_LINENO,
and use those numbers to tell which executable lines haven't been
covered.

This trick doesn't work for everything, but is better than nothing.


So, if you want, please take a gander at the code, at
   ftp://starship.python.net/pub/crew/dalke/trace.py


						Andrew
						dalke at bioreason.com



From chris at frostnet.advicom.net  Wed Jul 21 09:46:14 1999
From: chris at frostnet.advicom.net (Chris Frost)
Date: Wed, 21 Jul 1999 08:46:14 -0500
Subject: Merging two lists as a dictionary
In-Reply-To: <19990721153539.S6170@xs4all.nl>; from Thomas Wouters on Wed, Jul 21, 1999 at 03:35:39PM +0200
References: <19990720211946.A12202@pooh.frostnet.net> <19990721153539.S6170@xs4all.nl>
Message-ID: <19990721084614.C15502@pooh.frostnet.net>

On Wed, Jul 21, 1999 at 03:35:39PM +0200, Thomas Wouters wrote:
> >>> keywords = ['s','p','a','m']           
> >>> values = ['eggs', 'ham', 'toast', 'omelet']
> >>> result = {}
> >>> for i in range(min(len(keywords), len(values))):
> ...     result[keywords[i]] = values[i]
> ... 
> >>> result
> {'m': 'omelet', 's': 'eggs', 'p': 'ham', 'a': 'toast'}
> 
> the 'min()' is unecessary if you are 100% certain the lists are same length,
> or if you know which list will be the shorter one. (iter over the shortest
> one, in that case)
Great, that works better than my error checking method.

-- 
Chris Frost  |  <http://www.frostnet.advicom.net/chris/>
-------------+------------------------------------------
Public PGP Key:
   Email chris at frostnet.advicom.net with the subject "retrieve pgp key"
   or visit <http://www.frostnet.advicom.net/chris/about/pgp_key.phtml>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 256 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19990721/4427a611/attachment.sig>

From balaji at platinum.com  Tue Jul  6 17:14:52 1999
From: balaji at platinum.com (b s)
Date: 06 Jul 1999 17:14:52 -0400
Subject: popen2 on Windows
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90EC1@RED-MSG-50> <87hfnit1tc.fsf@linux01.prc.tno.nl>
Message-ID: <uk8sd5vlf.fsf@mailserver.ab.platinum.com>

John van der Koijk <vdkoijk at linux01.prc.tno.nl> writes:

All these things will work only if your application is a console application.
If a service is trying to spawn a child process and you want to capture the
output in the parent and wait for childs termination (by getting an EOF on the
read end of the pipe), it gets a little more complicated.
check out
"Creating a Child Process with Redirected Input and Output"
in the MSDN. It is not completely working, but gives you the mechanics of the
equivalent of popen().
|"BT" == Bill Tutt <billtut at microsoft.com> writes:
|
|    >> After coping with my initial frustration, I've decided to make
|    >> an unthreaded stub program talking to the DDE client, using
|    >> popen2 to do the IPC with the GUI and controller
|    >> program. Sadly, this approach also does not seem to work,
|    >> because using popen2 gives me:
|    >> 
|    >> [....] os.fork() AttributeError: fork
|    >> 
|    >> This is amazing, since the manual for 1.5.2 says:
|    >> 
|    >> 8.13 popen2 -- Subprocesses with accessible standard I/O
|    >> streams
|    >> 
|    >> Availability: Unix, Windows.
|    >> 
|
|    BT> Well, its wrong. :) Windows doesn't have fork().
|
|    BT> The win32pipe module exposes popen2() that does what you want
|    BT> it to do, however you need to avoid a bug in Win9x. The
|    BT> knowledge base article at:
|    BT> http://support.microsoft.com/support/kb/articles/q150/9/56.asp
|    BT> covers what you need to do to avoid this dilemma.
|
|Wow. I'm completely flabbergasted...
|
|I have realised that I could also use a socket interface. Works great,
|and speed is not one of my concerns, right now.
|
|Thanks,
|-- 
|John van der Koijk.
|--
|TNO Institute of Industrial Technology
|
|PO Box 5073, 2600 GB, Delft, The Netherlands



From mstenber at cc.Helsinki.FI  Tue Jul 27 23:41:52 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 28 Jul 1999 06:41:52 +0300
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <p51zdvn4c7.fsf@bidra241.bbn.hp.com> <14236.32616.980578.536614@dolphin.mojam.com> <379D89BC.DA91BA8C@hp.com> <14237.60407.130426.461846@dolphin.mojam.com> <14238.335.458728.140169@lrz.de> <14238.1372.334823.378192@dolphin.mojam.com>
Message-ID: <al8zp0hjv9b.fsf@sirppi.helsinki.fi>

Skip Montanaro <skip at mojam.com> writes:
>     Eugene> Skip Montanaro writes:
>     >> True.  I was thinking about the situation I'm in (small developer, a few
>     >> machines to maintain/upgrade/replace).
> 
>     Eugene> What's wrong with recoding the hot spots in C? Python is
>     primarily a rapid prototyping/scripting/glue language, it was never
>     meant as a sole implementation language for applications.
> There's nothing wrong with recoding hot spots in C.  Many people do it all
> the time.  Practically speaking, however, many Python users don't have the
> technical depth or other resource (VC++, for instance) to write chunks of
> their applications in C or C++.

That isn't (IMNSHO) the greatest problem; for me, main problem lies in the
fact that when you bring C/C++ to bear, you lose most of the 'totally
portable'-characteristics of normal Python code and get to test same
program on multiple platforms to see if the C/C++ extensions work as
advertised there, etc. Generally speaking, when developing Windows-Python
programs, you can do development mostly on UNIX and *poof*, it works on
Windows. 

On the other hand, try porting say, C/C++-based socket programs from UNIX
to Windows.. *brr*

> Skip Montanaro	| http://www.mojam.com/
> skip at mojam.com  | http://www.musi-cal.com/~skip/
> 847-475-3758

-Markus Stenberg

-- 
	Linux!  Guerrilla UNIX Development     Venimus, Vidimus, Dolavimus.



From mstenber at cc.Helsinki.FI  Thu Jul 29 03:30:39 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 29 Jul 1999 10:30:39 +0300
Subject: HPUX and threads?
References: <000b01bed969$f5b62120$71a22299@tim>
Message-ID: <al86733ud40.fsf@sirppi.helsinki.fi>

"Tim Peters" <tim_one at email.msn.com> writes:
> [Markus Stenberg]
> > Is HPUX-11 thread support tested/implemented/whatnot?
> 
> You may want to ask on the Thread-SIG if you don't get an answer here.
> 
> > This at least does not look promising..
> >
> > # /opt/python/bin/python
> > pthread_mutex_init: Invalid argument
> > Memory fault(coredump)
> >
> > (also tests failed with the same message, obviously)
> Are you running pthreads?  Python is sure trying to <wink>.

I am totally clueless as to C-based threads in general and pthreads in
particular. There are manual pages for pthread* things, and the library,
thus, I suppose I am.

> The msg is coming from line 263 of Python/thread_pthread.h.  It means the
> platform's
> 
> 		status = pthread_mutex_init(&lock->mut,
> 					    pthread_mutexattr_default);

Manual page says this.. wonder what lock->mut is, hopefully this type ;)

      int pthread_mutex_init(
         pthread_mutex_t *mutex,
         const pthread_mutexattr_t *attr
      );

> returned a non-zero (failure) status.  "Invalid argument" is what the
> platform perror() produced.  A coredump after an error here isn't
> surprising.
> 
> This is likely the first call made to any pthreads function, and is so
> vanilla that I'd guess pthreads itself doesn't work on your platform (does
> it from C?), or it's some non-std pthreads variant that thread_pthread.h
> doesn't know about.

Seems like non-std pthreads variant that thread_pthread.h doesn't know
about (see my other post).

> not-that-insight-implies-a-cure-ly y'rs  - tim

-Markus

-- 
We are Pentium of Borg. Division is futile. You will be approximated.



From kernr at mail.ncifcrf.gov  Fri Jul  2 18:11:38 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Fri, 02 Jul 1999 22:11:38 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net> <377B711A.E8B65E0B@cfdrc.com> <377BC620.F5553FF5@easystreet.com> <b%Pe3.4033$lr3.164497@news2.jacksonville.net> <14204.4556.177837.851763@lrz.de> <Pine.LNX.4.10.9907012309060.6263-100000@us2.mayo.edu>
Message-ID: <377d35d7.3612161@news.erols.com>

On Thu, 1 Jul 1999 23:16:00 -0500, Travis Oliphant <olipt at mayo.edu>
wrote:

>
>This has been a very interesting discussion on GUI's.
>
>It struck me that all this talk of a cross-platform GUI might be
>preventing a lot of very good GUI's from being used.
>
>So, I looked in to PyGtk and PyGnome (with its Gnome Canvas) and I'm
>impressed.  Right now it fits what I want to do very well and it works on
>systems with an X server (aren't X servers available for MAC's and
>Windows?)  

Not for free.  Well, Microimage's X server, MI/X, is freely avialable
for Windows and Mac, but it's only approximately X11R5 compliant
(http://www.microimages.com/freestuf/mix/mix.htm).  PyGTK and PyGnome
are unlikely to work with it satisfactorily.  Heck, tkinter doesn't
work satisfactorily with it.

Starnet does have an X11R6 server for Windows for which they give away
a free demo that has a 2 hour time limit per run, although it can be
restarted again.  (http://www.starnet.com/demo.htm)

There is a project in progress to port XFree86 to Windows with the
Cygwin tools.  Everything compiles now; we just need to wait a little
before it runs.  :-)

GTK+ for Windows (http://www.gimp.org/~tml/gimp/win32/) still hasn't
had enough of the libraries ported to compile PyGTK.  This could
change soon, though.

>So, I'm happy with it. If you're in the same situation you
>might like it too.  I tend to agree that there is plenty of room for more
>than one GUI.
>
>Travis Oliphant

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From gmcm at hypernet.com  Wed Jul 14 09:14:37 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 14 Jul 1999 08:14:37 -0500
Subject: Recursive method
In-Reply-To: <378C214F.3CF9CDD3@eml.villa-bosch.de>
Message-ID: <1280168234-10452044@hypernet.com>

Ralph Gauges wrote:

> Hans Nowak wrote: 
> > You have a nested function here. (I say function, because it does not
> > look like a method... it's missing the self argument!
> 
> Actualy it is a method, but JPython doesn't seem to have a
> self. I was missing that as well.

JPython 1.0b3 on jdk1.1
Copyright 1997-1998 Corporation for National Research Initiatives 
>>> class A: 
...  def foo(self): 
...    print `self` 
... 
>>> a = A() 
>>> a.foo() <__main__.A instance at 111> 
>>>

Nothing missing.

> > 
> > So, to solve this, you could try a different approach... maybe
> > 
> > (you said it's a method, right?)
> > 
> > class Blahblah:
> >     ...
> >     def makeTree(self):
> >          ....
> >     def makenodes(self, C):
> >          ....
> > 
> 
> 
> 
> I tried that as well, but I got the same error message. Only
> after I moved the makenodes outside the class it worked.

Also from JPython:
>>> class A:
...  def foo(self, seq):
...   if seq:
...    print seq[0]
...    self.foo(seq[1:])
...
>>> a = A()
>>> a.foo([1,2,3])
1
2
3
>>>

No problem with recursion. You've made some other error for which 
you'll kick yourself when you finally see it <wink>!


- Gordon



From Georg.Mischler at p98.f112.n480.z2.fidonet.org  Fri Jul  2 05:32:48 1999
From: Georg.Mischler at p98.f112.n480.z2.fidonet.org (Georg Mischler)
Date: Fri, 02 Jul 1999 10:32:48 +0100
Subject: Python GUIs: Portability or not?
Message-ID: <0000165e@bossar.com.pl>

From: Georg Mischler <schorsch at schorsch.com>

Mikael Lyngvig wrote:
> Travis Oliphant wrote:
>
>[...]
> >So, I looked in to PyGtk and PyGnome (with its Gnome Canvas) and I'm
> >impressed.  Right now it fits what I want to do very well and it
> >works on systems with an X server (aren't X servers available for
> >MAC's and Windows?)
>
> I don't know about the Mac, but I'm sure there are X servers available
> for NT.  However, not like they are available for Linux and other Unix
> boxes - you don't just download the X11 source, compile it, and that's
> it (unless a lot has changed the last few years).
>
> You'd probably have to buy a commercial X server for NT or, even
> worse, ask your users to do so.


Have a look at  http://www.microimages.com/www/html/freestuf/mix/.
This is a free X sever for both the mac and NT. It works pretty well
for what it is. But if you have a choice, and assuming that you plan
to work with X and non-X applications at the same time, then you
probably want a native GUI for your application.

There are other X servers around, but those who blend into the
host GUI (instead of creating their own X desktop within a window)
all seem to be commercial.

Now *this* would be a task for an open software project... ;)


Have fun!

-schorsch

--
Georg Mischler  --  simulation developper  --  schorsch at schorsch.com
+schorsch.com+  --  lighting design tools  --  http://www.schorsch.com/


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From bwarsaw at cnri.reston.va.us  Fri Jul 30 18:39:23 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Fri, 30 Jul 1999 18:39:23 -0400 (EDT)
Subject: DCPIGgies Get Together
Message-ID: <14242.10651.804230.338090@anthem.cnri.reston.va.us>

Notice to all Washington DC area Python users:

We are pleased to be resurrecting the DC area Python interest group
meetings, now called DCPIGgies.  We will be having our first meeting
at CNRI in Reston, Virginia on Tuesday August 31, 1999 from 7:30pm to
9:30pm.  The aim is to keep this meeting pretty technical.

To that goal, we're going to have two exciting presentations.  For the
first hour, the Digital Creations guys will be giving a talk on Zope
2.0's unique object model (acquisition, run-time classes,
transactional objects, etc.) including how Python programmers might
use this in their own non-Zope projects.

In the second hour, Andrew Kuchling will be talking about something
XML-ish, probably processing the Open Directory Project's
(www.dmoz.org) RDF dumps.

There will be question and answer periods after each talk, and we may
have an "open phones" going later than 9:30 if there's interest and if
Guido or I am still awake. :)

Pizza, salad, and soda will be provided.  We are asking for a
voluntary $5 donation at the door to cover the costs of food.  It is
very important that you RSVP by email to bwarsaw at cnri.reston.va.us so
that I can get an accurate head count (otherwise you may go hungry!).
We also need to make sure that we can fit everyone into our conference
room, so let me know if you're coming, even if you aren't going to
eat.

Directions to CNRI can be found on the Web:

    http://www.cnri.reston.va.us/directions.html

See you then!
-Barry



From arcege at shore.net  Sat Jul 10 06:20:10 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Sat, 10 Jul 1999 10:20:10 GMT
Subject: console show in Python?
References: <378626CF.9495FAB0@mry.rbd.com>
Message-ID: <u7Fh3.1487$6M6.461536@news.shore.net>

Scott Shillcock <shillcoc at mry.rbd.com> wrote:
: I am moving to Python from Tcl/Tk.
: Is there something similar to the Tcl/Tk 'console show' command in
: Python.

: I am using Tkinter for my gui and once I call mainloop() I no longer
: have access to the Python command line. I want to be able to interact
: with my gui from the command line to test new code.

: From Tcl/Tk I could turn on and off the interactive command line with
: 'console show' and 'console hide'. That is all I want to do with Python.
: Can anyone help me with this or point me in the right direction.

On UNIX anyway, if you use the quit() method from the mainloop, then
you can restart from Python's interactive prompt by calling the
mainloop() method again.

The code for Tkinter should make this behavior consistant on all
platforms, but has anyone tested this there?

  -Arcege




From arcege at shore.net  Mon Jul 12 16:13:53 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Mon, 12 Jul 1999 20:13:53 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787D8DA.3774AA7D@ssec.wisc.edu> <7m8p7t$o7s@dfw-ixnews10.ix.netcom.com> <3787FB51.12FE0048@ssec.wisc.edu> <7m923i$dem@dfw-ixnews5.ix.netcom.com> <14217.63931.75257.566266@weyr.cnri.reston.va.us> <7md2ip$19q2$1@nntp6.u.washington.edu> <14218.7724.879301.831590@weyr.cnri.reston.va.us>
Message-ID: <50si3.1643$6M6.541997@news.shore.net>

Fred L. Drake <fdrake at cnri.reston.va.us> wrote:

: Donn Cave writes:
:  > For what it's worth, me too:  current working directory is a process
:  > setting.  It's relevant to files and directories, to be sure, but whoever
:  > put chdir() and getcwd() in process parameters knew what he was doing.

:   I am vindicated!

For what it's worth, I'm with you both.

  -Arcege




From guido at CNRI.Reston.VA.US  Sat Jul 10 10:21:55 1999
From: guido at CNRI.Reston.VA.US (Guido van Rossum)
Date: Sat, 10 Jul 1999 10:21:55 -0400
Subject: builtin sequence subscript patch
In-Reply-To: Your message of "Fri, 09 Jul 1999 08:42:27 EDT."
             <199907091242.IAA07794@northshore.shore.net> 
References: <199907091242.IAA07794@northshore.shore.net> 
Message-ID: <199907101421.KAA04351@eric.cnri.reston.va.us>

> Someone posted that a class instance, with a __int__ method, could not
> be passed as a subscript of a built-in sequence.  I think this is a bug
> since the build-in insert method of list can take a like object
> correctly.
> 
> I'm enclosing a patch to correct this, based on the latest CVS code.
> I'll post the code to the newsgroup as well.

Michael, thanks for your patch.  It solves your problem.  But it
creates a new problem for me: you can now use floating point numbers
as list indices and it will truncate them.  I don't like this feature,
and I'm worried that it will be hard to remove once it's there.  So I
think it's better to use int() to cast your list indexing arguments
when you want to support objects with __int__ methods.

In Python 2.0, I want to change the __int__ interface into two
different ones: one that converts an existing integer-like value to
a real int, and one that tries to convert any (non-complex) numerical
value to a nearby integer.  Then list indices will use the former
conversion, and you will be able to do what you want without an int()
cast yourself.

(And yes, I know that the insert() method sets a precedent, but I
don't want to move any further until I have the interface that lets me 
do what's right!)

--Guido van Rossum (home page: http://www.python.org/~guido/)




From jam at p98.f112.n480.z2.fidonet.org  Thu Jul  1 13:10:34 1999
From: jam at p98.f112.n480.z2.fidonet.org (jam)
Date: Thu, 01 Jul 1999 18:10:34 +0100
Subject: please help with simple regex
Message-ID: <0000164a@bossar.com.pl>

From: jam <jam at newimage.com>

On Fri, Jul 02, 1999 at 12:35:51AM +0000, viofis at my-deja.com wrote:
> 
> I'm a beginner at Python, so this may be a stupid question:
> 
> I have a string, for example, myString="the Potato".
> 1. Convert it to all lowercase ("the potato")
> 2. Match it against a list of strings - "potato,tomato,corn,etc...",
> extract "potato" (if it matches with something on the list), then save
> into myNewString.
> 
> Any help is greatly appreciated.
> 
> thanks,
> Vadim
> 

hi.

I don't quite understand what it is you are trying to do. if you want to
take a string and find out if it contains some other string, you probably
don't need to worry about regular expressions at all.

what kind of problem are you trying to solve? do you 'need' to use a
specific solution for some reason?

I tried this:

[~] [9:02pm] [jam at toast-pts/6] % python
Python 1.5.2 (#1, Apr 18 1999, 16:03:16)  [GCC pgcc-2.91.60 19981201
(egcs-1.1.1  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import string
>>> buf = "The Carrot"
>>> validveggies = ("carrot", "corn", "spam")
>>> buf = string.lower(buf)
>>> for w in string.split(buf):
...     if w in validveggies:
...             print "valid veggie '%s' found." % (w)
... 
valid veggie 'carrot' found.

does that help?

regards,
Jeff
-- 
|| visit gfd <http://quark.newimage.com/> 
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>




From wolford at enews.nrl.navy.mil  Thu Jul  8 12:49:12 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Thu, 08 Jul 1999 12:49:12 -0400
Subject: Python and Fortran code
References: <7m23i3$ktr$1@bignews.shef.ac.uk>
Message-ID: <3784D688.A6541F39@enews.nrl.navy.mil>

Look at a combination of the SWIG tool (I can't remember where it's at but it
stands for Simplified Wrapper Interface Generator so you can search for it) and
a C wrapper for you fortran function. Basically, make it so you can call the
fortran from C. Then swig the C function and link the cstuff.o and fstuff.o with
cstuff_wrap.o (from SWIG) into a shared library then you can follow choice 2
from the other response to this post.

Scott

Jose L Gomez Dans wrote:

> Hi!
>         I have a quetion on using python to do some simple processing on
> some data I pre-process with some Fortran code. I wanted to be able to call
> this subroutine from within python, as it is already "tried and tested" and
> optimized. Basically, my subroutine looks like this:
>
>         subroutine ex1(exdata)
>         dimension exdata(256,200)
>         print*,'Doing some processing in FORTRAN'
>  C      We said we were doing some processing, so we might as well do it :)
>  C      Processing of exdata
>         return
>         end
>
>         This subroutine would need to get input from the user and would also
> need to print stuff out for the user to see several parameters as the
> software runs. I'd like to pass exdata as a Python object (NumPy array
> object), and then recover it with the results. I wonder if I could somehow
> do that compiling this subroutine as an object file and then calling it from
> python (not that i know how to do it, but with scilab, it works :D).
>
>         If it's of any use, I am using Debian GNU/Linux and g77. And using
> loads of complex numbers :))))
>
>         Thanks in advance!
>         Jose
>
> --
> Jose L Gomez Dans                       PhD student
>                                         Radar & Communications Group
>                                         Department of Electronic Engineering
>                                         University of Sheffield UK




From donn at u.washington.edu  Mon Jul 12 11:45:29 1999
From: donn at u.washington.edu (Donn Cave)
Date: 12 Jul 1999 15:45:29 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
References: <3787D36A.815B7C29@ssec.wisc.edu> 	<3787D8DA.3774AA7D@ssec.wisc.edu> 	<7m8p7t$o7s@dfw-ixnews10.ix.netcom.com> 	<3787FB51.12FE0048@ssec.wisc.edu> 	<7m923i$dem@dfw-ixnews5.ix.netcom.com> <14217.63931.75257.566266@weyr.cnri.reston.va.us>
Message-ID: <7md2ip$19q2$1@nntp6.u.washington.edu>

"Fred L. Drake" <fdrake at cnri.reston.va.us> writes:
| Aahz Maruch writes:
|> I think I understand how it happened: os.getcwd() is (probably
|> rightfully) documented there, so someone thought os.chdir() belonged
|> there, too.  AFAICT, that's the only mistake in that area of the docs,
|> though I could make a case that os.getcwd() should be moved (or at least
|> have a pointer in "F&D").
|
| Aahz,
|   Are you saying that (either of) these are in the wrong place?  The
| notion of "current directory" is tied only to the entry in the process 
| table; should it be documented otherwise?
|   The "Process Parameters" section contains functions with provide any
| direct access to the process table entry, both queries and changes, so
| I think it really is the right place for these two functions.
|   I can easily add a textual reference to the right place within the
| "Files and Directories" section, but you'll need to explain very
| clearly why the descriptions should be moved.

For what it's worth, me too:  current working directory is a process
setting.  It's relevant to files and directories, to be sure, but whoever
put chdir() and getcwd() in process parameters knew what he was doing.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From a.eyre at optichrome.com  Mon Jul 19 08:39:43 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 19 Jul 1999 12:39:43 GMT
Subject: [Q] Type checking...
In-Reply-To: <7mv000$sun$1@feed.teaser.fr>
References: <7mv000$sun$1@feed.teaser.fr>
Message-ID: <000401bed1e3$c70caa70$3acbd9c2@peridot.optichrome.com>

> class ClassA
> ...
> 
> 
> class ClassB :
>    def myMethod(self, anInstance)
         if anInstance.__class__ != ClassA:
              raise "Oh no!"

This won't work for other classes derived from ClassA, however.
To do that, check out:

anInstance.__class__.__bases__

There may also be some cunning library hack to do this automatically.




From vaton at postoffice.pacbell.net  Mon Jul 26 17:38:15 1999
From: vaton at postoffice.pacbell.net (Rob)
Date: Mon, 26 Jul 1999 14:38:15 -0700
Subject: Python and Red Hat Linux 6
Message-ID: <379CD547.22A10A50@postoffice.pacbell.net>

When programming in Perl or Python, I am using a header like the
following on an executable file to cause the shell (bash or csh) to run
the scripts as Perl/Python files:
#!/usr/bin/perl   for Perl   and #!/usr/bin/python   for Python. I've
checked and the compilers are located in the appropriate directories. I
keep getting this when I run scripts:
bash: <script name>: command not found. Does Linux support the '#!'
notation in the file header, or do I always have to run the scripts by
explicitly typing in 'perl' or 'python' before the script name?
~Rob~




From Eric.Kidd at p98.f112.n480.z2.fidonet.org  Sat Jul  3 06:02:17 1999
From: Eric.Kidd at p98.f112.n480.z2.fidonet.org (Eric Kidd)
Date: Sat, 03 Jul 1999 11:02:17 +0100
Subject: Curses, Ncurses, Slang, etc?
Message-ID: <000016ad@bossar.com.pl>

From: Eric Kidd <eric.kidd at pobox.com>

I'd like to write a Python application which uses either curses, ncurses,
slang or a similar tty interface library. Unfortunately, the built-in
curses module is hopeless broken on my RedHat 6.0 system (it doesn't export
'addch', for example).

Neither the ncurses module or the pwin module work out of the box on my
Linux system. Are there any well-supported Python modules for
curses/ncurses/slang which do work out of the box?

Thank you for any help you can provide!

Cheers,
Eric




From news at dorb.com  Tue Jul  6 14:16:37 1999
From: news at dorb.com (Darrell)
Date: Tue, 6 Jul 1999 14:16:37 -0400
Subject: Python under Java?
References: <MrUJKLEB0KGWFViX4p@holmes.parc.xerox.com>
Message-ID: <bLrg3.6122$U6.39610@newsr2.twcny.rr.com>

Don't forget this one.
http://www.ndim.edrc.cmu.edu/dougc/jpi/Home.html

Sorry if this comes though twice. NT rebooted in the middle of things.
--
--Darrell
Bill Janssen <janssen at parc.xerox.com> wrote in message
news:MrUJKLEB0KGWFViX4p at holmes.parc.xerox.com...
> Has anyone tried embedding Python under Java, using JNI?  When I try it
> (with JDK 1.2 on Solaris 2.6 and Python 1.5), I get SEGV inside
> Py_Initialize().
>
> Bill





From Isao.Akiyama at unisys.co.jp  Mon Jul 19 05:56:25 1999
From: Isao.Akiyama at unisys.co.jp (Isao AKIYAMA)
Date: Mon, 19 Jul 1999 18:56:25 +0900
Subject: Signature 0.1 does not work well
Message-ID: <3792F649.36C69183@unisys.co.jp>

Hi,

I tried Signature 0.1, but it did not work well :-(

T:\>c:\progra~1\python\python Signature.py
ordinary arglist: ('x', 'y', 'z')
special_args:
Traceback (innermost last):
  File "Signature.py", line 204, in ?
    print "special_args:", f.special_args()
  File "Signature.py", line 121, in special_args
    if ((self.func.func_code.co_flags & POS_LIST) and
NameError: POS_LIST

POS_LIST and KEY_DICT must be declared in the global scope or modified
by 'Signature' in special_args().

Cheers,

Isao AKIYAMA




From andy at robanal.demon.co.uk  Tue Jul 27 15:42:44 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Tue, 27 Jul 1999 19:42:44 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <7nhseu$gf0$1@nnrp1.deja.com> <al8d7xen311.fsf@sirppi.helsinki.fi>
Message-ID: <37a60ac7.6444596@news.demon.co.uk>

Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:

>aaron_watters at my-deja.com writes:
>> In article <al81zdxkktf.fsf at sirppi.helsinki.fi>,
>>   Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
>> > Even in websites' case, I would worry; Python doesn't use multiCPU
>> very
>
>Single select/update's on Solid or MySQL with Python interface do easily
>1000/second on trivial tables. (on most boxen)
>
>-Markus Stenberg
True, but part of the point of Gadfly is that you can so easily put it
in places you can't put the others - like inside your client app
instead of on a remote server.

Comparing like-for-like, Python is inevitably slower.  But Python
often makes it very easy to cache data locally, and try out totally
different architectures for a given business problem which can more
than redress the balance. in some cases.

Regards,

Andy




From Sven.Drescher at dlr.de  Tue Jul 20 05:51:56 1999
From: Sven.Drescher at dlr.de (Sven Drescher)
Date: Tue, 20 Jul 1999 11:51:56 +0200
Subject: popen3 not works with xterm???
Message-ID: <7n1g3h$pou@news.go.dlr.de>

Hi!

I have a question to the following situation.
I wrote a little C-programm to test the popen3() class.
That works fine. I write into the pipe and I can read from the pipe, what my
programm wants to print to stdout.
But now I thought it's also possible with the xterm. But nothing happens.
There was only a comment: "pipe was broken - error", but with my C-programm
it works errorless.
What did I wrong???

Thanks for comments and help!!!

Sven





From jblake at stamp-coin.com  Fri Jul 30 04:42:39 1999
From: jblake at stamp-coin.com (Jonathon)
Date: Fri, 30 Jul 1999 08:42:39 +0000 (UTC)
Subject: [Tutor] I want to learn to hack
In-Reply-To: <14240.48630.244929.427876@dolphin.mojam.com>
Message-ID: <Pine.LNX.4.04.9907300840420.21236-100000@natal.stamp-coin.com>

On Thu, 29 Jul 1999, Skip Montanaro wrote:

>    Skip> Eric, just think how much more they'll be hacking in Python

	Think of that army slogan:  Be all you can be.

	xan

	jonathon

		




From fredrik at pythonware.com  Tue Jul 13 14:39:28 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 13 Jul 1999 18:39:28 GMT
Subject: MySQL Module
References: <19990713122901.A2538@pooh.frostnet.net>
Message-ID: <00fb01becd5f$0af777d0$f29b12c2@secret.pythonware.com>

> Are than any modules for connecting to a MySQL sever? I didn't see
> anything in the contributed section of python.org. (If there aren't, what
> other databases do people think are better, why?)

http://starship.python.net/crew/adustman/MySQLdb.html
(found via www.python.org => search => site search "mysql")

</F>




From skip at mojam.com  Thu Jul 22 01:10:09 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 22 Jul 1999 00:10:09 -0500 (CDT)
Subject: Maintainability (was Re: Stackless & String-processing)
In-Reply-To: <7n675p$qsk@dfw-ixnews7.ix.netcom.com>
References: <001101bed3f2$5e8874a0$642d2399@tim>
	<7n675p$qsk@dfw-ixnews7.ix.netcom.com>
Message-ID: <14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>

>>>>> "Aahz" == Aahz Maruch <aahz at netcom.com> writes:

    Aahz> Tim Peters <tim_one at email.msn.com> wrote:

    >> That's fine for a one-man show.  "Maintainability" in my life cuts
    >> across programmers, and it's Not Good for each person to reinvent
    >> their own basic machinery -- even to have the ability to do so.  One
    >> reason I like Python is that it *doesn't* let the jerk down the hall
    >> invent their own control structures <0.5 wink>.

    Aahz> Yup.  Today I had to lay down the law to our programming team: use of
    Aahz> the construct "from foo import *" is now expressly, explicitly, and
    Aahz> completely forbidden.  "from foo import bar" is deprecated; "from foo
    Aahz> import foo" is acceptable; and "import foo" is strongly preferred.

    Aahz> I'd be in favor of getting "from foo import *" entirely out of the
    Aahz> language, but I suspect it's just too convenient for short hacks.

What Tim said about programmers not inventing control structures is somewhat
different than what Aahz put his foot down about.  There's nothing wrong
with "from foo import *" when it's used in the right context.  Some modules
in the standard distribution were designed specifically to support this
(types and Tkinter come to mind).  Used willy-nilly, you can get into
trouble with it, one-man show or not.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From jmaskell at nospam.saskpublishers.sk.ca  Tue Jul 20 18:02:18 1999
From: jmaskell at nospam.saskpublishers.sk.ca (Jason Maskell)
Date: Tue, 20 Jul 1999 16:02:18 -0600
Subject: Python suitability questions
References: <0CCE781E44F2C872.F528377A62B41DCA.73CAA63AD6E79314@lp.airnews.net> <7n2a9p$trq$1@news.accu.uu.nl>
Message-ID: <C8A42E1E0DB3D792.D5899D7736E26702.F072FD528818FB27@lp.airnews.net>

Martijn Faassen <m.faassen at vet.uu.nl> wrote in message
news:7n2a9p$trq$1 at news.accu.uu.nl...
> Jason Maskell <jmaskell at saskpublishers.sk.ca> wrote:
> [snip]
> > and the scripting language has to be comfortable being completely
> > embedded in my C++ framework,
>
> There are various ways to integrate Python with C or C++ code. You can
> hand code things of course, and I've heard Python's extending/embedding
> API is quite nice. You can also use various tools for it. SWIG is very
> handy, having toyed with it myself (www.swig.org).

Ok, will take a look.

> > I know that Python can handle the OO stuff,
> > but is there any kind of internal
> > object messaging, or do I roll my own?
>
> What do you mean by an 'internal object messaging'? What I do know that if
> you decide to roll your own you'll be fine with Python -- it's very
dynamic
> that way.

Ah yes, I thought I might have been unclear. Well, to use MtG for an example
once again, there is a card called manabarbs that causes 1 pt of damage to
any player who taps a land for mana. So, when the player taps a land for
mana, the land will send a message or two saying that it has been tapped,
which the dispatcher will then send to the appropriate cards, which will
then cause the damage. Thinking about it some more, I see that I will have
to write my own manager/dispatcher. Not a big hardship,  I just wasn't
thinking that clearly this morning (surprise).

> > How flexible can I lay out my source,
> > is it as stiff as say, C?
>
> This depends on what you mean. Python uses whitespace to indicate block
> structure, so you could see that as 'strict' (though not actually more

Actually, not worried so much about indents and braces and all that, just
thinking that I might have to have some ungodly source file organisation
where the cards are in one or more text files, and then one or two other
text files for various other things. I guess I'm just more interested in
whether or not I can point the interpreter to an arbitrary chunk of code and
 let it do it's thing, or what exactly is involved. I'll do some more
research. ;>

> Of course I don't want to say solid design of classes isn't necessary;
it's
> still good to do.

Ya, good OO design will be fairly essential in this project.

>
> Good luck!
>

Thanks a lot. ;>

Jason






From billtut at microsoft.com  Tue Jul  6 18:37:00 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Tue, 6 Jul 1999 22:37:00 GMT
Subject: popen2 on Windows
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90EC5@RED-MSG-50>


> From: b s [mailto:balaji at platinum.com]

> All these things will work only if your application is a 
> console application.

That's true of os.popen(), but this is certainly not true of
win32pipe.popen*(). win32pipe.popen* will work from a console application,
an NT service, or from within PythonWin itself. (a GUI application).

Bill



From neelk at brick.cswv.com  Tue Jul 20 20:00:56 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 20 Jul 1999 19:00:56 -0500
Subject: Signature 0.1 does not work well
References: <3792F649.36C69183@unisys.co.jp>
Message-ID: <7n32jo$9qu$1@brick.cswv.com>

In article <3792F649.36C69183 at unisys.co.jp>,
Isao AKIYAMA  <Isao.Akiyama at unisys.co.jp> wrote:
>Hi,
>
>I tried Signature 0.1, but it did not work well :-(
>
> [Error snipped]
>
>POS_LIST and KEY_DICT must be declared in the global scope or modified
>by 'Signature' in special_args().

I changed those two variables from module globals to class variables,
and then forgot to test make sure I changed everything. You can either
download version 0.1.1 from my website, you can change the references
to "POS_LIST" and "KEY_LIST" in Signature.special_args() to
"self.POS_LIST" and "self.KEY_LIST" in your current version.

Let me know if this doesn't clear up your problem.


Neel
http://www.sff.net/people/neelk/free-software/



From wtanksle at dolphin.openprojects.net  Sun Jul  4 17:35:43 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Sun, 04 Jul 1999 21:35:43 GMT
Subject: Tuples cf. Expanded objects??
References: <377bbde1@news.xtml.co.uk> <7lh375$5ea$1@news.udel.edu> <377CAC46.508D502F@iname.com> <slrn7nprts.3a4.wtanksle@dolphin.openprojects.net> <377F4F8B.10C82257@iname.com>
Message-ID: <slrn7nvktf.hlt.wtanksle@dolphin.openprojects.net>

On Sun, 04 Jul 1999 22:11:55 +1000, Gary Stephenson wrote:
>William Tanksley wrote:

>> On Fri, 02 Jul 1999 22:10:46 +1000, Gary Stephenson wrote:

>> An instance of an immutable type may be created in expanded form.

>Yes, but can they also  be created in an non-expanded form?  If not, then
>there is more of a relationship that your message implies - yes?

Of course they can -- that's how Eiffel works by default and how Python
works no matter what.

I wonder whether you know what "expanded" form means, though.  It means
that the object itself is stored in local storage, rather than a reference
to the object.  Of course, it's obvious that location of storage has
little to do with immutability of type (although it can take a lot of work
to store certain mutable types in local storage).

>gary

-- 
-William "Billy" Tanksley



From gmcm at hypernet.com  Mon Jul 26 15:59:02 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 26 Jul 1999 14:59:02 -0500
Subject: Embedding python/multi-thread env
In-Reply-To: <933013656.237.57@news.remarQ.com>
Message-ID: <1279107155-5981362@hypernet.com>

Rich Salz wrote:

> I believe that you cannot create multiple interepreters in a single
> executable. The python global lock is to prevent *threads* within
> the interpreter from stomping on each other's critical sections.

Sure can; it's part of the API. They still share the global lock, but 
they can have completely different sets of modules loaded, etc. 

- Gordon



From mithy at my-deja.com  Wed Jul  7 13:29:19 1999
From: mithy at my-deja.com (Preston Landers)
Date: Wed, 07 Jul 1999 17:29:19 GMT
Subject: hilighting 'self' (was: Re: total idiot question: +=, .=, etc...)
References: <000015b4@bossar.com.pl> <76u2rha5qr.fsf@four.npl.uiuc.edu> <14211.28251.23566.522880@anthem.cnri.reston.va.us>
Message-ID: <7m02ot$ise$1@nnrp1.deja.com>

In article <14211.28251.23566.522880 at anthem.cnri.reston.va.us>,
  bwarsaw at cnri.reston.va.us (Barry A. Warsaw) wrote:

> You could probably add "self" to python-font-lock-keywords.  I'm not
> entirely sure this is a good idea since self is just a convention, but
> if a lot of people would like to see this, I might consider adding it.

Well, even though self is a convention, its use is 'strongly encouraged'
and I don't think anybody would be terribly confused by Emacs
highlighting it.  Maybe if this was added to python-mode it could be
optional.

My personal opinion is that "self" is much more 'like' a type than a
keyword (like if, elif, for, and so on.)  I decided to color it like a
class declaration.

I inserted these lines into my python-mode.el at the appropriate place:

     ;; also color "self" as a 'type'
     '("\\b\\(self\\)[ .,]?"
       1 font-lock-type-face)

Sjoerd Mullender <sjoerd.mullender at oratrix.com> suggests adding this to
your .emacs instead:

(add-to-list 'after-load-alist
	     (cons "python-mode"
		   '((add-to-list (quote python-font-lock-keywords)
				  (quote ("\\b\\(self\\)[ .,]?"
					  1 font-lock-type-face))))))

which is a slightly more clever way of doing it. I'm not a terribly
experienced Emacs hacker.

--
||  Preston Landers <mithy at my-deja.com>  ||
||        http://askpreston.com          ||


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From emile at fenx.com  Fri Jul  2 11:02:15 1999
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 2 Jul 1999 08:02:15 -0700
Subject: Alternatives to pickle/class nesting
References: <Ch2f3.2457$VL2.321674@news.direcpc.com>
Message-ID: <Fw4f3.2458$VL2.322515@news.direcpc.com>

Well, I've spent the last couple of hours trying to piece a small sample
together that illustrates my problem.  Good news... No success!  ;-)
This probably means I've got some squirrelly code in the application
somewhere, or that I can restructure towards my sample and get things
going.

Comments-still-welcome-but-I'm-out-of-my-pickle-ly yr's

--

Emile van Sebille
emile at fenx.com
-------------------





From jugonzal at globalserve.net  Sat Jul  3 01:22:13 1999
From: jugonzal at globalserve.net (Juan M. Gonzalez)
Date: Fri, 02 Jul 1999 22:22:13 -0700
Subject: Waiting for an external process
Message-ID: <377D9E05.E4F91D2F@globalserve.net>

Is there a better way to determine when an external process has finished
than using
win32process.CreateProcess / GetExitCodeProcess ?




From jam at newimage.com  Fri Jul  2 11:29:33 1999
From: jam at newimage.com (jam)
Date: Fri, 2 Jul 1999 11:29:33 -0400
Subject: RFC822 & Text of Mail Message
In-Reply-To: <7likid$e23$1@nnrp1.deja.com>; from gcollins99@my-deja.com on Fri, Jul 02, 1999 at 03:07:03PM +0000
References: <7likid$e23$1@nnrp1.deja.com>
Message-ID: <19990702112933.A9007@toast.internal>

On Fri, Jul 02, 1999 at 03:07:03PM +0000, gcollins99 at my-deja.com wrote:
> 
> Greetings!
> 
>   Here is what I want to do but cannot find out how. I am using RFC822
> to parse out the headers and store them somewhere which works just fine.
> But I am also in need of getting the text of the incoming message. So
> what module would I use to also capture the message text. I see no way
> of doing this via rfc822 module since it's not designed to do such
> things. WOuld I use mimetools? And if so, how?
> 
> Thanks!
> 
> -Glen Collins
> 

hi.

take a look at 'mailbox.py' in the standard library distribution for an example
of how to do this.  it boils down to calling 'rewindbody' and the 'read'ing
the data like any normal file.

good luck ! ;)

regards,
J
-- 
|| visit gfd <http://quark.newimage.com/> 
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From nathan.froyd at rose-hulman.edu  Tue Jul  6 20:56:46 1999
From: nathan.froyd at rose-hulman.edu (Nathan Froyd)
Date: 06 Jul 1999 19:56:46 -0500
Subject: Multiple program defined types -- access from C?
Message-ID: <m3908tp99t.fsf@froydnj.laptop.rose-hulman.edu>

I'm writing a program that needs a lot of program-defined types that are
also exposed as python classes to the user. The problem is, I also need to
access the innards of a lot of these classes from C. Are there any elegant
ways to do this? It doesn't seem as though Python likes for class functions 
to be used from outside the C file they're defined in.
-- 
Nathan  |  froydnj at rose-hulman.edu  |  http://www.rose-hulman.edu/~froydnj/

Evolution is a million line computer program falling into place by accident.



From tim_one at email.msn.com  Thu Jul 22 23:26:17 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 22 Jul 1999 23:26:17 -0400
Subject: Dreaded Newbie Question
In-Reply-To: <K4tK+KAA26l3EwvK@epstein.co.uk>
Message-ID: <000701bed4bb$20388360$2c2d2399@tim>

[Dani Epstein]
[skipping questions well-answered already]
> ...
> For the moment though my grateful thanks goes to anyone who can tell me
> how to get a file's size, and where that info is to be found in the
> references.
>
> BTW, I'm running PythonWin 1.5.1

First download 1.5.2 (1.5.1 is more than a year old, and in ways that count
<wink>).

After that, ignore all the "os.stat" stuff you've been hearing and use

    os.path.getsize(filepath)

Or use os.stat if you're more comfortable with obscurity <wink>.

&-keep-a-printout-of-the-library-manual-next-to-the-toilet-ly y'rs  - tim





From coryalbrecht at hotmail.com  Thu Jul  1 16:37:39 1999
From: coryalbrecht at hotmail.com (Cory C. Albrecht)
Date: Thu, 01 Jul 1999 20:37:39 GMT
Subject: Compiling Python for Win32
Message-ID: <7lgjij$180bi_004@news.sentex.ca>

Hello all,

I'm trying to compile Python myself, so I can embed it in my own 
programmes. I can compile my own C/C++ files properly, but then linking 
with python15.lib (or with python15.dll to get a fresh import library 
right then and there) produces many undefined references such as the 
following:

Error! E2028: _object near * near spam_system( _object near *, _object 
              near * ) is an undefined reference
Error! E2028: _object near * near SpamError is an undefined reference
Error! E2028: PyMethodDef near SpamMethods[] is an undefined reference
Error! E2044: exported symbol W?spam_system$n(pn$_object$$pn$1$)pn$1$ 
              not found
Error! E2044: exported symbol W?SpamError$npn$_object$$ not found
Error! E2044: exported symbol W?SpamMethods$n[]$PyMethodDef$$ not found

I'm doubting that this is caused by different name-mangling between 
compilers (I use Watcom C/C++ 11.0) since otherwise my other programmes 
would never be able to use the various standard Win32 API DLLs in my 
<windows>\system32 directory.

What I would like is if somebody who has successfully built Python 1.5.2 
using MSVC++ could do the "Export Makefile" command and send me the 
result. A simple text list of the individual source files that make all 
of the Windows versions of the various python executables, libraries and 
DLLs, along with accompanying defines would also be welcome, and either 
will earn somebody my undying thanks. :-)

Thanks in advance, eh!

--
Cory C. Albrecht
http://www.sentex.ca/~cory/



From Thomas.A..Bryan at p98.f112.n480.z2.fidonet.org  Fri Jul  2 16:16:17 1999
From: Thomas.A..Bryan at p98.f112.n480.z2.fidonet.org (Thomas A. Bryan)
Date: Fri, 02 Jul 1999 21:16:17 +0100
Subject: Getting to know Python
Message-ID: <000016a0@bossar.com.pl>

From: "Thomas A. Bryan" <tbryan at arlut.utexas.edu>

Arnaldo wrote:

> In the mean-time I'm eager to have with the interactive interpreter.
> How do I access the shell from the interpreter to enter commands like
> ,clear, ls, cp, etc.
> Is it posible, how do I do that?

Yes.  os.system() or os.popen()
Of course, many of these commands are available in the standard 
Python library.

For exmple, os.listdir() and shutil.copy() for ls and cp.

Python 1.5.2 (#1, Apr 18 1999, 16:03:16)  [GCC pgcc-2.91.60 19981201 
(egcs-1.1.1  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import os
>>> retval = os.system('ls /')
bin  boot  dev  etc  home  lib  lost+found  misc  mnt  opt  proc
  root  sbin  tmp  usr  var
>>> cmd = os.popen('ls /')
>>> output = cmd.readlines()
>>> cmd.close()
>>> output
['bin\012', 'boot\012', 'dev\012', 'etc\012', 'home\012', 'lib\012', 
'lost+found\012', 'misc\012', 'mnt\012', 'opt\012', 'proc\012', 
'root\012', 'sbin\012', 'tmp\012', 'usr\012', 'var\012']

If you're playing in the interpreter, learn to use the doc strings 
for the modules that have them:

>>> print os.system.__doc__
system(command) -> exit_status
Execute the command (a string) in a subshell.
>>> print os.popen.__doc__
popen(command [, mode='r' [, bufsize]]) -> pipe
Open a pipe to/from a command returning a file object.

To find out what's available, do a dir(os).

You may also want to read through/download the library reference from 
www.python.org

Here's a handy function...you'll probably want to modify it so that 
it writes to a file unless you have a lot of "history" on your terminal
window.

def get_docs(modulename):
    for el in dir(eval(modulename)):
        try:
            print el
            exec('print %s.%s.__doc__' % (modulename,el))
        except: #catch all exceptions
            print '%s has no __doc__' % el

#for example
import string
get_docs('string')



From jfisher at are.berkeley.edu  Fri Jul  9 19:05:54 1999
From: jfisher at are.berkeley.edu (John Fisher)
Date: Fri, 09 Jul 1999 23:05:54 GMT
Subject: Ref counts: bug? feature? gotcha?
Message-ID: <7m9tn6$6qg$48@rks1.urz.tu-dresden.de>

Hey folks,

I just tracked down a bug in a program of mine and, realizing what
caused it, I've begun to wonder if it really should have been an error.

In a nutshell, a C extension did some work and packed the results into a
NumPy array to return to Python.  It went something like this:
	return Py_BuildValue("OOOlld", PyArray_Return(array1),
			     .... );
The other two arrays were returned the same way as the first; the three
other arguments are inconsequential.  The error came once back in the
interpreter: any attempts to access any of the arrays resulted in a seg
fault.

My solution was to manually Py_INCREF() each array, and then change the
return to:
	return Py_BuildValue("NNNlld", PyArray_Return(array1),
			     .... );

This worked fine.  But I'm unclear on some things, and would love a bit
of insight from those more experienced than I.

First... why was the change necessary?  Py_BuildValue() increments the
ref count for objects declared "O" in the format string.  The arrays
going into PyArray_Return() were not improper in any form; presumably
PyArray_Return() simply spat back a duplicate pointer.  I fail to see
why Py_BuildValue() did not increment the desired object's ref count.

Secondly, was it a correct, or safe, change to turn the "O"s into "N"s
in the format string?  I'm assuming that the pointers given by
PyArray_Return() were simply duplicates, and thus I don't want to
increment their reference counts again -- but assuming the pointers were
duplicates is what got me into trouble in the first place.  Yet it seems
to work here.

Anyway, has anyone had similar problems before?  It seems like a bit of
a gotcha in the API that might bear documenting.  Or maybe I was just
slow.  Heh.

Thanks!
John


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From fredrik at pythonware.com  Mon Jul 26 14:38:27 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 18:38:27 GMT
Subject: createfilehandler, how to use it with Tk?
References: <7nht1s$grn$1@nnrp1.deja.com> <012c01bed77a$d562b370$f29b12c2@secret.pythonware.com> <7ni8ga$p0k$1@nnrp1.deja.com>
Message-ID: <00cc01bed796$0d8b6ce0$f29b12c2@secret.pythonware.com>

algaba at my-deja.com wrote:
> asyncore is a new module ? I can't find it in my python 1.5.1

then why are you using 1.5.1? ;-)

asyncore and friends are shipped with 1.5.2,
but you can get a stand-alone distribution from
http://www.nightmare.com/software.html

</F>




From nikolai.kirsebom at siemens.no  Thu Jul  1 07:18:23 1999
From: nikolai.kirsebom at siemens.no (Nikolai Kirsebom)
Date: Thu, 01 Jul 1999 11:18:23 GMT
Subject: Python, ASP - Namespace (again)
References: <O18e3.449$U%.3201@news1.online.no> <7lbj2t$2es$1@m2.c2.telstra-mm.net.au>
Message-ID: <377b4c67.12387312@news.mch.sni.de>

On Wed, 30 Jun 1999 09:00:20 +1000, "Mark Hammond"
<MHammond at skippinet.com.au> wrote:

>There is no obvious reason to me why this wont work.  Note that if you are
>not running win32all-125, then I recommend you upgrade - there have been
>some fixes (but none that adddress this, as far as I can tell).
>
>The best I can suggest is that you use the win32trace module and print
>statements to work out what the hell is going on.
>
>Eg, making your first line "import win32traceutil" will mean that all print
>statement go to a debugging collector.  You can view the print statements by
>starting Pythonwin on the same machine, and selecting "Tools/Trace Collector
>Debugging Output".
>
>Then use lots of print statements - specifically print dir() etc, to see
>exactly what the namespace is...
>
>Mark.
>

Mark, thanks for hint on using the trace utility.

The next time I checked the code written in my first message, it
worked ok !!!!

As I indicated in the message, I have one ASP file and two PY files.
Updating the ASP file (saving it) followed by a refresh in the browser
results in updated browser page content.  However, updating any of the
PY files does not result in any changes in the browser output.  Could
anyone answer when the PY modules are loaded.  I thought the PY
modules automatically are reloaded if the datestamp of the *.py file
is newer than the *.pyc file.

I'm using the statement:

from <module> import *

How (if possible) would I reload explicitly such a module ?


Nikolai




From tismer at appliedbiometrics.com  Wed Jul 21 15:06:39 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Wed, 21 Jul 1999 19:06:39 GMT
Subject: pow()
References: <37961668.FDF78B51@callware.com>
Message-ID: <37961A3F.1AD602E1@appliedbiometrics.com>


Ivan Van Laningham wrote:
> 
> Hi All--
> I couldn't find an explanation for the three-argument version of pow()
> on DejaNews, the Python 1.5.2 docs or the Python FAQ, so I'll ask now.
> 
> Can someone explain (in simple terms, for a bear of small brain) to me
> what the modulo argument to __pow__ is for?
> 
>         __pow__(self,other[,modulo])

This is in the docstring of the pow function.
If the result is to be modulo a modulus,
the computation might take advantage of it (I don't know wether it
does).

>>> print pow.__doc__
pow(x, y[, z]) -> number

With two arguments, equivalent to x**y.  With three arguments,
equivalent to (x**y) % z, but may be more efficient (e.g. for longs).
>>> 

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From robin at jessikat.demon.co.uk  Sat Jul  3 21:35:17 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sun, 4 Jul 1999 02:35:17 +0100
Subject: ordered threading
References: <TbLVhCAmDPf3EwPU@jessikat.demon.co.uk>
 <000a01bec583$30332b40$769e2299@tim>
Message-ID: <Y2tnuCAVprf3EwNj@jessikat.demon.co.uk>

In article <000a01bec583$30332b40$769e2299 at tim>, Tim Peters
<tim_one at email.msn.com> writes
>[Robin Becker]
>> I have an unbounded number of worker threads which occasionally require
>> access to a unique server thread. I can easily mutex access to the
>> server thread, but is there any obvious/easy way to ensure that the
>> blocked workers wake up in the right order?
>
>No threading system on earth allows specifying the order in which blocked
>threads will wake up, although some systems (not Python's) allow you to
>*influence* the order via assorted "priority" gimmicks.
I was being naive.
>
>You could easily enough build an OrderedLock class whose acquire and release
yes I guess that's what I'll try and do
>methods implement whatever scheme you want, hiding the unordered nature of
>an internal mutex under a layer of serial-number cruft.  But then why not do
>these things serially in a single thread to begin with?
the workers are HTTP requests in separate threads from Zope I guess. I
want to serialise access to the time/space hog. I guess it could run
uniquely in the worker threads ie the locked resource would be access to
the hog code. I'm not sure that would let me guarantee unlocking as
there are ways for the worker threads to suicide.
>
>If you have frequent rigid ordering requirements, threads are likely the
>wrong approach.
>
>> Also if blocked can a thread be killed in any way?
>
>No, not short of killing the whole program.  A thread blocked on a condition
>that will never become true is a symptom of confused application design, not
>of a weakness in the thread model <wink>.
>
>expand-the-condition-to-include-"or-i-should-die-now"-ly y'rs  - tim
>
>
>
>

-- 
Robin Becker



From news at dorb.com  Tue Jul 20 16:58:15 1999
From: news at dorb.com (Darrell)
Date: Tue, 20 Jul 1999 16:58:15 -0400
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk>
Message-ID: <mq5l3.1161$u74.20521@newsr1.twcny.rr.com>

> sreekant wrote:
> I know there is Pyapache. I am looking for it's binary. Is there any
> other way of doing it faster ?
>
Find a way to profile this system of yours.
That is find out where the time is being spent.
Then ask the more specific question about optimizing that section.

--
--Darrell





From Jonathan.W.Hendry at p98.f112.n480.z2.fidonet.org  Thu Jul  1 02:53:32 1999
From: Jonathan.W.Hendry at p98.f112.n480.z2.fidonet.org (Jonathan W Hendry)
Date: Thu, 01 Jul 1999 07:53:32 +0100
Subject: Installing Python/XML on Windows?
Message-ID: <0000162d@bossar.com.pl>

From: Jonathan W Hendry <jhendry at shrike.depaul.edu>

Does any documentation exist on this? How do I set up the
paths so that Python can use all the various components of
the package?

Thanks,

- Jon



From 71351.356 at compuserve.com  Tue Jul 27 20:36:21 1999
From: 71351.356 at compuserve.com (T. C. Mits)
Date: Tue, 27 Jul 1999 20:36:21 -0400
Subject: make or similar in Python
References: <378390DD.58CA6BD5@nevex.com>
Message-ID: <7nljh9$gfg$1@ssauraab-i-1.production.compuserve.com>

Hi again:

    There is also a commercial product that does this called OpenMake.
Haven't evaluated yet.

Greg Wilson <gvwilson at nevex.com> wrote in message
news:378390DD.58CA6BD5 at nevex.com...
> Hi, folks.  Has anyone ever written a dependency management/update tool
> (like 'make') in Python?
> I'm not looking for something that reads makefiles, and happens to be
> written in Python; I'm looking
> for a Python module that users import, then provide data to, in order to
> manage dependencies.  (The
> answer I got from the Perl crowd was 'cool idea, but no'; I'm hoping
> I'll have more luck here.)
> Thanks,
> Greg (who'd really, really like to override some of the things that
> GNU make does)
>
>





From BVStraalen at lbl.gov  Sat Jul 17 03:14:29 1999
From: BVStraalen at lbl.gov (Brian Van Straalen)
Date: Sat, 17 Jul 1999 07:14:29 GMT
Subject: Running 2 gui libs simultaneously
References: <1280251163-5463910@hypernet.com> <378BEC9C.271A69EB@compaq.com> <5lwvw1czzf.fsf@eric.cnri.reston.va.us>
Message-ID: <37901F54.D8E41EF1@lbl.gov>

Guido van Rossum wrote:
> 

> I don't think that's the problem, at least not on Unix.  Each GUI
> package has its own connection (a socket) to the X server, and the X
> server knows which connection to send events to.  No different that
> having two different apps.  If wxWindows exposes the socket and the
> raw event handling function, you can use Tk's createfilehandler to add
> wxWindows's socket to the set of sockets on which it listens (default
> only its own connection to the X server) and in the callback call
> wxWindows's event handler.  Theoretically this should work fine.
> 

Maybe not that easy.  The XServer sends events to ALL widgets that
claim to have a piece of real estate, the event struct also
tells it where that widget is in the 'application hierarchy'.  The 
problem is tk stuff inside a wxWindow, or vice-versa:  BOTH 
UI event loops get the event (handed off to them without
you asking..ha ha).  I had a similar tough time getting
the Java AWT and OpenGL talking to each other a year ago.

Don't remember how I got this working eventually, but
it was platform specific (ungh!, so much for Java).

Brian Van Straalen



From p.t at iname.com  Sat Jul 31 04:43:55 1999
From: p.t at iname.com (Peter Torstenson)
Date: Sat, 31 Jul 1999 05:43:55 -0300
Subject: How to find out the pid of a running program?
Message-ID: <7nucu2$11ko@enews1.newsguy.com>

I'd like tknow if there is a way of getting the pid of a program running,
using python commands?

/Peter Torstenson






From kistler at fnmail.com  Wed Jul  7 07:41:56 1999
From: kistler at fnmail.com (Per Kistler)
Date: Wed, 07 Jul 1999 13:41:56 +0200
Subject: GD/ImageMagick/PIL
Message-ID: <37833D04.8367425A@fnmail.com>

Hi All

I'm trying PIL right now, but was wondering, whether
there is a Module around the libgd as well. 
In Perl it's called GD. 
And: Does PIL all, that ImageMagick does, or does one
need both?

Thanks, Per.

PS: I'm on a PC with SuSE6.1 Linux (2.2.5)
    Python 1.5.2
-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net
------------------------------------------------------------



From pereira at research.att.com  Sat Jul 17 00:20:38 1999
From: pereira at research.att.com (Fernando Pereira)
Date: Sat, 17 Jul 1999 04:20:38 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com> <7mnirc$t2b$2@cronkite.cc.uga.edu>
Message-ID: <170719990020380651%pereira@research.att.com>

In article <7mnirc$t2b$2 at cronkite.cc.uga.edu>, Graham Matthews
<graham at sloth.math.uga.edu> wrote:

> Fernando Pereira (pereira at research.att.com) wrote:
> : but, as far as I know, nothing beats
> : the automata-theoretic optimization techniques for pattern matching and
> : parsing, of which the typical modern compiler for regular expressions
> : is just the best-known instance.
> 
> In what sense are you using the words "nothing beats". I am sure that
> the automata-theoretic approach produces the fastest pattern matchers
> and parsers. So on this sense "nothing beats them".
I made the sense clear. For regular and deterministic CF languages,
although my main use of finite-state techniques these days is in speech
and text processing, involving some less known and some new techniques
for combining and optimizing weighted finite-state transducers
<http://www.research.att.com/sw/tools/fsm/>. 
> But the class of 
> languages recognisable by automaton doesn't cover all the languages of
> interest (in fact my guess is it covers very few ... context ...).
In what sense of "very few"? Most programming languages are designed to
have an LR(k) or LL(k) backbone, and regular lexical structure (those
that aren't are cursed by their unhappy implementors). {S|HT|X}ML and
other Internet protocols obey similar or even stronger restrictions as
far as I know.

Naturally occurring sequences (text, speech, proteins, DNA, ...) do not
fit neatly into the standard formal language classes, and for those the
question is that of what language class can best approximate the string
distribution, which pushes one quickly to very restrictive language
classes (eg. bounded-context regular languages, or lexicalized
stochastic CFGs) learnable automatically from data. 
> So
> in this sense lots of things "beat" the automata approach.
Opportunities for optimization => efficiency. Crucial in text and
speech indexing, Web search, biological sequence processing, event log
analysis, to name just a few.
> And then
> there are software engineering issues ...
Agreed. However, more traditional lexer- and parser-generators had
sensible answers to many of those issues, that might not be as elegant
as monadic approaches, but that paid serious attention to optimization
in the regular and CF backbones. By the way, I find it faintly amusing
to see the functional programming community rediscovering and dressing
in categorial garb a simple idea that was a main motivator for the
development of logic programming in the early 70s (see
<http://akpublic.research.att.com:9000/~pereira/bib.html#Pereira+Warren-
80:DCGs> for a survey). It *is* a good idea for rapid prototyping, but
interspersing pattern matching/parsing with arbitrary computations
(even purely functional ones) effectively blocks optimization, which is
required in large-scale applications.

-- F



From gregm at iname.com  Tue Jul 20 00:50:34 1999
From: gregm at iname.com (Greg McFarlane)
Date: Tue, 20 Jul 1999 14:50:34 +1000
Subject: Tk: Temporarily change cursor?
In-Reply-To: <3793AADF.5EDBAFBC@compaq.com>; from Greg Ewing on 20 Jul 1999 at 10:46:55AM
References: <3793AADF.5EDBAFBC@compaq.com>
Message-ID: <19990720145034.04299@nms.otc.telstra.com.au>

As far as I am aware, there is no (easy, reliable) way to do what you
want (create a busy cursor) using Tk.  The Blt extension to Tk
provides the "busy" Tcl command to do this.  You will need to install
Blt and then you can access this from python.  Pmw provides a python
interface to the Blt busy command.

    Blt: http://www.tcltk.com/blt/
    Pmw: http://www.dscpl.com.au/pmw/

On 20 Jul, Greg Ewing wrote:
> Is there an easy way to temporarily change the
> cursor in Tk while a piece of code runs? Or
> even a hard way that always works?
> 
> There doesn't seem to be any way to directly
> change the cursor currently displayed -- you
> have to change the cursor of a particular widget.
> 
> I've had some success with changing the cursor
> of a widget and then doing an update, but it
> only works reliably if I pick exactly the right
> widget, which is difficult, especially if the
> spot in the code where I want to do this is
> far removed from the user action which
> initiated it.
> 
> This is such a commonly done thing, surely
> there must be some provision for it in Tk?
> 
> Windows, Python 1.5.2 and Tk 8.0, by the way.
> 
> Thanks for any help,
> Greg
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From tim_one at email.msn.com  Tue Jul 20 02:32:00 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 20 Jul 1999 02:32:00 -0400
Subject: How fast can we multiply?
In-Reply-To: <37936AE4.810F2152@appliedbiometrics.com>
Message-ID: <001801bed279$92c348a0$139e2299@tim>

[Christian Tismer]
> ...
> Also I was tempted to use "&" on the numbers, in order to save
> one nbits call. It was just unclear what happens with negative
> numbers and dropped it.

It's clear enough what happens with positive numbers, though <wink>:

m = 1L << 10000
n = m - 1
print m & n  # prints 0

Use "|" instead and you're golden.  But that takes time proportional to the
number of bits in the larger input, so-- again --we need to fix nbits.  Then

nx = nbits(x)
ny = nbits(y)
if nx == 0 or ny == 0:
    return 0

would be constant-time.

> Was happy that negatives went correct without special treatment.

Yup!  It's a delight.

> ...
> [nbits with hex...]
>
> brrr. I can't help it. This should really be a quick log2()
> call, this seems quite natural to me.

It should be constant-time.

> ...
> At the same time, the reacton of the list was a little
> disappointing to me. There seems to be near to zero interest.
> Is this the hot summer? Holidays? Bad title of my post?

Na, I think "little interest" is it.  Outside of crypto, prime-number fans,
and people studying Calendars from Hell <wink>, nobody cares about really
big ints.  The first two groups-- who care a lot --care *so* much they're
using hyper-optimized bigint pkgs anyway.  Andrew Kuchling's wrapping gmp
for Python is the most useful thing that's ever been done in this area.

> Then I also had a look into NumPy and its matrix multiplication.
> I stunned. Again no attempt to optimize big O. It could be done
> with Python, the algorithms are known, the fruit is waiting
> to be picked.

Large dense matmult isn't very common.  Sing along:

Lemon tree, very pretty,
And de lemon flower is sweet.
But de fruit of de poor lemon,
Is impossible to eat.

> Challenge: I claim Python could be a competitor to Mathematica,
> if only some people were interested.

Give me one bored programmer with a few weeks to spare, and you can keep a
whole web full of people who are interested <wink>.

> *** Hey Python, where are your ambitioned Math students? ***
>
> (did they wake up, finally?)
>
> waiting-for-the-beginning-to-show-up-ly y'rs  - chris

i'll-join-you-in-the-bar-ly y'rs  - tim





From Jean-Michel.Bruel at univ-pau.fr  Thu Jul 22 05:16:13 1999
From: Jean-Michel.Bruel at univ-pau.fr (Jean-Michel BRUEL)
Date: Thu, 22 Jul 1999 09:16:13 GMT
Subject: [CFP:] OOPSLA'99 Workshop
Message-ID: <199907220916.LAA12476@crisv4.univ-pau.fr>

[apologies if you receive multiple copies of this announcement]

=================================================================
                               Call for Papers
=================================================================

                            OOPSLA'99 Workshop on 

             Rigorous Modeling and Analysis with the UML: Challenges
             			and Limitations
             			
             			Workshop #23

                               DENVER, COLORADO

                          Tuesday, November 2nd, 1999
                        ------------------------------

                 workshop website: http://www.univ-pau.fr/OOPSLA99

                                  - o 0 o -

WORKSHOP OVERVIEW
==================
The UML is a standard object modeling language that is 
targeted at large, complex systems. The high quality experiences 
embedded in the UML certainly makes
its application to complex systems desirable, but the lack of precise
semantics that support rigorous semantic analyses can limit its effectiveness.
The modeling of complex systems requires the use of 
modeling constructs that provide complexity management mechanisms and that
allow for the early detection of errors in requirements and designs.
The separation of views principle has proven to be an effective means of
controlling complexity and is well-supported by the UML. On the other hand,
the formality and rigor principle that facilitates the detection of errors in
requirements and designs is not well-supported in the UML.

The purpose of the proposed workshop is to bring together 
researchers and practitioners from academia and industry to 
discuss the semantic foundations of the UML as described in 
the OMG-UML documentation (current version is 1.3).
Presentations and discussions will focus on identifying deficiencies
in the UML semantics foundation, analyzing proposed approaches
to addressing the deficiencies, understanding UML limitations and
identifying the challenges that will have to be met in making the
UML a rigorous modeling notation. Particular attention will be paid
to the meta-modeling approach to semantics and to the OCL.

TOPICS
=======
The workshop topics include:
* Semantic foundations for UML static and dynamic models
* Use of meta-models to express semantics
* Semantic foundations for the OCL
* Using the OCL to support rigorous reasoning
* UML support for rigorous development of software (including
support for the use of software architectures and frameworks)

FORMAT
=======
This workshop is the second of a planned series of workshops on
strengthening the UML semantic foundation. The first workshop was
held at ECOOP'99 (see
http://www.cs.ukc.ac.uk/people/staff/sjhk/detail/ecoop99/UMLsemanticsFAQ/index.html).
The first workshop focused primarily on the utility of having precise
semantics. In the OOPSLA'99 workshop the focus will be on the semantic
foundations of the UML notations.

The workshop will consist mainly of working groups that
focus on a specific problem. There will be at least three invited
presentations that highlight UML deficiencies and proposed development
paths for the UML. The presentations, given in the first
hour of the workshop, will provide some of the
context for much of the discussions in the working groups. The
working groups will be determined by the submissions to the workshop
and will be announced at least 1 week before the workshop starts. 


SUBMISSIONS & PARTICIPATION
============================
Attendance will be by invitation only. If you would like to be considered 
for participation you are asked to prepare a statement outlining your
perception of the UML semantics and the use of the UML for
rigorous modeling. Statements on the following topics are especially
welcome: 
(1) deficiencies in the UML semantics, 
(2) expressing semantics using meta-models, 
(3) OCL semantics, 
(4) analysis of OCL expressions (including
proposals for automated support for analysis), 
(5) using the UML for rigorous specification and analysis, 
(6) providing automated support for rigorous analysis of UML models.

Send submissions via email to oopsla99 at in.tum.de. Submissions
should not exceed 2 pages (10 pt., single space) and
should be in one of the following formats:
* Plain text
* Postscript
* .pdf formats
Accepted submissions will be placed on the workshop website 
(http://www.univ-pau.fr/OOPSLA99).

IMPORTANT DATES
================
Submission deadline:     September 10th, 1999
Notification date:       October 1st, 1999
Workshop date:           November 2nd, 1999

ORGANIZERS
===========
This workshop is organized by members the Precise UML (pUML) group
(http://www.cs.york.ac.uk/puml).

Workshop Coordinators
----------------------
Robert France, Organizing Chair
Department of Computer Science
Colorado State University 
Fort Collins, CO 80523-1873.
phone: 970-491-6356 
fax: 970-491-2466
Email: france at cs.colostate.edu

Bernhard Rumpe, Program Co-Chair 
Faculty for Computer Science
Technische Universit?t M?nchen 
D-80333 Munich, Germany.
Phone:   +49 -89 -289-28129
Fax:     +49 -89 -289-28183
Email:  rumpe at in.tum.de

Brian Henderson-Sellers, Program Co-Chair
Director, Centre for Object Technology Applications and Research
    and Professor of Information Systems 
School of Computing Sciences 
University of Technology, Sydney, Australia.
phone: +61 (0)2 9514 1687
fax: +61 (0)2 9514 1807
Email: brian at socs.uts.edu.au

Jean-Michel Bruel, Publicity Chair & On-site Coordinator
Laboratoire T.A.S.C. 
Universit? de Pau et des Pays de l'Adour 
F-64000 Pau, France. 
phone: +33(0)5.59.92.31.89 
fax: +33(0)5.59.80.83.74
Email: Jean-Michel.Bruel at univ-pau.fr

Ana Moreira, On-site Coordinator
Departamento de Inform?tica
Faculdade de Ciencias e Tecnologia
Universidade Nova de Lisboa
2825 Monte da Caparica, Portugal
phone: +351-1-294 85 36 Ext. 0716 
fax: +351-1-294 85 41
email: amm at di.fct.unl.pt
       






From greg.ewing at compaq.com  Mon Jul 19 17:46:42 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 20 Jul 1999 09:46:42 +1200
Subject: [Q] Accessors...
References: <7mvbkj$pl$1@feed.teaser.fr>
Message-ID: <37939CC2.BD5A1AC9@compaq.com>

Olivier Deckmyn wrote:
> 
> This way, because it exists two methods get_XXX and set_XXX, a special
> proporty XXX would be implicitly created, and would make the user call
> "automatically" the get/set methods.

Yes, there is a way. If a class has a method
called __getattr__, it is automatically called
whenever you try to get the value of an attribute
which doesn't exist. So you can do something like

  class AutoAccess:

    def __getattr__(self, name, *args, **kw):
      return apply(getattr(self, "get_"+name), args, kw)

then any class which inherits from AutoAccess will
have any attempt to get a non-existent XXX attribute
redirected to a get_XXX method. (Note that the above
implementation is a bit simplistic -- it will enter an 
infinite loop if there is no get_XXX method defined
in the class.)

There is a corresponding feature for setting an
attribute: if there is a __setattr__ method, it gets
called on every attempt to set the value of an
attribute, so you can do

    def __setattr__(self, name, value):
      try:
        setter = getattr(self, "set_"+name)
      except AttributeError:
        setter = None
      if setter:
        setter(value)
      else:
        self.__dict__[name] = value

This is more complicated because the __setattr__
method, if present, is called *before* checking
whether the attribute exists (by necessity, since
setting an attribute which doesn't already exist
is perfectly legal). Hence the code to catch the
case where there is no set_XXX method and set the
attribute directly.

As you can see, what you are asking for can be
done, but it is expensive -- attribute access
through these methods will be considerably slower
than just calling get_XXX and set_XXX methods
directly. You'll have to decide whether it's
worth the price in your application.

Warning: The above code is for illustration
purposes only -- I haven't tested it!

Greg



From drew.mcdowell at msfc.nasa.gov  Fri Jul  9 14:55:03 1999
From: drew.mcdowell at msfc.nasa.gov (Drew McDowell)
Date: 9 Jul 1999 18:55:03 GMT
Subject: Python Code Repository?
Message-ID: <378645AA.5430A026@msfc.nasa.gov>

After seeing the "Brensenham line algorithm" post above, I'm curious
if there is a public "python code repository" (something akin to CPAN)
out there?  Or does Starship assume that role?

-Drew



From larsga at ifi.uio.no  Tue Jul 20 14:56:29 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 20 Jul 1999 20:56:29 +0200
Subject: [Q] Accessors...
References: <7mvbkj$pl$1@feed.teaser.fr> <37939CC2.BD5A1AC9@compaq.com> <7n16ue$66u$1@feed.teaser.fr>
Message-ID: <wkso6j880y.fsf@ifi.uio.no>

Olivier,

Python has a number of these 'magic' methods, and they are very much
worth knowing about. The complete list and descriptions of them are
hiding in section 3.3 of the language reference, which is well worth
bookmarking.

<URL: http://www.python.org/doc/current/ref/specialnames.html >

--Lars M.



From jackklein at att.net  Thu Jul 15 00:07:31 1999
From: jackklein at att.net (Jack Klein)
Date: Thu, 15 Jul 1999 04:07:31 GMT
Subject: Embedding Python in C
References: <7mi084$r80$1@kopp.stud.ntnu.no>
Message-ID: <37915e4e.168305575@netnews.worldnet.att.net>

On Wed, 14 Jul 1999 14:36:14 +0200, "Thomas R?berg"
<thomar at ifi.ntnu.no> wrote in comp.lang.c:

> I am writing a program in C/C++ and want to embed Python. The one problem I
> encounter is that when I try to import a module using
> Pyrun_SimpleString("import mymodule\n"), this module is not found. I have
> specified the correct paths at the correct places in my Makefile (at least I
> think I have). Anyone had problem with this??
> 
> Thomas R?berg
> thomar at ifi.ntnu.no

#1.  There is no such language as C/C++.

#2.  Please leave comp.lang.c and comp.lang.c++ out of your cross post
list.  Neither of these languages defines any way to embed python.

Jack Klein
-- 
Home: http://home.att.net/~jackklein
New Stuff: http://home.att.net/~jackklein/newstuff.html
Updated 11-Jul-1999




From ivanlan at callware.com  Thu Jul  8 20:32:06 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 9 Jul 1999 00:32:06 GMT
Subject: [PSA MEMBERS] Using (physical) units and quantities
References: <19990707181324.5617.qmail@lannert.rz.uni-duesseldorf.de> <199907080950.LAA32357@chinon.cnrs-orleans.fr>
Message-ID: <37854306.1BB7FC20@callware.com>

Hi All--

Konrad Hinsen wrote:
> 

[snip]

> 
> BTW, the best place for announcing Python modules is the Usenet
> group comp.lang.python.announce or its e-mail list equivalent.
> 

But for those of us who don't get the newsgroup, or don't know about the
announce list, it doesn't hurt to *also* announce on python-list and
psa-members too. ...

<suspenders-and-a-belt>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From jepler at inetnebr.com  Tue Jul  6 10:13:36 1999
From: jepler at inetnebr.com (Jeff Epler)
Date: Tue, 06 Jul 1999 14:13:36 GMT
Subject: range(Instance) vs. foo[Instance]
Message-ID: <slrn7o43v5.qkm.jepler@craie.inetnebr.com>

Suppose I have a class with __int__ defined:

class c:
	def __init__(self, val):
		self.val = int(val)

	def __int__(self): return self.val

>>> x=c(1)
>>> range(x)
[0]
>>> (0,1,2)[x]
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: sequence index must be integer

It seems desirable that the subscript operator call x.__int__ just like
range() must effectively do.  Is there some reason why not?

(Not the subscript operator in general, but the subscript operator for
the builtin types of lists and tuples)

I ask because I recently wrote a 5-minute hack on top of bytecodehacks that
makes "integers" function as rational numbers in the interactive Python
console---functions like "range" work, but subscripting of tuples and lists
is broken.

Jeff



From aahz at netcom.com  Mon Jul 19 20:15:41 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 20 Jul 1999 00:15:41 GMT
Subject: python + bash + new user
References: <3793BE52.43E023BB@netcard.com>
Message-ID: <7n0f3d$1d7@dfw-ixnews10.ix.netcom.com>

In article <3793BE52.43E023BB at netcard.com>,
Joyce Stack  <joyce at netcard.com> wrote:
>
>I am not new to programming but new to python...i want to know how do i
>call functions from the prompt....

>From which prompt, python or bash?  The python prompt should look
something like this:

% python
Python 1.5.2 (#6, Jun 29 1999, 15:08:27)  [GCC 2.7.2] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>

You can then type any valid python command or expression at the ">>>"
python prompt.  For example:

>>> int("3")
3
>>> x=int("3")
>>> x*x
9

Note that commands do not print output, but expressions do.  And, yes,
startling as it is for a C programmer, "=" is *not* an expression but a
statement/command.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From ivanlan at callware.com  Sat Jul 17 09:42:34 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sat, 17 Jul 1999 13:42:34 GMT
Subject: I envy newbies
References: <000001bed00d$a9a70480$d2a22299@tim>
Message-ID: <3790884A.A58864B@callware.com>

Hi All--

Tim Peters wrote:
> 
> In 5 years of running 95 and NT, only one program (a virus scanner) has ever
> had the audacity to change my autoexec.bat.  This isn't DOS!  Well, OK, 95
> is.  But 95 has other mechanisms too, and if you install software that edits
> yours it must be old enough to vote <wink>.
> 

I just looked at my autoexec.bat.  Borland C++ modifies the path.  The
NetWare client modifies your path and sets an environment variable.  The
Creative Ensoniq install sets some environment variables.  Most installs
that change autoexec.bat ask your permission first, and make backups.

> Changing that file is dangerous, in part because bat files have secret
> line-length limitations, the installer can't possibly know where in the path
> to place a new directory (potential shadowing), and an uninstaller would
> have to be smart to undo its changes (e.g. blindly restoring the file it
> started from would be wrong).
> 

Usually, these things modify the path by simply adding a line like this:

	PATH=L:\BC5\BIN;%PATH%

which does get around the simple line length limitation.  The actual
length of the path variable in the environment seems to not suffer the
255 character limits that W3.1 had; I was pushing or exceeding that
limit with 3.1, but I have no such problems now.

> So modern a-go-go installations do the safe thing instead:  wildly overwrite
> every DLL in the System directory <wink>.
> 

Hmmm, yes.  I think the moral here is obvious.

> > Besides, I don't think IDLE will work if you don't have Tcl in your
> > path,
> 
> Poor Guido recently wrote FixTk.py (in Lib/lib-tk/) to try to guess where
> Tcl lives under Windows (the Tcl/Tk installer doesn't touch autoexec.bat
> either); it's invoked automatically as soon as IDLE tries to import Tkinter.
> 

I had to add it to my NT path environment variable before IDLE would
work.

If newbies are *ever* going to use Python, the path has to be correct.

> > and if you have to add that, why not add Python?
> 
> Because "24 Hours" is in the title of your book, and it can take a newbie
> twice that long just to recover from a botched edit of this critical shared
> system file <0.7 wink>.
> 

But there may be no alternative.

> The installation script is in (the source distribution's)
> PCbuild/python15.wse -- although reading that will increase your
> appreciation for root canals <wink>.  If you want to check it out, just edit
> your autoexec.bat then reinstall Python over your current installation.
> 

I shall.  I've got a new system that I can afford to mess up.

Does or does not Win32All modify your path?

> > ...
> > I still maintain that the installer *ought* to modify your path; really,
> > that should be safe enough,
> 
> Afraid not -- and that's why nothing else (written this decade) does it
> anymore either.
> 

Borland was written this decade--at least the install was.  Novell's
NetWare client has a spiffy new install.  The ancestry of these things
may be as old as you are, Timmy, but their installs are fairly new.

Hacking the autoexec.bat is safer than dumping DLLs into the system32
directory.  Have you tried installing AOL4.0 by downloading it
recently?  Not only does it overwrite all your TCP/IP DLLs with old gunk
that f**ks every other form of networking you've got, but IT WON'T
BELIEVE THAT IT HAS ALREADY SCREWED YOU UP, so it keeps on trying to
re-copy the bad DLLs to your system32 directory every time you re-boot.

> making-newbies-do-it-by-hand-cleverly-shifts-the-blame-for-the-
>     problems-onto-them-ly y'rs  - tim

Honestly, I can't tell them, "cd to your Python directory for every
exercise in this book."  I'd rather tell my readers to edit their
autoexec.bat than tell them how to cd to a directory with spaces in the
path.  Bleaugh;-(>

<shifting-the-blame-to-timmy-is-fun-easy-and-educational-too!>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From wtanksle at dolphin.openprojects.net  Thu Jul  1 15:20:36 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Thu, 01 Jul 1999 19:20:36 GMT
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> <wkzp1g2t2e.fsf@ifi.uio.no>
Message-ID: <slrn7nnfs3.19k.wtanksle@dolphin.openprojects.net>

On 01 Jul 1999 19:11:53 +0200, Lars Marius Garshol wrote:
>* William Tanksley

[I made the joking claim that Lisp was an improvement on Forth, then to
defuse criticism I claimed:]
>| (Yes, I know -- the ancestry goes the other way around: Lisp ->
>| Scheme -> Forth.)

>Huh? Forth dates back to the 60s, whereas Scheme is from 1975, and I'm
>quite unsure of whether Chuck Moore knew Lisp at all.

I don't think Forth is that old, certainly not in any usable form.  Here,
according to the history on www.forth.com, "the first program to be called
Forth was written in about 1970."  That substantiates what you're saying
about Scheme -- I'm a little suprised.  Okay, toss Scheme out of the
ancestry ;-).

Chuck Moore, the inventor of Forth, got a BA in physics from MIT and went
into grad school at Stanford.  He claims to have taken classes from and
learned a lot from Lisp.

Forth is very much like Lisp, with the subtraction of memory management
and the addition of implicit parameter passing.  Forth is Lisp for people
who prefer "as simple as possible" to "but no simpler."  (And it IS a very
good language; I use it more than Scheme.)

A paragraph like that is neccesarily an oversimplification.  Perhaps I can
get away with saying that Forth learned more from Lisp than it did from
any other language existing at the time?  Yes, I think that's tenable.

>--Lars M.

-- 
-William "Billy" Tanksley



From jfisher at are.berkeley.edu  Fri Jul  9 19:05:54 1999
From: jfisher at are.berkeley.edu (John Fisher)
Date: Fri, 09 Jul 1999 23:05:54 GMT
Subject: Ref counts: bug? feature? gotcha?
Message-ID: <7m5v8c$nmr$1@nnrp1.deja.com>

Hey folks,

I just tracked down a bug in a program of mine and, realizing what
caused it, I've begun to wonder if it really should have been an error.

In a nutshell, a C extension did some work and packed the results into a
NumPy array to return to Python.  It went something like this:
	return Py_BuildValue("OOOlld", PyArray_Return(array1),
			     .... );
The other two arrays were returned the same way as the first; the three
other arguments are inconsequential.  The error came once back in the
interpreter: any attempts to access any of the arrays resulted in a seg
fault.

My solution was to manually Py_INCREF() each array, and then change the
return to:
	return Py_BuildValue("NNNlld", PyArray_Return(array1),
			     .... );

This worked fine.  But I'm unclear on some things, and would love a bit
of insight from those more experienced than I.

First... why was the change necessary?  Py_BuildValue() increments the
ref count for objects declared "O" in the format string.  The arrays
going into PyArray_Return() were not improper in any form; presumably
PyArray_Return() simply spat back a duplicate pointer.  I fail to see
why Py_BuildValue() did not increment the desired object's ref count.

Secondly, was it a correct, or safe, change to turn the "O"s into "N"s
in the format string?  I'm assuming that the pointers given by
PyArray_Return() were simply duplicates, and thus I don't want to
increment their reference counts again -- but assuming the pointers were
duplicates is what got me into trouble in the first place.  Yet it seems
to work here.

Anyway, has anyone had similar problems before?  It seems like a bit of
a gotcha in the API that might bear documenting.  Or maybe I was just
slow.  Heh.

Thanks!
John


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From C.Kotsokalis at ece.ntua.gr  Wed Jul  7 07:53:27 1999
From: C.Kotsokalis at ece.ntua.gr (Constantinos A. Kotsokalis)
Date: 7 Jul 1999 11:53:27 GMT
Subject: GD/ImageMagick/PIL
References: <37833D04.8367425A@fnmail.com>
Message-ID: <slrn7o6fsq.e5.C.Kotsokalis@ppp032.dialup.ntua.gr>

On Wed, 07 Jul 1999 13:41:56 +0200, Per Kistler wrote:
>I'm trying PIL right now, but was wondering, whether
>there is a Module around the libgd as well. 
>In Perl it's called GD. 

Go to the python website, to the contributed modules page, graphics
section. There are two such modules there.

-- 
Constantinos A. Kotsokalis || C.Kotsokalis at ece.ntua.gr
National Technical University of Athens - GREECE
Electrical and Computer Engineering Department
"Bus error -- driver executed."



From m.faassen at vet.uu.nl  Mon Jul 19 04:53:08 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 19 Jul 1999 08:53:08 GMT
Subject: what is python?
References: <7mt81n$s6e$1@bgtnsc03.worldnet.att.net>
Message-ID: <7mup1k$195i$1@news.accu.uu.nl>

Paul Johnston <johnston.p at worldnet.att.net> wrote:
> Hello people.

Hi Paul, welcome to this newsgroup. I replied to your post on comp.lang.xml
(in a thread about Python) and now you're here, cool! :)

[snip]

> So what are people using python for?  What types of problems is python
> better suited relative to, say, java?  I do alot of server-side programming
> & xml stuff.

People use Python for lots of things. I haven't used Java myself (though I
have used C++ so I'd be up to speed with Java quickly enough), so it's 
hard for me to compare the two. I'll just describe what I do with Python:

I use Python at work for primarily two things:

* Scripting around databases (for conversion etc)

I've used Python a lot to manipulate databases (through ODBC). It's very
easy to write a script that talks to a database through SQL. Python is
good at working with data; its list and dictionary constructs make
life very easy. Python can also easily send off emails or talk to an
FTP server, so the scripts can fetch their data by themselves and mail
people when they're done.
 
* Web publishing with Zope

Something quite different (though one can talk to databases with Zope as well)
is Zope. Zope is a web application platform written in (mostly) Python, and
is flexible and powerful, see http://www.zope.org

I'm currently investigating how to best integrate some XML code I wrote
(in Python, using the XML-sig libraries) into Zope.

During play, I also use Python for:

* Interfacing with lots of neat stuff. GUIs, libraries for all kinds of things.

* Prototyping neat stuff. One can write something new very quickly, without
getting lost in designing interfaces for all your classes (I always get
stranded there in C++ -- designing classes can be fun, but it gets old
after a while if you don't get your application to actually do something :)

I'm planning some apps that will have everything not speed critical (which
is a lot; GUI, analysis, loading/saving) in Python, with only the core
in C or C++ for speed when necessary.

Regards,

Martijn




From booysejh at telkom.co.za  Tue Jul 13 06:06:01 1999
From: booysejh at telkom.co.za (Riaan Booysen)
Date: Tue, 13 Jul 1999 12:06:01 +0200
Subject: Running 2 gui libs simultaneously
Message-ID: <7mf2rn$dj9$1@news1.saix.net>

Hello all,

How difficult would it be to run 2 window libraries simultaneously (Tcl/Tk /
wxPython) in one Python program.
The problem is that both need an event processing loop ( lib.mainloop() ).
Are threads my only solution?

Riaan Booysen






From tim_one at email.msn.com  Sun Jul 18 23:36:43 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 18 Jul 1999 23:36:43 -0400
Subject: IDLE help 
In-Reply-To: <000c01becf52$e4198c20$2acb94d1@darylh.ppp>
Message-ID: <000101bed197$ebb7a2e0$f0a02299@tim>

FYI, Daryl Harms has contributed an excellent intro to IDLE, available from

    http://www.python.org/doc/howto/idle/

Pictures & everything.  Check it out!

enablingly y'rs  - tim






From jeremy at cnri.reston.va.us  Fri Jul 23 16:42:41 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Fri, 23 Jul 1999 16:42:41 -0400 (EDT)
Subject: Python documentation available as ZIP files
In-Reply-To: <14232.48512.648609.496768@lrz.de>
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
	<14232.48512.648609.496768@lrz.de>
Message-ID: <14232.54209.732073.516014@bitdiddle.cnri.reston.va.us>

>>>>> "EL" == Eugene Leitl <eugene.leitl at lrz.uni-muenchen.de> writes:

  EL> I think this really means lowering the standards. I wouldn't
  EL> advocate using a more efficient (but obscure) standard as bzip2
  EL> (at least not yet), but everybody who is unable to unpack
  EL> .tar.gz/.tgz archives (you did include a pointer to WinZip in
  EL> the documentation, didn't you?) is perhaps not the right person
  EL> to read Python documentation anyway.

(I hope you forgot a <wink> somewhere.)

There is no reason for snobbery.  If some people prefer Zip archives
to gzipped tar files, we should accomodate them.  Python is a good
programming language for everyone, not just people who haved wasted
time and effort to learn about various obscure packaging details.

Jeremy




From Dani at SpamsNotKosher.com  Thu Jul 22 19:47:12 1999
From: Dani at SpamsNotKosher.com (Dani Epstein)
Date: Fri, 23 Jul 1999 00:47:12 +0100
Subject: Dreaded Newbie Question
Message-ID: <K4tK+KAA26l3EwvK@epstein.co.uk>

Hi!

A newbie to Python (though currently coding in Delphi), I'm having a
little difficulty tracking down stuff, although I have the tutorial and
manuals. Where is the best place to get more info, and why is there no
newsgroup for newbies like myself?

For the moment though my grateful thanks goes to anyone who can tell me
how to get a file's size, and where that info is to be found in the
references.

BTW, I'm running PythonWin 1.5.1
-- 

Dani Epstein



From eugene.leitl at lrz.uni-muenchen.de  Mon Jul 26 03:27:04 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Mon, 26 Jul 1999 00:27:04 -0700 (PDT)
Subject: mutating python VM opcodes
Message-ID: <14236.3528.398388.925726@lrz.de>

What's Python's reaction when presented with random bytecode sequence?
Slight mutation of a valid bytecode sequence?

Core dump, error message, something else?



From ajung at sz-sb.de  Fri Jul 16 09:33:24 1999
From: ajung at sz-sb.de (Andreas Jung)
Date: Fri, 16 Jul 1999 13:33:24 GMT
Subject: [XML-SIG] sgmllib has problems with dots in tag names
In-Reply-To: <14223.12495.910694.301072@weyr.cnri.reston.va.us>; from Fred L. Drake on Fri, Jul 16, 1999 at 09:17:03AM -0400
References: <19990716125230.A14994@sz-sb.de> <14223.12495.910694.301072@weyr.cnri.reston.va.us>
Message-ID: <19990716153323.A23945@sz-sb.de>

On Fri, Jul 16, 1999 at 09:17:03AM -0400, Fred L. Drake wrote:
> 
>   Ok, I've poked at the standard sgmllib a bit to see what the problem
> is.  The parser is recognizing the start and end tags.  Once
> recognized, it is looking for the handler methods start_*() / end_*()
> or do_*().  Since there's a dot in the name, these methods are not
> defined, and the unknown_*tag() methods are called instead of the
> handle_*tag() methods.


>   It should be easy to override the unknown_*tag() methods to use a
> table-based dispatcher or performs some form or name mangling, then
> passes known tags through to the handle_*tag() methods or whatever.
> This seems to be the easiest way to deal with the situation in the
> short term.

That's a solution that works with the standard sgmllib from the Python
distribution. However this solution does not work with sgmllib from 
xml.parsers. I'm not sure if tag names with dots in their names are valid in XML 
or not. So this might explain the different behaviour however I don't think
that's the reason. Maybe I'll find the real reason over the weekend.

Andreas
 

-- 
                                _\\|//_
                               (' O-O ')
------------------------------ooO-(_)-Ooo--------------------------------------
   Andreas Jung, Saarbr?cker Zeitung Verlag und Druckerei GmbH
   Saarbr?cker Daten-Innovations-Center
   Gutenbergstr. 11-23, D-66103 Saarbr?cken, Germany
   Phone: +49-(0)681-502-1563, Fax: +49-(0)681-502-1509
   Email: ajung at sz-sb.de (PGP key available)
-------------------------------------------------------------------------------



From tismer at appliedbiometrics.com  Tue Jul 13 08:01:44 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 13 Jul 1999 12:01:44 GMT
Subject: making a Windows NT service out of a python program
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F20@RED-MSG-50>
Message-ID: <378B2AA8.58477120@appliedbiometrics.com>


Bill Tutt wrote:
> 
> Take a look at the stuff in win32\demos\service\ of your friendly latest
> Win32 extension install tree.
> 
> It shows you how to wrap Python code into an NT service quite easily.
> 
> The code in win32\lib\win32serivceutil.py appears to do almost everything
> you would ever want to do with a service. Very cool stuff...

This is all great stuff if you have understood it already.
For me, it is still unclear what makes up a service, and
what I need at the minimum to turn a Python program into
an NT service.

I need nothing more than to run Medusa as a service. Did
anybody do this before? Do I need special calls at all?

thanks - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From janssen at parc.xerox.com  Tue Jul  6 21:07:39 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Wed, 7 Jul 1999 01:07:39 GMT
Subject: make clean bug in 1.5.2
In-Reply-To: <99Jun24.102155pdt."3693"@watson.parc.xerox.com>
References: <99Jun24.102155pdt."3693"@watson.parc.xerox.com>
Message-ID: <0rUeVP8B0KGWFkQ3Ak@holmes.parc.xerox.com>

Excerpts from ext.python: 24-Jun-99 make clean bug in 1.5.2 Bill
Janssen at parc.xerox. (171*)

> Running "make clean" seems to leave a couple of problematic files,
> SRC/libpython1.5.a, and getbuildinfo.o, which then cause problems
> when I try to "make" again.

My mistake.  Should have used "make clobber", not "make clean".

Bill








From Phil.Mayes at p98.f112.n480.z2.fidonet.org  Fri Jul  2 13:00:55 1999
From: Phil.Mayes at p98.f112.n480.z2.fidonet.org (Phil Mayes)
Date: Fri, 02 Jul 1999 18:00:55 +0100
Subject: Mailbox, tell and read
Message-ID: <000016a5@bossar.com.pl>

From: "Phil Mayes" <nospam at bitbucket.com>

Paul Prescod wrote in message <377AAE02.E8591727 at prescod.net>...
>The mailbox message objects simulate a file interface by seeking through a
>mailbox. A read() returns the entire message. It does this by seeking to
>the start, reading to the message end, and returning the data. The problem
>is that message offsets are derived by file.tell() which I presume is in
>terms of bytes. But then it does a read() based on the offsets and read()
>is in terms of characters.
>
>These don't match up on Windows because of CR/LF pairs. I think that the
>right fix is to never use tell() and instead keep track of the location by
>counting characters. Does that make sense?


I am working with multiple mails in a single file, and found that using
seek/tell on the file in text mode didn't allow me to extract a single mail,
so I use binary, but I have to convert internally else quopri leaves =CRLF
inside the string.  Code fragment:
    # read the body from the mbx file with 'rb' to get the correct
    # amount of data in, then convert CRLF to LF else quopri fails.
    fmbx = open(mbxfile, 'rb')
    fmbx.seek(self._offbody)
    s = fmbx.read(self._offend - self._offbody)
    return string.replace(s, '\r\n', '\n') # runs in 2/3 the time of
    # ... joinfield/splitfield
--
Phil Mayes    pmayes AT olivebr DOT com






From mal at lemburg.com  Thu Jul 29 07:45:01 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 29 Jul 1999 13:45:01 +0200
Subject: any way to know when the program is exiting?
References: <199907290135.SAA22743@hpcll563.cup.hp.com>
Message-ID: <37A03EBC.21E689E6@lemburg.com>

Ovidiu Predescu wrote:
> 
> Hi,
> 
> I was wondering if there is any way in a __del__ method to know when it's
> invoked on the way out, as a result of the program going away?

It's a hack, but you can check sys.path: if it's None, then your program
is about to exit. To see why, check the output of 'python -vv -c "None"'.
 
Would probably be a good idea to add some flag to the sys module to be
able to check the current running state of the interpreter:

E.g. sys.state:
	'startup' - interpreter is intializing
	'ready' - interpreter is up and running
	'exiting' - interpreter is exiting

> I have a __del__ method of a class that sends a message to a global variable.
> It happens that on the way out, that value in that variable is destroyed before
> the __del__ methods of my instances are invoked. The result is that I have a
> bogus value that doesn't respond to the usual methods anymore. How can I
> prevent sending messages to that object?
> 
> A possible solution would be to use atexit() but this doesn't work if the user
> program that's using my library replaces it with its own function.

You can use sys.exitfunc() for Python programs.

Hope that helps.
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   155 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From bduncan1 at san.rr.com  Wed Jul 21 17:10:46 1999
From: bduncan1 at san.rr.com (Bruce Duncan)
Date: Wed, 21 Jul 1999 14:10:46 -0700
Subject: Pre-compiled Python binaries
Message-ID: <bIql3.4597$2D1.40180@newsr2.san.rr.com>

Greetings,

My Internet Web hosting provider is using an unusual hardware/software
configuration,
namely Sun Solaris on a Pentium box.  Any chance of getting this in the
pre-compiled
Python archives?

-bsd-

bsd at san.rr.com





From movits at lockstar.com  Tue Jul 13 12:22:15 1999
From: movits at lockstar.com (Mordy Ovits)
Date: 13 Jul 1999 09:22:15 PDT
Subject: XML-DOM slow?
Message-ID: <378B681C.E8CFCC7F@lockstar.com>

I've been playing around with the XML libraries, to great effect.  I
can't believe how useful it is.   However, it seems to be very slow. 
The (short) program at the bottom of this email takes about 10 seconds
to run!
What am I doing wrong?  Am I defaulting to a slow parser?
Thanks!
Mordy
-----------------------------------------
#!/usr/bin/python
                   
from xml.dom.html_builder import HtmlBuilder
import urllib
from sys import argv

s = urllib.urlopen(argv[1])

htmlstr = s.read()

b = HtmlBuilder()

b.ignore_mismatched_end_tags = 1
b.feed(htmlstr)  
doc = b.document

anchors = doc.getElementsByTagName('A')
for a in anchors:
    if a._node.attributes.has_key("HREF"):
        href = a._node.attributes["HREF"]
        print "Title: ", a._node.children[0].value
        print "Link: ", href.children[0].value
        print
    else:
        print "error in <A> (?)"
        print a._node.attributes
-------------------
-- 
o Mordy Ovits
o Cryptographic Engineer
o LockStar Inc.
---------------------------------------------------------------------------
#!/usr/local/bin/python 
from sys import*;from string import*;a=argv;[s,p,q]=filter(lambda
x:x[:1]!=
'-',a);d='-d'in
a;e,n=atol(p,16),atol(q,16);l=(len(q)+1)/2;o,inb=l-d,l-1+d
while s:s=stdin.read(inb);s and map(stdout.write,map(lambda
i,b=pow(reduce(
lambda x,y:(x<<8L)+y,map(ord,s)),e,n):chr(b>>8*i&255),range(o-1,-1,-1)))



From dek at nano.ucsf.edu  Wed Jul 21 16:50:03 1999
From: dek at nano.ucsf.edu (David Konerding)
Date: 21 Jul 1999 20:50:03 GMT
Subject: how do I find memory leaks?
References: <7n4skv$8do@thalamus.wustl.edu>
Message-ID: <slrn7pccjc.23g.dek@nano.nmr.ucsf.edu>

On 21 Jul 1999 11:31:27 -0500, Heather A. Drury <heather at thalamus.wustl.edu> wrote:
>Hi,
>
>I'm relatively new to python and am using it (and Tkinter, 
>VTK, NumPy, numpyio, and some other modules that I've probably
>forgotten) to write a volume/slice visualization package.
>
>Here's what I see on a SGI Indigo2 Max Impact with 448 Meg of memory:
>
>	before I start application	350 meg free
>	start application		342 
>	read volume 1			284
>	re-read volume 1		229	
>	re-read volume 1		192	
>	re-read volume 1		173	
>	re-read volume 1		155	


Are you storing data as textures, or display lists?
If VTK is an interface to OpenGL, then you need to clean up
some resources.  In my program, if I create a large number
of display lists with lots of geometry or texture data, they
take up a lot of RAM.  

I haven't used VTK with python, only PyOpenGL.

Dave
-- 
--------------------------------------------------------------------------------
Email: dek at cgl.ucsf.edu    David Konerding     WWW: http://picasso.ucsf.edu/~dek
--------------------------------------------------------------------------------
Snail: Graduate Group in Biophysics
Medical Sciences 926, Box 0446
University of California
San Francisco, CA 94143



From m.faassen at vet.uu.nl  Wed Jul 28 08:05:50 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 28 Jul 1999 12:05:50 GMT
Subject: IDLE mini-hack & ? (adding menu items, understanding *args, self)
References: <199907272152.QAA01078@cro.ots.ac.cr>
Message-ID: <7nmrmu$bcu$2@newshost.accu.uu.nl>

Hi again :)

Hm, I replied to Isidor explaining various ways to call arguments, but
I hadn't read his post well enough -- he figured that out for himself.

I'll just blame my error on the monstrous length of Isidor's post. :)

Regards,

Martijn
 



From thomas at xs4all.nl  Mon Jul 19 08:20:24 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Mon, 19 Jul 1999 14:20:24 +0200
Subject: [Q] Type checking...
In-Reply-To: <slrn7p637a.4v2.kas@titan.magnetic-ink.dk>; from Klaus Alexander Seistrup on Mon, Jul 19, 1999 at 01:32:58PM +0200
References: <7mv000$sun$1@feed.teaser.fr> <slrn7p637a.4v2.kas@titan.magnetic-ink.dk>
Message-ID: <19990719142024.C6170@xs4all.nl>

On Mon, Jul 19, 1999 at 01:32:58PM +0200, Klaus Alexander Seistrup wrote:
> Olivier Deckmyn <olivier.deckmyn at mail.dotcom.fr> wrote:

> > Abstract : [Q]How to check the type of an instance before(after?)
> > sending it to a method...

> One possible solution is:

> MAGIC="I'm in class A :-)"

> class ClassA:
>   def __init__(self):
>     self.magic = MAGIC

[etc]

Actually... there is a builtin function 'isinstance':

|>>> class ClassA:
|...     def spam(a):
|...             print a*20
|...
|>>> class ClassB:
|...     def mymethod(self, anInstance):
|...             if isinstance(anInstance, ClassA):
|...                     print "Yes :)"
|...             else:
|...                     print "No :("
|...
|
|>>> b=ClassB()
|>>> a=ClassA
|>>> c="eggs"
|
|>>> b.mymethod(a)
|Yes :)
|>>> b.mymethod(c)
|No :(

|>>> print __builtins__.isinstance.__doc__
|isinstance(object, class-or-type) -> Boolean
|
|Return whether an object is an instance of a class or of a subclass thereof.
|With a type as second argument, return whether that is the object's type.

See, no need to fuzzle with magic, Python does it for you :-) There's also
'issubclass' btw:

|>>> print __builtins__.issubclass.__doc__
|issubclass(C, B) -> Boolean
|
|Return whether class C is a subclass (i.e., a derived class) of class B.

__module__.function.__doc__.rocks()'ly yours,

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From gmcm at hypernet.com  Tue Jul 27 19:35:10 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 27 Jul 1999 18:35:10 -0500
Subject: 1.5.2 for: else:
In-Reply-To: <slrn7ps8ej.vm0.wtanksle@dolphin.openprojects.net>
Message-ID: <1279007788-11957995@hypernet.com>

William Tanksley wrote:
> On Tue, 27 Jul 1999 14:29:43 +0200, Thomas Wouters wrote:

> >4.4 break and continue Statements, and else Clauses on Loops 
> 
> >Loop statements may have an else clause; it is executed when the loop
> >terminates through exhaustion of the list (with for) or when the condition
> >becomes false (with while), but not when the loop is terminated by a break
> >statement. 
> 
> Well, I'm suprised.  I had also expected the else clause to run on
> empty loop.  It seems so natural, in line with the meaning of a loop
> (loop on this; otherwise do that).
> 
> Fortunately I've never actually used it and thus been disenchanted.

Maybe it violated your expectations, but it's much more valuable than 
what you expected. After all, testing for an empty sequence is a 
no-brainer. But finding a match in a list, and testing whether you 
fell off the end without finding one, is (without the else clause) a 
much messier proposal.

your-disenchantment-is-my-pleasure-<wink>-ly y'rs



- Gordon



From sweeting at neuronet.com.my  Mon Jul 26 16:16:53 1999
From: sweeting at neuronet.com.my (sweeting at neuronet.com.my)
Date: Mon, 26 Jul 1999 20:16:53 GMT
Subject: "libmysqlclient.so.6" not found. (what does /sbin/ldconfig /usr/local/lib/mysql do ?)
References: <Pine.LNX.4.04.9907231351440.30978-100000@emerald.netskate.ru>
Message-ID: <7nifnd$u1a$1@nnrp1.deja.com>

Thank you Klaus & Oleg & Thomas for the instructions on updating
/etc/ld.so.conf

>    You havn't mention your OS and version, looking for
>    ldconfig I think it is Linux.

Whilst it looked like a very Linux-life-thing, I'm actually
using FreeBSD.  We don't have /etc/ld.so.conf but I did find
'ldconfig' in the /etc/rc boot script.

Then again, Thomas' idea of the symlink was also very appealing :

> Alternatively (and imho preferable over adding to your ld.so.conf
> each time) place a symbolic link in /usr/local/lib or /usr/lib to
> point to /usr/local/lib/mysql/libmysqlclient.so.6. That way you
> dont suffer the clutter of dos-like PATH addenda :)

A simple :

ln -s /usr/local/lib/mysql/libmysqlclient.so.6
/usr/lib/libmysqlclient.so.6

did the trick :-)

Thank you very much, to one and all, for your help,

chas

>    To permanently add /usr/local/lib/mysql to ld search directory
>    path, add it to /etc/ld.so.conf file (it is just plain text file
>     - add yet another line to the end of the file) and run ldconfig
>     without parameters to rebuild ld.so.cache.



> On Fri, 23 Jul 1999 sweeting at neuronet.com.my wrote:
> > When building Python from source, and with the MySQLmodule in the
> > Modules directory, I got the following error :
> >
> > Shared object "libmysqlclient.so.6" not found
> > *** Error code 1
> >
> > But libmysqlclient.so.6 does indeed exist in /usr/local/lib/mysql.
> >
> > Fortunately, somebody had  come across this in a previous
post/thread
> > and the cure was to run " /sbin/ldconfig /usr/local/lib/mysql"
before
> > compiling. Sure enough this worked and now I have a fully
functioning
> > Python interpreter (and environment)  :)
> >
> > However, upon rebooting the machine, if I try to run a Python
program
> > it will again complain that "libmysqlclient.so.6" can not be found
> > (since the MySQLmodule is shared). Once again
> > running " /sbin/ldconfig /usr/local/lib/mysql"  cures the problem
but
> > does this mean that I have to run it each and every time I restart
the
> > computer, as part of my startup scripts ? Is this error indicative
of
> > there being a larger problem or misconfiguration ?
>
>    It is not misconfiguration - but ld should know where to find .so
files.
>
> > Thanking you in advance,
> >
> > chas
>
> Oleg.
> ----
>     Oleg Broytmann        Netskate/Inter.Net.Ru
phd at emerald.netskate.ru
>            Programmers don't die, they just GOSUB without RETURN.
>
>


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From janssen at parc.xerox.com  Tue Jul 27 18:59:40 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Tue, 27 Jul 1999 22:59:40 GMT
Subject: Python and Red Hat Linux 6
In-Reply-To: <XFMail.990727125533.nathan@islanddata.com>
References: <XFMail.990727125533.nathan@islanddata.com>
Message-ID: <IrbXbQsB0KGWFZV7RK@holmes.parc.xerox.com>

Excerpts from ext.python: 27-Jul-99 Re: Python and Red Hat Linux 6
Nathan Clegg at islanddata. (501*)

> You have to give #! an absolute pathname for it to work.  env is
> (somewhat) reliably at /usr/bin and is a program used exclusively for
> running other programs in certain environments.  In this case, it is used
> because it will search your path for python rather than requiring it to be
> explicit.

Of course, it only works if python is on the PATH of the program
executing the script.  Personally, I don't find the /usr/bin/env hack
useful.

Bill







From eugene.leitl at lrz.uni-muenchen.de  Tue Jul 20 01:45:13 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Mon, 19 Jul 1999 22:45:13 -0700 (PDT)
Subject: python to perl conversion?
In-Reply-To: <3793F935.C4CEFD44@callware.com>
References: <slrn7p7q8j.n4.dyoung@baltazar.baltazar.org>
	<3793F935.C4CEFD44@callware.com>
Message-ID: <14228.3147.448161.442831@lrz.de>

Ivan Van Laningham writes:

 > Oooh, yuck!  We've had requests for Perl-to-Python before, but never for
 > this.  (Gag me with a Perl-handled spoon.)

What's the consensus for the best (=less painful) way to call a perl
module from Python? There is this interesting thing called bioperl
http://bio.perl.org/ , which is both 1) very useful 2) in vigorous
development, so porting it is out of question.



From heiland at ncsa.uiuc.edu  Tue Jul  6 09:00:08 1999
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Tue, 06 Jul 1999 08:00:08 -0500
Subject: importdl problem on Linux
Message-ID: <3781FDD7.CDCD824A@ncsa.uiuc.edu>

I've built 1.5.1 on Linux (RH5.2), but get unresolved symbols on dlopen,
etc. in Python/importdl.c when I try to link an app together.  I
configured with "--without-gcc".  Am I missing something obvious?

thanks,
Randy




From jcw at equi4.com  Fri Jul  2 10:41:25 1999
From: jcw at equi4.com (Jean-Claude Wippler)
Date: Fri, 02 Jul 1999 16:41:25 +0200
Subject: Why does shelve make such large files?
References: <19990701211450.A6827@optiplex.palga.uucp> <377BF01D.16F8AA33@cup.hp.com> <19990702123152.B9975@optiplex.palga.uucp> <377CC17C.4487A36A@easics.be>
Message-ID: <377CCF95.DECB8CF5@equi4.com>

Johan Wouters wrote:
> 
> Gerrit Holl wrote:
> 
> > Ah, I understand.
> > So pickle is useful for very small datases, but when they're really
> > huge, one should use shelve. Isn't it?
> 
> Pickle will serialise your data into a certain format. This way you
> can "store" objects like lists as a whole. The serialising has the
> benefit that you can eg. put your object (be it a list or something
> else) through a pipe or socket and depickle it at the other side. This
> way you can have python processes that exchange data in a sort of
> native python format.
> 
> Shelve gives you alot more than just storing the data as a long
> sequence. Imagine the extra space being the administration that is
> needed for all the extra functionality (like tagging what data items
> are valid etc)
> 
> The overhead induced by a database management system can be enormous
> compared to the bare data you want to access, but this is mostly a
> space/time tradeoff: the more space you use the faster you can do
> things like searching and sorting etc etc.

Ahem, with all due respect to everyone... this is humbug.

Pickle serializes your data in one sweep to/from disk.  It's compact.

Shelve stores serialized pieces in a keyed-access form, and uses file
space allocation to support add/modify/delete.  It starts out with more
empty space, but as the amount of stored data grows, that space does not
expand linearly.  Consider shelves to have half their file space being
used, on *average* (plus or minus a factor 2 perhaps).

But the last claim is horrendously misleading, I'm afraid: data storage
is not a space/time tradeoff.  It's about throughput (I/O bottlenecks)
and overhead in managing the supported data and indexing schemes.  There
are order-of-magnitude performance differences in how several solutions
work, because of this.

Shudder.  The notion that a large database package, or a large datafile,
is faster, is so far from reality that it has to be corrected, even in
this Python-oriented newsgroup.  My apologies for the S/N ratio drop.

-- Jean-Claude

P.S.  Geloof niet alles wat je leest, Gerrit.



From rceci at adams.com  Sun Jul  4 13:44:47 1999
From: rceci at adams.com (rceci at adams.com)
Date: Sun, 04 Jul 1999 17:44:47 GMT
Subject: Moving the python distribution
Message-ID: <7lo6id$bl$1@nnrp1.deja.com>

Hello,

I have a question about relocating a python distribution (1.5.1) that
was built and installed in the normal area ( i.e. prefix=/usr/local ).

I have moved the bin/*, include/python1.5/* and lib/python1.5/* to a NFS
mount that will support multiple architectures, e.g.:

bin/python -> $mount_pnt/python/bin/$ARCH/python1.5/python
include/python1.5/* -> $mount_pnt/python/include/$ARCH/python1.5/*
lib/python1.5/* ->  $mount_pnt/python/lib/$ARCH/python1.5/*

I have set up PYTHONHOME and PYTHONPATH to support this.  When I run
python on the build machine, everything works as planned.  When I tag
the "-v" on to python, I can see the expected messages that show the
built-in modules being loaded from my NFS "$mount_pnt" rather that the
build directory "/usr/local".  Thats great.  But when I log onto another
system of identical architecture ( in this case both machines are IRIX
6.4 MIPS4 systems), I get:

Could not find platform independent libraries <prefix>
Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]

And yet the "-v" still prints it is loading from the $mount_pnt.  This
second system does not have python installed in "/usr/local..." at all.
I am guessing that specifiying PYTHONHOME/PYTHONPATH works, but the
iterpreter will still print warnings if it cannot find python runtimes
under the original exec prefix location.  Is there a way to suppress
these warnings or fix them in the source?

Thanks

Rob Cecil
rceci at adams.com


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From fredrik at pythonware.com  Wed Jul 14 21:20:52 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Jul 1999 01:20:52 GMT
Subject: Expand tuple into argument list?
References: <87aesy4w3f.fsf@solano.in.renlabs.com>
Message-ID: <003d01bece60$48955100$f29b12c2@secret.pythonware.com>

Steven Work <steve at renlabs.com> wrote:
> I'm new to Python.  I want a dispatch function -- takes a function
> name and other arguments, calls the named function with the remaining
> arguments.  This works for limited length argument lists:
> 
> def dispatch(fnname, *rest):
>     fn = dispatch.func_globals[fnname]
>     if len(rest)==0:
> return fn()
>     elif len(rest)==1:
> return fn(rest[0])
>     elif len(rest)==2:
> return fn(rest[0], rest[1])
>     elif len(rest)==3:
> return fn(rest[0], rest[1], rest[2])
>     elif len(rest)==4:
> return fn(rest[0], rest[1], rest[2], rest[3])
>
> but is there a general way to expand the tuple into an argument list?

http://www.python.org/doc/FAQ.html#4.31
"How do I call a function if I have the arguments in a tuple?"

...

in your case, something like:

    function = global()[fnname]
    result = apply(function, arguments)

should do the trick.  you can pass a dictionary with
keyword arguments as a third argument to apply. see
the built-in functions section in the library reference
for more info.

</F>




From skip at mojam.com  Mon Jul 26 11:30:07 1999
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 26 Jul 1999 10:30:07 -0500 (CDT)
Subject: Access dictionary in order?
In-Reply-To: <slrn7po8t3.6k.diego@blanka2.blankanet.it>
References: <roy-2507991300370001@mc-as01-p61.med.nyu.edu>
	<14235.18248.924848.52652@dolphin.mojam.com>
	<slrn7po8t3.6k.diego@blanka2.blankanet.it>
Message-ID: <14236.31923.854644.753949@dolphin.mojam.com>

    Diego> But how much time requires the keylist generation?

On my 233MHz P-II it takes about 0.07s to generate the list of keys for a
dictionary with 100,000 int keys.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From MHammond at skippinet.com.au  Thu Jul 22 19:13:06 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 23 Jul 1999 09:13:06 +1000
Subject: install shield & python
References: <3797323B.64357B18@mmrd.com>
Message-ID: <7n88dm$m3m$1@m2.c2.telstra-mm.net.au>

Lori Ludick wrote in message <3797323B.64357B18 at mmrd.com>...

>Has anyone had any experience with install shield calling python?
>
>I have written a simple script for a co-worker to call from within
>install shield.The script will return some interger value depending on
>what is located in a particular file.
>Basically the script looks like the following:
...

> sys.exit(retVal)


>Do I need to anything special for her to read this return value other
>that using sys.exit()?

That depends.  Your code will work assuming you will invoke this script
using Python.exe, and InstallShield can see external program exit codes.
For more information, you wil need to indicate exactly how you expect
InstallShield to execute this Python code.

Can you guarantee that Python.exe and everything else you needs is installed
on the machine?

Mark.






From kperacles at geocities.com  Thu Jul  8 21:26:56 1999
From: kperacles at geocities.com (Karl Putland)
Date: Thu, 8 Jul 1999 20:26:56 -0500
Subject: Python/PythonWin and DDE?
References: <3784b800.0@news.dbtech.net> <7m39vq$r5o$1@m2.c2.telstra-mm.net.au>
Message-ID: <3785515c.0@news.dbtech.net>

Thanks for the reply, Mark.

Mark Hammond <MHammond at skippinet.com.au> wrote in message news:7m39vq$r5o$1 at m2.c2.telstra-mm.net.au...
> > Karl Putland wrote in message <3784b800.0 at news.dbtech.net>...
> > Are there any docs about PythonWin and DDE?
> 
> No, although DejaNews may find some examples.

Tried DejaNews and drew blanks.  I couldn't find anything with dde or DDE.

> 
> > Specifically can Python be used as a DDE client?
> 
> Yes.  Pythonwin acts as both a client as a server as it communicates between
> processes.

Where to start?  Does PythonWin's DDE usage follow that found in M$ VC++?  If so those docs are weak.  I only have examples in VB and VBA.  Where would be the best place to start with PythonWin and DDE?

> Mark.
>


Karl Putland

From emile at fenx.com  Mon Jul 12 21:01:51 1999
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 12 Jul 1999 18:01:51 -0700
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
References: <14217.63931.75257.566266@weyr.cnri.reston.va.us><000101becc7c$fe2c9c20$8e9e2299@tim><3787D36A.815B7C29@ssec.wisc.edu><3787FB51.12FE0048@ssec.wisc.edu><7m923i$dem@dfw-ixnews5.ix.netcom.com><199907121445.HAA09860@netcom9.netcom.com> <14218.7197.219561.903989@weyr.cnri.reston.va.us>
Message-ID: <wfwi3.2961$VL2.372747@news.direcpc.com>

Well, some trains don't care if you don't brake...

--

Emile van Sebille
emile at fenx.com
-------------------





From loril at mmrd.com  Thu Jul 22 11:01:15 1999
From: loril at mmrd.com (Lori Ludick)
Date: Thu, 22 Jul 1999 11:01:15 -0400
Subject: install shield & python
Message-ID: <3797323B.64357B18@mmrd.com>

Has anyone had any experience with install shield calling python?

I have written a simple script for a co-worker to call from within
install shield.The script will return some interger value depending on
what is located in a particular file.
Basically the script looks like the following:




if __name__ == '__main__':

 retVal = versionCheck(dataPath)

 if (retVal == RUN_TEST1):
     retVal=checkVal(dataPath, retVal)

 sys.exit(retVal)



Do I need to anything special for her to read this return value other
that using sys.exit()?


Thanks,
Lori

--
--------------------------------------------------------------------
Lori Ludick                             loril at mdmgr.com
Programmer/Analyst
Medical Manager Research and Development
--------------------------------------------------------------------





From William.Tanksley at p98.f112.n480.z2.fidonet.org  Fri Jul  2 12:49:34 1999
From: William.Tanksley at p98.f112.n480.z2.fidonet.org (William Tanksley)
Date: Fri, 02 Jul 1999 17:49:34 +0100
Subject: Python GUIs: Summary and Conclusion
Message-ID: <0000167c@bossar.com.pl>

From: wtanksle at dolphin.openprojects.net (William Tanksley)

On Fri, 02 Jul 1999 03:28:45 GMT, Mikael Lyngvig wrote:
>On Thu, 01 Jul 1999 20:13:59 GMT, "Michael P. Reilly"
><arcege at shore.net> wrote:

>>Been there, done that.  My pyirc client (alpha release) was designed
>>for just about any user-interface, and comes with three: dummy (no
>>input or output, just batch-mode)), curses and Tkinter.  You just can't
>>switch between them at runtime, ;)  I've done it for other apps too.

>Well, the only reason I haven't been there is that I've only worked on
>one GUI app _ever_ - a 2D tile-based strategy game [1].  I did plan
>support for console apps (curses, etc., using a client-server
>architecture), but I didn't much fancy the idea of playing with a
>command line interface :)

I'm working on a 2D tile-based adventure game, which right now is text
only.  I'm expanding it so that it can use graphics, but I need some
libraries.  Let me know what you find!

>[1] Which I expect to change in the future, which is why I'm reviewing
>portable GUIs (I don't do non-portable stuff, if there's anything I
>can do to avoid it).

That's right.

>-- Mikael

-- 
-William "Billy" Tanksley



From aahz at netcom.com  Thu Jul  8 19:50:17 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 8 Jul 1999 23:50:17 GMT
Subject: Python and Paradox
References: <7m2mrj$hes$1@nnrp1.deja.com> <7m2o2b$aq0@dfw-ixnews12.ix.netcom.com> <37852804.54D5D2AD@quasar.ipa.nw.ru>
Message-ID: <7m3dfp$b1k@dfw-ixnews4.ix.netcom.com>

In article <37852804.54D5D2AD at quasar.ipa.nw.ru>,
Alexander V. Voinov <avv at quasar.ipa.nw.ru> wrote:
>Aahz Maruch wrote:
>> In article <7m2mrj$hes$1 at nnrp1.deja.com>,
>> Kaine  <kaine_the_bloody at my-deja.com> wrote:
>>>
>>>I am looking for a way to access Paradox tables from Python without
>>>using ODBC. Does anyone have any ideas?
>>
>> Nope, from an ex-Paradox tech support engineer.
>
>Why nope? What is the platform? If it's win16, the direct wrapping is
>possible.

Oh, okay, if you purchase the API.  I was assuming that Kaine didn't
want to buy anything.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From gerrit.holl at pobox.com  Fri Jul 23 06:58:28 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 23 Jul 1999 12:58:28 +0200
Subject: cin
In-Reply-To: <37976CF0.3DC02CBA@weihenstephan.org>; from Dr. Peter Stoehr on Thu, Jul 22, 1999 at 09:11:44PM +0200
References: <3srl3.1302$Gj8.191985152@newsa.telia.net> <cCtl3.818$ui4.256788@news.shore.net> <37976CF0.3DC02CBA@weihenstephan.org>
Message-ID: <19990723125828.A29666@optiplex.palga.uucp>

On Thu, Jul 22, 1999 at 09:11:44PM +0200, Dr. Peter Stoehr wrote:
> 
> # -----------------------------------------
> aString = raw_input("Give me input :")
> print aString
> # -----------------------------------------
> 

print raw_input("Give me input :") # This seems much easier!

regards,
Gerrit.



From sdm7g at virginia.edu  Fri Jul 23 18:38:52 1999
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Fri, 23 Jul 1999 22:38:52 GMT
Subject: Thinking in Python ? 
Message-ID: <Pine.A32.3.90.990723183138.15084B-100000@elvis.med.Virginia.EDU>


I was just looking on BruceEckel.com for the latest online versions
of "Thinking in C++" &  "Thinking in Java" when I saw: 

<http://www.BruceEckel.com/PythonProjectWorkshop/index.html>

The Python Project Workshop -- 

A 5-day, Hands-on Intensive Program Creation Experience using Python
Coached by Bruce Eckel

(using "Learning Python" and "Programming Python" BTW.) 

This post is not intended as an advertisement -- but I liked his
Java book enough to go out and buy a paper copy, and  maybe 
"Thinking in Python" is next! 


---|  Steven D. Majewski   (804-982-0831)  <sdm7g at Virginia.EDU>  |---
---|  Department of Molecular Physiology and Biological Physics  |---
---|  University of Virginia             Health Sciences Center  |---
---|  P.O. Box 10011            Charlottesville, VA  22906-0011  |---




From Jeffrey.Kunce at p98.f112.n480.z2.fidonet.org  Fri Jul  2 18:08:19 1999
From: Jeffrey.Kunce at p98.f112.n480.z2.fidonet.org (Jeffrey Kunce)
Date: Fri, 02 Jul 1999 23:08:19 +0100
Subject: XML-RPC and firewall [solution]
Message-ID: <000016a6@bossar.com.pl>

From: "Jeffrey Kunce" <kuncej at mail.conservation.state.mo.us>

--=_693F2E73.E584ED6B
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

I have been playing with XML-RPC on my intranet, but
found that Fredrik's xmlrpclib.py wouldn't talk to external=20
servers through our firewall.

Since urllib has proxy-server support built-in, I wrote a=20
urllib-based transport plugin for xmlrpclib.py (attached).
Now I can speak XML-RPC with the rest of the world :-)

Testing on this has been minimal - use at your own risk.
If it checks out, maybe /F will include it in the distribution.

  --Jeff




--=_693F2E73.E584ED6B
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable
Content-Disposition: attachment; filename="xmlrpc_urllib_transport.py"

'''\
urllib-based transport plugin for xmlrpclib.py (with test code)
xmlrpc_urllib_transport.py
jjk  07/02/99  001  from xmlrpc_test_urllib.py 002

urllib transport was needed to get past proxy server

for xmlrpclib __version__ =3D "0.9.8"

****************************
*** USE AT YOUR OWN RISK ***
****************************
'''

import xmlrpclib

class UrllibTransport(xmlrpclib.Transport):
	'''Handles an HTTP transaction to an XML-RPC server via urllib
	(urllib includes proxy-server support)
	jjk  07/02/99'''

	def request(self, host, handler, request_body):
		'''issue XML-RPC request
		jjk  07/02/99'''
		import urllib
		urlopener =3D urllib.FancyURLopener()
		urlopener.addheaders =3D [('User-agent', self.user_agent)]
		# probably should use appropriate 'join' methods instead =
of 'http://'+host+handler
		f =3D urlopener.open('http://'+host+handler, request_body)
		return(self.parse_response(f))

def test():
	'''test UrllibTransport for xmlrpclib
	Use proxy for external address, no proxy for local address
	jjk  07/02/99'''
	import os
	print 'testing xmlrpc_urllib_transport.py'
	raw_input('start xmlrpcserver.py (test mode) on local machine, =
then press enter')
	testhosts =3D [
		(0, "http://localhost:8000"),	   # don't use proxy, =
local host
		#(0, "http://mdcmap.mocons.gov:8000"),   # don't use =
proxy, another host on local network
		(1, "http://betty.userland.com"),	#use proxy, =
userland's XML-RPC server
		]
	try: proxies =3D os.environ['http_proxy']
	except KeyError: proxies =3D None
	for useproxy, host in testhosts:
		print=20
		print 'testing host:', host
		if (useproxy):=20
			if (proxies):
				os.environ['HTTP_PROXY'] =3D proxies
		else:
			try: del os.environ['HTTP_PROXY']
			except KeyError: pass
		server =3D xmlrpclib.Server(host, transport=3DUrllibTranspo=
rt())
		print server
		try: print server.examples.getStateName(41)
		except xmlrpclib.Error, v: print "ERROR", v
	=09
if (__name__=3D=3D'__main__'):
	import pdb
	test()

--=_693F2E73.E584ED6B--



From gerrit.holl at pobox.com  Thu Jul  8 16:38:57 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 8 Jul 1999 20:38:57 GMT
Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
In-Reply-To: <tot2m7.su.ln@final.math.uni-bonn.de>; from Thomas Ackermann on Thu, Jul 08, 1999 at 09:22:05PM +0200
References: <tot2m7.su.ln@final.math.uni-bonn.de>
Message-ID: <7m9tlr$6qg$8@rks1.urz.tu-dresden.de>

On Thu, Jul 08, 1999 at 09:22:05PM +0200, Thomas Ackermann wrote:
> From: Thomas Ackermann <tgm at math.uni-bonn.de>
> Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
> Date: 	Thu, 8 Jul 1999 21:22:05 +0200
> To: python-list at python.org
> 
> Hi!
> 
> I have not much time to experiment with those modules and would like
> to hear opinions about to use what module     ;-)
> 

Well,
I used HTMLgen and I don't really understand why to use it. If you don't
know HTML, you can't use it: you need to understand tags like <B>, <I>, <U>
and such. I strongly recommend to have a look at HTML: if you look at it only
one hour, you'll already understand the basics. Another hour and you understand
frames and tables, and another hour to understand how it interacts with CGI.
So you'll only need 3 hours, if not less, to learn web programming. When you're
able to do that, you'll understand much more about HTMLgen (but won't need it,
I think...)

> Also, i am not exactly sure what Zope is - it seems to be something to
> create WWW-Pages and allow access to databases and it has it's own
> WebServer (a variant of Medusa) ...
> 

Me neither.

> So, what to use to allow for an Internet-Shop with pages created from
> databases and allowing customers to purchase over net?
> 

If you're doing such serious things, you should 

                                     **really**

learn web programming!!

> Can do Zope such things? Is it overkill at first? What can it do
> beyond? (And yes, i read some documentation, but i do not even
> understand what an "Application Server" is    ;-)
> 

Zope seems to be quite powerful.

>                GNU LINUX Python gtk pygtk MySQL FUDGE GURPS

GNU LINUX Python Civ III discoverb X much more.

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From movits at lockstar.com  Thu Jul  1 11:13:11 1999
From: movits at lockstar.com (Mordy Ovits)
Date: 01 Jul 1999 08:13:11 PDT
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain>
Message-ID: <377BB027.79A1331A@lockstar.com>

Alan Daniels wrote:
> 
> I'm working on a Quake-style game engine, where my goal is to give it
> a fully integrated editor, similar to the excellent ones in Starcraft
> and Unreal. I'm planning on doing this by giving the game and the
> editor as much back-end code in common as possible.

The game Extreme Warfare was a realtime 3D videogame that used Python as its
scripting language for AI and game logic.  Take a good look at it.

Does this project of yours have a homepage I could look at?

Mordy
-- 
o Mordy Ovits
o Cryptographic Engineer
o LockStar Inc.
---------------------------------------------------------------------------
#!/usr/local/bin/python 
from sys import*;from string import*;a=argv;[s,p,q]=filter(lambda x:x[:1]!=
'-',a);d='-d'in a;e,n=atol(p,16),atol(q,16);l=(len(q)+1)/2;o,inb=l-d,l-1+d
while s:s=stdin.read(inb);s and map(stdout.write,map(lambda i,b=pow(reduce(
lambda x,y:(x<<8L)+y,map(ord,s)),e,n):chr(b>>8*i&255),range(o-1,-1,-1)))



From clarkeia at logica.com  Fri Jul 30 05:02:03 1999
From: clarkeia at logica.com (Ian Clarke)
Date: Fri, 30 Jul 1999 10:02:03 +0100
Subject: Is bsddb.pyd thread safe in 1.5.2 (win95)
References: <7npimc$l9b@romeo.logica.co.uk> <Z37o3.41724$jl.29736754@newscontent-01.sprint.ca>
Message-ID: <7nrpm0$hcq@romeo.logica.co.uk>

Bruce Dodson wrote in message ...
>I see a bsddb.pyd included with Python 1.5.1, and I assume with 1.5.2.  I
>think this means Python shouldn't fall back to dumbdbm on a normal
>installation.  If I am wrong about that, e.g. if bsddbm.pyd is based on the
>1.x series dbm that didn't support concurrent use, I think you can get a
>safe version of dbm from www.SleepyCat.com, or maybe the gdbm from
>www.roth.net/libs.  It should be quite simple to plug in one of those to
the
>bsddbm or gdbm Python module source.


Is there any way to find out which version of dbm the 1.5.2 bsddb.pyd file
is based on?  I would rather stick to using standard Python libraries if
possible for deployment purposes, but finding out whether a program is
thread-safe is a bit of a black art unless you are told in documentation
(experimentation, for example, is not really a satisfactory method!).

Also, what is a .pyd file - and should I be able to find bsddbm.pyd in my
installation (because I can't).

Cheers,

Ian.





From kernr at mail.ncifcrf.gov  Tue Jul 27 17:54:00 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Tue, 27 Jul 1999 21:54:00 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <p51zdvn4c7.fsf@bidra241.bbn.hp.com> <14236.32616.980578.536614@dolphin.mojam.com> <379D89BC.DA91BA8C@hp.com> <14237.60407.130426.461846@dolphin.mojam.com> <14238.335.458728.140169@lrz.de> <14238.1372.334823.378192@dolphin.mojam.com>
Message-ID: <379e2890.6069849@news.erols.com>

On Tue, 27 Jul 1999 14:19:06 -0500 (CDT), Skip Montanaro
<skip at mojam.com> wrote:

>
>    Eugene> Skip Montanaro writes:
>    >> True.  I was thinking about the situation I'm in (small developer, a few
>    >> machines to maintain/upgrade/replace).
>
>    Eugene> What's wrong with recoding the hot spots in C? Python is
>    primarily a rapid prototyping/scripting/glue language, it was never
>    meant as a sole implementation language for applications.
>
>There's nothing wrong with recoding hot spots in C.  Many people do it all
>the time.  Practically speaking, however, many Python users don't have the
>technical depth or other resource (VC++, for instance) 

Just to address this for-instance, you can compile Python extensions
for Win32 systems using the free egcs compiler.
See <URL: http://starship.python.net/crew/kernr/mingw32/Notes.html>
for more information.

List of modules that I've compiled using this setup:
  * MMTK 2.0a9
  * Konrad Hinsen's NetCDF interface
  * Zope 2.0a3 (which includes ExtensionClass et al.)
  * mxCrypto
  * mxTextTools
  * mxTools
  * SAML interface
  * Numerical Python
  * Multipack
  * FFTW-numpy
  * Readline module
  * sparsemodule
  * gmpmodule
  * PyGLTT
  * and probably some more that I've forgotten about

>to write chunks of
>their applications in C or C++.
>
>Skip Montanaro	| http://www.mojam.com/
>skip at mojam.com  | http://www.musi-cal.com/~skip/
>847-475-3758

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From Andrew.Csillag at p98.f112.n480.z2.fidonet.org  Thu Jul  1 14:31:27 1999
From: Andrew.Csillag at p98.f112.n480.z2.fidonet.org (Andrew Csillag)
Date: Thu, 01 Jul 1999 19:31:27 +0100
Subject: Name of function within the function?
Message-ID: <0000162c@bossar.com.pl>

From: Andrew Csillag <andrew at starmedia.net>

"Dula, Debbie (EXCHANGE:RICH2:2C25)" wrote:
> 
> I've looked and looked, but can't find the answer to this ... I'm
> sure it's right in front of me ...
> 
> I want the current function's name to be output in my error
> messages, but can't figure out how a function can get its own name
> (without already knowing it).  For example,
> 
>         def f1():
>                 print "Error in function", f1.__name__
> 
> is not really useful for what I need as I might as well print "Error
> in function f1".  Is there any generic way I can get the current
> function's name?
> 
> Thanks,
> Debbie

How's about this:
import sys
def y():
    print callerFunc()
def y2():
    print callerFunc()

def callerFunc():
    try:
        raise "arg"
    except "arg":
        func=sys.exc_traceback.tb_frame.f_back.f_code.co_name
    return func

y()
y2()

Should print:
y
y2

Cheers,
Drew Csillag
-- 
  Writing perl programs is like pounding stakes into your feet:
  You could if you had to, but would be better off if you didn't.
Corallary:
  Using Microsoft products is like pounding stakes into your feet:
  You could if you had to, but would be better off if you didn't.



From dalke at bioreason.com  Tue Jul 20 00:14:46 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Mon, 19 Jul 1999 22:14:46 -0600
Subject: bug in regrtest
Message-ID: <3793F7B6.482F02A3@bioreason.com>

Hello,

  We use a modified version of regrtest (Lib/test/regrtest.py)
for internal regression tests.  I just tracked down a bug in
our version which seems to be present in the original one.


  The regression run is compared to golden output from a
known, correct file.  This is done in "runtest" using an instance
of the Compare class, which compares each output data with
the corresponding data from the standard.  If they are different,
an exception is raised.

  The problem comes when the regression run is finished while
golden data is still available.  There is a `close' method in the
Compare class which can check for this sort of error and raise
an exception, but it is never called.

  The relevant code from regrtest.runtest looks like (with some
editing for clarity):

     if generate:
         cfp = open(outputfile, "w")
     elif verbose:
         cfp = sys.stdout
     else:
         cfp = Compare(outputfile)
     try:
        save_stdout = sys.stdout
        try:
            if cfp:
                sys.stdout = cfp
                print test              # Output file starts with test name
            __import__(test, globals(), locals(), [])
        finally:
            sys.stdout = save_stdout
     except ....
        return 0
     else:
        return 1


It looks like the easiest, correct fix is to add the following:

            if cfp and not generate and not cfp:
               cfp.close()

just after the __import__ statement.

I tested against the Python 1.5.1 distribution, which has a
slightly older version of regrtest than 1.5.2, although the latest
CVS version still has the flaw.  There are two tests which now fail
the regression tests, test_re and test_select.

I looked more closely at test_select, since it is smaller than
test_re.  The output/test_select file contains output which
agrees with the output of
  python regrtest.py -v test_select.py

*However*, if you look at test_select.py, the proper output will
only be generated when the verbose flag is set, because of constructs
like:

>                 if verbose:
>                         print 'timeout =', tout

In normal tests, test_select.py should generate no output.  This
can be verified by using -g to generate a new output/test_select
file, which will contain only the test name.

So this output difference was masked by the Compare problem
mentioned above.

Looking at the CVS logs, it has been that way for a long time,
1996/12/11.

						Andrew Dalke
						dalke at bioreason.com



From Vladimir.Marangozov at inrialpes.fr  Wed Jul 21 12:11:25 1999
From: Vladimir.Marangozov at inrialpes.fr (Vladimir Marangozov)
Date: Wed, 21 Jul 1999 17:11:25 +0100
Subject: IDLE's PyShell
References: <000801bed335$8c233e80$a72d2399@tim>
Message-ID: <3795F12D.91B820BB@inrialpes.fr>

Tim Peters wrote:
> 
> > I'm very far from this code, so don't blame me for not proposing a
> > patch, but here are two things I experimented with that seem to help
> > partially:
> >
> > 1. Clear the linecache when closing the PyShell window
> 
> Submit a patch!

If I knew where's the appropriate place to put the linecache.clearcache()
function call, I would have submitted a patch. But I don't know very well
the shell logic, so I'll let somebody else do it for me.

Actually, I really don't understand very well how the PyShell works.
Here's another little exercise that causes some weird behavior. If you can
reproduce it, then it's most probably a bug unrelated to my personal install.

1) start a brand new idle.py (the PyShell welcomes you to Pythonland)
2) press RETURN a couple of times (just to warm the linecache ;-)
3) See the contents of the cache dict by typing "PyShell.linecache.cache"
4) Now provoke a name error by typing, say, "bonk"

we get the usual Traceback ... NameError: bonk.
Everything seems to be perfect, until...

5) See again the contents of the cache with PyShell.linecache.cache

I get a huge output, containing the entire PyShell.py source...
With all due respect to PyShell.py, its source shouldn't be there, isn't it?

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov at inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252



From mwh21 at cam.ac.uk  Sun Jul  4 18:24:42 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 04 Jul 1999 23:24:42 +0100
Subject: Why have 300 articles just appeared in c.l.py?
Message-ID: <m33dz4oxxx.fsf@atrus.jesus.cam.ac.uk>

I just had the slightly alarming experience of about 300 articles, all
of which I had read, appearing unthreaded in this newsgroup.

Did this happen to everyone, or is it some local snafu?

TIA
Michael



From mal at lemburg.com  Tue Jul  6 17:08:12 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 6 Jul 1999 21:08:12 GMT
Subject: mxODBC and MySQL access to remote DB
References: <Pine.LNX.4.05.9907062021240.29446-100000@ns.ac-copy.net>
Message-ID: <3782703C.67DDB283@lemburg.com>

Joachim Schmitz wrote:
> 
> On Tue, 6 Jul 1999, M.-A. Lemburg wrote:
> 
> > >     ,DBtype="MySQL")
> > >   File "Form.py", line 45, in __init__
> > >     ,"pw="+self.pw)^M
> > > mxODBC.OperationalError: ('S1000', 1045, "[TCX][MyODBC]Access denied for
> > > user: 'user=perry at localhost' (Using password: YES)", 4237)
> >
> > This a problem with MySQL, not mxODBC. Please check the MySQL
> > logs for more detailed information about access failures.
> >
> 
> The problem was that mxODBC or better the ODBC-driver myodbc_mysql didn?t
> connect to the host "test.ac-copy.net" but to localhost. I finally found
> the solution in the connect.c module of the driver: (version
> iodbc-2.50.22a/myodbc-2.50.22 )

Ah, ok, that sounds reasonable.
 
> in  SQLConnect around line 183.
> 
> #ifndef _UNIX_
>   SQLGetPrivateProfileString(fix_str(dsn,szDSN,cbDSN),"server",
>                              "localhost", host, sizeof(host),"ODBC.INI");
>   SQLGetPrivateProfileString(dsn,"database",
>                              dsn, database, sizeof(database),"ODBC.INI");
>   SQLGetPrivateProfileString(dsn,"port",
>                              "0", port, sizeof(port),"ODBC.INI");
>   port_nr=(uint) atoi(port);
>   SQLGetPrivateProfileString(dsn,"flag",
>                              "0", flag, sizeof(flag),"ODBC.INI");
>   flag_nr=(ulong) atol(flag);
>   SQLGetPrivateProfileString(dsn,"stmt",
>                              "0", init_stmt,sizeof(init_stmt),"ODBC.INI");
> #else
>   strmov(database,szDSN);
>   {
>     char *cp;
>     strmov(host, (cp = getenv("MYSQL_HOST")) ? cp : "localhost");
>   }
>   flag_nr=0;
>   init_stmt[0]=0;
> #endif /* IS UNIX */
> 
> After setting the enviromentvariable MYSQL_HOST to test.ac-copy.net it
> worked.
> I?m not sure if this is the intended behavior, cause it?s nowhere
> documented.

Anyway, it shows Open Source at it's best ;-)

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   178 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From Larry.W..Virden at p98.f112.n480.z2.fidonet.org  Fri Jul  2 00:28:57 1999
From: Larry.W..Virden at p98.f112.n480.z2.fidonet.org (Larry W. Virden)
Date: Fri, 02 Jul 1999 05:28:57 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001669@bossar.com.pl>

From: lvirden at cas.org (Larry W. Virden)

On Jul 2,  8:16, Gerrit Holl wrote:
} Subject: Re:  Python GUIs: Abandoning TkInter and welcoming wxPython?
--> On Fri, Jul 02, 1999 at 11:56:10AM +0000, lvirden at cas.org wrote:
--> > From:	lvirden at cas.org
--> > According to Mikael Lyngvig <mikael at pobox.com>:
--> > :Well, TkInter is extremely slow, but not because of the Python part.
--> > When I run the Tk test suite  on my SPARC, Tk cranks thru creating and
--> > deleting windows as fast as the eye can track.
--> 
--> I think you should also say what kind of processor you have if you're
--> talking about this:
--> 
--> "Windows NT on a Pentium IV 500 MHZ runs over 10 times faster than a
--> Linux 386 30 MHZ" (is it?)

Probably worthwhile to indicate what OS, what version of Tk (and Python
if that's factoring into the discussion), and probably amount of memory.
Another factor may be the kind of video card - 8 bit vs 32 bit vs something
else...

I know I was referring to an Ultra 5 / Solaris 2.6 / Tk any version /
128 meg of memory, 8 bit color 19" monitor.


-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.




From chapman at bioreason.com  Fri Jul 16 12:46:05 1999
From: chapman at bioreason.com (Mitch Chapman)
Date: Fri, 16 Jul 1999 10:46:05 -0600
Subject: Getters and Setters
References: <000701bece99$6dbeede0$51a22299@tim> <378EB666.BBEC521D@compaq.com>
Message-ID: <378F61CD.145EC60B@bioreason.com>

Greg Ewing wrote:
> There's something to be said for designing a
> language so that the same syntax is used for
> accessing an attribute whether it's directly
> exposed or got at through a method.

You've been reading Bertrand Meyer again,
haven't you? ;)

(Object-Oriented Software Construction makes the same argument.)

-- 
Mitch Chapman
chapman at bioreason.com



From tim_one at email.msn.com  Sat Jul 17 00:34:38 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Jul 1999 00:34:38 -0400
Subject: Meta-language for mxTextTools
In-Reply-To: <010001becf64$a615fa50$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <000301bed00d$ae0a1080$d2a22299@tim>

[M.-A. Lemburg]
> Hmm, I guess Tim will like this ;-)

[Tony J. Ibbs (Tibs)]
> That would be nice - I bet he won't comment until some code is produced,
> though!

No, no, comments *have* to be made before code is produced.  Afterwards,
they're called complaints <0.9 wink>.

I think you're doing fine here, and symbolic label names are 80% of the tag
table battle, "block if" another 15%.  The biggest caution is that no little
language stays little forever -- unless it's a failure.  So things you'd
never do in a "real language" are suspect in any language; e.g., if I have
to write "lower+upper" instead of "lower + upper" because whitespace is your
only means of separating fields, that doesn't bode well for future feature
bloat.

face-it-your-life-is-over<wink>-ly y'rs  - tim





From tim_one at email.msn.com  Mon Jul 12 11:41:21 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 12 Jul 1999 11:41:21 -0400
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
In-Reply-To: <14217.63931.75257.566266@weyr.cnri.reston.va.us>
Message-ID: <000101becc7c$fe2c9c20$8e9e2299@tim>

[Fred L. Drake]
>   Are you saying that (either of) these [chdir, getcwd] are in the
> wrong place?  The notion of "current directory" is tied only to the
> entry in the process table; should it be documented otherwise?
>   The "Process Parameters" section contains functions with provide any
> direct access to the process table entry, both queries and changes, so
> I think it really is the right place for these two functions.
>   I can easily add a textual reference to the right place within the
> "Files and Directories" section,

That would do!

> but you'll need to explain very clearly why the descriptions
> should be moved.

They don't have to be moved, provided they're also mentioned where people
first look <wink>.  The notion that the current directory is "a process
table thing" is a bit Unix-centric, and e.g. a Windows person is more likely
to think of it as "a directory" first.

currently y'rs  - tim





From gmcm at hypernet.com  Mon Jul 19 09:23:37 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 19 Jul 1999 08:23:37 -0500
Subject: designing question
In-Reply-To: <3792D572.7F223EC3@worldnet.fr>
Message-ID: <1279735663-4809742@hypernet.com>

Pierrette writes:

> I have two questions:
> 1. The database will hold abouth 50.000 lines in its biggest table,
> can gadfly handle this without problems, what abouth acces time ??

Gadfly is in-memory, and (to support transactions) has 2 copies of 
the db. If the machines have the memory, access time will be fine.

> 2. The application will have abouth 50 different inputscreens wich
> acces all to different tables. I will create a class for every
> window wich inheret from a general 'input-screen'-class. I wonder if
> a should place the db-access code (insert-update-delete) in every
> specific window or if i should  create a big db class wich has all
> possible db-acces in it ? 

Completely philosophical question. For flexibility, I favor the GUI 
making no assumptions about how the data is stored. I would have the 
GUI making calls like "app.addthingie(thingie)", where the app has an 
"abstract" UI.

Of course, it almost nevers works out that cleanly in practice, but 
it's certainly a worthy goal.

Now if each screen maps exactly to a table, this is probably 
overkill. But if your data is normalized, this makes a lousy UI - 
users usually want to know all kinds of silly things, like whether 
there's a "next" record, or the owning customer's name (not the 
foreign key in this row, which is just a number) that violate 
normalization.

- Gordon



From billtut at microsoft.com  Fri Jul 30 19:17:44 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Fri, 30 Jul 1999 23:17:44 GMT
Subject: Need help with TCP/IP client access from Windows
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90FF6@RED-MSG-50>


> From: Bill Tutt 
> 
> > From: Eric S. Raymond [mailto:esr at thyrsus.com]
> > 
> > 
> > Gordon McMillan <gmcm at hypernet.com>:
> > > > OK.  Will giving the filename ".\\intriguer" work for 
> this?  And,
> > > > given that the filename is going through the C library, 
> > will leaving
> > > > it as "./intriguer" work?
> > > 
> 
> Well, as long as your current working directory is what you 
> want it to be, sure it'll work, otherwise you'll need to pull 
> the standard "where did I get run from" nonsense.
> 

Something like the following usually does the trick:
import os
import sys

(head, tail) = os.path.split(sys.argv[0])
# If tail is None, then we just got invoked in the current working
directory.
if tail is None:
	path = os.getcwd()
else:
# Otherwise, head now contains a possibly relative path to where we were 
# invoked from, normalize it.
	path = os.sys.abspath(head)
inifile = open('%s/blah.ini' % path, 'r')

Bill



From emile at fenx.com  Fri Jul 30 08:40:19 1999
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 30 Jul 1999 05:40:19 -0700
Subject: os.name
References: <1278785134-25350540@hypernet.com>
Message-ID: <V5ho3.4340$VL2.463886@news.direcpc.com>

on both my 95 and 98 boxes:

>>> os.name
'nt'
>>> os.environ('OS')
Traceback (innermost last):
  File "<pyshell#8>", line 1, in ?
    os.environ('OS')
AttributeError: no __call__ method defined

--

Emile van Sebille
emile at fenx.com
-------------------


Gordon McMillan <gmcm at hypernet.com> wrote in message
news:1278785134-25350540 at hypernet.com...
> Donn Cave writes:
>
> > Holger Jannsen <holger at phoenix-edv.netzservice.de> writes:
>
> > | It's because I need to differ between Win95, Win98 and WinNT....
> >
> > Not clear, missing parts of speech, however, os.uname()...  Windows
> > behavior unknown.
>
> Windows gives  an AttributeError.
>
> os.environ['OS'] gives "Windows_NT" on my NT box. Maybe some Win9xers
> could try it on theirs...
>
> - Gordon
>
>





From jeff at popmail.med.nyu.edu  Wed Jul 21 10:28:32 1999
From: jeff at popmail.med.nyu.edu (Jeff Berliner)
Date: Wed, 21 Jul 1999 09:28:32 -0500
Subject: where is gfplus?
References: <19990721100617.07994@nms.otc.telstra.com.au>
Message-ID: <jeff-2107990928320001@antwerp.med.nyu.edu>

In article <19990721100617.07994 at nms.otc.telstra.com.au>, Greg McFarlane
<gregm at iname.com> wrote:
> I have been trying to get through to the gfplus site
> <http://shamrock.med.nyu.edu/~jeff/gfplus/> but it seems to be down. 
> Does anyone know of anywhere else I can get gfplus (an interactive
> gadfly shell)?  Or can anyone send it too me (it can't be that big:).

   Sorry, the machine which was hosting gfplus is currently restricted
from the net.  I've moved the gfplus pages to:

   http://endeavor.med.nyu.edu/~jeff/gfplus/

   So, you can download gfplus 1.1 from there.  Also, 1.2 will hopefully
be out in the next few days.  

> Also, is there a gui interface to gadfly?  Something that lists all
> the tables, all the columns in a table, all the data in a table,
> indexes, etc.
   
   As well as Richard Jone's GTk based one, I believe Marc Risney was
developing a GUI Gadfly shell as well, BarFly, which I don't know the
status of.  Maybe if he's listening he can tell us :)

   Apologies again!

                                                - Jeff

-- 
Jeff Berliner                                   jeff at popmail.med.nyu.edu
Academic Computing,                             Phone: (212) 263-2501
New York University School of Medicine          Fax:   (212) 263-8542

-- 
Jeff Berliner                            jeff at popmail.med.nyu.edu 
Academic Computing,                      Phone: (212) 263-5744
New York University School of Medicine   Fax:   (212) 263-8542



From ivanlan at callware.com  Wed Jul 21 15:49:02 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 21 Jul 1999 19:49:02 GMT
Subject: pow()
References: <1279537084-16754559@hypernet.com>
Message-ID: <3796242E.AEBC536@callware.com>

Hi All--

Gordon "He's been eating those funny mushrooms again" McMillan wrote:
> 
> Ivan Van Laningham wrote:
> 
> > I couldn't find an explanation for the three-argument version of
> > pow() on DejaNews, the Python 1.5.2 docs or the Python FAQ, so I'll
> > ask now.
> >
> > Can someone explain (in simple terms, for a bear of small brain) to
> > me what the modulo argument to __pow__ is for?
> >
> >  __pow__(self,other[,modulo])
> 
> It's for the number of points on the stars that revolve around your
> head. bam() needs no such arg, because birdies all have exactly 2
> wings. socko() is no longer supported; which is probably why
> Universal Studios dropped out of the Python Consortium.
> 
> leaving-a-bunch-of-looney-tunes-behind-ly y'rs
> 

You forgot taz() and let's-kick-it-up-a-notch().  But I suppose you're
going to tell me the Food Channel dropped out of the Consortium too. ...

<you-thought-maybe-he-got-those-recipes-from-Guido?>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From aaronl at cannontech.com  Mon Jul 19 15:08:31 1999
From: aaronl at cannontech.com (Aaron Lauinger)
Date: Mon, 19 Jul 1999 14:08:31 -0500
Subject: Embedding python/multi-thread env
Message-ID: <379377AF.E8E7351D@cannontech.com>

I am looking for a suitable scripting language for a multi-threaded
server application.  I am wondering if python might be a fit.  The
server must be able to run multiple scripts concurrently (in separate
worker threads) and must be able to stop any of these scripts cleanly at
any time.  The scripting language must also allow writing C extensions -
I am believe python meets this req.  I am not so sure about the threaded
part.  If python cannot meet the threading requirement is there a
scripting language that can?
Tcl also seems like a candidate, but all of the Tcl docs i've found
avoids the topic of embedding multiple interpreters in a multi-threaded
application like the plague.  Does what i'm looking for even exist?
Any insights are much appreciated.

Aaron Lauinger
Cannon Technologies
http://www.cannontech.com




From bradlosh at my-deja.com  Mon Jul 26 13:35:41 1999
From: bradlosh at my-deja.com (Brad)
Date: Mon, 26 Jul 1999 17:35:41 GMT
Subject: Python GUI
Message-ID: <7ni69d$nc7$1@nnrp1.deja.com>

Is there a Python GUI builder out there anywhere?


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From aahz at netcom.com  Thu Jul  1 12:19:58 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 1 Jul 1999 16:19:58 GMT
Subject: Mailbox, tell and read
References: <1281323312-36476914@hypernet.com> <377B7742.F39CB0A4@prescod.net>
Message-ID: <7lg4fe$e9f@dfw-ixnews5.ix.netcom.com>

In article <377B7742.F39CB0A4 at prescod.net>,
Paul Prescod  <paul at prescod.net> wrote:
>
>First, are there platforms other than DOS/Windows/Win32 where there is a
>difference between text and binary mode? I didn't want to munge things
>myself because I'm not sure if what is the "right" munge on the Mac, for
>instance. Can I just do a search and replace for CR/LF -> CR on all
>platforms, all of the time?

I ran into this last year (while I was still a Perl heretic).
Unfortunately, you *have* to use binary mode in opening a mail file
because one of the legal MIME attachment types is non-encoded binary --
and that means you can't just blindly do a search/replace, either.
Never did come to a completely satisfactory answer.  :-(

I think CR will break on Unix, but I'm not sure.
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From sreekantk at freenet.co.uk  Sun Jul 11 21:40:13 1999
From: sreekantk at freenet.co.uk (Sreekant Kodela)
Date: Mon, 12 Jul 1999 02:40:13 +0100
Subject: Help with strings please
References: <378555CC.8B65F809@freenet.co.uk>
Message-ID: <3789477D.E040B9E6@freenet.co.uk>

I just completed the project. I used &S for ' , &D for " and &R for `
as & is probably not used.

Thanks
sreekant




From fdrake at cnri.reston.va.us  Tue Jul 20 11:55:21 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Tue, 20 Jul 1999 15:55:21 GMT
Subject: GNU info archive updated
Message-ID: <14228.39913.756024.165983@weyr.cnri.reston.va.us>

  If you've been trying to use the GNU "info" version of the Python
documentation but don't want to become bogged down in figuring out how 
to install info files, you're in luck!
  Well, ok, this isn't such a big deal.  I've added a README to the
distributed package of info files which tells how to install them into 
an existing info directory.  This is in response to a question from a
user; if you've been meaning to ask about how to get more out of the
documentation, let me know!
  Please, don't hesitate to send mail to python-docs at python.org if you 
have any questions or suggestions regarding the documentation.  I try
to make sure all questions are adequately answered, and always welcome 
suggestions for improvement.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From gmcm at hypernet.com  Wed Jul 21 10:58:01 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 21 Jul 1999 09:58:01 -0500
Subject: VC++ 6.0 larger exe files explained
In-Reply-To: <cxsW$GA6scl3EwNs@jessikat.demon.co.uk>
Message-ID: <1279557219-15543469@hypernet.com>

Robin Becker wrote:
> <rant>
> As part of the Wintel push to take over the world it seems as though
> VC++ 6.0 will now only generate files with sections 0x1000 (4k) or
> longer. Under 5.0 and before the linker would allow much smaller exe
> file sections. So python.exe which used to be 5k becomes 20k.
> Needless to say this should also apply to dll's etc and appears to
> offer no obvious advantage. Bill has never been other than cavalier
> with our space time resources. </rant>

Ay yi yi. The whole PE header is full of convolutions so the loader 
can take a packed disk image and align the sections on the fly.

I guess this "feature" designed to "optimize" load time. Or maybe MS 
has been investing in drive manufacturers...

- Gordon



From wtanksle at dolphin.openprojects.net  Tue Jul 27 20:15:20 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Wed, 28 Jul 1999 00:15:20 GMT
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com>
Message-ID: <slrn7psisn.da.wtanksle@dolphin.openprojects.net>

On Tue, 27 Jul 1999 18:35:10 -0500, Gordon McMillan wrote:
>William Tanksley wrote:
>> On Tue, 27 Jul 1999 14:29:43 +0200, Thomas Wouters wrote:

>> >4.4 break and continue Statements, and else Clauses on Loops 

>> >Loop statements may have an else clause; it is executed when the loop
>> >terminates through exhaustion of the list (with for) or when the condition
>> >becomes false (with while), but not when the loop is terminated by a break
>> >statement. 

>> Well, I'm suprised.  I had also expected the else clause to run on
>> empty loop.  It seems so natural, in line with the meaning of a loop
>> (loop on this; otherwise do that).

>> Fortunately I've never actually used it and thus been disenchanted.

>Maybe it violated your expectations, but it's much more valuable than 
>what you expected. After all, testing for an empty sequence is a 
>no-brainer. But finding a match in a list, and testing whether you 
>fell off the end without finding one, is (without the else clause) a 
>much messier proposal.

First of all, meeting expectations is one reason why I like Python, and
therefore to me is a very valuable quality.  I would go so far as to say
that it's more valuable than any amount of utility to people who have
learned and happen to remember the shibboleth.

I would also disagree with you that my interpretation would result in
disutility: you'd merely have to use exceptions instead of 'break'.  This
is more in line with the way the rest of Python works, and makes the
effect of the abnormal termination of the loop perfectly clear.

The very worst part of the current else: behavior is that it changes the
meaning of else.  In other constructs, else is an alternate path to take
if the data being tested fails a single expression test.  In this
contruct, else is a path taken if the code block belonging to the previous
test executes a certain instruction.

Pretty odd, eh?

class neverMind(): pass

try:
 for x in [1,2,3]:
  print x
  if x == 2: raise neverMind
except neverMind:
  print "modern else clause here"

>- Gordon

-- 
-William "Billy" Tanksley



From pduffin at mailserver.hursley.ibm.com  Wed Jul 21 07:59:27 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Wed, 21 Jul 1999 12:59:27 +0100
Subject: Tcl/Python: Who was first?
References: <932552685.620225@news.adtranzsig.de> <3795ACD5.FCFF67A1@equi4.com>
Message-ID: <3795B61F.71EB@mailserver.hursley.ibm.com>

Jean-Claude Wippler wrote:
> 
> Dirk-Ulrich Heise wrote:
> >
> > I just saw a message called
> >
> > README: Tcl-URL! - the weekly guide to Tcl resources
> >
> > on comp.lang.tcl.announce. (Reading it for the first time in my life)
> > Did Python steal this (good) idea from Tcl or was it the other way round?
> 
> No and no.
> 
> This was first started for Tcl by yours truly last year, and quickly

Oops, looks like I was wrong.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From gmcm at hypernet.com  Fri Jul  2 16:22:44 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 2 Jul 1999 15:22:44 -0500
Subject: execfile(), import and namespaces
In-Reply-To: <000201bec4a3$0f909700$c7326880@goldeneye>
Message-ID: <1281178908-45167423@hypernet.com>

Nikolas Kauer wrote:

> We use a modification of "Listing One" of Andrew Kuchling's "A CGI
> Framework in Python" article in Web Techniques (February 1998) for
> our dynamic Web site. [...]  The object files would be loaded as
> modules with 'import'.
> 
> Problem: In the following code snipet environment variables are
> passed to the executed Python file via the corresponding global
> namespace argument of the execfile() function; whereas the 'import'
> statement does not allow for such arguments.

But __import__ does. Check the library docs under built-in-functions.

- Gordon



From breadly at my-deja.com  Tue Jul 20 08:14:34 1999
From: breadly at my-deja.com (breadly at my-deja.com)
Date: Tue, 20 Jul 1999 12:14:34 GMT
Subject: debugging MSVC DLLs
References: <C2AA61E5CAA7D211832F0008C7A4138E01B486DB@eurmsgneu01.eur.autodesk.com>
Message-ID: <7n1p73$pvt$1@nnrp1.deja.com>

Cool,

I know nothing about python and was just looking around and I get to
post an answer ;)

  mark.mcmahon at eur.autodesk.com wrote:
> I think that this is possible using the Additional DLL's in the
> Debug tab in MSVC...

When I was trying to do this sort of stuff, I tried to use that tab in
MSVC, but unfortunately (no Microsoft flaming required) it worked
inconsistently. By that I mean it worked on an NT machine but not
another NT machine and not another W98 machine. I never figured out why.
>
> -----Original Message-----
> From: Tom Adelman [mailto:ta16 at cornell.edu]
> Sent: Monday, July 19, 1999 5:08 PM
> To: python-list at python.org
> Subject: debugging MSVC DLLs
>
> I'm wondering if someone can help me with this (or should I direct
this
> question to a Microsoft group?):

don't bother, I don't read the microsoft group ;)

My experience with this stuff should apply to you.
As long as you compile all your DLLs with the debug flags and include
database information (Project Settings -> C/C++ -> Debug-Info)
then the debugger will be able to pick them up if you open the code and
set a breakpoint.

As I said before I had some problems/inconsistencies. Try the
Additional DLL tab first - add the full path of the DLL and make sure a
debug database file (I think it is .pdb) is in the same place.

If that doesn't work, try putting the DLL and .pdb (if thats right) in
a directory in the path.

If this is working right, you should be able to step in from one to the
other.

hope that helps...
now back to python stuff...

brad.

--
------------------
Note this is not an email account, just used for
posting


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From tim_one at email.msn.com  Wed Jul 21 00:57:25 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 21 Jul 1999 00:57:25 -0400
Subject: How fast can we multiply?
In-Reply-To: <m2wvvvqwjb.fsf@netmeg.net>
Message-ID: <000401bed335$862ee420$a72d2399@tim>

[Les Schaffer]
> i dont see the connection between floating point matrix multiplies and
> the integer arithmetic tim sketched out in the previous post.

Knuth Volume 2 is a good reference, for this and all other vital questions
in life <wink>.

Turns out it's possible to multiply two 2x2 matrices using 7 multiplies.
Since matmult can be "blocked", a larger matrix can be treated as a 2x2
array of smaller submatrices, and the same trick recursively applied.  In
the end you get a method with runtime O(n**log2(7)) ~= O(n**2.81) instead of
O(n**3).

The intmult trick is less work to code, and yields a bigger payoff earlier.

much-like-python-itself-ly y'rs  - tim





From bruce_dodson at bigfoot.com  Thu Jul 29 21:14:19 1999
From: bruce_dodson at bigfoot.com (Bruce Dodson)
Date: Thu, 29 Jul 1999 22:14:19 -0300
Subject: Is "shelve" thread-safe on Windows?
References: <7npimc$l9b@romeo.logica.co.uk>
Message-ID: <Z37o3.41724$jl.29736754@newscontent-01.sprint.ca>

I see a bsddb.pyd included with Python 1.5.1, and I assume with 1.5.2.  I
think this means Python shouldn't fall back to dumbdbm on a normal
installation.  If I am wrong about that, e.g. if bsddbm.pyd is based on the
1.x series dbm that didn't support concurrent use, I think you can get a
safe version of dbm from www.SleepyCat.com, or maybe the gdbm from
www.roth.net/libs.  It should be quite simple to plug in one of those to the
bsddbm or gdbm Python module source.




From salzr at certco.com  Tue Jul 27 16:40:05 1999
From: salzr at certco.com (Rich Salz)
Date: Tue, 27 Jul 1999 16:40:05 -0400
Subject: Embedding python/multi-thread env
References: <1279107155-5981362@hypernet.com>
Message-ID: <rpu7ag$0$37nspbi$7gc@news.supernews.com>

You're right, I wasn't clear enough.  You can create multiple interpreters,
but
you still must serialize all python calls -- e.g., don't have two threads
trying to
create an interpreter at the same time.

Gordon McMillan wrote in message <1279107155-5981362 at hypernet.com>...
Rich Salz wrote:

> I believe that you cannot create multiple interepreters in a single
> executable. The python global lock is to prevent *threads* within
> the interpreter from stomping on each other's critical sections.

Sure can; it's part of the API. They still share the global lock, but
they can have completely different sets of modules loaded, etc.

- Gordon







From heiland at ncsa.uiuc.edu  Wed Jul  7 10:12:46 1999
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Wed, 07 Jul 1999 09:12:46 -0500
Subject: Py_BuildValue, PyArg_ParseTuple
References: <37835CB3.2B876BE3@ncsa.uiuc.edu>
Message-ID: <3783605D.D063A377@ncsa.uiuc.edu>

as an addendum to my question, I should have checked the return code of
PyArg_ParseTuple... it seems that it always fails.  Why?





From gmcm at hypernet.com  Tue Jul 13 16:41:08 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 13 Jul 1999 15:41:08 -0500
Subject: Recursive method
In-Reply-To: <378B499B.4301F893@eml.villa-bosch.de>
Message-ID: <1280227844-6866579@hypernet.com>

Ralph Gauges writes:

> ... (see code below). Unfortunately all I get is a
> 'NameError' for makenodes, when the function is supposed to
> call itself. 
...
> 
>  def makeTree():
>                 .
>                 .
>             def makenodes(C):
> 

The problem is the nested definition. The name "makenodes" is local 
to makeTree and makeTree's locals are not visible inside makenodes.

Unnest makenodes and it should work.

- Gordon



From stidolph at origin.ea.com  Thu Jul 22 18:00:57 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Thu, 22 Jul 1999 22:00:57 GMT
Subject: how do I find memory leaks?
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC23@molach.origin.ea.com>

I posted a module called Debug to track down memory leaks.  It requires
specific code be added to your __init__ and __del__ routines, as well as
having a Destroy method we use to eliminate circular references.  If you
want it I will post it again.

-----Original Message-----
From: Greg Ewing [mailto:greg.ewing at compaq.com]
Sent: Thursday, July 22, 1999 4:04 PM
To: python-list at cwi.nl
Subject: Re: how do I find memory leaks?


"Heather A. Drury" wrote:
> 
> I thought using python/vtk
> would take care of all this memory garbage for me

A reasonable expectation, but unfortunately Python
doesn't quite deliver. Garbage collection in Python
is by reference counting only, so if there are any
cycles in your garbage, it never goes away.

You have to design your program so that it either
doesn't create any cycles or ensures that they are
broken when the data is no longer needed.

There was a module called Cyclops announced here
by Tim Peters recently for finding cycles in data 
structures. Take a look back in dejanews.

Greg



From tim_one at email.msn.com  Thu Jul  1 11:53:50 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 1 Jul 1999 11:53:50 -0400
Subject: TopSort in Python?
In-Reply-To: <7lfjn0$b1v$1@nnrp1.deja.com>
Message-ID: <000701bec3d9$ea2586c0$5d9e2299@tim>

[Dinu C. Gherman]
> Does anybody have a simple-minded-but-working full-Python
> implementation of topsort, the topological sorting algorithm?
> Or maybe *any* topsort? I remember only one such algorithm...
> python.org/search doesn't reveal any...

Searching for "topological" instead turns up two hits there, a few more on
DejaNews.  Apparently "the std" algorithm I posted years ago predates
DejaNews, though.  So here's a fancier version ...

check-out-aaron's-kjbuckets-too-ly y'rs  - tim

# topsort takes a list of pairs, where each pair (x, y) is taken to
# mean that x <= y wrt some abstract partial ordering.  The return
# value is a list, representing a total ordering that respects all
# the input constraints.
# E.g.,
#    topsort( [(1,2), (3,3)] )
# may return any of (but nothing other than)
#    [3, 1, 2]
#    [1, 3, 2]
#    [1, 2, 3]
# because those are the permutations of the input elements that
# respect the "1 precedes 2" and "3 precedes 3" input constraints.
# Note that a constraint of the form (x, x) is really just a trick
# to make sure x appears *somewhere* in the output list.
#
# If there's a cycle in the constraints, say
#    topsort( [(1,2), (2,1)] )
# then CycleError is raised, and the exception object supports
# many methods to help analyze and break the cycles.  This requires
# a good deal more code than topsort itself!

from exceptions import Exception

class CycleError(Exception):
    def __init__(self, sofar, numpreds, succs):
        Exception.__init__(self, "cycle in constraints",
                           sofar, numpreds, succs)
        self.preds = None

    # return as much of the total ordering as topsort was able to
    # find before it hit a cycle
    def get_partial(self):
        return self[1]

    # return remaining elt -> count of predecessors map
    def get_pred_counts(self):
        return self[2]

    # return remaining elt -> list of successors map
    def get_succs(self):
        return self[3]

    # return remaining elements (== those that don't appear in
    # get_partial())
    def get_elements(self):
        return self.get_pred_counts().keys()

    # Return a list of pairs representing the full state of what's
    # remaining (if you pass this list back to topsort, it will raise
    # CycleError again, and if you invoke get_pairlist on *that*
    # exception object, the result will be isomorphic to *this*
    # invocation of get_pairlist).
    # The idea is that you can use pick_a_cycle to find a cycle,
    # through some means or another pick an (x,y) pair in the cycle
    # you no longer want to respect, then remove that pair from the
    # output of get_pairlist and try topsort again.
    def get_pairlist(self):
        succs = self.get_succs()
        answer = []
        for x in self.get_elements():
            if succs.has_key(x):
                for y in succs[x]:
                    answer.append( (x, y) )
            else:
                # make sure x appears in topsort's output!
                answer.append( (x, x) )
        return answer

    # return remaining elt -> list of predecessors map
    def get_preds(self):
        if self.preds is not None:
            return self.preds
        self.preds = preds = {}
        remaining_elts = self.get_elements()
        for x in remaining_elts:
            preds[x] = []
        succs = self.get_succs()

        for x in remaining_elts:
            if succs.has_key(x):
                for y in succs[x]:
                    preds[y].append(x)

        if __debug__:
            for x in remaining_elts:
                assert len(preds[x]) > 0
        return preds

    # return a cycle [x, ..., x] at random
    def pick_a_cycle(self):
        remaining_elts = self.get_elements()

        # We know that everything in remaining_elts has a predecessor,
        # but don't know that everything in it has a successor.  So
        # crawling forward over succs may hit a dead end.  Instead we
        # crawl backward over the preds until we hit a duplicate, then
        # reverse the path.
        preds = self.get_preds()
        from random import choice
        x = choice(remaining_elts)
        answer = []
        index = {}
        in_answer = index.has_key
        while not in_answer(x):
            index[x] = len(answer) # index of x in answer
            answer.append(x)
            x = choice(preds[x])
        answer.append(x)
        answer = answer[index[x]:]
        answer.reverse()
        return answer

def topsort(pairlist):
    numpreds = {}   # elt -> # of predecessors
    successors = {} # elt -> list of successors
    for first, second in pairlist:
        # make sure every elt is a key in numpreds
        if not numpreds.has_key(first):
            numpreds[first] = 0
        if not numpreds.has_key(second):
            numpreds[second] = 0

        # if they're the same, there's no real dependence
        if first == second:
            continue

        # since first < second, second gains a pred ...
        numpreds[second] = numpreds[second] + 1

        # ... and first gains a succ
        if successors.has_key(first):
            successors[first].append(second)
        else:
            successors[first] = [second]

    # suck up everything without a predecessor
    answer = filter(lambda x, numpreds=numpreds: numpreds[x] == 0,
                    numpreds.keys())

    # for everything in answer, knock down the pred count on
    # its successors; note that answer grows *in* the loop
    for x in answer:
        assert numpreds[x] == 0
        del numpreds[x]
        if successors.has_key(x):
            for y in successors[x]:
                numpreds[y] = numpreds[y] - 1
                if numpreds[y] == 0:
                    answer.append(y)
            # following "del" isn't needed; just makes
            # CycleError details easier to grasp
            del successors[x]

    if numpreds:
        # everything in numpreds has at least one predecessor ->
        # there's a cycle
        if __debug__:
            for x in numpreds.keys():
                assert numpreds[x] > 0
        raise CycleError(answer, numpreds, successors)
    return answer





From skip at mojam.com  Mon Jul 26 17:07:03 1999
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 26 Jul 1999 16:07:03 -0500 (CDT)
Subject: filter() + simple program not outputting
In-Reply-To: <379CC972.750BD42D@netcard.com>
References: <379CC972.750BD42D@netcard.com>
Message-ID: <14236.52569.79310.914826@dolphin.mojam.com>

    Joyce> can someone tell me why this simple program is not outputting
    Joyce> when i run from the BASH prompt

    Joyce> #!/usr/bin/env python
    Joyce> def f(x):
    Joyce>         return x%2 != 0 and x%3 != 0
    Joyce>         filter(f, range(2, 25))
    Joyce>         f( )

You need to call f from the module level (the calls to filter and f above
are from inside the scope of the function f).  Try:

    #!/usr/bin/env python
    def f(x):
	    return x%2 != 0 and x%3 != 0
    print filter(f, range(2, 25))

instead.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From aaron_watters at my-deja.com  Tue Jul 20 09:54:29 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Tue, 20 Jul 1999 13:54:29 GMT
Subject: Python/Gadfly and SQL wildcards
References: <1279735665-4809662@hypernet.com>
Message-ID: <7n1v2d$sd6$1@nnrp1.deja.com>

In article <1279735665-4809662 at hypernet.com>,
  gmcm at hypernet.com wrote:
> Colm Rafferty writes:
>
> > Does anyone know how to implement the wildcard in gadfly SQL?
> > e.g.
> > Select * from table where fieldname = 'A*'
> > or
> > Select * from table where fieldname like 'A*'
> >
> > * does not work!
> > Though it does in other SQL languages
>
> In ANSI SQL, that would be "fieldname like 'A%'"
>
> Last I used Gadfly, it didn't support 'like'

Yes and no.  I actually posted an extension view for gadfly that
allowed a more powerful, but highly awkward, analogue to "LIKE"
which is based on the python "re" regular expression matching
library.

  http://www.egroups.com/group/gadfly-rdbms/149.html?raw=1

In this case you'd translate

  Select * from table where fieldname like 'A*'

into

  SELECT * FROM table WHERE EXISTS
  (SELECT * FROM re where table.fieldname=re.string AND
                          re.pattern='A.*')

Here we use a "virtual infinite table" re(string, pattern)
which "includes" all possible pairs where pattern is an "re"
syntax regular expression and string matches the pattern.

Unfortunately you have to use "EXISTS" to force the gadfly
optimizer to not screw up at present.  I'm working on fixing
the query engine and optimizer to work better with predicates
like LIKE and other extensions -- so I'm holding off on adding
features to the current way of doing things.

   Aaron Watters http://www.chordate.com

===

What's a chicken crossed with a dog?
dog chicken sine theta


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gerrit.holl at pobox.com  Thu Jul  1 11:44:16 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 1 Jul 1999 17:44:16 +0200
Subject: getopt and Non-Unixes?
Message-ID: <19990701174416.A27152@optiplex.palga.uucp>

Hello all,

I was wondering how getopt worked on dos/windows (and other operating systems).
While in dos, the default way to handle arguments is "/v /d /f" (isn't it?).

I don't have access to any micro$oft machine (it doesn't make me unhappy...),
but I just wanted to know how my program works on Dos/windows/macintosch and
other non-Unix systems.

regards,

Gerrit.



From larsga at ifi.uio.no  Wed Jul 14 15:33:36 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 14 Jul 1999 21:33:36 +0200
Subject: WWW/urllib.urlretrieve problems
References: <Pine.LNX.4.04.9907141751130.9626-100000@emerald.netskate.ru>
Message-ID: <wk908jkovj.fsf@ifi.uio.no>

* Oleg Broytmann
| 
|    First problem URL is http://www.expert.ru/ - it just timeouts.
| When I pointed Netscape to the address - the page appeared well. Lynx and
| telnet timed out too!
|    Don't understand it. What in Netscape Communicator is so good?

It works for me when using telnet...
 
| Second problem is with http://w3.one.net/~alward/. Netscape and lynx
| showed the page, urllib.urlretrieve() and telnet returned error403 -
| forbidden:
| 
| ---------- Session ----------
| phd at emerald 204 >>> t w3.one.net 80
| Trying 206.112.192.125...
| Connected to w3.one.net.
| Escape character is '^]'.
| GEt /~alward/ HTTP/1.0
| Host: w3.one.net

Insert 'User-agent: Mozilla/3.0' and you get the page. The page
doesn't seem to contain anything special, so I've no idea why they
block non-Mozilla user-agents.

An email to the server admin might be an idea.
 
--Lars M.



From rceci at my-deja.com  Fri Jul  2 02:27:07 1999
From: rceci at my-deja.com (rceci at my-deja.com)
Date: Fri, 02 Jul 1999 07:27:07 +0100
Subject: 1.5.1 on IRIX
Message-ID: <00001655@bossar.com.pl>

From: rceci at my-deja.com

Andrew,

Took your suggestions - moved the libs back under a python1.5 subdir,
and set PYTHONHOME.  Here is my results (this the output of my csh
script running the relocated python runtime ):

PYTHONPATH=/usr/mdi_dev//python/lib/irix32/python1.5
PYTHONHOME=/usr/mdi_dev//python/lib/irix32
python=/usr/mdi_dev//python/bin/irix32/python
Could not find platform independent libraries <prefix>
Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
# /usr/mdi_dev//python/lib/irix32/python1.5/exceptions.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/exceptions.py
import exceptions # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/exceptions.pyc
# /usr/mdi_dev//python/lib/irix32/python1.5/site.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/site.py
import site # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/site.pyc
# /usr/mdi_dev//python/lib/irix32/python1.5/os.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/os.py
import os # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/os.pyc
import posix # builtin
# /usr/mdi_dev//python/lib/irix32/python1.5/posixpath.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/posixpath.py
import posixpath # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/posixpath.pyc
# /usr/mdi_dev//python/lib/irix32/python1.5/stat.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/stat.py
import stat # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/stat.pyc
# /usr/mdi_dev//python/lib/irix32/python1.5/UserDict.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/UserDict.py
import UserDict # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/UserDict.pyc
Python 1.5.1 (#3, May 21 1999, 16:24:24) [C] on irix646-32-mips2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
# /usr/mdi_dev//python/lib/irix32/python1.5/string.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/string.py
import string # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/string.pyc
import strop # builtin
>>>

I have a PYTHONSTARTUP that simply imports os,sys,string.

It looks like python is loading all of its modules from the correct
place (*/usr/mdi_dev/python"), as opposed to where the interpreter was
told during buildtime ("/usr/local").

I am wondering - on this second machine that is causing the warnings to
be printed out, there is no python installation in /usr/local.  Does
python check its configured buildtime prefix/exec_prefix locations, EVEN
though PYTHONHOME and PYTHONPATH are setup so python is instructed to
get its runtimes from the new location?  That would be a bummer, because
we are
distributing binaries to our customers, and need to put the distro in
any arbitrary customer-specified location, and having these warnings pop
up would probably cause a lot of confusion.

Thanks

Rob




Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From graham at sloth.math.uga.edu  Sat Jul 17 12:21:17 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 17 Jul 1999 16:21:17 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com> <7mnirc$t2b$2@cronkite.cc.uga.  edu> <170719990020380651%pereira@research.att.com>
Message-ID: <7mqaht$r27$2@cronkite.cc.uga.edu>

<graham at sloth.math.uga.edu> wrote:
: > But the class of 
: > languages recognisable by automaton doesn't cover all the languages of
: > interest (in fact my guess is it covers very few ... context ...).
Fernando Pereira (pereira at research.att.com) wrote:
: In what sense of "very few"? Most programming languages are designed to
: have an LR(k) or LL(k) backbone, and regular lexical structure 

So the regular expressions help you with the regular lexical structure.
But they don't help with the LR(n) bit. They also don't help with context
dependencies. That leaves out a lot of languages.

<graham at sloth.math.uga.edu> wrote:
: > And then
: > there are software engineering issues ...
Fernando Pereira (pereira at research.att.com) wrote:
: Agreed. However, more traditional lexer- and parser-generators had
: sensible answers to many of those issues, that might not be as elegant
: as monadic approaches, but that paid serious attention to optimization
: in the regular and CF backbones. 

I don't really see why using parser combinators means that you cannot
optimise. You are not precluded from using regular expression parsers
as primitive parsers, nor are you precluded from using LALR(n) parsers
as primitive parsers. Hence you can optimise those parts of your
parsing with automaton if you want. So I don't really understand your
criticism.

Also as I said parser combinators are really more of a sequencing 
technique -- one that applies nicely to parsing.

Fernando Pereira (pereira at research.att.com) wrote:
: By the way, I find it faintly amusing
: to see the functional programming community rediscovering and dressing
: in categorial garb a simple idea that was a main motivator for the
: development of logic programming in the early 70s (see
: <http://akpublic.research.att.com:9000/~pereira/bib.html#Pereira+Warren-
: 80:DCGs> for a survey). 

Yes combinators are much like definite clause grammars in prolog. But if
I remember correctly from my use of DCG's, they gave you a fixed set of
combinators? Moreover they were limited to string parsing. The combinator
approach in the fp world has neither limitation.

Fernando Pereira (pereira at research.att.com) wrote:
: It *is* a good idea for rapid prototyping, but
: interspersing pattern matching/parsing with arbitrary computations
: (even purely functional ones) effectively blocks optimization, which is
                                -------------------------------

Can you explain why? I don't see it, since you can have reg expr parsers
and other such fast parsers as primitive parsers that you combine ...

graham
-- 
            The girls is all salty, the boys is all sweet, 
                      the food ain't too shabby 
                    an' they piss in the streets
    In France, way down in France, way on down, way on down in France



From sprout at dok.org  Tue Jul 27 15:35:53 1999
From: sprout at dok.org (Chris Green)
Date: Tue, 27 Jul 1999 19:35:53 GMT
Subject: newbie namespace question
Message-ID: <m3u2qp9981.fsf@ghostplanet.dok.org>

A long time ago, I swore I read something that answered this but I am
unable to find it in any of the docs or faq.

How can a subfunction of a method of a class access the other members
of a class? I'm playing with pygtk and more or less munging what was
already in the test gtk example.

class TimeCalc:
	def __init__:
		[ snip ]
		self.make_list()
	def make_list():
		s=[0 , 0]

		def select_clist(self, row, col, event, selected=s):
			print self.get_text(row,col)	
               self.m_clist.connect("select_row", select_clist)
	def grok_data():
		pass

Ok. Excuse the bad example but when select_clist is called, how can I
access the previous self so that I can use self.grok_data().

I vaguely remember running into this problem before and just
reorganizing my class to get around it.  What is the proper way to do
this? (and if they aren't the same, how do you bump up one level in
namespaces? :) )

Thanks,
Chris
-- 
Chris Green <sprout at dok.org> <grapeape at uab.edu>
Let not the sands of time get in your lunch.



From jsight at mindspring.com  Thu Jul 22 19:01:40 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Thu, 22 Jul 1999 19:01:40 -0400
Subject: Fork, Threading, and Select madness.  Please help!
Message-ID: <3797A2D4.DB293C40@mindspring.com>

Ok, I have been working on a website which is now available at
http://www.biddin.com/ that performs a search of multiple auction sites
all at once, and then returns the results.  The problem is that it is
very slow, and it appears that the primary bottleneck is in waiting on
the search engines at the various sites to begin reading data.

Well, this makes sense because the code to implement the search is
currently pure sequential with no threading or non-blocking IO
whatsoever.  This means that a search request on Amazon can only occur
after a successful completion of a search on EBay (which is notoriously
slow).

What I'm getting around to asking (slowly<g>) is, what is the best way
to implement an algorithm that can retrieve the data from ALL 3 sites
simultaneously, so that the code is never blocked up waiting on one
particular bad site?  Note that it should also be possible to do
timeouts if search site is failing, and it should be easily extensible
to handle more than 3 simultaneous requests in the future.  Also, keep
in mind that threading is my LAST possible choice as the Python
implementation at my presence provider doesn't support Threading.  :)

Suggestions as to where I can read up on implementing this as a series
of forked processes, and whether or not this would be practical are
certainly welcome.  Also, if anyone has any brilliant ideas for how it
could easily be done with the Select mechanism, I'd love to hear it.

---------------
Jesse D. Sightler
http://www3.pair.com/jsight/

PS - Not that it matters, but much of this code will probably be
released under GPL eventually.  :)



From algaba at my-deja.com  Mon Jul 26 10:58:08 1999
From: algaba at my-deja.com (algaba at my-deja.com)
Date: Mon, 26 Jul 1999 14:58:08 GMT
Subject: createfilehandler, how to use it with Tk?
Message-ID: <7nht1s$grn$1@nnrp1.deja.com>

I have to communicate a tk front end with another modules.
And I'm trying:
... tk-inter code

class lectora:
    def __init__(self):
	from socket import *
	HOST = ''    # The remote host
	PORT = 50008              # The same port as used by the server
	self.s = socket(AF_INET, SOCK_STREAM)
	self.s.connect(HOST, PORT)
	self.s.send('Hello, world')

    def el_socket(self, handle, event):
	pass
        #data = handle.recv(1024)

    def close(self):
	self.s.close()
	print 'Received', `data`

l = lectora()

tk_mask =  _tkinter.READABLE

_tkinter.createfilehandler(l.s, tk_mask, l.el_socket )

test = Test()
test.mainloop()

But this segfaults. :(

I've been browsing  fnorb example of use of createfilehandler
, but it's as complex as hell.
has anyone a simple,straightforward example of it?

Thanks


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gjh at bel-epa.delete-this.com  Sun Jul 18 07:34:00 1999
From: gjh at bel-epa.delete-this.com (Graham Higgins)
Date: Sun, 18 Jul 1999 12:34:00 +0100
Subject: PyApache binaries
References: <7mjtlk$6v3$1@nnrp1.deja.com>
Message-ID: <180719991234008478%gjh@bel-epa.delete-this.com>

In article <7mjtlk$6v3$1 at nnrp1.deja.com>, <max at rightworks.com> wrote:

> Could anyone point me to where I can find NT binaries for PyApache.
> thanks
> max

You might care to ask this question of the pyapache-devel list on:

pyapache-devel at msg.com.mx

I maintain the pyapache ftp distribution archives but I don't keep
binaries, sorry.



From jonathan at onegoodidea.com  Fri Jul 30 10:33:21 1999
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Fri, 30 Jul 99 14:33:21 GMT
Subject: Python/PostgreSQL interface module
Message-ID: <mt2.0-24150-933345201@news.informatik.uni-bonn.de>

well after much lurking and occasional commenting on these groups, here is
my first contribution to the community. it's a module for accessing PostgreSQL
databases from Python.

yes, i know there already is one, but i wanted to play with some different
features and to get the hang of writing extension modules. you are free to not
use this module. please don't email me to tell me i've wasted my time - all
learning is worthwhile and maybe someone else will get something useful from
this.

i'm afraid it doesn't have a proper name yet because i can't think of one - a
free copy of the source for the best idea ;-)

i welcome feedback,

:-j



(extracted from the README:)


Python/PostgreSQL Interface Module                         (Version: 1.0)
----------------------------------


Summary:
--------

PostgreSQL is a neat, free, relational database system which can be found at:

   <http://www.postgresql.org/>

This is a module that provides the PostgreSQL programmers' API to Python
programmers. It is written in C and compiles down to a shared-object file
that can be dynamically loaded into the Python interpreter. It is basically
a wrapper around the 'libpq' library that comes with PostgreSQL.


Limitations:
------------

This interface has not been used in anger yet. It's been written as part of
a one good idea project and I'm releasing it more widely because it may be
of use to other people and might result in some useful feedback.


License:
--------

There should be a LICENSE file accompanying this distribution (and at the top
of each source file). I've used the Berkeley license because I've always
liked it. This code is not copylefted, feel free to stick it into a commercial
product and sell it, become rich, buy a small island in the pacific, retire
young and happy. However, you have to retain the copyright on the code, and
it would be polite to send me a note of thanks ;-)


Obtaining The Latest Distribution:
----------------------------------

You should be able to find it by following the obvious links on the one good
idea webpages for 'Projects':

   <http://www.onegoodidea.com/>

As of this distribution, the appropriate URL was:

   <http://www.onegoodidea.com/downloads/python-postgresql-1.0.tgz>


PyGreSQL
--------

This module is not PyGreSQL. PyGreSQL is written by D'Arcy J.M. Cain,
<darcy at druid.net>. It is another Python/PostgreSQL interface which has been
around longer than this one. I wrote this module because I wanted a different
mechanism for accessing named fields in query results (and partly as an
exercise in extending Python).

== 
jonathan hogg, one good idea limited, 25 clouston street, glasgow g20 8qr, uk
jonathan at onegoodidea.com www.onegoodidea.com t:(0)976-614338 f:(0)7970-537451

<P><A HREF="http://www.onegoodidea.com/projects.html">Python/PostgreSQL
API</A> - access PostgreSQL programmers' API from Python.  (29-Jul-99)

-- 
----------- comp.lang.python.announce (moderated) ----------
Article Submission Address:  python-announce at python.org
Python Language Home Page:   http://www.python.org/
Python Quick Help Index:     http://www.python.org/Help.html
------------------------------------------------------------



From kas at maps.magnetic-ink.dk  Sat Jul 17 02:47:40 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Sat, 17 Jul 1999 08:47:40 +0200
Subject: Snippet: bitcount()
References: <slrn7oukop.68v.kas@titan.magnetic-ink.dk> <14223.51006.317025.189165@lrz.de>
Message-ID: <slrn7p09o9.ahf.kas@titan.magnetic-ink.dk>

Eugene Leitl <eugene.leitl at lrz.uni-muenchen.de> wrote:

> Ouch, ouch. That was really _inefficient_. At least you could nybble
> or byte-wise lookup, and there should be lots of bigmask magic which
> does in in O(logN). Consider bitcounting L's.
>
>  > def bitcount(x):
>  >     n = 0
>  >     while x:
>  >         x = x & (x-1)
>  >         n = n + 1
>  >     # end while
>  >     return n
>  > # end def bitcount

I don't know about nybble or byte-wise lookups.  The algorithm above does
the job in O(N) where N is the number of bits in the number, even on huge
numbers -- or, to put it in simple words: it doesn't spend time counting
all the 0's.

Let's consider the two big twin primes:

  BIGP1 = 1706595L * (1L << 11235) - 1L
  BIGP2 = 1706595L * (1L << 11235) + 1L

Both numbers have 3389 digits, BIGP1 has 11243 bits whereas BIGP2 has 10.
If you time the execution you will see that counting the bits in BIGP2 is
*very* close to 11243/10 times faster than counting the bits in BIGP1.

Now, I'm not a mathematician and so it is difficult for me to imagine how
I can come up with an algorithm that is actually faster than only counting
the 1's and essentially ignoring the 0's -- which is what the algorithm
above claims to do -- unless I decide to count several bits at a time, but
I wouldn't know how to construct such a thing.

Please prove me wrong. :-)


Cheers,

  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From joe at strout.net  Tue Jul 13 11:38:40 1999
From: joe at strout.net (Joe Strout)
Date: Tue, 13 Jul 99 15:38:40 GMT
Subject: PIDDLE 1.0a3
Message-ID: <mt2.0-29824-931880320@news.informatik.uni-bonn.de>

I'm pleased to announce the third alpha release of PIDDLE (Plug-In
Drawing, Does Little Else).  PIDDLE is a cross-platform, cross-media
drawing API; by choosing the desired backend renderer, your Python
program can generate output to the PDF, EPS, TK, MacOS QuickDraw, a PIL
canvas (and from there to PNG, GIF, JPEG, etc.), and others.

This version adds a unified testing framework, and a number of bug
fixes.  To download, visit the PIDDLE home page:

      http://www.strout.net/python/piddle/


PIDDLE is written in 100% Python, but some backends require extensions
(such as PIL) or may work on only certain platforms (such as MacOS
QuickDraw).

Comments and feedback are welcome; visit the mailing list at
http://www.egroups.com/list/pythonpiddle/info.html to join the
discussion.

Cheers,
-- Joe

== 
,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe at strout.net             http://www.strout.net              |
`------------------------------------------------------------------'

<P><A HREF="http://www.strout.net/python/piddle/">PIDDLE v1.0a3</A> -
drawing API with backends that do the actual drawing to a variety of
formats, including PDF, Postscript, MacOS QuickDraw, Tk, PIL, and
eventually OpenGL and Windows.  (12-Jul-99)

-- 
----------- comp.lang.python.announce (moderated) ----------
Article Submission Address:  python-announce at python.org
Python Language Home Page:   http://www.python.org/
Python Quick Help Index:     http://www.python.org/Help.html
------------------------------------------------------------



From John.Michelsen at p98.f112.n480.z2.fidonet.org  Sat Jul  3 13:06:28 1999
From: John.Michelsen at p98.f112.n480.z2.fidonet.org (John Michelsen)
Date: Sat, 03 Jul 1999 18:06:28 +0100
Subject: Python GUIs: Summary and Conclusion
Message-ID: <000016ab@bossar.com.pl>

From: "John Michelsen" <john.michelsen at gte.net>


>>Well, the only reason I haven't been there is that I've only worked on
>>one GUI app _ever_ - a 2D tile-based strategy game [1].  I did plan
>>support for console apps (curses, etc., using a client-server
>>architecture), but I didn't much fancy the idea of playing with a
>>command line interface :)
>
>I'm working on a 2D tile-based adventure game, which right now is text
>only.  I'm expanding it so that it can use graphics, but I need some
>libraries.  Let me know what you find!


I'm writing a Tkinter/PIL based client for the Worldforge project.
Check out:
http://www.worldforge.org/website/client/tkClient/

John




From yishai at platonix.com  Tue Jul  6 08:35:43 1999
From: yishai at platonix.com (Yishai Beeri)
Date: Tue, 06 Jul 1999 15:35:43 +0300
Subject: Newbie question: embedding python in a c++ main
Message-ID: <3781F81F.D8730AEE@platonix.com>

Hello,

[If the answers to this questions are available in the python docs,
please direct me. I did not find direct answers there].

I am trying to embed the python interpreter in a c++ program. All it
does is a simple main, which initializes python, initializes several
python modules (extensions written in c++), and run a python program.

When I read the python program into a char* and use PyRun_SimpleString,
it seems to work fine. However, when I try to
use PyRun_SimpleFile, after opening the file, I get a crash somewhere
inside the PyParser functions.

Does anyone have a working example of what I'm trying to do? I suspect
the problem might lurk in c/c++ incompatibilities, but I'm not sure.

Thanks,
Yishai Beeri




From kperacles at geocities.com  Fri Jul 23 11:10:07 1999
From: kperacles at geocities.com (Karl Putland)
Date: Fri, 23 Jul 1999 09:10:07 -0600
Subject: COM bites - help please
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F9C@RED-MSG-50>
Message-ID: <iG%l3.18$CP5.3091@news.uswest.net>

Bill Tutt <billtut at microsoft.com> wrote in message news:4D0A23B3F74DD111ACCD00805F31D8100DB90F9C at RED-MSG-50...
> 
> 
> > From: Karl Putland [mailto:kperacles at geocities.com]
> 
> > I've got a couple of pyCOM things that are kicking my @$$! 
> > 
> > 1.  Trying to wrap a list of instances into a 
> > util.Collection.  All instances have been prepped with 
> > _public_methods_ and _public_attrs_ and have been wrapped 
> > with a call to win32com.server.util.wrap(field) Trying to 
> > access the Collection from a com instance produces the 
> > following error.
> > 
> [...]
> > >>> c1.fields        # This is a util.Collection(list of 
> > GMField instances)
> > Traceback (innermost last):
> >   File "<pyshell#56>", line 1, in ?
> >     c1.fields
> >   File "D:\Python\win32com\client\dynamic.py", line 394, in 
> > __getattr__
> >     raise pythoncom.com_error, details
> > com_error: (-2147467259, 'Unspecified error', None, None)
> > 
> 
> Not having the source makes it more difficult to help here...
> If memory serves given a list of Python instances that can be wrapped as COM
> objects, the way to make a collection is:
> 

I can e-mail the source if you would like to take a look.  I thought it was going to be to large.  There are four tightly connected modules and most of it is specific to GoldMine.  I couldn't really reduce it beyond what it is.

> collection = util.Collection(map(lambda x: util.wrap(x), list)
> 

Thanks,
Kind of what I've done, but this is how I had it done with the error being produced.

def gm_com_table_prep(table_inst):
    """Preppare for COM wrapper of gmtasble.GMTable

    Need to see if we can generate the COM specific attributes on a per
    instance basis.

    The GMTable class is not pulicly accesable.  GMTables are only created by
    the factory in GMcomConnector.  They get wrapped in the COM interface
    there.  This just gives each instance the appropriate COM attributes

    """

    # Changes the date format to a string from a date tuple
    table_inst.COM = 1
    
    table_inst._public_methods_ = ['append',
                        'bottom',
                        'close',
                        'delete',
                        'filter',
                        'goto',
                        'is_sql',
                        'new',
                        'open',
                        'range',
                        'read',
                        'replace',
                        'search',
                        'seek',
                        'set_order',
                        'skip',
                        'tables',
                        'top',
                        'unlock']

    table_inst._public_attrs_ = ['name',
                                 'handle',
                                 'records',
                                 'rec_no',
                                 'fields',
                                 'fieldnames']
    if debug: print "\nLooping over fields to set _public_attrs_ for table", table_inst.name
    for name in table_inst.fieldnames:
        table_inst._public_attrs_.append(string.lower(name))
    if debug: print "\nWrapping Fields"
    tmpfields=[]
    for field in table_inst.fields:
        field._public_methods_ = ['read',
                                  'write',
                                  'append']
        field._public_attrs_ = ['name',
                                'size',
                                'type'
                                'value']
        name = string.lower(field.name)
        field = win32com.server.util.wrap(field)
        tmpfields.append(field)
        setattr(table_inst, name, field)
    table_inst.fields=tmpfields


> > 3. PyUnicode objects bit me in the ass when strings are 
> > passed into methods and manipulated.
> > class Foo:
> >     <COM stuff>
> >     def bar(self, somestring):
> >         somestring = string.upper(somestring)
> >         return somestring
> > 
> > >>>foo = win32com.client.Dispatch("Python.Foo")
> > >>>foo.bar("Hello World")
> > 
> 
> PyUnicode strings have an upper and a lower method so this becomes:
> def bar(self, somestring):
> somestring = somestring.upper()
> return somestring
> 

Sorry simplistic example.  Don't remember the exact error but I had to do it here and I don't even manipulate the string.

    def seek(self, indexvalue=''):
        """Not quite sure how to handle this.
        
        Currently no err is exact match.
        0 == Error
        1 == Exact match - no error
        2 == Close match - no error???
        """

        # String conversion for COM
        indexvalue = str(indexvalue)
        
        res = self.gmdll.db_seek(self.handle, indexvalue)
        if res == 0:
            raise GMSeekError, 0
        elif res == 2:
            raise GMMatchError, indexvalue

> > Some traceback about TypeError PyUnicode.
> > 
> > Causing a need to change bar to
> >     def bar(self, somestring):
> >         somestring=str(somestring)
> >         somestring = string.upper(somestring)
> >         return somestring
> > This change needs to be made in ALL methods that accept and 
> > manipulate strings???
> > 
> 
> This is the bane of Unicode, the string module can't handle them.
> You either have to pass Unicode into functions that like Unicode, or convert
> the Unicode into a normal Python string. Not much else you can do about
> this.
> It's been rumored that Unicode support will get better in Python 1.6.
> 

That would be good.

> > Is there a better method than COM for allowing someone who 
> > uses VB access to Python objects, short of converting all of 
> > the VB users to Python or resorting to reimplementing in VB 
> > or C++?  
> 
> Using Python is good. :) But yes, this is the easiest way for Python and VB
> to talk to each other.
> 
> > The COM interface I've been able to expose is 
> > sufficient for now but does not allow some of the access that 
> > I have in Python.  Maybe that's good in the end?
> > 
> 
> I think so. :)
> 
> Hope this helps out a little bit...
> 
> Bill

Thanks Bill.  Just got a little frustrated having to pollute the Python too much.  I would love it if you or someone else could look over what I have so far, but I know all to well the time constraints of family and job.  

I know that I have a couple things to work out.  Like right now each field holds a reference to it's parent table.  And there might be another reference floating around that doesn't let it die.  TP's Cyclops might help me here.  Other than that most of it is pretty solid.

Thanks for the help.

Karl Putland
kperacles at geocities.com




From Jean-Michel.Bruel at univ-pau.fr  Thu Jul  8 03:58:25 1999
From: Jean-Michel.Bruel at univ-pau.fr (Jean-Michel BRUEL)
Date: Thu, 8 Jul 1999 07:58:25 GMT
Subject: [Call for Participation: <<UML>>'99]
Message-ID: <199907080758.JAA10373@crisv4.univ-pau.fr>

[apologies if you receive multiple copies of this announcement]

=================================================================
     Preliminary Call for Participation      <<UML>>'99
=================================================================

 Second International Conference on the
      Unified Modeling Language

 October 28-30, 1999, Fort Collins, Colorado, USA
           (just before OOPSLA)

=================================================================
       http://www.cs.colostate.edu/UML99
=================================================================

Invited Speakers: Grady Booch and David Harel

Registration:

  -------------------------------------------------------------------
  Registration  | IEEE/ACM Members | Non-Members | Full-time students
  -------------------------------------------------------------------
  Before Oct. 6 |       $405       |    $505     |      $ 175       |
  -------------------------------------------------------------------
  After Oct. 6  |       $490       |    $605     |      $ 190       |
  -------------------------------------------------------------------

  Full registration includes the following: session attendance,
  conference lunches and breaks on each day of the conference (3 days), 
  conference reception, conference banquet, and conference proceedings.
  Student registration includes only session attendance and breaks.

Hotel Information (tentative):

  University Park Holiday Inn, 425 W. Prospect Road,
  Fort Collins, CO 80526.
  Single, double, triple, quad rate: $92 per night.
  Cut-off date: Sept. 13, 1999.
							
Further Information:
   Robert B. France             E-mail: france at cs.colostate.edu
   Computer Science Department  Tel:    970-491-6356
   Colorado State University    Fax:    970-491-2466
   Fort Collins, CO 80523, USA

   Bernhard Rumpe               E-mail: rumpe at in.tum.de
   Institut fuer Informatik     Tel:    0049-89-289-28129
   T. Universitaet Muenchen     Fax:    0049-89-289-28183
   80290 Muenchen, Germany

=================================================================
Sponsored by IEEE CS Technical Committee on Complexity in Computing.
In Cooperation with ACM SIGSOFT.
With the Support of the OMG and Rational Software. 
UML is a trademark of the OMG.
=================================================================



From tony at lsl.co.uk  Thu Jul 15 11:33:38 1999
From: tony at lsl.co.uk (Tony J. Ibbs (Tibs))
Date: Thu, 15 Jul 1999 15:33:38 GMT
Subject: Meta-language for mxTextTools
Message-ID: <002001beced7$692bc800$f0c809c0@lslp7o.lsl.co.uk>

Christian Tismer <tismer at appliedbiometrics.com> wrote:
> I got the impression: Why another language?

Yes, that always makes me uncomfortable too. Although, really, the original
tuple form is *already* another language (just a very low level one).

> Can't we express the tagging tuples by Python programs?
> Marc's tuples are nothing more than predefined actions,
> conditions, and jumps to labels.
>
> Given a restricted subset of Python, and a couple of known
> functions which are to be used, the tags could be expressed
> as a small number of Python functions, and
> it would be a real application for the ByteCodeHacks
> to turn these Python functions back into tagging tables.
> We would go without any parsing since Python does it for us.


Strangely enough, I have had exactly this thought myself.

I guess the reply is in two parts:

1. I wanted to keep this to something simple that I could have a hope of
implementing myself. In the very, very restricted spare time I have (we have
two small children!) I *might* just be able to use this little language as a
way into learning mxTextTools and produce something I can write quicker than
the tuples themselves. Simple text transforms to produce a "mini-language"
is something I've done before and can give back seemingly amazing progress
for little effort.

2. I can't help feeling that "jump-to-label" (aka
skip-forwards/backwards-N-statements) is intrinsically necessary in such an
engine, and I can't see any way of doing that sensible whilst keeping to
Python syntax.

3. I rather wanted not to hack Python itself, and to keep independent of
whether one is using CPython, JPython (which OK, can't use the current
mxTextTools anyway) or some other language accessing the DLL. Perhaps a bit
spurious, this one.

(BTW, there is no comfy chair)

Oh, I just thought of:
4. Also, for me, I think that the way to go from the original tuple-form to
(perhaps) a final form as Python programs would be via something I can think
in more easily than the original tuples - so perhaps in the end this "little
language" will be no more than a half-way house.

Having said all of the above, I am definitely not averse to someone else
coming up with something better, or explaining to me why what I am doing is
the hard way to do it and, look, here is a lazier way...

> Whoops - I could make the tagging engine into a pluggable
> interpreter for stackless Python, this is just another engine.


Please feel free - you do that and I will happily use it...

> yes-I-have-to-duck-and-cover-ly y'rs - chris :-)


Oh, I don't know, it seemed sensible to me (hint: not a recommendation!)

--
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home





From herzog at online.de  Sun Jul 11 19:39:39 1999
From: herzog at online.de (Bernhard Herzog)
Date: 12 Jul 1999 01:39:39 +0200
Subject: Tk-Problem?
References: <lz1zeej0lk.fsf@netcologne.de>
Message-ID: <m3wvw6u56s.fsf@greebo.nodomain.de>

Hans Gubitz <gubitz at netcologne.de> writes:

> I need some help. What am I doing wrong?
> 
> gubitz at pumuckl:~/python/apk > python
> Python 1.5.1 (#1, Apr  5 1999, 00:08:45)  [GCC egcs-2.91.66 19990314 (e on linux2

You seem to be running the Python from SuSE Linux 6.1.

> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> float("1")
> 1.0
> >>> from Tkinter import *
> >>> root=Tk()
> >>> button=Button(root, text="Quit", command=root.destroy)
> >>> button.pack()
> >>> root.mainloop()
> >>> float("1")
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> ValueError: float() literal too large: 1

I had the same problem on the same system, although this happened with
pygtk. I finally found that the culprit was the environment variable
LC_CTYPE which has the value de_DE by default on SuSE 6.1, at least on
the German version. Setting it to C should help.

It seems to me that this is really a bug in the C-library.


-- 
Bernhard Herzog	  | Sketch, a python based drawing program
herzog at online.de  | http://www.online.de/home/sketch/



From gmcm at hypernet.com  Thu Jul 22 23:22:54 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 22 Jul 1999 22:22:54 -0500
Subject: Maintainability (was Re: Stackless & String-processing)
In-Reply-To: <14231.42621.76462.103289@96.chicago-25-30rs.il.dial-access.att.net>
References: <7n82u1$jid$1@nntp.Stanford.EDU>
Message-ID: <1279426126-23428601@hypernet.com>

Skip Montanaro wrote:
 
> Two points.  One, the string module wasn't designed to be imported
> using "from string import *".  In my previous message I mentioned
> the types and Tkinter modules.  I'm not aware of any others that
> were explicitly designed for that.  

You can add threading to your list.

someones-got-a-fondness-for-'t'-ly y'rs

- Gordon



From jon at rdt.monash.edu.au  Wed Jul 28 16:45:17 1999
From: jon at rdt.monash.edu.au (Jonathan Giddy)
Date: 28 Jul 1999 20:45:17 GMT
Subject: HPUX and threads?
References: <al8yag0x66w.fsf@sirppi.helsinki.fi>
Message-ID: <7nnq4t$viq$1@towncrier.cc.monash.edu.au>

Markus Stenberg <mstenber at cc.Helsinki.FI> writes:

>Is HPUX-11 thread support tested/implemented/whatnot? 
Implemented - yes, tested - not in your environment (until now).

>This at least does not look promising..

># /opt/python/bin/python
>pthread_mutex_init: Invalid argument
>Memory fault(coredump)

OK, this is probably due to an incorrect guess about which version of 
pthreads is being used.

If you look at the beginning of the source file Python/thread_pthread.h
you'll see a series of preprocessor tests, defining PY_PTHREAD_D4 and 
friends.

It seems HPUX selects either Draft 4 or the final standard, based on whether
_DECTHREADS_ is defined.  Try placing the following after the #endif

#if defined(PY_PTHREAD_D4)
#error PY_PTHREAD_D4
#elif defined(PY_PTHREAD_D6)
#error PY_PTHREAD_D6
#elif defined(PY_PTHREAD_D7)
#error PY_PTHREAD_D7
#else
#error PY_PTHREAD_STD
#endif

make clean and recompile.  This will (hopefully) give you an error like:
In file included from thread.c:146:
thread_pthread.h:88: #error PY_PTHREAD_STD

which tells you which version was assumed.  Then, try undefining the
assumed value, and replace it with the likely correct value.

So, if you get PY_PTHREAD_STD, then replace the code above with:
#undef PY_PTHREAD_STD
#define PY_PTHREAD_D4

then make clean; make

That will hopefully get you running.  A more permanent solution requires
someone who knows the history of pthreads on HPUX.

Jon.



From peterh at sapros.com  Thu Jul  1 12:54:27 1999
From: peterh at sapros.com (Peter Haight)
Date: Thu, 01 Jul 1999 09:54:27 -0700
Subject: Why is no exception raised for file.read() when file is only writeable?
Message-ID: <199907011654.JAA02271@wartch.sapros.com>

I've been helping someone learn the language and he was having trouble
understanding why this didn't work:

test=open('testfile','w')
testread=test.readline()

I explained to him the problem, but I was wondering why Python doesn't raise
an exception when you do this. He probably wouldn't have had to ask me
anything if Python printed something like this:

IOError: [Errno X] You can't read from a file only open for writing.





From Ivan.Van.Laningham at p98.f112.n480.z2.fidonet.org  Fri Jul  2 18:22:10 1999
From: Ivan.Van.Laningham at p98.f112.n480.z2.fidonet.org (Ivan Van Laningham)
Date: Fri, 02 Jul 1999 23:22:10 +0100
Subject: Slicing beyond list --  bug or featurette?
Message-ID: <00001697@bossar.com.pl>

From: Ivan Van Laningham <ivanlan at callware.com>

Hi All--

Pierre Fortin wrote:
> 
> Just ran into a situation which can be summarized as follows:
> 

[bobbitt]

> list[2:50] returns [ 2,3,4 ]
> list[20:] or list[20:30] returns []
> 
> Can I count on this, or should I code try/except for the day it's
> fixed?  This 'feature' reduces the error situations; but is it
> correct/intended?
> 

I don't presume to speak for Guido, but the way slicing is implemented
it seems likely that you can count on it: list[:] is implemented as if
it were list[0:MAXINT].  I doubt that Guido will change it, but then,
the Guido is the Guido.

<but-they-don't-call-him-guido-svetz-for-nothing>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From bhoel at server.python.net  Thu Jul  8 03:25:36 1999
From: bhoel at server.python.net (Berthold =?iso-8859-1?Q?H=F6llmann?=)
Date: Thu, 08 Jul 1999 09:25:36 +0200
Subject: Trouble with HTMLGen, HTMLtest
References: <e3UDcTMy#GA.342@nih2naae.prod2.compuserve.com>
Message-ID: <37845270.BD3E08DD@starship.python.net>

Dave Borgelt wrote:
> 
> I am having a problem with HTMLGen that is driving me nuts.  It has to do
> with running the test code in HTMLtest.  (HTMLtest.test()).
> 
> I expect the solution is simple but I can't see it.  The path to the HTMLGen
> files is in my PythonPath (hence the import works) but as you can see the
> test fails.
> 
> -----------------------------
> 
> >>> import HTMLtest
> >>> HTMLtest.test()
> Traceback (innermost last):
>   File "<interactive input>", line 1, in ?
>   File "c:\programming\python\HTMLGen\HTMLtest.py", line 26, in test
>     'overview.html')      # manual top
>   File "c:\programming\python\HTMLGen\HTMLtest.py", line 35, in overview
>     doc = SeriesDocument('HTMLgen.rc')
>   File "c:\programming\python\HTMLGen\HTMLgen.py", line 256, in __init__
>     if resource: execfile(mpath(resource), self.__dict__)
> IOError: [Errno 2] No such file or directory
> >>>
> 
> ------------------------------
> 
> Any help will be appreciated.
> 
> Dave Borgelt

This only works in the directory where HTMLgen.rc (and probably other
files) lie. This is usually the directory generated when unpacking
HTMLge.

Greetings 

Berthold

-- 
bhoel at starship.python.net / http://starship.python.net/crew/bhoel/
        It is unlawful to use this email address for unsolicited ads
        (USC Title 47 Sec.227). I will assess a US$500 charge for
        reviewing and deleting each unsolicited ad.



From sreekantk at freenet.co.uk  Tue Jul 13 07:08:54 1999
From: sreekantk at freenet.co.uk (Sreekant Kodela)
Date: Tue, 13 Jul 1999 12:08:54 +0100
Subject: Zope newbee help please.
Message-ID: <378B1E46.338A2CFD@freenet.co.uk>

Hi folks

I just installed Zope. Looks good. Is there any good tutorial or guide
for the Zope to create a good site.

Thanks
sreekant





From tbryan at arlut.utexas.edu  Mon Jul  5 11:05:50 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Mon, 05 Jul 1999 10:05:50 -0500
Subject: Problem with informixdb and python1.5.2
References: <7ljsbc$srn$1@news.harvard.net> <377E2D7B.BEBC065C@arlut.utexas.edu> <7llpke$m2r$1@news.fas.harvard.edu>
Message-ID: <3780C9CE.EE7B1E68@arlut.utexas.edu>

Ken Liu wrote:

> > Have you tried recompiling informixdb with the 1.5.2 header files?
> > I don't think that too much has changed between 1.5.1 and 1.5.2, but
> > it couldn't hurt to try.
> 
> That sounds like the thing to do.  In fact, I did change the 
> PYTHONINCL and PYTHONLIB dirs to point to the 1.5.2 directories 
> and recompiled informixdb, but the same problem persists.
> 
> so is the common experience that the modules don't work when you have 
> multiple versions of python installed?  Why is that?

I don't think it's common.  I have one system with two Pythons, and 
all of the modules seem to find the correct Python.  I just know that
I've seen some posts that mention such problems.  (Perhaps only on 
Windows systems?)  

The best venue for this question may be the db-sig mailing 
list.  Try asking your question again there.  

 
-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From holger at phoenix-edv.netzservice.de  Thu Jul 29 10:08:12 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Thu, 29 Jul 1999 14:08:12 GMT
Subject: os.name
Message-ID: <37A0604C.322192FE@phoenix-edv.netzservice.de>

High,

someone working on a more specific version of os.name or better:
something inside of 'sys.builtin_module_names'.

It's because I need to differ between Win95, Win98 and WinNT....

Thanx,

Holger



From Alexander.V..Voinov at p98.f112.n480.z2.fidonet.org  Sat Jul  3 02:12:24 1999
From: Alexander.V..Voinov at p98.f112.n480.z2.fidonet.org (Alexander V. Voinov)
Date: Sat, 03 Jul 1999 07:12:24 +0100
Subject: Curses, Ncurses, Slang, etc?
Message-ID: <000016b4@bossar.com.pl>

From: "Alexander V. Voinov" <avv at quasar.ipa.nw.ru>

Eric Kidd wrote:

> Neither the ncurses module or the pwin module work out of the box on my
> Linux system. Are there any well-supported Python modules for
> curses/ncurses/slang which do work out of the box?

SLang itself is supported, and even a higher-level set of widgets, written
purely in Python, is underway, but slowly :-(.
You may pick up a binary for Red Hat 6.0 at:

ftp://starship.python.net/pub/crew/avv/pyslang/work/slang.so

or recompile it out-of-box from
ftp://starship.python.net/pub/crew/avv/pyslang/pyslang-0.2.0.tgz.

As usual, I remind that Red Hat, as well as other distributions, comes with
the Python wrapper for the 'snack' library, which provides higher-level
widgets than bare pyslang.

The documentation on pyslang is mostly one for the SLang itself, with the
modifications mentioned in my readme's.
Examples may give more information, as usual.

The principal advantage of pyslang I'm aware of is that you may use the same
script under win32 without any changes.

Next bug-fix release of pyslang is very close to completion.

Alexander






From arcege at shore.net  Fri Jul 16 21:19:57 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Sat, 17 Jul 1999 01:19:57 GMT
Subject: propapating exceptions
References: <slrn7ovfss.4bc.quinn@lira.ugcs.caltech.edu>
Message-ID: <1TQj3.294$ui4.84182@news.shore.net>

Quinn Dunkan <quinn at lira.ugcs.caltech.edu> wrote:
: I've lately found myself in the position of wanting to catch exceptions
: based on their attributes, rather than on the exception itself.  So (in
: mal's odbc module):

: try:
:     some sql command
: except ODBC.MySQL.OperationalError, msg:
:     if msg[1] == CR_SERVER_GONE_ERROR: # (snarfed out of errmsg.h)
:         restart server, try again
:     else:
:         raise sys.exc_info()[0], sys.exc_info()[1]

: It's a rather wordy way to pass the exception on without touching it,
: and I still lose the traceback info.  Is there a better way to pass on
: an exception as if it had never been caught?

In Python 1.5, using "raise" by itself will reissue the exception with
the same value.  Fortunately (or unfortunately), the traceback does not
seem to be modified.

  -Arcege




From achatham at hotmail.com  Mon Jul 12 21:56:39 1999
From: achatham at hotmail.com (achatham at hotmail.com)
Date: Tue, 13 Jul 1999 01:56:39 GMT
Subject: Reading from a COM port?
Message-ID: <7me6ci$5dg$1@nnrp1.deja.com>

What would be a good way to access a COM port in Solaris? I'm trying to
read data from a temperature probe and filter it. Can I just access the
device as a file or is there a more elegant solution? Sorry, I'm just a
beginner at Python. Any help would be appreciated.

Thanks,
Andrew


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Gerrit.Holl at p98.f112.n480.z2.fidonet.org  Thu Jul  1 19:14:50 1999
From: Gerrit.Holl at p98.f112.n480.z2.fidonet.org (Gerrit Holl)
Date: Fri, 02 Jul 1999 00:14:50 +0100
Subject: Why does shelve make such large files?
Message-ID: <00001630@bossar.com.pl>

From: Gerrit Holl <gerrit.holl at pobox.com>

Hi,

Is it really necesarry for shelve to make such large files?
Have a look at this:
/tmp> python
Python 1.5.2 (#1, Apr 18 1999, 00:16:12)  [GCC 2.7.2.3] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import shelve
>>> d = shelve.open('database')
>>> d['key'] = 'value'
>>>
/tmp> ls -l database
-rw-rw-r--   1 gerrit   gerrit      16384 Jul  1 21:13 database
                                    ^^^^^

16 KB for only one key!!?
pickle seems to make _much_ smaller files!

Why is this?

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!




From heiland at ncsa.uiuc.edu  Thu Jul  8 11:36:28 1999
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Thu, 08 Jul 1999 10:36:28 -0500
Subject: PyImport_ImportModule, PyModule_GetDict
Message-ID: <3784C57B.59DF0D33@ncsa.uiuc.edu>

If I understand this correctly, the equivalent of doing this:

>>> import mymodule
>>> dir(mymodule)

is sort of this, in the C API:

PyObject *mod = PyImport_ImportModule("mymodule");
 PyObject *myDict = PyModule_GetDict(mod);

Q:  what is the C API equivalent of this:

>>> from mymodule import *
>>> dir()


--Randy





From gregm at iname.com  Tue Jul 27 00:24:14 1999
From: gregm at iname.com (Greg McFarlane)
Date: Tue, 27 Jul 1999 14:24:14 +1000
Subject: Python Performance
In-Reply-To: <7nhtiv$h8f$1@nnrp1.deja.com>; from aaron_watters@my-deja.com on 26 Jul 1999 at 03:07:17PM
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com> <al86737ocwo.fsf@sirppi.helsinki.fi> <7nhtiv$h8f$1@nnrp1.deja.com>
Message-ID: <19990727142414.01582@nms.otc.telstra.com.au>

On 26 Jul, aaron_watters at my-deja.com wrote:
> ... Perhaps it is a perversion of some
> kind but I simply don't use deep hierarchies, &c.  I find them
> confusing too often anyway.

At the same time I was reading this post from Aaron, I was also
browsing his gadfly source.  I noticed the following class heirachy:

    SimpleRecursive
	BTPredicate
	    NontrivialEqPred
		QuantEQ
		    QuantLT
			QuantLE

So I suppose his definition of "deep" is at least 7 :^)=

Plug:  To anyone who needs a good, free, robust, flexible, sql-based,
python-only database, they should check out gadfly <http://www.chordate.com>.

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From aahz at netcom.com  Thu Jul 22 20:08:09 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 23 Jul 1999 00:08:09 GMT
Subject: python + bash + new user
References: <3793BE52.43E023BB@netcard.com> <3793C729.55DE42D0@netcard.com> <7n5nfq$j0$1@vvs.superst.iae.nl> <3797AE3B.687785E5@netcard.com>
Message-ID: <7n8bp9$37@dfw-ixnews12.ix.netcom.com>

In article <3797AE3B.687785E5 at netcard.com>,
Joyce Stack  <joyce at netcard.com> wrote:
>
>I did exactly as you said and chmod's etc but it just says that it
>doesnt recognise "./simple.py" as a command.???..what am i doing work??
>anyother simple programs will work...

That's because you don't have Python installed at /usr/local/bin/python
(or you're on a Windoze box).  Try 'python simple.py' instead.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From olli at rhein-zeitung.de  Fri Jul 23 04:21:03 1999
From: olli at rhein-zeitung.de (olli at rhein-zeitung.de)
Date: 23 Jul 1999 08:21:03 GMT
Subject: int(long(-sys.maxint-1)) fails on Linux
References: <000001bed49f$c052e820$2c2d2399@tim> <3797D8A1.E6D95CCB@vision.arc.nasa.gov>
Message-ID: <7n98lf$ii5$1@news.rhein-zeitung.DE>

> This is on a Pentium III system running straight Mandrake 6.0.  I think, as Christian
> and Oliver pointed out, it is pgcc related.  I tried it on a Mandrake 5.3 system (using
> the normal gcc, not pgcc, and it worked correctly)

Yep, that is exactly what happens when you use pgcc. I moved to pgcc some time
ago and caused a lot of problems. The distribution itself worked fine, but as
soon as someone updated a package and recompiled it with egcs or gcc. That
module was not importable by the distribution I created. I moved back to egcs
without this pentium/pentiumII stuff, and everything worked. And if this is
the case again, I have to look into this more then I like to. ;-))) 

I reported the ASM code to a (P)GCC/EGCS professional or poweruser and he is
looking into the issue right now. I believes that it is either pgcc or pgcc in
combination with glibc2.1 or so. But I hopefully get some info until the
weekend. I like to have this issue fixed before releasing any new distribution
release. ;-))

> It sounds as if Oliver's very-soon-to-be-released versions of the packages may have
> fixed the problem w/ compiler flags.  I'm anxious to try it out (Oliver, thanks for all
> the effort, BTW.  I got impatient waiting for your new releases, and have been making
> my own based on you specs and packages...  LOT of work getting everything just right!
> I figured you were busy, as we all are, but I'm happy to see that you'll be releasing soon.
> I will definately join the python4linux mailing list, as I would like to help out in the
> future)

That is fine and promising. Do you mind to send me your spec file or SRPM? May
be we can integrate the two files into one and get an even better solution for
some tasks. I am good adt rpming but I am by no means godlike in this field.

Bye, Oliver


-- 
Oliver Andrich, RZ-Online, Schlossstrasse Str. 42, D-56068 Koblenz
Telefon: +49-261-39210-40 Fax: -33 |       http://rhein-zeitung.de 
                                   |            http://andrich.net
GMC d- s++:++ a- C++ UL++++ P-- L+++ E--- W++ N+ w-- O- M-- 
V- PS++ PE Y+ PGP+ t 5+ X R+ tv+ b++ DI++ D- G e h- r z+



From hassan-2 at egroups.net  Wed Jul 21 18:37:39 1999
From: hassan-2 at egroups.net (hassan-2 at egroups.net)
Date: Wed, 21 Jul 1999 22:37:39 GMT
Subject: python voice chat
Message-ID: <7n5i3j$190l@eGroups.com>

Go here to talk about python with our new real time Internet voice
feature:

  http://www.egroups.com/VoiceChatPage?listName=python-list

Scott Hassan
eGroups.com





From tre17 at pc142.cosc.canterbury.ac.nz  Tue Jul  6 20:45:23 1999
From: tre17 at pc142.cosc.canterbury.ac.nz (Timothy R Evans)
Date: 07 Jul 1999 12:45:23 +1200
Subject: Getting contents of Text / Entry widgets
References: <7ltmfu$nbt$1@nnrp1.deja.com>
Message-ID: <c75pv252spo.fsf@pc142.cosc.canterbury.ac.nz>

catlee at my-deja.com writes:

> Hi,
> 
> I'm writing a little utility that uses a Tkinter Text widget (and an
> Entry widget).  I want to perform an action for every key that is
> entered.  So I do something like:
> 
> def callback(event):
>    print event.widget.get()
> 
> widget=Entry()
> widget.bind("<Key>",callback)
> 
> The problem with this is that the output generated by callback is
> always one character behind.  I assume this is because when the event
> is generated, the character you've typed hasn't been entered into the
> buffer yet.  I could modify callback to look something like:
> def callback(event):
>    print event.widget.get() + event.char
> 
> But this doesn't handle backspace and delete characters well.
> 
> Is there a nice way of retrieving the contents of a Text / Entry widget
> every time the widget's text buffer is changed (as opposed to whenever
> a key is pressed)?
> 
> Thanks,
> Chris
> 
> 
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.

The class below does this, see the example at the bottom of the script
for how to use it.  This example just prints the contents of the entry
widget.

==============================================================
#!/usr/bin/python

from Tkinter import *

class CallbackEntry(Entry):
    '''Replacement of Entry widget.  Set the "callback" attribute to
    your function, which will be called whenever the entry widget
    changes and will be passed the new contents of the widget'''

    def __init__(self, *args, **kw):
        apply(Entry.__init__, (self,)+args, kw)
        self.stringvar = StringVar(self)
        self.stringvar.trace_variable('w', self.changed)
        self.configure(textvariable=self.stringvar)
        self.callback = None

    def changed(self, *args):
        'Note that the args given to this method are useless'
        if self.callback != None:
            self.callback(self.stringvar.get())

if __name__ == '__main__':
    def test(string):
        print string
    mw = Tk()
    e = CallbackEntry(mw)
    e.pack()
    e.callback = test
    mw.mainloop()

==============================================================
--
Tim Evans



From fredrik at pythonware.com  Tue Jul 13 18:33:41 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 13 Jul 1999 22:33:41 GMT
Subject: XML-RPC
References: <931899884snz@vision25.demon.co.uk>
Message-ID: <003a01becd7f$c3303650$f29b12c2@secret.pythonware.com>

Phil Hunt <philh at vision25.demon.co.uk> wrote:
> Does anyone know if an XML-RPC server exists for Python?
> 
> (XML-RPC is a protocol for making remote procedure calls, where
> the parameters are sent in XML. The XML-RPC spec is at www.xmlrpc.com
> There's a python XML-RPC client at www.pythonware.com/products/xmlrpc/
> but I'm not aware of a Python server).

www.pythonware.com/products/xmlrpc/ says:

    "Sample code for Medusa and Python's SocketServer
    module is included in the current release."

</F>




From larsga at ifi.uio.no  Thu Jul 29 03:09:37 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 29 Jul 1999 09:09:37 +0200
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com> <379f06e9@news.xtml.co.uk>
Message-ID: <wkk8rkx77y.fsf@ifi.uio.no>

* Nick Trout
| 
| Are there any pieces of test code anywhere which could be used as a
| good test for Pythons speed (encompasing all of its language
| features) so that if anyone tinkered with the bytecode execution
| unit they could measure the speed improvement. It would be useful if
| it also tested correctness to ensure no mistakes?!

My xmlproc XML parser might be a good test of string processing at
least, and I have regression and speed testing harnesses for it.
However, it's already been optimized by hand, so it may not be the
best choice.

Anyway, if you want the harnesses, just let me know.

--Lars M.



From tim_one at email.msn.com  Tue Jul 20 02:31:57 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 20 Jul 1999 02:31:57 -0400
Subject: J, you can't pass your own test.
In-Reply-To: <7mum92$15a0$1@news.accu.uu.nl>
Message-ID: <001601bed279$90d4e620$139e2299@tim>

[Martijn Faassen]
> Cool, I didn't know I *had* a political standing.

Of course you do.  Everyone's political standing is continuously updated in
real time -- see the Political Standing Wizard at python.org.

> So what's my political standing?

Sorry, the Wizard is write-only for everyone except Guido.  Not even bots
have been able to crack it!

don't-suspect-a-friend-turn-him-in-ly y'rs  - tim





From kuncej at mail.conservation.state.mo.us  Fri Jul  2 18:08:19 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Fri, 2 Jul 1999 22:08:19 GMT
Subject: XML-RPC and firewall [solution]
Message-ID: <s77cf213.010@mail.conservation.state.mo.us>

I have been playing with XML-RPC on my intranet, but
found that Fredrik's xmlrpclib.py wouldn't talk to external 
servers through our firewall.

Since urllib has proxy-server support built-in, I wrote a 
urllib-based transport plugin for xmlrpclib.py (attached).
Now I can speak XML-RPC with the rest of the world :-)

Testing on this has been minimal - use at your own risk.
If it checks out, maybe /F will include it in the distribution.

  --Jeff



-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: xmlrpc_urllib_transport.py
URL: <http://mail.python.org/pipermail/python-list/attachments/19990702/fdb4a9e6/attachment.ksh>

From mikael at pobox.com  Fri Jul  2 11:28:25 1999
From: mikael at pobox.com (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 15:28:25 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773CA67.7057F797@quasar.ipa.nw.ru> <3773D55B.8C966174@spacenet.tn.cornell.edu> <slrn7naiqb.8ll.zeitlin@seth.lpthe.jussieu.fr> <7li9t6$j2k$1@srv38.cas.org>
Message-ID: <30CF6C97EED47EAD.8A8612E26FB53A0A.B0666516E77A0F15@lp.airnews.net>

On 2 Jul 1999 12:04:54 GMT, lvirden at cas.org wrote:

>I wonder why various industry rags have proclaimed during the past
>10 months about Apple's G3 machines outselling other vendor
>machines.  ...

That makes two of us.


-- Mikael




From tjs at longford.cs.monash.edu.au  Thu Jul  8 00:26:14 1999
From: tjs at longford.cs.monash.edu.au (Toby J Sargeant)
Date: Thu, 8 Jul 1999 14:26:14 +1000
Subject: Coredump: python 1.5.2
Message-ID: <19990708142614.A3595@longford.cs.monash.edu.au>

The following code causes a coredump under python 1.5.2

class Stuffed:
   def __getattr__(self,key):
      print self

print Stuffed()

[toby at ephedra]$ python 
Python 1.5.2 (#3, Jun 17 1999, 22:05:33)  [GCC egcs-2.91.66 19990314/Linux (egcs- on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 

(Part of the) Backtrace:

#0  0x4009fc16 in chunk_alloc (ar_ptr=0xbfe02050, nb=18031028) at malloc.c:2637
#1  0x4009fb8a in __libc_malloc (bytes=20) at malloc.c:2616
#2  0x8073800 in PyTuple_New (size=2) at tupleobject.c:85
#3  0x805de9b in do_mktuple (p_format=0xbfe020c4, p_va=0xbfe020c8, endchar=41, 
    n=2) at modsupport.c:246
#4  0x805df75 in do_mkvalue (p_format=0xbfe020c4, p_va=0xbfe020c8)
    at modsupport.c:275
#5  0x805e15d in Py_VaBuildValue (format=0x808a51e "(OO)", va=0xbfe020e8)
    at modsupport.c:426
#6  0x805e0fb in Py_BuildValue (format=0x808a51e "(OO)") at modsupport.c:399
#7  0x80662d6 in instance_getattr (inst=0x80be9b0, name=0x80a5b78)
    at classobject.c:632
#8  0x8071a31 in PyObject_GetAttrString (v=0x80be9b0, name=0x808baf9 "__str__")
    at object.c:368
#9  0x8071784 in PyObject_Str (v=0x80be9b0) at object.c:255
#10 0x807163f in PyObject_Print (op=0x80be9b0, fp=0x401301a0, flags=1)
    at object.c:186
#11 0x8069d28 in PyFile_WriteObject (v=0x80be9b0, f=0x80a11f8, flags=1)
    at fileobject.c:1044
#12 0x80521a4 in eval_code2 (co=0x80be888, globals=0x80a69c8, locals=0x0, 
    args=0x82771cc, argcount=2, kws=0x0, kwcount=0, defs=0x0, defcount=0, 
    owner=0x0) at ceval.c:1054
#13 0x805421b in call_function (func=0x80be9f0, arg=0x82771c0, kw=0x0)
    at ceval.c:2484
#14 0x8053e3a in PyEval_CallObjectWithKeywords (func=0x80be9f0, arg=0x82771c0, 
    kw=0x0) at ceval.c:2322
#15 0x80662ec in instance_getattr (inst=0x80be9b0, name=0x80a5b78)
    at classobject.c:635
#16 0x8071a31 in PyObject_GetAttrString (v=0x80be9b0, name=0x808baf9 "__str__")
    at object.c:368
#17 0x8071784 in PyObject_Str (v=0x80be9b0) at object.c:255
#18 0x807163f in PyObject_Print (op=0x80be9b0, fp=0x401301a0, flags=1)
    at object.c:186
#19 0x8069d28 in PyFile_WriteObject (v=0x80be9b0, f=0x80a11f8, flags=1)
    at fileobject.c:1044
#20 0x80521a4 in eval_code2 (co=0x80be888, globals=0x80a69c8, locals=0x0, 
    args=0x8277064, argcount=2, kws=0x0, kwcount=0, defs=0x0, defcount=0, 
    owner=0x0) at ceval.c:1054
#21 0x805421b in call_function (func=0x80be9f0, arg=0x8277058, kw=0x0)
    at ceval.c:2484
#22 0x8053e3a in PyEval_CallObjectWithKeywords (func=0x80be9f0, arg=0x8277058, 
    kw=0x0) at ceval.c:2322
#23 0x80662ec in instance_getattr (inst=0x80be9b0, name=0x80a5b78)
    at classobject.c:635
#24 0x8071a31 in PyObject_GetAttrString (v=0x80be9b0, name=0x808baf9 "__str__")
    at object.c:368

Toby.



From Greg.Ewing at p98.f112.n480.z2.fidonet.org  Mon Jul  5 16:53:53 1999
From: Greg.Ewing at p98.f112.n480.z2.fidonet.org (Greg Ewing)
Date: Mon, 05 Jul 1999 21:53:53 +0100
Subject: printing with print
Message-ID: <000017ce@bossar.com.pl>

From: Greg Ewing <greg.ewing at compaq.com>

Angus MacKay wrote:
> 
> I like the printing of any object ability of "print" so I
> do not want to use "" % style printing

The "%s" format will handle any object, too -
not just strings:

>>> x = [1, 2]
>>> y = ("buckle", ("my", "shoe"))
>>> print "%s%s" % (x, y)
[1, 2]('buckle', ('my', 'shoe'))

Look, ma, no space!

Greg



From nick at NOSPAMvideosystem.co.uk  Mon Jul 12 12:12:20 1999
From: nick at NOSPAMvideosystem.co.uk (Nick Trout)
Date: Mon, 12 Jul 1999 17:12:20 +0100
Subject: idle
References: <7m9tng$6qg$58@rks1.urz.tu-dresden.de>
Message-ID: <378a1221@news.xtml.co.uk>

I can't get breakpoints to work in the version I have, 0.4. I'm running it
under Windows.

Is this just me?!

Tim Hamza <thamza.nospam at geocities.com> wrote in message
news:7m9tng$6qg$58 at rks1.urz.tu-dresden.de...
>Have there been any updates to idle.  I thought I saw something about a 0.5
>version, but I haven't been able to find anything at python.org.
>
>





From fredrik at pythonware.com  Wed Jul 21 12:32:35 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Jul 1999 16:32:35 GMT
Subject: String2List: odd function
References: <199907211304.IAA14937@cro.ots.ac.cr>
Message-ID: <002a01bed396$a4de2690$f29b12c2@secret.pythonware.com>

Isidor <rodisi01 at my-deja.com> wrote:
> ALTERNATIVELY:
> 
> If someone will point me to (or give me) some suggestions re.
> posting code to the newsgroup, I could just do that. I'm worried
> about line breaks and all that stuff. Would an ascii attachment
> work in deja-world? (I've never had much access to newsgroups, so
> this is all sort of new to me, and I have to use the free and
> powerful (many many thanks!) but somewhat frustrating deja-news
> interface.)

afaik, all major newsgroup/mailing list archives can deal
with text attachments.  smaller code snippets are ok, in
other words.

...

another option is to upload stuff to ftp.python.org;
see ftp://ftp.python.org/incoming-instructions

    The 'incoming' directory is available for upload submissions.  Stuff
    you leave there is not externally accessible until someone at
    python.org moves it the publically-readable parts of the hierarchy.

    To have that happen, put your stuff in /incoming under a distinctive
    name.  Include a brief "<stuff>.README" file, identifying the stuff
    *and including a contact address for it*.  Then send email to:

            ftpmaster at python.org

    notifying us of the deposit.  Tell us who you are, what the stuff is for,
    and the names of the pieces you put here.  (When contributing patches to the
    python distribution, see also 'incoming-patches-instructions' in the
    top-level ftp dir.)  Make sure that your return address is valid, so we can
    contact you if necessary!

    Questions or suggestions?  Email 'ftpmaster at python.org'...
                                            |
    (In case of emergency, pull handle:     |  )
                                            |
                                         =======

...

and yes, if you join the PSA, you can get your own webspace
on the python starship.  see:

http://starship.python.net/ (the starship)
http://www.python.org/psa/ (psa membership info)

</F>




From greg.ewing at compaq.com  Thu Jul 22 18:58:10 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 23 Jul 1999 10:58:10 +1200
Subject: Stackless & String-processing
References: <001101bed3f2$5e8874a0$642d2399@tim>
Message-ID: <3797A202.FA5B6A06@compaq.com>

Tim Peters wrote:
> 
> In a way deeper than it is in a functional program not using monads?  IOW,
> is the clarity of "control flow" due to monads specifically or just to
> functional semantics in general?

Monads are really only a big deal in the context of
functional languages, where there is no such thing
as control flow. Monads provide an astoundingly
elegant way of writing what is effectively procedural
code in a functional language. If you've ever used
one of the pre-monad systems for doing I/O in a
functional language, you'll appreciate what a
massive achievement this is.

However, Python is already a procedural language,
so there's no need to use monads there to fake
procedural code. It would be as pointless as
putting training wheels on a tricycle.

With regard to parser combinators, there are various
things at work in the functional language context that
don't apply to Python.

Using monads and parser combinators in a functional
language, you can write parsers that look very much
like a BNF grammar or whatever your favourite notation
is -- *without* needing any special tool to compile
and execute them.

This is largely due to the general sparseness of the
syntax, and the ability to define new infix operators,
neither of which apply to Python. You can't translate
the code into Python without losing most of the notational
convenience which accounts for a large part of the
technique's attractiveness.

The idea of composing parsing functions can still
be useful, however. For instance you might have a
function with the signature

  def p_separated_list(p_item, p_separator):

where p_item and p_separator are parsers for
list items and separator tokens. Which incidentally
could be an answer to

> What I haven't seen is an example of something clearly useful
> (wrt parsing specifically) *beyond* those.

Note that the implementation of p_separated_list
need not be quite the standard version you might
find in a library. In a parser for Python code, for
example, it would take care of that quirky comma-at-
the-end-of-a-sequence thing.

> I will give parser combinators a try; they do sound interesting to me.  More
> interesting would be to hear about that "new programming language"!

If you're interested in examples of monad use for
parsing and compiling, I had a fit of insanity last
year in which I implemented a compiler in Hugs using
monads for several different things, including parsing.
If you'd like to see it, I could dig it out and put it
on a web page somewhere.

Greg



From sreekantk at freenet.co.uk  Sun Jul 25 17:19:22 1999
From: sreekantk at freenet.co.uk (sreekantk at freenet.co.uk)
Date: Sun, 25 Jul 1999 22:19:22 +0100
Subject: How to confirm if mod_pyapache is working.
Message-ID: <379B7F5A.B0843658@freenet.co.uk>

Hi all

I for the first time compiled, installed etc of Apache 1.3.6 and
PyApache module. Then configured the AddHandler etc. It is running now.
However when I move then Python executable in to a hiding place all the
cgi files fail . I thought the python has been embedded in to
mod_pyapache and it only needs library files but not python it self.
Well I may be wrong.

Anyway is there a way I can confirm that the .py cgi files are going
through the pyapache module rather than
regular python interpreter.

Any advice greatly appreciated.

Thanks
sreekant





From alex at somewhere.round.here  Wed Jul 21 16:08:40 1999
From: alex at somewhere.round.here (Alex)
Date: 21 Jul 1999 16:08:40 -0400
Subject: pow()
References: <37961668.FDF78B51@callware.com>
Message-ID: <etdk8rthik7.fsf@m2-225-11.mit.edu>

>>> pow (20,5,3)
2
>>> pow(20,5)%3
2
>>> pow(20,5)-(pow(20,5)/3)*3
2

It returns the remainder you get when you divide by the third argument
the power specified by the first two arguments.  You can compute the
power faster if you know that you're only going to need that remainder
in the end.

Alex.



From mgushee at havenrock.com  Tue Jul 20 23:52:05 1999
From: mgushee at havenrock.com (Matt Gushee)
Date: 21 Jul 1999 12:52:05 +0900
Subject: Need feedback on Tkinter project
Message-ID: <m3yaga4q3e.fsf@havenrock.com>

Greetings, Pythoneers--

I've created a Tkinter solitaire game which more-or-less works and can
be enjoyed as is, if you're into that sort of thing.

But what I really want to do is use it as the basis of a Tkinter
tutorial. I don't intend it to be all-inclusive or to replace Frederik 
Lundh's excellent tutorial, but it will cover a couple of topics that
(last I knew) aren't covered in any depth by Lundh, namely the Canvas
and Menu widgets.

So at this point I'd like a few hardy souls to check it out and tell
me what you think. If I've written any horrible or bizarre code,
certainly tell me about that, but I'm especially interested in your
comments on my comments -- and docstrings -- and places that don't
have them but should. That will help me understand what to write when
I begin the actual tutorial.

You may download the package (.tar.gz or .zip) from 

	http://www.havenrock.com/softlab/pegboard/index.html

I look forward to hearing from you!

Matt Gushee
Portland, Maine, USA
mgushee at havenrock.com



From jbauer at rubic.com  Thu Jul  8 19:42:04 1999
From: jbauer at rubic.com (Jeff Bauer)
Date: Thu, 8 Jul 1999 23:42:04 GMT
Subject: Python for the PalmPilot and Other Devices
References: <37852547.E5AC00F4@ssec.wisc.edu>
Message-ID: <3785374C.A5B7D34B@rubic.com>

Nick,

A Python port for the CE devices is available, and it's
not (hardly) scaled down at all.  Further, there's active
development going on -- including Mark Hammond's participation.

well-you-did-say-"and Other Devices"-ly yr's

-Jeff



From skodela at photon.sindh.com  Wed Jul 28 09:56:31 1999
From: skodela at photon.sindh.com (skodela at photon.sindh.com)
Date: 28 Jul 1999 14:56:31 +0100
Subject: Am stupid and now going crazy...
References: <Pine.LNX.4.10.9907272336360.4460-100000@vagabond.longval.ca>
Message-ID: <m3oggwyj1s.fsf@photon.sindh.com>

If you can't make it work , give me a shout and I will try and find the
modules I compiled earlier in my garbage directory.[ not an easy job - you
can lose the whole os there.]

Regards
sreekant
sreekantk at freenet.co.uk



From tismer at appliedbiometrics.com  Tue Jul  6 16:32:38 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 6 Jul 1999 20:32:38 GMT
Subject: Python suitable for a game engine?
References: <0000163d@bossar.com.pl> <378238af@news.xtml.co.uk>
Message-ID: <378267E6.9F757E21@appliedbiometrics.com>


Nick Trout wrote:

[I wrote, thinking of coroutines but not talking of it loud]

> >I had a quick look, but not the time to check it completely.
> >Does it have coroutines or small threads to do very
> >quick context switches? This is something which
> >Python will be able to do very fast, soon. You can have
> >thousands of current contexts and switch them in no time.
> >This can simplify the implementation of games in a scale,
> >that the speed penalty might be compensated by that.
> >Well, we have to try...
> 
> You don't need a proper threaded system to do game AI. Every object just get
> s X ticks each frame. This is maybe getting a bit off topic. A context switc
> h can be as quick as changing a global pointer to the new current context. I
> think Unreal works as you explain, running lots of objects in parallel. When
> you create a new "live" object it spawns a new "thread". Essentially this is
> just a list of objects that have a member function tick() which gets called
> every frame. Please correct me if I'm wrong.

Yes. I thought of coroutines without spelling them. These are no
threads. In a simple world, you coud have one main coroutine
and thousands of "thing" coroutines which would switch forth and back.
This is extremely fast and convenient, since all the routines are
already running and don't need to do any setup each time they
resume. They also share a global namespace, so they can communicate.
This was kind cooperative multitasking.

Alternatively, preemptive multitasking can also be done, with the
same low cost of switching time, since we always stay in the language
and never need to take care of system stacks, real threads
and so on.

I'm about to publish the building blocks for all flavors of
these, and one can build the most suitable system in Python
itself.

Why I said this can really a saving: In time critical situations
like games and simulations, the Python fuction call overhead,
setup up temp variables, even loading of often used functions
into local variables, costs time.
But if you have simple handlers which are initialized once
and then only transfer control to colleagues when it's time,
you save almost all of it.

> The way I'd look at is: which languages are suitable ie. can do what you wan
> t, can be embedded conveniently. And for games, which is fastest and most co
> mpact?!! Lua and Python can do what you want and can be embedded but Pythons
> byte code execution is slower than Lua's by about 50% or more and its a lot
> smaller. Python is streets ahead for writing tools/parsers/simple utils with
> a GUI. Although you can do all that in Lua as well!

I agree that Python is not too small, although it isn't relly big.
But for games, you could go without many of its types, so much
of Python's code is just hanging around. But development
goes towards much more modularity and pluggability, so I guess
it will not take so much time until you can select if you want
Python with head or without tail.

> I'm interested in your ideas. I have thought about doing similar things. I c
> ame to the conclusion that Lua was more suitable for embedding in a game. Pe
> rhaps you could elaborate on your ideas.

I cannot compare what I don't know. I'm just so pleased
with Python that I put all my energy into it, to help
it suit all the areas where it is still suboptimum.
You may expect something to happen with the speed issue
by end of this year.

later - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From cbarry at 2xtreme.net  Sat Jul  3 04:24:16 1999
From: cbarry at 2xtreme.net (Christopher R. Barry)
Date: Sat, 03 Jul 1999 08:24:16 GMT
Subject: Language extensibility
References: <1283099377-77267484@hypernet.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> <wkzp1g2t2e.fsf@ifi.uio.no> <ey3r9mskvja.fsf@lostwithiel.tfeb.org> <x27lojoums.fsf@todday.aiai.ed.ac.uk> <gZ6f3.650$6M6.197790@news.shore.net> <ey3oghulrui.fsf@lostwithiel.tfeb.org>
Message-ID: <871zeqi161.fsf@2xtreme.net>

Tim Bradshaw <tfb at tfeb.org> writes:

> * Michael P Reilly wrote:
> 
> > Supposedly, Lisp was created in 1953, a year after Fortran, and is said
> > to be the second higher-level (non-assembly) language written.
> 
> No, I think it's 1958 -- the 40th anniversary conference was last
> year, after all!
> 
> (A good year for programming languages and guitars both).

Hmm... L-4 CES, Les Paul... it seems every classic Gibson was
made in 1958....

Christopher



From jhspies at futurenet.co.za  Tue Jul  6 15:08:44 1999
From: jhspies at futurenet.co.za (Johann Spies)
Date: Tue, 6 Jul 1999 21:08:44 +0200 (SAST)
Subject: AttributeError - a dumb question
In-Reply-To: <1280843419-65357321@hypernet.com>
Message-ID: <Pine.LNX.3.96.990706210804.1040A-100000@Johann>

On Tue, 6 Jul 1999, Gordon McMillan wrote:


> Look at the case of the "L" in Krylys and KryLys.
> 
> Now kick yourself, hard!

Indeed!

Sorry for the waste of bandwidth.  Today is not my day :(

Johann

 --------------------------------------------------------------------------
| Johann Spies                                 Windsorlaan 19              |
| jhspies at futurenet.co.za                3201 Pietermaritzburg             |
| Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
 --------------------------------------------------------------------------

     "Ask, and it shall be given you; seek, and ye shall
      find; knock, and it shall be opened unto you."        
                                  Matthew 7:7 




From aahz at netcom.com  Sat Jul 17 14:55:05 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 17 Jul 1999 18:55:05 GMT
Subject: I envy newbies
References: <000101bed082$747b0de0$27a22299@tim>
Message-ID: <7mqji9$q1c@dfw-ixnews14.ix.netcom.com>

In article <000101bed082$747b0de0$27a22299 at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>
>I think the best advice you can give a newbie regardless is to force the
>install to put Python in a directory *they* choose, and one without spaces
>in its name.  Like C:\Python.  This is easier than telling them how to find
>python.exe in the (ill-defined) "default" location, and will simplify later
>tasks whether or not they muck with autoexec.bat.

The one thing I *hate* about this advice is that it's a Bad Idea to put
too many files/directories into the root.  (I *think* FAT-32 gets rid of
the 64-entry limitation, but I'm not sure, and anyone on Win95 or NT has
a decent chance of being on FAT-16.  But even without the limitation, I
still think it's a Bad Idea.)
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From samschul at pacbell.net  Sat Jul  3 16:32:56 1999
From: samschul at pacbell.net (Sam Schulenburg)
Date: Sat, 3 Jul 1999 13:32:56 -0700
Subject: "Is theit an interest in a  SCSI Hard drive test environment under Python?"
Message-ID: <wruf3.171$vI3.34037@typhoon-sf.snfc21.pbi.net>

I have been developing a SCSI diagnostic test environment under Python. It
has the following features:

1) Works under any Windows 32 os that supports ASPI
A(dvanced)(S)ystem(P)rogramming(I)nterface:
         a) Win95/98/NT

2) Provides for rapid development of drive test scripts using Python

3) Supports low level drive test requirements:
         a) Buffer management
         b)Separate read/write buffers for each TargetID
         c)Precision timing loop on Pentium systems using Windows High
performance counters(10 micro-second resolution)
         d)Can issue any SCSI command with io6(),io10(),and io12() functions
         e)Support for up to 15 drives

4)Full integrated help system
         a) Access to python HTML documentation using IDLE 0.5
         b) Access to ASPIpython help for both the ASPIPy.dll and ASPI.py
functions using autoduck help authoring system
         c) All functions documented using Python documentation strings
         d) Ability to add user defined documentation to help systems

5)Works under the following environments:
          a)  Win32 console mode (For high speed and full buffer editing
capabilities)
          b) Python IDLE environment (Recommended environment to use)

If their is sufficient interest in this environment I would like to have
someone volunteer a server where I could locate the installation and source
code for this test platform. I plan on having a releasable beta within two
days, and will be providing updates as they occur.

Sam Schulenburg
samschul at pacbell.net






From kperacles at geocities.com  Thu Jul 22 15:40:52 1999
From: kperacles at geocities.com (Karl Putland)
Date: Thu, 22 Jul 1999 13:40:52 -0600
Subject: Python modules and COM for GoldMine
Message-ID: <9yKl3.356$wm4.44170@news.uswest.net>

Hello All,

If your company uses GoldMine(www.goldminesw.com) I want to know.  I have been developing some modules for accessing GoldMine from Python and PyCOM.  I would like some help with testing these objects.  <fluff>GoldMine is the #1 rated Contact Manager and SFA tool. You can learn more at www.goldminesw.com.</fluff>  If you are interested in lending a hand please e-mail me at kperacles at geocities.com

Thanks,
Karl Putland
kperacles at geocities.com





From 71351.356 at compuserve.com  Thu Jul 29 22:22:57 1999
From: 71351.356 at compuserve.com (T. C. Mits)
Date: Thu, 29 Jul 1999 22:22:57 -0400
Subject: Python and SlickEdit
References: <379CA3F9.A746217E@iRenaissance.com> <7njrob$oq6$1@mailusr.wdf.sap-ag.de>
Message-ID: <7nr2h6$llq$1@ssauraab-i-1.production.compuserve.com>

Howdy:

    I'm using that aforementioned macro extension with Vslick.   Works well,
except for comments.  Can't get that to work and trying to puzzle out the
internals of the editor takes too much time.

    Since Python doesn't use brackets (which I'm starting to like ) I wrote
a macro to do a jump to end of structure, which is the next dedented line.

Its ugly, there's probably a neater way to do it, but it works so far.  I
need to add a simple way to goto the prev dedent; currently I have a similar
macro for this.  I assigned them to tool bar buttons:

/**
  *
  * Go to the next line where cur column is non-space.
  *
  * This is especially useful for the Python language which uses indentation
  * as a structuring device (which has a certain advantage, "a wysiwyg
style")
  *
  * @author JB
  * @date   7/28/99
  * @version 0.01
  *
  */
_command next_dedent()
{  
   cursor_down();
   ch = get_text(1);

   while ( !pos( '['p_word_chars'\!\@\#\$\%\^\&\*\(\)\_\+\-\=\{\}\[\]\|\:\;\"\<\>\?\,\.\/\~\`]',
      ch,1,'r') || !_dbcsStartOfDBCSCol(p_col) ) {

      if( cursor_down()){
         break;
      }
      ch = get_text(1);

   } // while

}  // next_dedent().


Daniel Dittmar <daniel.dittmar at sap-ag.de> wrote in message news:7njrob$oq6$1 at mailusr.wdf.sap-ag.de...
> Rowland Smith <rowland at iRenaissance.com> wrote:
> 
> >Does anyone know if there are any Python plug-ins for Visual SlickEdit
> >and where they are?
> 
> There's a Python module in the contributed area
> (ftp://ftp.slickedit.com/pub/usermacs/python.zip). 
> 
> It does code coloring, but no procedure outlining (at least not
> working). 
> 
> Daniel Dittmar
> 
> 






From akuchlin at mems-exchange.org  Fri Jul 16 10:25:52 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: Fri, 16 Jul 1999 10:25:52 -0400 (EDT)
Subject: Pythoneers at Ottawa Linux conference?
Message-ID: <199907161425.KAA00162@amarok.cnri.reston.va.us>

I'm going to be at the Ottawa Linux Symposium next week.  There's a
space on the schedule open for BOFs, so if any other Pythoneers will
be there and are interested in a Python BOF (or going out to dinner,
whatever), please track me down at the conference and let me know;
we'll see what can be arranged.

      The conference's Web site is http://www.ottawalinuxsymposium.org/ .

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
France has culture but no civilization. England has civilization but no
culture. The United States has neither. Canada has both.
    -- Robin Mathews






From perry at ac-copy.net  Sun Jul  4 14:31:39 1999
From: perry at ac-copy.net (Joachim Schmitz)
Date: Sun, 4 Jul 1999 18:31:39 GMT
Subject: mxODBC and MySQL access to remote DB
Message-ID: <Pine.LNX.4.05.9907042015320.15753-100000@ns.ac-copy.net>

Hello,

I?m trying to connect from a linux-box with mxODBC to a MySQL db on a
remote SUN. I can access the db with mysql-client from linux. I can also
connect with the odbctool from the iodbc-mysql package.

When I try it with mxODBC.iODBC, which uses the same "iodbc.ini" I always
get this:

line: 38
-- in file: Form.py; function: __init__; line: 40
arg #0 = DSN=immobilien;UID=perry;PWD=pwpwpwpw
Traceback (innermost last):
  File "immoform.py", line 153, in ?
    ,DBtype="MySQL")
  File "Form.py", line 45, in __init__
    ,"pw="+self.pw)^M
mxODBC.OperationalError: ('S1000', 1045, "[TCX][MyODBC]Access denied for
user: 'user=perry at localhost' (Using password: YES)", 4237) 

The iodbc.ini entry looks like this:

[immobilien]
Trace    = Off
TraceFile= stderr
Driver   = /usr/lib/mysql/libmyodbc_mysql.so
DSN      = immobilien
SERVER   = test.ac-copy.net
UID      =
PWD      =
PORT     = 3306
#OPTION  = 1
#DB      = test
#SOCKET  = /tmp/mysql.sock
 
with:

odbctool -u perry -p pwpwpwpw -d immobilien "show tables"
<table border=1>
<th>Tables in immobilien</th>
<tr>
<td>IMMOAUFTRAG</td>
</tr>
<tr>
<td>IMMOKUNDE</td>
</tr>
</table>                           

Any help is welcome.



Gru?

Joachim Schmitz

WWW-Consultant

email: js at ac-copy.net
tel: +49-241-89491-0
fax: +49-241-89491-29





From herzog at online.de  Wed Jul 21 16:34:25 1999
From: herzog at online.de (Bernhard Herzog)
Date: 21 Jul 1999 22:34:25 +0200
Subject: Embedding: Creating / Deleting CObjects?!
References: <7n2p2b$bmt$1@news.rz.uni-karlsruhe.de>
Message-ID: <m3ogh5rbce.fsf@greebo.nodomain.de>

Bjoern Giesler <un4e at rz.uni-karlsruhe.de> writes:

> I'm truly baffled. I'm trying to define a backend for my embedded Python
> system, and the corresponding C routines seem to work, but the program
> crashes somewhere in the Python library routines, dances the Fandango all
> over my internal object representation, and whatnot. 
> 
> I'm posting a couple of fragments of my code, just so that somebody can tell
> me whether this is the canonical way of handling things (this side of Python
> is not exactly well documented...) 
> 
> /* Content destructor function */
> static void
> PLPythonDestructor(proplist_t obj)
> {
>   PLRelease(obj);
> }
> 
> /* CObject Creator */
> static PyObject *
> PLPythonNewInstance(PyObject *self, PyObject *args)
> {
>   proplist_t plClassName, plArgs;
>   instance_t instance;
>   char *className;
>   PyObject *pyArgs, *retval;
> 
>   printf("NewInstanceVar\n");
> 
>   if(!PyArg_ParseTuple(args, "sO", &className, &pyArgs))
>     return NULL;
> 
>   plArgs = PLPy2PLObject(pyArgs); /* NULL is OK here */
>   plClassName = PLMakeString(className);
>   instance = PLObjectNewInstance(plClassName, plArgs); /* [My own braindead
> 							  class system.
> 							  Works, though.] */
>   if(plClassName)
>     PLRelease(plClassName);
>   if(plArgs)
>     PLRelease(plArgs);
> 
>   if(!instance)
>     {
>       PyErr_SetString(PyExc_RuntimeError, "Couldn't create new instance");
>       return NULL;
>     }
>   retval = PyCObject_FromVoidPtr((void *)instance,
> 				 PLPythonDestructor);
>   Py_INCREF(retval);

This INCREF creates a memory leak. PyCObject_FromVoidPtr returns a new
CObject which you can return as is.

>   return retval;
> }
> 
> /* CObject Destructor */
> static PyObject *
> PLPythonDeleteInstance(PyObject *self, PyObject *args)
> {
>   instance_t instance;
>   PyObject *instanceObject;
> 
>   printf("DeleteInstance\n");
> 
>   if(!PyArg_ParseTuple(args, "O", &instanceObject))
>     return NULL;
> 
>   /*
>   instance = (instance_t) PyCObject_AsVoidPtr(instanceObject);
>   PLObjectDeleteInstance(instance);
>   */
> 
>   Py_XDECREF(instanceObject);

This DECREF is wrong here, because PyArg_ParseTuple does not increase
the refcount of *instanceObject.

>   return Py_None;
> }


-- 
Bernhard Herzog	  | Sketch, a python based drawing program
herzog at online.de  | http://www.online.de/home/sketch/



From paul at prescod.net  Tue Jul 13 18:29:43 1999
From: paul at prescod.net (Paul Prescod)
Date: Tue, 13 Jul 1999 22:29:43 GMT
Subject: ANN: Stackless Python 0.3
References: <378A40AD.B4FDB79F@appliedbiometrics.com>
		<m37lo5kpd0.fsf@atrus.jesus.cam.ac.uk>
		<378AFD45.8793D507@appliedbiometrics.com> <TaIi3.1701$6M6.569174@news.shore.net>
Message-ID: <378BBDD7.D472AD91@prescod.net>

"Michael P. Reilly" wrote:
> 
> And newbies wonder why Python doesn't have assignment expressions. ;)

Cheap shot. Its a problem with C's syntax, not the assignment expression
concept.

-- 
 Paul Prescod  - ISOGEN Consulting Engineer speaking for only himself
 http://itrc.uwaterloo.ca/~papresco

To me programming is more than an important practical art. It is
also a gigantic undertaking in the foundations of knowledge.
	- Grace Hopper



From moun at usenix.org  Thu Jul 22 17:01:18 1999
From: moun at usenix.org (Moun Chau)
Date: Thu, 22 Jul 1999 21:01:18 GMT
Subject: USENIX Tcl/Tk Conference Call for Papers-Paper submissions due: 9/1/99
Message-ID: <FFAIE7.ArA@usenix.org>

The 7th USENIX Tcl/Tk Conference is a forum to:
 * bring together Tcl/Tk researchers and practition
 * publish and present current work involving Tcl/Tk
 * learn about the latest developments in Tcl/Tk
 * plan for future Tcl/Tk related developments

Tcl/2K : The 7th USENIX Tcl/Tk Conference
      February 14-18, 2000
      Austin, Texas, USA
Sponsored by USENIX, The Advanced Computing Systems Association
---------------------------------------------------------------------------- 
Please find the Call for Papers at
http://www.usenix.org/events/tcl2k/cfp
----------------------------------------------------------------------------
Paper submissions due: September 1, 1999
Demonstration, and Panel Proposals due: September 1, 1999
Poster submissions due: December 8, 1999

The conference program will include formal paper and panel presentations,
poster and both reviewed and informal demonstrations, works-in-progress,
Birds-of-a-Feather sessions, and two-days of high-quality tutorials.  All
forms of participation provide an opportunity to report on original Tcl/Tk
research.

Formal papers should address topics of interest to experienced Tcl/Tk
programmers; posters and informal demos may be geared to any level of user
from beginner to expert. Awards will be given for the best
paper and best student paper at the conference.
============================================================================
The USENIX Association's international membership includes engineers,
system administrators, and computer scientists working on the cutting edge
of  systems and software.  Our conferences are recognized for their
technical excellence and pragmatic emphasis.

Summary: 
Followup-To: 
Distribution: world
Organization: USENIX Association
Keywords: 





From perry at ac-copy.net  Tue Jul  6 14:37:57 1999
From: perry at ac-copy.net (Joachim Schmitz)
Date: Tue, 6 Jul 1999 18:37:57 GMT
Subject: mxODBC and MySQL access to remote DB
In-Reply-To: <37821BDE.61CD46D6@lemburg.com>
References: <37821BDE.61CD46D6@lemburg.com>
Message-ID: <Pine.LNX.4.05.9907062021240.29446-100000@ns.ac-copy.net>

On Tue, 6 Jul 1999, M.-A. Lemburg wrote:

> >     ,DBtype="MySQL")
> >   File "Form.py", line 45, in __init__
> >     ,"pw="+self.pw)^M
> > mxODBC.OperationalError: ('S1000', 1045, "[TCX][MyODBC]Access denied for
> > user: 'user=perry at localhost' (Using password: YES)", 4237)
> 
> This a problem with MySQL, not mxODBC. Please check the MySQL
> logs for more detailed information about access failures.
>  

The problem was that mxODBC or better the ODBC-driver myodbc_mysql didn?t
connect to the host "test.ac-copy.net" but to localhost. I finally found
the solution in the connect.c module of the driver: (version
iodbc-2.50.22a/myodbc-2.50.22 ) 

in  SQLConnect around line 183.

#ifndef _UNIX_
  SQLGetPrivateProfileString(fix_str(dsn,szDSN,cbDSN),"server",
                             "localhost", host, sizeof(host),"ODBC.INI");
  SQLGetPrivateProfileString(dsn,"database",
                             dsn, database, sizeof(database),"ODBC.INI");
  SQLGetPrivateProfileString(dsn,"port",
                             "0", port, sizeof(port),"ODBC.INI");
  port_nr=(uint) atoi(port);
  SQLGetPrivateProfileString(dsn,"flag",
                             "0", flag, sizeof(flag),"ODBC.INI");
  flag_nr=(ulong) atol(flag);
  SQLGetPrivateProfileString(dsn,"stmt",
                             "0", init_stmt,sizeof(init_stmt),"ODBC.INI");
#else
  strmov(database,szDSN);
  {
    char *cp;
    strmov(host, (cp = getenv("MYSQL_HOST")) ? cp : "localhost");
  }
  flag_nr=0;
  init_stmt[0]=0;
#endif /* IS UNIX */

After setting the enviromentvariable MYSQL_HOST to test.ac-copy.net it
worked.
I?m not sure if this is the intended behavior, cause it?s nowhere
documented.

> 
> Note that I have a line stating the database in my odbc.ini
> file:
> 
> [mysql]
> DRIVER   = /usr/local/lib/libmyodbc.so
> SERVER   = localhost
> DATABASE = test 
> 
> Also note that iODBC switched from using a file named 'iodbc.ini'
> to 'odbc.ini' recently.
At least the odbctool is working with the iodbc.ini file.



Gru?

Joachim Schmitz

WWW-Consultant

email: js at ac-copy.net
tel: +49-241-89491-0
fax: +49-241-89491-29





From mgushee at ime.net  Wed Jul 21 18:53:10 1999
From: mgushee at ime.net (Matt Gushee)
Date: Thu, 22 Jul 1999 07:53:10 +0900
Subject: Test -- please ignore
Message-ID: <37964F56.9A4B3514@ime.net>

This one is from Netscape.



From sreekantk at freenet.co.uk  Mon Jul 12 00:04:37 1999
From: sreekantk at freenet.co.uk (Sreekant Kodela)
Date: Mon, 12 Jul 1999 05:04:37 +0100
Subject: database
References: <3787c551.1@ganymede.datanet.hu>
Message-ID: <37896955.A625D32@freenet.co.uk>

Hi

Yes to your question re: using databases. You can download postgresql or mysql
etc dbm interfaces. Eg: PyGreSQL et al. Then

import pg
con=pg.connect('skodela')
res=con.query('select * from uselessdb')
con.close()
res=res.getresult()
for n in res:
    print n

just an example. I am using postgreSQL and Python.

Regards
sreekant

dzsi wrote:

> hi,
>
> i am new to python.......
>
> i have a database (comma separated text), and i would like to get some data
> out of it based on some query. let's say there are seven fields, f1...f7,
> and i am searching for entries in f1, and want to display f1..f7  for the
> field that matches the query. how do i go about doing that?
>
> also, can i access db databases, or any other common pc type databases with
> perl scripts only on a unix system? without any server configuration,
> assuming python is installed. so far, the comma separated text file is the
> only way i could figure out.
>
> thanks,
>
> adam.




From mal at lemburg.com  Thu Jul 15 12:24:17 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 15 Jul 1999 16:24:17 GMT
Subject: Meta-language for mxTextTools
References: <08bb01beceb7$334a58c0$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <378E0B31.42A78E9C@lemburg.com>

Tony J. Ibbs (Tibs) wrote:
> 
> I am aiming to use Marc-Andr? Lemburg's mxTextTools package for an
> application in my (hah!) spare time. Being me, the first thing I've set out
> to do is to provide the missing meta-language so that I can avoid typing in
> the tag tables as "bare" tuples.
> 
> What I've done so far is to produce an alpha version of the meta-language
> documentation, which can be found at
> 
>     http://www.tibsnjoan.demon.co.uk/mxtext/metalang.html
> 
> Translations of the examples from the mxTextTools distribution are also
> supplied.
> 
> I still have to actually *implement* the thing (I expect to do this with
> mxTextTools itself, with appropriate hand-bootstrapping), but since I'm
> about to go off on holiday for a couple of weeks, it seems worth
> "publishing" my thoughts so far and getting any comments.
> 
> Note that I am *not* claiming that this is a particularly sophisticated
> animal, nor that it is necessarily the best beast that could be dreamt of -
> just that it is something I can work with and should be easy to implement.

This looks very promising :-)

One itch that I have with the syntax: since you are
already using Pythonish syntax, why not simply write

def keyword:
    ...

instead of

keyword is:
    ...

An alias "back" for -1 would help reading the tables:

Skip back 
vs.
Skip -1

Hmm, I guess Tim will like this ;-)

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   169 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From tony at lsl.co.uk  Fri Jul 16 04:37:17 1999
From: tony at lsl.co.uk (Tony J. Ibbs (Tibs))
Date: Fri, 16 Jul 1999 08:37:17 GMT
Subject: Meta-language for mxTextTools
Message-ID: <010501becf66$6a316b30$f0c809c0@lslp7o.lsl.co.uk>

Christian Tismer wrote:
>Can you express the parser for the little language with the tagging
>engine? That would be a funny bootstrap :-)


My intention is to write the parser to the little language *in* the little
language, and then to hand-translate it to tuples. It may be that I don't
have to implement all of the little language before I can use it to
bootstrap the rest (I seem to remember languages like BCPL worked in this
manner).

> me> 2. I can't help feeling that "jump-to-label" (aka
> me> skip-forwards/backwards-N-statements) is intrinsically necessary in
such an
> me> engine, and I can't see any way of doing that sensible whilst keeping
to
> me> Python syntax.
>
>if target_of_jump_is("thisone"):
>   # do this
>if target_of_jump_is("thatone"):
>   # do that
>
>... and so on.

Hmm. I think that this requires slightly more complex translations to
achieve. For this first project, I want to keep close to the "flavour" of
the original tag tables. This is partly because I have a sneaky feeling that
there are going to be cases where it is easier (for some sense of "easier")
to generate the tuples by hand, and in that case keeping the two
representations close is an advantage (the obvious case would be if I
*can't* represent all possible tables with the little language). Proof will
be in the pudding...

>This is the syntax which I use for my computed goto example with
>continuations. Man, this is all so similar. Where does the yourney go.


I don't pretend to understand where stackless Python might lead us, except
to say that I am boggled by the implications - it's certainly always
reassuring when apparently unrelated things turn out to be able to use each
other and make life simpler in unexpected ways. But I still want something
that works in (conventional) CPython first...

>Is it allowed to go until 4 in this joke? I don't remember :-)


In Python, I don't know. In the Goons or Firesign Theatre, it definitely
would be...

>Isn't a tagging engine a number of comparisons, head
>movements, and gotos? Continuations could do that.


I believe you. Honestly, I do, I do. Really. You can explain in detail
later, when I've written some code for this vapourware. But I suspect that I
would categorise understanding that as "needs actual thought", whereas what
I'm trying to do definitely doesn't (it's actually just an excuse to
understand how mxTextTools works, believe it or not!).

Tibs
--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.demon.co.uk/
Feet first with 5 wheels...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)





From markusk at bidra241.bbn.hp.com  Tue Jul 20 05:58:12 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 20 Jul 1999 11:58:12 +0200
Subject: Python and Boehm-Demers GC, I have code.
References: <000901becffa$b0f25e00$91a22299@tim>
Message-ID: <p5k8rvmymj.fsf@bidra241.bbn.hp.com>

"Tim Peters" <tim_one at email.msn.com> writes:

[deletia]
> 
> My guess is that simple refcounting has excellent cache behavior in Python,
> and GC doesn't.  Unlike as in almost any other language, the little loop
> 
>     for i in xrange(10000000):
>         pass
> 
> creates a huge amount of heap trash at a furious pace (even ints "are boxed"
> in Python).  RC can reuse the same little bit of storage over & over (the
> trash is recycled as soon as it's created), while delayed GC chews up some
> large amount of address space over & over instead.  You can test this by
> timing that loop alone (with and without refcounting).

True. 
IMHO this is one of Pythons major design flaws. There is no need to create a new
object for each step of the loop. 
Smalltalk has loops at least as flexible as Python and it doesn't have this problem. 
Loops in squeak (www.squeak.org) are already 2 to 3 times faster than in Python. 

> 
> Note too that Python does no stack allocation of locals or frames or
> anything else:  everything is heap-based, with the most speed-critical
> allocations augmented by assorted internal fast free lists.  Disable the
> refcounting, and the finalizers on the "intense" objects don't get called,
> and then the fast free lists don't get used either.  So the allocation speed
> optimizations Python implements get castrated too.
> 
> I've always said it would take a very sophisticated mark-&-sweep to beat
> RC's performance for *Python* (it's not Lisp, or Smalltalk, or C++, or ...
> under the covers); 

Is there anything that would prevent someone to reimplement loops such that only one
 variable is allocated ?


> if true (& I haven't seen anything yet that changes my
> mind <wink>), you're at the first step of very long road if you want to make
> BDW the primary approach; but it would be great if BDW could be invoked
> "just sometimes" to reclaim the cycles RC can't catch.
> 
> semi-encouraging-ly y'rs  - tim
> 
> 
> 
> 


Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From paul at digicool.com  Sat Jul 10 10:14:21 1999
From: paul at digicool.com (Paul Everitt)
Date: Sat, 10 Jul 1999 10:14:21 -0400
Subject: Python Code Repository?
References: <378645AA.5430A026@msfc.nasa.gov> <19990709163127.A1964@toast.internal> <slrn7od2d6.dl.amk@207-172-57-127.s127.tnt2.ann.va.dialup.rcn.com>
Message-ID: <7m9tnn$6qg$64@rks1.urz.tu-dresden.de>

"A.M. Kuchling" wrote:
> Michel Pelletier (the one at Digital Creations) was working on a
> ZTrove at one point, I believe.  (For information on the whole idea of
> Trove, see http://www.tuxedo.org/~esr/trove/).  I don't know what's
> happened to ZTrove; anyone from DC care to comment?

OK, lonnnggg post ahead.

Zope Product Library

  With the progress made on Zope2 and the Zope Portal Toolkit (PTK),
  we've made enough progress to begin construction of a "Freshmeat"
  package library for Zope software.  (Note: the jargon of "product",
  "package", "release", and "version" is awfully contentious.)

  Since this is effectively the same ground being covered here, I'd
  like to take a moment to describe what we are doing and see if folks
  here on the Python list find it interesting.  I'll also cross-post
  to the Zope list.  The desired outcome is for our experience and
  software to provide a boost to a real effort.  (Historical note: I
  was the shepherd for the original ill-fated "Locator-SIG".)

  Goals

    o Fast location of desired software

    o Keep central admin requirements *low* by distributing control

    o Allow package authors to administer their entries

    o Adhere to appropriate standards and initiatives

  Zope Product Library

    In our system, Authors create Products and provide Releases of
    those Products.  A Release is a version of a Product for a
    platform (if binary) or a source version in some packaging scheme
    (such as RPM).  Thus, a Release is the bits and a Product is the
    container for all information about those bits.

    The things capitalized in the above are "ZClasses".  For you
    non-Zopistas, a ZClass is like a Python class that you build
    through the web and store in the Zope object database.  It is an
    instance that can make instances.  It also provides a simple way
    to store "your kinds of things" in the Zope object database.

    The private, under-construction new zope.org site has a membership
    system where we give people a Zope home folder to stash their
    stuff.  They pick from a list of things they can add, such as
    uploaded images, HTML documents, etc.

    People with higher privilege -- such as an Author -- can also
    choose "Product" from the list of things they can add to their
    folder.  When they do so, the Product gets added to the site
    catalog.

    A Product is like a folder -- you can stick a home page, other
    docs, PDF files, etc. in it.  More importantly, inside a Product
    you can add a Release.  This will provide the kind of thing seen
    in Freshmeat.

    Thanks to acquisition, much of the information about a Release
    (author name and email, description, etc. ) comes from the Product
    and doesn't have to be repeated.

    The ZCatalog is awfully useful for this kind of system.  First, it
    is (almost) always up-to-date.  Whenever you add a Product or
    Release, or edit one, or delete one, the Zope transaction also
    updates the catalog.  Next, ZCatalog's indexing engine does both
    fielded and full-text searches.

    Another nice feature is that objects control how they are indexed,
    as the Catalog is calling methods on the site content to get at
    indexed content.  For instance, if you decided you'd like to untar
    the release and index the docstrings, go right ahead.  As long as
    you can implement a Python method to do it.  If you want the
    full-text to actually come from four properties about a Product,
    piece of cake.  If you want to create a Remote Product listing,
    where the content is retrieved by urllib for indexing, you can
    quickly do that as well.

    Once the content is in the ZCatalog, well, Bob's your uncle.  You
    can create multiple interactive search forms.  You can query the
    catalog to insert a tree on a page for browsing the contents.  You
    can create multiple presentations (by date like Slashdot and
    Freshmeat, by category like ODP, by author, by platform, etc.)

    More interestingly, you can get the Catalog to dish up requests as
    RDF.  Imagine browsing the package library using the tree control
    in Mozilla, which gets fed by RDF.  Or, imagine a command-line
    Python script that acts like rpmfind.

  What About Trove and ZTrove?

    This isn't Trove.  This *could* be Trove, but we are focused on
    getting something in place fast on zope.org.  We are solving our
    own problems first.

    Once we have made all our mistakes and learned from them, it would
    definately be time to talk about Trove and a Python package
    locator.

  ZCatalog

    The main advance that allows a high-end package system is the
    "ZCatalog",
    http://www.zope.org/Download/ZCatalog/ZCatalogTutorial.  This
    tutorial provides a good rundown of all the neat things about the
    catalog.

    Does the Catalog scale?  We think so, but that's a hard question.
    We have a *lot* of painful, painful experience in this area, based
    on a big consulting job we did a few years ago, and those lessons
    are reflected in the software.  Our indexing works very hard at
    keeping memory footprint moderately in check and dishing up
    incredibly fast searches.

    As a datapoint, we took a 15 Mb chunk of the rufus RPM database
    and loaded it into the indexer.  A combination fielded and
    full-text search took 0.004 seconds on a PII-350 (inside the
    indexer, formatting the response and returning it took a lot
    longer).

  CPAN and rufus.w3.org

    It appears to me that CPAN does the following functions:

      o Software repository for Perl code

      o Mirrored worldwide

      o Routes requests to a correct, available archive

    A related system that I know more about is rufus.w3.org's RPM
    database.  It adds a particularly useful feature that I *think*
    CPAN has: a command-line tool to query repositories and fetch
    packages for installation.  I unfortunately don't know much about
    the workflow of how package listings get into the repository, then
    how they get updated.

    I'll observe that the searchability for both of these systems is
    awfully weak.  For us in Zopeland, this is a major opportunity.

  The Mythical Python Locator

    So, what exactly *is* being looked for here?  A simple package
    registry?  Something integrated with dist-utils?  How much should
    this system scale?  What standards should it piggy-back?  Is it
    centralized or decentralized, both in implementation and
    administration?  What jargon do we agree to and what is the
    metadata we will choose?

    And most important, is anybody willing to actually work on it?

    In closing, though I'm skeptical that this will move from "brief
    flurry of activity" into accomplished project, I think this is an
    area where a definate disadvantage can be turned into a clear
    advantage.  The goal shouldn't be just to catch up.  The goal
    should be go farther, to show off and taunt, then to rub the
    others' noses in it. :^)



From Gerrit.Holl at p98.f112.n480.z2.fidonet.org  Fri Jul  2 10:31:52 1999
From: Gerrit.Holl at p98.f112.n480.z2.fidonet.org (Gerrit Holl)
Date: Fri, 02 Jul 1999 15:31:52 +0100
Subject: Why does shelve make such large files?
Message-ID: <00001663@bossar.com.pl>

From: Gerrit Holl <gerrit.holl at pobox.com>

On Thu, Jul 01, 1999 at 10:47:57PM +0000, Ovidiu Predescu wrote:
> Gerrit Holl wrote:
> 
> > Is it really necesarry for shelve to make such large files?
> > Have a look at this:
> > /tmp> python
> > Python 1.5.2 (#1, Apr 18 1999, 00:16:12)  [GCC 2.7.2.3] on linux2
> > Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> > >>> import shelve
> > >>> d = shelve.open('database')
> > >>> d['key'] = 'value'
> > >>>
> > /tmp> ls -l database
> > -rw-rw-r--   1 gerrit   gerrit      16384 Jul  1 21:13 database
> >                                     ^^^^^
> > 
> > 16 KB for only one key!!?
> > pickle seems to make _much_ smaller files!
> > 
> > Why is this?
> 
> The shelve module uses DBM which is like a small database that allows       
<
> you to store objects and search it for a given key. DBM allows you to       
<
> store millions of objects and search for them later without requiring       
<
> you to load all the file in memory first.                                   <
> 

Interesting...

> The pickle module on the other hand is serializing the objects with the
> purpose of deserializing them _all_ from the file later. Pickle does not
> offer you any way to search for data based on a key, you have to do this
> yourself after the objects have been created from the file. This is
> opposed to the way shelve handles this, all the key accesses and
> insertions in a shelve object are actually reads or writes to or from
> the DBM file.
> 
> And to answer your question, DBM is creating these big files because of
> the way it manages the database. The data in the database file could
> have gaps as a result of multiple insertions and deletions. Pickle's
> data in files is a simple representation of the objects that were
> written and there is no way to update the file other than rewriting it
> entirely.
> 

Ah, I understand.
So pickle is useful for very small datases, but when they're really huge, one
should use shelve. Isn't it?

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!




From herzog at online.de  Thu Jul 15 06:19:47 1999
From: herzog at online.de (Bernhard Herzog)
Date: 15 Jul 1999 12:19:47 +0200
Subject: Getters and Setters
References: <7mjrqg$e4o@ds2.acs.ucalgary.ca> <7mjune$e2g@ds2.acs.ucalgary.ca> <378D8E21.957C6AEA@bioreason.com>
Message-ID: <m3btde43lo.fsf@greebo.nodomain.de>

Andrew Dalke <dalke at bioreason.com> writes:

> Neil Schemenauer <nascheme at ucalgary.ca> proposed:
> > # A mixin class to automagicly create getter/setter methods for
> > # class attributes.

> Here's one optimization, don't recreate getter/setter objects.
> How about this:
> 
> class GetterSetter:
>     def __getattr__(self, name):
>         try:
>             if name[:3] == 'get':
>                 x = _Getter(self.__dict__[name[3:]])
>                 self.__dict__[name] = x
>                 return x

It seems to me that this way, the get* method will always return the
same value, the value of the instance variable when the _Getter was
first instantiated, even if the instance variable changed later.

>             elif name[:3] == 'set':
>                 x = _Setter(self, name[3:])
>                 self.__dict__[name] = x
>                 return x

And this one will introduce a circular reference

>         except KeyError:
>             pass
>         raise AttributeError

 
> Since your timing test does so many repeats (esp. when I increase
> n :), that suggests I can take a 1-time cost per attribute, so
> I can use an eval.
> 
> class GetterSetter:
>     def __getattr__(self, name):
>         try:
>             if name[:3] == 'get':
>                 #x = _Getter(self.__dict__[name[3:]])
> 		val = self.__dict__[name[3:]]
> 		x = eval("lambda :x", {"x": val})
>                 self.__dict__[name] = x
>                 return x

This has the same problems as above, i.e. the method always returns the
same value.

>             elif name[:3] == 'set':
>                 x = _Setter(self, name[3:])
>                 self.__dict__[name] = x
>                 return x
>         except KeyError:
>             pass
>         raise AttributeError

> For my final trick, I'll do the same to the Setter:
> import operator
> class GetterSetter:
>     def __getattr__(self, name):
>         try:
>             if name[:3] == 'get':
>                 #x = _Getter(self.__dict__[name[3:]])
> 		val = self.__dict__[name[3:]]
> 		x = eval("lambda :x", {"x": val})
>                 self.__dict__[name] = x
>                 return x
>             elif name[:3] == 'set':
> 		x = eval("lambda val: set(dict, %s, val)" % `name[3:]`,
> 			 {"dict": self.__dict__, "set": operator.setitem})
>                 #x = _Setter(self, name[3:])
>                 self.__dict__[name] = x

And again a circular reference.

>                 return x
>         except KeyError:
>             pass
>         raise AttributeError
> It's only about 10% slower, though for n==10 it's still about
> 8 times slower.  And it has a nasty case of cyclical references
> (since __dict__ caches the lambda functions, which contain
> a reference to __dict__).  I'll leave cleaning that up as an
> exercise to the student :)

Oh. I should have read the entire post before writing my reply :)


-- 
Bernhard Herzog	  | Sketch, a python based drawing program
herzog at online.de  | http://www.online.de/home/sketch/



From aahz at netcom.com  Sat Jul 10 19:21:58 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 10 Jul 1999 23:21:58 GMT
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu>
Message-ID: <7m9to2$6qg$77@rks1.urz.tu-dresden.de>

In article <3787D36A.815B7C29 at ssec.wisc.edu>,
Nick Bower  <nick.bower at ssec.wisc.edu> wrote:
>
>I feel really dumb asking this question, but how do I change the current
>working directory?  For browsing the file system I mean.

os.chdir()
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From g_will at cyberus.ca  Mon Jul 19 13:40:43 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 19 Jul 1999 17:40:43 GMT
Subject: tkinter call backs
References: <01bed202$58879280$e855c3d1@Cyberus.cyberus.ca> <00ca01bed209$5543adf0$f29b12c2@secret.pythonware.com>
Message-ID: <01bed20d$bef09c00$e855c3d1@Cyberus.cyberus.ca>

Hi,

In fileopen(), root is a unknown name because (I believe) I dont pass root
to it.  Fileopen() is a call back (and I dont know how to pass an instance
or variable through a call back).  I get an name error when I then use

fileName=tkFileDialog.askopenfilename(parent=root, title="Open")

This part is in a module to make a menu.  When the module is tested by
itself it runs fine and root does not give and error.  print root returns .
(period).  When the module is imported into the main module and then used,
I now get an error.  root=Tk() is defined in the main module.  Is this a
problem of local and global names?  How do I overcome it?




Fredrik Lundh <fredrik at pythonware.com> wrote in article
<00ca01bed209$5543adf0$f29b12c2 at secret.pythonware.com>...
> Gordon Williams <g_will at cyberus.ca> wrote:
> > I have a simple problem that I cant find the solution.  I have a menu
item
> > to open a file which has a command "openfile".  I would like to
activate
> > "tkFileDialog.askopenfilename" to get the name of the file.  This
requires
> > "parent=root" or something like that for the position of the message
box.
> > 
> > How do I pass "root" into openfile?  
> 
> after which he promptly answered his own
> question:
> 
> >     fileName=tkFileDialog.askopenfilename(parent=root, title="Open")
> 
> what am I missing here?
> 
> </F>
> 
> 



From jam at newimage.com  Fri Jul 16 17:39:36 1999
From: jam at newimage.com (jam)
Date: Fri, 16 Jul 1999 17:39:36 -0400
Subject: J, you can't pass your own test.
In-Reply-To: <7mnj5k$id9$1@nnrp1.deja.com>; from srenner@lycosmail.com on Fri, Jul 16, 1999 at 03:30:14PM +0000
References: <378E3481.EE31C97@uswest.net> <19990715175320.B20155@toast.internal> <7mnj5k$id9$1@nnrp1.deja.com>
Message-ID: <19990716173936.F20935@toast.internal>

On Fri, Jul 16, 1999 at 03:30:14PM +0000, srenner at lycosmail.com wrote:
> 
> The original post, while confused and not giving enough information, and
> somewhat resentful in tone, did not deserve the spleen heaped upon it in
> the first response, which was unrepresentative of the general tone of
> this newsgroup.
> 
> 

I completely understand. however, that is the way I felt about the tone of
the post to which I was responding. someone was complaining about the price
of the python books, and I took major offense to that. it is and remains one
voice amongst many, and the fact that more respected contributors than
myself *did* take the time to respond to such a post says a lot for the
community as a whole, which is far more important than my own political
standing in the group at the moment.

regards,
J
-- 
|| visit gfd <http://quark.newimage.com/> 
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From ivanlan at callware.com  Mon Jul 19 16:06:58 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 19 Jul 1999 20:06:58 GMT
Subject: Installation Aborted! on NT
References: <7mvtbm$55r$1@nnrp1.deja.com>
Message-ID: <37938562.943B1919@callware.com>

Hi All--

Ross Boylan wrote:
> 
> I've run into a strange problem I thought I'd report, in case it might
> affect other users.  When I try to install py152.exe or win32all-
> 125.exe on my system from my account, I get the blue installation
> background and then a box with the title "Installation Aborted!" and
> the message "Internal Error"
> 

[snip]

> The system in NT 4.0 WS, SP5.  Software is minimal, including MSIE 5.0,
> Exceed 6.1 (X window system), 3Com Nic software, Matrox drivers, and
> Lexmark print utility.  There was a password authentication system on
> the machine (Ascend AppServNT v2.0Ci1), but we removed it.  40 bit
> security.
> 

The only thing I can offer here is that I did my installations on NT4.0
workstation with service pack 4 installed, and only upgraded to SP5
quite a while later (128-bit, by the way).


<bit-by-the-bits>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From viofis at my-deja.com  Thu Jul  1 20:35:51 1999
From: viofis at my-deja.com (viofis at my-deja.com)
Date: Fri, 02 Jul 1999 00:35:51 GMT
Subject: please help with simple regex
Message-ID: <7lh1h6$t84$1@nnrp1.deja.com>

I'm a beginner at Python, so this may be a stupid question:

I have a string, for example, myString="the Potato".
1. Convert it to all lowercase ("the potato")
2. Match it against a list of strings - "potato,tomato,corn,etc...",
extract "potato" (if it matches with something on the list), then save
into myNewString.

Any help is greatly appreciated.

thanks,
Vadim



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From joyce at netcard.com  Mon Jul 19 20:05:00 1999
From: joyce at netcard.com (Joyce Stack)
Date: Tue, 20 Jul 1999 00:05:00 GMT
Subject: python + bash + new user
Message-ID: <3793BE52.43E023BB@netcard.com>

hello,

I am not new to programming but new to python...i want to know how do i
call functions from the prompt....

joyce




From fredrik at pythonware.com  Mon Jul 12 19:37:36 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 12 Jul 1999 23:37:36 GMT
Subject: Zope newbee help please.
References: <378B1E46.338A2CFD@freenet.co.uk>
Message-ID: <001201beccbf$86a2ab20$f29b12c2@secret.pythonware.com>

Sreekant Kodela <sreekantk at freenet.co.uk> wrote:
> I just installed Zope. Looks good. Is there any good tutorial or guide
> for the Zope to create a good site.

I think you'll find that you get much better
Zope support on the Zope mailing lists than
on the Python language newsgroup.

see
http://www.zope.org

and especially
http://www.zope.org/Community

for more info.

</F>




From tismer at appliedbiometrics.com  Fri Jul 16 05:08:57 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Fri, 16 Jul 1999 09:08:57 GMT
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net> <FExMpA.Juv@midway.uchicago.edu> <378EA68A.F2D8A7E@callware.com>
Message-ID: <378EF6A9.E35E3C97@appliedbiometrics.com>


Ivan Van Laningham wrote:

[every real content snipped:]

> As Christian Tismer says, "flea dirt."  And a more apt description of
> Perl I have yet to encounter.

Well, actually I have two killer categorizations for languages.
The other one is "write only".
Perl is living happily in the intersection (beneath APL which I like).

this-is-my-private-optinion-and-not-my-
   -employers-(who-doesn't-exist)-ly y'rs - chris->I#love at perl$

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From Haimo.Zobernig at cern.ch  Tue Jul 13 04:51:36 1999
From: Haimo.Zobernig at cern.ch (Haimo G. Zobernig)
Date: Tue, 13 Jul 1999 10:51:36 +0200
Subject: making a Windows NT service out of a python program
Message-ID: <378AFE18.F02306AC@cern.ch>

I need to turn a few python programs into Windows NT services. Has anyone on
this newsgroup got any advice for that? I know about SRVANY.EXE from the NT
resource kit. Is saw there is a module called win32service in the win32
extensions. The docs give just syntactic info, but I need to understand the
relevant concepts first.
I presume I'll need to get hold of some MS docs, but I have no idea where to
start. Any pointers?

-- 
Haimo G. Zobernig                                  tel   +41 22 7677361
Univ. of Wisconsin                                 fax   +41 22 7678370
c/o CERN Bld. 32 R-A02
1211 Geneve 23, Switzerland                email Haimo.Zobernig at cern.ch



From gmcm at hypernet.com  Fri Jul 30 09:39:28 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 30 Jul 1999 08:39:28 -0500
Subject: NT select.select?
In-Reply-To: <al8zp0epeje.fsf@sirppi.helsinki.fi>
Message-ID: <1278784323-25399323@hypernet.com>

Markus Stenberg wrote:

> Is there some low limit on maximum number of sockets you can have in
> the Python-NT's select call? A program that happens to work
> perfectly on Linux seems to die on NT around 64(?) sockets to the
> 'too many file descriptors in call' error.

Correct. Windows has the same limit (64) in the native 
WaitForMultipleObjects calls.
 
> Any portable ways to bypass it?

Doubt it. A Windows developer would probably have N threads each
handling M sockets.

FWIW, HPUX comes with 3 different flavors of sockets and select. One
version has a low limit, the second a moderate limit and the third
has a limit so huge I can't imagine testing it. 

- Gordon



From tim_one at email.msn.com  Fri Jul 23 04:35:07 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 23 Jul 1999 04:35:07 -0400
Subject: int(long(-sys.maxint-1)) fails on Linux
In-Reply-To: <37981790.24835500@vision.arc.nasa.gov>
Message-ID: <000301bed4e6$44f499c0$392d2399@tim>

[Chad Netzer]
> ...
> Here is some assembly of the same region, compiled with the -O1
> flag, which DOES exhibit the incorrect behavior.  Notice that it is
> quite different from the previous code,

Yes, but despite being wrong, it's much faster <wink>.

> ... and I don't know what condition codes "lea" sets,

That's the problem:  lea doesn't change any flags (it's the "load effective
address" instruction, and when used for its *intended* purpose (loading
effective addresses <wink>), it would get in the way if it did alter any
flags).

So it's as if "x << 1" had never been computed (which the lea does, btw --
it's tricky!) or tested against 0.  They (mis)optimized a necessary "test"
away; see the first version of the working code you posted, which does a
very similar thing in this section (but has the crucial "test" after the
lea).

> clearly the problem is with pgcc w/ optimization turned on, which is a
> Mandrake 6.0 issue.

Yup!  You should report it to "them", of course -- such things rarely fix
themselves ...

people-also-introduce-bugs-when-optimizing-so-view-it-as-an-example-
    of-advanced-humanistic-ai<wink>-ly y'rs  - tim





From graham at sloth.math.uga.edu  Fri Jul 16 11:17:06 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 16 Jul 1999 15:17:06 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com>
Message-ID: <7mnidi$t2b$1@cronkite.cc.uga.edu>

Graham Matthews <graham at sloth.math.uga.edu> wrote:
: >[Parser Combinators] 
Neel Krishnaswami (neelk at brick.cswv.com) wrote:
: Thanks for the tip -- it led me to do some digging on the web and I
: found a nice description of this in a paper called "Monadic Parser
: Combinators" by Graham Hutton and Erik Meijer. Its URL is:
: 
:   http://www.cs.nott.ac.uk/Department/Staff/gmh/monparsing.ps

Yes this is a nice place to start. Another nice place to look is in
the Clean documentation. Somewhere in their docs (I think in the old
book) there is a Clean implementation of combinators (without all the
"do" notation).

The trickiest part for me was to get the error handling to work well,
since Clean/Haskell don't have exceptions. This should be "easier"
in Python, although you are welcome to see my code if you want. Also
feel free to email me questions.

Neel Krishnaswami (neelk at brick.cswv.com) wrote:
: This really seems like a clean and straightforward way of doing
: things.  My experience with recursive descent parsers has been pretty
: bad, but now I suspect that this is because I wasn't organizing my
: work in a systematic enough fashion. 

I had similar experiences with rec.descent parsers. By the way this
technique is not limited to parsing. I use essentially the same 
code to implement a type checker too. A parser has input a string
or a file, and output an AST. The type checker has the same structure
as a parser but it's input (the thing it parses) is an AST and it's
output an AST annotated with types. Combinators are more general
than just parsing and patterns. They are a form of sequencing
construct.

graham

-- 
              This ain't no technological breakdown
                 Oh no, this is the road to hell



From fdrake at cnri.reston.va.us  Thu Jul 29 10:56:51 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Thu, 29 Jul 1999 14:56:51 GMT
Subject: where is nsremote?
In-Reply-To: <yhs9080jfxv.fsf@haplix8.jlab.org>
References: <yhs9080jfxv.fsf@haplix8.jlab.org>
Message-ID: <14240.27571.7650.562566@weyr.cnri.reston.va.us>


Jeff Templon writes:
 > I found a tool called nsremote by Scott Hassan that does something I
 > want.  However, the link is broken on the python www site.  Does anyone
 > know where I can download this?

Jeff,
  You didn't say what functionality you're looking for.  If this is a
Netscape remote-control module, that's not too hard, at least on Unix
(don't know about other platforms, but suspect those would be better
documented).
  I've attached a module I threw together that can be used to control
Netscape and Grail on Unix.  I'd welcome additions for other platforms 
and browsers.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives

-------------- next part --------------
A non-text attachment was scrubbed...
Name: browser.py
Type: text/x-python
Size: 1833 bytes
Desc: browser remote control module
URL: <http://mail.python.org/pipermail/python-list/attachments/19990729/e79c40ff/attachment.py>

From ze5yr at p98.f112.n480.z2.fidonet.org  Thu Jul  1 12:32:47 1999
From: ze5yr at p98.f112.n480.z2.fidonet.org (ze5yr)
Date: Thu, 01 Jul 1999 17:32:47 +0100
Subject: theads & global namespaces
Message-ID: <00001622@bossar.com.pl>

From: ze5yr <ze5yr at my-deja.com>

In article <7ldqi2$nc1$1 at nnrp1.deja.com>,
  rdev at my-deja.com wrote:

> - If you're running on Solaris, the thread won't get to run
> unless you give up some time.  So try putting a time.sleep(.001)
> into your while loop.  On NT, this is not necessary (don't know
> why.)

It works on NT, but the looping thread will eat up horrendous amounts
of CPU time (at least it does if under pthreads, don't know if Python
threads would do the same).  So it's a good idea to use time.sleep
under NT as well.

--
Cliff Crawford                           -><-
http://www.people.cornell.edu/pages/cjc26
                            empty-handed I go,
)O(       and behold the spade is in my hands


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From cgw at fnal.gov  Wed Jul  7 20:19:34 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Thu, 8 Jul 1999 00:19:34 GMT
Subject: Two Pythons talking to each other?
Message-ID: <199907080019.TAA22244@buffalo.fnal.gov>

There's no reason, just because the two programs are running on the same
machine, that they can't talk to each other using network protocols (TCP/IP,
sockets, etc).  You can use the "loopback" address of 127.0.0.1 to make
all connections.  You really don't want to develop and test using a 
completely different communications scheme (e.g. pipes, files) then 
re-implement everything for the networked version, and debug it all
over again.  Client and server can both run on the same box.



From Ovidiu.Predescu at p98.f112.n480.z2.fidonet.org  Thu Jul  1 18:47:57 1999
From: Ovidiu.Predescu at p98.f112.n480.z2.fidonet.org (Ovidiu Predescu)
Date: Thu, 01 Jul 1999 23:47:57 +0100
Subject: Why does shelve make such large files?
Message-ID: <00001642@bossar.com.pl>

From: Ovidiu Predescu <ovidiu at cup.hp.com>

Gerrit Holl wrote:

> Is it really necesarry for shelve to make such large files?
> Have a look at this:
> /tmp> python
> Python 1.5.2 (#1, Apr 18 1999, 00:16:12)  [GCC 2.7.2.3] on linux2
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import shelve
> >>> d = shelve.open('database')
> >>> d['key'] = 'value'
> >>>
> /tmp> ls -l database
> -rw-rw-r--   1 gerrit   gerrit      16384 Jul  1 21:13 database
>                                     ^^^^^
> 
> 16 KB for only one key!!?
> pickle seems to make _much_ smaller files!
> 
> Why is this?

The shelve module uses DBM which is like a small database that allows
you to store objects and search it for a given key. DBM allows you to
store millions of objects and search for them later without requiring
you to load all the file in memory first.

The pickle module on the other hand is serializing the objects with the
purpose of deserializing them _all_ from the file later. Pickle does not
offer you any way to search for data based on a key, you have to do this
yourself after the objects have been created from the file. This is
opposed to the way shelve handles this, all the key accesses and
insertions in a shelve object are actually reads or writes to or from
the DBM file.

And to answer your question, DBM is creating these big files because of
the way it manages the database. The data in the database file could
have gaps as a result of multiple insertions and deletions. Pickle's
data in files is a simple representation of the objects that were
written and there is no way to update the file other than rewriting it
entirely.

-- 
Ovidiu Predescu <ovidiu at cup.hp.com>
http://www.geocities.com/SiliconValley/Monitor/7464/



From greg.ewing at compaq.com  Sun Jul  4 16:53:53 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Mon, 05 Jul 1999 08:53:53 +1200
Subject: printing with print
References: <377925EF.1CBB640F@starvision.com>
Message-ID: <377FC9E1.21B2C433@compaq.com>

Angus MacKay wrote:
> 
> I like the printing of any object ability of "print" so I
> do not want to use "" % style printing

The "%s" format will handle any object, too -
not just strings:

>>> x = [1, 2]
>>> y = ("buckle", ("my", "shoe"))
>>> print "%s%s" % (x, y)
[1, 2]('buckle', ('my', 'shoe'))

Look, ma, no space!

Greg



From dalke at bioreason.com  Thu Jul 15 03:30:41 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Thu, 15 Jul 1999 01:30:41 -0600
Subject: Getters and Setters
References: <7mjrqg$e4o@ds2.acs.ucalgary.ca> <7mjune$e2g@ds2.acs.ucalgary.ca>
Message-ID: <378D8E21.957C6AEA@bioreason.com>

Neil Schemenauer <nascheme at ucalgary.ca> proposed:
> # A mixin class to automagicly create getter/setter methods for
> # class attributes.

> class GetterSetter:
>     def __getattr__(self, name):
>         try:
>             if name[:3] == 'get':
>                 return _Getter(self.__dict__[name[3:]])
>             elif name[:3] == 'set':
>                 return _Setter(self, name[3:])
>         except KeyError:
>             pass
>         raise AttributeError

Here's one optimization, don't recreate getter/setter objects.
How about this:

class GetterSetter:
    def __getattr__(self, name):
        try:
            if name[:3] == 'get':
                x = _Getter(self.__dict__[name[3:]])
                self.__dict__[name] = x
                return x
            elif name[:3] == 'set':
                x = _Setter(self, name[3:])
                self.__dict__[name] = x
                return x
        except KeyError:
            pass
        raise AttributeError

For your timing test that reduced the overhead to a factor of
about 2.5 (I upped n to 10000 to get better times on my machine):

  Your version
val> python getattr.py
Comparing 10000 iterations
GetterSetter 1.78183102608
explicit function 0.201846957207

  Cached objects
val> python getattr.py
Comparing 10000 iterations
GetterSetter 0.455724000931
explicit function 0.207740902901

Still, you're doing this __call__ lookup every time, so why
not use something which doesn't need that lookup, like a function?
My first thought was to create a lambda function or a function
in local scope, but the argument passing trick (lamba val=val: val)
would mess up the API by allowing an argument to be passed in.

Since your timing test does so many repeats (esp. when I increase
n :), that suggests I can take a 1-time cost per attribute, so
I can use an eval.

class GetterSetter:
    def __getattr__(self, name):
        try:
            if name[:3] == 'get':
                #x = _Getter(self.__dict__[name[3:]])
		val = self.__dict__[name[3:]]
		x = eval("lambda :x", {"x": val})
                self.__dict__[name] = x
                return x
            elif name[:3] == 'set':
                x = _Setter(self, name[3:])
                self.__dict__[name] = x
                return x
        except KeyError:
            pass
        raise AttributeError

val> python getattr.py
Comparing 10000 iterations
GetterSetter 0.359815955162
explicit function 0.210291981697

Now, only 70% slower.

For my final trick, I'll do the same to the Setter:
import operator
class GetterSetter:
    def __getattr__(self, name):
        try:
            if name[:3] == 'get':
                #x = _Getter(self.__dict__[name[3:]])
		val = self.__dict__[name[3:]]
		x = eval("lambda :x", {"x": val})
                self.__dict__[name] = x
                return x
            elif name[:3] == 'set':
		x = eval("lambda val: set(dict, %s, val)" % `name[3:]`,
			 {"dict": self.__dict__, "set": operator.setitem})
                #x = _Setter(self, name[3:])
                self.__dict__[name] = x
                return x
        except KeyError:
            pass
        raise AttributeError


Notice that there are now no helper classes, so you can mitigate
some of the eval() time by not doing the class creation + __init__.

The results?

val> python getattr.py
Comparing 10000 iterations
GetterSetter 0.236621975899
explicit function 0.20987200737

It's only about 10% slower, though for n==10 it's still about
8 times slower.  And it has a nasty case of cyclical references
(since __dict__ caches the lambda functions, which contain
a reference to __dict__).  I'll leave cleaning that up as an
exercise to the student :)


						Andrew Dalke
						dalke at acm.org



From kas at maps.magnetic-ink.dk  Sat Jul 17 05:26:19 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Sat, 17 Jul 1999 11:26:19 +0200
Subject: Snippet: bitcount()
References: <000601bed02a$2235af20$06a22299@tim> <slrn7p0hsv.e38.kas@titan.magnetic-ink.dk>
Message-ID: <slrn7p0j1r.eaf.kas@titan.magnetic-ink.dk>

A few minutes ago I wrote:

> the method fails for signed ints, like e.g. 0x8000000

Erratum: should of course read "the methods fails for unsigned ints, like
e.g. 0x80000000, because python perceives them as signed ints (use longs
instead)"...


  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From Michael.P..Reilly at p98.f112.n480.z2.fidonet.org  Sun Jul  4 17:05:37 1999
From: Michael.P..Reilly at p98.f112.n480.z2.fidonet.org (Michael P. Reilly)
Date: Sun, 04 Jul 1999 22:05:37 +0100
Subject: Is it just me, or..
Message-ID: <000017cd@bossar.com.pl>

From: "Michael P. Reilly" <arcege at shore.net>

Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
: Is there a bug in the handling of pipes with the Python? Esp. in
: non-blocking way.

: Example:
:  (r,w)=os.pipe()
:  r = os.fdopen(r, "r", 0)
:  w = os.fdopen(w, "w", 16384)
:  w.write('foobarbaz')
:  select.select([r],[],[]) # returns r, as it should
:  r.read(1) # returns 'f'
:  select.select([r],[],[]) # blocks!

: Now select _blocks_, despite there being obviously some data that at least
: should not have been buffered elsewhere. Also, fdopened stuff's reading
: features are bit annoying, as doing read(bigNumber) results in blocking
: call, as opposed to recv(bigNumber)'s semantics. Thus instead of pipes
: (=fast) I am using TCP/UNIX sockets (=slow) but they seem to work better
: for what I need. Am I missing something really obvious?

: This is on Linux, if it matters.

At first glance, I'd say that it is the buffering in the fdopen(w, "w",
16384) call.  Put the bufsize to 0 or call the flush method.

  -Arcege



From bwinton at tor.dhs.org  Tue Jul 13 23:15:41 1999
From: bwinton at tor.dhs.org (Blake Winton)
Date: Wed, 14 Jul 1999 03:15:41 GMT
Subject: XML-RPC
References: <931899884snz@vision25.demon.co.uk> <378BBD4D.9DA43FE5@prescod.net>
Message-ID: <slrn7ont1m.i9e.bwinton@tor.dhs.org>

On Tue, 13 Jul 1999 22:27:25 GMT, Paul Prescod <paul at prescod.net> wrote:
>Phil Hunt wrote:
>> Does anyone know if an XML-RPC server exists for Python?
>> There's a python XML-RPC client at www.pythonware.com/products/xmlrpc/
>> but I'm not aware of a Python server).
>I think that you want to look at Zope.

I think that's a bit misleading...  The XML-RPC client also contains an
implementation of an XML-RPC server.  You don't need to bring Zope into
the picture at all.

Later,
Blake.

-- 
One Will.  One Dream.  One Truth.  One Destiny.  One Love.



From clgonsal at kami.com  Fri Jul  9 07:16:22 1999
From: clgonsal at kami.com (C. Laurence Gonsalves)
Date: Fri, 09 Jul 1999 11:16:22 +0000
Subject: VIM and [Python] block jumping
References: <3781347D.CB2FBAB3@starvision.com> <Pine.SOL.4.05.9907051548470.13052-100000@saturn.radsci.uci.edu> <37815B80.B4B51F94@starvision.com> <3785D753.DB2FF1ED@kami.com>
Message-ID: <7m9tmo$6qg$34@rks1.urz.tu-dresden.de>

Of course, right after I posted that, realized that there's a problem.
The blockMotion function doesn't account for lines that are
continuations of other lines. That means that triple-quoted string
literals, lines with more {[('s that }])'s, and lines joined with \ can
screw it up. Are there any other situations? I'll try to post a
corrected version later. Some of those problems are pretty nasty though
(triple-quoted strings in particular).

-- 
  C. Laurence Gonsalves              "Any sufficiently advanced
  clgonsal at kami.com                   technology is indistinguishable
  http://www.cryogen.com/clgonsal/    from magic" -- Arthur C. Clarke



From simngolda at voyager.net  Wed Jul 14 00:00:25 1999
From: simngolda at voyager.net (Sim & Golda Zacks)
Date: Wed, 14 Jul 1999 00:00:25 -0400
Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
References: <tot2m7.su.ln@final.math.uni-bonn.de>
Message-ID: <7mh1qu$c6m@newsops.execpc.com>

I am in the exact same situation as you are. I am a web programmer and I'm
trying to implement the CGI and database stuff with Python. I am using the
HTMLFMT module from the INTERNET PROGRAMMING book and the cgi module from the
standard library. What the HTMLFMT library does for you is just that you don't
have to type in all the tags, basically it's nothing magical, if I didn't have
it I would have to make something up and it probably wouldn't be half as good.
the standard cgi unit gives you all the fields from the form, and I haven't
looked at the cgi modules from the book yet to see if they give me any added
benefit. The big problem I came across was my web host, and all of the other
ones I talked to, refused to  install the mysql interface to Python, and it has
to be included in the build (or something like that) So I just installed gadfly,
which seems to be working great for me right now. I'm still playing with it not
in production yet.
I have no idea what ZOPE does, but everyone who talks about it seems to love it.

Hope this helps

Sim Zacks

Thomas Ackermann wrote:

> Hi!
>
> I need to generate some hompages and cgi-script for database access,
> forms and such and want to do it in Python, of course. As it is i have
> no experience with it's WWW-related stuff ...
>
> There are htmllib and cgi from the library, HTMLfmt and genCGI from
> the "Internet Programming with Python" book, HTMLgen itself and
> something just announced called "html", which seems to be a variant of
> HTMLgen(?!)
>
> I have not much time to experiment with those modules and would like
> to hear opinions about to use what module     ;-)
>
> Also, i am not exactly sure what Zope is - it seems to be something to
> create WWW-Pages and allow access to databases and it has it's own
> WebServer (a variant of Medusa) ...
>
> So, what to use to allow for an Internet-Shop with pages created from
> databases and allowing customers to purchase over net?
>
> Can do Zope such things? Is it overkill at first? What can it do
> beyond? (And yes, i read some documentation, but i do not even
> understand what an "Application Server" is    ;-)
>
>           Thanx in advance and
>           Byebye,
> --
>   Thomas Ackermann | Tel. +49-(0)228/631369|73-7773 | <tgm at math.uni-bonn.de>
>              finger tgm at rhein.math.uni-bonn.de for public key
>                GNU LINUX Python gtk pygtk MySQL FUDGE GURPS




From Gary.Pennington at uk.sun.com  Wed Jul  7 04:11:30 1999
From: Gary.Pennington at uk.sun.com (Gary Pennington - UK Performance Centre)
Date: Wed, 07 Jul 1999 09:11:30 +0100
Subject: python core dump (SIGBUS) on Solaris
References: <3F8BA119B48AD11197470008C7A493CB98B443@HQOBGEX1>
Message-ID: <37830BB2.94C4613D@uk.sun.com>

ScherBi at BAM.com wrote:

> I have a bus error problem with this module too.  Some cookies cause the
> error when trying to create an instance for reading cookies; eg:
>
> ck = Cookie.Cookie(os.environ['HTTP_COOKIE'])
>
> Not all cookies cause the error.  If I copy a suspect cookie string into a
> variable in an interactive session and try to create an instance as above,
> Python dumps core with a Bus Error.
>
> I've been able to create a work-around for now by parsing the cookie string,
> but just wanted to say that Mr. Pennington is not alone in this...
>
> PS.
> This is also on Solaris 2.6, compiled with gcc 2.8.1
>
> Bill Scherer
>

<snip>

I've done (a lot) more investigation since yesterday. It appears that the
problem is being generated inside cPickle in the loads function. I altered the
Cookie._debabelize function to trap the error more explicitly and found the
following:-

Test Code :-

    str = _unquote(val)
    try:
        return loads(str)
    except Exception:
        return str

Python 1.5.2 (#1, Jul  6 1999, 16:50:14)  [GCC egcs-2.91.66 19990314
(egcs-1.1.2  on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import Cookie
>>> C=Cookie.Cookie()
>>> C.load("webJudgeUser=garyp;")
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python1.5/Cookie.py", line 544, in load
    self.__ParseString(rawdata)
  File "/usr/local/lib/python1.5/Cookie.py", line 572, in __ParseString
    M.set(K, apply(self.net_setfunc, (V,)), V)
  File "/usr/local/lib/python1.5/Cookie.py", line 342, in _debabelize
    return loads(str)
cPickle.BadPickleGet: path
>>>


Hmmm, something going wrong in the cPickle loads function not being trapped in
the exception handler unless you specifically state Exception.

I haven't got the skills/time to look at cPickle, so as a work-around I use the
pickle module and this works fine. Just modify the import statement at the top
of the Cookie.py module.

I hope this is helpful.

Yours,

Gary





From Tim.Peters at p98.f112.n480.z2.fidonet.org  Fri Jul  2 18:03:15 1999
From: Tim.Peters at p98.f112.n480.z2.fidonet.org (Tim Peters)
Date: Fri, 02 Jul 1999 23:03:15 +0100
Subject: How do you make clickable Python/Win9x scripts?
Message-ID: <000016a4@bossar.com.pl>

From: "Tim Peters" <tim_one at email.msn.com>

[Rob]
> Is there a way to make a Python script file a clickable .BAT file in
> Win9x?

This kind of thing is the best I've ever come across:

rem=''' Here's a cheery greeting you can't get rid of, so use it
@echo off
:any other bat cmds you'd like to do first
:and then execute python on this file:
%PATH_TO_YOUR_PYTHON_DIR%python %0
goto endofpython
'''

name = raw_input("What's your name? ")
print "Hi,", name

rem = '''
:endofpython
:any other bat cmds you'd like to do afterwards
:'''

The batcrap is hidden from Python in multiline triple-quoted strings
assigned to "rem", the initial lines of which W9X treat as comments.  Note
the use of label notation (:''') to hide the final string closer from the
bat processor.  BTW, you're usually better off starting bat comments via ":"
than "rem"!  To understand why, try sticking

: <hi>

in one bat file and

rem <hi>

in another <wink>.

command.com-leaves-the-teensiest-bit-to-be-desired-ly y'rs  - tim






From gerrit.holl at pobox.com  Fri Jul  2 06:27:49 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 2 Jul 1999 12:27:49 +0200
Subject: print color strings?
In-Reply-To: <7lgo5b$15is$1@nntp6.u.washington.edu>; from Donn Cave on Thu, Jul 01, 1999 at 09:55:55PM +0000
References: <19990701213528.A8877@optiplex.palga.uucp> <14203.53386.710688.87678@weyr.cnri.reston.va.us> <7lgo5b$15is$1@nntp6.u.washington.edu>
Message-ID: <19990702122749.A9975@optiplex.palga.uucp>

On Thu, Jul 01, 1999 at 09:55:55PM +0000, Donn Cave wrote:
> "Fred L. Drake" <fdrake at cnri.reston.va.us> writes:
> |Gerrit Holl writes:
> |> is there an easy way to print color strings in python?
> |> 
> |> A function like printcolor('brightyellow', 'red', 'spam'), that finds
> |> out on what kind of terminal you are and prints it to stdout?
> |
> | Gerrit,
> |   If you're doing this on Unix, take a look at the ncurses library;
> | there's a Python interface to that.  (I think Oliver Andrich did it;
> | searching DejaNews should find it fairly quickly since this keeps
> | coming up.)
> |   The standard curses module will be documented in the next release of 
> | the Library Reference.
> 
> And if that doesn't work, probably your termcap or terminfo definitions
> lack the necessary information about color escape sequences.  That's a
> widespread problem.
> 
> The color escape sequences that have worked for me on a couple of terminal
> types have been of the form <ESC>[32m, where 32 is a number from 30 to 37.
> I use only 31 to 36.  30 and 37 are more or less black and white, but the
> details vary, so I switch from color to non-color with <ESC>[m instead of
> trying to sort out the details of black and white vs. background and
> foreground.
> 

Ah, thanks. I was forgotten that would work on python the same as in the
bash-shell...

> Of course these are terminal specific escape sequences.  They're evil and
> you must not use them, especially if you have a choice.
> 

Thanks. They're only for my sys.ps1, so it'll be good enough for me..,

But I thought there might be a simpler interface than a whole curses library,
but there's not.

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From www at OpsResearch.com  Mon Jul 12 19:01:56 1999
From: www at OpsResearch.com (www at OpsResearch.com)
Date: Mon, 12 Jul 1999 23:01:56 GMT
Subject: Resume Referral
Message-ID: <378A73E0.904F8823@OpsResearch.com>

For more information see:

   http://OpsResearch.com





From rceci at adams.com  Thu Jul 29 17:58:50 1999
From: rceci at adams.com (rceci at adams.com)
Date: Thu, 29 Jul 1999 21:58:50 GMT
Subject: ImportError on SWIGged C++ modules.
Message-ID: <7nqiqk$8sd$1@nnrp1.deja.com>

Hello,
I have a c++ library that runs great on HPUX (10.20) and IRIX (6.4), in
a native (ie non-Python) environment.  I have also been successful at
getting the library to import w/o problems on IRIX, as a SWIGged module
in Python.  Now, I am having difficulty getting the library to import on
HP.  I found an old reference to C++ modules in HP needing
"Python/importdl.c" to have shl_load changed to cxxshl_load and what
not.  I am not sure that is still valid, or if I have some other build
problem, specific to HPUX.  I am using Python 1.5.1. and SWIG1.1p5.
When I import the python library I get the following errors:

Python 1.5.1 (#1, Jul 22 1999, 13:51:48) [C] on hp-uxB
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
Traceback (innermost last):
  File "/vobs/release//run_python.py", line 7, in ?
    import MDIreg
  File "/vobs/release//python/lib/hpux10/python1.5/MDIreg.py", line 2,
in ?
    import MDIregc
ImportError: Failed to load
/vobs/release//python/lib/hpux10/python1.5/MDIregcmodule.sl
>>>

======

adding -v to the python invocation tells me more:

# /vobs/release//python/lib/hpux10/python1.5/exceptions.pyc has bad
mtime
import exceptions # from
/vobs/release//python/lib/hpux10/python1.5/exceptions.py
# can't create /vobs/release//python/lib/hpux10/python1.5/exceptions.pyc
Python 1.5.1 (#1, Jul 22 1999, 13:51:48) [C] on hp-uxB
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
# /vobs/release//python/lib/hpux10/python1.5/MDIreg.pyc matches
/vobs/release//python/lib/hpux10/python1.5/MDIreg.py
import MDIreg # precompiled from
/vobs/release//python/lib/hpux10/python1.5/MDIreg.pyc
shl_load /vobs/release//python/lib/hpux10/python1.5/MDIregcmodule.sl
/usr/lib/dld.sl: Unresolved symbol: __StaticCtorTable_Start (data)  from
/usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: __ZStaticCtorTable_End (data)  from
/usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: __StaticCtorTable_End (data)  from
/usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: __ZStaticCtorTable_Start (data)
from /usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: U_update_state_vector (code)  from
/usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: U_get_shLib_unw_tbl (code)  from
/usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: U_resume_execution (code)  from
/usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: U_get_previous_frame_x (code)  from
/usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: U_get_unwind_table (code)  from
/usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: U_get_shLib_text_addr (code)  from
/usr/lib/libCsup.1
/usr/lib/dld.sl: Unresolved symbol: U_get_unwind_entry (code)  from
/usr/lib/libCsup.1
shl_findsym initMDIregc
import MDIregc # dynamically loaded from
/vobs/release//python/lib/hpux10/python1.5/MDIregcmodule.sl
Contacting Registry ...
Bus error

By the way, I am using static C++ constructors in a few cases and am not
sure that is part of the problem.

Many Thanks


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From bkc at Murkworks.com  Thu Jul 22 12:58:30 1999
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 22 Jul 1999 12:58:30 -0400
Subject: Wasted two days debugging my RefCount error : read my tale of woe so you don't do the same (LONG)
Message-ID: <7n7iks$cpr$1@news.clarkson.edu>

Yes, it's true. I must hang my head in shame for not refcounting correctly,
thereby causing my Python module (DLL) to crash when it's unloaded.

After two days of tracing around with the Borland Debugger (also having
entirely recompiled Python to use CodeGuard under Borland), I have the
source of my bug.

It all started so innocently. You see, I wanted to be able to call Novell
NDS Apis from Python, so I looked to swig for a solution. What a great
little program!

But, my program kept crashing everytime the DLL module got unloaded.

The crash I got was "code at offset 0x0000000 referenced memory at
0x000000". Pretty useful, eh?

Now, I know this means that I either over-wrote the stack and returned to 0,
or I had a pointer to a function that got overwritten with 0. Naturally I
suspected that the Novell APIs where trashing memory ( I have to handle a
lot of arbitrary data buffers of large size).

After much fussing and tracing and debugging I just couldn't get a
breakpoint to capture a memory overwrite. Eventually I whittled it down to a
dictionary cleanup, and finally got a break point set in dictobject.c

static void
dict_dealloc(mp)
 register dictobject *mp;
{
 register int i;
 register dictentry *ep;
 for (i = 0, ep = mp->ma_table; i < mp->ma_size; i++, ep++) {
  if (ep->me_key != NULL) {
   Py_DECREF(ep->me_key);
  }
  if (ep->me_value != NULL) {
   Py_DECREF(ep->me_value);   <-------- DIES HERE
  }
 }


The DecRef of the value would eventually make a call to 0x00000000.

Aha, I thought, something is overwriting a type definition structure with
NULLs! Now, if I could find out what object type it was I could set a data
write breakpoint on the structure and I could catch the culprit function.

Lets see ... run/break/ not null, ok, run/break/ not null, ok.  run/break
not null  (keep this up about 250 times).

run/break its null. Yeah! lets see.. object type -> Py_None

Hmm. Where is Py_None?  Its actually PyNothing_Type.. Find the structure
address and set a breakpoint on it. Where's that global stored.. Ah, I see
it..

What the f***

Looky here:

static PyTypeObject PyNothing_Type = {
 PyObject_HEAD_INIT(&PyType_Type)
 0,
 "None",
 0,
 0,
 0,  /*tp_dealloc*/ /*never called*/


Suddenly, bells and whistles start going off.. Hmm. I recall something about
Py_None.. Must think.. lessee, a week ago, I said to myself "I don't think
this is quite right".. Now, what was that exactly... "yeah, refcount Py_None
in that swig typemap"..

Which brings me to the source of the problem, to wit:

Given an NDS function like this:

%apply STRINGBUF {nptr attrVal};

NWDSCCODE
NWDSGetAttrVal
(
   NWDSContextHandle context,
   pBuf_T            buf,
   nuint32           syntaxID,
   nptr              attrVal
);
%clear STRINGBUF;

Swig creates a nice Python function to call NWDSGetAttrVal and return the
error code (NWDSCCODE) as an integer.

I thought, "hey, all NDS functions return 0 on success and non-zero on
error, so lets throw an exception if the return code is not 0".

So I wrote a typemap like this:

%typemap(python,out) NWDSCCODE  {
  if(0 != $source) {
    ThrowException($source,"$name",NULL);
      return NULL;
   } else {
    $target = Py_None;
   }
 }

The function ThrowException() sets a module global (argh! wish this was
per-thread, but I don't unlock the interpreter in there anyway).

So, given a function that has output args, Swig produces code that looks
somewhat like this:

(the NDS call goes here)

  if(0 != _result) {
    ThrowException(_result,"NWDSGetAttrName",NULL);
      return NULL;
   } else {
    _resultobj = Py_None;
   }
 }
{
    PyObject *o;

    o = Py_BuildValue("s",_arg2);
    if ((!_resultobj) || (_resultobj == Py_None)) {
      _resultobj = o;
    } else {
  _resultobj = t_output_helper(_resultobj, o);
    }
}

In the above case, this function always returns an output variable, so
_resultobj is over-written by the t_output_helper functions.

But what happens when the NDS API function would only return NWDSCCODE? Then
swig produces this:

    _result = (NWDSCCODE )NWDSGetAttrVal(_arg0,_arg1,_arg2,_arg3);
{
  if(0 != _result) {
    ThrowException(_result,"NWDSGetAttrVal",NULL);
      return NULL;
   } else {
    _resultobj = Py_None;
   }
 }
    return _resultobj;
}

Notice something funny?

THERES NO INCREF!

  Py_INCREF(Py_None);


Without the INCREF, Py_None would eventually get garbage collected when it's
refcount went to 0 (which it normally never would go to zero because it's
artificially referenced somehwere I suspect)

It seems that the simple solution is to change the typemap to this:

%typemap(python,out) NWDSCCODE  {
  if(0 != $source) {
    ThrowException($source,"$name",NULL);
      return NULL;
   } else {
      Py_INCREF(Py_None);
    $target = Py_None;
   }
 }


But then, I'll get code like this:

  if(0 != _result) {
    ThrowException(_result,"NWDSGetAttrName",NULL);
      return NULL;
   } else {
    Py_INCREF(Py_None);
    _resultobj = Py_None;
   }
 }
{
    PyObject *o;

    o = Py_BuildValue("s",_arg2);
    if ((!_resultobj) || (_resultobj == Py_None)) {
      _resultobj = o;
    } else {
  _resultobj = t_output_helper(_resultobj, o);
    }
}

See the problem? I'm incrementing Py_None but then not referencing it. So
the increment count keeps going up. Eventually it'll wrap (heh, well
probably not for a long time).

I suspect that not free-ing Py_None is ok, since the source says that it'll
never be called, but I'm unhappy about incrementing endlessly.

Anyone have any suggestions on how to fix this with better typemaps? What I
need is a typemap for "functions with argout" and a different one for
"functions without argout".

NOTES AND COMMENTS:

I would have saved a lot of time if Py_None's dealloc function pointed to a
real function that would print a message to stderr saying "You dummy, you're
trying to garbage collect Py_None".


If you're compiling Python under Borland C, you can pretty much copy the VC
options in config.h, but you need to add this:

#define MALLOC_ZERO_RETURNS_NULL 1

Enjoy the rest of the week, I've wasted mine :-(

--
Brad Clements,
bkc at murkworks.com





From dalke at bioreason.com  Wed Jul  7 17:36:56 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Wed, 07 Jul 1999 15:36:56 -0600
Subject: scoping assertion
References: <37839491.3F7C1F6@mikemccandless.com>
Message-ID: <3783C878.B8C3790B@bioreason.com>

Michael McCandless <mail at mikemccandless.com> said:
> But because there are no macros in Python, and because you can't
> make this a function (?), I think you'd need to type this out every
> time.

Try this:

import sys

def nonexist(name):
  try:
    1/0
  except ZeroDivisionError:
    assert not sys.exc_info()[2].tb_frame.f_back.f_locals.has_key(name),\
      "variable %s already exists" % `name`

def test():
    nonexist("x")
    print "x does not exist"
    x = 1
    print "What about now?"
    nonexist("x")
    print "x does not exist"

>>> test()
x does not exist
What about now?
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 6, in test
  File "<stdin>", line 5, in nonexist
AssertionError: variable 'x' already exists

  There is a lot of tricky things you can do by peering at the
stack.


						Andrew Dalke
						dalke at acm.org



From vaton at postoffice.pacbell.net  Mon Jul 26 17:01:42 1999
From: vaton at postoffice.pacbell.net (Rob)
Date: Mon, 26 Jul 1999 14:01:42 -0700
Subject: Python and Red Hat Linux 6.0
Message-ID: <379CCCB6.A90DCE48@postoffice.pacbell.net>

When programming in Perl or Python, I am using a header like the
following on an executable file to cause the shell (bash or csh) to run
the scripts as Perl/Python files:
#!/usr/bin/perl   for Perl   and #!/usr/bin/python   for Python. I've
checked and the compilers are located in the appropriate directories. I
keep getting this when I run scripts:
bash: <script name>: command not found. Does Linux support the '#!'
notation in the file header, or do I always have to run the scripts by
explicitly typing in 'perl' or 'python' before the script name?






From puetsch at gmx.de  Sat Jul 31 03:59:01 1999
From: puetsch at gmx.de (Felix Puetsch)
Date: 31 Jul 1999 07:59:01 GMT
Subject: os.name
References: <1278785134-25350540@hypernet.com> <V5ho3.4340$VL2.463886@news.direcpc.com> <001e01beda8e$2111d2e0$f29b12c2@secret.pythonware.com> <Ggto3.4361$VL2.467372@news.direcpc.com>
Message-ID: <7nuac5$3rs$1@chief.death.de>

emile at fenx.com (Emile van Sebille) wrote in 
<Ggto3.4361$VL2.467372 at news.direcpc.com>:

>By the way, in windows at a command prompt:
>
>95,98:    net config
>nt:       net config [server|workstation]
>
>gives a line that starts : Software version

Wouldn't "ver" be easier?

-- 
Felix P?tsch



From esr at thyrsus.com  Fri Jul 30 00:52:11 1999
From: esr at thyrsus.com (Eric S. Raymond)
Date: Fri, 30 Jul 1999 00:52:11 -0400
Subject: Need help with TCP/IP client access from Windows
In-Reply-To: <1278839420-22085362@hypernet.com>; from Gordon McMillan on Thu, Jul 29, 1999 at 05:21:12PM -0500
References: <slrn7q1ej3.18d.esr@snark.thyrsus.com> <1278839420-22085362@hypernet.com>
Message-ID: <19990730005211.T7149@thyrsus.com>

Gordon McMillan <gmcm at hypernet.com>:
> No problem there. (The only difference - when using Python - is that 
> you can only select on sockets on Windows).

So you're saying my blocking-socket code will work as-is?  Great.  The
Python manual implies otherwise...
  
> > (2) How does one read and write the Windows equivalent of a dotfile?
> >     Must this involve the (shudder) registry?
> 
> You could, by using Mark's win32 extensions. MS would prefer you
> kept this kind of info in the registry, but that doesn't stop
> anybody, least of all you <g>.
> 
> The issue is really what is meant by a HOME directory. On NT, this 
> has a certain amount of meaning - there's probably a 
> %WINDIR%/profiles/%USERNAME% directory tree for that user (where 
> those funky %...% things are environment vars), but probably it's 
> safer to get that path out of the registry. On Win9x, it's pretty 
> meaningless, since a logon and username are pretty much an empty 
> formality. There's almost no way to keep multiple users on a Win9x 
> box from stepping on each other. (Of course, I mean multiple serial 
> users; the only possible kind in Windows).

Hm..."serial users".  I like that.  It just...*sounds right* for Windows.

:-)
 
> Probably better just to have a default file, and let the user 
> override that if need be. Which may well never happen.

OK, but where to put it?  I don't know what the conventions are.

> > (3) Is there a reliable way for Python programs under Windows to
> > point a local
> >     browser at a specified URL?
> 
> The best way is to use WinExec from Mark's win32 extensions. You just 
> give it the URL and it loads the user's default browser. Using 
> os.system is harder, because most of the time a browser is not on the 
> user's path.

Are these in the stock 1.5.2 Windows distribution?

> However, this worked fine for me:
> 
> browse='"C:\\Program Files\\Netscape\\Communicator\\Program\\netscape"
> %s'
> 
> Note the double quotes around the nowhere-but-Windows <snicker> path 
> with spaces in it, and that argv[1] is expected to be a URL (same for 
> IE). Also note that the path doesn't go thru the c runtime lib, so 
> those wonderful backslashes can't be avoided.

Got it.  Thanks.
-- 
		<a href="http://www.tuxedo.org/~esr">Eric S. Raymond</a>

Our society won't be truly free until "None of the Above" is always an option.



From avv at quasar.ipa.nw.ru  Thu Jul  1 19:11:00 1999
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Thu, 01 Jul 1999 16:11:00 -0700
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net> <377B711A.E8B65E0B@cfdrc.com> <377BC620.F5553FF5@easystreet.com>
Message-ID: <377BF583.6846FBF9@quasar.ipa.nw.ru>

Al Christians wrote:

> I looked at Fox and put it aside because I need a grid control.
> wxPython seems to be the winner for Windows in that category, even
> though the grid is pretty minimal.
>
> Any other candidates with a grid?

Did you look at the demos for the underlying toolkit? Some of them presents a
'table' which is probably less rich than that of wxWindows, but it does exist.

What it really lacks is a 'richtext'/html widget, sort of which is present and
actively developed in wxWindows. As I understand it's just about to become
available in wxPython.

Alexander





From ritm at gnet.tn  Tue Jul 13 13:42:02 1999
From: ritm at gnet.tn (MB)
Date: Tue, 13 Jul 1999 19:42:02 +0200
Subject: newbie questions about sys.path, wxpython, IDLE and pythonwin
Message-ID: <378B7A69.12874AAC@gnet.tn>

I need to know exactly how the sys.path is built. In my system I have no
pythonpath variable but the sys.path returns a list. I suppose it was
built by default. This list doesn't contain all my python
subdirectories. So I want to know when and how this list was built ?
(yes I examined the registry and found references to python directories
but they don't match my sys.path)
Another question in relation with the first. I downloaded wxPython an
tried to run the demo.py from the pythonwin GUI (opening
c:\wxpython\demo\demo.py and running it). It runs perfectly. Then I
tried the same thing through the IDLE GUI and received this error:
------------------
Traceback (innermost last):
  File "C:\PROGRA~1\PYTHON\TOOLS\IDLE\ScriptBinding.py", line 131, in
run_module_event
    execfile(filename, mod.__dict__)
  File "C:\Program Files\Python\wxPython\demo\demo.py", line 3, in ?
    import Main
ImportError: No module named Main
-----------------

Obviously it was a problem of path. So I added the directory an its
subdirectories to sys.path and tried again. I received this

---------------------------------
Traceback (innermost last):
  File "C:\PROGRA~1\PYTHON\WXPYTHON\DEMO\Main.py", line 186, in
OnSelChanged
    self.GetDemoFile('Main.py')
  File "C:\PROGRA~1\PYTHON\WXPYTHON\DEMO\Main.py", line 221, in
GetDemoFile
    except IOException:
NameError: IOException
Traceback (innermost last):
  File "C:\PROGRA~1\PYTHON\WXPYTHON\DEMO\Main.py", line 186, in
OnSelChanged
    self.GetDemoFile('Main.py')
  File "C:\PROGRA~1\PYTHON\WXPYTHON\DEMO\Main.py", line 221, in
GetDemoFile
    except IOException:
NameError: IOException
---------------------------------

Other thing. I examined demo.py through the debugger under pythonwin I
found this statement:
'import miscc'
The interpreter import this module but is not able to locate it. When I
type >>> miscc it says that is a built-in module. Are there modules that
are not python files ? What is a built-in module ? I didn't find its
definition in the docs.
Last question. About the  syntaxic ':' at the end of a line when it
begins with def, class,if ,when,etc... Is it absolutely necessary for
the language ?
Thanks





From mikael at pobox.com  Thu Jul  1 10:38:53 1999
From: mikael at pobox.com (Mikael Lyngvig)
Date: Thu, 01 Jul 1999 14:38:53 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <slrn7nai1j.8ll.zeitlin@seth.lpthe.jussieu.fr> <64F6457BE0585BB3.1CF2F67BC8A680BD.A19A9FEE23BC9E58@lp.airnews.net> <7lfm5q$bra$1@nnrp1.deja.com>
Message-ID: <F17A0E3671B7D39C.03C1D18540AB728A.FC25CA1FDC9EE114@lp.airnews.net>

On Thu, 01 Jul 1999 12:16:00 GMT, Andrew Cooke
<andrew at andrewcooke.free-online.co.uk> wrote:

>I'm confused.  I'm sitting at a moderate NT box (266MHz, 96Mb) - my
>work machine, and running the Tk demos.  There is no problem at all -
>they are instantaneous.
>
>Is this problem to do with memory?  Or is it a W95 thing?  If so,
>is it due to thread scheduling?

Not memory, all machines I've tried it on had 128 MB.  Not Windows 95,
I never tried Tk (or anything but a game) on a Windows 9x machine
(only NT).



-- Mikael



From tim_one at email.msn.com  Sat Jul  3 14:37:59 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 3 Jul 1999 14:37:59 -0400
Subject: Slicing beyond list --  bug or featurette?
In-Reply-To: <377D234A.4C9BDAC0@cisco.com>
Message-ID: <000401bec583$2d05e020$769e2299@tim>

[Pierre Fortin]
> Just ran into a situation which can be summarized as follows:
>
> list = [ 0,1,2,3,4 ]
>
> list[<0..4>] returns the appropriate item
>
> list[m] where m is >= len(list) expectedly fails
>
> list[2:50] returns [ 2,3,4 ]
> list[20:] or list[20:30] returns []
>
> Can I count on this, or should I code try/except for the day it's
> fixed?

It's guaranteed by footnote 2 to the table in the Library Ref's section
2.1.5 (Sequence Types).  I think the Language Ref leaves it open to debate,
though (if anything ever changes here, it's almost certainly that the
Language Ref will get more words blessing the current implementation).

> This 'feature' reduces the error situations; but is it correct/intended?

All three <wink>.  Slicing returns a subsequence, and the possibility to
return an empty subsequence works out very smoothly in practice.

More dubious is that negative indices index "from the right end":

    x[-1] == x[len(x)-1]

This is too handy to give up, but in my experience *has* let some errors
slip by; while I can't recall any case where

    x[insane_1 : insane_2] == []

didn't simply handle an expected endcase gracefully.  Doesn't mean we
couldn't construct one, just saying I've never seen one occur naturally.

BTW, in some older code you may see the strange idiom

    prefix[sys.maxint:] = suffix

or more likely

    prefix[HUGE:] = suffix

in a loop where HUGE = sys.maxint is computed once outside the loop.  In
1.5.2 it's spelled

    prefix.extend(suffix)

the-set-of-all-i-less-than-2-where-i-greater-than-100-is-just-empty-ly
'rs  - tim





From max at rightworks.com  Thu Jul 15 02:04:40 1999
From: max at rightworks.com (max at rightworks.com)
Date: Thu, 15 Jul 1999 06:04:40 GMT
Subject: PyApache binaries
Message-ID: <7mjtlk$6v3$1@nnrp1.deja.com>

Could anyone point me to where I can find NT binaries for PyApache.
thanks
max


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From godzilla at netmeg.net  Tue Jul 20 09:29:47 1999
From: godzilla at netmeg.net (Les Schaffer)
Date: Tue, 20 Jul 1999 13:29:47 GMT
Subject: How fast can we multiply?
References: <001801bed279$92c348a0$139e2299@tim>
Message-ID: <m2wvvvqwjb.fsf@netmeg.net>

Tim and Christian spoke:

C> Then I also had a look into NumPy and its matrix multiplication.  I
C> stunned. Again no attempt to optimize big O. It could be done with
C> Python, the algorithms are known, the fruit is waiting to be picked.

T> Large dense matmult isn't very common.  Sing along:

i dont see the connection between floating point matrix multiplies and 
the integer arithmetic tim sketched out in the previous post.

Spock, explain!

les 'my-denseness-is-multiplying' schaffer
-- 
____        Les Schaffer              ___| --->> Engineering R&D <<---
Theoretical & Applied Mechanics          |  Designspring, Inc. 
Center for Radiophysics & Space Research |  http://www.designspring.com/
Cornell Univ.  schaffer at tam.cornell.edu  |  les at designspring.com



From fdrake at cnri.reston.va.us  Fri Jul 23 10:45:00 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 23 Jul 1999 14:45:00 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
In-Reply-To: <14231.42621.76462.103289@96.chicago-25-30rs.il.dial-access.att.net>
References: <001101bed3f2$5e8874a0$642d2399@tim>
	<7n675p$qsk@dfw-ixnews7.ix.netcom.com>
	<14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>
	<7n82u1$jid$1@nntp.Stanford.EDU>
	<14231.42621.76462.103289@96.chicago-25-30rs.il.dial-access.att.net>
Message-ID: <14232.32748.787261.820359@weyr.cnri.reston.va.us>

Skip Montanaro writes:
 > Again, the string.re thing was a bug.  Most system modules do follow this
 > convention.

  Many older ones do not, unfortunately.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From eugene.leitl at lrz.uni-muenchen.de  Sun Jul  4 20:22:05 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Sun,  4 Jul 1999 17:22:05 -0700 (PDT)
Subject: Language extensibility (was: Why is tcl broken?)
In-Reply-To: <wk3dz4u320.fsf@ifi.uio.no>
References: <1283099377-77267484@hypernet.com>
	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com>
	<375FDCA7.62FF@mailserver.hursley.ibm.com>
	<7josm1$bgc$1@Starbase.NeoSoft.COM>
	<37628EBF.207F7A8F@iname.com>
	<37682DD3.43923243@compaq.com>
	<7ka2be$qja$1@nntp6.u.washington.edu>
	<37776FFA.2D85@mailserver.hursley.ibm.com>
	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it>
	<876748mcxg.fsf@home.ivm.de>
	<37788BB9.ABD@mailserver.hursley.ibm.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.ibm.com>
	<nkjzp1hizvb.fsf@tfeb.org>
	<377A3383.D453468@iname.com>
	<377aa228.31380423@news.jpl.nasa.gov>
	<slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net>
	<wkzp1g2t2e.fsf@ifi.uio.no>
	<ey3r9mskvja.fsf@lostwithiel.tfeb.org>
	<x27lojoums.fsf@todday.aiai.ed.ac.uk>
	<wmcn1xenodh.fsf@polytechnique.fr>
	<14206.52213.119553.809177@lrz.de>
	<wk3dz4u320.fsf@ifi.uio.no>
Message-ID: <14207.64080.35726.524863@lrz.de>

Lars Marius Garshol writes:
 
 > * Eugene Leitl
 > | 
 > | Church, A., The Calculi of Lambda Conversion, Princeton University
 > | Press, Princeton, N.J., 1941
 > 
 > According to David Harel's Algorithmics, the correct reference is
 > actually S.C. Kleene "A Theory of Positive Integers in Formal Logic",
 > Amer. J. Math. 57 (1935), pp. 153-173,219-244.

Wildly off-topic, but fun: http://www.cl.cam.ac.uk/users/mgh1001/limericks.html

One of the most prolific writers of Limericks was John von Neumann, 
who is said to have written thousands of them. One of my favorites 
of his augurs the development of VR ... 

     There was a young man called Kleene, 
     Who invented a f@#&ing machine, 
         Concave, or convex, 
         It fit either sex, 
     And was remarkably easy to clean! 

     - John von Neumann 

Thanks to Jim Horning for this version, which may well be the more correct: 

     There was a young man called Kleene, 
     Who invented a f@#&ing machine, 
         Concave and convex, 
         It could screw either sex, 
     And diddle itself in between! 

     - John von Neumann 



From Jerome.Kalifa at polytechnique.fr  Sat Jul  3 04:12:10 1999
From: Jerome.Kalifa at polytechnique.fr (Jerome Kalifa)
Date: 03 Jul 1999 10:12:10 +0200
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> 	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> 	<375FDCA7.62FF@mailserver.hursley.ibm.com> 	<7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> 	<37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> 	<37776FFA.2D85@mailserver.hursley.ibm.com> 	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it> 	<876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> 	<lw67471fyi.fsf@copernico.parades.rm.cnr.it> 	<3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> 	<377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> 	<slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> 	<wkzp1g2t2e.fsf@ifi.uio.no> <ey3r9mskvja.fsf@lostwithiel.tfeb.org> <x27lojoums.fsf@todday.aiai.ed.ac.uk>
Message-ID: <wmcn1xenodh.fsf@polytechnique.fr>

Jeff Dalton <jeff at todday.aiai.ed.ac.uk> writes:
[...]
> 
> Well *Lisp* dates back to the late 50s.  Scheme may be from 75,
> but Lisp is a different matter.  And lambda calculus is, what,
> 1949?
> 

No, as a mathematical theory, it was developed before the war.
-- 
Jerome Kalifa
Centre de Mathematiques Appliquees,  Ecole Polytechnique.
91128 Palaiseau Cedex, France.  (33)169333981



From tgabriel at bellsouth.net  Mon Jul 19 06:49:54 1999
From: tgabriel at bellsouth.net (Xenophanes)
Date: Mon, 19 Jul 1999 06:49:54 -0400
Subject: OK python gurus!
References: <7mfn4c$m80$1@nnrp1.deja.com>
Message-ID: <379302D2.D9023EED@bellsouth.net>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19990719/757eae4b/attachment.html>

From fredrik at pythonware.com  Wed Jul 14 14:55:55 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 14 Jul 1999 18:55:55 GMT
Subject: Terminating Python...
References: <BFC0587634C3D111AACB00805FA766FA02C1C652@exchny12.ny.smb.com>
Message-ID: <001b01bece2a$819d23a0$f29b12c2@secret.pythonware.com>

Velasquez, Carlos <carlos.velasquez at ssmb.com> wrote:
> I just downloaded and built Python v1.51 for the Sun Solaris platform. All
> the tests went well and I created some simple scripts to verify its
> operation. One question is, how do I exit python? The only thing that works
> is <Ctrl> Z but that just stops the job. if I execute Python again and Ctrl
> Z again, and if I do a ps -ef to see all processes, there are 2 python
> instances.
> 
> Is there a correct way to exit from Python?

the tutorial has the answer:
http://www.python.org/doc/current/tut/node4.html

    Typing an EOF character (Control-D on Unix, Control-Z on
    DOS or Windows) at the primary prompt causes the inter-
    preter to exit with a zero exit status.

which is a slightly convoluted way to say:

    "to get out, press Control-D if you're on a Unix box.  On
    Windows and MS-DOS, use Control-Z followed by Return
    instead.

</F>




From tismer at appliedbiometrics.com  Fri Jul 16 04:27:15 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Fri, 16 Jul 1999 08:27:15 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim>
		<7mlu8n$6ug$1@brick.cswv.com> <378EAC15.20269642@compaq.com>
Message-ID: <378EECE3.3851C55D@appliedbiometrics.com>


Greg Ewing wrote:
> 
> Neel Krishnaswami wrote:
> >
> > What's the difference, exactly? AFAICT you need to save the execution
> > state when suspending both coroutines and generators, but I might be
> > missing something obvious....
> 
> The crucial difference is that a generator is never resumed
> after its caller has returned. This means that the generator's
> state can be pushed onto the same stack as the caller's.

Fine.

> A coroutine, on the other hand, can outlive the context in
> which it was created, and therefore needs a stack all of
> its own.

Smell of danger...

> Another way to think about it is that a generator call is
> equivalent to an ordinary call where one of the parameters
> is a procedure. For example, where in "generator python"
> you might write

...

> If Python had something akin to Smalltalk code blocks,
> generators wouldn't be needed. It would be nifty to
> be able to write something like
> 
>   for_each_even_number_up_to(42) -> (e):
>     print e, "is an even number!"

Traceback (innermost last):
  File "<inbox:Re: Stackless & String-processing>", line 22, in claim
SemanticMisconceptionError: 
     things should be made as simple as possible, but not any simpler
>>> #:-)

>>> # sorry did you really mean that? 
>>> # Where is your generator's state?
>>> # how would you do backtracking?

hoping-that-I'm-misconcepted-and-not-you-ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From markus_kohler at hp.com  Tue Jul  6 05:15:05 1999
From: markus_kohler at hp.com (Markus Kohler)
Date: Tue, 06 Jul 1999 11:15:05 +0200
Subject: Python and QT 2.0 anyone ?
Message-ID: <3781C919.BAD892BB@hp.com>

Hi,

Is there someone working on or has already finished QT 2.0 
support for Python ?

Greetings,
Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From arcege at shore.net  Mon Jul 19 20:40:28 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Tue, 20 Jul 1999 00:40:28 GMT
Subject: [Q] Type checking...go further...
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AABFC@molach.origin.ea.com>
Message-ID: <0APk3.528$ui4.177861@news.shore.net>

Stidolph, David <stidolph at origin.ea.com> wrote:
: You cannot have typed method parameters (Python is typeless), but can only
: check the types of parameters within the method.

: from types import * #Needed for different types like IntType

: def foo(a, b, c):
:   if type(a) is IntType:
:     print a,'is a integer'
:   elif type(a) is LongType:
:     print a,'is a long integer'
:   elif type(a) is StringType:
:     print a,'is a integer'
:   elif type(a) is FloatType:
:     print a,'is a floating point number'
:   elif type(a) is ListType:
:     print a,'is a list'
:   elif type(a) is TupleType:
:     print a,'is a tuple'

: or

: def foo(a,b,c):
:   if not (type(a) is IntType):
:     raise "Param a must be integer"
:   if not (type(b) is StringType):
:     raise "Param b must be string"
:   if not (type(c) is ListType):
:     raise "Param c must be list"
:   ...

The "isinstance()" builtin function also works on types:

  def foo(a, b, c):
    if not isinstance(a, types.IntType):
      raise TypeError, "Param a must be integer"
    if not isinstance(b, types.StringType):
      raise TypeError, "Param b must be string"
    if not isinstance(c, types.ListType):
      raise TypeError, "Param c must be list"

It is often better not to call "from types import *" since that can
clutter the namespace unnecessarily.

  -Arcege




From Florian.W. at p98.f112.n480.z2.fidonet.org  Thu Jul  1 21:08:21 1999
From: Florian.W. at p98.f112.n480.z2.fidonet.org (Florian W.)
Date: Fri, 02 Jul 1999 02:08:21 +0100
Subject: Newbie has trouble with in/output
Message-ID: <0000163a@bossar.com.pl>

From: "Florian W." <flowit at hotmail.com>

Hello Python-people

I have two little problems with Python (using WinNt )

1.) Being in a pipe:

I won't to filter the output of the compiler (e.g. only errors and warning
are reported, no logo etc.)
using the pipe-mechanismen :

cl.exe blabla | python filter_compiler_output.py

How do I catch the input and how do print the output ?
The processing requires line-oriented handling, so that I can't catch all
input first, process it as a block and finally display it all at once.
Is the way of calling correct at all ?

2.) Catch the output of an application:

This time I call the compiler myself (e.g. os.system('c:\cl.exe blabla'))
and
want to process the output of the compiler. But how do I catch it ?
os.system just returns the error code. And hacks like c:\cl.exe blabla >
output.txt
don't look very good. (Especially if there are some perl-hackers)

Mfg Florian







From mal at lemburg.com  Fri Jul  9 03:53:00 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 9 Jul 1999 07:53:00 GMT
Subject: Why are index() and count() only for mutable sequences?
References: <004701bec95f$186faa80$a0ee6dd1@BobAlexDesk> <3784F778.6A6B4568@lemburg.com> <19990708224110.B29441@optiplex.palga.uucp>
Message-ID: <3785AA5C.58FB2F5C@lemburg.com>

Gerrit Holl wrote:
> 
> On Thu, Jul 08, 1999 at 09:09:44PM +0200, M.-A. Lemburg wrote:
> > Date:         Thu, 08 Jul 1999 21:09:44 +0200
> > From: "M.-A. Lemburg" <mal at lemburg.com>
> > To: Bob Alexander <bobalex at ix.netcom.com>
> > Cc: python-list at python.org
> > Subject: Re: Why are index() and count() only for mutable sequences?
> >
> > Bob Alexander wrote:
> > >
> > > There is no modification of the sequence performed by these functions
> > > -- it seems as though they could be available for all sequences. It
> > > would certainly be useful for tuples as well as lists (strings, too,
> > > for that matter, even though we have the more general string.find()).
> >
> > You could try count(), index(), forall() and exists() from mxTools:
> >
> >       http://starship.skyport.net/~lemburg/mxTools.html
> >
> > Albeit, they work with conditions (functions that return 1/0) to
> > implement the testing process.
> >
> 
> Is this an anwser??
> He asked: "why"...

Well, count() and index() are methods not functions, so the
question was a little misleading. You cannot add methods to
*all* sequences, the only way to do this is by defining a
function which takes a sequence argument (*). That's why I gave
the hint to mxTools.

(*) This might change in future versions of Python though,
once types (classes coded in C) support the normal class
mechanisms too -- you could then add the methods to the base
class.

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   175 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From news at dorb.com  Thu Jul  8 12:58:16 1999
From: news at dorb.com (Darrell)
Date: Thu, 8 Jul 1999 12:58:16 -0400
Subject: Efficient string manipulation , Help please
References: <37856BF1.BD046003@freenet.co.uk>
Message-ID: <GN4h3.6634$D5.30915@newsr1.twcny.rr.com>

SGML gets around these kinds of problems with entities.
http://www.topdat.com/java/html/SGML/ENTITIES.HTM

Entities are like macros.
&lt; defines <
&gt; defines >
&quot; defines "

I'm suggesting that you replace the offending symbols as they come in with
an entity.
And replace the entities when you output the text.

Of course I have almost no idea about the details of your problem.
--
--Darrell
Sreekant Kodela <sreekantk at freenet.co.uk> wrote in message
news:37856BF1.BD046003 at freenet.co.uk...
>
> Dear all
>
> I seem to be in trouble here. I am trying to make some cgi scripts for
> my web site. I am using the postgresql as the backend db and a form
> based address book and news forum etc. The problem is everything is
> working ok as long as noone uses ' and " in the text. If they do, bingo
> , postgres refuses to accept. I tried using string library [mainly
> string.join(string.split('awkward'"'st'ring')). It works sometimes and
> doesn't some times.
>
> Any better and efficient way to deal with this situation !
>
> Any suggestions and ideas will be greatly appreciated.
>
> Thanks
> sreekant
>
>
>





From Darrell at p98.f112.n480.z2.fidonet.org  Thu Jul  1 05:38:17 1999
From: Darrell at p98.f112.n480.z2.fidonet.org (Darrell)
Date: Thu, 01 Jul 1999 10:38:17 +0100
Subject: getopt and Non-Unixes?
Message-ID: <00001629@bossar.com.pl>

From: "Darrell" <news at dorb.com>

You end up using -v -d -f or -vdf this last one will throw Dos people for a
loop.
I supply the help message if the command line isn't understood.

def runIt():
 try:
  opts, args = getopt.getopt(sys.argv[1:], "l:hdpf:o:")
 except getopt.error, msg:
  sys.stderr.write("Error: %s\n" % str(msg))
  print help_str
  sys.exit(1)


--
--Darrell
Gerrit Holl <gerrit.holl at pobox.com> wrote in message
news:19990701174416.A27152 at optiplex.palga.uucp...
> Hello all,
>
> I was wondering how getopt worked on dos/windows (and other operating
systems).
> While in dos, the default way to handle arguments is "/v /d /f" (isn't
it?).
>
> I don't have access to any micro$oft machine (it doesn't make me
unhappy...),
> but I just wanted to know how my program works on Dos/windows/macintosch
and
> other non-Unix systems.
>
> regards,
>
> Gerrit.
>
>




From tismer at appliedbiometrics.com  Fri Jul 23 07:31:43 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Fri, 23 Jul 1999 11:31:43 GMT
Subject: int(long(-sys.maxint-1)) fails on Linux
References: <000001bed49f$c052e820$2c2d2399@tim>
		<3797D8A1.E6D95CCB@vision.arc.nasa.gov> <37981790.24835500@vision.arc.nasa.gov>
Message-ID: <3798529F.C77FEAEE@appliedbiometrics.com>


Chad Netzer wrote:
> 
> Here I go, replying to my own posts again...
...
> Here is some assembly of the same region, compiled with the -O1 flag, which DOES
> exhibit the incorrect behavior.  Notice that it is quite different from the previous code,
> and I think the bug is that it uses the "LEA" instruction improperly:
> 
>  <PyLong_AsLong+89>: test   %edx,%edx                            ; x < 0
>  <PyLong_AsLong+91>: jge    0x806c22a <PyLong_AsLong+106>
>  <PyLong_AsLong+93>: test   %ebx,%ebx                            ; sign > 0
>  <PyLong_AsLong+95>: jg     0x806c231 <PyLong_AsLong+113> ; goto overflow
>  <PyLong_AsLong+97>: lea    0x0(,%edx,2),%eax             ; ?? Where is (x << 1)?

lea doesn't change any flags. It optimizes the shift to an add
which is fine. The implied test of the result is just not done
and this is a false optimization.

Somehing similar happened to me when I hand-optimized assembly
code. It is tempting to replace
   add ax, 1
by
   inc ax

but beware, that doesn't set flags, and it's likely that you forget
about it when changing tests.

>  <PyLong_AsLong+104>: jne    0x806c231 <PyLong_AsLong+113> ; goto overflow
>  <PyLong_AsLong+106>: mov    %edx,%eax
>  <PyLong_AsLong+108>: imul   %ebx,%eax
>  <PyLong_AsLong+111>: jmp    0x806c246 <PyLong_AsLong+134>
> 
> As you can see, the original code had add %eax, %eax instruction to shift
> left by one, the optimized one doesn't.  I don't know what the heck it
> was trying to do, and I don't know what condition codes "lea" sets, so
> I can't know what it intended.  In any case, I speculate that line +104
> was causing a jump to the overflow handler when it should not.
> 
> Anyway, I apologize for cluttering the newsgroup with assembly;  clearly the
> problem is with pgcc w/ optimization turned on, which is a Mandrake 6.0 issue.

I can't see any bad about assembly. This is just another data type,
and Python can be a fine tool as a generic, 
table driven disassembler ;-)

but that's another project which I will continue next year - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From Gaetan_Corneau at baan.com  Wed Jul 21 11:24:30 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 21 Jul 1999 15:24:30 GMT
Subject: String2List: odd function
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF4ED@ex-quebec-u1.baan.com>

Isidore,

This one is easy  :)

Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> s = '[1, 2, 3, 4, 5, ["a", "b", "c"], 6, 7, 8]'
>>> s = "l = " + s
>>> exec(s)
>>> l
[1, 2, 3, 4, 5, ['a', 'b', 'c'], 6, 7, 8]
>>>

Bye,
______________________________________________________
   Gaetan Corneau
   Software Developer (System integration Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From gmcm at hypernet.com  Fri Jul  9 15:50:27 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 9 Jul 1999 14:50:27 -0500
Subject: Win32 installer / "standalone" packager
Message-ID: <1280576364-81429328@hypernet.com>

This is to announce that I have made a major upgrade to my Win32 
Installer and that it is now available on python.org's contrib site.

What's New
 Updated for 1.5.2 
 Automatic handles Python/TK apps
 Support for console / console-less apps
 New one-step generation (no config file needed)
 Detects binary (dll dependencies) even without dumpbin.exe 
installed.

Where is it?
 ftp://ftp.python.org/pub/python/contrib/System/Installer_r_01.exe

How's it licensed?
 As loosely as possible.

More Info:

Installer is a set of utilities for creating installation
packages and quasi-standalone python executables for Win32. No
compiler is required.

Win32Installer steals from freeze, Christian Tismer's sqfreeze,
and Greg Stein's imputils.py. It will detect TK usage, and 
automatically install the necessary files. It can run console
apps or pure GUIs.

It can create a single-file self-extracting installation package.

It can also create quasi-standalone executables. These are not
single file; but everything exists in one directory tree with
no dependencies outside that tree, and no registry dependencies.
So uninstalling consists of deleting the directory tree, and it
shouldn't get messed up by other versions of python or TCL/TK
installed on the user's machine.

It also creates the smallest possible distributions - Python
modules are distributed as compressed archives (no source) and
imported directly from the archive. For example, Python/Lib
(excluding the test package) fits in a 475K Python_lib.pyz
file.

Win32Installer works with Python 1.5.2. It will not work with
Python 1.5.1 or earlier, mostly because of the fact that in
1.5.2, zlib.pyd is statically linked to zlib, while in 1.5.1
it was dynamically linked.

Files: Installer_r_01.exe
Contact: gmcm at hypernet.com

Copyright McMillan Enterprises; licensed on the "who cares" 
model.

- Gordon



From pphalen at teleo.net  Fri Jul 30 11:15:40 1999
From: pphalen at teleo.net (Patrick Phalen)
Date: Fri, 30 Jul 1999 08:15:40 -0700
Subject: 1.5.2 for: else:
References: <00ad01beda66$d6bdb9b0$f29b12c2@secret.pythonware.com>
Message-ID: <9907300817450G.00576@quadra.teleo.net>

On Fri, 30 Jul 1999, Fredrik Lundh wrote:

> huh?  there's only ONE *controlling* expression in
> the loop: the internal test that checks if the
> sequence has any more items. the else clause
> is executed if and only if THAT expression eval-
> Traceback (innermost last):
>   File "effbot.py", line 12032, in effbot
>   File "effbot.py", line 3428, in generate_reply
>     raise ImpedanceMismatch # bailing out
> effbot.ImpedanceMismatch

The truth is out. "Frederick Lundh" is a bot!



From andrew at starmedia.net  Thu Jul  1 11:05:30 1999
From: andrew at starmedia.net (Andrew Csillag)
Date: Thu, 1 Jul 1999 15:05:30 GMT
Subject: coverage tool for Python?
References: <377ABE31.249B1E51@cup.hp.com>
Message-ID: <377B83BA.448D6525@starmedia.net>

Ovidiu Predescu wrote:
> 
> Hi,
> 
> Are there any coverage tools for Python?
> 
> Thanks,
> --
> Ovidiu Predescu <ovidiu at cup.hp.com>
> http://www.geocities.com/SiliconValley/Monitor/7464/
Theres pycover also at
ftp://ftp.python.org/pub/python/contrib/All/pycover-0.1.tar.gz which,
while not all that great, should give you some good stuff.
-- 
Writing perl programs is like pounding stakes into your feet:
You could if you had to, but would be better off if you didn't.



From robin at jessikat.demon.co.uk  Wed Jul 21 09:29:30 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Wed, 21 Jul 1999 14:29:30 +0100
Subject: VC++ 6.0 larger exe files explained
Message-ID: <cxsW$GA6scl3EwNs@jessikat.demon.co.uk>

<rant>
As part of the Wintel push to take over the world it seems as though
VC++ 6.0 will now only generate files with sections 0x1000 (4k) or
longer. Under 5.0 and before the linker would allow much smaller exe
file sections. So python.exe which used to be 5k becomes 20k. Needless
to say this should also apply to dll's etc and appears to offer no
obvious advantage. Bill has never been other than cavalier with our
space time resources.
</rant>
-- 
Robin Becker



From timo at bbn.com  Mon Jul 12 14:13:25 1999
From: timo at bbn.com (Tim O'Malley)
Date: Mon, 12 Jul 1999 18:13:25 GMT
Subject: python core dump (SIGBUS) on Solaris
In-Reply-To: <14212.42641.217500.947671@amarok.cnri.reston.va.us>
References: <3F8BA119B48AD11197470008C7A493CB98B443@HQOBGEX1>
	<37830BB2.94C4613D@uk.sun.com>
	<37837154.7A6889A@ix.netcom.com>
	<slrn7o7nb9.eo.amk@207-172-111-58.s58.tnt1.ann.va.dialup.rcn.com>
	<37849166.DD658F5@uk.sun.com>
	<14212.42641.217500.947671@amarok.cnri.reston.va.us>
Message-ID: <14218.11963.905182.216092@ghost.bbn.com>

hola.

Sorry for my delay in seeing this thread...

Andrew M. Kuchling writes:
 > Which is fine if loads() just chokes on bad input and raises an
 > exception, but not if it dumps core.  I don't see how to change this
 > behaviour in Cookie.py without breaking backward compatibility by
 > turning unpickling off by default,

The latest version of Cookie.py (Feb 23, 1999) contains a fix of sorts
to allow working around the problem with pickling.  This version of
Cookie.py has 3 different types of Cookie objects.  The first type will
never use pickles; the second type always uses pickles; and the third
type follows the old behavior:

class SimpleCookie(BaseCookie):
    """SimpleCookie
    SimpleCookie supports strings as cookie values.  When setting
    the value using the dictionary assignment notation, SimpleCookie
    calls the builtin str() to convert the value to a string.  Values
    received from HTTP are kept as strings.
    """

class SerialCookie(BaseCookie):
    """SerialCookie
    SerialCookie supports arbitrary objects as cookie values. All
    values are serialized (using cPickle) before being sent to the
    client.  All incoming values are assumed to be valid Pickle
    representations.  IF AN INCOMING VALUE IS NOT IN A VALID PICKLE
    FORMAT, THEN AN EXCEPTION WILL BE RAISED.
    """

class SmartCookie(BaseCookie):
    """SmartCookie
    SmartCookie supports arbitrary objects as cookie values.  If the
    object is a string, then it is quoted.  If the object is not a
    string, however, then SmartCookie will use cPickle to serialize
    the object into a string representation.
    """


In order to preserve backward compatibility:

  Cookie = SmartCookie        


		----------------------------------------
Does this adequately address the concerns?
  If not, I'm certainly open to other ideas...

-- 
.....................................................................
.   happy daze            .   Age and treachery will beat           .
.        -tim O           .       youth and skill every time.       .
.....................................................................



From donn at u.washington.edu  Fri Jul  9 12:31:16 1999
From: donn at u.washington.edu (Donn Cave)
Date: 9 Jul 1999 16:31:16 GMT
Subject: Help! File Locking
References: <37856BF1.BD046003@freenet.co.uk> <7m53s4$cmq$1@nnrp1.deja.com>
Message-ID: <7m584k$k06$1@nntp6.u.washington.edu>

chacs <cs at ctzen.com> writes:
| Need some help in file locking.
|
| x1.py:
|   import fcntl, FCNTL
|   f = open('myfile', 'r+')
|   fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
|   input()
|   x = str(int(f.readline()) + 1)
|   f.seek(0)
|   f.write(x)
|   fcntl.flock(f.fileno(), FCNTL.LOCK_UN)
|   f.close()

I'd guess there may be a hole in the file interlocks here.  The file
object "f" buffers writes (in the underlying C library stdio).  The
real I/O to the file doesn't happen until the close(), which happens
after the file lock has been released and the other process waiting
for the lock wakes up.

Solution would be to insert f.flush() before releasing the lock.
Or just use posix open/read/write functions directly.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu
.............................
| x2.py:
|   import fcntl, FCNTL
|   f = open('myfile', 'r+')
|   fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
|   x = str(int(f.readline()) + 1)
|   f.seek(0)
|   f.write(x)
|   fcntl.flock(f.fileno(), FCNTL.LOCK_UN)
|   f.close()
|
| Both script are the same except x1 has an additional input() statement.
|
| Basically, what the script does is to open a file, lock it and update
| the first line to the next higher number.
|
| My problem is (assume file content is "0"):
|
| 1. in shell 1, run x1.py, x1 got lock and block at input().
| 2. in shell 2, run x2.py, x2 block at flock().
| 3. in shell 1, input something to unblock input().
| 4. x1 update "0" to "1", unlock and exit.
| 5. x2 gets lock, BUT instead of getting the new file content (i.e. "1")
| written by x1, it got the old content and update the file to "1" again
| (instead of reading "1" and update to "2").
|
| Any idea what is wrong ?
|
| Thanx in advance.
|
| C S Chang
| cs at ctzen.com



From Alexander.V..Voinov at p98.f112.n480.z2.fidonet.org  Thu Jul  1 05:11:00 1999
From: Alexander.V..Voinov at p98.f112.n480.z2.fidonet.org (Alexander V. Voinov)
Date: Thu, 01 Jul 1999 10:11:00 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001644@bossar.com.pl>

From: "Alexander V. Voinov" <avv at quasar.ipa.nw.ru>

Al Christians wrote:

> I looked at Fox and put it aside because I need a grid control.
> wxPython seems to be the winner for Windows in that category, even
> though the grid is pretty minimal.
>
> Any other candidates with a grid?

Did you look at the demos for the underlying toolkit? Some of them presents a
'table' which is probably less rich than that of wxWindows, but it does exist.

What it really lacks is a 'richtext'/html widget, sort of which is present and
actively developed in wxWindows. As I understand it's just about to become
available in wxPython.

Alexander




From Daniel_Faken at brown.edu  Wed Jul 14 11:05:47 1999
From: Daniel_Faken at brown.edu (Daniel Faken)
Date: Wed, 14 Jul 1999 11:05:47 -0400
Subject: Python interface to GLUI?
Message-ID: <378CA74B.AD128F28@brown.edu>

Hello fellow OpenGL + Python enthusiasts,

  Has anyone out there made a python interface to GLUI (Paul
Rademacher's free C++ GLUT-based GUI library) yet?

  (for info about GLUI see http://www.cs.unc.edu/~rademach/glui/)

  Just trying to avoid redoing it if someone else has..

thanks for any help,
Daniel

-- 
Daniel B. Faken
Daniel_Faken at brown.edu
http://www.chem.brown.edu/brochure/people/grads/faken/



From conrad at srl.caltech.edu  Fri Jul  9 20:34:04 1999
From: conrad at srl.caltech.edu (Conrad Steenberg)
Date: Fri, 09 Jul 1999 17:34:04 -0700
Subject: Printing returned object in embedded app
Message-ID: <7m9tn9$6qg$52@rks1.urz.tu-dresden.de>

Hi

I'm trying to construct an interactive front-end for Python, using GTK.

My C program passes strings to PyRun_String, and obtains the returned object.
Now in the normal interactive Python on a terminal, a command like
1+1
would print out the value, but using my method it does not. Using
print(1+1) works like a charm. However, when I try to have the returned object
printed using eg. 
PyObject_CallMethod(object,"print","") 
I get an error, "AttributeError: print". Can anybody tell me what I'm doing
wrong? Or how to get the result printed automatically like in interactive
Python?

Thanks!

Conrad
--



From bwinton at tor.dhs.org  Thu Jul 22 23:41:58 1999
From: bwinton at tor.dhs.org (Blake Winton)
Date: Fri, 23 Jul 1999 03:41:58 GMT
Subject: python + bash + new user
References: <3793BE52.43E023BB@netcard.com> 	<7n0f3d$1d7@dfw-ixnews10.ix.netcom.com> 	<3793C729.55DE42D0@netcard.com> 	<7n5nfq$j0$1@vvs.superst.iae.nl> <3797AE3B.687785E5@netcard.com> <3797C75B.2CFC9529@rubic.com>
Message-ID: <slrn7pfpl5.f9h.bwinton@tor.dhs.org>

On Fri, 23 Jul 1999 01:37:31 GMT, Jeff Bauer <jbauer at rubic.com> wrote:
>Joyce Stack wrote:
>> I did exactly as you said and chmod's etc but it just
>> says that it doesnt recognise "./simple.py" as a command.???
>> >         $ cat simple.py
>> >         #!/usr/bin/python
>Your python interpreter is not located in /usr/bin.  It may
>be in /usr/local/bin.  On many systems you can use the 'which'
>command to determine where the executable is.
>Alternatively, you can use the following line which works on
>many Unix systems:
>  #!/usr/bin/env python
>symlink-it-there-when-your-sysadmin's-not-looking-ly yr's,

So while we're on this topic, I seem to have hit a small snag...  I'm
using RedHat 6.something at home, and it ships with Python 1.5.1.  I've
also got an NT box at work where I'm using Python 1.5.2.  In the
interests of being able to move scripts from one box to the other, I've
installed Python 1.5.2 in /opt/{bin,lib/python1.5,man/man1}.  If I log
in myself, I can run /opt/bin/python, and get 1.5.2.  I can also create
a program with "#!/opt/bin/python", and if I type in "python spam.py" at
the command prompt, it works just fine.  Unfortunately, if I type in
"./spam.py", it doesn't do quite what I want it to.  I've traced my
problems down to sys.path (when it works, it points at /opt, when it
doesn't, it points at /usr), but I can't seem to figure out where it's
getting set, or how I can change it.  (I've tried setting $PYTHONHOME,
which works for me, but the whole reason I started this was to run cgi
scripts, and it's not working for the cgi scripts.)

So my question is this.  Can I get python working in such a way that if
it's being run from /opt, it will check /opt before /usr?  I'm not
averse to changing c files, and recompiling.

(Oh, yeah, the script is at http://tor.dhs.org/cgi-bin/daily.py  It just
grabs a bunch of comics, the weather in sunny Toronto, and the word of
the day, and puts them all up on one convienent page.  If any of the
people maintaining the script archives want a copy of it, I'ld be more
than happy to comment it, and send it off.)

Thanks,
Blake.

-- 
One Will.  One Dream.  One Truth.  One Destiny.  One Love.



From Thomas.S..Strinnhed at p98.f112.n480.z2.fidonet.org  Fri Jul  2 13:13:24 1999
From: Thomas.S..Strinnhed at p98.f112.n480.z2.fidonet.org (Thomas S. Strinnhed)
Date: Fri, 02 Jul 1999 18:13:24 +0100
Subject: Why does shelve make such large files?
Message-ID: <0000166d@bossar.com.pl>

From: "Thomas S. Strinnhed" <thstr at serop.abb.se>

Hi

Ovidiu Predescu wrote:
>[...]
> 
> The shelve module uses DBM which is like a small database that allows
> you to store objects and search it for a given key. DBM allows you to
> store millions of objects and search for them later without requiring
> you to load all the file in memory first.
> 
> The pickle module on the other hand is serializing the objects with the
> purpose of deserializing them _all_ from the file later. Pickle does not
> offer you any way to search for data based on a key, you have to do this
> yourself after the objects have been created from the file. This is
> opposed to the way shelve handles this, all the key accesses and
> insertions in a shelve object are actually reads or writes to or from
> the DBM file.
> 
> And to answer your question, DBM is creating these big files because of
> the way it manages the database. The data in the database file could
> have gaps as a result of multiple insertions and deletions. Pickle's
> data in files is a simple representation of the objects that were
> written and there is no way to update the file other than rewriting it
> entirely.
> 
> --
> Ovidiu Predescu <ovidiu at cup.hp.com>
> http://www.geocities.com/SiliconValley/Monitor/7464/

So, just to make shure I follow, in short terms 
 * pickle makes "persistant objects" flushed to a file 
 * shelve + dbm is in fact a (relational) database storing
   (arbitary**) objects in some hash-searchable order

(**) By arbitary I mean _any_ kind of object, or do they need to be
     of the same class?? (In order for searching to work)

About usage: shelves when I need to search and pickle when I just need
to save my objects?? Are there other considerations based on number and
size of objects?

Best regards
 -- Thomas S. Strinnhed, thstr at serop.abb.se



From kuncej at mail.conservation.state.mo.us  Tue Jul  6 11:20:07 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Tue, 6 Jul 1999 15:20:07 GMT
Subject: XML-RPC and firewall [solution]
Message-ID: <s781d84b.054@mail.conservation.state.mo.us>

>> ...I wrote a  urllib-based transport plugin for xmlrpclib.py...

>(I suppose I don't have to mention that I factored out the
>Transport stuff just to make it easier to do things like this...)

Yes! Thanks for the excellent foresight. Glad to make use of it.

>just one question: does this generate the right content-type?
>(the spec says text/xml ...

Darn! It looks like it's sending application/x-www-form-urlencoded.
The code is buried in the massive open_http() method of the 
urllib.URLopener class. I don't see any easy solution. I guess one
could subclass FancyURLopener and reimplement open_http()
with one line changed...

It bad content-type han't caused me any problems yet, though.

  --Jeff





From Robin.Becker at p98.f112.n480.z2.fidonet.org  Fri Jul  2 15:03:34 1999
From: Robin.Becker at p98.f112.n480.z2.fidonet.org (Robin Becker)
Date: Fri, 02 Jul 1999 20:03:34 +0100
Subject: ordered threading
Message-ID: <0000167f@bossar.com.pl>

From: Robin Becker <robin at jessikat.demon.co.uk>

I have an unbounded number of worker threads which occasionally require
access to a unique server thread. I can easily mutex access to the
server thread, but is there any obvious/easy way to ensure that the
blocked workers wake up in the right order?

Also if blocked can a thread be killed in any way?
-- 
Robin Becker



From amaranda at nospam.com  Thu Jul  1 07:12:56 1999
From: amaranda at nospam.com (Alex Maranda)
Date: Thu, 01 Jul 1999 12:12:56 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <slrn7nai1j.8ll.zeitlin@seth.lpthe.jussieu.fr> <64F6457BE0585BB3.1CF2F67BC8A680BD.A19A9FEE23BC9E58@lp.airnews.net>
Message-ID: <377B4D38.4CFC@nospam.com>

Mikael Lyngvig wrote:
 
> In contrast, the wxPython demo is so fast that you probably wouldn't
> realize it is implemented in Python, unless you're used to do GUI
> programming yourself.
I'll second that.

> The wxWindows performance problems must be more of the "this could be
> done better" than "this has to be done better" nature.
I'll second that.

> A very strong plus for wxWindows is that the Python API and the native
> C++ API are so close that it actually makes sense to use Python in the
> design and prototyping phase and then rewrite the Python code to C++,
> if you want to, when you've reached a stable design.
This is the way to go indeed. Of course only the performance sensitive
bits - there's no point in rewriting everything. Then put freeze to work
on the remaining python code, and here you go. I've actually done that
with the (sizable) wxPython demo on Linux (linking everything statically
- libpython, wxPython wrappers, wxGTK, GTK, GLIB, even stdc++; in the
end the only dynamic dependencies of the binary were libc, libdl,
libX11). I got a 4.5 Meg binary, from which 0.5 Meg was the demo
bytecode. 4 Meg of 'runtime' it's not much nowadays, and boy it's fast.
I'm not advocating this for casual use, I just explored packaging
possibilities for the average U*ix platform. The Windows version is
smaller 'cause GDI and the other libs are always there :)

Cheers,
-- 
Alex Maranda         mailto: amaranda at spider dot com
Spider Software Ltd. Tel: +44 (0)131 4757036
Edinburgh, UK        http://members.xoom.com/Alex_Maranda
STREAMS based communications protocols for embedded systems



From reason at shadow.net  Tue Jul 20 17:47:30 1999
From: reason at shadow.net (Vladik)
Date: Tue, 20 Jul 1999 17:47:30 -0400
Subject: [Q] In-Browser technology
References: <932337314.611.103@news.remarQ.com> <BE2245BA1DCDAE31.8804258C1F60F9A7.2EB4BCD57DBA05D1@lp.airnews.net>
Message-ID: <932522335.684.43@news.remarQ.com>

OK, I will try to clarify my question.
With the warning that I may not be speaking
authoritatively about the web technologies yet:
I am just starting this research.


See below.....


Cameron Laird <claird at starbase.neosoft.com> wrote in message
news:BE2245BA1DCDAE31.8804258C1F60F9A7.2EB4BCD57DBA05D1 at lp.airnews.net...
>
> In article <932337314.611.103 at news.remarQ.com>,
> Vladik <reason at shadow.net> wrote:
> >Hello,
> >Sorry for cross-posting, but
> >my question is really to the people on the news groups
> >who know about Eiffel, Python and Perl.
> >
> >Basically,
> >We are to develop a web client (a program that
> >can be ran witting a web browser) for our otherwise n-tier
> >cross-platform system, our research department gave the
> >recommendation and actually a prototype of the client
> >in MS ASP (active server pages) and, I think but not
> >sure, blended with Visual Basic. Talk about bias...
> >But this means that it can only run in MS IE and not
> >in Netscape, also it limits the number of platforms not only
> >for the client on which web browser runs, but also for
> >the web server (MS platforms only, because web server with
> >active pages can only run there)
> >There is some talk about Java, but our web client
> >will also be ran via Serial links (not T1 connections)
> >on relatively inexpensive systems -- therefore Java is
> >something to look into but it may be slow.
> >So I am set on a quest to find something
> >that
> >a) can be executed within IE or Netscape,
> >b) be faster and overall more resource friendly then JavaScript
technology
> >c) supports OO programming
> >d) does not depend on the platform on which web server is running.( if
the
> >web server is running NT or UNIX (sun, linux)
> >
> >I have found one thing that almost satisfies the requirements:
> >it is based on Oberon-2 programming language and so far sounds
> >VERY promising.
> >  The reason why I am still looking is because
> >there are more books and internet support for Python,Perl and Eiffel
> >then for Oberon (this will be a factor when presenting the proposal
> >to the management).  The other problem is that there is plugin
> >for this technology available for NT and Mac platforms (for both
> >IE and Netscape) and is not available for UNIX platforms (which means
that a
> >web browser can not be ran on UNIX). There is source code,
> >though.
> >If interested: http://caesar.ics.uci.edu/juice/
> >
> >So, finally, my question is is there a technology
> >I am looking for available for either Perl or Eiffel or Python
> .
> .
> .
> No.
>
> That's the short answer.  Slightly longer is this:  I
> don't understand--although I'm sufficiently intrigued
> to answer.
>
> Let's first dispose of the part of my reply where I
> pick nits from your description:
> 1.  "We are to develop a web client (a
>     program that can be ran witting a
>     web browser) ..." evokes to me a
>     plugin.  Is a plugin (supposing a
>     sufficiently potent one exists)
>     exactly what meets your require-
>     ments, or is there a point I'm
>     missing?
I do not know what it is supposed to be
a plugin or not --- but a technology that
allows to develop not-trivial GUI applications
that run withing a browser.

> 2.  "... a prototype of the client
>     in MS ASP (active server pages)
>     and, I think but not sure, blended
>     with Visual Basic ..."  I have no
>     definite idea of what you're say-
>     ing here.  Is it that your
>     customer has a picture of the
>     visual appearance of the desired
>     deliverable?  What is there about
>     the ASP prototype that does *not*
>     meet the requirements?  Is VB
>     blended on the server or client
>     side?  Does it matter?  Is it really
>     VB, or VBScript?

 I meant that VB script was used in the prototype, not VB

> 3.  Are you aware that *lots* of browsers
>     read pages served as ASP?  You can
>     use ASP without requiring IE.
NO I was not aware.  Thanks for pointing out
I will try to find more about netscape (the only
two browser we are to support are Netscape and IE).

> 4.  "... Java ... may be slow."  Lots of
>     things may be slow.  The market-lead-
>     ing browsers are definitely pigs, by
>     my standards, but they've already
>     been mandated, according to your
>     description.
>
>     Technical point:  Java applets work
>     well for me.  Invariably, when I hear
>     people complain about the size of
>     Java applets, they're really com-
>     plaining about ancillary images.
>     Java executable classes themselves
>     are admirably compact, in my experi-
>     ence.
OK, I will take this point into consideration.
I guess what I was looking for is something like
this:
"Here is a technology like Java (or different
but accomplises the same thing), and it is based on this
OO programing language and here is the page where you
can read more about that also includes technical comparasings
between this technology and JavaScript (or other)
"
> 5.  Of what resource is JavaScript not
>     sufficiently conservative?  Do you
>     seriously have performance problems
>     with it?  There are plenty of aspects
>     of JavaScript that deserve scorn,
>     but, unless I have a specific issue,
>     I happily use it to accomplish real
>     work on the client side.
I have never used JavaScript for development, for comparising Juice
applet gets to my computer faster and draws the wireframe
example faster.
The client app we will have to develop will have to
a) maintain a limited set of bussiness rules
b) apply a to relatively large sets of data from the database


> 6.  How much client-side programming are
>     you doing?  Why does object-orienta-
>     tion there matter to you?

Yes it does. Because the app will contain some bussiness
rules, some data access and data presentation layers.

> 7.  What technologies do you want that
>     *do* link client and server platform
>     compatibility?  I can't make this
>     constraint informative in the context
>     of everything else you've written.
Well, I guess what I was asking the ability to have
the server run on any platform and this web client
be executable withing a browser that runs on any platform.

> 8.  What's the part about Juice that you
>     like--is it the promise of superior
>     performance?  Do Franz and Kistler
>     truly make source available?  I
>     hadn't noticed that.  In any case, if
>     I thought it'd help with one of my
>     projects, I'd write them directly.
They make SRC available, I downloaded it,
the license does not seem to be restrictive, but I am
clear on that 100% yet.

>     I'm fond of Oberon, and I wish Juice
>     well.  I don't understand your inter-
>     est in it.
>
> It sounds as though you're involved in quite an ambi-
> tious project, of the sort I like.  I wish you well.
> I suspect you'll do well to clarify several of your
> requirements; this will make it much more inviting for
> comp.lang.* readers to help you.
> --
>
> Cameron Laird

Thanks,
Vladislav





From r.m.b.lindgren at usit.uio.no  Wed Jul 21 14:26:36 1999
From: r.m.b.lindgren at usit.uio.no (Rolf Marvin =?iso-8859-1?q?B=F8e?= Lindgren)
Date: 21 Jul 1999 20:26:36 +0200
Subject: python vs. lisp (whereis quote?)
In-Reply-To: Eugene Leitl's message of "Wed, 21 Jul 1999 11:00:47 -0700 (PDT)"
References: <x88r9m3cyje.fsf@rpppc2.hns.com> <lbzk8rva3to.fsf@morgoth.uio.no> <x88zp0rbd1x.fsf@rpppc2.hns.com> <mtlncbhv94.fsf@astron.berkeley.edu> <lbzogh635zl.fsf@morgoth.uio.no> <14230.2767.18575.943818@lrz.de>
Message-ID: <lbzvhbdeu5f.fsf@morgoth.uio.no>

[ Eugene Leitl

| Why? Because it's Scheme, or because it's Guile?

because too many of those involved in Guile don't understand the
language. 

-- 
Rolf Lindgren
roffe at tag.uio.no



From mal at lemburg.com  Thu Jul 22 09:12:47 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 22 Jul 1999 15:12:47 +0200
Subject: Pre-compiled Python binaries
References: <bIql3.4597$2D1.40180@newsr2.san.rr.com>
Message-ID: <379718CF.EE4BECBB@lemburg.com>

Bruce Duncan wrote:
> 
> Greetings,
> 
> My Internet Web hosting provider is using an unusual hardware/software
> configuration,
> namely Sun Solaris on a Pentium box.  Any chance of getting this in the
> pre-compiled
> Python archives?

Sure, here a whole bunch of them:

	http://starship.skyport.net/~lemburg/mxCGIPython.html

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   162 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From olli at rhein-zeitung.de  Sat Jul  3 14:14:02 1999
From: olli at rhein-zeitung.de (olli at rhein-zeitung.de)
Date: 3 Jul 1999 18:14:02 GMT
Subject: Curses, Ncurses, Slang, etc?
References: <19990703140217.D14374@moebius.dartmouth.edu>
Message-ID: <7lljta$bnj$1@news.rhein-zeitung.DE>

Hi,

I can only tell you that the ncurses module I wrote works for me on my RedHat 6 and Mandrake 6
systems. What problems do you have?

If support and so on is an issue, then I would use the slang bindings at the moment.

Bye, Oliver

Eric Kidd <eric.kidd at pobox.com> wrote:
> I'd like to write a Python application which uses either curses, ncurses,
> slang or a similar tty interface library. Unfortunately, the built-in
> curses module is hopeless broken on my RedHat 6.0 system (it doesn't export
> 'addch', for example).

> Neither the ncurses module or the pwin module work out of the box on my
> Linux system. Are there any well-supported Python modules for
> curses/ncurses/slang which do work out of the box?

> Thank you for any help you can provide!

> Cheers,
> Eric




-- 
Oliver Andrich, RZ-Online, Schlossstrasse Str. 42, D-56068 Koblenz
Telefon: +49-261-39210-40 Fax: -33 |       http://rhein-zeitung.de 
                                   |            http://andrich.net
GMC d- s++:++ a- C++ UL++++ P-- L+++ E--- W++ N+ w-- O- M-- 
V- PS++ PE Y+ PGP+ t 5+ X R+ tv+ b++ DI++ D- G e h- r z+



From cfelling at iae.nl  Sun Jul 11 18:19:56 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 12 Jul 1999 00:19:56 +0200
Subject: Extending ConfigParser
References: <3788BE09.26379078@starship.python.net>
Message-ID: <7mb5ac$i8m$1@vvs.superst.iae.nl>

In article <3788BE09.26379078 at starship.python.net> you wrote:
...
> So it seems, that class members whos names begin with "__" can't be
> overloaded? Is this right, or have I overseen something?

class members of class CLASS that are of the form __X get mangled into
_CLASS__X in class statements in class CLASS. aka private attributes.

so in your case do something like

class MyParser...:
	...
	old_ConfigParser__read = ConfigParser._ConfigParser__read
	def _ConfigParser__read(...

On the other hand, as ConfigParser so furiously tries to hide its read
method, I would think twice before using this trick:)

-- 
groetjes, carel
-- 
groetjes, carel



From Christian.Tismer at p98.f112.n480.z2.fidonet.org  Sat Jul  3 19:06:24 1999
From: Christian.Tismer at p98.f112.n480.z2.fidonet.org (Christian Tismer)
Date: Sun, 04 Jul 1999 00:06:24 +0100
Subject: Python suitable for a game engine?
Message-ID: <000016ba@bossar.com.pl>

From: Christian Tismer <tismer at appliedbiometrics.com>



Nick wrote:
> 
> The language doesn't necessarily need to support it. You could implement the
> threaded system yourself in Lua, you then have more control over what goes
> on. Proper threads are relatively slow in Windoze (I believe) and would be a
> nightmare to debug? If Lua wasn't fast enough you could write a scheduler
> with priority assessment in C/C++ and attached Lua script callbacks to this.

I'm not discussing what I can write in C/C++ but what LUA has.

> Personally I wouldn't dream of useing proper threads for game logic/AI I
> would write a basic scheduler, it wouldn't be difficult. Grim Fandango uses
> Lua as its scripting language and I think they did something along these
> lines. I think Balders Gate is another game that is rumoured to use it.

Which doesn't help me very much. I'd like to figure out
why I should use LUA instead of Python. Please give me real facts.

At the moment, I see no reason to switch, but a big
advantage using Python, since I can extend it extremely.
Please let me understand your point.

Still unsatisfied -ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From wblake at fau.campuscw.net  Tue Jul  6 23:03:44 1999
From: wblake at fau.campuscw.net (wil blake)
Date: Tue, 6 Jul 1999 23:03:44 -0400
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> <wkzp1g2t2e.fsf@ifi.uio.no> <slrn7nnfs3.19k.wtanksle@dolphin.openprojects.net> <9kgf3.301$F7.23516@news3.giganews.com>
Message-ID: <7lug3l$lra$1@news.campuscwix.net>

Christopher Browne wrote in message
<9kgf3.301$F7.23516 at news3.giganews.com>...

>Yes, please.  If there's relationship between Forth and Lisp, it does
>not lie in the Scheme direction.

>I don't remember seeing any indication of Moore having encountered
>Lisp; if so, that is very interesting.

In the Special Forth issue from Byte Magazine(August 1980?) an article ("The
Evolution of Forth, and Unusual Language"?) quoting Chuck Moore has a Forth
history that has one Forth code listing containing the word "atoms" and a
quote from Moore saying in effect "The Lisp influence is evident." I think
that the article may have originated from a Rochester Forth Conference
discussion.
-Wil Blake






From fredrik at pythonware.com  Fri Jul 23 08:37:11 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 23 Jul 1999 12:37:11 GMT
Subject: Extreme Newbie Question
References: <MPG.1202263db22c797c989681@news.globix.net>
Message-ID: <008201bed508$18bf7dd0$f29b12c2@secret.pythonware.com>

Robert Zurer <bob at bluestarscrap.com> wrote:
> It's probably ridiculously obvious but it's my first hour working with 
> Python...
> 
> In interactive mode how do you save the script that you've entered line 
> by line to a file?

well, how do you usually save batchfiles/shell scripts
you've entered line by line to a file? :-)

...

the Python FAQ basically says "Use an external editor";
see entries:

http://www.python.org/doc/FAQ.html#1.19 (unix)
http://www.python.org/doc/FAQ.html#7.10 (windows, mac)

depending on your platform.

however, it fails to mention that Python 1.5.2 comes with
the IDLE environment, which includes a python-aware editor,
a nifty interactive "shell" window, and lots of other goodies:

http://www.python.org/doc/howto/idle/

well worth trying.

...

btw, most platforms come with some kind of trivial text
editor (e.g. Notepad on Windows), and supports cut and
paste from console windows, so you can always save stuff
you've written into Python's basic command-line interface
window...

</F>




From mstenber at cc.Helsinki.FI  Wed Jul 28 09:19:35 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 28 Jul 1999 16:19:35 +0300
Subject: HPUX and threads?
Message-ID: <al8yag0x66w.fsf@sirppi.helsinki.fi>

Is HPUX-11 thread support tested/implemented/whatnot? 

This at least does not look promising..

# /opt/python/bin/python
pthread_mutex_init: Invalid argument
Memory fault(coredump)
 
(also tests failed with the same message, obviously)

Python without threads seems to work fine.

-- 
   "I am Grey. I stand between the candle and the star.
  We are Grey. We stand between the darkness and the light."
			- B5



From mal at lemburg.com  Fri Jul 16 04:11:35 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 16 Jul 1999 10:11:35 +0200
Subject: mxODBC and MS SQL Server
Message-ID: <378EE937.5D0E5801@lemburg.com>

Just thought, I'd inform you of an inconvenience that occurs
when using mxODBC to access MS SQL Server.

The ODBC driver of SQL Server does not LONG columns to appear
anywhere but the end of a SELECT columns list, e.g.
SELECT anInteger, aString, aLONG FROM ... will work, but
SELECT aString, aLONG, anInteger FROM ... won't.

For now the work-around is simple: always put LONG columns at the
end of the SELECT column list.

The next version of mxODBC will include a fix for this.

BTW, using the EasySoft ODBC bridge, you can now even access
MS SQL Server running on a WinNT from any Unix machine, e.g.
your web-server or Python development machine.

More infos on mxODBC and all of the above can be found at:

	http://starship.skyport.net/~lemburg/mxODBC.html

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   168 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From gmcm at hypernet.com  Sun Jul  4 08:43:49 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 4 Jul 1999 07:43:49 -0500
Subject: ordered threading
In-Reply-To: <Y2tnuCAVprf3EwNj@jessikat.demon.co.uk>
Message-ID: <1281033720-53905059@hypernet.com>

Robin Becker writes:

> In article <000a01bec583$30332b40$769e2299 at tim>, Tim Peters
> <tim_one at email.msn.com> writes

> >[Robin Becker]
> >> I have an unbounded number of worker threads which occasionally require
> >> access to a unique server thread. I can easily mutex access to the
> >> server thread, but is there any obvious/easy way to ensure that the
> >> blocked workers wake up in the right order?
> >
> >No threading system on earth allows specifying the order in which blocked
> >threads will wake up, although some systems (not Python's) allow you to
> >*influence* the order via assorted "priority" gimmicks.

> the workers are HTTP requests in separate threads from Zope I guess.
> I want to serialise access to the time/space hog. I guess it could
> run uniquely in the worker threads ie the locked resource would be
> access to the hog code. I'm not sure that would let me guarantee
> unlocking as there are ways for the worker threads to suicide.

If that's what you want, you don't really need to order the 
unblocking. Just give each thread an input queue protected by a 
Condition. Worker threads put requests on the Hog thread's input 
queue (including a ref to the return queue). If you find that some 
worker threads seem to get better access to the Hog's queue than 
others, a time.sleep(0.001) will probably fix it - since it will 
pretty much force a context switch.


- Gordon



From Gaetan_Corneau at baan.com  Thu Jul 22 03:54:26 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Thu, 22 Jul 1999 07:54:26 GMT
Subject: Embedding question
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF4EF@ex-quebec-u1.baan.com>

Hi,

I have never embedded Python in a program directly (always have used ActiveX
Scripting) and I have a simple question: 
can Python be put in a Windows NT program's DLL? The "main" program should
know nothing about the Python interpreter. it is a program that is almost
entirely made of "plug ins".

Thanks,
______________________________________________________
   Gaetan Corneau
   Software Developer (System integration Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From poinot at onera.fr  Mon Jul 26 10:18:39 1999
From: poinot at onera.fr (Marc POINOT)
Date: Mon, 26 Jul 1999 16:18:39 +0200
Subject: Configure and cross-compiler
Message-ID: <379C6E3E.78CEB6DD@onera.fr>

Hi,

I've compiled Python for a  NEC super-computer,
using a cross compiler.

The configure script failed for the tests, it
leaves with a exit 1 emergency-escape-do-not-use-elevators.
Maybe a slight modification can allow
to produce Python with less verification, at least
for cross-compilation.

I've replaced the corresponding exit 1 with "true".

Marcvs [alias Yes, don't leave, I have the truth...]






From chad at vision.arc.nasa.gov  Thu Jul 29 22:12:08 1999
From: chad at vision.arc.nasa.gov (Chad Netzer)
Date: Thu, 29 Jul 1999 19:12:08 -0700
Subject: Problem compiling Python on Suse 6.1
References: <37A01B2B.2CF78A89@gmx.net>
Message-ID: <37A109F8.F5FEDBA0@vision.arc.nasa.gov>

Tomaz Ficko wrote:

> I have a problem compiling Python 1.5.2 with Tk and readline support.

Can you post the section of your Modules/Setup file that involves Tkinter
and readline?  In which directory are your libtcl and libtk files (and are
they
in the same directory)?  And where is readline?

Chad





From BgPorter at NOacmSPAM.org  Wed Jul 28 16:39:02 1999
From: BgPorter at NOacmSPAM.org (Brett g Porter)
Date: Wed, 28 Jul 1999 20:39:02 GMT
Subject: a little motivation, anyone?
References: <19990728052704.10608.00002260@ng-fq1.aol.com> <MBJn3.859$Yk3.6771@newsr1.twcny.rr.com>
Message-ID: <GTJn3.79756$eF3.24937@news.rdc1.nj.home.com>

Darrell <news at dorb.com> wrote in message
news:MBJn3.859$Yk3.6771 at newsr1.twcny.rr.com...
> Writing code is hard.
> It's hard for everyone, I think :)
> Writing code is fun.
> It's not fun for everyone.
> If it's fun for you then no doubt you are cut out to be a coder:
> Else maybe coding isn't for you.
>
> Hint #1: Look for and steal patterns of how to do things from
>     the code of others. A pattern can be as simple as how
>     to handle command line switches.
>
> Hint #2: Write lots of code.
>
> Hint #3: Read lots of code.

Hint #4: Read lots of books about writing code. Some of my faves:
'The Practice of Programming', Kernighan & Pike
'Code Complete', Steve McConnell
'The Zen of Code Optimization', Michael Abrash








From nick.bower at ssec.wisc.edu  Sat Jul 10 19:35:54 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sat, 10 Jul 1999 18:35:54 -0500
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu>
Message-ID: <3787D8DA.3774AA7D@ssec.wisc.edu>

thanks for the help all.  exactly what i thought would have existed.

Incidentally, I went back to the library reference and still couldn't find
os.chdir

cheers, nick.


Nick Bower wrote:

> I feel really dumb asking this question, but how do I change the current
> working directory?  For browsing the file system I mean.
>
> I've found exec commands, but these would surely be device dependent and
> there doesn't seem to be anything obvious in the os or os.path
> documentation.  i looked through the tutorials too.
>
> thanks in advance, nick

--

Nick Bower
Space Science and Engineering Center
University of Wisconsin, Madison, USA
Phone: (608) 265 8007
Email: nick.bower at ssec.wisc.edu
Web: http://arm1.ssec.wisc.edu/~nickb





From joakal at my-deja.com  Thu Jul  1 23:11:36 1999
From: joakal at my-deja.com (joakal at my-deja.com)
Date: Fri, 02 Jul 1999 04:11:36 +0100
Subject: Tkinter just WON'T work right!
Message-ID: <00001650@bossar.com.pl>

From: joakal at my-deja.com

Sorry to be so inspecific about my subject =) Also, I'm sure this
question is asked frequently, but I couldn't find what I was looking for
in any of the 1400 or so messages here (too many of them start with
``Help!'' or ``Python Newbie''). Anyway, here it is:

I tried to use the Grail browser under Slackware 4.0 and Python 1.5.1. I
got the error saying it's not configured for Tkinter, so I went into my
/usr/lib/python1.5/config/Setup file and uncommented the appropriate
stuff.

I tried again, same problem. I read a bit more, and it said I have to
rebuild Python to add Tkinter (don't remember having to do that before).
So, argh, I tried to do that. Now I get an error when I use 'make'
saying there's an error when it gets to sedscript.

Can anyone help?
Thanks in advance!


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From graham at sloth.math.uga.edu  Thu Jul 15 10:39:57 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 15 Jul 1999 14:39:57 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim>
Message-ID: <7mkrrt$a81$1@cronkite.cc.uga.edu>

[Neel Krishnaswami]
: > If we could compose generators just like functions, then the bookkeeping
: > can be abstracted away and the same approach will work for arbitrarily
: > complicated parsing tasks.
Tim Peters (tim_one at email.msn.com) wrote:
: The latter is one of the arguments Griswold gave in the paper that
: introduced Icon, contrasting Icon's uniform approach to SNOBOL4's distinct
: pattern and procedural sublanguages.
: 
: I don't think he'd make the same argument today!  Icon is an idiosyncratic
: delight, but most string pattern matching was in fact easier (to write,
: modify, or read) in SNOBOL4.  Once you start using Icon for complex pattern
: matching, you'll soon discover that it's very hard to go back to old code
: and disentangle "the pattern" from "the processing" -- *because* everything
: looks the same, and it's all intermixed.  The distinct sublanguages in
: SNOBOL4 enforced a clean separation; and that was more often an aid than a
: burden.

As far as I understand this thread I think Neel is right, and Tim and
Griswold are wrong. I have been recently doing a lot of programming with
parser combinators in Clean and Haskell. They are something like what
Neel is talking about (as the name "parser combinators" suggests you
write small "string parsers" as functions, and then have "parser
combinators" that stitch these together). Note that in this parser
combinator approach there is no distinction between the code for "the
pattern" and the code for "the processing" -- they are all written in
the same language (Clean/Haskell in my case). At first I found the
idea of having the two types of code (the pattern and the processing)
in the same language (rather than in 2 distinct languages as in the
traditional way) a bit difficult. But I perservered and I am glad I
did. I found that once you reach a certain level of familiarity with
combinators you can write new pattern and processing code very
quickly and very reliably. The trick is to get the combinators right.

graham
-- 
              This ain't no technological breakdown
                 Oh no, this is the road to hell



From fdrake at cnri.reston.va.us  Tue Jul  6 11:32:51 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Tue, 6 Jul 1999 15:32:51 GMT
Subject: popen2 on Windows
In-Reply-To: <4D0A23B3F74DD111ACCD00805F31D8100DB90EC1@RED-MSG-50>
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90EC1@RED-MSG-50>
Message-ID: <14210.8611.982259.413266@weyr.cnri.reston.va.us>

From: John van der Koijk [mailto:vdkoijk at linux01.prc.tno.nl]
 > This is amazing, since the manual for 1.5.2 says:
 > 
 > 8.13 popen2 -- Subprocesses with accessible standard I/O streams
 > 
 >    Availability: Unix, Windows.

Bill Tutt writes:
 > Well, its wrong. :) Windows doesn't have fork().

  This was also fixed in the recent release of the documentation!  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From R.Hooft at EuroMail.com  Thu Jul  1 04:30:24 1999
From: R.Hooft at EuroMail.com (Rob Hooft)
Date: 01 Jul 1999 10:30:24 +0200
Subject: Tkinter: why does an empty frame not shrink?
References: <19990625162938.39336@nms.otc.com.au>
Message-ID: <m3d7yc6acf.fsf@devel.no.rwd.dinet>

>>>>> "GM" == Greg McFarlane <gregm at iname.com> writes:

 GM> If you have several child widgets packed in a frame and you
 GM> unpack the children one by one, the frame will gradually get
 GM> smaller, except when the last child is unpacked, in which case
 GM> the frame does not resize.  Does anyone know why the frame does
 GM> not shrink to 0x0 when you unpack the last child?  Any soluton?

In one of my applications, I put the frame inside another frame, and
resize the empty container frame to 1x1 manually to solve this. Ugly,
but it works:

    def makeWidgets(self,fn=None):
        self.rightbarvar=variables.booleanvariable() # used in a menu entry
        self.rightbarvar.set(1)
        ....
        self.rightbarc=Tkinter.Frame(self)
        self.rightbarc.pack(expand=Tkinter.NO,fill=Tkinter.Y,side=Tkinter.RIGHT)
        self.rightbar=Tkinter.Frame(self.rightbarc)
        self.rightbar.pack(expand=Tkinter.NO,fill=Tkinter.Y,side=Tkinter.RIGHT)
        ....

    def togglerightbar(self):
        if self.rightbarvar.get():
            self.rightbar.pack(expand=Tkinter.NO,fill=Tkinter.Y,side=Tkinter.RIGHT)
        else:
            self.rightbar.pack_forget()
        self.rightbarc.configure(width=1,height=1)

Regards,

Rob Hooft
-- 
=====   R.Hooft at EuroMail.net   http://www.xs4all.nl/~hooft/rob/  =====
=====   R&D, Nonius BV, Delft  http://www.nonius.nl/             =====
===== PGPid 0xFA19277D ========================== Use Linux! =========



From John.Hirsch at p98.f112.n480.z2.fidonet.org  Thu Jul  1 06:19:00 1999
From: John.Hirsch at p98.f112.n480.z2.fidonet.org (John Hirsch)
Date: Thu, 01 Jul 1999 11:19:00 +0100
Subject: Python in the public eye !
Message-ID: <000015f8@bossar.com.pl>

From: John Hirsch <johnh at phm.gov.au>


Thanks to Drew and the Crew at the ANU :) who created a fantastic exhibit   
which has been very popular with all the visitors.
This is not the first exhibit at the Powerhouse Museum that has been   
developed using Python. Mainly due to the programming prowess of one Tim   
Docker (Ex engineer at the Museum) who is now on a mission to convert   
fellow programmers at his new job.

>Anyone who visits the Powerhouse Museum (Sydney, Australia) might be
>interested to know that the "Wedge" virtual environment exhibit in the
>new "Universal Machine" permanent exhibition which opened last night
>is Python powered.  The whole program for the exhibit was put together
>over the last fortnight on top of our in house "pSpace" VE system
>which use python as it's embedded language. I got to watch primary
>school kids using it for the first time today and it was a real
>buzzzz. Thanks Guido and Co !!!


 ----------------------------------------
Jonathan Hirsch
Powerhouse Museum
Sydney, Australia
E-mail : johnh at phm.gov.au
Phone: (02) 92170233
 ----------------------------------------  




From fredrik at pythonware.com  Mon Jul 26 22:07:08 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 27 Jul 1999 02:07:08 GMT
Subject: Newbie question
References: <379d0ee6@news.gci.net>
Message-ID: <001501bed7d4$bc4723d0$f29b12c2@secret.pythonware.com>

Arnaldo <javanet at sinbad.net> wrote:
> I'm just trying out some examples from the book Learning Python and I run
> into a exercise that is not working for me.
> It's on page 95 , exercise 4b.
> This is the answer the book has
> ===========================================
> l = [1, 2, 4, 8, 16, 32, 64]
> x = 5
> i = 0
> while i < len(l):
>     if 2 ** x == l[i]:
>         print 'at index', i
>         break
>         i = i + 1
> else:
>      print x , 'not found'
> ============================================

either you or the LP authors have messed up;
for the loop to ever terminate, the "i = i + 1"
statement should at the "while" level, not in-
side the "if" statement.

try changing the loop to:

    ...
    while i < len(l):
        if 2 ** x == l[i]:
            print 'at index', i
            break
        i = i + 1
    ...

</F>




From nick at NOSPAMvideosystem.co.uk  Thu Jul  1 15:28:32 1999
From: nick at NOSPAMvideosystem.co.uk (Nick)
Date: Thu, 1 Jul 1999 20:28:32 +0100
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain> <377B7E17.5BD7@nospam.com>
Message-ID: <377bbf9b@news.xtml.co.uk>

Don't want to put the cat amongst the pigeons but Lua
(http://www.tecgraf.puc-rio.br/lua/), is a language specifically designed to
be embedded and it runs about 50-100% faster than Python and is a lot more
compact (30k or something). It has a lot of similarities syntactically as
far as I can see.

 If I was to use an embedded scripting language in a game I would use Lua
not Python. If I was to write tools for the game I would use Python.

Cheers,
Nick.



Alex Maranda <amaranda at nospam.com> wrote in message
news:377B7E17.5BD7 at nospam.com...
>> their leader is killed, etc. The events would be things such as
>> "create", "monster sees player", "clock tick", etc. All the functions
>> in the script would be callbacks written in C++, to manipulate the
>> monster to do things such as "patrol area", "play an animation",
>> "sleep for X seconds", etc. In other words, the scripts would be glue,
>> not the main thread of execution.
>you use confusing terms; I understand what you're saying, but you
>shouldn't use the word 'thread' for anything else than its OS meaning.
>You could say the 'the bulk of processing'. You might very well end up
>using real Python threads; don't forget in this case to release/reaquire
>the global interpreter lock in your C++ callbacks, if you make blocking
>syscalls (unlikely in a game).
>
>>
>> I'm been using Python for a couple of years now, and it is my favorite
>> scripting language. But, I don't have much experience with embedding
>> it in a another program, and I don't know enough about the speed
>> issues. So, my questions are...
>>
>> - Will Python be fast enough? Even with possibly 100 or 200 active
>>   items on a level?
>Unlikely. You won't know until you try. To my knowledge UnrealScript is
>eventually translated to C++.
>
>> - Especially when all of the scripts will be live, and usually in the
>>   middle of a callback?
>> - Should I use the interpreter in the library, or should I take an
>>   approach of crunching the byte-codes myself?
>??
>
>Cheers,
>--
>Alex Maranda         mailto: amaranda at spider dot com
>Spider Software Ltd. Tel: +44 (0)131 4757036
>Edinburgh, UK        http://members.xoom.com/Alex_Maranda
>STREAMS based communications protocols for embedded systems





From Tim.Peters at p98.f112.n480.z2.fidonet.org  Sat Jul  3 06:37:56 1999
From: Tim.Peters at p98.f112.n480.z2.fidonet.org (Tim Peters)
Date: Sat, 03 Jul 1999 11:37:56 +0100
Subject: Upgrade or not ??
Message-ID: <000016af@bossar.com.pl>

From: "Tim Peters" <tim_one at email.msn.com>

[Pierrette Serrand]
> First a general question :
> Actually we're working with Python 1.5.1 and we're satisfied with it.
> Should we upgrade to 1.5.2

Of course.

> and if yes WHY?

Carrots and sticks.  And why wouldn't you?  It's free!

More than a year passed between 1.5.1 and 1.5.2, spanning hundreds of little
improvements and bugfixes, along with a few major ones.  So, e.g., if you
find a bug in 1.5.1, the response will be (http://www.python.org/1.5/):

    Forget about reporting a bug until you've confirmed that the bug
    still occurs in the 1.5.2 release

IOW, 1.5.1 is no longer supported.  And while the dozen <wink> lines of code
you've already written for 1.5.1 are very likely to work fine under 1.5.2,
as time goes on more & more new useful modules and code snippets are likely
to require new-in-1.5.2 features.

> No a Tkinter guestion :

Not a Tk guy, so I'll skip to the end:

> ...
> - Not write in python ? In this case we lose the work done for the
> prototype

If you can get a prototype done in Python 10x faster than in anything else,
and get the first crucial rounds of redesign done at that speed too, you win
big overall even if you do end up recoding the refined design in some other
language.

IOW, be realistic here:  you lose the work done for a prototype no matter
what language you write it in.  That's why they call it a prototype instead
of a product <wink>.

would-you-rather-throw-away-500-lines-or-5000?-ly y'rs  - tim






From andrew at andrewcooke.free-online.co.uk  Thu Jul  1 08:16:00 1999
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Thu, 01 Jul 1999 12:16:00 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <slrn7nai1j.8ll.zeitlin@seth.lpthe.jussieu.fr> <64F6457BE0585BB3.1CF2F67BC8A680BD.A19A9FEE23BC9E58@lp.airnews.net>
Message-ID: <7lfm5q$bra$1@nnrp1.deja.com>

In article
<64F6457BE0585BB3.1CF2F67BC8A680BD.A19A9FEE23BC9E58 at lp.airnews.net>,
  mikael at pobox.com (Mikael Lyngvig) wrote:
> On 26 Jun 1999 21:37:12 GMT, zeitlin at seth.lpthe.jussieu.fr (Vadim
> Zeitlin) wrote:
>
> > While I don't think wxPython can replace TkInter as a de-facto GUI
standard
> >for Python I still think that it has several advantages compared to
TkInter.
> >Instead of saying how much better, prettier, faster (this one has
really made
> >me smile, BTW: I don't think wxWin is currently very fast, if it has
been
> >cited as its advantage it must really mean that TkInter is extremely
slow.
>
> Well, TkInter is extremely slow, but not because of the Python part.
> Even the Tk widget tour, included with Tk, is so slow that you almost
> think "Please, draw the next character - you can do it." whenever a
> window with some text needs to be redrawn.
>
> In contrast, the wxPython demo is so fast that you probably wouldn't
> realize it is implemented in Python, unless you're used to do GUI
> programming yourself.
>
> One important point is that I probably experience Tk(Inter) as slow
> because I'm used to Win32 desktops with very fast graphics hardware
> and no X/Windows style client/server architecture.  I imagine that
> many X/Windows users experience Tk(Inter) as fast enough.

I'm confused.  I'm sitting at a moderate NT box (266MHz, 96Mb) - my
work machine, and running the Tk demos.  There is no problem at all -
they are instantaneous.

Is this problem to do with memory?  Or is it a W95 thing?  If so,
is it due to thread scheduling?

Andrew



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From MHammond at skippinet.com.au  Tue Jul  6 08:15:12 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Tue, 6 Jul 1999 22:15:12 +1000
Subject: COM with Python?
References: <al8wvwe6z80.fsf@myntti.helsinki.fi>
Message-ID: <7lsrsp$fh2$1@m2.c2.telstra-mm.net.au>

Markus Stenberg wrote in message ...
>Is there some form of _somewhat_ more in-depth, but less abstract source of
>information for COM+Python than the 'quick client/server'/the slides
>respectively.

Nope.

>Spesifically, I'm looking for information on activation policies; how to
>implement idispatch objects that are singletons for a single machine.

Your best bet is DejaNews - the singleton subject was done fairly thourghly
a while ago, and no significan changes have been made to the COM package to
make it any eaiser (alas)

Mark.








From rfreedm1 at nospam.rochester.rr.com  Sun Jul 25 15:24:24 1999
From: rfreedm1 at nospam.rochester.rr.com (Benjamin Schollnick)
Date: Sun, 25 Jul 1999 19:24:24 GMT
Subject: Anti-Port scan.... (Port Monitor / Logger)
Message-ID: <rNHvjEdhm5Pp-pn2-vyeUbHbYW9pR@d185d18f8.rochester.rr.com>

Folks,

	I know this is a fairly simple program, but this is just a thought 
right now.

I'm having some activity on my modem that I can't explain, but I run 
OS/2....
So I can't use NUKENABBER, etc, for Windows to monitor the ports on my
"server" box.

I was thinking, it'd be easy to write a python program to grab a port,
and
log the access's.....

But then I was thinking, but how would I grab all the ports 
conviently, and
not have to write a billizion sock call statements....

Anyone got a suggestion on how to do this....I would also probably 
want to
make it selective, for example, so that I can have my web server still
operational
while this is online....

I'm stumped, but of course, I just got done with moving, and need to 
finish unpacking,
etc....

			Benjamin



From phd at emerald.netskate.ru  Tue Jul  6 12:27:31 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Tue, 6 Jul 1999 20:27:31 +0400 (MSD)
Subject: AttributeError - a dumb question
In-Reply-To: <Pine.LNX.3.96.990706182058.780A-100000@Johann>
Message-ID: <Pine.LNX.4.04.9907062026150.23992-100000@emerald.netskate.ru>

Hi!

   Python is case-sensitive language.

On Tue, 6 Jul 1999, Johann Spies wrote:
> class F:
>     def __init__(self):
> 
>         self.x = self.Krylys()
                           ^
                           lowercase

>     def KryLys(self):
             ^
            uppercase

>         l = [1,2,3,4,5]
>         return l
>     
> 
> l = F()
> 
> Now this assignment fails:
> 
> Traceback (innermost last):
>   File "<stdin>", line 11, in ?
>   File "<stdin>", line 4, in __init__
> AttributeError: Krylys
> 
> I hope you can help me with this one.
> 
> Johann
> 
>  --------------------------------------------------------------------------
> | Johann Spies                                 Windsorlaan 19              |
> | jhspies at futurenet.co.za                3201 Pietermaritzburg             |
> | Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
>  --------------------------------------------------------------------------
> 
>      "Ask, and it shall be given you; seek, and ye shall
>       find; knock, and it shall be opened unto you."        
>                                   Matthew 7:7 

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From tim_one at email.msn.com  Thu Jul 22 20:10:19 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 22 Jul 1999 20:10:19 -0400
Subject: int(long(-sys.maxint-1)) fails on Linux
In-Reply-To: <3796DC4F.224795DA@vision.arc.nasa.gov>
Message-ID: <000001bed49f$c052e820$2c2d2399@tim>

[Chad Netzer]
> Compiling python 1.5.2 on Linux Mandrake 6.0,  the "make test" step fails
> because it gets an OverflowError during test_types.py on the
> following line:
>
> int(long(-sys.maxint-1))
>
> test_long.py fails also for the same reason.

Chad, let's break this down to see where the overflow is coming from.  Try
this line by line and post your results:

>>> import sys
>>> sys.maxint
2147483647
>>> -sys.maxint
-2147483647
>>> -sys.maxint-1
-2147483648
>>> long(-sys.maxint-1)
-2147483648L
>>> int(long(-sys.maxint-1))
-2147483648
>>>

Neither the OS nor the libraries should have anything to do with this.  You
didn't mention the important part, though <wink>:  what kind of HW is this
running on?  If it has a "funny" wordsize, and limits.h is missing or
foolish, Python may have a bogus idea of what the largest int is (which is
why I want to see what sys.maxint returns for you).

boy-that-linux-is-unstable<wink>-ly y'rs  - tim





From MHammond at skippinet.com.au  Tue Jul 13 19:06:10 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Wed, 14 Jul 1999 09:06:10 +1000
Subject: making a Windows NT service out of a python program
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F20@RED-MSG-50> <378B2AA8.58477120@appliedbiometrics.com> <378B2C50.B7C2134A@cern.ch> <7mfot5$n30$1@nnrp1.deja.com>
Message-ID: <7mgglt$akv$1@m2.c2.telstra-mm.net.au>

Andrew Cooke wrote in message <7mfot5$n30$1 at nnrp1.deja.com>...
>In article <378B2C50.B7C2134A at cern.ch>,
>  "Haimo G. Zobernig" <Haimo.Zobernig at cern.ch> wrote:


>They are useful because you can specify an account for them to run in
>(assuming you have sufficient priveliges), set whether they should be
>started automatically, and let them access the user's screen.

Actually, "let them access the user's screen" is not technically correct.
The term they use is "desktop", and for good reason.  The desktop the
service has access to may _not_ be the user's desktop!!  If anyone is
interested, I can point to you a magazine with a very good series on NT
Services and how they interact with NT.

>to stop, start, pause them etc.)  Writing programs to install/start
>services is more complicated, but hopefully you won't need to do that
>(the Control panel can stop and start them, and I assume the Python
>package others mentioned has an installer).  If you poke around in the

Indeed it does.  Installing a Python service is trivial.  The Python
interface also allows you to specify the account, desktop interaction
options, startup options, etc, just as you would expect (also as you would
expect of Python, this is command-line only :-)  Then again, Python _can_
display the standard NT services dialog!

>The worst thing about them is that they can be tricky to debug.  If they
>are running in the system account then they have a restricted
>environment and the standard output channels send useful data to

he he - not with Python :-)  Just import win32traceutil at the top of your
service, and you can see all your print statements in Pythonwin :-)

>All this is from C and Java knowledge, I've never written a Python
>service, so I can't give more specific help.

You should spend an hour having a play - hopefully you will be impressed :-)
Ive never played with Java's services, but Python is light-years ahead of
C/C++ in terms of ease in creating services.

Mark.





From joeld at anisci.com  Tue Jul 13 16:52:30 1999
From: joeld at anisci.com (Joel Davis)
Date: 13 Jul 1999 13:52:30 PDT
Subject: Wrapping C++ w/inheritance
References: <378BA78E.B6C19463@anisci.com>
Message-ID: <378BA879.DFD8BC0D@anisci.com>

oops.. that should be
http://www.flash.net/~joeldav/pywrap/

--
----------------------------------------------------------------
                                                joeld at anisci.com





From peter.stoehr at weihenstephan.org  Thu Jul 22 15:11:44 1999
From: peter.stoehr at weihenstephan.org (Dr. Peter Stoehr)
Date: Thu, 22 Jul 1999 21:11:44 +0200
Subject: cin
References: <3srl3.1302$Gj8.191985152@newsa.telia.net> <cCtl3.818$ui4.256788@news.shore.net>
Message-ID: <37976CF0.3DC02CBA@weihenstephan.org>

Hi everyone,

"Michael P. Reilly" wrote:
> 
> Joacim L*wgren <fire at netlimit.com> wrote:
> : Hi,
> 
> : I'm new to Python and I'm wondering what the C++ "cin" (or C
> : "scanf()")equivalent in Python is. Doesn't seem there's a quick function for
> : this.
> 
> Actually it is not a function, it is an operation (%).
> 
> >>> name, age = 'Michael', 30
> >>> s = 'My name is %s and I am %d years old.'
> >>> print s % (name, age)
> My name is Michael and I am 30 years old.
> 
> Take a look at the Library Reference Manual, Section 2.1.5.1 "More String
> Operations" (http://www.python.org/doc/current/lib/typesseq-strings.html).
> 
>   -Arcege

this emulates the C-function sprintf and not scanf.
What Joacim is looking for is something like raw_input :-)

# -----------------------------------------
aString = raw_input("Give me input :")
print aString
# -----------------------------------------

Greetings from Erding/Bavaria/Germany/Europe/Earth/Universe
    Peter	
-- 
---------------------------------------------------------------------------
   Dr. Peter Stoehr --- Teisenbergweg 6 --- 85435 Erding --- 08122/47232
---------------------------------------------------------------------------
		I'm the terror that flaps in the night



From kranio at nospam.nospam.it  Tue Jul 27 05:51:40 1999
From: kranio at nospam.nospam.it (Kranio)
Date: Tue, 27 Jul 1999 09:51:40 GMT
Subject: Python and Red Hat Linux 6
References: <379CD547.22A10A50@postoffice.pacbell.net> <379CD5C2.5BA096A2@mail.danbbs.dk>
Message-ID: <379d7fcf.52695@news.tiscalinet.it>

On Mon, 26 Jul 1999 21:40:18 GMT, Brian Pedersen
<brian.pedersen at mail.danbbs.dk> wrote:
>Try 
>
>!#/usr/bin/env python 
>
>instead (don't ask me why, i read it in a FAQ somewhere :-)
>--                                                             
>Brian Pedersen, M.Sc.E.E.  

I have the same problem and this solution won't works. If you type

$ file.py
bash: ./file.py : Command not found

is a normal behaviour. But the correct syntax should be

$ ./file.py

but the output is

bash: file.py : Permission denied

Any good solution?

Marco 



From Thomas.Bryan at p98.f112.n480.z2.fidonet.org  Sat Jul  3 01:34:19 1999
From: Thomas.Bryan at p98.f112.n480.z2.fidonet.org (Thomas Bryan)
Date: Sat, 03 Jul 1999 06:34:19 +0100
Subject: Problem with informixdb and python1.5.2
Message-ID: <000016aa@bossar.com.pl>

From: Thomas Bryan <tbryan at arlut.utexas.edu>

Ken Liu wrote:
> 
> I've been using the informixdb module (version 1.2a1) with python1.5.1 for a
> while now with no problems.  Just a few days ago we installed python1.5.2,
> and now informixdb won't work any more.  when I attempt to construct a
> Connection object, I get an error about "unable to get name of current
> working directory."  Here's a log of my interaction with the interpreter

Do you now have both python 1.5.1 and python1.5.2 installed?

> which does not make any sense.  Our Informix setup has not changed for the
> last few months, and the informixdb modules continues to work great 
> under python1.5.1.
> 
> Does anyone have any suggestions on what might be wrong?

informixdb might be getting confused about which Python 
version/libraries it should be using.  I'm not sure how to 
determine whether that's the problem, but I know that others
have had problems with various packages when they have two 
versions of Python installed at the same time.

Have you tried recompiling informixdb with the 1.5.2 header files?  
I don't think that too much has changed between 1.5.1 and 1.5.2, but 
it couldn't hurt to try.



From ddainese97x at x18dsi.unive.it  Mon Jul 26 05:00:19 1999
From: ddainese97x at x18dsi.unive.it (Diego Dainese)
Date: 26 Jul 1999 09:00:19 GMT
Subject: Access dictionary in order?
References: <roy-2507991300370001@mc-as01-p61.med.nyu.edu> <14235.18248.924848.52652@dolphin.mojam.com>
Message-ID: <slrn7po8t3.6k.diego@blanka2.blankanet.it>

On Sun, 25 Jul 1999 12:22:03 -0500 (CDT), Skip Montanaro wrote:
> You demonstrated the best way.  There is no need to save the key list unless
> the dictionary is very large though.  Just grab 'em and sort 'em when you
> need:
> 
>     keylist = dict.keys()
>     keylist.sort()
>     for key in keylist: do_stuff(dict[key])
> 

But how much time requires the keylist generation?

That is, the keylist is made traversing all the dictionary nodes or is
cached inside it?

-- 
Diego Dainese
--
To reply remove the numbers and the `x' from my address
--
Sorry for my bad English!



From fredrik at pythonware.com  Tue Jul 20 04:40:31 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 20 Jul 1999 08:40:31 GMT
Subject: Temporarily change cursor?
References: <3793AADF.5EDBAFBC@compaq.com>
Message-ID: <006801bed28b$87756f20$f29b12c2@secret.pythonware.com>

Greg Ewing <greg.ewing at compaq.com> wrote:
> Is there an easy way to temporarily change the
> cursor in Tk while a piece of code runs? Or
> even a hard way that always works?

http://www.deja.com/getdoc.xp?AN=392027946

</F>




From m.faassen at vet.uu.nl  Wed Jul 21 04:57:28 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 21 Jul 1999 08:57:28 GMT
Subject: Python suitability questions
References: <0CCE781E44F2C872.F528377A62B41DCA.73CAA63AD6E79314@lp.airnews.net> <7n2a9p$trq$1@news.accu.uu.nl> <C8A42E1E0DB3D792.D5899D7736E26702.F072FD528818FB27@lp.airnews.net>
Message-ID: <7n421o$6ug$1@news.accu.uu.nl>

Jason Maskell <jmaskell at nospam.saskpublishers.sk.ca> wrote:
> Martijn Faassen <m.faassen at vet.uu.nl> wrote in message

[snip]

>> What do you mean by an 'internal object messaging'?

[snip answer]

I think you'll find Python very nice to implement dispatcher classes like
this.
  
> thinking that I might have to have some ungodly source file organisation
> where the cards are in one or more text files, and then one or two other
> text files for various other things. I guess I'm just more interested in
> whether or not I can point the interpreter to an arbitrary chunk of code and
>  let it do it's thing, or what exactly is involved. I'll do some more
> research. ;>

I think you won't be disappointed. Python's module organisation is the 
normal way to organize code. You can put everything in a single vast
module if you like, or distribute things across modules. There are also
ways to point Python at arbitrary chunks of code, but I'd try going with 
Python's module system first if I were you.

Regards,

Martijn




From mgushee at havenrock.com  Thu Jul 22 20:54:07 1999
From: mgushee at havenrock.com (Matt Gushee)
Date: 23 Jul 1999 09:54:07 +0900
Subject: how do I find memory leaks?
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC23@molach.origin.ea.com>
Message-ID: <m34siwkwy8.fsf@havenrock.com>

"Stidolph, David" <stidolph at origin.ea.com> writes:

> I posted a module called Debug to track down memory leaks.  It requires
> specific code be added to your __init__ and __del__ routines, as well as
> having a Destroy method we use to eliminate circular references.  If you
> want it I will post it again.

Please do!

Matt Gushee
Portland, Maine, USA



From bmatt at server.python.net  Thu Jul 15 09:40:13 1999
From: bmatt at server.python.net (Bruno Mattarollo)
Date: Thu, 15 Jul 1999 09:40:13 -0400
Subject: httplib supporting HTTP/1.1 [Q]
Message-ID: <19990715094013.A29589@starship.python.net>

Hi Pythoneers!

	I was wondering if someone was working on a module to support the HTTP/1.1 protocol... I saw in the AMK page (at Python.Net) that there is a ToDo item ... But is there a module that fully supports this protocol?

	TIA

/B



From sennomo at my-deja.com  Thu Jul 15 12:53:45 1999
From: sennomo at my-deja.com (sennomo at my-deja.com)
Date: Thu, 15 Jul 1999 16:53:45 GMT
Subject: MIME coding problems
Message-ID: <7ml3mj$l37$1@nnrp1.deja.com>

I am having problems with MIME encoding and decoding.  The following
code is supposed to take a binary file, base64-encode it, then decode
the encoded form, producing a copy of the original file.  Can anyone
explain to me why it doesn't work?

------------------------------------------
import mimetools

# open a binary file for reading
input_file = open("any.gif", 'r')

# open a transition file for writing
# to store base64 encoded form of input_file
transition_file = open("transition.gif", 'w')

# encode
mimetools.encode(input_file, transition_file, 'base64')

# close files
transition_file.close()
input_file.close()

# open encoded transition file for reading
transition_file = open("transition.gif", 'r')

# open new file for writing
# to store decoded version of transition_file
final_file = open("new.gif", 'w')

# decode, essentially creating a new copy of
# input_file
mimetools.decode(transition_file, final_file, 'base64')

# close files
final_file.close()
transition_file.close()
------------------------------------------

For some reason, "new.gif" is always shorter than the
original "any.gif".  I've tried it with different files, and changing
the order of file openings and closings, but the results are always the
same.  I am using WinNT Server 4.0.  Could that have anything to do
with the problem?  I'm kind of in a hurry to resolve this, so I would
greatly appreciate any help I can get.

sennomo


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From fredrik at pythonware.com  Mon Jul 19 05:59:13 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Jul 1999 09:59:13 GMT
Subject: Strange side effect from reading a list of numbers from argv
References: <O3sk3.6047$Sg6.52046@news1.rdc1.sdca.home.com>
Message-ID: <005701bed1cd$5c12f640$f29b12c2@secret.pythonware.com>

Harold Gottschalk <heg at softbrew.com> wrote:
> I just wrote this program to become familiar with python and I do not
> understand why it works with a list I create with in the program and not
> when I use sys.argv[1:].
> 
> I am sure there is some nuance I am missing, but an explanation to its
> behavior would be helpful.

adding a "print type(value)" or "print repr(value)" inside
the loop helps you figure out what's going on.  then read
http://www.python.org/doc/current/ref/comparisons.html
to find the rules for comparing apples and oranges:

    "Most other types compare unequal unless
    they are the same object; the choice whether
    one object is considered smaller or larger than
    another one is made arbitrarily but consistently
    within one execution of a program."

adding a "value = int(value)" inside the loop helps you
do what you want.  if you're feeling fancy, you can
also write that as:

    for value in map(int, sys.argv[1:]):
        ...

</F>




From cgw at fnal.gov  Wed Jul 28 10:42:36 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Wed, 28 Jul 1999 14:42:36 GMT
Subject: beep
In-Reply-To: <379F0751.71214802@lri.fr>
References: <379F0751.71214802@lri.fr>
Message-ID: <14239.5852.727974.172767@buffalo.fnal.gov>

Stephane Conversy writes:
 > Hi all.
 > 
 > How to make the system beep with python ?

Not guaranteed to work on all systems or in all situations (e.g. if
output is redirected to a file), but "print chr(7)" works most of the
time.

You could get around the I/O redirection problem by doing something
like this:

def beep()
    f=open('/dev/tty','w')
    f.write(chr(7))
    f.close()
 
but it would not be portable to Windows.





From phil at ricochet.net  Thu Jul  8 03:04:44 1999
From: phil at ricochet.net (phil)
Date: Thu, 08 Jul 1999 07:04:44 GMT
Subject: Two Pythons talking to each other?
References: <199907080019.TAA22244@buffalo.fnal.gov> <199907080529.HAA19502@axil.hvision.nl>
Message-ID: <3784486a.7993113@news.ricochet.net>

On Thu, 8 Jul 1999 05:30:40 GMT, "Hans Nowak" <ivnowa at hvision.nl>
wrote:

>
>On 7 Jul 99, Charles G Waldman wrote:
>
<snip good solution>
>
<snip good question>
>
>So my question is, how does one generally deal with this? Should I 
>attach a newline (\n) (or maybe another separator character) after 
>every string? Or are there other ways to guarantee that two commands 
>will arrive at the other side as two strings?
>
The primary protocols are TCP and UDP. UDP is non-connection oriented
and will deliver a datagram. Sending a packet from one side to the
other followed by another packet will not coalesce the packets into
one. However, seems to me that you really want a session oriented
solution. I suspect that the guys in the sockets group will tell you
that a TCP connection is a stream, so all bets are off. Sending "hello
wordl. " followed by sending "hello snailor" might result in receiving
"hello wordl.hel" and "lo snailor".  The solution is to impose your
own protocol on top of this using TCP as a delivery mechanism.
Something as simple as <NL> delimiting  might do it for you. 

If  the localhost address works ok, you can also use the local machine
name or the local IP address to open connections and this should ease
the move to running on separate machines.

TCP/IP seems to work in lots of places'ly yrs.
(did i get that right?).

phil.




From MHammond at skippinet.com.au  Tue Jul 13 10:27:49 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Wed, 14 Jul 1999 00:27:49 +1000
Subject: Python and COM?
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F0F@RED-MSG-50> <378aa53f.0@news.dbtech.net>
Message-ID: <7mfi9q$t73$1@m2.c2.telstra-mm.net.au>

[Re adding typeinfo support to Python COM objects]
> that C++ or VB for creating objects.  If anyone knows how far along
> Christian Tismer is with what he was working on, or what's left to be
done.

I dont think Chris got far - he got distracted by stackless Python :-)

There was basically code presented that worked.  However, it was not
integrated with the existing win32com code, but instead was split out into
"proof-of-concept".  Some ugliness exists, but there was general agreement
(I think) that fixing that required enough work that it may never happen.
The only things remaining were agreement on the attribute names (and we can
assume silence was approval, else we wont get anywhere either :) and an
implementation!

Mark.





From alex at somewhere.round.here  Wed Jul 21 10:37:43 1999
From: alex at somewhere.round.here (Alex)
Date: 21 Jul 1999 10:37:43 -0400
Subject: Merging two lists as a dictionary
References: <19990720211946.A12202@pooh.frostnet.net> <199907210323.NAA12503@piglet.dstc.edu.au>
Message-ID: <etdlncagjbc.fsf@m2-225-11.mit.edu>

Would 

import operator
destination={}
assert len(list_of_keys)==len(list_of_values)
map(operator.setitem,len(list_of_keys)*[destination],list_of_keys,
    list_of_values)

do the trick?

Maybe the "for" loop way is faster, though.

Alex.



From m.faassen at vet.uu.nl  Tue Jul 20 12:36:27 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 20 Jul 1999 16:36:27 GMT
Subject: Stackless & String-processing
References: <001101bed26d$a5bc4580$139e2299@tim> <7n27vi$mm4$2@cronkite.cc.uga.edu>
Message-ID: <7n28ib$6v8$5@news.accu.uu.nl>

Graham Matthews <graham at sloth.math.uga.edu> wrote:
> Tim Peters (tim_one at email.msn.com) wrote:
> : Who said anything about agreeing?  You two obviously disagree, but
> : disagreement doesn't imply misunderstanding.  Heck, you've both repeated the
> : same unbudging positions so often even *I* understand what you're both
> : saying <wink>.
> :
> : + other equally useful contributions.

> What is the point of your post Tim? Getting your jollies from adding
> useless material to a thread?

Actually this makes Tim go up in his political standing in this newsgroup,
if this is possible. :)

Actually, no. This kind of post is actually socially useful. By adding
some lightness to a heated discussion it's sometimes possible to prevent
nastiness from occuring.

Of course I shouldn't give such Grand Usenet Secrets away; my Pythonic
political standing is sure to drop. :)

Regards,

Martijn




From smalleys at gte.net  Sat Jul  3 18:08:24 1999
From: smalleys at gte.net (Sue)
Date: Sat, 03 Jul 1999 22:08:24 GMT
Subject: How do you make clickable Python/Win9x scripts?
References: <000301bec519$bdd41640$d69e2299@tim>
Message-ID: <377E89FF.54AB2F2D@gte.net>

I have an NT hack - not sure about Win95.  Note that this
does NOT require registry/admin access, as far as I know.

1.  Make sure the .pyc version of your file exists using CLI
"python.exe compileall.py $cwd" or whatever.

2. Create a new filetype (Python Executable), pointing to
your python binary, with suffix .pyc.

3. Add .pyc to CMD_SUFFIXES (not 100% sure)

4. Drop a link to the pyc file on your desktop.

5.  Pass script parameters using standard Win hack.

6. I tried a PIF file but gave up.  Looked like overkill.

Sue.



From johann at physics.berkeley.edu  Thu Jul  1 04:20:00 1999
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 01 Jul 1999 01:20:00 -0700
Subject: sound-playing modules?
Message-ID: <mtiu84zsr3.fsf@astron.berkeley.edu>

Hi,

Are there any modules out there for playing back sound files, such as
.au, .wav, .mp3, and other formats?

Are any of these modules portable between Windows and Linux?

-- Johann

-- 
Johann Hibschman                           johann at physics.berkeley.edu



From frohne at gci.net  Thu Jul 29 12:20:56 1999
From: frohne at gci.net (Ivan Frohne)
Date: Thu, 29 Jul 1999 08:20:56 -0800
Subject: Bit-twiddling
References: <000f01bed969$fbed9f00$71a22299@tim> <wklnc0x7bf.fsf@ifi.uio.no>
Message-ID: <37a10adc@news.gci.net>

> Based on this advice I took a stab at the rest of the algorithm and
> now it actually works! Wheee! So if anyone wants a string of random
> numbers with a period of 2**199937-1 and equi-distributed in 623
> dimensions, just let me know. :-)
>
> | if-you're-really-after-random-numbers-see-ivan-frohne-ly y'rs  - tim

I have a new version of the Mersenne Twister just around the corner
that uses the Numeric module if it's available.  This makes it almost
as fast as whrandom.

To get a right shift for ordinary integers that always shifts in high-order
zeros, this
works (x is an ordinary 32-bit signed integer):

 x >> 1 & 0x7FFFFFFF  #  Logical right shift 1
 x >> 2 & 0x3FFFFFFF  #  Logical right shift 2
.....
 x >> 11 & 0x001FFFFF  # Logical right shift 11

etc.

--Ivan Frohne





From poinot at onera.fr  Wed Jul 28 07:15:56 1999
From: poinot at onera.fr (Marc POINOT)
Date: Wed, 28 Jul 1999 13:15:56 +0200
Subject: Gadfly DBase and CGI
References: <7nmkpg$cjl$1@gruvel.une.edu.au>
Message-ID: <379EE66B.C1122073@onera.fr>

David Miron wrote:

> Hi All,
>
> I am trying to use the KWP database within a cgi script.  This is on a dec
> unix machine running apache.
>
> Has anyone ever done this?

Yes, Gadfly, CGI, DEC and Apache too...
What is the error ? (you can have  look at
the appache server errors, somewhere
like /usr/local/etc/httpd/logs ).
Can you read plain html files from a distant machine
with this server ?

Marcvs [alias Is it an Apache, a Python-cgi or a Gadfly error ?]






From tismer at appliedbiometrics.com  Sat Jul 17 09:49:28 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Sat, 17 Jul 1999 13:49:28 GMT
Subject: Snippet: bitcount()
References: <000601bed02a$2235af20$06a22299@tim>
		<m33dynlhsg.fsf@atrus.jesus.cam.ac.uk>
		<379074EE.B6D563F6@appliedbiometrics.com> <m3n1wva221.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <379089E8.CBC2BCB7@appliedbiometrics.com>


Michael Hudson wrote:
> 
> Christian Tismer <tismer at appliedbiometrics.com> writes:
> 
> > > Mini-proposal: len(a) for a long returns the length of the number in
> > > radix-256.
> >
> > bitcount2 is not very slow, see here:

> Still chews gobs of memory though... is there any way of converting
> longs <-> arrays? That'd do the trick, though you'd still end up
> copying, no doubt. Still, I don't think I'm being realistic to talk of
> dealing with numbers so large copying them is an issue in Python. It's
> all just for play, anyway.

No way to do that. Longs are exposed to marshal only at this
time. I guess it could be made visible as a read-only buffer
if we extend the buffer function (i.e. extend longs with
buffer methods), but is it worth it?
The single copy for marshal is very cheap, compared to hex.

if-we-had-this-trick-you-would-have-seen-it-in-the-thread - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From fredrik at pythonware.com  Tue Jul 27 20:50:14 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Jul 1999 00:50:14 GMT
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <slrn7psisn.da.wtanksle@dolphin.openprojects.net>
Message-ID: <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com>

William Tanksley <wtanksle at dolphin.openprojects.net> wrote:
> The very worst part of the current else: behavior is that it changes the
> meaning of else.  In other constructs, else is an alternate path to take
> if the data being tested fails a single expression test.  In this
> contruct, else is a path taken if the code block belonging to the previous
> test executes a certain instruction.

nope.  you've got it all backwards.  consider this:

    for an "if" statement, "else" is a path taken if the
    expression evaluates to false.

    for a "while" statement, "else" is a path taken if the
    expression evaluates to false.  or in other words,
    when the loop terminates by natural causes.

    for a "for" statement, "else" is a path taken when
    there are no more elements to loop over.  or
    in other words, when the loop terminates by
    natural causes.

not that complicated, was it?

now, what you seem to have trouble with isn't
the "else"-clause -- it's the behaviour of "break":

    "break" doesn't just break out of the inner
    body (the portion between while/for and
    else), it breaks out of the ENTIRE construct.
    including the "else" clause.  just as the docs
    say.

"else" always works the same way.  and so does
"break".  what else did you expect from Guido?

</F>




From coll at tomcat.ru  Thu Jul 15 23:50:31 1999
From: coll at tomcat.ru (coll at tomcat.ru)
Date: 16 Jul 1999 03:50:31 GMT
Subject: Simple The Best !!!
Message-ID: <7393HYV1.2A3ZSPG3@tomcat.ru>

Looking for absolutely new ways to earn money?
I got them for you!
That's it: http://jump.to/webmoney
There are no BS ! only approved methodes.
Why don't you visit this site right now?...



From claird at starbase.neosoft.com  Mon Jul 19 07:03:47 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Jul 1999 11:03:47 GMT
Subject: what is python?
References: <7mt81n$s6e$1@bgtnsc03.worldnet.att.net> <7970018B1A94450D.D0FA47F7C6A2DEE5.9D19F01B85820D25@lp.airnews.net>
Message-ID: <258CC1FEDCF009C4.29FBCDEA7639F3B2.1F4FE0E20B13F6A6@lp.airnews.net>

In article <7970018B1A94450D.D0FA47F7C6A2DEE5.9D19F01B85820D25 at lp.airnews.net>,
I expostulated:
			.
			.
			.
>in several senses.  Lars Marius Garshol, by the way, has a few
>wise words to say on their relative performance.
Whoops!  I thought I'd mentioned <URL:http://
x10.dejanews.com/getdoc.xp?AN=387053216> here.
My mistake.
			.
			.
			.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From mithy at my-deja.com  Thu Jul 22 23:05:50 1999
From: mithy at my-deja.com (Preston Landers)
Date: Fri, 23 Jul 1999 03:05:50 GMT
Subject: Python, Linux 2.0, GCC 2.7.2.3 and threading
Message-ID: <7n8m6a$c7f$1@nnrp1.deja.com>

Hello everyone,

A friend of mine posted here recently describing a problem he was
having running my program Pagecast under Linux 2.0.  Pagecast is a
Python program that uses several threads.  He was using Linux 2.0.36,
glibc 2.0.7, and Python 1.5.2.  I can't confirm at this time what
compiler version he was using (to compile Python) but I think it was
probably GCC 2.7.2.3.

I know this because I could reproduce his errors on my machine when I
rebooted into Linux 2.0.36.  At some point in my program, after a
couple of threads ran and exited, Linux would 'oops' then continue on
its merry way.  I can post a kernel oops that Klaus (my friend) sent
me; mine were very similiar. I had compiled my Python with GCC
2.7.2.3.

Then I recompiled Python with pgcc-2.91.66 19990314 (egcs-1.1.2
release) for unrelated reasons.  I was no longer able to reproduce the
problem; everything worked great, and Python was *much* peppier to
boot.

I'm just wondering if anyone can confirm having a similar experience;
does Python (especially 1.5.2) need to be compiled with egcs on Linux
for threading to work properly?  If so, I would like to put words to
that effect in my program's README to warn my users.

Never mind that Red Hat apparently didn't feel the need to include
threading support in their 5.x release Python packages.  (Nor
slackware.)  I get more email about that than anything else.  The
answer: get a real distribution.  No really, the answer is recompile
Python.  I'm just wondering if I need to add a caveat that says "with
egcs" or perhaps "NOT with gcc 2.7.2.3."

Thanks for your time,

---Preston

preston at askpreston.com

Pagecast is at:
http://askpreston.com/projects/pagecast

The Pagecast 1.1 release is coming soon!  Watch Freshmeat.net for
details.


--
||  Preston Landers <mithy at my-deja.com>  ||
||        http://askpreston.com          ||


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From bruce_dodson at bigfoot.com  Fri Jul 30 14:16:44 1999
From: bruce_dodson at bigfoot.com (Bruce Dodson)
Date: Fri, 30 Jul 1999 15:16:44 -0300
Subject: Recursive constructors?
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC70@molach.origin.ea.com>
Message-ID: <X1mo3.42931$jl.30215469@newscontent-01.sprint.ca>

It is legal, but you had better make sure that (whatever) becomes false
somewhere along the line, as a side-effect of foo::__init__.  Otherwise, I
pity the foo at the bottom, who has to carry an entire heap full of bars.
(okay,okay)

>class foo:
>   def __init__(self, x):
>      if (whatever):
>         self.bar = foo(y)





From joelg at alum.mit.edu  Sat Jul 31 14:05:20 1999
From: joelg at alum.mit.edu (Joel Gould)
Date: Sat, 31 Jul 1999 18:05:20 GMT
Subject: CGI Database Question
Message-ID: <37a33a28.23844403@news.ne.mediaone.net>

This seems to come up every now and then but I was unable to find an
answer to this questions.

I would like to use Python for CGI scripts.  I am going to ask my ISP
to install Python (they run an NT server with IIS), but what database
package should I use?  Specifically, what should I tell my ISP to
install so I can use a (simple) database from my CGI scripts?

Also, do I have to worry about thread safety?  In other words, can
multiple requests come into the web server causing multiply copies of
my Python scripts to be run concurrently?  Does this impact the
database selection?

Joel Gould



From gmcm at hypernet.com  Sat Jul 24 00:33:10 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 23 Jul 1999 23:33:10 -0500
Subject: Memory leaks whilst embedding python
In-Reply-To: <8E0CE4469aymaratcpcouk@news.tcp.co.uk>
Message-ID: <1279335510-28879097@hypernet.com>

Chris Williamson discovers that:

>      Py_Initialize();
>      Py_Finalize();

leaks.

Well, if MSVC reported on all of MFC's static allocations, you'd see 
that it leaks, too.

These are nothing to worry about. It's static data that Python needs 
to run. It won't continually grow. If you use the debug memory stuff, 
you should take your checkpoint after Py_Initialize and report before 
Py_Finalize. In fact, you should probably take your checkpoint after 
you've done your imports, because a lot of the memory they use won't 
get released until Py_Finalize.



- Gordon



From Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org  Thu Jul  1 10:38:53 1999
From: Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org (Mikael Lyngvig)
Date: Thu, 01 Jul 1999 15:38:53 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001613@bossar.com.pl>

From: mikael at pobox.com (Mikael Lyngvig)

On Thu, 01 Jul 1999 12:16:00 GMT, Andrew Cooke
<andrew at andrewcooke.free-online.co.uk> wrote:

>I'm confused.  I'm sitting at a moderate NT box (266MHz, 96Mb) - my
>work machine, and running the Tk demos.  There is no problem at all -
>they are instantaneous.
>
>Is this problem to do with memory?  Or is it a W95 thing?  If so,
>is it due to thread scheduling?

Not memory, all machines I've tried it on had 128 MB.  Not Windows 95,
I never tried Tk (or anything but a game) on a Windows 9x machine
(only NT).



-- Mikael



From news at dorb.com  Thu Jul  1 15:14:38 1999
From: news at dorb.com (Darrell)
Date: Thu, 1 Jul 1999 15:14:38 -0400
Subject: Installing Python/XML on Windows?
References: <377bab1c@news.depaul.edu>
Message-ID: <N8Pe3.4327$U6.23654@newsr2.twcny.rr.com>

Check the doc for the details but in short.
Use site.pth which for me on windows is in my python directory.
It looks something like this.

d:/python/site-packages
d:/python/site-packages/other
d:/python/site-packages/Aycock
d:/python/site-packages/kj

In this same vein I use this to avoid adding each one of these to my system
path.

#!/usr/bin/env python
"""
The .py you want to run is in the python path but not
in your shell path.

I know everyone but me already has this problem solved
in a better way than this. If so please show me the way.

Assuming this is runIt.py then
runIt.py someScript.py arg1 arg2
"""
import sys
import imp

fp, pathname, description = imp.find_module(sys.argv[1])
sys.argv=sys.argv[2:]
md=imp.load_module('__main__', fp, pathname, description)



--
--Darrell
Jonathan W Hendry <jhendry at shrike.depaul.edu> wrote in message
news:377bab1c at news.depaul.edu...
> Does any documentation exist on this? How do I set up the
> paths so that Python can use all the various components of
> the package?
>
> Thanks,
>
> - Jon





From nathan at islanddata.com  Tue Jul 27 15:55:33 1999
From: nathan at islanddata.com (Nathan Clegg)
Date: Tue, 27 Jul 1999 12:55:33 -0700 (PDT)
Subject: Python and Red Hat Linux 6
In-Reply-To: <379CD5C2.5BA096A2@mail.danbbs.dk>
Message-ID: <XFMail.990727125533.nathan@islanddata.com>

On 26-Jul-99 Brian Pedersen wrote:
> !#/usr/bin/env python 
> 
> instead (don't ask me why, i read it in a FAQ somewhere :-)


You have to give #! an absolute pathname for it to work.  env is
(somewhat) reliably at /usr/bin and is a program used exclusively for
running other programs in certain environments.  In this case, it is used
because it will search your path for python rather than requiring it to be
explicit.


----------------------------------
Nathan Clegg
 nathan at islanddata.com





From Gary.Pennington at uk.sun.com  Thu Jul  8 12:47:11 1999
From: Gary.Pennington at uk.sun.com (Gary Pennington - UK Performance Centre)
Date: Thu, 08 Jul 1999 17:47:11 +0100
Subject: Memory management & Garbage Collection - pickle/cPickle
Message-ID: <3784D60F.49AFFDCB@uk.sun.com>

Hi,

I've written some server side code which makes extensive use of shelve
(and indirectly pickle). I've noticed that over time the amount of heap
used by the server grows considerably. Investigation showed that objects
create from pickle or cPickle cause memory to be allocated that is never
freed.

The following test program soon uses up all the memory on my machine :-

import sys
sys.path.insert(0,"/home/garyp/dev/webJudge/dev/cgi")
import shelve
d=shelve.open("gameDB")
print d.keys()
def getObj():
                return d["remote1"]
for i in range(0,100000):
                x=getObj()
                del x
                x=None

gameDB contains some objects previously pickled and about 30k in size.
Running the above and using truss I can see memory being allocated (lots
of brk calls) as the data is continuosly re-read from the file.

The del calls and x=None seem to make no difference. Also, it makes no
difference whether I go through shelve or access the data files direct.
My system is using dumdbm.py, it's Solaris 2.6.

Any suggestions as to what the problem is?

Gary Pennington






From guido at cnri.reston.va.us  Thu Jul 15 18:25:24 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 15 Jul 1999 18:25:24 -0400
Subject: Running 2 gui libs simultaneously
References: <1280251163-5463910@hypernet.com> <378BEC9C.271A69EB@compaq.com>
Message-ID: <5lwvw1czzf.fsf@eric.cnri.reston.va.us>

Greg Ewing <greg.ewing at compaq.com> writes:

> Gordon McMillan wrote:
> > 
> > You don't have to use Tk's mainloop. You can call update() in a loop.
> > Perhaps wxWindows has something similar, or you can call Tk's update
> > from within some idletask in wxWindows.
> 
> I don't fancy your chances of getting that to work.
> The problem isn't just getting two main loops to
> run at the same time -- it's making sure that each
> event goes to the right UI manager.

I don't think that's the problem, at least not on Unix.  Each GUI
package has its own connection (a socket) to the X server, and the X
server knows which connection to send events to.  No different that
having two different apps.  If wxWindows exposes the socket and the
raw event handling function, you can use Tk's createfilehandler to add 
wxWindows's socket to the set of sockets on which it listens (default
only its own connection to the X server) and in the callback call
wxWindows's event handler.  Theoretically this should work fine.

> You might be able to do it by reading events
> yourself, figuring out somehow which UI they
> are meant for, and somehow passing them on to
> the relevant toolkit. Would be platform-dependent,
> since you'd have to deal with events at a low
> level.

This should not be necessary at all.

(On the other hand, I don't know if this perhaps is a real problem on
Windows?)

--Guido van Rossum (home page: http://www.python.org/~guido/)



From Tom.Loredo at p98.f112.n480.z2.fidonet.org  Fri Jul  2 08:17:40 1999
From: Tom.Loredo at p98.f112.n480.z2.fidonet.org (Tom Loredo)
Date: Fri, 02 Jul 1999 13:17:40 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001691@bossar.com.pl>

From: Tom Loredo <loredo at spacenet.tn.cornell.edu>

lvirden at cas.org wrote:
> 
> According to Vadim Zeitlin <zeitlin at dptmaths.ens-cachan.fr>:
> :
> : The sad truth is that macs are not very common nowadays. I would like to
help
> :develop wxMac, but I don't have one (nor at home nor anywhere near me in the
> 
> I wonder why various industry rags have proclaimed during the past
> 10 months about Apple's G3 machines outselling other vendor
> machines.  ...

Make me a third.  About a million iMacs have been sold in less than
a year, and although the Mac market share is only around 10% right
now (that's a huge number of Macs, by the way!), the fraction
of all personal computers *in use* that are Macs is actually about 20%.
It is simply way off the mark to believe that Macs are not very common
nowadays.  I spend most of my time at a Unix box, and work on all
three major platforms regularly (Unix, Mac, Win32), so please don't
dismiss this as just another Mac-only user mouthing off.  It's a great
platform, and I simply wouldn't want to see python GUI development
ignoring it.

-Tom Loredo



From aaron_watters at my-deja.com  Fri Jul 23 09:30:12 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Fri, 23 Jul 1999 13:30:12 GMT
Subject: gadfly RDBMS ???
References: <Pine.LNX.4.04.9907211629470.12211-100000@emerald.netskate.ru> <3795C21F.A0F13648@onera.fr> <3795EA7F.4A452490@arcus.tel.hr> <37969382.C25210F4@compaq.com>
Message-ID: <7n9qp5$m8q$1@nnrp1.deja.com>

In article <37969382.C25210F4 at compaq.com>,
  Greg Ewing <greg.ewing at compaq.com> wrote:
> The same goes for indexes, key declarations,
> integrity constraints, etc. etc. They're optional
> extras to the basic relational model which
> Gadfly doesn't provide -- but that doesn't mean
> Gadfly isn't relational!

Ahem.  Gadfly *does* provide indices.  And a unique index
functions the same as a key constraint.

More complex integrity constraints are not provided,
but would be nice.

   -- Aaron Watters

===

Doing business with some people is like
petting a skunk.
It's a learning experience
you don't care to repeat.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gregm at iname.com  Wed Jul 21 01:09:57 1999
From: gregm at iname.com (Greg McFarlane)
Date: Wed, 21 Jul 1999 15:09:57 +1000
Subject: Tk: Temporarily change cursor?
In-Reply-To: <3794EC17.64249239@compaq.com>; from Greg Ewing on 21 Jul 1999 at 09:37:27AM
References: <3793AADF.5EDBAFBC@compaq.com> <19990720145034.04299@nms.otc.telstra.com.au> <3794EC17.64249239@compaq.com>
Message-ID: <19990721150957.10799@nms.otc.telstra.com.au>

On 21 Jul, Greg Ewing wrote:
> Greg McFarlane wrote:
> > 
> >     Blt: http://www.tcltk.com/blt/
> 
> I've just looked at the web page for Blt, and it says
> this about the "busy" command:
> 
>   Manages an invisible "busy" window which prevents further
>   user device (keyboard, mouse, button, etc.) interactions.
> 
> I don't like the sound of that. I don't want to lock
> up the whole machine and prevent interaction with
> windows belonging to other processes.

Blt busy does not "lock up the whole machine".  It simply prevents any
interaction with the windows you specify (until you call
busy_release).  For example, you can display a busy cursor over (and
stop user input from) some of the windows in your application and only
let the user interact with one window.  You can implement modal
dialogs this way.

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From aarhodes at cisco.com  Thu Jul  8 16:23:25 1999
From: aarhodes at cisco.com (Aaron Rhodes)
Date: Thu, 08 Jul 1999 20:23:25 +0000
Subject: "Wrapping" Functions in classes
Message-ID: <378508BC.4ECD2149@cisco.com>

How do I wrap the methods in a class so that before
they are called, another function gets called first.
Is there a special __variable__ that I can use?
This other function should be called for each of the
methods in the class.


Aaron

aarhodes at cisco.com






From mfletch at vrtelecom.com  Thu Jul  1 10:22:54 1999
From: mfletch at vrtelecom.com (Mike C. Fletcher)
Date: Thu, 01 Jul 1999 14:22:54 GMT
Subject: TopSort in Python?
References: <7lfjn0$b1v$1@nnrp1.deja.com>
Message-ID: <2RKe3.1724$_A4.30209@news2.rdc1.on.home.com>

Well, it's not topsort-the-official-algo, but when I first described it, Tim
Peters said it was a type of topological sort.  Warning: it's extremely
inefficient, using recursive calls of all things!  I did this without any
knowledge of the computer science in the field (have since learned some, but
didn't fix this code), it was just a description of an algorithm I needed to
get a job done, and I don't use it any more.

Cheers,
Mike


Dinu C. Gherman <gherman at my-deja.com> wrote in message
news:7lfjn0$b1v$1 at nnrp1.deja.com...
> Does anybody have a simple-minded-but-working full-Python
> implementation of topsort, the topological sorting algorithm?
> Or maybe *any* topsort? I remember only one such algorithm...
> python.org/search doesn't reveal any...
>
> Thanks,
>
> Dinu
>
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.
>


begin 666 dsort.py
M;G5L;'9A;" ]("@Q+"D-"@T*8VQA<W, at 1%-O<G0Z#0H))R<G#0H)02 B9&5P
M96YD96YC>2(@<V]R=&EN9R!C;&%S<RP@=7-E9"!T;R!O<F1E<B!E;&5M96YT
M<PT*"6%C8V]R9&EN9R!T;R!D96-L87)E9" B9&5P96YD96YC:65S(B H;6%N
M>2UT;RUO;F4@<F5L871I;VYS:&EP<RD-"@E)<R!N;W0 at 82!B96%U=&EF=6P@
M86QG;RP at 8G5T(&ET('=O<FMS("AO<B!S965M<R!T;RD-"@E297%U:7)E<R!H
M87-H86)L92!V86QU97, at 9F]R(&%L;"!E;&5M96YT<RX-"@D-"@E4:&ES(&ES
M(&$@<75I8VL@:&%C:RP@=7-E(&%T('EO=7(@;W=N(')I<VLA#0H)#0H)0F%S
M:6,@=7-A9V4Z#0H)"4-R96%T92!A($13;W)T(&UY<V]R=&5R#0H)"69O<B!E
M86-H(&5L96UE;G0@<2!W:&EC:"!I<R!P87)T(&]F('1H92!S970@=&\@<V]R
M="P at 8V%L;#H-"@D)"6UY<V]R=&5R+G)U;&4H(&1S;W)T+FYU;&QV86PL('$I
M#0H)"0DC('1H:7,@:7,@;F]T('-T<FEC=&QY(&YE8V5S<V%R>2!F;W(@96QE
M;65N=',@=VAI8V@@87)E#0H)"0DC(&1E<&5N9&5N="!O;B!O=&AE<B!O8FIE
M8W1S+"!B=70@:70@:7,@;F5C97-S87)Y(&9O<@T*"0D)(R!T:&]S92!W:&EC
M:"!A<F4@;F]T+B @1V5N97)A;&QY(&ET)W, at 96%S:65S="!T;R!C86QL#0H)
M"0DC('1H92!N=6QL(')U;&4 at 9F]R(&5A8V@@96QE;65N="X-"@D)9F]R(&5A
M8V@@<G5L92!X(&1E<&5N9',@;VX@>2P at 8V%L;#H-"@D)"6UY<V]R=&5R+G)U
M;&4H('@L('DI#0H)"7=H96X at 7V%L;%\@<G5L97, at 87)E(&5N=&5R960L(&-A
M;&P-"@D)=')Y. at T*"0D)<V]R=&5D;&ES=" ](&UY<V]R=&5R+G-O<G0H*0T*
M"0EE>&-E<'0 at 5F%L=65%<G)O<CH-"@D)"6AA;F1L92!R96-U<G-I=F4 at 9&5P
M96YD96YC:65S(&AE<F4N+BX-"@D)#0H)"0T*"49O<B!A;B!E>&%M<&QE(&]F
M(')E86PM;&EF92!U<V4L('-E92!T:&4 at 5E)-3"!L:6YE87)I<V5R+ at T*"0T*
M"2<G)PT*"61E9B!?7VEN:71?7RAS96QF+"!R96-U<G-E17)R;W(]3F]N92 I
M. at T*"0ES96QF+F1E<&5N9&]N(#T@>VYU;&QV86PZ6S!=?0T*"0ES96QF+G)E
M8W5R<V5%<G)O<B ](')E8W5R<V5%<G)O<@T*"61E9B!R=6QE*"!S96QF+"!D
M97!O;BP at 9&5P<RDZ#0H)"2<G)PT*"0E296=I<W1E<B!A(")R=6QE(BX@($)O
M=&@@96QE;65N=',@;75S="!B92!H87-H86)L92!V86QU97,N#0H)"5-E92!T
M:&4 at 8VQA<W,G(&1O8W5M96YT871I;VX at 9F]R('5S86=E+ at T*"0DG)R<-"B,)
M"7!R:6YT("<G)W)E9VES=&5R:6YG(')U;&4Z)R<G+"!D97!O;BP at 9&5P<PT*
M"0EI9B!S96QF+F1E<&5N9&]N+FAA<U]K97DH(&1E<',@*2!A;F0 at 9&5P;VX@
M:7,@;F]T(&YU;&QV86PZ#0H)"0ES96QF+F1E<&5N9&]N6R!D97!S(%TN87!P
M96YD*"!D97!O;B I#0H)"65L:68 at 9&5P;VX@:7,@;F]T(&YU;&QV86PZ#0H)
M"0ES96QF+F1E<&5N9&]N6R!D97!S(%T@/2!;+3$L(&1E<&]N70T*"0EE;&EF
M(&YO="!S96QF+F1E<&5N9&]N+FAA<U]K97DH(&1E<',@*3H-"@D)"7-E;&8N
M9&5P96YD;VY;(&1E<', at 72 ](%LM,2!=#0H)9&5F('-O<G0H('-E;&8@*3H-
M"@D))R<G#0H)"4=E="!T:&4@<V]R=&5D(')E<W5L=', at 87, at 82!L:7-T#0H)
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M+'D@:6X@=&5M<#H-"@D)"71E;7 R+F%P<&5N9"@@>2 I#0H)"2, at 9F]L;&]W
M:6YG(&%D9',@=&AE(&5L96UE;G1S('=I=&@@;F\@9&5P96YD96YC:65S#0H)
M"71E;7 R6VQE;BAT96UP,BDZ72 ]('-E;&8N9&5P96YD;VY;(&YU;&QV86P@
M75LQ.ET-"@D)<F5T=7)N('1E;7 R#0H)9&5F(%]D<V]R="@@<V5L9BP@:V5Y
M+"!V86QU92 I. at T*"0EI9B!V86QU95LP72 ]/2 M,CH-"@D)"6EF('-E;&8N
M<F5C=7)S945R<F]R. at T*"0D)"7)A:7-E(%9A;'5E17)R;W(L("<G)T1E<&5N
M9&5N8VEE<R!W97)E(')E8W5R<VEV92$G)R<-"@D)"65L<V4Z#0H)"0D):68@
M7U]D96)U9U]?. at T*"0D)"0EP<FEN=" G)R=296-U<G-I=F4 at 9&5P96YD96YC
M>2!D:7-C;W9E<F5D(&%N9"!I9VYO<F5D(&EN(&1S;W)T+D1S;W)T+E]D<V]R
M="!O;B E<SHE<R<G)R4H:V5Y+"!V86QU92D-"@D)"0ER971U<FX@,2 C('=E
M(&MN;W<@:70@:&%S(&%T(&QE87-T(&]N92!R969E<F5N8V4N+BX-"@D)96QI
M9B!V86QU95LP72 ]/2 M,3H@(R!H879E;B=T('EE="!C86QC=6QA=&5D('1H
M:7,@<F1E<'1H#0H)"0EV86QU95LP72 ]("TR#0H)"0ET96UP=F%L(#T at 6S!=
M#0H)"0EF;W(@>"!I;B!V86QU95LQ.ETZ#0H)"0D)=')Y. at T*"0D)"0ET96UP
M=F%L+F%P<&5N9"@@,2 K('-E;&8N7V1S;W)T*"!X+"!S96QF+F1E<&5N9&]N
M6WA=*2 I#0H)"0D)97AC97!T($ME>45R<F]R. at T*"0D)"0ES96QF+F1E<&5N
M9&]N6R!N=6QL=F%L(%TN87!P96YD*"!X("D@(R!I<R!A;B!U;G)E9F5R96YC
M960 at 96QE;65N= T*"0D)"0ET96UP=F%L+F%P<&5N9"@@,2 I#0H)"0EV86QU
M95LP72 ](&UA>"@@=&5M<'9A;" I#0H)"0ER971U<FX@=F%L=65;,%T-"@D)
M96QS93H-"@D)"7)E='5R;B!V86QU95LP70T*)R<G#0IF<F]M(&UC9BYU=&EL
M<R!I;7!O<G0 at 9'-O<G0-"CX^/B!X(#T at 9'-O<G0N1%-O<G0H*0T*/CX^(&UA
M<"@@>"YR=6QE+"!;,2PR+#(L-"PU+#1=+"!;,BPS+#0L-2PV+#-=("D-"EM.
M;VYE+"!.;VYE+"!.;VYE+"!.;VYE+"!.;VYE+"!.;VYE70T*/CX^('@N<V]R
(="@I#0HG)R<`
`
end




From gmcm at hypernet.com  Fri Jul 16 10:21:03 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 16 Jul 1999 09:21:03 -0500
Subject: I envy newbies
In-Reply-To: <378EA68A.F2D8A7E@callware.com>
Message-ID: <1279991448-21085550@hypernet.com>

Ivan "in 24 Hours" Van Laningham wrote:
> 
> ``The first thing you need to do is to find where IDLE lives, and to
> do that you need to know where Python lives; this should be simple,
> since the Python directory should be in your path.

Unlike Tim, it's on my path (NT), but I fairly sure I put it there, 
not the installer.

> On all my Windows
> systems, Python is in "C:\Python", 

Installer's default is C:\Program Files\Python. And Tim's trick 
appears to require Active Desktop, so you can ignore him...

BTW, I thought I saw you mention PYTHONPATH in another post on 
Windows installations.  The installer does not create one, and one is 
not necessary to use the official installation. In fact, the only 
time I use it on my Windows box is so I can use an alternate 
installation (3 at the moment, but I've had as many as 7).

but-you-are-at-the-right-level-of-detail-for-your-publisher-ly y'rs 

- Gordon



From larsga at ifi.uio.no  Wed Jul 21 14:07:50 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 21 Jul 1999 20:07:50 +0200
Subject: how do I find memory leaks?
References: <7n4skv$8do@thalamus.wustl.edu>
Message-ID: <wk4six98qx.fsf@ifi.uio.no>

* Heather A. Drury
| 
| Any hints/suggestions/pointers would be greatly appreciated.

One easy way is to get my plumbo module and see if it can find any
cycles in your program. It only understands what happens in
Python-space and it's not complete, but several people have found it
useful.

<URL: http://www.stud.ifi.uio.no/~larsga/download/python/plumbo.py >


If Plumbo doesn't solve your problem (or is too slow), you can try Tim
Peter's more complex, complete and faster Cyclops, which he released
on Monday.

<URL: http://www.python.org/ftp/python/contrib/System/Cyclops.py >

--Lars M.



From wtanksle at dolphin.openprojects.net  Tue Jul 27 20:55:20 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Wed, 28 Jul 1999 00:55:20 GMT
Subject: 1.5.2 for: else:
References: <Pine.LNX.4.04.9907271612140.9849-100000@emerald.netskate.ru> <19990727142943.V4068@xs4all.nl> <slrn7ps8ej.vm0.wtanksle@dolphin.openprojects.net> <7nlfl6$7te$1@nntp6.u.washington.edu>
Message-ID: <slrn7psl7n.da.wtanksle@dolphin.openprojects.net>

On 27 Jul 1999 23:33:58 GMT, Donn Cave wrote:
>wtanksle at dolphin.openprojects.net (William Tanksley) writes:
>|On Tue, 27 Jul 1999 14:29:43 +0200, Thomas Wouters wrote:

>|> http://www.python.org/doc/current/tut/node6.html#SECTION006200000000000000000

>|> 4.4 break and continue Statements, and else Clauses on Loops 

>|> Loop statements may have an else clause; it is executed when the loop
>|> terminates through exhaustion of the list (with for) or when the condition
>|> becomes false (with while), but not when the loop is terminated by a break
>|> statement. 

>| Well, I'm suprised.  I had also expected the else clause to run on empty
>| loop.  It seems so natural, in line with the meaning of a loop (loop on
>| this; otherwise do that).

>| Fortunately I've never actually used it and thus been disenchanted.

>[(I'm confused? else clause _does_ execute after an empty loop, like
>   for x in ():
>       pass
>   else:
>       print 'for else'
>)]

Of course.  It also executes on a full loop, or a half-full one, or for
that matter a half-empty one.  The only time it doesn't execute is when
the code block _inside_ the loop happens to execute a break.  Pretty odd
when you think about it.

>It seems to me to work the only useful way,

I'm sorry, I'm allergic to that phrase.  I can't deny that it's useful
now, but anyone claiming that any current condition is the ONLY useful
way...  Ugh.  Obviously hasn't bothered even thinking about alternatives.

If there existed even a _less_ useful way you'd be wrong.  And even
supposing myself to be wrong, there does exist such a way.  QED.

>so you almost can't
>go wrong - at worst, if you mistake its meaning, you won't use it.

That's the problem!  You _will_ use it, under the mistaken impression that
it performs some kind of useful 'else' function.  Instead it performs a
hidden exception-handling.  Was this leftover from some time when Python
didn't have exception handling or something?

>That may not be the same thing as "natural", but it's close.

I'm also allergic to "intuitive", and since I nearly used the word myself,
my sinuses are completely plugged.  ;-)

I can't really claim natural behavior, but I do claim that both of our
problems are easily solved using other means, and my solution is
consistent with other uses of else while the current solution isn't.  On
your side, of course, is the fact that the current solution IS the current
solution, no ifs ands ors or elses about it.

>It's one of the neat little things that make me pause for a moment
>of appreciation when I get to use it.

I'm pausing right now for a moment of appreciation.  No, this is
incredulity.  What is it about this that makes you appreciate it?  I can't
see anything in it except a world-class kludge.  The alternate solution,
throwing an exception, is SO much nicer that very few people even USE the
current behavior.

>| Understanding the current semantics seems to require understanding the
>| nature of loop tests in Python.  Since I understand them I can agree that
>| this makes sense, but I'm still a bit leery, because you're not 'elsing'
>| on the data but rather on code inside the loop.

>| I would _definitely_ use exceptions to handle this.

>Note that "try/except" also takes an "else", which also happens to have
>the only useful sense

Hack, ack, sniffle.  ;)

>and executes when the try block executes with no
>exception.  It's handy for code that should execute in the normal case
>but does not itself need to be enclosed in an exception handler, so you
>can focus the try block more precisely on the section that needs it.

I agree.  It's a wonderfully useful tool, and unlike either of our
solutions is pretty much irreplacable.  Not as useful as 'finally', of
course.

It shares a characteristic with my solution: it executes only when the
previous control structure's block has failed to execute.

>	Donn Cave, University Computing Services, University of Washington

-- 
-William "Billy" Tanksley



From alan at littleford.net  Sat Jul 10 23:12:54 1999
From: alan at littleford.net (alan at littleford.net)
Date: Sun, 11 Jul 1999 03:12:54 GMT
Subject: Another reason assignment expressions would be nice (?)
References: <000301becb50$8c2045a0$479e2299@tim>
Message-ID: <37880adc.1505773814@news.slip.net>

On Sat, 10 Jul 1999 23:50:42 -0400, "Tim Peters"
<tim_one at email.msn.com> wrote:

........
>How about
>
>    a = foo()
>    a.setAttr('i', 'self.SomeFunction()')
>    a.setTemplate("'A test line %d' % self.i")
>    a.setTemplate("'Same value as above %d' % self.i")
>
>?  Then (a) the code is easier to write; and, (b) when it all falls apart
>you can at least find your mutations in an editor via searching for
>"setAttr", or even turn on a debug flag causing setAttr to log what it's
>doing.
>
>different-operations-different-methods-ly y'rs  - tim
>

I'd thought along this line -- I'd even though about implementing a
stack within foo and then you could push, pop+push and pop args to and
fro -- this would do away with having to declare stuff (which the
setAttr() is really still doing) but would lead to the usual stack
management problems (as well as being hellish to understand, as you
intimated). Still and all, having i = 3 return 3 would be the nicest
for this app ..

Tnx
Alanl



From ivanlan at callware.com  Fri Jul  2 18:22:10 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 2 Jul 1999 22:22:10 GMT
Subject: Slicing beyond list --  bug or featurette?
References: <377D234A.4C9BDAC0@cisco.com>
Message-ID: <377D3B92.A5109CB9@callware.com>

Hi All--

Pierre Fortin wrote:
> 
> Just ran into a situation which can be summarized as follows:
> 

[bobbitt]

> list[2:50] returns [ 2,3,4 ]
> list[20:] or list[20:30] returns []
> 
> Can I count on this, or should I code try/except for the day it's
> fixed?  This 'feature' reduces the error situations; but is it
> correct/intended?
> 

I don't presume to speak for Guido, but the way slicing is implemented
it seems likely that you can count on it: list[:] is implemented as if
it were list[0:MAXINT].  I doubt that Guido will change it, but then,
the Guido is the Guido.

<but-they-don't-call-him-guido-svetz-for-nothing>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From emile at fenx.com  Wed Jul 21 10:31:52 1999
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 21 Jul 1999 07:31:52 -0700
Subject: String2List: odd function
References: <199907211304.IAA14937@cro.ots.ac.cr>
Message-ID: <lTkl3.3901$VL2.416769@news.direcpc.com>

Isidor,

Maybe I'm missing something, but:

>>> a = eval ("[['itema1', 'itema2'], 'itemb1', ['cheese1',
['cheese1a','cheese1b'], '2'], 'blah', 'rest of itemb2', 'zib']")
>>> a
[['itema1', 'itema2'], 'itemb1', ['cheese1', ['cheese1a', 'cheese1b'],
'2'], 'blah', 'rest of itemb2', 'zib']


--

Emile van Sebille
emile at fenx.com
-------------------


Isidor <rodisi01 at my-deja.com> wrote in message
news:199907211304.IAA14937 at cro.ots.ac.cr...
> Dear Pythonographers -
>
> Here is a very odd little something I have written in the process
> of trying to make ModuleWriter module (see previous message) deal
> with "documentary" strings.
>
> Takes a list-like string, e.g.,:
>
> "[['itema1', 'itema2'], 'itemb1', ['cheese1', ['cheese1a',
> 'cheese1b'], '2'], ''blah' rest of itemb2', 'zib']"
>
> and returns *real* lists:
>
> [['itema1', 'itema2'], 'itemb1', ['cheese1', ['cheese1a',
> 'cheese1b'], '2'], "'blah' rest of itemb2", 'zib']
>
> Which, in text-world, looks like nothing more than the removal of
> some quotation marks, but in Python-world...well, I don't have to
> tell you what the difference is.
>
> This function is probably a poster child for code-in-need-of-
> optimization. I have probably written the slowest possible
> version (actually, that's a boast to which I don't have the right
> to make either!)...I have written the most obvious clunker-headed
> version. I will be interested to hear suggestions for making it
> speedier and svelter. (I suspect that Guido's optimizations
> essay(s) may be useful here. I will have to review them.)
>
> GETTING IT: (repeated from my ModuleWriter message)
>
> I don't have web space where I can put this right now, so e-mail
> me at rodisi01 at my-deja.com, and every day or every other day, I
> will send a copy off to whoever has requested it. I know this is
> a hassle, I'm sorry. If it looks like there is serious interest
> in this thing, I will find some web space for it so you can click
> and save to your heart's content. In fact, I thought I heard
> someone in this list talk about offering web space for python
> projects....;)...
>
> ALTERNATIVELY:
>
> If someone will point me to (or give me) some suggestions re.
> posting code to the newsgroup, I could just do that. I'm worried
> about line breaks and all that stuff. Would an ascii attachment
> work in deja-world? (I've never had much access to newsgroups, so
> this is all sort of new to me, and I have to use the free and
> powerful (many many thanks!) but somewhat frustrating deja-news
> interface.)
>
> I hope someone finds this interesting or at least amusing!
>
> Isidor
>
>
>
>





From greg.ewing at compaq.com  Wed Jul  7 23:07:24 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 08 Jul 1999 15:07:24 +1200
Subject: Bizarre Tkinter radio button behaviour
Message-ID: <378415EC.9F7A87A5@compaq.com>

I have some sets of radio buttons linked to
Tkinter variables. I set the initial values of
the variables, expecting the corresponding 
buttons to come up selected when the window
first appears.

But it doesn't happen.

I mess around trying different things. I
add a command to one of the radio buttons to
print out the values of the variables when
they change. All of a sudden the buttons
start appearing with default selections.

Great, I think, and take out my debugging code...
and they stop working again!

After some more trial and error I find that
if *one* of the radio buttons has a command
attached -- even just an empty procedure --
then *all* of the default buttons come up
properly selected. Otherwise, none of them
do.

Has anyone else noticed this behaviour?
Is it a bug in Tk or Tkinter? Or have I done
something silly somewhere?

Python 1.5.2 and Tk 8.0, btw.

Shaking his head in bewilderment,
Greg



From cbbrowne at news.hex.net  Wed Jul 14 22:03:53 1999
From: cbbrowne at news.hex.net (Christopher Browne)
Date: Thu, 15 Jul 1999 02:03:53 GMT
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <377B454C.6F59@mailserver.hursley.ibm.com> <7liguf$58o$1@Starbase.NeoSoft.COM> <m0JTEFA0DTf3EwLV@ugglan.demon.co.uk> <slrn7nss0n.5u0.wtanksle@dolphin.openprojects.net> <378077E1.5E65@mailserver.hursley.ibm.com> <slrn7o58l7.4vr.wtanksle@dolphin.openprojects.net> <3783305C.33FA@mailserver.hursley.ibm.com> <slrn7o7d7o.6la.wtanksle@dolphin.openprojects.net> <3785B997.2001@mailserver.hursley.ibm.com>  
  <w6btdmw4mm.fsf@wallace.nextel.no> <3789ABAB.1E71@mailserver.hursley.ibm.com> <slrn7okdnn.ep3.wtanksle@dolphin.openprojects.net> <n7vi3.21348$AU3.605286@news2.giganews.com> <378CCAD2.97D4FEA7@research.att.com>
Message-ID: <dkbj3.37476$AU3.735796@news2.giganews.com>

On Wed, 14 Jul 1999 17:37:22 GMT, Garrett G. Hodgson
<garry at research.att.com> wrote: 
>Christopher Browne wrote:
>> On Mon, 12 Jul 1999 18:41:59 GMT, William Tanksley
>> <wtanksle at dolphin.openprojects.net> wrote:
>> >(although I am curious as to whether an indent/dedent based syntax might
>> >be amusing in Lisp).
>> 
>> That would provide opportunity for the implementation to be criticized
>> inanely by *absolutely everyone.*
>> 
>> Not only those that see Lisp as "too many parens," but also those that
>> knee-jerk into thinking that Python must be worthless due to
>> "indentation-sensitivity."
>
>if you could only add in the required use of all combinations of
>[$&@_;{}] as well, you might have the perfect language for flamebait.

You mean I *forgot* to indicate that the type of each object needs to
be attributed with its unique punctuation?

How forgetful of me.

defun myfunc
   let
       name  @_[1]
       dest  @_[2]
     if
        exist-p $dest
	print "$name" >> *dest
        error (string-append "Error: " $dest" " does not exist!\n")
-- 
"Terrrrrific." -- Ford Prefect
cbbrowne at ntlug.org- <http://www.ntlug.org/~cbbrowne/langscript.html>



From fdrake at cnri.reston.va.us  Fri Jul 23 10:42:13 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 23 Jul 1999 14:42:13 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
In-Reply-To: <7n82u1$jid$1@nntp.Stanford.EDU>
References: <001101bed3f2$5e8874a0$642d2399@tim>
	<7n675p$qsk@dfw-ixnews7.ix.netcom.com>
	<14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>
	<7n82u1$jid$1@nntp.Stanford.EDU>
Message-ID: <14232.32581.185996.363686@weyr.cnri.reston.va.us>

Amit Patel writes:
 > Here's a gotcha with import *:
 > 
 > import re
 > from string import *  # Great whenever you're doing string processing
 > re.compile('abc')

Amit,
  This specific problem isn't present in the version in the CVS
repository; was it still a problem with the 1.5.2 release?  Looking
through string.py, I didn't see any other problems of unintentionally
exported symbols.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From tim_one at email.msn.com  Wed Jul 28 22:28:03 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 28 Jul 1999 22:28:03 -0400
Subject: Bit-twiddling
In-Reply-To: <wkso68xzat.fsf@ifi.uio.no>
Message-ID: <000f01bed969$fbed9f00$71a22299@tim>

[Lars Marius Garshol]
> I'm translating a numerical algorithm from C to Python, but have run
> into a problem in that the algorithm seems to rely on the fact that
> shifting, addition and multiplication all 'wrap' when the result
> exceeds 32 bits.
>
> Since the Python equivalents do not,

int left shift in Python is end-off and 0-fill and non-complaining (same as
in C); int right shift in Python is sign-extending (and undefined in C for
negative signed ints).  Python int + and * do gripe about overflow.

> does there exist a common workaround for this?

The only easy workaround is to use longs, converting to int only when
necessary or expedient.  C defines unsigned int arithmetic to act "as if"
modulo the wordsize, and that's what long Python arithmetic (followed by
low-bit masking) does.

> For example, I'd like
>
>     mt[0]= seed & 0xffffffff
>     for mti in range(1,n):
>         mt[mti]=(69069 * mt[mti-1]) & 0xffffffff
>
> to never put anything that doesn't fit in an unsigned long in the mt
> list.

Provided that seed was a long to begin with, this code almost works as is.
You just need to append "L" to the literals (for 69069 that isn't necessary,
it just avoids the repeated expense of runtime widening; for 0xffffffff it
*is* necessary on a 32-bit machine, else 0xffffffff == -1 gets widened
to -1L and then the masking doesn't accomplish anything (i & -1L == i for
all long i)).

The elements of mt are then all longs, but each one's *value* fits in a
32-bit C unsigned long:  numerically, it computes the same values as would
the C code.

if-you're-really-after-random-numbers-see-ivan-frohne-ly y'rs  - tim





From fredrik at pythonware.com  Wed Jul  7 08:17:39 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 7 Jul 1999 12:17:39 GMT
Subject: Win32: silent script
References: <37809062.6D463928@phoenix-edv.netzservice.de> <7ls61c$3nu$1@Starbase.NeoSoft.COM> <37820897.93F4ED5D@phoenix-edv.netzservice.de>
Message-ID: <002701bec872$b7931810$f29b12c2@secret.pythonware.com>

Holger Jannsen <holger at phoenix-edv.netzservice.de> wrote:
> Problem is that Windows link files with extensions like .py and .py
> to python.exe (take a look at registry!). This causes in installation
> of python under Windows. Don't like to change registry!
> 
> To start pythonw.exe in that links in windows-menu I've to specify
> the whole path to that file... Ouch, I really didn't want that!;-(

from Misc/NEWS:

    - The Windows configuration adds a new main program, "pythonw", and
    registers a new extension, ".pyw" that invokes this.  This is a
    standard Python interpreter that does not pop up a console window;
    handy for pure Tkinter applications.  All output to the original
    stdout and stderr is lost; reading from the original stdin yields
    EOF.  Also, both python.exe and pythonw.exe now have a pretty icon
    (a green snake in a box, courtesy Mark Hammond).

</F>




From mithy at my-deja.com  Wed Jul  7 02:06:38 1999
From: mithy at my-deja.com (Preston Landers)
Date: Wed, 07 Jul 1999 06:06:38 GMT
Subject: Emacs: highlighting "self" tweak
Message-ID: <7luqpa$3s9$1@nnrp1.deja.com>

Hello,

This is what I did to highlight 'self' in my Python files.

It's a small (unified context) patch against python-mode.el version
3.90, the lastest as far as I know.  It shouldn't be too hard to patch
it against older versions of python-mode (you only need to insert 3
lines at a certain spot.)

I wish I knew of an easy way to make it a hook instead of a patch.  If
someone can figure that out, I would *love* to hear from you.

I decided to make it color 'self' as a type (same color as classes),
because that's the best I could think of.

If this patch comes across mangled or whatever, you can always pick
it up at my Python page:

http://askpreston.com/projects/python/

You can also find my silly little Emacs/Python shortcut hook there.

cheers,

---Preston

--------cut here

--- python-mode.el	Tue Jul  6 23:37:55 1999
+++ /usr/share/xemacs20/site-lisp/python-elisp/python-mode.el	Tue Jul  6
23:30:48 1999
@@ -316,6 +316,9 @@
      ;; functions
      '("\\bdef[ \t]+\\([a-zA-Z_]+[a-zA-Z0-9_]*\\)"
        1 font-lock-function-name-face)
+     ;; also color "self" as a 'type'
+     '("\\b\\(self\\)[ .,]?"
+       1 font-lock-type-face)
      ))
   "Additional expressions to highlight in Python mode.")
 (put 'python-mode 'font-lock-defaults '(python-font-lock-keywords))



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From larsga at ifi.uio.no  Wed Jul  7 15:04:38 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 07 Jul 1999 21:04:38 +0200
Subject: xml-0.5.1 and comments
References: <378217AA.2ED7E731@pcraster.nl>
Message-ID: <wk908sqo1l.fsf@ifi.uio.no>

* Cees Wesseling
| 
| But the DOM walker seems not to pass comment's through.  What I want
| is an exact copy of the input xml minus the transformations I want
| to made. But in Walker overriding doComment does not trigger any
| comments.

Probably because the DOM builder you use does not pass on comments.
Most of them are built on top of SAX, which does not report comments
(see below).

If you really want this you can write a DOM builder on top of your
favourite parser. Should be pretty easy to do and you can look at
xml.dom.builder and the SAX driver for that parser for guidance.
 
| Also wading to some of the SAX code, I started wondering if I can do
| this with SAX. 

SAX 1.0 does not report comments, since they're not part of the
logical document. SAX 2, which will have lexical information beyond
the pure logical stuff, will be able to report comments.

| Does anyone has some sample code, for DOM or SAX that does nothing
| else then replicating the xml input exactly, is that possible?

No. Stuff like whitespace in tags etc etc is not available through the
DOM or SAX, so you can't do that. With xmlproc 0.62 you should be able
to do it since it will let you peek at the raw markup (although I
think you'll lose the entity structure).

BTW: you may want to subscribe to the XML-SIG to discuss this sort of
thing. 

--Lars M.



From gerrit.holl at pobox.com  Thu Jul  1 15:14:50 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 1 Jul 1999 21:14:50 +0200
Subject: Why does shelve make such large files?
Message-ID: <19990701211450.A6827@optiplex.palga.uucp>

Hi,

Is it really necesarry for shelve to make such large files?
Have a look at this:
/tmp> python
Python 1.5.2 (#1, Apr 18 1999, 00:16:12)  [GCC 2.7.2.3] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import shelve
>>> d = shelve.open('database')
>>> d['key'] = 'value'
>>>
/tmp> ls -l database
-rw-rw-r--   1 gerrit   gerrit      16384 Jul  1 21:13 database
                                    ^^^^^

16 KB for only one key!!?
pickle seems to make _much_ smaller files!

Why is this?

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From stalnaker at acm.org  Thu Jul  1 01:11:10 1999
From: stalnaker at acm.org (Max M. Stalnaker)
Date: Wed, 30 Jun 1999 22:11:10 -0700
Subject: Using freeze
References: <EoBd3.6644$rZ4.56779@news4.mia>
Message-ID: <377AF86D.B9C7FCD5@acm.org>

Not an expert, but I have done it using vc++5.

In the freeze directory, under tools, which directory you get when to
download and install the source, which you have to do, there is a html file
that is where you need to go next.

My project plan would be :

1) compile source (to prove things are set up and you know what to do)
2) execute freeze.py on your program, with a command line parameter to put
the result in c:\temp
3) goto c:\temp and type nmake
4) test the result

You might have problems with 6.0 since  5.0 is what is supported out of the
box.
The resulting executionable will be dependent on some external programs.

SunitJoshi wrote:

> Hello
>     I'm trying to make exe's from python scripts using VC6++. However I'm
> lost as to how to use freeze to accomplish this. If someone could include
> some details I would really appreciate it.
>
> thanks
> Sunit

--
Max M. Stalnaker  stalnaker at acm.org
Astar Computer Consulting:  Networking and Custom Accounting Software





From p.t at iname.com  Mon Jul 26 09:13:14 1999
From: p.t at iname.com (Peter Torstenson)
Date: Mon, 26 Jul 1999 10:13:14 -0300
Subject: How to use a variable form main script in a module.
Message-ID: <7nhmt5$su0@enews3.newsguy.com>

This is probaly a silly question from a newbie.
In my main script I put a value for to a variable called "x".
Now I want to use the "x" variable in a module.
The module is imported into the main script. How to do this?

I've been trying to find out how to do it in tutorials an the reference
manual but I can?t find anyting written about this. Perhaps because it's a
to silly question.

/Peter Torstenson

Rio de Janeiro





From cfelling at iae.nl  Sun Jul 11 15:53:33 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 11 Jul 1999 21:53:33 +0200
Subject: Simple Tk8 Problems
References: <37880ED5.2BD7C347@ssec.wisc.edu>
Message-ID: <7masnt$hqc$1@vvs.superst.iae.nl>

Nick Bower <nick.bower at ssec.wisc.edu> wrote:

> 1.  I can't control the size of the Tk root window.  I've tried putting
> ".width=xxx" statements everywhere in the following snippet, but nothing
> seems to work.  (The FileList is a ListBox next to a scrollbar).  Is
> there something I'm missing?.

try to add something like

      self.filelist.config(width=10,heigth=12)

this turns the filelist widget into a 10 chars wide and 12 lines high blob.


> 2. Also, the Listbox in the above code is scrolling all over the place
> tracking the mouse movement as it runs across the top of it.  I don't
> even need to touch the scrollbar and it seems to track mouse movement so
> somehow it's been bound to the mouse as well as the scroll bar.  Weird.

for what it's worth, the snippet you posted behaves as it should, on
my machine atleast.

-- 
groetjes, carel



From rasumner at math5-pc.wisdom.weizmann.ac.il  Wed Jul 14 09:54:28 1999
From: rasumner at math5-pc.wisdom.weizmann.ac.il (Reuben Sumner)
Date: 14 Jul 1999 13:54:28 GMT
Subject: memory leaking?!
Message-ID: <slrn7op5jj.cva.rasumner@math5-pc.wisdom.weizmann.ac.il>

I have some code creating a somewhat complicated data structure.
When I am using Python 1.5.2 compiled with Py_REF_DEBUG I can see
that reference counts are dissapearing (although I can't figure out
what is getting lost, all the objects that I can think of I have
confirmed are being released).  When I compile with Py_TRACE_REFS
then all of a sudden the reference counts don't increase between
subsequent calls.  What on earth is going on here?  Running "top"
and generating a lot of very large structures then memory does seem
to leak no matter which way I compile python (although I think it is
slower with Py_TRACE_REFS).

Reuben

P.S. I did apply a small debugging patch (that I posted here) that just
makes the debugging output optional depending on a new command line switch.




From emile at fenx.com  Sun Jul 25 06:37:12 1999
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 25 Jul 1999 03:37:12 -0700
Subject: Python info please
References: <379AD667.434E9A47@cal.vsnl.net.in>
Message-ID: <9QBm3.4153$VL2.438705@news.direcpc.com>

Hi Arup,

Take a look at http://www.python.org/mailman/listinfo

See you,

--

Emile van Sebille
emile at fenx.com
-------------------


Arup Bhanja <abhanja at cal.vsnl.net.in> wrote in message
news:379AD667.434E9A47 at cal.vsnl.net.in...
> Hi Pythoners,
>
> Goede Morgon,
>
> is there a list for python programming ? Can you give me the
> subscription details?
>
> Regards,
>
> Arup Bhanja
> Calcutta,INDIA.
>
> Currently executing...
> >>>myquotes=[]
> >>>myquotes.append('I love monty python')
> >>>myquotes
> ..done enjoy Python
>
>





From greg.ewing at compaq.com  Tue Jul 20 18:09:38 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 21 Jul 1999 10:09:38 +1200
Subject: Cyclops 0.9.4
References: <000c01bed257$84f52a80$139e2299@tim>
Message-ID: <3794F3A2.DF8BE8FA@compaq.com>

Seems to me that Cyclops (or any other such thing
implemented in pure Python) is incapable of finding
what you *really* want to find -- which is groups
of objects not reachable from anywhere.

Greg



From roy at popmail.med.nyu.edu  Mon Jul 26 16:03:07 1999
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Mon, 26 Jul 1999 16:03:07 -0400
Subject: Parsing oracledb errors
Message-ID: <roy-2607991603070001@qwerky.med.nyu.edu>

I've built my database with lots of named constraints.  I'd like to be
able to figure out, from the error message I get in a try/except
oracledb.error, exactly what constraint I violated.  For example, if I
get:

EXECUTE caused a ORA-00001: unique constraint (NETDB.U_MACHINE_HOSTNAME)
violated

I want to pull out U_MACHINE_HOSTNAME, which tells me exactly what went
wrong.  Is there any easy way to do this, without having to parse every
error message string?  It's not so much that it's a pain to have to parse
them, but I'm not even sure that they'll all be in a regular format, so
parsing them may be impossible.

-- 
Roy Smith <roy at popmail.med.nyu.edu>
New York University School of Medicine




From paul at prescod.net  Mon Jul  5 19:02:27 1999
From: paul at prescod.net (Paul Prescod)
Date: Mon, 5 Jul 1999 23:02:27 GMT
Subject: Unicode via COM
Message-ID: <37813983.7F7D677A@prescod.net>

I can write characters into a PowerPoint slideshow via range.InsertText
but it takes a Python string as an argument. How do I get a >255 character
into Unicode. I'm looking for something like unicode_chr( 8132 )

I guess it has something to do with UnicodeFromRaw but I don't know what
"Raw" means? UTF-8? Do I have to memorize the details of some encoding?
How do I convert 8132 into a "raw" string?

-- 
 Paul Prescod  - ISOGEN Consulting Engineer speaking for only himself
 http://itrc.uwaterloo.ca/~papresco

Math -- that most logical of sciences -- teaches us that the truth can
be highly counterintuitive and that sense is hardly common.
	K.C.Cole, "The Universe and the Teacup"



From mithy at my-deja.com  Sun Jul 18 01:37:23 1999
From: mithy at my-deja.com (Preston Landers)
Date: Sun, 18 Jul 1999 05:37:23 GMT
Subject: Python 1.5.2 + Linux 2.0.36 == Bad Combo?
References: <slrn7ou0ru.3sp.kas@titan.magnetic-ink.dk> <378F31D6.99BF0197@arlut.utexas.edu> <slrn7oufnu.5ng.kas@titan.magnetic-ink.dk>
Message-ID: <7mrp6i$rhd$1@nnrp1.deja.com>

In article <slrn7oufnu.5ng.kas at titan.magnetic-ink.dk>,
  kas at maps.magnetic-ink.dk wrote:
> Considering that Preston is able to reproduce the Kernel Oops using
the
> same kernel and python version as I do, I belive it's something deeper
> than the python installation.  But thanks anyway, I'll take a look at
the
> URL. :-)

Hey, this is Preston, I wrote one of the programs Klaus is having
trouble with.

I think his original post didn't mention that Pagecast uses threads --
though there is little to no "voodoo" involved; it's pretty
straightforward.

this particular problem (the segfault and kernel oops with Python 1.5.2
and Linux 2.0.36) seems to happen when the number of threads increases
above four.  In fact, it seems to happen when the fourth thread *exits*.
GDB sez: "error accessing memory address 0x4008A408 : no such process."

It's not a memory or hardware problem as someone else suggested.  I'm
able to reproduce the bug exactly on my own machine, with a different
distribution no less.  And I've stress tested my RAM extensively.

I still haven't had a chance to really dive into debugging Pagecast
under Python 1.5.2 and Linux 2.0.36 yet but I will (probably tomorrow.)
I normally run Linux 2.2.x.   Klaus, I'll do that soon and post to the
Pagecast list and/or here.

BTW if anyone is interested (quick plug) Pagecast is a Python program
that submits lists of URLs to internet spiders like ALtaVista, Hotbot,
etc.  http://askpreston.com/projects/pagecast

Thanks,

--
||  Preston Landers <preston at askpreston.com>  ||
||           http://askpreston.com            ||


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gmcm at hypernet.com  Sat Jul 17 10:06:41 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 17 Jul 1999 09:06:41 -0500
Subject: I envy newbies
In-Reply-To: <000101bed00d$ab59e4a0$d2a22299@tim>
References: <1279991447-21085630@hypernet.com>
Message-ID: <1279905910-26230576@hypernet.com>

[Gordon McMillan]
> > ...
> > I think the Tcl install does modify your path.

[Tim] 
> It may under NT (can't remember), but definitely does not under W95.

Perhaps leftover from an install of an earlier version.

> > But Python will go looking for it in the default places. Which is,
> > of course, C:\Program Files\...
> 
> Which should really be taken from the registry, at key
> 
>     SOFTWARE\Microsoft\Windows\CurrentVersion\ProgramFilesDir

Ah, there it is, key  HKLM\SOFTWARE\Microsoft\Windows\CurrentVersion
value ProgramFilesDir. I wonder who else doesn't know that <wink>.

> > [I think it's hilarious that MS was so enamored of their "advance" in
> > being able to put spaces in paths that they made the "proper" place
> > to install everything a demo of this "feature"...
> > Which demos the fact that spaces in path names are a pain in the 
> > ass].
> 
> yup!-and-they're-regretting-it-all-the-way-to-the-bank-ly y'rs  -
> tim

Well they fixed it before producing /F's version of NT.

best-argument-for-being-a-swede-<wink>-ly y'rs

- Gordon



From arcege at shore.net  Fri Jul 23 08:06:44 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Fri, 23 Jul 1999 12:06:44 GMT
Subject: string.replace
References: <Pine.BSF.4.10.9907221514460.24906-100000@soda.CSUA.Berkeley.EDU>
Message-ID: <oVYl3.990$ui4.311547@news.shore.net>

marlon <marlon at CSUA.Berkeley.EDU> wrote:
: I tried the same routine on a machine with 1.5.1 and it worked fine,
: you're right about the version being the problem. Thanks for your help.
:   marlon

: On Thu, 22 Jul 1999, Gordon McMillan wrote:

:> marlon asks:
:> 
:> > can anyone tell me why this breaks?
:> > 
:> > Python 1.4 (Nov  4 1997)  [GCC 2.7.2.3]
:>          ^^^
:> 
:> > Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
:> > >>> import sys, string
:> > >>> line = 'cat "in" the hat'
:> > >>> line = string.replace( line, '"', '')
:> > Traceback (innermost last):
:> >   File "<stdin>", line 1, in ?
:> > AttributeError: replace
:> > >>> 
:> 
:> I believe string.replace came with 1.5.
:> 
:> - Gordon

You can use string.translate and string.maketrans instead.

In Python 1.5:
  exclamation = 'nu'
  exclamation = string.replace(exclamation, 'u', 'i')

In Python 1.4 (and earlier):
  exclamation = 'nu'
  knights_map = string.maketrans('u', 'i')
  exclamation = string.translate(exclamation, knights_map)

  -Arcege




From gmcm at hypernet.com  Tue Jul 27 22:43:00 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 27 Jul 1999 21:43:00 -0500
Subject: 1.5.2 for: else:
In-Reply-To: <slrn7psisn.da.wtanksle@dolphin.openprojects.net>
Message-ID: <1278996512-12636218@hypernet.com>

William Tanksley wrote:
> On Tue, 27 Jul 1999 18:35:10 -0500, Gordon McMillan wrote:

> >Maybe it violated your expectations, but it's much more valuable than 
> >what you expected. After all, testing for an empty sequence is a 
> >no-brainer. But finding a match in a list, and testing whether you 
> >fell off the end without finding one, is (without the else clause) a 
> >much messier proposal.
[rant]
> I would also disagree with you that my interpretation would result
> in disutility: you'd merely have to use exceptions instead of
> 'break'.  
[more rant]
> class neverMind(): pass
> 
> try:
>  for x in [1,2,3]:
>   print x
>   if x == 2: raise neverMind
> except neverMind:
>   print "modern else clause here"

--------billy.py--------------------
class neverMind: pass

try:
 for x in [1,2,3]:
  print x
  if x == 2: raise neverMind
except neverMind:
  print "modern else clause here"

for x in [1,2,3]:
 print x
 if x == 2:
  break
else:
 print "Hi Billy!"
---------------------------------------------------

C:\TEMP>python billy.py
1
2
modern else clause here
1
2

C:\TEMP>

Notice the difference?
You need to raise the exception if you've gone through the loop and 
not found 2. The best way to do that <snicker> is with an else 
clause.

Or will you now claim that:

lst = [1,2,3]
for i in range(len(lst)):
  if lst[i] == 2: break
if i == len(lst):
  print "Tanksley else clause here"

is a better solution <wink>?

insufferably-sanctimoniously-y'rs

- Gordon



From thornley at visi.com  Sun Jul  4 20:31:33 1999
From: thornley at visi.com (David Thornley)
Date: Mon, 05 Jul 1999 00:31:33 GMT
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <m3btdvhrlt.fsf@localhost.localdomain> <d2Of3.1233$U5.237262@ptah.visi.com> <377fc323.69785805@nntp.ix.netcom.com>
Message-ID: <F1Tf3.1270$U5.242370@ptah.visi.com>

In article <377fc323.69785805 at nntp.ix.netcom.com>,
Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote:
>On Sun, 04 Jul 1999 18:50:49 GMT, thornley at visi.com (David Thornley)
>declaimed the following in comp.lang.python:
>
>
>> "Thinking Forth" is an excellent book to read.  So is Stroustrup's
>> "Design and Evolution of C++".  I haven't seen a similar book about
>> the Lisp way of thinking, not counting books specifically on learning
>> Lisp and doing it well.  Then again, I find Common Lisp revolutionary
>> enough sometimes.
>>
>	Unfortunately, it predates much of what is now considered to be
>LISP.
>
>	"Anatomy of LISP", John Allen, (1978 McGraw-Hill Computer
>Science Series).
>
>	While using LISP, it was started as a data structures type
>course...
>
I am unfamiliar with that.  I suppose I could track it down.

As for the suggestion of SICP, all I can do is plead a temporary
memory failure.  I should have thought of that unprompted.


--
David H. Thornley                        | If you want my opinion, ask.
david at thornley.net                       | If you don't, flee.
http://www.thornley.net/~thornley/david/ | O-



From aaron_watters at my-deja.com  Wed Jul 14 09:03:13 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Wed, 14 Jul 1999 13:03:13 GMT
Subject: Gadfly accessing results
References: <7mh2ed$hrb@newsops.execpc.com>
Message-ID: <7mi1qe$gi3$1@nnrp1.deja.com>

In article <7mh2ed$hrb at newsops.execpc.com>,
  Sim & Golda Zacks <simngolda at voyager.net> wrote:
> When I do a Select query in Gadfly how would I then be able to use
> individual fields from the results? In Delphi (my primary programming
> language) I would use a query and then call
> query.fieldbyname('fielname').value and I have access to all the
fields
> that were selected. How would I do this in Gadfly? For example my
select
> is for 5 fields and it returns 2 records. I want to be able to get
each
> field individually from each record

For gadfly, like other dbapi conformant python-database interfaces
you execute a query via a cursor

    curs = connection.cursor()
    ...
    curs.execute("select col1, col2, col3 from table")

Get the results of the select as a list of tuples via

    resultlist = cursor.fetchall() # and similar

Get the 1st result row via

    first = resultlist[0]

the third via

    third = resultlist[2]

to get the 2nd column data value of the third row:

    col_3_2 = third[1]

to loop through all the data values by row and then by column:

    for row in resultlist:
        for col in row:
            print col,
        print

Please see the gadfly documentation (with the package)
and the dbapi specification available in the db-sig area
of http://www.python.org.

   -- Aaron Watters http://www.chordate.com

===

Do you expect me to talk?
No, Mr. Bond, I expect you to die.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Brian at digicool.com  Fri Jul  9 09:58:17 1999
From: Brian at digicool.com (Brian Lloyd)
Date: Fri, 9 Jul 1999 13:58:17 GMT
Subject: Python for the PalmPilot and Other Devices
Message-ID: <613145F79272D211914B0020AFF6401914DBC1@gandalf.digicool.com>

> You'd have to ask Brian Lloyd the original reasons for porting
> Python to his pathetic little 2MB CE device. ;-)  Now that CE's
> got 16-32MB direct from the factory, well, grins all around.

Well, to see if I could do it, of course :^) And because I happened
to have a CE device, and not a Pilot.

> The CE devices have some advantages.  There is a wide diversity 
> in models (palm sized, handhelds, even laptop sized units) and
> vendors (at least a dozen or so).  

They also have some disadvantages - for example, there is a wide
diversity in models (palm sized, handhelds, even laptop sized units)
and vendors (at least a dozen or so)...  :^)  Configuration mgmt
has been a big pain in the a** for the CE port, and it's only made
worse by the fact that everyone working on it has a pesky "real job" 
and tend to be very busy.


> The Win32 development heritage
> is reasonably well-understood, if often cursed.  The CE platform 
> includes a lot of built-in features (support for TCP/IP, sockets, 
> IrDa).  Microsoft is using its clout to move CE into automobiles 
> and cable set-tops.  True, CE represents a capitulation to a well
> known software vendor, but ...

He he - when I first embarked on that, I thought that the built in
features would be great help -- until I realized that there were
some missing features like, oh, most of the std C library that
Python takes for granted... :^)

Seriously though, the successive generations of CE devices have
been steadily piling more and more desktop win32 functionality
in - you could say that this is good or bad, but in the long run
it is probably good. Once CE finishes turning into NT in a handheld
(never mind that it will need to come with 256M of ram and need to
be powered by a carry-along car battery), we'll be able to use most
of the existing win32 port code pretty much unchanged :)

There are actually some potentially interesting opportunities for Python

on CE. For example, I believe that the upcoming sega Dreamcast video 
game unit will run CE as it's os, so it's not out of the realm of 
possibility that one could use Python for video game prototyping
or something.

I personally am waiting for a decent AutoPC so I can write a Python
script to drive me to work so I can get some shuteye on the way 
in... :^)


Brian Lloyd        brian at digicool.com
Software Engineer  540.371.6909              
Digital Creations  http://www.digicool.com 






From kranio at nospam.nospam.it  Fri Jul 23 10:10:47 1999
From: kranio at nospam.nospam.it (Kranio)
Date: Fri, 23 Jul 1999 14:10:47 GMT
Subject: new italian python group
Message-ID: <37987773.6511212@news.tiscalinet.it>

This is the link to italian python group
http://web.tiscalinet.it/python


Marco Buzzo



From quinn at lira.ugcs.caltech.edu  Fri Jul 16 19:07:57 1999
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 16 Jul 1999 23:07:57 GMT
Subject: Accessing an object name ?
References: <378F0BB5.9AD878D@cellware.de>
Message-ID: <slrn7oveqh.4bc.quinn@lira.ugcs.caltech.edu>

On Fri, 16 Jul 1999 10:38:45 +0000, bernard <bh at cellware.de> wrote:
>
>Is there any way of mapping from an object id to an object name
>so that given the object declaration:
>
>obj1 = myClass()
>
>I can, within an instance method, access the object name 'obj1' ?
>
>bernard.

No, not really.  Actually, you can, but it's a bad idea.  You could do this:

class Foo:
	def name(self):
		g = globals()
		for key in g.keys():
			if g[key] is self:
				return key

but you really don't want to, because this will break as soon as you use
Foo from a different module.  This is because there's no real "global" in
python, just module global, so Foo won't be able to get ahold of the
right namespace.  Besides, an object can be bound to more than one name,
or to no name at all (array[0] = Foo()).

In simililar circumstances, I've done stuff like:

obj1 = myClass('obj1')

Which is cleaner anyway.

OTOH, you can get class and method names:

myClass.__name__ == 'myClass'
myClass.method.__name__ == 'method'

Binding via class or def is more special than =, I guess.



From dwolff567 at my-deja.com  Tue Jul 27 15:50:03 1999
From: dwolff567 at my-deja.com (dwolff567 at my-deja.com)
Date: Tue, 27 Jul 1999 19:50:03 GMT
Subject: [Tkinter] Inheriting from Toplevel
Message-ID: <7nl2h8$keo$1@nnrp1.deja.com>

I am having some fairly weird problems with my Tkinter app.  I am
spawning a new Toplevel window where I display a graph in a
canvas instance.  However, when I send the data to the widget to
be plotted, the whole thing freezes.  Since the plotting class works
fine when run by itself, I can only assume that I have some problems
with the way that the windows are instantiated.  Here is a general
outline:

The RootInterface inherits from Frame:

class RootInterface(Frame):

    def __init__(self,parent=None):
	Frame.__init__(self,parent)
	self.pack()
        ...

which is instantiated in the usual way:

if __name__ == '__main__':
    root = Tk()
    main = RootInterface(root)
    main.pack()
    root.mainloop()

Now, when a menu item is chosen from RootInterface, the Plot window
is brought up:
self.dfWindow = DFWindow(self)

where DFWindow inherits from Toplevel:

class DFWindow(Toplevel):

    def __init__(self,master):
        Toplevel.__init__(self,master)
        self.plot = Plotter(self,0.0,md.cvar.dfcutoff,0.0,0.0,
                            "Distribution Function")
        self.plot.pack()

and the self.plot is an instance of Plotter which inherits from
Frame and contains the canvas and plotting methods.  When I want to
plot something, a method in RootInterface is called which calls a
method in DFWindow which finally calls the plotting method in Plotter.

My question is, am I doing something wrong that would cause the
strange behavior that I described earlier?  I apologize for the length
of this message, but I am at my wits end.

Thanks,

   Dave




Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From werner at mosaic.co.za  Thu Jul 22 15:07:37 1999
From: werner at mosaic.co.za (Werner Fouche)
Date: Thu, 22 Jul 1999 21:07:37 +0200
Subject: Python / MS ADO
Message-ID: <7n7p91$p92$1@hermes.is.co.za>

Hi,

In article http://www.python.org/pipermail/db-sig/1997-April/000197.html I
read the following:

> One interesting note: the ODBC module is apparently quite fast
> relatively to other mechanisms available to Python for accessing
> databases (e.g. ADO). You can *make* ADO go faster, but it takes a good
> amount of fine tuning and voodoo. Empirically, the ODBC module is simple

I am interested in using ADO from Python on Windows NT. Any directions to
where I can obtain a demo Python ADO application would be greatly
appreciated.

Thanks,

Werner






From mwh21 at cam.ac.uk  Tue Jul 27 07:28:45 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: Tue, 27 Jul 1999 12:28:45 +0100
Subject: Python and Red Hat Linux 6
In-Reply-To: <379D84E0.5F872B@arlut.utexas.edu>
References: <379CD547.22A10A50@postoffice.pacbell.net> <379CD5C2.5BA096A2@mail.danbbs.dk> <379d7fcf.52695@news.tiscalinet.it> <379D84E0.5F872B@arlut.utexas.edu>
Message-ID: <Pine.LNX.4.10.9907271220420.19094-100000@atrus.jesus.cam.ac.uk>

On Tue, 27 Jul 1999, Thomas Bryan wrote:

> Kranio wrote:
> 
> > I have the same problem and this solution won't works. If you type
> > 
> > $ file.py
> > bash: ./file.py : Command not found
> 
> file.py is not in your PATH and isn't found.
> 

No, no, that would make sense; you get an error like this if file.py is
found and is executable and has a first line of the form

#!/path/to/executable

and /path/to/executable doesn't exist.

HTH
Michael




From Carey.Evans at p98.f112.n480.z2.fidonet.org  Sun Jul  4 18:24:25 1999
From: Carey.Evans at p98.f112.n480.z2.fidonet.org (Carey Evans)
Date: Sun, 04 Jul 1999 23:24:25 +0100
Subject: Python GUIs: Summary and Conclusion
Message-ID: <000016b9@bossar.com.pl>

From: Carey Evans <c.evans at clear.net.nz>

Greg Ewing <greg.ewing at compaq.com> writes:

> I'd certainly be in favour of a native
> Tk interface, especially if it became the 
> "standard" one included in the distribution.
> Even if it wasn't any faster, I'd feel
> better while using it!

There's another place this has been done, that I haven't seen
mentioned.  Perl/Tk has no dependencies on Tcl - it's source is
generated by a collection of Perl scripts that work on the distributed
Tcl C source and convert it to XS, and all the Tcl parts have been
reimplemented in Perl.

It might be reasonable to do the same for Python.

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

			 print [<>]->[rand$.]



From spamkiller at epstein.co.uk  Wed Jul 28 09:24:14 1999
From: spamkiller at epstein.co.uk (Dani Epstein)
Date: Wed, 28 Jul 1999 14:24:14 +0100
Subject: Dreaded Newbie Question
References: <S3i8NRAzoFm3EwEN@epstein.co.uk>
 <000701bed723$82f0d100$492d2399@tim>
Message-ID: <vSwjkBA+Rwn3Ew5p@epstein.co.uk>

In article <000701bed723$82f0d100$492d2399 at tim>, Tim Peters
<tim_one at email.msn.com> writes
>It's a little less Unix-centric now, but that does go deep.  Python was
>first written on a Mac!  Which is why Unix shows through everywhere <wink>.

How marvellously straightforward this whole computing thing is.

>> Thanks to all for your replies. This adventure looks like its going to
>> mean good company as well as fun!
>
>I found both to be true.  Let us know when you get stuck.

I most certainly will - thanks once again.
-- 

Dani Epstein




From gward at cnri.reston.va.us  Wed Jul 21 17:44:44 1999
From: gward at cnri.reston.va.us (Greg Ward)
Date: 21 Jul 1999 21:44:44 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk>
Message-ID: <7n5f0c$snu$1@ffx2nh5.news.uu.net>

sreekant kodela <sreekantk at freenet.co.uk> wrote:
> - Is there anyway I can improve the performance of python code.

It depends what the performance problem you're worried about is.  If
you're concerned about startup overhead of CGI scripts (and you should
be), then Python is certainly guilty, but a damn sight better than a
certain other popular scripting language widely used for CGI programming
(mainly because Python caches compiled bytecode files).  Apart from
PyApache, one possibility is Persistent CGI, which arose in Zope-land.
Andrew Kuchling recently wrote an Apache module (mod_pcgi.c) for this,
see ftp://starship.python.net/pub/crew/amk/new/.  Sounds like a good
idea, but I've never used it.

If you're concerned about runtime performance -- ie. how long it takes
to mangle strings and loop over lists and so forth -- I wouldn't worry
about it.  Yes, Python is slow, but the Internet is so much slower that
it hardly matters.

        Greg
-- 
Greg Ward <gward at cnri.reston.va.us>
Corporation for National Research Initiatives, Reston, VA, USA
http://starship.python.net/~gward/



From Gary.Stephenson at p98.f112.n480.z2.fidonet.org  Mon Jul  5 04:11:55 1999
From: Gary.Stephenson at p98.f112.n480.z2.fidonet.org (Gary Stephenson)
Date: Mon, 05 Jul 1999 09:11:55 +0100
Subject: Tuples cf. Expanded objects??
Message-ID: <000016c1@bossar.com.pl>

From: Gary Stephenson <stephensong at iname.com>



William Tanksley wrote:

> On Fri, 02 Jul 1999 22:10:46 +1000, Gary Stephenson wrote:
>
> An instance of an immutable type may be created in expanded form.
>

Yes, but can they also  be created in an non-expanded form?  If not, then
there is more of a relationship that your message implies - yes?

cheers,

gary



From marlon at CSUA.Berkeley.EDU  Thu Jul 22 18:15:27 1999
From: marlon at CSUA.Berkeley.EDU (marlon)
Date: Thu, 22 Jul 1999 15:15:27 -0700 (PDT)
Subject: string.replace
In-Reply-To: <1279441446-22507099@hypernet.com>
Message-ID: <Pine.BSF.4.10.9907221514460.24906-100000@soda.CSUA.Berkeley.EDU>

I tried the same routine on a machine with 1.5.1 and it worked fine,
you're right about the version being the problem. Thanks for your help.
							marlon

On Thu, 22 Jul 1999, Gordon McMillan wrote:

> marlon asks:
> 
> > can anyone tell me why this breaks?
> > 
> > Python 1.4 (Nov  4 1997)  [GCC 2.7.2.3]
>          ^^^
> 
> > Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> > >>> import sys, string
> > >>> line = 'cat "in" the hat'
> > >>> line = string.replace( line, '"', '')
> > Traceback (innermost last):
> >   File "<stdin>", line 1, in ?
> > AttributeError: replace
> > >>> 
> 
> I believe string.replace came with 1.5.
> 
> - Gordon
> 

_




From curt at www.journyx.com  Mon Jul 12 20:41:16 1999
From: curt at www.journyx.com (curt at www.journyx.com)
Date: Tue, 13 Jul 1999 00:41:16 GMT
Subject: free python web timesheet service http://pnk.com
Message-ID: <7me1va$3v2$1@nnrp1.deja.com>

http://pnk.com/freetime.html

goto the above url to start using our free web service
install nothing
just use your browser

app totally written in python

http://pnk.com/freetime.html


___________________________________________________________________
FREE Project Management Software??? --> http://journyx.com/wts.html
-------------------------------------------------------------------


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From tismer at appliedbiometrics.com  Tue Jul 20 06:09:05 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 20 Jul 1999 10:09:05 GMT
Subject: How fast can we multiply?
References: <001801bed279$92c348a0$139e2299@tim>
Message-ID: <37944AC1.F913550D@appliedbiometrics.com>


Tim Peters wrote:
> 
> [Christian Tismer]
> > ...
> > Also I was tempted to use "&" on the numbers, in order to save
> > one nbits call. It was just unclear what happens with negative
> > numbers and dropped it.

Waaah, did I say "&"? I used "|" or course.

> It's clear enough what happens with positive numbers, though <wink>:
> 
> m = 1L << 10000
> n = m - 1
> print m & n  # prints 0
> 
> Use "|" instead and you're golden.  But that takes time proportional to the
> number of bits in the larger input, so-- again --we need to fix nbits.  Then

But your "min" suggestion makes still a lot more sense to me.

> > Was happy that negatives went correct without special treatment.
> 
> Yup!  It's a delight.

My first attempt had special "-" treatment, until I recognized
that the shift and mask operations already do all we need. :-)

...
> Na, I think "little interest" is it.  Outside of crypto, prime-number fans,
> and people studying Calendars from Hell <wink>, nobody cares about really
> big ints.  The first two groups-- who care a lot --care *so* much they're
> using hyper-optimized bigint pkgs anyway.  Andrew Kuchling's wrapping gmp
> for Python is the most useful thing that's ever been done in this area.

I see. The gmp makes very much sense. Until now.

On the other hand, if the hard coded "C" basic elements are
made fine and tiny, they give us nuts and bolts to build
the real algorithms on top.

We can therefore do what gmp does, not so much slower, but
readable, in Python. No porting, no nothing, just using.

And again I see a win for Python. It is able to show how
algorithms work, in a very clear way using its clean
syntax, and you can use the implementation of math directly
as highschool text. *And* it just works, and is fast.
Fells very very promising to me.

[matmult]
> Large dense matmult isn't very common.  Sing along:

[lemon tree - sung]

It is just so easy. Factoring matrix multiplication
into its submatrices and then save (I think it was one)
multiplications - it would be a dlight to write in
Python. Not doing so appears to me as to know
quicksort, but using bubble sort all the time,
not just at break-even (10-20).
But I'm telling nothing new to you.

> > Challenge: I claim Python could be a competitor to Mathematica,
> > if only some people were interested.
> 
> Give me one bored programmer with a few weeks to spare, and you can keep a
> whole web full of people who are interested <wink>.

I'm bored enough, just the few weeks are missing. Sigh...

> > *** Hey Python, where are your ambitioned Math students? ***
> >
> > (did they wake up, finally?)
> >
> > waiting-for-the-beginning-to-show-up-ly y'rs  - chris
> 
> i'll-join-you-in-the-bar-ly y'rs  - tim

Reading the last posts, I think we aren't as alone as I thought.

As another little challenge:

Tim already mentioned Jurjen N.E. Bos' high precision math
library. It is in Python's contrib site.
I have the strong impression that it was not used by anybody
since a year now. python.org fails also to advertize it.

It could need a little rework, to make use of 1.5.2's features.
And if someone marries it with a fast multiplication, that
could make a neat tool.

If I were a student, I would propose my advisor to let me do such
things as my semester homework. What a delight :-)

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From Michael.P..Reilly at p98.f112.n480.z2.fidonet.org  Thu Jul  1 11:17:40 1999
From: Michael.P..Reilly at p98.f112.n480.z2.fidonet.org (Michael P. Reilly)
Date: Thu, 01 Jul 1999 16:17:40 +0100
Subject: Why must path be in args in os.execv?
Message-ID: <0000161a@bossar.com.pl>

From: "Michael P. Reilly" <arcege at shore.net>

Holger Jannsen <holger at phoenix-edv.netzservice.de> wrote:
: High,

: there's again something I don't understand. Hope you do.;-)

: I'd like to start a windows-executable with some arguments.
: Ok. 'os.execv' or anything like that is the thing.

: But if I don't repeat the path with the arguments, the execution fails.
: Better: Something happens, but not the correct execution.

: For testing I constructed an easy batch named test.bat with only one command:
: echo %0 %1 %2 %3 %4 %5 >result.log

: Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
: Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
:>>> import os
:>>> os.chdir("c:\\mypython")
:>>> os.execv("test.bat",["test.bat","1","2","3","4"])

: When I start with:
:>>> os.execv("test.bat",["1","2","3","4"])
: only the interpreter ends. No Log-File has been written...

: I really spent a long time since I found out. Perhaps the documentation
: have to be changed?
:>>>
: execv (path, args) 
: Execute the executable path with argument list args, replacing the current
process (i.e.,
: the Python interpreter). The argument list may be a tuple or list of strings.
: Availability: Unix, Windows. 
:>>>

os.execv is a construct originating on UNIX, where the program name
does not always reflect the program file's name (with reason).  It is
common to create an executable that can react differently when called
with a different name.  As a simple example, gzip, gunzip and gzcat are
all the same physical file on a UNIX box, but called with those names,
it will compress, uncompress and uncompress-and-stream-output,
respectively.

Including the program name in the execv call allows this from a calling
process.

But in most cases, you can just use the program file:
  def myexec(program, *args):
    if len(args) == 1 and type(args[0]) == type(args):
      args = args[0]
    os.execv(program, (program,) + args)

This places the program filename in the arg0 position (see section 6.1.5
"Process Management" in the Python Library Reference.

  -Arcege



From wware-nospam at world.std.com  Fri Jul  9 22:42:27 1999
From: wware-nospam at world.std.com (Will Ware)
Date: Sat, 10 Jul 1999 02:42:27 GMT
Subject: Python Code Repository?
References: <378645AA.5430A026@msfc.nasa.gov> <19990709163127.A1964@toast.internal>
Message-ID: <FEMvIr.3DF@world.std.com>

jam (jam at newimage.com) wrote:
: I don't think anything as nice as CPAN exists for python ...
: I know enough about python and cgi scripting and sql databases to come
: up with something pretty quick, and all we'd need is submissions. there
: might already be some examples of code snippets out on the net in various
: places..
: does anyone have any further suggestions or ideas on how we can get this
: rolling?

There is alot of merit to the idea, and probably if you implement it,
getting contributions won't be a problem. It might be worthwhile to
set up a voting system that allows people to rate contributions for
interesting-ness. You might look at the slashdot.org moderation scheme,
that guy has thought long and hard about the problem of rating content
in a way that is both fair and informative.

There might be a few different ways people could search for something.
The first could be the Yahoo-style hierarchy, where you'd see something
like this:
AI -> neural nets -> back propogation -> you get the picture...
The second would be searching on keywords, which would be found either
in the source code itself or an associated text description or readme
file. I think freshmeat.net implements something like this.

What would be an interesting opportunity for innovation would be to
allow people to add new keywords after the fact. Suppose I'm looking at
some piece of code that integrates numerical solutions to differential
equations, and I realize that it has some bearing on some non-obviously
related thing like a problem in cellular automata. As a user of the
system, it would be good if I could add "cellular automata" as a search
keyword to the numerical integration code. Just a thought, but you might
end up finding some interesting connections that wouldn't otherwise
emerge.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com



From Gareth.McCaughan at pobox.com  Thu Jul  1 12:35:39 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 01 Jul 1999 17:35:39 +0100
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com>
	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com>
	<375FDCA7.62FF@mailserver.hursley.ibm.com>
	<7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com>
	<37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu>
	<37776FFA.2D85@mailserver.hursley.ibm.com>
	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it>
	<876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.ibm.com>
	<lwoghyexy5.fsf@copernico.parades.rm.cnr.it>
	<377B454C.6F59@mailserver.hursley.ibm.com>
	<377B593B.F7F7504B@iname.com>
	<377B8CFE.6113@mailserver.hursley.ibm.com>
Message-ID: <86pv2cwco4.fsf@g.pet.cam.ac.uk>

Paul Duffin wrote:

>>> Mistake, I think I meant dynamic scoping.
>> 
>>> (defun inner (b)
>>> (+ a b))
>>> 
>>> (defun outer (a)
>>> (inner 3))
>>> 
>>> (outer 2)
>> 
>>> 
>>> A simple macro version of setq would convert from
>>> (setq symbol '(list))
>>> to
>>> (set 'symbol '(list))
>> 
>> All 3 the same instance of the first misconception.
>> That's OLD, BROKEN Lisp!!
> 
> Please enlighten me as to what my misconception is. I am obviously not
> a Lisp expert and I realise that there are plenty of different flavours
> around but the above code works inside Emacs which says that it is
> based on Common Lisp.

Emacs is not based on Common Lisp. Common Lisp is not dynamically
scoped (though you can request specific variables to be scoped
dynamically).

The other main modern Lisp dialect, namely Scheme, is also not
dynamically scoped.

Old versions of Lisp were dynamically scoped because the first
Lisps were interpreted only and it's easier to do dynamic scoping
than lexical scoping in an interpreter. But with dynamic scoping
it's hard to get really good compiled code, and modularity is
harder too (because any function you call can mess with your
variables). So recent Lisps have switched to lexical scoping.

Emacs Lisp, however, is not a recent Lisp.

-- 
Gareth McCaughan            Dept. of Pure Mathematics & Math. Statistics,
Gareth.McCaughan at pobox.com  Cambridge University, England.



From Emile.van.Sebille at p98.f112.n480.z2.fidonet.org  Thu Jul  1 18:29:25 1999
From: Emile.van.Sebille at p98.f112.n480.z2.fidonet.org (Emile van Sebille)
Date: Thu, 01 Jul 1999 23:29:25 +0100
Subject: Alternatives to pickle/class nesting
Message-ID: <0000166a@bossar.com.pl>

From: "Emile van Sebille" <emile at fenx.com>

Good morning,

Unlike Gerrit's problem with seemingly large amounts of space being
consumed for storage, I cpickled my object and was impressed that only
81 bytes were used.  Of course, the entire object didn't get pickled.
In browsing dejanews, I found a post by Guido saying that nested class
structures, being rare, are not supported by pickle.  Being new to this,
I didn't anticipate the pending pickle problem, and now need to either
restructure the classes into one pickle-friendly object with the same
functionality, or look at alternatives for swizzling these objects?

Essentially, my class instances consist of a few variables, and two
dictionaries of string keys and UserList values modified for write-only
stack-like behavior.  If I do restructure the classes, by retaining the
same/similar structure without external class instances in the
structure, is it likely pickle will deal with it properly?

Can __getinitargs__, __getstate__ and __setstate__ be used to correct
this?

I have a debug display routine that dumps the contents of these
instances.  Any way to use this routine to create a local __dump__ /
__load__ pair.  Dumping I think I get, but are there traps in the
loading?

Any suggestions?

--

Emile van Sebille
emile at fenx.com
-------------------






From avv at quasar.ipa.nw.ru  Thu Jul  8 18:36:52 1999
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Thu, 08 Jul 1999 15:36:52 -0700
Subject: Python and Paradox
References: <7m2mrj$hes$1@nnrp1.deja.com> <7m2o2b$aq0@dfw-ixnews12.ix.netcom.com>
Message-ID: <37852804.54D5D2AD@quasar.ipa.nw.ru>

Aahz Maruch wrote:

> In article <7m2mrj$hes$1 at nnrp1.deja.com>,
> Kaine  <kaine_the_bloody at my-deja.com> wrote:
> >
> >I am looking for a way to access Paradox tables from Python without
> >using ODBC. Does anyone have any ideas?
>
> Nope, from an ex-Paradox tech support engineer.

Why nope? What is the platform? If it's win16, the direct wrapping is possible.

For DOS, there are several independent ports of Python to djgpp, and there is a
recipe, how to call functions from a real-mode DOS library. But this is definitely
for those, who likes to make sure they are true hackers.

Alexander





From Al.Christians at p98.f112.n480.z2.fidonet.org  Thu Jul  1 01:48:48 1999
From: Al.Christians at p98.f112.n480.z2.fidonet.org (Al Christians)
Date: Thu, 01 Jul 1999 06:48:48 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001632@bossar.com.pl>

From: Al Christians <achrist at easystreet.com>

I looked at Fox and put it aside because I need a grid control.
wxPython seems to be the winner for Windows in that category, even
though the grid is pretty minimal.

Any other candidates with a grid?

Al




Lyle Johnson wrote:
> 
> I've already worked up the binary distribution of FXPy for Windows; I just
> forgot to mention it in the previous post. The binary installer can be found
> at:
> 
>     ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42-win32.exe
> 
> I haven't worked up a binary distribution of the (C++) FOX distribution
> itself, primarily because it changes most every day :) But that is in the
> plans as well.
> 
> Thanks,
> 
> Lyle
> 
> Mikael Lyngvig wrote:
> 
> > On Wed, 30 Jun 1999 09:07:51 -0500, Lyle Johnson <jlj at cfdrc.com>
> > wrote:
> >
> > >Just to add fuel to the fire, we now have a Python interface to FOX, a
> > >cross-platform GUI toolkit:
> > >
> > >    http://www.cfdrc.com/pub/FOX/fox.html
> > >
> > >The latest version of the Python wrappers can be picked up at:
> > >
> > >    ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42.tar.gz
> > >
> > >I don't know how the speed of FXPy would compare with some of the others
> > >mentioned (Tk, wxWindows, Qt, GTK, ...) but it would be worthwhile for
> > >some independent opinions.
> >
> > I'm downloading it now.  The http URL is wrong, it should be
> > http://www.cfdrc.com/FOX/fox.html.
> >
> > The presentation seems promising, but I'd like to suggest a binary
> > distribution for PC/Intel computers.  I know this is not common for
> > Unix distributions, but it's the norm for PCs.
> >
> > -- Mikael



From gvwilson at nevex.com  Wed Jul  7 13:39:42 1999
From: gvwilson at nevex.com (Greg Wilson)
Date: Wed, 07 Jul 1999 13:39:42 -0400
Subject: make or similar in Python
Message-ID: <378390DD.58CA6BD5@nevex.com>

Hi, folks.  Has anyone ever written a dependency management/update tool
(like 'make') in Python?
I'm not looking for something that reads makefiles, and happens to be
written in Python; I'm looking
for a Python module that users import, then provide data to, in order to
manage dependencies.  (The
answer I got from the Perl crowd was 'cool idea, but no'; I'm hoping
I'll have more luck here.)
Thanks,
Greg (who'd really, really like to override some of the things that
GNU make does)





From gerrit.holl at pobox.com  Wed Jul 21 10:59:04 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Wed, 21 Jul 1999 14:59:04 GMT
Subject: uu.py - get result of uuencode() in a string
In-Reply-To: <3795CDC1.40F5E36@freemail.nl>; from Richard van de Stadt on Wed, Jul 21, 1999 at 03:40:17PM +0200
References: <3795CDC1.40F5E36@freemail.nl>
Message-ID: <19990721165903.A5108@optiplex.palga.uucp>

On Wed, Jul 21, 1999 at 03:40:17PM +0200, Richard van de Stadt wrote:
> Is it possible to get the output of the uuencode() function
> into a string, instead of a file or stdout?
> 

use the stringIO module.

regards,
Gerrit.



From ivanlan at callware.com  Wed Jul 21 15:15:13 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 21 Jul 1999 19:15:13 GMT
Subject: pow()
References: <37961668.FDF78B51@callware.com> <37961A61.64480218@aw.sgi.com>
Message-ID: <37961C41.482BDE17@callware.com>

Hi All--

Gary Herron wrote (and Christian Tismer did too):

Thanks to both of you.

> 
> This is from the library manual named "2.3 Built-in Functions":
> 
> pow (x, y[, z])
> 
> Return x to the power y; if z is present, return x to the power y,
> modulo z (computed more efficiently than pow(x, y) % z). The arguments
> must have numeric types. With mixed operand types, the rules for binary
> arithmetic operators apply.  The effective operand type is also the type
> of the result; if the result is not expressible in this type, the
> function raises an exception; e.g., pow(2, -1) or pow(2, 35000) is not
> allowed.
> 

So why didn't this turn up when I searched www.python.org for references
to "pow" ???

The explanation is perfectly clear, but I didn't see the reference to
the Built-in Functions page; when I searched, nothing came up (that was
pertinent, anyway).

<there-might-be-doc-bugs-on-summa-you-mugs>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From Darrell at p98.f112.n480.z2.fidonet.org  Fri Jul  2 05:58:39 1999
From: Darrell at p98.f112.n480.z2.fidonet.org (Darrell)
Date: Fri, 02 Jul 1999 10:58:39 +0100
Subject: C shared object files
Message-ID: <00001685@bossar.com.pl>

From: "Darrell" <news at dorb.com>

Check the docs under "Extending and Embedding" and search those key words at
dejanews.com.
Also check out some examples of how other people have done it, which will
show up in your search.
--
--Darrell


Martin Casado <casado at ogl.llnl.gov> wrote in message
news:199907021717.KAA03837 at ogl.llnl.gov...
>
>   Python Guru's,
>
>    I am trying to get python to dynamically load a C object module, and
> first off I can't figure out how to compile myCollProggy.c into a
> shareable object file, and secondly I was curios if there is a way to
check
> at runtime if a module is on the system or not.
>
>   I have made arrangements with the proper authorities and there is a
veritable
> wealth of Karma points waiting for anyone who can help me :-)  Hehe thanks
> very much in advance.
>
> -Martin
>
>




From larsga at ifi.uio.no  Wed Jul 28 17:03:06 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 28 Jul 1999 23:03:06 +0200
Subject: Bit-twiddling
Message-ID: <wkso68xzat.fsf@ifi.uio.no>

I'm translating a numerical algorithm from C to Python, but have run
into a problem in that the algorithm seems to rely on the fact that
shifting, addition and multiplication all 'wrap' when the result
exceeds 32 bits.

Since the Python equivalents do not, does there exist a common
workaround for this?

For example, I'd like

    mt[0]= seed & 0xffffffff
    for mti in range(1,n):
        mt[mti]=(69069 * mt[mti-1]) & 0xffffffff

to never put anything that doesn't fit in an unsigned long in the mt
list.

Any ideas?

--Lars M.



From victor at idaccr.org  Fri Jul  2 17:46:57 1999
From: victor at idaccr.org (Victor S. Miller)
Date: 02 Jul 1999 17:46:57 -0400
Subject: python -c not working
Message-ID: <ulr9mqlo6m.fsf@runner.princeton.idaccr.org>

I built Python on various of our systems (1.5b2) (Solaris 2.7 and AIX
4.x), and everything seemed to be ok, until today when I was trying to
install Numerical Python.  I tracked down the error to the fact that

python -c

doesn't work

for example

python -c 'print 1'

gets no output, and

python -c 'import path; print sys.path'

get a syntax error at import.  Does anybody have a clue as to what can
be wrong?

I built it in without such problems on another set of Solaris 2.6
systems.
-- 
Victor S. Miller     | " ... Meanwhile, those of us who can compute can hardly
victor at idaccr.org    | be expected to keep writing papers saying 'I can do the
CCR, Princeton, NJ   | following useless calculation in 2 seconds', and indeed
    08540 USA        | what editor would publish them?"  -- Oliver Atkin



From emile at fenx.com  Mon Jul 26 10:58:08 1999
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 26 Jul 1999 07:58:08 -0700
Subject: Access dictionary in order?
References: <roy-2507991300370001@mc-as01-p61.med.nyu.edu> <14235.18248.924848.52652@dolphin.mojam.com> <slrn7po8t3.6k.diego@blanka2.blankanet.it> <Wy_m3.4215$VL2.444209@news.direcpc.com>
Message-ID: <WK_m3.4216$VL2.444194@news.direcpc.com>

( oops   ;-(  )

Of-course,-had-I-allowed-both-tests-to-do-the-same-number-of-iterations,
-the-results-may-have-been-different-ly yr's

--

Emile van Sebille
emile at fenx.com
-------------------





From greg.ewing at compaq.com  Tue Jul 20 18:29:38 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 21 Jul 1999 10:29:38 +1200
Subject: Stackless & String-processing
References: <001101bed26d$a5bc4580$139e2299@tim> <7n27vi$mm4$2@cronkite.cc.uga.edu> <7n28ib$6v8$5@news.accu.uu.nl>
Message-ID: <3794F852.B984B059@compaq.com>

If you're trying to parse something using a grammar
that needs heavy transformations to make it tractable,
then obviously using monadic parsers with arbitrary
code embedded in them is going to get in the way
of those transformations.

However, not all parsing problems are like that.
Programming languages are usually designed to be
straightforward to parse efficiently. Monadic 
techniques can allow such parsers to be written 
very elegantly.

Greg



From nick.bower at ssec.wisc.edu  Tue Jul 20 10:31:01 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Tue, 20 Jul 1999 09:31:01 -0500
Subject: Scientific Plotting?
References: <379392E7.EB0F29E7@ssec.wisc.edu>
Message-ID: <37948825.50CCABE0@ssec.wisc.edu>

Thanks for the suggestions everyone!!

I took a look at the GIST documentation.  Bunch of plot-types, but the
graphs in the documentation aren't exactly what I'd think about putting
in a report.  They'd look better with axes, which is perhaps a feature
anyway.

Printing back-ends are certainly very important.  I didn't see much
attention paid to ps fonts and general layout in most of the packages. 
But there were a couple of new suggestions that I hadn't checked out
yet.

> Fortunately, Python interfaces very cleanly with Matlab.  With Andrew
> Sterian's nice PyMat module

Thanks!  Looks very cool.

nick



From gmcm at hypernet.com  Wed Jul 28 17:38:24 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 28 Jul 1999 16:38:24 -0500
Subject: 1.5.2 for: else:
In-Reply-To: <slrn7pun42.2ro.wtanksle@dolphin.openprojects.net>
Message-ID: <1278928389-16733806@hypernet.com>

William Tanksley wrote:
> On Tue, 27 Jul 1999 21:43:00 -0500, Gordon McMillan wrote:

[snip Billy's silly error and his entirely insufficient embarassment 
<wink>]

> However, you're wrong about the best way to detect a 2 in the list
> using this style.  To steal the same example:
> 
> try:
>  for x in [1,2,3]:
>   print x
>   if x == 2: raise neverMind
>  print "modern else clause here"
> except neverMind: print "found two"
> else:
>  print "another place to put the modern else clause"
> 
> Wow, TMTOWTDI.  Oops, did I say the wrong thing?
> 
> Of course, by demonstrating both ways to do it, I'm making the code
> more complex; but you can see that this solution solves the problem
> quite elegantly.

WIth an extra class, and 3 or 4 more lines of code, not to mention 
treading on the toes of all who think exceptions should only be used 
for exceptional conditions (or maybe it _is_ exceptional when 
something you wrote works <1e3 wink>).

> The place to use MY definition of 'else', OTOH, is when you're
> building a structure during the iteration, but the result of an
> empty loop is different than the initial value the variable has to
> have when going through the loop.
> 
> Here you go.  This example might convert a Python list into a Lisp
> tree-list -- but to imitate Lisp (okay, it's a toy problem, I'm
> cheating) an empty list produces NIL ~= None.
> 
> var = Node(None,None)
> for x in pyList:
>   var.car = x
>   var.cdr = Node()
> else:
>   var = None

Not a Lisper, but isn't what you're doing better done thusly:

def test1(lst):
 var = None
 for x in lst:
  var = (x, var)
 return var

print `test1([1,2,3])`

print `test1([])`
 
Produces:
(3, (2, (1, None)))
None
 
> >insufferably-sanctimoniously-y'rs
> 
> >- Gordon
> 
> unstoppably-confidantly y'rs
> 
> -- 
> -William "Billy" Tanksley

so-self-righteous-without-the-winks-I-might-almost-be-mistaken
 -for-that-other-GM-ly y'rs

- Gordon



From kperacles at geocities.com  Sat Jul 10 21:38:54 1999
From: kperacles at geocities.com (Karl Putland)
Date: Sat, 10 Jul 1999 20:38:54 -0500
Subject: Python and COM?
Message-ID: <3787f71e.0@news.dbtech.net>

For a COM server implemented in python, can a type library be generated?

I would like to be able to distribute some of the work I've done, but most of the people that would use it would be using VB/VBA so I would like expose the Python objects as COM objects and generate a type library so that they can take advantage of the QuickInfo and parameter hints.

TIA

Karl Putland
kperacles at geocities.com


From bmcd at es.co.nz  Tue Jul 27 17:53:56 1999
From: bmcd at es.co.nz (Ben Caradoc-Davies)
Date: 27 Jul 1999 21:53:56 GMT
Subject: Python and Red Hat Linux 6.0
References: <379CCAC0.AD5D1C2F@postoffice.pacbell.net>
Message-ID: <slrn7psasp.d2.bmcd@ripley.sulaco.net>

On Mon, 26 Jul 1999 13:53:20 -0700, Rob <vaton at postoffice.pacbell.net> wrote:
>When programming in Perl or Python, I am using a header like the
>following on an executable file to cause the shell (bash or csh) to run
>the scripts as Perl/Python files:
>#!/usr/bin/perl   for Perl   and #!/usr/bin/python   for Python. I've
>checked and the compilers are located in the appropriate directories. I
>keep getting this when I run scripts:
>bash: <script name>: command not found. Does Linux support the '#!'
>notation in the file header, or do I always have to run the scripts by
>explicitly typing in 'perl' or 'python' before the script name?

Every decent Unix supports #! shell scripts. Linux is one of them. As far as I
understand it, #! is just a different magic number that allows the kernel's
binary format handlers to detect the script in the same way that 0x7f then
"ELF" marks an elf executable.

If you're interested, check out
  /usr/src/linux/fs/binfmt_script.c
When an image is exec'ed, the kernel applies all the binary format handlers it
knows in succession, until one works. 

As for your problem:

1)	Check your PATH. If . is not in your path, you have to run the script
	as ./scriptname
	This is my best guess as to your problem, given your symptoms.

2)	Check your script. It's easy to write !#/usr/bin/python by accident,
	with your mind moving faster than your fingers. But for this problem
	you get different errors.

3)	chmod +x scriptname
	However, this will give you "Permission denied." if you forget. This is
	a common problem, especially if you use an editor which resets the
	permissions whenever you save.

-- 
Ben Caradoc-Davies <bmcd at es.co.nz>



From Joachim.Schmitz at p98.f112.n480.z2.fidonet.org  Sun Jul  4 14:31:39 1999
From: Joachim.Schmitz at p98.f112.n480.z2.fidonet.org (Joachim Schmitz)
Date: Sun, 04 Jul 1999 19:31:39 +0100
Subject: mxODBC and MySQL access to remote DB
Message-ID: <000017cb@bossar.com.pl>

From: Joachim Schmitz <perry at ac-copy.net>

Hello,

I?m trying to connect from a linux-box with mxODBC to a MySQL db on a
remote SUN. I can access the db with mysql-client from linux. I can also
connect with the odbctool from the iodbc-mysql package.

When I try it with mxODBC.iODBC, which uses the same "iodbc.ini" I always
get this:

line: 38
-- in file: Form.py; function: __init__; line: 40
arg #0 = DSN=immobilien;UID=perry;PWD=pwpwpwpw
Traceback (innermost last):
  File "immoform.py", line 153, in ?
    ,DBtype="MySQL")
  File "Form.py", line 45, in __init__
    ,"pw="+self.pw)^M
mxODBC.OperationalError: ('S1000', 1045, "[TCX][MyODBC]Access denied for
user: 'user=perry at localhost' (Using password: YES)", 4237) 

The iodbc.ini entry looks like this:

[immobilien]
Trace    = Off
TraceFile= stderr
Driver   = /usr/lib/mysql/libmyodbc_mysql.so
DSN      = immobilien
SERVER   = test.ac-copy.net
UID      =
PWD      =
PORT     = 3306
#OPTION  = 1
#DB      = test
#SOCKET  = /tmp/mysql.sock
 
with:

odbctool -u perry -p pwpwpwpw -d immobilien "show tables"
<table border=1>
<th>Tables in immobilien</th>
<tr>
<td>IMMOAUFTRAG</td>
</tr>
<tr>
<td>IMMOKUNDE</td>
</tr>
</table>                           

Any help is welcome.



Gru?

Joachim Schmitz

WWW-Consultant

email: js at ac-copy.net
tel: +49-241-89491-0
fax: +49-241-89491-29




From ze5yr at my-deja.com  Sat Jul 24 22:50:55 1999
From: ze5yr at my-deja.com (ze5yr)
Date: Sun, 25 Jul 1999 02:50:55 GMT
Subject: Q: anyone running Python on OS X server?
References: <004401bed44f$1a5c6400$f29b12c2@secret.pythonware.com> <37979524.C9F4F6BC@freenet.co.uk> <008c01bed508$e4e85a80$f29b12c2@secret.pythonware.com>
Message-ID: <7ndu2e$41o$1@nnrp1.deja.com>

In article <008c01bed508$e4e85a80$f29b12c2 at secret.pythonware.com>,
  "Fredrik Lundh" <fredrik at pythonware.com> wrote:
> sreekant kodela <sreekantk at freenet.co.uk> wrote:
> > Do you mean Linux / Xfree ?
> > If that is what you asked then answer is Yes.
>
> no, I meant "OS X," as in "Mac OS X Server":
> http://www.apple.com/macosx/server/

Ah, so I'm not alone in having to babysit one of these beasts :-)


> turns out that the answer is "yes, works right
> out of the box!".  (which is pretty cool, IMO).

Yeah, I was worried about installing it too, but after
>SIX< >HOURS< of debugging a simple Perl script, I broke down and
downloaded it..the install turned out to be much easier than I
expected!

Now if I can only get ssh to do the same :-(


--
Cliff Crawford                           -><-
http://www.people.cornell.edu/pages/cjc26
                            empty-handed I go,
)O(       and behold the spade is in my hands


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Randall.Hopper at p98.f112.n480.z2.fidonet.org  Fri Jul  2 01:49:28 1999
From: Randall.Hopper at p98.f112.n480.z2.fidonet.org (Randall Hopper)
Date: Fri, 02 Jul 1999 06:49:28 +0100
Subject: newbie question about tkFileDialog.py
Message-ID: <0000166f@bossar.com.pl>

From: Randall Hopper <aa8vb at yahoo.com>

Dinesh Shanbhag (c):
 |Did not find anything like tk_getOpenFile in the Tcl/Tk man pages for
 |selecting Directories...
 |
 |Am i missing something obvious ?

I don't think so.  I looked for the same and didn't find it either.

Randall




From wtanksle at dolphin.openprojects.net  Thu Jul 29 15:48:49 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Thu, 29 Jul 1999 19:48:49 GMT
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <slrn7psisn.da.wtanksle@dolphin.openprojects.net> <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com> <slrn7pum9t.2ro.wtanksle@dolphin.openprojects.net> <001501bed935$561b8210$f29b12c2@secret.pythonware.com>
Message-ID: <slrn7q1c10.5jp.wtanksle@dolphin.openprojects.net>

On Wed, 28 Jul 1999 20:11:10 GMT, Fredrik Lundh wrote:
>William Tanksley <wtanksle at dolphin.openprojects.net> wrote:
>> >    for an "if" statement, "else" is a path taken if the
>> >    expression evaluates to false.

>> Close enough -- I would say it's the path taken if the 'if' block isn't
>> executed.

>you can say what you want, but it isn't defined
>that way; the language reference says:

>    "If all expressions are false, the suite of
>    the else clause, if present, is executed." 
>    (where "all expressions" are the expressions
>    for all if/elif clauses in the statement)

>if you continue reading, you'll find:

>    "if the expression is false (which may be
>    the first time it is tested) the suite of the
>    else clause, if present, is executed"

>and

>    "when the items are exhausted (which is
>    immediately when the sequence is empty),
>    the suite in the else clause, if present, is
>    executed"

>for ALL these statements, the else clause is exe-
>cuted if and only if ALL controlling expressions
>evaluate to FALSE.

But that's not even superficially true for the current behavior of a loop.
A loop's else clause is executed even when there were TRUE results of the
controlling expressions.

In fact, the loop's else clause is executed when a single FALSE is
returned from an often hidden evaluation.

>huh?  the "else" clause isn't involved in the loop's
>termination at all; it's executed when the con-
>trolling expression becomes false.  if you raise
>an exception, or use the break statement, or
>return to the caller, you don't end up in the
>"else" clause by a very simple reason:  the con-
>trolling expression isn't false.

This actually makes a lot of sense for a while statement, since there's an
explicit single-valued controlling expression.  If the while statement
were the only loop statement with an else, I wouldn't bother arguing about
it; I would just think that I would rather not use a break for that
purpose (but scripting languages are wonderful that way, you get to write
code that makes sense to YOU, not to me or Guido).

>I repeat:

>for ALL these statements, the else clause is exe-
>cuted if and only if ALL controlling expressions
>evaluate to FALSE.

Except 'for', where it executes if and only if _one_ of the "controlling
expressions" (all of which are implicit) evaluates to FALSE.

If it truly was consistent with your own definition, then it would execute
if and only if none of the controlling expressions ever evaluated to TRUE
-- in other words, if the for loop never iterated.

I think I got you by your very own words.

>you can argue as much as you want, but that's the
>way it is.  you gotta live with it (unless you have your
>own time machine, of course ;-)

I agree with this, of course :).

I'm nore agitating for Python 2.0.  But then I'd be happy to see else go
away or stay the same, too; it's too minor of a feature.  I guess, then,
that I'm merely defending my point of view.

></F>

-- 
-William "Billy" Tanksley



From mal at lemburg.com  Fri Jul 16 05:58:57 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 16 Jul 1999 09:58:57 GMT
Subject: Meta-language for mxTextTools
References: <010001becf64$a615fa50$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <378F0261.3A84A45C@lemburg.com>

Tony J. Ibbs (Tibs) wrote:
> 
> M.-A. Lemburg wrote:
> > One itch that I have with the syntax: since you are
> > already using Pythonish syntax, why not simply write
> >
> > def keyword:
> >     ...
> >
> > instead of
> >
> > keyword is:
> >     ...
> 
> "is" was what I originally thought of, because I was trying to think of
> "reading" the tag table definition back to myself, and that sounded better.
> When I realised that I was wanting to be "Python-esque" I realised that
> using "def" would perhaps be more consistent, but by then I'd got used to
> "is", and also I couldn't help feeling that "def" was less clear *to me* - I
> wanted it to be obvious to me that I wasn't defining a function (perhaps I'm
> still (vaingloriously) hoping to allow Python code in tag-table files...).
> However, it's probably not a religious issue, and may perhaps change before
> things get released. How strong an itch is it?

Well, it's just that "def" state exactly what you are doing
when you define a table. "is" is only used in comaprisons in
Python, so the analogy doesn't go too far.

You might even consider using "table" to start the section.
This wouldn't conflict with functions/methods in Python and
still express what you are aiming at.
 
> > An alias "back" for -1 would help reading the tables:
> >
> > Skip back
> > vs.
> > Skip -1
> 
> I would be happy with that - would one bother with "forward" as well, even
> though I doubt that is as often used (I tend to assume not)? Or should one
> reuse "previous":
> 
>     Skip previous
> 
> which is explicit in meaning -1 (whereas to me "back" just indicates the
> direction), but seems to be missing the word "to".

There is a small difference in semantics between "previous" and
"back": you are using "previous" to move the instruction pointer
in the table, while "back" moves the engine's current position
in the text. So having to different synonyms may be a good idea
(the wording is secondary, IMHO).
 
> Hmm. I'll think on it - both have disadvantages for me, but I definitely
> like "hiding" that -1 (although since this is moving the read-head it is
> less important to me than losing what I think of as "line numbers" in the
> tag table).

Right. That's my major gripe with the tables: you constantly
have to adjust offsets whenever you insert things. A meta language
ought to take care of this by introducing jump targets.

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   168 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From bucodi at worldnet.fr  Wed Jul 28 03:04:12 1999
From: bucodi at worldnet.fr (pserrand)
Date: Wed, 28 Jul 1999 09:04:12 +0200
Subject: Designing questions
Message-ID: <379EAB6B.2AD7E3C0@worldnet.fr>

I'm writing  a application wich contains abouth 50 inputscreen linked to
db-tables.
Each inputscreen has 2 classes
- 1 GUI class wich  construct the screen.
- 1 db class wich  do the db acces.

All the gui-classes are done at the moment, so i have the prototype of
the application.
And here's my problem :
The source code becomes realy long, wich gives a long time to load it. I
was thinking of spliting it in modules (1 module for each screen wich
contains the gui-class and the db-class) and then importing it when i
needed it, this is when the user makes the choice in the main-menu.
The question i have is how to 'unload' a module when i don't need it
anymore ? If i don't unload a module i would  have as many modules in
memory as the user has made choises in the main menu.
I think this could become a memory problem especailly since i'm using
gadfly as db, wich allso holds the tables in memory.

I really would appreciate some hints and comments here from experienced
guys who allready wrote BIG real world applications in PYTHON.

Pierrette
A kiss from a french lady for every answer :)

If possible send a copy of your answer to my e-mail bucodi at worldnet.fr





From thomas at xs4all.nl  Mon Jul 19 10:12:26 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Mon, 19 Jul 1999 16:12:26 +0200
Subject: [Q] Type checking...go further...
In-Reply-To: <7mv9n8$md$1@feed.teaser.fr>; from Olivier Deckmyn on Mon, Jul 19, 1999 at 03:35:35PM +0200
References: <7mv9n8$md$1@feed.teaser.fr>
Message-ID: <19990719161226.F6170@xs4all.nl>

On Mon, Jul 19, 1999 at 03:35:35PM +0200, Olivier Deckmyn wrote:

> Thank you all for your answers !
> I finally use isinstance in the methods of my classes...

> But to go further, is there a way in python to have typed methods parameters ?
> aka :

> class MyClass:
>     def myMethod(string myString):
>         print myString

Nope, there isn't. Python is fully dynamically typed. That means you have to
expect everything anytime anywhere, but it also means you wont have to worry
about what you send to where when. If you want to be sure an argument is of
a specific type (say, a string) either make sure by hand all calls are
strings and add a typecheck in the call, or convert the argument to a
string.

Or, if you want a function to do something very different depending on what
was passed, use a dispatch table:

class Eggs:
	def __init__(self, value):
		self.value = value

class MyClass:
	def spam_string(self, egg):
		print egg

	def spam_eggs(self, egg):
		print egg.value

	def spam_other(self, egg):
		if isinstance(egg, Eggs):
			self.spam_eggs(egg)
		else:
			self.spam_string(egg)

	def spam_list(self, spamlist):
		for egg in spamlist:
			self.spam(egg)

	dispatch = { type(""): spam_string,
		     type([]): spam_list }

	def spam(self, egg):
		if self.dispatch.has_key(type(egg)):
			self.dispatch[type(egg)](self, egg)
		else:
			self.spam_other(egg)

(this is a very simple example. Most of this can be done with a simple
wrapper classes with __repr__ or __str__ functions. But it's just for
basics, anyway.)

>>> a=MyClass()
>>> b=Eggs("sausages")
>>> c="Ham"
>>> a.spam(b)     
sausages
>>> a.spam(c)
Ham
>>> a.spam([["spam"]*5, b, b, c*10, [c]*10])
spam
spam
spam
spam
spam
sausages
sausages
HamHamHamHamHamHamHamHamHamHam
Ham
Ham
Ham
Ham
Ham
Ham
Ham
Ham
Ham
Ham

damned-if-i-know-why-i-dont-work-with-python-dai'ly yours,

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From rozzin at geekspace.com  Thu Jul  1 21:31:17 1999
From: rozzin at geekspace.com (Joshua Rosen)
Date: Thu, 01 Jul 1999 21:31:17 -0400
Subject: scope-resolution
References: <000d01bebeda$1b7430a0$d29e2299@tim> <slrn7n8qiv.kl.behrends@allegro.cse.msu.edu>  <3778C2C7.3B101FD0@geekspace.com> <slrn7nk3gv.18i.behrends@allegro.cse.msu.edu>
Message-ID: <377C1665.274FD346@geekspace.com>

Eckh--apparently, I had a dead-brain day, and took you much too
seriously.
Of course, there's got to be someone out there who needed to here that
stuff....

I imagine that you were using typing as a metaphor for scoping, and,
now that I see it, I can appreciate it.

It would be nice to have some non-mandatory scope-resolution
capabilities beyond self-referencing in instance-methods and `global'
statements.

What would be a different way of getting the effect of:

	ClassX.foo(ClassY())
?

Have bound and unbound methods be different creatures, perhaps....

No, never mind, I suppose--thinking about it, so long as you can
assure `this variable is a class variable', then everything should be
fine. Actually, just adding the instance's namespace to the list of
namespaces searched, in methods, shouldn't be too bad to implement,
and probably wouldn't break any of the existing code, eh? [function
locals, instance locals, globals]? 

		-Rozzin.




From aahz at netcom.com  Thu Jul 22 11:08:59 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 22 Jul 1999 15:08:59 GMT
Subject: Cyclops 0.9.4
References: <001301bed410$e59eafe0$642d2399@tim>
Message-ID: <7n7c6b$o7i@dfw-ixnews4.ix.netcom.com>

In article <001301bed410$e59eafe0$642d2399 at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>
>If you have no idea which objects to look for, you're not going to get help
>short of divine revelation.  An object living after you expect it to die is
>an error in program logic, and as hard to track down as any other kind of
>programmer flub; heck, harder than most, because usually an error of
>omission (you neglected to lose all the references) -- there's no line of
>code "to blame".

Going off on a tangent, I want to whine a bit about a "bug" in my code
that I came across recently:  I'm downloading web pages, cross-checking
HREFs against URLs already in the database.  One URL on one particular
web page keeps causing my code to crash.  Netscape's View | Source
didn't show anything wrong.  Finally, I used urllib.urlopen() to get a
raw look at the page -- and it turned out that the URL had a null byte
at the end!  (Yes, '<a href="http://foo/bar/\000">')

A null byte in the middle of a web page.  Who'd'a thunk?  <grrrrrr!>
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From Fredrik.Lundh at p98.f112.n480.z2.fidonet.org  Fri Jul  2 04:17:30 1999
From: Fredrik.Lundh at p98.f112.n480.z2.fidonet.org (Fredrik Lundh)
Date: Fri, 02 Jul 1999 09:17:30 +0100
Subject: TIFF-images
Message-ID: <0000165a@bossar.com.pl>

From: "Fredrik Lundh" <fredrik at pythonware.com>

From: <olevi at my-deja.com>
> Is there any way to present TIFF-images through tkinter? After what I've
> read, only GIF, PPM and PGM are supported, but maybe there's a way
> around?

http://www.pythonware.com/products/pil/
(software distribution)

http://www.pythonware.com/library/pil/handbook/index.htm
http://www.pythonware.com/library/pil/handbook/imagetk.htm
(docs)

</F>



From emile at fenx.com  Sun Jul 25 13:21:06 1999
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 25 Jul 1999 10:21:06 -0700
Subject: Access dictionary in order?
References: <roy-2507991300370001@mc-as01-p61.med.nyu.edu>
Message-ID: <OKHm3.4158$VL2.439920@news.direcpc.com>

Roy,

You don't have to maintain a separate keylist.  Try:

>>> d = {1:11,2:22,3:33}
>>> keylist = d.keys()
>>> keylist.sort()

then do for key in keylist:

--

Emile van Sebille
emile at fenx.com
-------------------


Roy Smith <roy at popmail.med.nyu.edu> wrote in message
news:roy-2507991300370001 at mc-as01-p61.med.nyu.edu...
> I've got a dictionary of items which I will sometimes want random
access
> to, but sometimes I'll want to be able to step through the items in a
> fixed order as well.  Any good way to do that?
>
> The best I've come up with so far is to keep a parallel list of the
> dictionary keys in sorted order and use that to step through when I
need
> to:
>
>    for key in keylist:
>       do_stuff (dict[key])
>
> which will work, but seems kind of clumsy.  Anything smoother I can
do?





From gmcm at hypernet.com  Thu Jul 22 23:45:58 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 22 Jul 1999 22:45:58 -0500
Subject: Fork, Threading, and Select madness.  Please help!
In-Reply-To: <3797A2D4.DB293C40@mindspring.com>
Message-ID: <1279424742-23511862@hypernet.com>

Jesse D. Sightler wrote:

> Ok, I have been working on a website which is now available at
> http://www.biddin.com/ that performs a search of multiple auction
> sites all at once, and then returns the results.  The problem is
> that it is very slow, and it appears that the primary bottleneck is
> in waiting on the search engines at the various sites to begin
> reading data.
> 
> Well, this makes sense because the code to implement the search is
> currently pure sequential with no threading or non-blocking IO
> whatsoever.  
> 
> What I'm getting around to asking (slowly<g>) is, what is the best
> way to implement an algorithm that can retrieve the data from ALL 3
> sites simultaneously, so that the code is never blocked up waiting
> on one particular bad site?  

In theory, select (multiplexing) is the best solution, especially 
since you are just gathering data. But this might well mean trashing 
all of your existing code. I'd recommend using asyncore if you're 
doing this from scratch.

If you're using one of the higher level protocol (urllib, 
httplib) modules, a quick glance (I'm no expert on these guys) says 
you'll have to thread or use subprocesses. Multiplexing usually means 
turning your logic inside out (you're no longer driven by meaningful 
exchanges, but by drips and drabs of data appearing in network 
buffers). So the transformation from blocking sockets to multiplexed 
non-blocking sockets is major surgery.

- Gordon



From catlee at my-deja.com  Fri Jul  9 17:44:02 1999
From: catlee at my-deja.com (catlee at my-deja.com)
Date: Fri, 09 Jul 1999 21:44:02 GMT
Subject: Python re module
Message-ID: <7m9tn4$6qg$46@rks1.urz.tu-dresden.de>

Try the following:

import re
re.search("((.)\\1+)","a")

Now, I know this isn't proper syntax (you shouldn't reference a group
inside itself), but on my machine python hangs.

Is this a bug, or a feature? =)

Cheers,
Chris


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From peter.stoehr at weihenstephan.org  Wed Jul 28 13:18:33 1999
From: peter.stoehr at weihenstephan.org (Dr. Peter Stoehr)
Date: Wed, 28 Jul 1999 19:18:33 +0200
Subject: Memory leak in Tkinter.py ?
Message-ID: <379F3B68.4D533F3@weihenstephan.org>

Hi Python-Fans,

Greg McFarlane (the developer of Pmw) and I found a memory leak in
Tkinter.py of Python 1.5.2.

The tag_bind method of the Canvas class maintainece a dict called
self._tagcommands to store lists of function-ids returned from the _bind
call.
As far as I can see, this information not used by any other method.
Furthermore, the lists grow with every call of tag_bind and the memory
is never released.

After removing the
	if sequence and func and res:
block of the tag_bind method, the memory consumption of the enclosed
test program is constant and no longer growing.

Is it okay to remove the 
	if sequence and func and res:
block or did I miss something?

    Greetings
	Peter

#########################################################################################

import Tkinter

def callback(event):
    pass

def autoUpdate():
    buttonId = canvas.bind('<ButtonPress>', callback)
    canvas.unbind('<ButtonPress>', buttonId)
    root.after(10, autoUpdate)

root = Tkinter.Tk()

canvas = Tkinter.Canvas(root)
idn = canvas.create_rectangle(100,100,200,200, fill='green')
canvas.pack()
autoUpdate()

root.mainloop()

##########################################################################################
-- 
---------------------------------------------------------------------------
   Dr. Peter Stoehr --- Teisenbergweg 6 --- 85435 Erding --- 08122/47232
---------------------------------------------------------------------------
		I'm the terror that flaps in the night



From gmcm at hypernet.com  Sat Jul 24 08:56:29 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 24 Jul 1999 07:56:29 -0500
Subject: Tk and a little headache (installation troubles)
In-Reply-To: <3799183B.28530914@no_spam.uwplatt.edu>
Message-ID: <1279305310-30695550@hypernet.com>

Brett Gengler wrote:

> For some strange reason I'm having trouble getting the tk libray to
> work with python.  I've installed tk & tcl 8.2 with a standard
> install. 

>   Any ideas what I'm doing wrong?

Python still uses 8.0.

There were patches posted in early June (from Dieter Maurer?) for 
using 8.1.

- Gordon



From vdkoijk at linux01.prc.tno.nl  Tue Jul  6 08:11:59 1999
From: vdkoijk at linux01.prc.tno.nl (John van der Koijk)
Date: 06 Jul 1999 14:11:59 +0200
Subject: popen2 on Windows
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90EC1@RED-MSG-50>
Message-ID: <87hfnit1tc.fsf@linux01.prc.tno.nl>

"BT" == Bill Tutt <billtut at microsoft.com> writes:

    >> After coping with my initial frustration, I've decided to make
    >> an unthreaded stub program talking to the DDE client, using
    >> popen2 to do the IPC with the GUI and controller
    >> program. Sadly, this approach also does not seem to work,
    >> because using popen2 gives me:
    >> 
    >> [....] os.fork() AttributeError: fork
    >> 
    >> This is amazing, since the manual for 1.5.2 says:
    >> 
    >> 8.13 popen2 -- Subprocesses with accessible standard I/O
    >> streams
    >> 
    >> Availability: Unix, Windows.
    >> 

    BT> Well, its wrong. :) Windows doesn't have fork().

    BT> The win32pipe module exposes popen2() that does what you want
    BT> it to do, however you need to avoid a bug in Win9x. The
    BT> knowledge base article at:
    BT> http://support.microsoft.com/support/kb/articles/q150/9/56.asp
    BT> covers what you need to do to avoid this dilemma.

Wow. I'm completely flabbergasted...

I have realised that I could also use a socket interface. Works great,
and speed is not one of my concerns, right now.

Thanks,
-- 
John van der Koijk.
--
TNO Institute of Industrial Technology

PO Box 5073, 2600 GB, Delft, The Netherlands




From fredrik at pythonware.com  Thu Jul 29 11:12:23 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 29 Jul 1999 15:12:23 GMT
Subject: Is "shelve" thread-safe on Windows?
References: <7npimc$l9b@romeo.logica.co.uk> <14240.25927.863885.288661@dolphin.mojam.com>
Message-ID: <001701bed9d4$c3f72240$f29b12c2@secret.pythonware.com>

Skip Montanaro <skip at mojam.com> wrote:
>     Andrew> Can anyone suggest a simple way to gain the functionality of
>     Andrew> "shelve" on a Windows platform, but which is safe for handling
>     Andrew> concurrent accesses (like from CGI scripts)?
> 
> You might take a look at file locking.  On Unix there's the fcntl module.
> Not sure if there is similar functionality on Windows.

    http://www.python.org/doc/current/lib/msvcrt-files.html

    msvcrt.locking(file.fileno(), mode, bytes)

if anyone knows how to use the "mode" argument,
please let us know... (and send doc patches to
Fred!).

</F>




From mak at mikroplan.com.pl  Thu Jul 22 18:36:10 1999
From: mak at mikroplan.com.pl (Grzegorz Makarewicz)
Date: Fri, 23 Jul 1999 00:36:10 +0200
Subject: cin
In-Reply-To: <37976CF0.3DC02CBA@weihenstephan.org>
Message-ID: <000001bed492$987a7af0$129e75c3@mikroplan.com.pl>

Hi,

Try very old module named sscanf sources in
ftp.python.org/pub/python/contrib/misc/sscanf.c and sscanf.README
#--snip--
import sscanf
fmt='My name is %s and I am %d years old.'
data=('Michael', 30)
result=fmt %data
print result
print sscanf.sscanf(result,fmt)
#/--snip--

output:
My name is Michael and I am 30 years old.
['Michael', 30]

Grzegorz Makarewicz
mak at mikroplan.com.pl

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Dr. Peter Stoehr
> Sent: 22 July 1999 21:12
> To: fire at netlimit.com
> Subject: Re: cin
>
>
> Hi everyone,
>
> "Michael P. Reilly" wrote:
> >
> > Joacim L*wgren <fire at netlimit.com> wrote:
> > : Hi,
> >
> > : I'm new to Python and I'm wondering what the C++ "cin" (or C
> > : "scanf()")equivalent in Python is. Doesn't seem there's a
> quick function for
> > : this.
> >
> > Actually it is not a function, it is an operation (%).
> >
> > >>> name, age = 'Michael', 30
> > >>> s = 'My name is %s and I am %d years old.'
> > >>> print s % (name, age)
> > My name is Michael and I am 30 years old.
> >
> > Take a look at the Library Reference Manual, Section 2.1.5.1
> "More String
> > Operations"
> (http://www.python.org/doc/current/lib/typesseq-strings.html).
> >
> >   -Arcege
>
> this emulates the C-function sprintf and not scanf.
> What Joacim is looking for is something like raw_input :-)
>
> # -----------------------------------------
> aString = raw_input("Give me input :")
> print aString
> # -----------------------------------------
>
> Greetings from Erding/Bavaria/Germany/Europe/Earth/Universe
>     Peter
> --
> ------------------------------------------------------------------
> ---------
>    Dr. Peter Stoehr --- Teisenbergweg 6 --- 85435 Erding --- 08122/47232
> ------------------------------------------------------------------
> ---------
> 		I'm the terror that flaps in the night
>
> --
> http://www.python.org/mailman/listinfo/python-list
>




From tgabriel at bellsouth.net  Fri Jul  9 11:44:49 1999
From: tgabriel at bellsouth.net (Xenophanes)
Date: Fri, 09 Jul 1999 11:44:49 -0400
Subject: Decent Tutorial
References: <000015af@bossar.com.pl>
Message-ID: <378618F1.5564ED47@bellsouth.net>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19990709/f940235e/attachment.html>

From da at ski.org  Tue Jul 20 00:19:50 1999
From: da at ski.org (David Ascher)
Date: Mon, 19 Jul 1999 21:19:50 -0700 (Pacific Daylight Time)
Subject: Call for Panelists
Message-ID: <Pine.WNT.4.05.9907192110440.174-100000@david.ski.org>

Call for Panelists

The program for the O'Reilly Python Conference includes the following
moderated, 50-minute panels:

o Python Success Stories
	o Date: Monday, 23 August 1999
	o Moderator: Uche Ogbuji of FourThought LLC

o Joys and Pains of Building an Open-Source Community
	o Date: Tuesday, 24 August 1999
	o Moderator: Paul Everitt of Digital Creations

These panels are intended to be informal, structured discussions of the
topics by members of the python community.

In addition to panel members already determined, the program committee
invites conference attendees and members of the Python community to
participate.  We seek individuals who have successfully applied Python to
interesting computer problems in business, academics, or government; and
individuals who have been involved in Python-based open-source projects.

Interested parties should send e-mail by August 1 to monterey at egroups.com
with the following information:

 o Name, affiliation, phone number and e-mail address of the panelist

 o A paragraph on the successful use of Python, or relevant open-source
   project.

 o A brief biography of the panelist

 o Whether or not the panelist has registered to attend the conference

The O'Reilly Python conference is part of the O'Reilly Open Source
Convention (http://conferences.oreilly.com/) from August 21-24, 1999 in
Monterey, California.  This conference includes six concurrent conferences
featuring a wide array of exciting open-source technologies and
personalities.  The Python conference includes four tutorials, 15
presentations, birds-of-a-feather sessions, and more.

The Program Committe
Python Conference






From billtut at microsoft.com  Fri Jul 30 19:52:03 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Fri, 30 Jul 1999 23:52:03 GMT
Subject: Need help with TCP/IP client access from Windows
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90FF7@RED-MSG-50>


> -----Original Message-----
> From: Eric S. Raymond [mailto:esr at thyrsus.com]
> 
> 
> > > (2) How does one read and write the Windows equivalent of 
> a dotfile?
> > >     Must this involve the (shudder) registry?
> > 
> > You could, by using Mark's win32 extensions. MS would prefer you
> > kept this kind of info in the registry, but that doesn't stop
> > anybody, least of all you <g>.
> > 
> > The issue is really what is meant by a HOME directory. On NT, this 
> > has a certain amount of meaning - there's probably a 
> > %WINDIR%/profiles/%USERNAME% directory tree for that user (where 
> > those funky %...% things are environment vars), but probably it's 
> > safer to get that path out of the registry. On Win9x, it's pretty 
> > meaningless, since a logon and username are pretty much an empty 
> > formality. There's almost no way to keep multiple users on a Win9x 
> > box from stepping on each other. (Of course, I mean multiple serial 
> > users; the only possible kind in Windows).
> 

Just to be annoying... I thought I'd send out some info on how things are
supposed to work on Win32 boxes. 

The correct way to go about doing these things is via the SHGetFolderPath
API.
(Calling SHGetSpecialFolderPath is also ok, but SHGetFolderPath is the
preferred API entry point.)
Docs at: 
http://msdn.microsoft.com/library/sdkdoc/shellcc/Shell/Functions/ShGetFolder
Path.htm

A description of all of the special folders is at:
http://msdn.microsoft.com/library/sdkdoc/shellcc/shell/Functions/CSIDL.htm


Ok so can I use these in Python yet?
Unfortunately, not at the moment. If someone screams and yells I'm sure Mark
or I will write the necessary wrapping code.

The preferred place on NT (ignoring terminal server) for per user
application data files to reside in is:
"%WINDIR%\Profiles\%USERNAME%\Application Data\<appname>"
If you don't want the data to roam (assuming your account is setup that
way), the data should go in:
"%WINDIR%\Profiles\%USERNAME%\Local Settings\Application Data\<appname>"

Bill









From Sven.Drescher at dlr.de  Tue Jul 20 05:47:41 1999
From: Sven.Drescher at dlr.de (Sven Drescher)
Date: Tue, 20 Jul 1999 11:47:41 +0200
Subject: os.pipes() - need help!
Message-ID: <7n1frh$poo@news.go.dlr.de>

Hi!

I'm really hopeless. I tried for a week to use pipes, but nothing happend.
The information I found on the homepages and in Deja-News aren't enough for
me.
I found the following example:

#!/usr/local/bin/python

import os
import sys

#pipes definieren

pipe1 = os.pipe() # parent stdin, child stdout
pipe2 = os.pipe() # child stdin, parent stdout

pid=os.fork()
if pid:
    os.close(pipe1[1]) # child stdout - close
    os.close(pipe2[0]) # child stdin - close
    os.dup2(pipe1[0], 0) # sys.stdin = parent stdin
    os.dup2(pipe2[1], 1) # sys.stdout = parent stdout

else:
    os.close(pipe1[0]) # parent stdin - close
    os.close(pipe2[1]) # parent stdout - close
    os.dup2(pipe2[0], 0) # sys.stdin = child stdin
    os.dup2(pipe1[1], 1) # sys.stdout = child.stdout

    os.execv("/usr/openwin/bin/xterm", ["xterm"])

    sys.stdout.write("ls\n")
    sys.stdout.flush()

So, now I don't know why nothing happens. I tried various combinations and
also print and others - no result!!!

Thanks for help!

Sven





From wolfgang.grafen at de.bosch.com  Mon Jul 26 08:17:16 1999
From: wolfgang.grafen at de.bosch.com (Wolfgang Grafen)
Date: Mon, 26 Jul 1999 14:17:16 +0200
Subject: seqdict - 2 sequential dictionaries for python
References: <mt2.0-27614-932479015@news.informatik.uni-bonn.de>
Message-ID: <379C51CC.B496B966@de.bosch.com>

26.7.1999 

Package seqdict 

seqdict - single value dictionary 
msedict - multiple value dictionary 

The Python built-in dictionary has no reliable sequence order because
of performance reasons. 

The two new dictionaries introduced here keep their entries in sequence. 

* seqdict keeps one value for one key 
* mseqdict keeps multiple values per key and is the closest emulation 
  of areal world dictionary. 

As the sequential dictionary is a combination of a list and a dictionary
you can do most operations defined with lists as well and much more. 

Now you can slice, add, sort with dictionaries. The functional
operations 
map, filter, reduce are implemented as well. Use a dictionary like a
stack 
with push and pop. Split it, reverse it and swap keys and values. 

A small tutorial is provided as HTML and Postscript Source and is
subject
to become improved some other day. 

I think both dictionaries are candidates for the python standard
library. 

Use the seqdicts to emulate a printed dictionary. Use them for workinse 
them for working with csv-tables. Keep parsed code in a seqdict, modify
it
and write it back ... 

Download from 
http://www.germany.net/teilnehmer/100,366919/Python/Modules/Modules.html

Wolfgang Grafen 
mailto:WolfgangGrafen at gmx.de
-------------- next part --------------
A non-text attachment was scrubbed...
Name: wolfgang.grafen.vcf
Type: text/x-vcard
Size: 329 bytes
Desc: Card for Wolfgang Grafen
URL: <http://mail.python.org/pipermail/python-list/attachments/19990726/627f0d65/attachment.vcf>

From stidolph at origin.ea.com  Wed Jul 21 10:31:36 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Wed, 21 Jul 1999 14:31:36 GMT
Subject: VC++ 6.0 larger exe files explained
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC15@molach.origin.ea.com>

This might (stress this is a pure guess) have something to do with virtual
memory and its size being 4k pages.  The goal is not to make things bigger,
but to make them faster/more powerful and it is a common trade of space for
time.  I'm not saying they succeeded, but compilers used to live and die by
their benchmarks.

-----Original Message-----
From: Robin Becker [mailto:robin at jessikat.demon.co.uk]
Sent: Wednesday, July 21, 1999 8:30 AM
To: python-list at cwi.nl
Subject: VC++ 6.0 larger exe files explained


<rant>
As part of the Wintel push to take over the world it seems as though
VC++ 6.0 will now only generate files with sections 0x1000 (4k) or
longer. Under 5.0 and before the linker would allow much smaller exe
file sections. So python.exe which used to be 5k becomes 20k. Needless
to say this should also apply to dll's etc and appears to offer no
obvious advantage. Bill has never been other than cavalier with our
space time resources.
</rant>
-- 
Robin Becker



From mcfletch at vrtelecom.com  Sat Jul 31 02:33:32 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Sat, 31 Jul 1999 06:33:32 GMT
Subject: mxTextTools Package - Version 1.1.0
In-Reply-To: <37A05256.E18A9380@lemburg.com>
References: <37A05256.E18A9380@lemburg.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIIEECICAAA.mcfletch@vrtelecom.com>

Marc-Andr?'s mention at the bottom of his page shamed me into creating some
documentation and examples... so...

Announcing:
	SimpleParse -- A Parser Generator for mxTextTools -- version 1.0

What is it:
	A simple parser generation system based on Marc-Andr? Lemburg's mxTextTools
tagging engine.  The generator is a very small python module which creates
tuple trees suitable for use directly with the mxTextTools engine.

What's new:
	Mostly documentation and some reorganization.

Where can I get it:
	http://members.home.com/mcfletch/programming/simpleparse/simpleparse.html
	(I was going to put this on starship, but I can't seem to access that
account any more).

What does it cost:
	Python license. No cost. Use at your own risk.

References:
	http://starship.skyport.net/~lemburg/mxTextTools.html


-----Original Message-----
From: python-list-request at cwi.nl [mailto:python-list-request at cwi.nl]On
Behalf Of M.-A. Lemburg
Sent: July 29, 1999 9:09 AM
To: python-list at cwi.nl
Subject: ANN: mxTextTools Package - Version 1.1.0


ANNOUNCING:

                   mxTextTools Version 1.1.0

          A Python extension package providing fast text
                 scanning and processing tools
...




From mfletch at vrtelecom.com  Fri Jul  2 14:25:26 1999
From: mfletch at vrtelecom.com (Mike C. Fletcher)
Date: Fri, 02 Jul 1999 18:25:26 GMT
Subject: FOX-USERS: Re: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net><3773c3a0.102825445@news.oh.verio.com><xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net><xJRc3.292$Of3.67094@news.shore.net><7laovo$c4s$1@Starbase.NeoSoft.COM><377A24B7.AD6FAA1A@cfdrc.com><3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net><377B711A.E8B65E0B@cfdrc.com><377BC620.F5553FF5@easystreet.com> <377BF583.6846FBF9@quasar.ipa.nw.ru> <NF6f3.6498$_A4.93414@news2.rdc1.on.home.com> <377CFD82.8EA084A2@cfdrc.com>
Message-ID: <qu7f3.6732$_A4.94859@news2.rdc1.on.home.com>

I should clarify, when I'm speaking about "rich" text editing widgets, I'm
talking about colouring/styling, font selection, selection management, and
preferably some sort of layout engine.  Rich text as in RTF format, or HTML
can both easily be built on top of a generic "rich" text widget.

My personal ideal would be to allow a cascading style sheet-style layout
system.

Looking at the code for FXEditor, it seems to be very tightly focused on
source-code editing, but subclassing it to allow for a colouriser which also
changes fonts looks possible (it uses a low-level device context to draw
graphic fonts).  It is not a good base for a full layout engine, however, as
it rather lacks in abstraction, doesn't provide for formatting blocks, and
is generally a straight-forward-no-looking-back text-placement engine.
Possibly adapting the Mozzilla layout engine (Gecko?) might be a better
option.

Enjoy yourselves,
Mike


Lyle Johnson <jlj at cfdrc.com> wrote in message
news:377CFD82.8EA084A2 at cfdrc.com...
...
> I think any kind of rich text (i.e. RTF-compliant) or HTML widget is much
> farther off for FOX.
...





From tim_one at email.msn.com  Sun Jul 18 16:33:19 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 18 Jul 1999 16:33:19 -0400
Subject: Python and Boehm-Demers GC, I have code.
In-Reply-To: <o+zLoAAp4ak3Ewjk@jessikat.demon.co.uk>
Message-ID: <000101bed15c$c5685ca0$e7a22299@tim>

[Tim jabbers about cycles in IDLE, and a mysterious cycle-finding tool]

[Robin Becker]
> very interesting; I'm spending a bit of time on Zope at the moment. This
> if anything will be a long running thing which shouldn't leak memory.
> Any chance of an analysis?

Not from me -- I don't even know how to spell Xoap.

> Or the (I'm sure all the features are good) version of plumbo?

It's Cyclops.py, and I just uploaded it to python.org's ftp contrib site.
It should show up in the System directory later this week (send a complaint
every 5 minutes  to ftpmaster at python.org until it does <wink>).

that-barry-gets-too-much-sleep-ly y'rs  - tim





From alan.gauld at gssec.bt.co.uk  Wed Jul 28 11:36:22 1999
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Wed, 28 Jul 1999 16:36:22 +0100
Subject: another beginners tutorial
Message-ID: <379F2376.E7525AB5@gssec.bt.co.uk>

I am writing yet another absolute beginners tutorial 
for Python. (Actually its supposed to be language 
independant and does use BASIC and Tcl too but it 
majors on Python!)

I would appreciate any feedback from real newbies to 
programming - the aim is to get you up to speed so 
you can use the 'official tutorial' that comes with 
the Python download (or indeed the Java. Tcl Perl etc 
tutors too).

What distinguishes it from the other existing intros 
is that it tries to go into more detail on the concepts 
and only teach generic methods(for list access/searching etc) 
rather than Python best practice(slicing etc).

Its about 40-50% there, the url is:

http://members.xoom.com/alan_gauld/tutor/tutindex.htm

Please send feedback to me by mail at:

mailto:alan.gauld at btinternet.com
TIA,

Alan G.


-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.



From amullhau at nospam.ix.netcom.com  Sun Jul 11 23:34:49 1999
From: amullhau at nospam.ix.netcom.com (Andrew P. Mullhaupt)
Date: Sun, 11 Jul 1999 23:34:49 -0400
Subject: [Matrix-SIG] Re: Example Code: Bresenham Line Algorithm
References: <7m9tln$6qg$4@rks1.urz.tu-dresden.de> <Pine.LNX.4.10.9907112154520.15581-100000@us2.mayo.edu>
Message-ID: <7mbo2o$8ik@dfw-ixnews14.ix.netcom.com>

Travis Oliphant <Oliphant.Travis at mayo.edu> wrote in message
news:Pine.LNX.4.10.9907112154520.15581-100000 at us2.mayo.edu...
>
> I'm not sure what the Bresenham Line Algorithm does.

Bresenham's algorithm is a fast way to sequentially decide which pixels to
set in order to represent a line or ellipse on a rectangular grid.

> I could evaluate whether or not I'd like to include
> it in the package that I'm developing (Multipack),

I'd be surprised if Multipack was the right place for a Bresenham
implementation. Maybe PIL, or a more likely, a graphics device driver is
where you find these line and ellipse drawing things.

There is at least one alternative algorithm to Bresenham, which on some
architectures is significantly faster for sufficiently large display sizes.
My brother Patrick and I developed one a few years back, but found the
display size required to outperform Bresenham was too large given the
architectures of the time.

Later,
Andrew Mullhaupt





From Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org  Thu Jul  1 23:28:45 1999
From: Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 04:28:45 +0100
Subject: Python GUIs: Summary and Conclusion
Message-ID: <00001651@bossar.com.pl>

From: mikael at toolwood.com (Mikael Lyngvig)

On Thu, 01 Jul 1999 20:13:59 GMT, "Michael P. Reilly"
<arcege at shore.net> wrote:

>Been there, done that.  My pyirc client (alpha release) was designed
>for just about any user-interface, and comes with three: dummy (no
>input or output, just batch-mode)), curses and Tkinter.  You just can't
>switch between them at runtime, ;)  I've done it for other apps too.

Well, the only reason I haven't been there is that I've only worked on
one GUI app _ever_ - a 2D tile-based strategy game [1].  I did plan
support for console apps (curses, etc., using a client-server
architecture), but I didn't much fancy the idea of playing with a
command line interface :)

Anyway, it could be fun to have wxBatch and wxConsole variants of
wxWindows.  wxBatch would then, supposedly, map a wxCheckBox into a
boolean option, a wxTextCtrl into a file name or a string option, and
so on.  (I don't have any good ideas for wxPostScriptDC and
wxLayoutConstraints, though).  wxConsole should be fairly trivial,
except for the pixel-fiddling parts.

Now, there's a task!

[1] Which I expect to change in the future, which is why I'm reviewing
portable GUIs (I don't do non-portable stuff, if there's anything I
can do to avoid it).


-- Mikael



From Nick at p98.f112.n480.z2.fidonet.org  Thu Jul  1 17:28:32 1999
From: Nick at p98.f112.n480.z2.fidonet.org (Nick)
Date: Thu, 01 Jul 1999 22:28:32 +0100
Subject: Python suitable for a game engine?
Message-ID: <00001631@bossar.com.pl>

From: "Nick" <nick at NOSPAMvideosystem.co.uk>

Don't want to put the cat amongst the pigeons but Lua
(http://www.tecgraf.puc-rio.br/lua/), is a language specifically designed to
be embedded and it runs about 50-100% faster than Python and is a lot more
compact (30k or something). It has a lot of similarities syntactically as
far as I can see.

 If I was to use an embedded scripting language in a game I would use Lua
not Python. If I was to write tools for the game I would use Python.

Cheers,
Nick.



Alex Maranda <amaranda at nospam.com> wrote in message
news:377B7E17.5BD7 at nospam.com...
>> their leader is killed, etc. The events would be things such as
>> "create", "monster sees player", "clock tick", etc. All the functions
>> in the script would be callbacks written in C++, to manipulate the
>> monster to do things such as "patrol area", "play an animation",
>> "sleep for X seconds", etc. In other words, the scripts would be glue,
>> not the main thread of execution.
>you use confusing terms; I understand what you're saying, but you
>shouldn't use the word 'thread' for anything else than its OS meaning.
>You could say the 'the bulk of processing'. You might very well end up
>using real Python threads; don't forget in this case to release/reaquire
>the global interpreter lock in your C++ callbacks, if you make blocking
>syscalls (unlikely in a game).
>
>>
>> I'm been using Python for a couple of years now, and it is my favorite
>> scripting language. But, I don't have much experience with embedding
>> it in a another program, and I don't know enough about the speed
>> issues. So, my questions are...
>>
>> - Will Python be fast enough? Even with possibly 100 or 200 active
>>   items on a level?
>Unlikely. You won't know until you try. To my knowledge UnrealScript is
>eventually translated to C++.
>
>> - Especially when all of the scripts will be live, and usually in the
>>   middle of a callback?
>> - Should I use the interpreter in the library, or should I take an
>>   approach of crunching the byte-codes myself?
>??
>
>Cheers,
>--
>Alex Maranda         mailto: amaranda at spider dot com
>Spider Software Ltd. Tel: +44 (0)131 4757036
>Edinburgh, UK        http://members.xoom.com/Alex_Maranda
>STREAMS based communications protocols for embedded systems




From spud200 at earthlink.net  Mon Jul 26 19:37:37 1999
From: spud200 at earthlink.net (Spud)
Date: Mon, 26 Jul 1999 16:37:37 -0700
Subject: Python and Red Hat Linux 6
References: <379CD547.22A10A50@postoffice.pacbell.net> <379CD5C2.5BA096A2@mail.danbbs.dk>
Message-ID: <7nirk7$icq$1@oak.prod.itd.earthlink.net>

Yeah, using #!/bin/env/python should help, but Red Hat should support
pound-bang (#!) anyway. You might want to check in with tech support.

Brian Pedersen <brian.pedersen at mail.danbbs.dk> wrote in message
news:379CD5C2.5BA096A2 at mail.danbbs.dk...
> Rob wrote:
> >
> > When programming in Perl or Python, I am using a header like the
> > following on an executable file to cause the shell (bash or csh) to run
> > the scripts as Perl/Python files:
> > #!/usr/bin/perl   for Perl   and #!/usr/bin/python   for Python. I've
> > checked and the compilers are located in the appropriate directories. I
> > keep getting this when I run scripts:
> > bash: <script name>: command not found. Does Linux support the '#!'
> > notation in the file header, or do I always have to run the scripts by
> > explicitly typing in 'perl' or 'python' before the script name?
> > ~Rob~
>
> Try
>
> !#/usr/bin/env python
>
> instead (don't ask me why, i read it in a FAQ somewhere :-)
>
> --
> Brian Pedersen, M.Sc.E.E.                         /\
> DSP Software and Algorithm Development       .~. /  \ .~.
> URL: http://www.danbbs.dk/~kibria/brian/        ~    ~
> NOTE: New e-mail address !






From gmcm at hypernet.com  Fri Jul 30 12:49:38 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 30 Jul 1999 11:49:38 -0500
Subject: Need help with TCP/IP client access from Windows
In-Reply-To: <19990730113931.D17721@thyrsus.com>
References: <1278785133-25350622@hypernet.com>; from Gordon McMillan on Fri, Jul 30, 1999 at 08:25:58AM -0500
Message-ID: <1278772913-26085625@hypernet.com>

Eric S. Raymond wrote:

> Gordon McMillan <gmcm at hypernet.com>:
> > Probably the same directory where you put your script. (The Win31 
> > convention of putting all ini files in the Windows directory under 
> > <appname>.ini is now, thank goodness, passe).
> 
> OK.  Will giving the filename ".\\intriguer" work for this?  And,
> given that the filename is going through the C library, will leaving
> it as "./intriguer" work?

Yes and yes. File Manager doesn't know what to make of an 
extensionless file name, but that's OK. For years I created comma 
delimited files as ".cdf". Then one day I installed IE 4, and they 
magically became "Channel Definition Files".

This is good. It means that while Ionesco is dead, his spirit 
lives on.

> (Sorry for the newbie questions, but I have zero Windows-programming
> experience.) -- 

No prob. As long as you don't expect extreme conformance out 
of the c runtime lib, you'll do fine.

a-foolish-consistency-etc.-ly y'rs

- Gordon



From fredrik at pythonware.com  Mon Jul 26 17:27:22 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 21:27:22 GMT
Subject: filter() + simple program not outputting
References: <379CC972.750BD42D@netcard.com>
Message-ID: <000d01bed7ad$a6c20600$f29b12c2@secret.pythonware.com>

(posted and mailed)

Joyce Stack <joyce at netcard.com> wrote:
> New to python but not to programming...can someone tell me why this
> simple program is not outputting when i run from the BASH prompt....any
> pointers would be helpful...thanking you in advance....
> 
> Joyce
> 
> #!/usr/bin/env python
> 
> def f(x):
>         return x%2 != 0 and x%3 != 0
>         filter(f, range(2, 25))
>         f( )

dunno. what is it supposed to do?  there's
no print statement in there, nor anything
else related to printing.  and while "f" is
declared, it's never called.  and that
function looks a bit strange -- it starts
out by returning to the caller, and thus
never runs any code.

but it's close.  could the following be
what you tried to do:

--- cut ---
#!/usr/bin/env python

def f(x):
    return x%2 != 0 and x%3 != 0

print filter(f, range(2, 25))
--- cut ---

note the use of indentation -- things at
the left margin are executed when you
run this as a script ("def" is a statement
too!).

</F>




From claird at Starbase.NeoSoft.COM  Fri Jul 16 12:26:32 1999
From: claird at Starbase.NeoSoft.COM (Cameron Laird)
Date: 16 Jul 1999 11:26:32 -0500
Subject: I envy newbies
References: <1279991447-21085630@hypernet.com>
Message-ID: <2E4327EF0690EAF8.3DAF07256B1C5D39.F38C922CCDEC596E@lp.airnews.net>

In article <1279991447-21085630 at hypernet.com>,
Gordon McMillan <gmcm at hypernet.com> wrote:
			.
			.
			.
>Files\... [I think it's hilarious that MS was so enamored of their 
>"advance" in being able to put spaces in paths that they made the 
>"proper" place to install everything a demo of this "feature"... 
>Which demos the fact that spaces in path names are a pain in the 
>ass]. And if tcl80.dll and tk80.dll are on your path (which they 
Unarguable.  It's beyond my capacity to retain
my usual neutral mien whenever this comes to mind.
			.
			.
			.
>> "Yearning for NT when you're running Linux is like lusting for a Big
>> Wheel when you're driving a BMW." --Harrison Picot.
>
>More apt would be an cut-rate underpowered "luxury" car (you 
>know, where you have to turn off the A/C to go up a hill) and a John 
>Deere. Now Win9x, that's a tinker toy.
Gordon, we're going to have to continue to keep
you in the back room.  Retail consumers persist
in testifying they *want* those "underpowered
'luxury' cars".

Suggestive episode from the history of agricul-
tural mechanization:  one of the still
under-celebrated inventions was the three-point
hitch, which DRAMATICALLY simplified the quotidian
realities of hooking diverse implements to ones
chosen platform.  (Oh; and power-take-off coup-
lings, too.)  I leave implications to language
theory to the usual advocates of gluing
technology.


Most at-home on an iron-seated Farmall myself,
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From pserrand at p98.f112.n480.z2.fidonet.org  Fri Jul  2 08:29:45 1999
From: pserrand at p98.f112.n480.z2.fidonet.org (pserrand)
Date: Fri, 02 Jul 1999 13:29:45 +0100
Subject: test - ignore
Message-ID: <0000165b@bossar.com.pl>

From: pserrand <pserrand at worldnet.fr>

test with python



From avv at quasar.ipa.nw.ru  Fri Jul  9 12:27:09 1999
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Fri, 09 Jul 1999 09:27:09 -0700
Subject: Python and Paradox
References: <7m2mrj$hes$1@nnrp1.deja.com> <7m2o2b$aq0@dfw-ixnews12.ix.netcom.com> <37852804.54D5D2AD@quasar.ipa.nw.ru> <7m4q97$dpq$1@m2.c2.telstra-mm.net.au>
Message-ID: <378622DC.DD6EA3D@quasar.ipa.nw.ru>

Mark Hammond wrote:

> Alexander V. Voinov wrote in message <37852804.54D5D2AD at quasar.ipa.nw.ru>...
> >Aahz Maruch wrote:
>
> >> >I am looking for a way to access Paradox tables from Python without
> >> >using ODBC. Does anyone have any ideas?
> >>
> >> Nope, from an ex-Paradox tech support engineer.
>
> Yep, from someone who has :-)  Well, Im probably not technically correct,
> but if you:
> * Install MSAccess
> * Install Paradox drivers from the "Value Pack" (on the CD) - I think this
> does install an ODBC driver even tho we dont use it directly (hence my "not
> technically correct")

Another interesting question, for those who can check it: Does this Paradox DDL
also expose those older direct API, for those who still remember them? If this
engine was _not_ rewritten from the scratch, but just licensed, this would be
possible. But anyway, it's too slow even on the direct level. If one likes to
work with a C API database engine, I'd suggest to purchase Raima Data Manager.
Some time ago I moved a large office application from Paradox to db_Vista
relational submodel and gained great performance benefit. Certainly, I made
Python wrappers for a resonably higher level abstraction over its plain C API.
And RDM's network model would give you even more benefit if you realize how to
use it properly. (I speak it for myself, it's just an experience...)

Alexander





From tim_one at email.msn.com  Sat Jul 17 15:30:32 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Jul 1999 15:30:32 -0400
Subject: I envy newbies
In-Reply-To: <7mqji9$q1c@dfw-ixnews14.ix.netcom.com>
Message-ID: <000301bed08a$d63c68a0$27a22299@tim>

[Tim]
> I think the best advice you can give a newbie regardless is to force the
> install to put Python in a directory *they* choose, and one without
> spaces in its name.  Like C:\Python.  ..

[Aahz]
> The one thing I *hate* about this advice is that it's a Bad Idea to put
> too many files/directories into the root.  (I *think* FAT-32 gets rid of
> the 64-entry limitation, but I'm not sure, and anyone on Win95 or NT has
> a decent chance of being on FAT-16.  But even without the limitation, I
> still think it's a Bad Idea.)

FAT32 and NTFS don't care.  FAT16 actually has room for 512 entries,
although files/folders with "non-8.3 names" consume more than one entry.
Details at e.g.

    http://support.microsoft.com/support/kb/articles/q120/1/38.asp

You're right that you *can* get in trouble creating a file in a FAT16 root;
I do it routinely, though (well, just folders, not files), but I won't be
close to getting in trouble before I install about 300 more programs there
<wink>.

each-windows-user-finds-a-unique-way-to-survive-ly y'rs  - tim





From tim_one at email.msn.com  Mon Jul 19 22:04:00 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 19 Jul 1999 22:04:00 -0400
Subject: Killing a thread
In-Reply-To: <3793A6DB.731BDB83@compaq.com>
Message-ID: <000b01bed254$223f1d40$139e2299@tim>

[catlee at my-deja.com wrote:]
> - Is it conceivable that a user could make write some resource
> intensive code without using any kind of jump operations?

[Greg Ewing]
> Yes!
>
> def fib(n):
>     return {
>         0: lambda: 1,
>         1: lambda n=n: fib(n-1) + fib(n-2)
>     }[n>=3]()
>
> Although if you include function calls as well
> as jumps, you're probably safe (until Tim comes
> up with an even more perverted counterexample).

If the critical resource is time, measuring anything other than time is a
doomed strategy.  Consider this guy:

import re
from time import clock
for i in range(10):
    start = clock()
    re.search("x.*" * i + "y", "x" * 100)
    finish = clock()
    print i, round(finish - start)

This sets up some exponentially-expensive regexp searches.  Even as low as
i==4, the search

    re.search("x.*x.*x.*x.*y", "x" * 100)

consumes close to 40 seconds on my machine, and at i==5 well over 500.  I'm
not waiting for i==6 <wink>.

The first x matches at any of 100 places, the second at any from the first x
to the end, the third at any from the second to the end, and so on.  The "y"
at the end prevents any attempt from succeeding, and so forces all possible
choose(100, 4) combinations of "x" matches to be tried.  The search is
entirely in re's C code:  no bytecodes are interpreted-- nor function calls
made --for the duration.

To be really safe, I think you need to write some platform-specific C cruft
to spawn a distinct process, watch its progress, and kill it if it gets out
of hand.

Note too that it's no real trick to crash the Python interpreter (which is
why you need a separate process, not just a new thread):

a = []
b = []
a.append(a)
b.append(b)
print a == b

The number of ways you can fool Python similarly decreases every year; that
way in particular is widely known so I don't mind publicizing it even more
<wink>.

will-crash-python-for-big-bux-ly y'rs  - tim





From Mike.C..Fletcher at p98.f112.n480.z2.fidonet.org  Thu Jul  1 10:22:54 1999
From: Mike.C..Fletcher at p98.f112.n480.z2.fidonet.org (Mike C. Fletcher)
Date: Thu, 01 Jul 1999 15:22:54 +0100
Subject: TopSort in Python?
Message-ID: <00001612@bossar.com.pl>

From: "Mike C. Fletcher" <mfletch at vrtelecom.com>

Well, it's not topsort-the-official-algo, but when I first described it, Tim
Peters said it was a type of topological sort.  Warning: it's extremely
inefficient, using recursive calls of all things!  I did this without any
knowledge of the computer science in the field (have since learned some, but
didn't fix this code), it was just a description of an algorithm I needed to
get a job done, and I don't use it any more.

Cheers,
Mike


Dinu C. Gherman <gherman at my-deja.com> wrote in message
news:7lfjn0$b1v$1 at nnrp1.deja.com...
> Does anybody have a simple-minded-but-working full-Python
> implementation of topsort, the topological sorting algorithm?
> Or maybe *any* topsort? I remember only one such algorithm...
> python.org/search doesn't reveal any...
>
> Thanks,
>
> Dinu
>
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.
>


begin 666 dsort.py
M;G5L;'9A;" ]("@Q+"D-"@T*8VQA<W, at 1%-O<G0Z#0H))R<G#0H)02 B9&5P
M96YD96YC>2(@<V]R=&EN9R!C;&%S<RP@=7-E9"!T;R!O<F1E<B!E;&5M96YT
M<PT*"6%C8V]R9&EN9R!T;R!D96-L87)E9" B9&5P96YD96YC:65S(B H;6%N
M>2UT;RUO;F4@<F5L871I;VYS:&EP<RD-"@E)<R!N;W0 at 82!B96%U=&EF=6P@
M86QG;RP at 8G5T(&ET('=O<FMS("AO<B!S965M<R!T;RD-"@E297%U:7)E<R!H
M87-H86)L92!V86QU97, at 9F]R(&%L;"!E;&5M96YT<RX-"@D-"@E4:&ES(&ES
M(&$@<75I8VL@:&%C:RP@=7-E(&%T('EO=7(@;W=N(')I<VLA#0H)#0H)0F%S
M:6,@=7-A9V4Z#0H)"4-R96%T92!A($13;W)T(&UY<V]R=&5R#0H)"69O<B!E
M86-H(&5L96UE;G0@<2!W:&EC:"!I<R!P87)T(&]F('1H92!S970@=&\@<V]R
M="P at 8V%L;#H-"@D)"6UY<V]R=&5R+G)U;&4H(&1S;W)T+FYU;&QV86PL('$I
M#0H)"0DC('1H:7,@:7,@;F]T('-T<FEC=&QY(&YE8V5S<V%R>2!F;W(@96QE
M;65N=',@=VAI8V@@87)E#0H)"0DC(&1E<&5N9&5N="!O;B!O=&AE<B!O8FIE
M8W1S+"!B=70@:70@:7,@;F5C97-S87)Y(&9O<@T*"0D)(R!T:&]S92!W:&EC
M:"!A<F4@;F]T+B @1V5N97)A;&QY(&ET)W, at 96%S:65S="!T;R!C86QL#0H)
M"0DC('1H92!N=6QL(')U;&4 at 9F]R(&5A8V@@96QE;65N="X-"@D)9F]R(&5A
M8V@@<G5L92!X(&1E<&5N9',@;VX@>2P at 8V%L;#H-"@D)"6UY<V]R=&5R+G)U
M;&4H('@L('DI#0H)"7=H96X at 7V%L;%\@<G5L97, at 87)E(&5N=&5R960L(&-A
M;&P-"@D)=')Y. at T*"0D)<V]R=&5D;&ES=" ](&UY<V]R=&5R+G-O<G0H*0T*
M"0EE>&-E<'0 at 5F%L=65%<G)O<CH-"@D)"6AA;F1L92!R96-U<G-I=F4 at 9&5P
M96YD96YC:65S(&AE<F4N+BX-"@D)#0H)"0T*"49O<B!A;B!E>&%M<&QE(&]F
M(')E86PM;&EF92!U<V4L('-E92!T:&4 at 5E)-3"!L:6YE87)I<V5R+ at T*"0T*
M"2<G)PT*"61E9B!?7VEN:71?7RAS96QF+"!R96-U<G-E17)R;W(]3F]N92 I
M. at T*"0ES96QF+F1E<&5N9&]N(#T@>VYU;&QV86PZ6S!=?0T*"0ES96QF+G)E
M8W5R<V5%<G)O<B ](')E8W5R<V5%<G)O<@T*"61E9B!R=6QE*"!S96QF+"!D
M97!O;BP at 9&5P<RDZ#0H)"2<G)PT*"0E296=I<W1E<B!A(")R=6QE(BX@($)O
M=&@@96QE;65N=',@;75S="!B92!H87-H86)L92!V86QU97,N#0H)"5-E92!T
M:&4 at 8VQA<W,G(&1O8W5M96YT871I;VX at 9F]R('5S86=E+ at T*"0DG)R<-"B,)
M"7!R:6YT("<G)W)E9VES=&5R:6YG(')U;&4Z)R<G+"!D97!O;BP at 9&5P<PT*
M"0EI9B!S96QF+F1E<&5N9&]N+FAA<U]K97DH(&1E<',@*2!A;F0 at 9&5P;VX@
M:7,@;F]T(&YU;&QV86PZ#0H)"0ES96QF+F1E<&5N9&]N6R!D97!S(%TN87!P
M96YD*"!D97!O;B I#0H)"65L:68 at 9&5P;VX@:7,@;F]T(&YU;&QV86PZ#0H)
M"0ES96QF+F1E<&5N9&]N6R!D97!S(%T@/2!;+3$L(&1E<&]N70T*"0EE;&EF
M(&YO="!S96QF+F1E<&5N9&]N+FAA<U]K97DH(&1E<',@*3H-"@D)"7-E;&8N
M9&5P96YD;VY;(&1E<', at 72 ](%LM,2!=#0H)9&5F('-O<G0H('-E;&8@*3H-
M"@D))R<G#0H)"4=E="!T:&4@<V]R=&5D(')E<W5L=', at 87, at 82!L:7-T#0H)
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M+'D@:6X@=&5M<#H-"@D)"71E;7 R+F%P<&5N9"@@>2 I#0H)"2, at 9F]L;&]W
M:6YG(&%D9',@=&AE(&5L96UE;G1S('=I=&@@;F\@9&5P96YD96YC:65S#0H)
M"71E;7 R6VQE;BAT96UP,BDZ72 ]('-E;&8N9&5P96YD;VY;(&YU;&QV86P@
M75LQ.ET-"@D)<F5T=7)N('1E;7 R#0H)9&5F(%]D<V]R="@@<V5L9BP@:V5Y
M+"!V86QU92 I. at T*"0EI9B!V86QU95LP72 ]/2 M,CH-"@D)"6EF('-E;&8N
M<F5C=7)S945R<F]R. at T*"0D)"7)A:7-E(%9A;'5E17)R;W(L("<G)T1E<&5N
M9&5N8VEE<R!W97)E(')E8W5R<VEV92$G)R<-"@D)"65L<V4Z#0H)"0D):68@
M7U]D96)U9U]?. at T*"0D)"0EP<FEN=" G)R=296-U<G-I=F4 at 9&5P96YD96YC
M>2!D:7-C;W9E<F5D(&%N9"!I9VYO<F5D(&EN(&1S;W)T+D1S;W)T+E]D<V]R
M="!O;B E<SHE<R<G)R4H:V5Y+"!V86QU92D-"@D)"0ER971U<FX@,2 C('=E
M(&MN;W<@:70@:&%S(&%T(&QE87-T(&]N92!R969E<F5N8V4N+BX-"@D)96QI
M9B!V86QU95LP72 ]/2 M,3H@(R!H879E;B=T('EE="!C86QC=6QA=&5D('1H
M:7,@<F1E<'1H#0H)"0EV86QU95LP72 ]("TR#0H)"0ET96UP=F%L(#T at 6S!=
M#0H)"0EF;W(@>"!I;B!V86QU95LQ.ETZ#0H)"0D)=')Y. at T*"0D)"0ET96UP
M=F%L+F%P<&5N9"@@,2 K('-E;&8N7V1S;W)T*"!X+"!S96QF+F1E<&5N9&]N
M6WA=*2 I#0H)"0D)97AC97!T($ME>45R<F]R. at T*"0D)"0ES96QF+F1E<&5N
M9&]N6R!N=6QL=F%L(%TN87!P96YD*"!X("D@(R!I<R!A;B!U;G)E9F5R96YC
M960 at 96QE;65N= T*"0D)"0ET96UP=F%L+F%P<&5N9"@@,2 I#0H)"0EV86QU
M95LP72 ](&UA>"@@=&5M<'9A;" I#0H)"0ER971U<FX@=F%L=65;,%T-"@D)
M96QS93H-"@D)"7)E='5R;B!V86QU95LP70T*)R<G#0IF<F]M(&UC9BYU=&EL
M<R!I;7!O<G0 at 9'-O<G0-"CX^/B!X(#T at 9'-O<G0N1%-O<G0H*0T*/CX^(&UA
M<"@@>"YR=6QE+"!;,2PR+#(L-"PU+#1=+"!;,BPS+#0L-2PV+#-=("D-"EM.
M;VYE+"!.;VYE+"!.;VYE+"!.;VYE+"!.;VYE+"!.;VYE70T*/CX^('@N<V]R
(="@I#0HG)R<`
`
end



From gherron at aw.sgi.com  Wed Jul 21 15:07:13 1999
From: gherron at aw.sgi.com (Gary Herron)
Date: Wed, 21 Jul 1999 19:07:13 GMT
Subject: pow()
References: <37961668.FDF78B51@callware.com>
Message-ID: <37961A61.64480218@aw.sgi.com>

Ivan Van Laningham wrote:
> 
> Hi All--
> I couldn't find an explanation for the three-argument version of pow()
> on DejaNews, the Python 1.5.2 docs or the Python FAQ, so I'll ask now.
> 
> Can someone explain (in simple terms, for a bear of small brain) to me
> what the modulo argument to __pow__ is for?
> 
>         __pow__(self,other[,modulo])
> 

This is from the library manual named "2.3 Built-in Functions":


pow (x, y[, z]) 

Return x to the power y; if z is present, return x to the power y,
modulo z (computed more efficiently than pow(x, y) % z). The arguments
must have numeric types. With mixed operand types, the rules for binary
arithmetic operators apply.  The effective operand type is also the type
of the result; if the result is not expressible in this type, the
function raises an exception; e.g., pow(2, -1) or pow(2, 35000) is not
allowed. 

-- 
Dr. Gary Herron <gherron at aw.sgi.com>
206-287-5616
Alias | Wavefront
1218 3rd Ave, Suite 800, Seattle WA 98101



From amk at mira.erols.com  Wed Jul 14 21:46:59 1999
From: amk at mira.erols.com (A.M. Kuchling)
Date: 15 Jul 1999 01:46:59 GMT
Subject: Python re module
References: <199907120337.XAA06072@eric.cnri.reston.va.us> <14217.63630.458437.222314@amarok.cnri.reston.va.us> <7mfsgt$ood$1@nnrp1.deja.com>
Message-ID: <slrn7oqfq7.dn.amk@mira.erols.com>

On Tue, 13 Jul 1999 17:20:43 GMT, catlee at my-deja.com <catlee at my-deja.com> wrote:
>Ok, that makes sense.  Two questions: where can I get the latest
>version of PCRE and how can I tell (from python) what version of PCRE I
>am using?

Two negative answers; you can't upgrade PCRE easily, because various
Python-specific patches need to be made to it.  I have an alpha
release of a module based on a more recent version of PCRE, but it's
very early work.  Also, there isn't a way to tell what version of PCRE
you have; simply referring to the version of Python is sufficient,
because it's only going to change with new Python releases.  

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
    "Yes, but who do you work for?"
    "Work for? I don't work *for* anybody; I'm just having fun."
    -- Rigg and the Doctor, in "Nightmare of Eden"



From Gerrit.Holl at p98.f112.n480.z2.fidonet.org  Thu Jul  1 15:44:16 1999
From: Gerrit.Holl at p98.f112.n480.z2.fidonet.org (Gerrit Holl)
Date: Thu, 01 Jul 1999 20:44:16 +0100
Subject: getopt and Non-Unixes?
Message-ID: <0000161f@bossar.com.pl>

From: Gerrit Holl <gerrit.holl at pobox.com>

Hello all,

I was wondering how getopt worked on dos/windows (and other operating systems).
While in dos, the default way to handle arguments is "/v /d /f" (isn't it?).

I don't have access to any micro$oft machine (it doesn't make me unhappy...),
but I just wanted to know how my program works on Dos/windows/macintosch and
other non-Unix systems.

regards,

Gerrit.




From rgh at inmet.com  Fri Jul 23 13:00:11 1999
From: rgh at inmet.com (Randy Hudson)
Date: Fri, 23 Jul 1999 17:00:11 GMT
Subject: Jpython Question
References: <Pine.GSO.4.05.9907191525580.20433-100000@allegro> <wkr9m3846k.fsf@ifi.uio.no> <Pine.GSO.4.05.9907231035200.14333-100000@allegro>
Message-ID: <37989F9B.9DBBBA98@inmet.com>

Ian Kezsbom wrote:
> 
> Just out of curiosity...lets say I've written a Java program, with many
> classes and I want to run that program through JPython and have access to
> all its classes...how would I go about doing this...

I've done this; here's how:

First, define a thread that will become the main thread of the program.

  from java.lang import Thread, String
  from wherever import MyAppMainClass
  from jarray import array    # a method to convert a Python list to a Java array
  class AppThread(Thread):    # defines a class AppThread that extends java.lang.Thread
    def __init__(self,args):  # define the AppThread constructor
                              # This example takes the args list for the
                              # call to main as a constructor 
                              # "self" is the conventional ID for Java "this"
      self.args = args
    def run(self):            # override  Thread.run()
      MyAppMainClass.main( array(self.args,String) )

At this point we've define a class AppThread to run the program on another
thread. Create an instance of this thread, with the appropriate "command line"
arguments:

  appThread = AppThread(["-runSilent", "-runDeep"])

If you need to set any Java properties, do it now:

  from java.lang import System
  props = System.getProperties()
  props['maxDepth'] = '20000'    # like a -DmaxDepth=20000 switch to jre

Now start the program

  appThread.start() # standard Thread.start(), which calls the run() defined above

At this point you can look at anything publicly accessible

  from wherever import InstrumentPanel
  nicely_display( InstrumentPanel.getDepth() )

-- Randy Hudson



From earlybird at mop.no  Tue Jul 13 15:57:29 1999
From: earlybird at mop.no (Alexander Staubo)
Date: Tue, 13 Jul 1999 21:57:29 +0200
Subject: HTML parser/reformatter/filter wanted
Message-ID: <MPG.11f5b895bd447488989685@news.mop.no>

I'm looking for a piece of Python code that can filter HTML and tie up 
unclosed tags.

By way of explanation: I'm building a bulletin-board-ish solution where 
users should be able to post HTMl-formatted submissions; I only want to 
enable a subset of HTML (eg., <b>, <i> and so forth), plus I want to make 
sure text-level tags are ended correctly. I could sit down and write a 
smaller HTML parser, but I don't want to reinvent wheels if I can just 
grab this stuff off the shelf.

-- 
Alexander Staubo             http://www.mop.no/~alex/
"He could open a tin of sardines with his teeth, strike a Swan Vestas 
on his chin, rope steers, drive a steam locomotive and hum all the 
works of Gilbert and Sullivan without becoming confused or breaking 
down in tears."
--Robert Rankin, _The Book of Ultimate Truths_



From Arnaldo at p98.f112.n480.z2.fidonet.org  Fri Jul  2 04:20:31 1999
From: Arnaldo at p98.f112.n480.z2.fidonet.org (Arnaldo)
Date: Fri, 02 Jul 1999 09:20:31 +0100
Subject: Getting to know Python
Message-ID: <0000169a@bossar.com.pl>

From: "Arnaldo" <javanet at sinbad.net>

I just downloaded and compiled python in a sunos4.1 OS. I'm starting to have
fun with it but I'm really new
to the language.
I'm planning to buy a book this weekend so I can learn more about python's
syntax and posibly implement some of my c-shell/perl scripts in python.
In the mean-time I'm eager to have with the interactive interpreter.
How do I access the shell from the interpreter to enter commands like
,clear, ls, cp, etc.
Is it posible, how do I do that?

thanks guys
ajr





From gmcm at hypernet.com  Tue Jul  6 10:43:47 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 6 Jul 1999 09:43:47 -0500
Subject: changing some lines in a file
In-Reply-To: <3781FA33.92876E30@daimi.au.dk>
Message-ID: <1280853815-64731922@hypernet.com>

Peter Posselt Vestergaard writes:

> Hi, I'm quite a newbie in Python-programming, so I hope you'll
> forgive me asking about some pretty things. :-) My problem is that I
> have a file in which i want to find some lines and replace them with
> some other data. I've tried a lot of things but nothing seems to
> work.

Unfortunately, you haven't said what "doesn't work" means. 

> Below is my best try, but it seems that the script stops when
> it is about to execute the line:

Stops how? With a traceback?

> line=('fi='+Form["fi"].value)
> 
> I'll be thankfull for any kind of help.
> Best regards
> 
> Peter Vestergaard
> 
> ------------------------------------------------------------------
> 
> 
> db = open("db.dat","r+")		

Why are you opening for update?
				 
> data = db.readlines()
> db.close

MIssing something?

> 
> found=0
> for line in data:
>     print found
>     if found==1:
>         if line[0:2]=='fi':
>             line=('fi='+Form["fi"].value)

Ah. Well you'll never see any results, because you're rebinding the 
variable "line" to a new string. When you return to the "for", you 
lose all your hard work.

You can either modify the data list by using 

for i in range(len(data)):
  line = data[i]
  .....
     data[i] = 'fi=' + Form[whatever]....

Or create a new list and append the modified lines to it.

(Updating a file in place usually means lots of seek()s and tells() 
and is basically never worthwhile on text files.)

>         elif line[0:2]=='la':
>             line='la='+Form["la"].value
>         elif line[0:2]=='em':
>             line='em='+Form["em"].value
>         elif line[0:2]=='ho':
>             line='ho='+Form["ho"].value
>         elif line[0:2]=='st':
>             line='st='+Form["st"].value
>             found=2
>     if line[0:9]==('id='+Form["id"].value):
>         found=1
> if found==0:
>     print '<H1>Fejl i ?rskortnummeret!</H1>'
>     stop 
> 
> db = open("db.dat","w")
> db.writelines(data)
> db.close
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list

- Gordon



From Rob at p98.f112.n480.z2.fidonet.org  Fri Jul  2 10:17:20 1999
From: Rob at p98.f112.n480.z2.fidonet.org (Rob)
Date: Fri, 02 Jul 1999 15:17:20 +0100
Subject: How do you make clickable Python/Win9x scripts?
Message-ID: <0000169f@bossar.com.pl>

From: Rob <vaton at postoffice.pacbell.net>

Is there a way to make a Python script file a clickable .BAT file in
Win9x?
It's possible with Perl, and looks something like this:
@rem = '
@echo off
C:\perl\perl5\bin\perl -x -S %0 %1%
goto endofperl
@rem ';
#!.\perl\bin\perl -w

print "Hello World!\n";
__END__
:endofperl

The above Perl file would be named world.bat.




From Gary.Pennington at uk.sun.com  Thu Jul  8 13:16:49 1999
From: Gary.Pennington at uk.sun.com (Gary Pennington - UK Performance Centre)
Date: Thu, 08 Jul 1999 18:16:49 +0100
Subject: Memory management & Garbage Collection - pickle/cPickle
References: <3784D60F.49AFFDCB@uk.sun.com>
Message-ID: <3784DD01.8DC77DFD@uk.sun.com>

Ignore earlier post, I must have circular references in my code. I'll try
to re-design to get rid of them.

Gary

Gary Pennington - UK Performance Centre wrote:

> Hi,
>
> I've written some server side code which makes extensive use of shelve
> (and indirectly pickle). I've noticed that over time the amount of heap
> used by the server grows considerably. Investigation showed that objects
> create from pickle or cPickle cause memory to be allocated that is never
> freed.
>
> The following test program soon uses up all the memory on my machine :-
>
> import sys
> sys.path.insert(0,"/home/garyp/dev/webJudge/dev/cgi")
> import shelve
> d=shelve.open("gameDB")
> print d.keys()
> def getObj():
>                 return d["remote1"]
> for i in range(0,100000):
>                 x=getObj()
>                 del x
>                 x=None
>
> gameDB contains some objects previously pickled and about 30k in size.
> Running the above and using truss I can see memory being allocated (lots
> of brk calls) as the data is continuosly re-read from the file.
>
> The del calls and x=None seem to make no difference. Also, it makes no
> difference whether I go through shelve or access the data files direct.
> My system is using dumdbm.py, it's Solaris 2.6.
>
> Any suggestions as to what the problem is?
>
> Gary Pennington




From roy at popmail.med.nyu.edu  Sun Jul 25 13:00:36 1999
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Sun, 25 Jul 1999 13:00:36 -0400
Subject: Access dictionary in order?
Message-ID: <roy-2507991300370001@mc-as01-p61.med.nyu.edu>

I've got a dictionary of items which I will sometimes want random access
to, but sometimes I'll want to be able to step through the items in a
fixed order as well.  Any good way to do that?

The best I've come up with so far is to keep a parallel list of the
dictionary keys in sorted order and use that to step through when I need
to:

   for key in keylist:
      do_stuff (dict[key])

which will work, but seems kind of clumsy.  Anything smoother I can do?



From Greg.Ewing at p98.f112.n480.z2.fidonet.org  Mon Jul  5 16:33:22 1999
From: Greg.Ewing at p98.f112.n480.z2.fidonet.org (Greg Ewing)
Date: Mon, 05 Jul 1999 21:33:22 +0100
Subject: Binary pickle portability
Message-ID: <000017cc@bossar.com.pl>

From: Greg Ewing <greg.ewing at compaq.com>

Olaf Delgado wrote:
> 
> After looking a bit closer, I found that the problem is probably related
> to the Numeric module (i.e., pickling matrices) and doesn't depend on
> whether binary or ordinary mode is used.

I think you may have been right the first time.
The Numeric module always seems to pickle its 
matrices in a binary format, whichever pickler
is used. I suspect it's just dumping whatever
is in memory, so as you say, it's non-portable.

If you want a portable format, you may have to
convert the Numeric matrices to standard Python
lists or tuples before pickling.

Greg



From dmiron at metz.une.edu.au  Wed Jul 28 06:08:22 1999
From: dmiron at metz.une.edu.au (David Miron)
Date: Wed, 28 Jul 1999 20:08:22 +1000
Subject: Gadfly DBase and CGI
Message-ID: <7nmkpg$cjl$1@gruvel.une.edu.au>

Hi All,

I am trying to use the KWP database within a cgi script.  This is on a dec
unix machine running apache.

Has anyone ever done this?  I keep getting an error in the script every time
I go to create a new connection to an existing database.  What permissions
must I have (755)?

Thanks guys

David





From thomas at xs4all.nl  Fri Jul 23 05:57:52 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Fri, 23 Jul 1999 11:57:52 +0200
Subject: "libmysqlclient.so.6" not found. (what does /sbin/ldconfig /usr/local/lib/mysql do ?)
In-Reply-To: <7n9c7f$ihd$1@nnrp1.deja.com>; from sweeting@neuronet.com.my on Fri, Jul 23, 1999 at 09:21:55AM +0000
References: <7n9c7f$ihd$1@nnrp1.deja.com>
Message-ID: <19990723115752.D4068@xs4all.nl>

On Fri, Jul 23, 1999 at 09:21:55AM +0000, sweeting at neuronet.com.my wrote:

> However, upon rebooting the machine, if I try to run a Python program
> it will again complain that "libmysqlclient.so.6" can not be found
> (since the MySQLmodule is shared). Once again
> running " /sbin/ldconfig /usr/local/lib/mysql"  cures the problem but
> does this mean that I have to run it each and every time I restart the
> computer, as part of my startup scripts ? Is this error indicative of
> there being a larger problem or misconfiguration ?

Add /usr/local/lib/mysql to your /etc/ld.so.conf. It contains a list
(seperated by newlines, if i'm correct) of directories to add to the dynamic
linker/loader search path.

Alternatively (and imho preferable over adding to your ld.so.conf each time)
place a symbolic link in /usr/local/lib or /usr/lib to point to
/usr/local/lib/mysql/libmysqldclient.so.6. That way you dont suffer the
clutter of dos-like PATH addenda :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From billtut at microsoft.com  Tue Jul 27 15:20:26 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Tue, 27 Jul 1999 19:20:26 GMT
Subject: MS-SQL Server:Connectivity Help...
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90FC2@RED-MSG-50>

There are three approaches.

1) Use the Win32 COM support to connect to SQL Server via ADO.
This is perhaps the most functional access to the database you're likely to
get in Python atm.
2) Use the Win32 extension ODBC module. 
This only exposes a limited interface to the database, and has several known
annoying bugs. (Date/time type binding being the most annoying...)
3) Use the mxODBC extension. 
This still only exposes a limited interface to ODBC, but fixes some of the
annoying bugs #2 currently has.
http://starship.python.net/crew/lemburg/
mxODBC also has a curious license which you should investigate before using.

Personally, I'd vote for #2 or #3 if you can.
ADO slows things down just a tad due to it having to convert to Unicode and
back again for sane use in Python.

Bill

> -----Original Message-----
> From: Edward B. Wilson II [mailto:zero at ecom.net]
> Sent: Tuesday, July 27, 1999 9:09 AM
> To: python-list at cwi.nl
> Subject: MS-SQL Server:Connectivity Help...
> 
> 
> I would like to introduce Python with a bang to my new work 
> group.  Does
> anyone have a driver to "insert" a single row into an MS-SQL server7.0
> db.  It appears from the main Python page, one must compile their own
> ODBC drivers with VC++ to get the most robust solution.  Can anybody
> comment on this shortcoming.  Getting Python to work for 
> MS_SQL/NT would
> be a big entrance.
> 
> Thanx--
> 
> Ed
> 
> 
> 



From ikezsbom at eecs.tufts.edu  Mon Jul 19 15:27:14 1999
From: ikezsbom at eecs.tufts.edu (Ian Kezsbom)
Date: Mon, 19 Jul 1999 15:27:14 -0400
Subject: Jpython Question
Message-ID: <Pine.GSO.4.05.9907191525580.20433-100000@allegro>

I had a quick question on Jpython and I'm not sure if this is the right
place to post, but I figured I would try anyway...

I was just wondering if jpython can instantiate java classes and invokes
methods on them...any help would be welcome...

Thanks,

Ian




From nick at NOSPAMvideosystem.co.uk  Tue Jul  6 13:18:57 1999
From: nick at NOSPAMvideosystem.co.uk (Nick Trout)
Date: Tue, 6 Jul 1999 18:18:57 +0100
Subject: Python suitable for a game engine?
References: <0000163d@bossar.com.pl>
Message-ID: <378238af@news.xtml.co.uk>

>I had a quick look, but not the time to check it completely.
>Does it have coroutines or small threads to do very
>quick context switches? This is something which
>Python will be able to do very fast, soon. You can have
>thousands of current contexts and switch them in no time.
>This can simplify the implementation of games in a scale,
>that the speed penalty might be compensated by that.
>Well, we have to try...

You don't need a proper threaded system to do game AI. Every object just get
s X ticks each frame. This is maybe getting a bit off topic. A context switc
h can be as quick as changing a global pointer to the new current context. I
think Unreal works as you explain, running lots of objects in parallel. When
you create a new "live" object it spawns a new "thread". Essentially this is
just a list of objects that have a member function tick() which gets called
every frame. Please correct me if I'm wrong.

The way I'd look at is: which languages are suitable ie. can do what you wan
t, can be embedded conveniently. And for games, which is fastest and most co
mpact?!! Lua and Python can do what you want and can be embedded but Pythons
byte code execution is slower than Lua's by about 50% or more and its a lot
smaller. Python is streets ahead for writing tools/parsers/simple utils with
a GUI. Although you can do all that in Lua as well!

I'm interested in your ideas. I have thought about doing similar things. I c
ame to the conclusion that Lua was more suitable for embedding in a game. Pe
rhaps you could elaborate on your ideas.

N






From tim_one at email.msn.com  Sat Jul  3 14:38:05 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 3 Jul 1999 14:38:05 -0400
Subject: ordered threading
In-Reply-To: <TbLVhCAmDPf3EwPU@jessikat.demon.co.uk>
Message-ID: <000a01bec583$30332b40$769e2299@tim>

[Robin Becker]
> I have an unbounded number of worker threads which occasionally require
> access to a unique server thread. I can easily mutex access to the
> server thread, but is there any obvious/easy way to ensure that the
> blocked workers wake up in the right order?

No threading system on earth allows specifying the order in which blocked
threads will wake up, although some systems (not Python's) allow you to
*influence* the order via assorted "priority" gimmicks.

You could easily enough build an OrderedLock class whose acquire and release
methods implement whatever scheme you want, hiding the unordered nature of
an internal mutex under a layer of serial-number cruft.  But then why not do
these things serially in a single thread to begin with?

If you have frequent rigid ordering requirements, threads are likely the
wrong approach.

> Also if blocked can a thread be killed in any way?

No, not short of killing the whole program.  A thread blocked on a condition
that will never become true is a symptom of confused application design, not
of a weakness in the thread model <wink>.

expand-the-condition-to-include-"or-i-should-die-now"-ly y'rs  - tim





From amackay at starvision.com  Mon Jul  5 21:27:28 1999
From: amackay at starvision.com (Angus MacKay)
Date: Mon, 05 Jul 1999 18:27:28 -0700
Subject: VIM and block jumping
References: <3781347D.CB2FBAB3@starvision.com> <Pine.SOL.4.05.9907051548470.13052-100000@saturn.radsci.uci.edu>
Message-ID: <37815B80.B4B51F94@starvision.com>

I am not looking for indentation but just code navigation.

when code uses identifiers for blocks (such as "()" in lisp, "{}" is
most sane languages, "#if/#endif" for C preproc ...) vim can navigate
blocks very easily (with [{/]} for "{}", [#/]# for "#if/#endif" ...)

with python this becomes harder. is there any way to do it? (I think
I could write a function for it but I thought I'd see if it had
already been done)

cheers, Angus.

Fuming Wang wrote:
> 
> Hi, I have a small patch that can make vim do following indentation:
> def func():
>     print "hello world!"
>     if a == b and \
>        b == c or  \
>        e == f:
>         print "got it!"
> 
> I can send you the patch if this is what you like.
> 
> Cheers,
> Fuming Wang
> fuming at venus.radsci.uci.edu
> 
> On Mon, 5 Jul 1999, Angus MacKay wrote:
> 
> > this is a question for all of the VIMers out there.
> >
> > is there any way to get [# ]# style logical python block jumping
> > to work for python? (to make up for the lack of {} in python)
> >
> > cheers, Angus.
> >
> >



From paul at prescod.net  Thu Jul  1 10:12:18 1999
From: paul at prescod.net (Paul Prescod)
Date: Thu, 01 Jul 1999 10:12:18 -0400
Subject: Mailbox, tell and read
References: <1281323312-36476914@hypernet.com>
Message-ID: <377B7742.F39CB0A4@prescod.net>

Gordon McMillan wrote:
> 
> How 'bout opening in binary, and munging line endings yourself just
> before returning the message? 

First, are there platforms other than DOS/Windows/Win32 where there is a
difference between text and binary mode? I didn't want to munge things
myself because I'm not sure if what is the "right" munge on the Mac, for
instance. Can I just do a search and replace for CR/LF -> CR on all
platforms, all of the time?

> If you can't use tell(), you can't trust seek(), either.

Good point.

-- 
 Paul Prescod  - ISOGEN Consulting Engineer speaking for only himself
 http://itrc.uwaterloo.ca/~papresco

"I don't want you to describe to me -- not ever -- what you were doing
to that poor boy to make him sound like that; but if you ever do it
again, please cover his mouth with your hand," Grandmother said.
	-- John Irving, "A Prayer for Owen Meany"



From arcege at shore.net  Tue Jul 20 10:37:39 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Tue, 20 Jul 1999 14:37:39 GMT
Subject: Are postings to comp.lang.python.announce mirrored here?
References: <37933DB1.47814E84@cfdrc.com> <14227.22618.779710.843912@anthem.cnri.reston.va.us> <JNPk3.532$ui4.178537@news.shore.net> <14227.60175.280720.305790@anthem.cnri.reston.va.us>
Message-ID: <TQ%k3.598$ui4.194325@news.shore.net>

Barry A. Warsaw <bwarsaw at cnri.reston.va.us> wrote:

:>>>>> "MPR" == Michael P Reilly <arcege at shore.net> writes:

:     MPR> FYI, I've attempted some cross-posts to both which didn't
:     MPR> even make it to the moderator.  Cross-posting is the Right
:     MPR> Thing(TM), but not all servers seem to handle it when one is
:     MPR> moderated.  Seperate posts is at times the only option (if
:     MPR> you want it announced in both newsgroups).

: Ah!  You can always just email it to both python-list at python.org and
: python-announce at python.org.

*Whine&Pout* But then it's not cross-posting.

Winking-ly yrs
  -Arcege




From Ken.Liu at p98.f112.n480.z2.fidonet.org  Sat Jul  3 15:51:42 1999
From: Ken.Liu at p98.f112.n480.z2.fidonet.org (Ken Liu)
Date: Sat, 03 Jul 1999 20:51:42 +0100
Subject: Problem with informixdb and python1.5.2
Message-ID: <000016b3@bossar.com.pl>

From: Ken Liu <kyliu at idiomtech.com>

Thomas Bryan <tbryan at arlut.utexas.edu> wrote:

> Do you now have both python 1.5.1 and python1.5.2 installed?

Yes. 

> Have you tried recompiling informixdb with the 1.5.2 header files?  
> I don't think that too much has changed between 1.5.1 and 1.5.2, but 
> it couldn't hurt to try.

That sounds like the thing to do.  In fact, I did change the PYTHONINCL and
PYTHONLIB dirs
to point to the 1.5.2 directories and recompiled informixdb, but the same
problem persists.  

so is the common experience that the modules don't work when you have multiple
versions of
python installed?  Why is that?  Is there anyway for me to "unconfuse" the
module so that
it would ignore the 1.5.1 installation (short of actually removing it...)?


thanks much

k



From Gaetan_Corneau at baan.com  Wed Jul 14 14:52:31 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 14 Jul 1999 18:52:31 GMT
Subject: Redirect stdout to file
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF47E@ex-quebec-u1.baan.com>

Frederik,

now, keep in mind that all python variables
are just named pointers.  so:

stdout = sys.stdout

creates a variable called "stdout" in *your*
module, which points to the same thing as
"sys.stdout" currently points to...
[GC] 
OK, I didn't know how my "from ... import ..." was interpreted. I thought is
was really giving me the real object (sys.stdout in this case). 

I will remember that :)

Thanks again!
______________________________________________________
   Gaetan Corneau
   Software Developer (System integration Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"



From sjoerd at oratrix.nl  Wed Jul  7 09:31:45 1999
From: sjoerd at oratrix.nl (Sjoerd Mullender)
Date: Wed, 07 Jul 1999 15:31:45 +0200
Subject: Emacs: highlighting "self" tweak
In-Reply-To: Your message of Wed, 07 Jul 1999 06:06:38 +0000.
             <7luqpa$3s9$1@nnrp1.deja.com> 
References: <7luqpa$3s9$1@nnrp1.deja.com> 
Message-ID: <19990707133146.F3AD4301CF2@bireme.oratrix.nl>

On Wed, Jul 7 1999 Preston Landers wrote:

> Hello,
> 
> This is what I did to highlight 'self' in my Python files.
> 
> It's a small (unified context) patch against python-mode.el version
> 3.90, the lastest as far as I know.  It shouldn't be too hard to patch
> it against older versions of python-mode (you only need to insert 3
> lines at a certain spot.)
> 
> I wish I knew of an easy way to make it a hook instead of a patch.  If
> someone can figure that out, I would *love* to hear from you.

You can use this piece of magic in your .emacs file (tested on XEmacs
21.1):

(add-to-list 'after-load-alist
	     (cons "python-mode"
		   '((add-to-list (quote python-font-lock-keywords)
				  (quote ("\\b\\(self\\)[ .,]?"
					  1 font-lock-type-face))))))

This changes the python-font-lock-keywords variable right after
python-mode is loaded.

-- Sjoerd Mullender <sjoerd.mullender at oratrix.com>



From mark.mcmahon at eur.autodesk.com  Mon Jul 19 12:42:32 1999
From: mark.mcmahon at eur.autodesk.com (mark.mcmahon at eur.autodesk.com)
Date: Mon, 19 Jul 1999 18:42:32 +0200
Subject: debugging MSVC DLLs
Message-ID: <C2AA61E5CAA7D211832F0008C7A4138E01B486DB@eurmsgneu01.eur.autodesk.com>

I think that this is possible using the Additional DLL's in the 
Debug tab in MSVC...

Project -> Settings...
Debug tab
select Additional Dll's
- specify the dll's that you want to debug.

Note: I havn't done this myself so I dont know if it is what
I think it is.
 Please let me know

Mark

-----Original Message-----
From: Tom Adelman [mailto:ta16 at cornell.edu]
Sent: Monday, July 19, 1999 5:08 PM
To: python-list at python.org
Subject: debugging MSVC DLLs


I'm wondering if someone can help me with this (or should I direct this
question to a Microsoft group?):

What I'm doing now:
    I write classes in MSVC++, compile them as a DLL, and SWIG them.  I
then run python.exe from within the Microsoft Developer Studio, and then
within Python load this DLL.  This way, I can use the MS debugger to
look at what's going on within my classes while I manipulate them in
Python.
    The problem with this is that I can only debug the DLL whose project
I used to start Python; I don't know how to, for example, write a new
DLL and switch to debugging the new one.

What I want to do:
    I would rather run Python from within Windows, load various DLLs,
and debug whichever DLL I need, or a new one that I write, while
remaining in the same Python execution.

Does anyone know if it's possible to do this?  Can you tell me how?

Thanks for any help,
Tom Adelman
ta16 at cornell.edu





-- 
http://www.python.org/mailman/listinfo/python-list



From MHammond at skippinet.com.au  Mon Jul 26 20:04:08 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Tue, 27 Jul 1999 10:04:08 +1000
Subject: Python documentation available as ZIP files
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us> <7nal1u$6cb$1@tribune.oar.net>
Message-ID: <7nistg$p02$1@m2.c2.telstra-mm.net.au>

David Robinow wrote in message <7nal1u$6cb$1 at tribune.oar.net>...

> This is really sad.  You're not really doing Windows users any favors by
>doing this.
>"tar" and "gzip" are "must know" tools.
>Let me quote from the GNU macs on Windows FAQ

ROFL!!!

Let me guess - not only do you not use Windows, you don't know anyone who
does?

Mark.






From Nick at p98.f112.n480.z2.fidonet.org  Fri Jul  2 07:39:35 1999
From: Nick at p98.f112.n480.z2.fidonet.org (Nick)
Date: Fri, 02 Jul 1999 12:39:35 +0100
Subject: Python suitable for a game engine?
Message-ID: <0000165d@bossar.com.pl>

From: "Nick" <nick at NOSPAMvideosystem.co.uk>

The language doesn't necessarily need to support it. You could implement the
threaded system yourself in Lua, you then have more control over what goes
on. Proper threads are relatively slow in Windoze (I believe) and would be a
nightmare to debug? If Lua wasn't fast enough you could write a scheduler
with priority assessment in C/C++ and attached Lua script callbacks to this.
Personally I wouldn't dream of useing proper threads for game logic/AI I
would write a basic scheduler, it wouldn't be difficult. Grim Fandango uses
Lua as its scripting language and I think they did something along these
lines. I think Balders Gate is another game that is rumoured to use it.

N


Christian Tismer <tismer at appliedbiometrics.com> wrote in message
news:377BD789.C4A1D191 at appliedbiometrics.com...
>
>
>Nick wrote:
>>
>> Don't want to put the cat amongst the pigeons but Lua
>> (http://www.tecgraf.puc-rio.br/lua/), is a language specifically designed
to
>> be embedded and it runs about 50-100% faster than Python and is a lot
more
>> compact (30k or something). It has a lot of similarities syntactically as
>> far as I can see.
>>
>>  If I was to use an embedded scripting language in a game I would use Lua
>> not Python. If I was to write tools for the game I would use Python.
>
>I had a quick look, but not the time to check it completely.
>Does it have coroutines or small threads to do very
>quick context switches? This is something which
>Python will be able to do very fast, soon. You can have
>thousands of current contexts and switch them in no time.
>This can simplify the implementation of games in a scale,
>that the speed penalty might be compensated by that.
>Well, we have to try...
>
>ciao - chris
>
>--
>Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
>Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
>Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
>10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
>PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
>     we're tired of banana software - shipped green, ripens at home




From commando at netvigator.com  Sun Jul 18 08:11:29 1999
From: commando at netvigator.com (commando at netvigator.com)
Date: Sun, 18 Jul 1999 20:11:29 +0800
Subject: Please visit
Message-ID: <7msg8t$or7$1@imsp009a.netvigator.com>

http://www.digitalworld.com.hk/front_c.htm





From marcoxa at copernico.parades.rm.cnr.it  Thu Jul  1 02:54:44 1999
From: marcoxa at copernico.parades.rm.cnr.it (Marco Antoniotti)
Date: 01 Jul 1999 08:54:44 +0200
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.i <lwoghyexy5.fsf@copernico.parades.rm.cnr.it> <7ldt5p$srh$1@dragon.flightlab.com>
Message-ID: <lwhfnosvuz.fsf@copernico.parades.rm.cnr.it>

jenglish at flightlab.com (Joe English) writes:

> Marco Antoniotti wrote:
> 
> >BTW. Is TCL statically or dynamically scoped?
> 
> Neither, really, since Tcl doesn't have nested procedures, lambda
> expressions, "let/letrec", local bindings or the like.
> 
> I'd describe Tcl as "namespace-scoped": there is an open-ended set
> of top-level "namespace" environments, plus a distinguished "global"
> environment.  In addition, each Tcl "proc" has its own local environment,
> into which variables from other environments can be explicitly imported
> via "variable", "global", explicit namespace reference, or "upvar".
> "upvar" gives you something *sort* of like dynamic scope, but not
> exactly.

Now, who was saying that Common Lisp macros are a source of
inconsistencies? :)

Cheers

-- 
Marco Antoniotti ===========================================
PARADES, Via San Pantaleo 66, I-00186 Rome, ITALY
tel. +39 - 06 68 10 03 17, fax. +39 - 06 68 80 79 26
http://www.parades.rm.cnr.it/~marcoxa



From kikutani at sprintmail.com  Thu Jul 29 12:18:00 1999
From: kikutani at sprintmail.com (Kikutani Makoto)
Date: 29 Jul 1999 16:18:00 GMT
Subject: Qt
References: <7nkdce$t1c$1@oak.prod.itd.earthlink.net> <199907272335.JAA06521@envy.fulcrum.com.au> <7nm4i7$t59$1@oak.prod.itd.earthlink.net>
Message-ID: <7npuro$ga3$1@oak.prod.itd.earthlink.net>

On 28 Jul 1999 05:30:47 GMT, Kikutani Makoto <kikutani at sprintmail.com> wrote:

> > > mako% python aclock.py                /usr/local/src/Lang/pyKDEsrc-0.8/exampl
> > > es
> > > Traceback (innermost last):
> > >   File "aclock.py", line 4, in ?
> > >     from kde import *
> > >   File "/usr/lib/python1.5/kde.py", line 7, in ?
> > >     import libkdec
> > > ImportError: /usr/lib/libkimgio.so.2: undefined symbol: png_set_strip_16
> > > 
> > > What's wrong ?
> > > My Linux Box is Debian potato, and kdelibs2g is version 1.1.1-19990523.
> > > Python is 1.5.2 (python-gtk 0.6.1-1)
> > 
> >    Not sure - do you have the PNG library version that Qt requires? The example 
> > works for me under redhat 6.0 with qt 1.44 and png 1.0.3 (not sure what the 
> > actual requirements of Qt are though).
> 
> I have two PNG libraries (I don't know why).
> 
>   libpng2         1.0.2b-0.1     PNG library - runtime
>   libpng0g        0.96-5         PNG library - runtime
> 
> I didn't know Qt requires PNG library, because qt1g(1.44-3) package
> doesn't depend on libpng*.
> 
> Another related package is:
>   kdelibs2g       1.1.1-19990523 KDE core libraries (runtime files)

I thought current kdelibs2g had some inconsistent with PNG library.
So I got the source of kdelibs2g Debian package from:
ftp://kde.tdyc.com./pub/kde/debian
and rebuilt it.

Finally I could run the examples.
Thanks, Richard.

makoto

-- 
Kikutani, Makoto  kikutani at sprintmail.com




From daniels at mindspring.com  Sat Jul  3 00:49:51 1999
From: daniels at mindspring.com (Alan Daniels)
Date: 3 Jul 1999 04:49:51 GMT
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain>	<377B7E17.5BD7@nospam.com> <377bbf9b@news.xtml.co.uk> <377BD789.C4A1D191@appliedbiometrics.com> <377c870b@news.xtml.co.uk>
Message-ID: <slrn7nr5r7.qi.daniels@localhost.localdomain>

On Fri, 2 Jul 1999 10:39:35 +0100, the infinitely wise Nick
(nick at NOSPAMvideosystem.co.uk) spoke forth to us, saying...

[snip...]
>The language doesn't necessarily need to support it. You could implement the
>threaded system yourself in Lua, you then have more control over what goes
>on. Proper threads are relatively slow in Windoze (I believe) and would be a
>nightmare to debug? If Lua wasn't fast enough you could write a scheduler
>with priority assessment in C/C++ and attached Lua script callbacks to this.
>Personally I wouldn't dream of useing proper threads for game logic/AI I
>would write a basic scheduler, it wouldn't be difficult. Grim Fandango uses
>Lua as its scripting language and I think they did something along these
>lines. I think Balders Gate is another game that is rumoured to use it.

The only time I would consider adding threads is well after I had
everything else working (and I mean EVERYTHING), and wanted to
experiment around with it a bit. So far, my current plan is to use a
round-robin approach to keep the script processing from varying the
framerate too much, even if that means I can't use the interpreter but
have to translate the bytecodes myself.

The information on Lua is very useful, by the way. Thanks for this.

-- 
=======================
Alan Daniels
daniels at mindspring.com
daniels at cc.gatech.edu



From skip at mojam.com  Fri Jul 23 19:32:40 1999
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 23 Jul 1999 18:32:40 -0500 (CDT)
Subject: Converting itegers to strings
In-Reply-To: <19990723171921.A5205@pooh.frostnet.net>
References: <19990723171921.A5205@pooh.frostnet.net>
Message-ID: <14232.64055.800716.858880@227.chicago-26-27rs.il.dial-access.att.net>

>>>>> "Chris" == Chris Frost <chris at frostnet.advicom.net> writes:

    Chris> How can you convert numbers (an interger in this case) to
    Chris> strings? (i.e. I have some_number = 3, and would like to use it
    Chris> as a string)

There are a couple ways.  Try the % operator, backquotes or the str builtin
function.

    import math
    "%d" % 3
    type("%d" % 3)
    `math.pi`
    type(`math.pi`)
    ("%6.1f" % math.pi)
    type("%6.1f" % math.pi)
    str(math.pi)
    type(str(math.pi))

All should print
Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From nick.bower at ssec.wisc.edu  Thu Jul  8 18:25:12 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Thu, 08 Jul 1999 17:25:12 -0500
Subject: Python for the PalmPilot and Other Devices
Message-ID: <37852547.E5AC00F4@ssec.wisc.edu>

Is there any information, or even interest, in making a scaled-down
python interpretter for 3Com's PalmPilot series?  A new compact (<100k)
Micro-Java VM has just been released for it btw.

And more generally speaking, are there any plans for a micro-java-like
python version?  What is the feasibility of this?

nick

--

Nick Bower
Space Science and Engineering Center
University of Wisconsin, Madison, USA
Phone: (608) 265 8007
Email: nick.bower at ssec.wisc.edu
Web: http://arm1.ssec.wisc.edu/~nickb





From chris at frostnet.advicom.net  Wed Jul 21 00:09:30 1999
From: chris at frostnet.advicom.net (Chris Frost)
Date: Wed, 21 Jul 1999 04:09:30 GMT
Subject: Merging two lists as a dictionary
In-Reply-To: <14229.15052.805319.307262@buffalo.fnal.gov>; from Charles G Waldman on Tue, Jul 20, 1999 at 10:13:16PM -0500
References: <19990720211946.A12202@pooh.frostnet.net> <14229.15052.805319.307262@buffalo.fnal.gov>
Message-ID: <19990720230930.B12553@pooh.frostnet.net>

On Tue, Jul 20, 1999 at 10:13:16PM -0500, Charles G Waldman wrote:
> Unless I'm missing the point (which happens more than I'd like to
> admit!)  I think what you're looking for is just:
> 
> dicp={}
> for key,value in map(None,key_list,value_list):
>     dict[key]=value
Perfect! (I wasn't sure on how to do the last line...)

thanks,
-- 
Chris Frost  |  <http://www.frostnet.advicom.net/chris/>
-------------+------------------------------------------
Public PGP Key:
   Email chris at frostnet.advicom.net with the subject "retrieve pgp key"
   or visit <http://www.frostnet.advicom.net/chris/about/pgp_key.phtml>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 256 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19990721/4f247f79/attachment.sig>

From neelk at brick.cswv.com  Wed Jul 14 19:11:32 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 14 Jul 1999 18:11:32 -0500
Subject: Stackless & String-processing
References: <378A40AD.B4FDB79F@appliedbiometrics.com>
Message-ID: <7mj5f4$5pt$1@brick.cswv.com>

In article <378A40AD.B4FDB79F at appliedbiometrics.com>,
Christian Tismer  <tismer at appliedbiometrics.com> wrote:
>
>               Stackless Python 0.3
>
>The possibilities are for instance:
>
>Continuations, with which we will build Coroutines and Generators
>
>Coroutines are able to run at the speed of a single C function call,
>which makes them a considerable alternative in certain algorithms.
>This is no longer a speculation since I have working coroutine
>prototypes, written with continuations.

I've been looking at Icon, and it occurs to me that if coroutines and
generators were available at the Python level, it might yield a way of
doing string processing that is more "Pythonic" than regexps.

Regexps are nice because when you have a pattern that they can
directly represent, then you can simply specify a pattern and then you
don't have to worry about the tedious bookkeeping of looping over the
string and keeping track of the state variable.

However, when you need to match a pattern that a regexp can't match,
then suddenly you need to break the string into tokens with regexps
and then loop over the pieces and keep track of a bunch of state
variables that don't necessarily correspond to the pieces you are
actually interested in.

This is unpleasant, because a) clumsy bookkeeping is bad, and b)
there's two different sorts of syntax needed to do basically 
similar tasks. If we could compose generators just like functions,
then the bookkeeping can be abstracted away and the same approach
will work for arbitrarily complicated parsing tasks. 

So I think it would be nice if these lovely toys were available at
the Python level. Is this a possibility?

(I'd love to be able to define coroutines in Python like so:

def evens(z):
    for elt in z:
        if z % 2 == 0:
            suspend(z)

It would probably require some rethinking of Python's iteration
protocol though.)


Neel



From eugene.leitl at lrz.uni-muenchen.de  Thu Jul  1 21:14:07 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Thu,  1 Jul 1999 18:14:07 -0700 (PDT)
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
In-Reply-To: <b%Pe3.4033$lr3.164497@news2.jacksonville.net>
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net>
	<3773c3a0.102825445@news.oh.verio.com>
	<xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net>
	<xJRc3.292$Of3.67094@news.shore.net>
	<7laovo$c4s$1@Starbase.NeoSoft.COM>
	<377A24B7.AD6FAA1A@cfdrc.com>
	<3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net>
	<377B711A.E8B65E0B@cfdrc.com>
	<377BC620.F5553FF5@easystreet.com>
	<b%Pe3.4033$lr3.164497@news2.jacksonville.net>
Message-ID: <14204.4556.177837.851763@lrz.de>

Hank Fay writes:
 > Is it possible to use ActiveX controls in any of these GUI's?  Then you've
 > got a large choice of controls.

No system-dependencies here please. Particularly those relating to an 
obsolete system, doomed to die <0.02 wink>.



From mwh21 at cam.ac.uk  Sat Jul 17 08:33:26 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 17 Jul 1999 13:33:26 +0100
Subject: Snippet: bitcount()
References: <000601bed02a$2235af20$06a22299@tim> <m33dynlhsg.fsf@atrus.jesus.cam.ac.uk> <379074EE.B6D563F6@appliedbiometrics.com>
Message-ID: <m3n1wva221.fsf@atrus.jesus.cam.ac.uk>

Christian Tismer <tismer at appliedbiometrics.com> writes:

> > Mini-proposal: len(a) for a long returns the length of the number in
> > radix-256.
> 
> bitcount2 is not very slow, see here:
> 
> def bitlen1(x):
>     """number of bytes necessary to store a number"""
>     sign = x<0
>     x = abs(long(x))
>     h = hex(x)
>     l = (len(h)-3)*4
>     if sign and h[2] >= "8":
>         l = l+1
>     return (l+7)/8
> 
> import marshal
> 
> def bitlen2(x):
>     """number of bytes necessary to store a number"""
>     sign = x<0
>     x = long(x)
>     b=marshal.dumps(x)
>     l = (len(b)-5)*15 / 2
>     hi = ord(b[-1])*256 + ord(b[-2])
>     l = l + sign -15
>     while hi:
>         l = l + 1
>         hi = hi >> 1
>     return max(1, (l+7)/8)

Still chews gobs of memory though... is there any way of converting
longs <-> arrays? That'd do the trick, though you'd still end up
copying, no doubt. Still, I don't think I'm being realistic to talk of
dealing with numbers so large copying them is an issue in Python. It's
all just for play, anyway.

Cheers,
Michael

-- 
Oh, very funny. Very sar-cah-stic.         http://www.ntk.net
http://www.ntk.net/doh/options.html - ho ho



From Gerrit.Holl at p98.f112.n480.z2.fidonet.org  Fri Jul  2 10:27:49 1999
From: Gerrit.Holl at p98.f112.n480.z2.fidonet.org (Gerrit Holl)
Date: Fri, 02 Jul 1999 15:27:49 +0100
Subject: print color strings?
Message-ID: <00001662@bossar.com.pl>

From: Gerrit Holl <gerrit.holl at pobox.com>

On Thu, Jul 01, 1999 at 09:55:55PM +0000, Donn Cave wrote:
> "Fred L. Drake" <fdrake at cnri.reston.va.us> writes:
> |Gerrit Holl writes:
> |> is there an easy way to print color strings in python?
> |> 
> |> A function like printcolor('brightyellow', 'red', 'spam'), that finds
> |> out on what kind of terminal you are and prints it to stdout?
> |
> | Gerrit,
> |   If you're doing this on Unix, take a look at the ncurses library;
> | there's a Python interface to that.  (I think Oliver Andrich did it;
> | searching DejaNews should find it fairly quickly since this keeps
> | coming up.)
> |   The standard curses module will be documented in the next release of 
> | the Library Reference.
> 
> And if that doesn't work, probably your termcap or terminfo definitions
> lack the necessary information about color escape sequences.  That's a
> widespread problem.
> 
> The color escape sequences that have worked for me on a couple of terminal
> types have been of the form <ESC>[32m, where 32 is a number from 30 to 37.
> I use only 31 to 36.  30 and 37 are more or less black and white, but the
> details vary, so I switch from color to non-color with <ESC>[m instead of
> trying to sort out the details of black and white vs. background and
> foreground.
> 

Ah, thanks. I was forgotten that would work on python the same as in the
bash-shell...

> Of course these are terminal specific escape sequences.  They're evil and
> you must not use them, especially if you have a choice.
> 

Thanks. They're only for my sys.ps1, so it'll be good enough for me..,

But I thought there might be a simpler interface than a whole curses library,
but there's not.

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!




From larsga at ifi.uio.no  Fri Jul 30 03:28:24 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 30 Jul 1999 09:28:24 +0200
Subject: Need help with TCP/IP client access from Windows
References: <slrn7q1ej3.18d.esr@snark.thyrsus.com> <NDBBIAPHEKGOBIHCBLIIOECECAAA.mcfletch@vrtelecom.com> <19990730005536.U7149@thyrsus.com>
Message-ID: <wkr9lqwq93.fsf@ifi.uio.no>

* Mike Fletcher
| 
| import os
| os.system( 'start %s'%url )
| 
| Which does whatever the user has defined as the proper thing to do with
| urls, normally, launch IE or Netscape.

* Eric S. Raymond
| 
| Really?  That sounds perfect.  Just for my information, what is `start'?

A program that runs another program (according to the help text I got
when I ran it at the command-line prompt with no arguments). However,
on my old Win95 box it doesn't understand about URLs.

--Lars M.



From jepler at inetnebr.com  Sun Jul 18 18:29:02 1999
From: jepler at inetnebr.com (Jeff Epler)
Date: Sun, 18 Jul 1999 22:29:02 GMT
Subject: Strange side effect from reading a list of numbers from argv
References: <O3sk3.6047$Sg6.52046@news1.rdc1.sdca.home.com>
Message-ID: <slrn7p4lit.d2e.jepler@craie.inetnebr.com>

On Sun, 18 Jul 1999 21:55:58 GMT, Harold Gottschalk
 <heg at softbrew.com> wrote:
>I just wrote this program to become familiar with python and I do not
>understand why it works with a list I create with in the program and not
>when I use sys.argv[1:].

try it with
	for value in map(int, sys.argv[1:]):
or else you're working with strings.  It's come up before that "string <
int" doesn't cause an exception, but I don't remeber what the good reason
for it was.

Jeff
-- 
\/ http://www.infidels.org/                      Jeff Epler jepler at inetnebr.com
Ask yourself whether you are happy and you cease to be so.
		-- John Stuart Mill



From neelk at brick.cswv.com  Thu Jul 15 19:35:38 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 15 Jul 1999 18:35:38 -0500
Subject: Getters and Setters
References: <000701bece99$6dbeede0$51a22299@tim>
Message-ID: <7mlr8a$6o9$1@brick.cswv.com>

In article <000701bece99$6dbeede0$51a22299 at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>[Neil Schemenauer]
>> For those who are not familiar, CLOS (and Dylan) create a "getter"
>> and "setter" method for each attribute of a class. This saves the
>> programmer from writing "getX" and "setX" for every attribute when
>> no special action is needed.  Later, an explict method can be created
>> if needed.
>
>Python goes one better:  you can write an explicit method if needed, but it
>doesn't encourage bad practice by automatically supplying methods that
>expose every data attribute, crying "use me! abuse me!" <wink>.

Be fair, now -- Dylan doesn't try to do data hiding in the class
definition; it uses module export lists to determine what is visible
to users of the module. This isn't any worse that Python's leading-
underscore convention for data hiding.

Besides, Dylan has syntactically-significant whitespace just like
Python does, so it has to be a good language, right? :)

Neel



From billtut at microsoft.com  Mon Jul  5 19:08:38 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Mon, 5 Jul 1999 23:08:38 GMT
Subject: popen2 on Windows
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90EC1@RED-MSG-50>

> From: John van der Koijk [mailto:vdkoijk at linux01.prc.tno.nl]

> After coping with my initial frustration, I've decided to make an
> unthreaded stub program talking to the DDE client, using popen2 to do
> the IPC with the GUI and controller program. Sadly, this approach also
> does not seem to work, because using popen2 gives me:
> 
> [....] os.fork()
> AttributeError: fork
> 
> This is amazing, since the manual for 1.5.2 says:
> 
> 8.13 popen2 -- Subprocesses with accessible standard I/O streams
> 
>    Availability: Unix, Windows.
> 

Well, its wrong. :) Windows doesn't have fork().
The win32pipe module exposes popen2() that does what you want it to do,
however you need to avoid a bug in Win9x. The knowledge base article at:
http://support.microsoft.com/support/kb/articles/q150/9/56.asp
covers what you need to do to avoid this dilemma.

Bill





From Daniel_Faken at brown.edu  Fri Jul 23 11:02:55 1999
From: Daniel_Faken at brown.edu (Daniel Faken)
Date: Fri, 23 Jul 1999 11:02:55 -0400
Subject: Python interface to IBM Data Explorer?? (DX/OpenDX)
Message-ID: <3798841F.B9ED6E95@brown.edu>

Hello everyone,

  Greetings from Rhode Island.

  Has anyone developed a python interface (via SWIG or something else)
for the Data Explorer visualization system?
  IBM recently made it "OpenSource", and it seems very useful to me.
  There's a web site at http://www.opendx.org

thanks for any help,
Daniel Faken

-- 
Daniel B. Faken
Daniel_Faken at brown.edu
http://www.chem.brown.edu/brochure/people/grads/faken/



From gerrit.holl at pobox.com  Wed Jul 21 09:46:10 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Wed, 21 Jul 1999 13:46:10 GMT
Subject: [Tutor] I want to learn to hack
In-Reply-To: <001501bed279$8fc0b700$139e2299@tim>; from Tim Peters on Tue, Jul 20, 1999 at 02:31:55AM -0400
References: <14227.17376.361340.434400@nobot.cnri.reston.va.us> <001501bed279$8fc0b700$139e2299@tim>
Message-ID: <19990721154610.B3663@optiplex.palga.uucp>

On Tue, Jul 20, 1999 at 02:31:55AM -0400, Tim Peters wrote:
> > I want to learn to hack
> So, little Roger, you want to learn how to hack?  You came to the right
> place!  First walk down the hall and pick up a copy of Python from Guido.
> He'll have you cracking bank accounts in no time <wink>.
> 
> the-psa-is-the-richest-organization-on-earth-ly y'rs  - tim
> 

Well, I think Roger meant hack, not crack. Have a look at Eric S Raymonds
"How to become an hacker FAQ" at http://www.tuxedo.org/~esr/ (and some
further).

Hacking is good,
cracking is bad.

I also want to learn to hack, I'm learning it!
I *don't* want to learn to crack!!

regards,
Gerrit.



From fredrik at pythonware.com  Mon Jul 26 15:09:21 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 19:09:21 GMT
Subject: Extension module mystery
References: <379caeb9.0@cfanews.harvard.edu>
Message-ID: <001701bed79a$5efff240$f29b12c2@secret.pythonware.com>

Markus Demleitner <msdemlei at tucana.harvard.edu> wrote:
> Now, when I try the following:
> lis = ldw.ldw("Hol",["Kohl","Pfohl"]) 
> a = max(lis)
> (ldw.ldw being the glue function), I get
> Traceback (innermost last):
>   File "zw.py", line 3, in ?
>     a = max(lis)
> TypeError: argument 2: expected string, list found
> (that's python 1.5.2, FWIW).  Odd, isn't it?  Why
> should max insist on a string?

maybe it isn't max that generates that error;
instead, it looks like you've missed to test for
a Python exception somewhere (if you ignore
errors, and don't call PyErr_Clear() to clean up,
Python may raise them later...)

could it be that you forgot to change the glue
function to look for an object rather than a
string?

</F>




From gmcm at hypernet.com  Wed Jul  7 09:38:04 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 7 Jul 1999 08:38:04 -0500
Subject: Python array --> C
In-Reply-To: <3782CEF6.97D7C5FB@ncsa.uiuc.edu>
Message-ID: <1280771409-69691649@hypernet.com>

Randy Heiland writes:

> In my continuing quest to better understand embedding/extending, I
> have the following question...
> 
> in Demo/embed, I have modified demo.c so that the static module
> xyzzy_foo(self, args) now does the following (instead of returning
> 42L): {
>         unsigned char retval[5] = {0,2,4,8,16};
>         return Py_BuildValue("s#", retval,5);
> }
> i.e., it returns an array of unsigned chars.
> 
> From main(), I would like to invoke xyzzy.foo() and get this array
> of unsigned chars into a C array. How do I do this?  I assume I'll
> need to use PyRun_String(...) somehow - true?

Not at all. You'll want to use PyString methods to get its size and 
char *, etc. You can use python as a lib without ever involving the 
interpreter. See the C API reference manual (and be prepared to grep 
the objects and modules subdirectories).

- Gordon



From greg.ewing at compaq.com  Wed Jul 28 16:16:26 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 29 Jul 1999 08:16:26 +1200
Subject: IDLE's PyShell
References: <000801bed335$8c233e80$a72d2399@tim> <379689C1.B91E0D13@compaq.com> <mark.933063456@declan>
Message-ID: <379F651A.5ABA3818@compaq.com>

Mark C Favas wrote:
> 
> f = Tk.Toplevel()
> f.transient(parent)
> 
> to get a window without the close button

Doesn't work for me. It gets rid of the fullscreen
and iconify buttons, but it still has a close button.

Greg



From mvanier at bbb.caltech.edu  Thu Jul  8 21:34:08 1999
From: mvanier at bbb.caltech.edu (Michael Vanier)
Date: 08 Jul 1999 18:34:08 -0700
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.i <m3btdvhrlt.fsf@localhost.localdomain> <d2Of3.1233$U5.237262@ptah.visi.com>
Message-ID: <jln1x6wqr3.fsf@arathorn.bbb.caltech.edu>

thornley at visi.com (David Thornley) writes:
> In article <m3btdvhrlt.fsf at localhost.localdomain>,
> Lieven Marchand  <mal at bewoner.dma.be> wrote:
> >cbbrowne at news.brownes.org (Christopher B. Browne) writes:
> >
> >The difference between Moore and "just another of the opinionated
> >fools" is indeed that he gets results. I'm more into Common Lisp these
> >days but "Thinking Forth" is a book that has radically changed many of
> >my ideas about programming. 
> >
> I tend to learn new ways of thinking with new languages.  I came up with
> a really nice way of doing one particular thing in a C++ environment,
> and I'm positive it was inspired by "Lispy" thinking.  (Now if I can
> get my boss to buy in on it.....)
> 
> "Thinking Forth" is an excellent book to read.  So is Stroustrup's
> "Design and Evolution of C++".  I haven't seen a similar book about
> the Lisp way of thinking, not counting books specifically on learning
> Lisp and doing it well.  Then again, I find Common Lisp revolutionary
> enough sometimes.

IMHO the best book of this sort for Lisp is "On Lisp" by Paul Graham (and to a
lesser extent Graham's "ANSI Common Lisp").  After reading those books, you'll
never want to live without (Lisp) macros again ;-)  SICP is a *great* book,
but it's more about the Big Ideas In Computing than about Lisp/Scheme per se.
Graham's books describe how Lisp features (especially macros) allow you to
work at a higher level of abstraction than is possible in other languages, and
may evoke a religious transformation in susceptible readers ;-)

But-are-prefix-syntaxes-*really*-superior-to-infix-ones?-
You-be-the-judge-ly y'rs, Mike



-------------------------------------------------------------------------
Mike Vanier	mvanier at bbb.caltech.edu
Department of Computation and Neural Systems, Caltech 216-76
Will optimize nonlinear functions with complex parameter spaces for food.



From Vladimir.Marangozov at inrialpes.fr  Wed Jul 28 11:00:42 1999
From: Vladimir.Marangozov at inrialpes.fr (Vladimir Marangozov)
Date: Wed, 28 Jul 1999 16:00:42 +0100
Subject: beep
References: <379F0751.71214802@lri.fr>
Message-ID: <379F1B1A.2199B575@inrialpes.fr>

Perhaps printing the standard escape sequence '\a' (Alert) would do?

>>> print '\a'
<beep :->
>>>

Stephane Conversy wrote:
> 
> Hi all.
> 
> How to make the system beep with python ?
> I couldn't find anything about that in the faq and in the doc, and I'm
> almost
> sure I've seen something about that...
> 
>     stef
> 
> --
> St?phane Conversy
> http://www-ihm.lri.fr/~conversy/
> mailto:conversy at lri.fr

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov at inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252



From simngolda at voyager.net  Wed Jul 14 20:12:52 1999
From: simngolda at voyager.net (Sim & Golda Zacks)
Date: Wed, 14 Jul 1999 20:12:52 -0400
Subject: Terminating Python...
References: <BFC0587634C3D111AACB00805FA766FA02C1C652@exchny12.ny.smb.com> <378CFDDF.489CDCEF@compaq.com>
Message-ID: <7mj8s9$op2@newsops.execpc.com>

I use raise SystemExit which is documented in "Internet Programming with
Python"

Greg Ewing wrote:

> "Velasquez, Carlos" wrote:
> >
> > Hello,
> >
> > I just downloaded and built Python v1.51 for the Sun Solaris platform.
> > One question is, how do I exit python?
>
> Ctrl-D at the interpreter prompt should do the trick.
> Or if it's running code, Ctrl-C to get its attention
> and then Ctrl-D.
>
> Greg




From greg.ewing at compaq.com  Sun Jul  4 16:33:22 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Mon, 05 Jul 1999 08:33:22 +1200
Subject: Binary pickle portability
References: <Pine.OSF.4.05.9906291852100.1376-100000@pinax.Mathematik.Uni-Bielefeld.DE> <Pine.OSF.4.05.9906292018360.28230-100000@pinax.Mathematik.Uni-Bielefeld.DE>
Message-ID: <377FC512.604F8E34@compaq.com>

Olaf Delgado wrote:
> 
> After looking a bit closer, I found that the problem is probably related
> to the Numeric module (i.e., pickling matrices) and doesn't depend on
> whether binary or ordinary mode is used.

I think you may have been right the first time.
The Numeric module always seems to pickle its 
matrices in a binary format, whichever pickler
is used. I suspect it's just dumping whatever
is in memory, so as you say, it's non-portable.

If you want a portable format, you may have to
convert the Numeric matrices to standard Python
lists or tuples before pickling.

Greg



From conversy at lri.fr  Wed Jul 28 10:08:58 1999
From: conversy at lri.fr (Stephane Conversy)
Date: Wed, 28 Jul 1999 16:08:58 +0200
Subject: beep
References: <379F0751.71214802@lri.fr> <379F1B1A.2199B575@inrialpes.fr>
Message-ID: <379F0EF9.54CDDBEC@lri.fr>

Vladimir Marangozov wrote:

> Perhaps printing the standard escape sequence '\a' (Alert) would do?
>
>

is it portable ? (ie Mac OS)


--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr






From puetsch at gmx.de  Thu Jul  1 08:45:11 1999
From: puetsch at gmx.de (Felix Puetsch)
Date: 1 Jul 1999 12:45:11 GMT
Subject: Reflection in Python?
References: <7ldglr$ita$1@nnrp1.deja.com> <7ldm3s$led$1@nnrp1.deja.com>
Message-ID: <8DF696103424242@chief>

Why do you post it 3 times? Is it such important?

-- 
Felix Puetsch



From p.t at iname.com  Thu Jul 29 08:43:34 1999
From: p.t at iname.com (Peter Torstenson)
Date: Thu, 29 Jul 1999 09:43:34 -0300
Subject: Use of variables form outside the class.
Message-ID: <7npi9i$2j12@enews3.newsguy.com>

I declare a variable "x" right in the beginning of a module.
Later I declare a class with function defs.
I managed to declare varibles inside the class and address them by putting
"self." in front of the variable.
How do I reach the "x" variable that is outside any class from inside the
class functions?

Peter Thorstenson
The Newbie






From Haimo.Zobernig at cern.ch  Wed Jul 14 11:21:53 1999
From: Haimo.Zobernig at cern.ch (Haimo G. Zobernig)
Date: Wed, 14 Jul 1999 17:21:53 +0200
Subject: making a Windows NT service out of a python program
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F20@RED-MSG-50>
Message-ID: <378CAB10.C5E77604@cern.ch>

Bill Tutt wrote:
> 
> Take a look at the stuff in win32\demos\service\ of your friendly latest
> Win32 extension install tree.
> 
> It shows you how to wrap Python code into an NT service quite easily.
> 
> The code in win32\lib\win32serivceutil.py appears to do almost everything
> you would ever want to do with a service. Very cool stuff...
> 
> No, I didn't write it, it's all Mark's fault... :)
> 
> Bill

Hi,
I have been trying the pipeTestService.py in win32\demos\service, sending it
single messages with the pipeTestServiceClient.py script - that works fine.
But I get errors when trying out the multi-thread multi-message options
( -t <n> -m <k> ). I am running the script with ^R from within PythonWin.

pipeTestServiceClient.py -v 'test test test'

>>> Sending 'test test test'
Server sent back 'haimo (on thread 172) sent me 'test test test''


pipeTestServiceClient.py -t 10 -m 20 -v 'test test test'

>>> Sending 'test test test'
Server sent back 'haimo (on thread 132) sent me 'test test test''
Spawning 10 threads each sending 20 messages...
Traceback (innermost last):
  File "C:\Program Files\Python\win32\demos\service\pipeTestServiceClient.py",
line 68, in stressThread
    r = CallPipe(CallNamedPipe, ("\\\\%s\\pipe\\PyPipeTest" % server, "#" * 512,
1024, NMPWAIT_WAIT_FOREVER))
  File "C:\Program Files\Python\win32\demos\service\pipeTestServiceClient.py",
line 45, in CallPipe
    raise win32api.error, (rc, fnerr, msg)
api_error: (233, 'CallNamedPipe', 'No process is on the other end of the pipe.')
Failed after 0 messages
Traceback (innermost last):
  File "C:\Program Files\Python\win32\demos\service\pipeTestServiceClient.py",
line 68, in stressThread
    r = CallPipe(CallNamedPipe, ("\\\\%s\\pipe\\PyPipeTest" % server, "#" * 512,
1024, NMPWAIT_WAIT_FOREVER))
  File "C:\Program Files\Python\win32\demos\service\pipeTestServiceClient.py",
line 45, in CallPipe
    raise win32api.error, (rc, fnerr, msg)
api_error: (233, 'CallNamedPipe', 'No process is on the other end of the pipe.')
Failed after 2 messages
Traceback (innermost last):
  File "C:\Program Files\Python\win32\demos\service\pipeTestServiceClient.py",
line 68, in stressThread
    r = CallPipe(CallNamedPipe, ("\\\\%s\\pipe\\PyPipeTest" % server, "#" * 512,
1024, NMPWAIT_WAIT_FOREVER))
  File "C:\Program Files\Python\win32\demos\service\pipeTestServiceClient.py",
line 45, in CallPipe
    raise win32api.error, (rc, fnerr, msg)
api_error: (233, 'CallNamedPipe', 'No process is on the other end of the pipe.')
Failed after 4 messages

etc...

Am I doing something wrong, or does that really not work?
I'd like to be sure that the examples work before I adapt them to my own
programs.

-- 
Haimo G. Zobernig                                  tel   +41 22 7677361
Univ. of Wisconsin                                 fax   +41 22 7678370
c/o CERN Bld. 32 R-A02
1211 Geneve 23, Switzerland                email Haimo.Zobernig at cern.ch



From esr at snark.thyrsus.com  Thu Jul 29 16:11:09 1999
From: esr at snark.thyrsus.com (Eric S. Raymond)
Date: 29 Jul 1999 20:11:09 GMT
Subject: [Tutor] I want to learn to hack
References: <19990721154610.B3663@optiplex.palga.uucp> <Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org> <14231.11726.916541.492153@nobot.cnri.reston.va.us>
Message-ID: <slrn7q1dau.17t.esr@snark.thyrsus.com>

Roger Masse <rmasse at cnri.reston.va.us> wrote:
>Errr, it seems I've inadvertantly 'cracked' opened a can of worms.
>I really just wanted to know how the Python tutors deal with such
>requests, since they are arriving with alarming frequency to
>webmaster at python.org.

I fear this is my fault.  

Some time back, I published a page called "How To Become A Hacker"
<URL:http://earthspace.net/~esr/faqs/hacker-howto.html">.  As a result,
I get a couple dozen semi-illiterate email notes a week from teenagers
wanting me to teach them everything I know.  Instantly.

(I tolerate this because once a month or so I get a letter from a kid
who is well-spoken, intelligent, and thoughtful.  *Those* I help.)

Earlier this year I changed the document to recommend Python as a
first language.  I still think this was a good idea, but I see it has
had an unfortunate side-effect.
-- 
		<a href="http://earthspace.net/~esr">Eric S. Raymond</a>



From granadam at mail.datanet.hu  Sat Jul 10 18:11:04 1999
From: granadam at mail.datanet.hu (dzsi)
Date: Sun, 11 Jul 1999 00:11:04 +0200
Subject: database
Message-ID: <3787c551.1@ganymede.datanet.hu>

hi,

i am new to python.......

i have a database (comma separated text), and i would like to get some data
out of it based on some query. let's say there are seven fields, f1...f7,
and i am searching for entries in f1, and want to display f1..f7  for the
field that matches the query. how do i go about doing that?

also, can i access db databases, or any other common pc type databases with
perl scripts only on a unix system? without any server configuration,
assuming python is installed. so far, the comma separated text file is the
only way i could figure out.

thanks,

adam.







From larsga at ifi.uio.no  Thu Jul  1 10:17:02 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 01 Jul 1999 16:17:02 +0200
Subject: [Q] dbase-file/clipper-index read-module
References: <930819652.351193@services>
Message-ID: <wkaetg4fq9.fsf@ifi.uio.no>

* Timo Rainio
| 
| anyone around who knows about a Python-module for dbase-file (.dbf)
| access?  support for clipper-style NTX-index structures would be
| great. I only want to _READ_ clipper-files here.

I have a module that can handle .dbf and .fpt files at

<URL: http://www.stud.ifi.uio.no/~larsga/download/python/ >

It can't handle NTX files, since I didn't need that when I wrote
it. :)

--Lars M.



From dominic.binks at aethos.co.uk  Tue Jul 13 03:36:18 1999
From: dominic.binks at aethos.co.uk (Dominic Binks)
Date: Tue, 13 Jul 1999 08:36:18 +0100
Subject: Problem with informixdb and python1.5.2
References: <000016a3@bossar.com.pl>
Message-ID: <378AEC72.8D12D0A7@aethos.co.uk>

Ken Liu wrote:
> 

> Python 1.5.2 (#8, Jul  1 1999, 20:22:25)  [GCC 2.7.2.3] on linux2
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import informixdb
> >>> con = informixdb.informixdb('acme_online_demo')
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "informixdb.py", line 47, in __init__
>     self.conn = _informixdb.informixdb(logon)
> InformixdbError: Error -922 performing LOGON: Cannot get name of current
> working directory.
> >>>

Can informix read your current working directory?  I've certainly seen
this cause problems with dbaccess, so I don't see why the ESQL/C
libraries should be much different.

Dominic
-- 
Dominic Binks/Systems Engineer/Logica-Aldiscon
400 Park Avenue/Aztec West/Bristol/BS32 4TR/United Kingdom
Tel: +44 1454 614455/Fax: +44 1454 620527/Mobile: +44 498 693964
E-mail: dominic.binks at aethos.co.uk



From tjreedy at udel.edu  Thu Jul 15 18:33:10 1999
From: tjreedy at udel.edu (Terry Reedy)
Date: 15 Jul 1999 22:33:10 GMT
Subject: database
References: <3787c551.1@ganymede.datanet.hu>
Message-ID: <7mlnj6$rjl$1@news.udel.edu>

In article <3787c551.1 at ganymede.datanet.hu>, granadam at mail.datanet.hu says...

>i am new to python.......
>
>i have a database (comma separated text), and i would like to get some data
>out of it based on some query. let's say there are seven fields, f1...f7,
>and i am searching for entries in f1, and want to display f1..f7  for the
>field that matches the query. how do i go about doing that?

After reading in lines with readline() or readlines(), use 
string.splitfields(line,',') to split each line (this assumes that you do not 
have ,s embedded within quoted strings -- which is why people sometimes use 
something obscure like '|' as a field separator).  Then use == comparison or 
one of the string search functions to test f1.

Terry J. Reedy




From neelk at brick.cswv.com  Thu Jul  1 23:06:15 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 1 Jul 1999 22:06:15 -0500
Subject: total idiot question: +=, .=, etc...
References: <000d01bec11f$8f966e20$e19e2299@tim>
Message-ID: <7lhab7$2ha$1@brick.cswv.com>

In article <000d01bec11f$8f966e20$e19e2299 at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>[Neel Krishnaswami]
>
>> With the clear way of writing it, the NewArticle class would
>> mysteriously fail to do the right thing, because the __cmp__ method
>> would look in the Article class dictionary without bothering to check
>> NewArticle's.
>
>The trap is hard to avoid, though:  if a direct instance of Article is
>"self", and a direct instance of NewArticle is "other", even keying off
>self.__class__.etc is a surprise (well, at least to "other"!).  In part
>you're taking a single-dispatch approach to a problem with multiple-dispatch
>headaches.  

<light dawns> 

Of course! I had noticed I wasn't having this problem in Dylan, and 
it somehow failed to occur to me that the only significant difference
between Dylan and Python (Dylan's multiple-dispatch object model)
might have something to do with it. :)


Neel



From algaba at my-deja.com  Sat Jul 10 06:04:24 1999
From: algaba at my-deja.com (algaba at my-deja.com)
Date: Sat, 10 Jul 1999 10:04:24 GMT
Subject: Again, small utility docSor
Message-ID: <7m75r6$39o$1@nnrp1.deja.com>

Sorry if you get this twice !!

I've done this:

"""
1998 Manuel Gutierrez Algaba
You are free to use , modify, distribute and copy this piece of
python code, if you keep this free copyright notice in it.

arbor.py Version 1.0

It makes trees of anything you want to. Class depencencies, calls...

For more related doc py programs look at:
http:\\www.ctv.es\USERS\irmina\texpython.htm

or email me:
irmina at ctv.es

Something like this :
|||->a
|||  |||->1
|||  |||  |->e
|||  ||-->2
|||  |--->4
||-->b
|--->c

just describing it as:
    a.add(['a','b','c'],'letras', 'letras del abecedario')
    a.add([1,2,4], 'a' )
    a.add(['e'],1)

    a.dibuja('letras')

It's so extremely easy and small ( the python code )
that it doesn't deserve a translation...

TODO:
- Read from a properties file.

"""

import sys

class atree:
    def __init__(self, cabecera, hojas, diccionario, inf):
        self.cabecera = cabecera
        self.inf = inf
        self.hojas = hojas
        self.diccionario = diccionario
        self.anyadido = ""
        self.anyadido_hijo = ""

    def dibuja(self, anyadido=""):
        self.anyadido = anyadido
        if not self.inf is None:
            self.imprimeln( self.cabecera + "("+ self.inf +")")
        else:
            self.imprimeln( self.cabecera)
        k = 0
        cuantas = len(self.hojas)
        for i in self.hojas:
            self.dibuja_hoja(i, k, cuantas)
            k = k + 1

    def dibuja_hoja(self , hoja, pos, cuantas):
        self.anyadido_hijo = ""
        self.imprime( self.anyadido )
        composicion_vert = self.compone_vert(cuantas, pos)
        self.imprime( composicion_vert )
        composicion_vert = self.escribe_flecha(composicion_vert)
        if hoja in self.diccionario.keys():
            self.diccionario[hoja].dibuja(self.anyadido +
self.anyadido_hijo)
        else:
            self.imprimeln( hoja)

    def compone_vert(self, cuantas, pos):
        a = ""
        for i in xrange(0, cuantas):
            if (i < cuantas - pos) :
                b = "|"
                c = "|"
            else:
                b = " "
                c = "-"
            a = a + c
            self.anyadido_hijo = self.anyadido_hijo + b

        return a

    def escribe_flecha(self, anyadido):
        self.imprime( "->")
        self.anyadido_hijo = self.anyadido_hijo + "  "
        return anyadido + "  "


    def imprime(self, s):
        sys.stdout.write(str(s))

    def imprimeln(self, s):
        sys.stdout.write(str(s)+"\n")

class rep:
    def __init__(self):
        self.diccionario  = {}
        #self.anyade()

    def add(self, lista_hojas, cabecera, comentario = None ):
        self.diccionario[cabecera] = atree(cabecera,
                                           lista_hojas,
                                           self.diccionario,
                                           comentario)

    def dibuja(self, l):
        self.diccionario[l].dibuja()

if __name__=='__main__':
    a = rep()
    a.add(['a','b','c'],'letras', 'letras del abecedario')
    a.add([1,2,4], 'a' )
    a.add(['e'],1)

    a.dibuja('letras')

Sorry again, if you get this twice.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From rceci at adams.com  Sun Jul  4 13:44:47 1999
From: rceci at adams.com (rceci at adams.com)
Date: Sun, 04 Jul 1999 18:44:47 +0100
Subject: Moving the python distribution
Message-ID: <000017c8@bossar.com.pl>

From: rceci at adams.com

Hello,

I have a question about relocating a python distribution (1.5.1) that
was built and installed in the normal area ( i.e. prefix=/usr/local ).

I have moved the bin/*, include/python1.5/* and lib/python1.5/* to a NFS
mount that will support multiple architectures, e.g.:

bin/python -> $mount_pnt/python/bin/$ARCH/python1.5/python
include/python1.5/* -> $mount_pnt/python/include/$ARCH/python1.5/*
lib/python1.5/* ->  $mount_pnt/python/lib/$ARCH/python1.5/*

I have set up PYTHONHOME and PYTHONPATH to support this.  When I run
python on the build machine, everything works as planned.  When I tag
the "-v" on to python, I can see the expected messages that show the
built-in modules being loaded from my NFS "$mount_pnt" rather that the
build directory "/usr/local".  Thats great.  But when I log onto another
system of identical architecture ( in this case both machines are IRIX
6.4 MIPS4 systems), I get:

Could not find platform independent libraries <prefix>
Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]

And yet the "-v" still prints it is loading from the $mount_pnt.  This
second system does not have python installed in "/usr/local..." at all.
I am guessing that specifiying PYTHONHOME/PYTHONPATH works, but the
iterpreter will still print warnings if it cannot find python runtimes
under the original exec prefix location.  Is there a way to suppress
these warnings or fix them in the source?

Thanks

Rob Cecil
rceci at adams.com


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From da at ski.org  Wed Jul 14 12:11:23 1999
From: da at ski.org (David Ascher)
Date: Wed, 14 Jul 1999 09:11:23 -0700 (Pacific Daylight Time)
Subject: Pickling C objects
In-Reply-To: <378B27A9.4201@cs.bham.ac.uk>
Message-ID: <Pine.WNT.4.04.9907140909240.212-100000@rigoletto.ski.org>

On Tue, 13 Jul 1999, Stuart Reynolds wrote:

> I'm writing some new Python classes with a C implementation. How do you
> make a C object pickleable?

To my knowledge, the easiest way is to make them be ExtensionClass
objects and use the __getstate__/__setstate__ methods.

See http://www.digicool.com/releases/ExtensionClass

The latest version is not on the web, but it's available through the Zope
read-only CVS directory: http://www.zope.org/Community/CVS_public_access

--david




From mstenber at cc.Helsinki.FI  Tue Jul 27 00:13:30 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 27 Jul 1999 07:13:30 +0300
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <7nhseu$gf0$1@nnrp1.deja.com>
Message-ID: <al8d7xen311.fsf@sirppi.helsinki.fi>

aaron_watters at my-deja.com writes:
> In article <al81zdxkktf.fsf at sirppi.helsinki.fi>,
>   Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
> > Even in websites' case, I would worry; Python doesn't use multiCPU
> very
> > well, and with single CPU, it is fairly easy to saturate one with
> > pure-Python code, especially if it is badly designed to boot.
> Bad design aside.  If you have many CPU's for a web site you
> might consider using plain old CGI scripts, since python would then
> use all the cpu's.  In a single (multithreaded) process python will
> only use one (thanks to the global interpreter lock).  Also, I
> strongly feel CGI is more stable and reliable than other methods.

True, but overhead of one Python instance starting makes 2 CPU box fairly
prohibitive for CGI use, and thus requires 4+ to make it even give _some_
gains compared to constantly-running interpreter.

> > Using a Gadfly database, _with_ pre-compiled queries: (very simple
> > one-liners from table with one entry)
> > select: 21.4182/sec [3.64s] (46.689ms/call)
> > update: 14.3209/sec [3.98s] (69.828ms/call)
> Note please that gadfly queries have a certain amount of fixed overhead.
> You *might* find that queries over a table with one entry are not
> much faster than queries over tables with 1000 entries, or maybe even
> tables with 10000 entries that have appropriate indices.

True, but same numbers for non-Python-implemented databases are, even with
Python interfaces, 100+x times as fast. 

> However each gadfly query requires 10's of python function
> calls regardless of how simple the query is.  But more complex queries
> over more data don't necessarily require many more python function
> calls (this was one of the design goals for the engine).

Nod.. I didn't mean to particularly _blame_ Gadfly, just point out that
implementing some things in Python seems to be prohibitvely slow due to the
'speed'. 

> Also, I hope you have added the kjbuckets builtin...

Yes.

> On second thought 21 queries per second ain't too bad...
> What are you comparing it to?  Updates are naturally slower
> since they involve writes to a log file for recovery purposes.

Single select/update's on Solid or MySQL with Python interface do easily
1000/second on trivial tables. (on most boxen)

> > .. and so on. Even more distressing results lie in the speed of say,
> > implementing protocols in Python (when we're talking _nearly_ realtime
> > action).
> Be careful here.  Protocol speed can be effected by a lot of stuff.
> Some of it can be python and some of it can be other system overhead.
> If a profiler or unix "top" says it's the python process and it's
> implemented well, then you've got a point.  But until I see the
> code I personally would withhold assignment of blame.

Well, I might share the code to get some insights, but according to 'top'
Python process is hogging the CPU, _and_ primarily thanks to number of
method calls in the OO-based protocol model with two layers (TLS-lookalike
protocol with 100 percent Python implementation).

I've speeded it up by factor of N by just removing the layers and merging
all functions without more than 5 lines to their calling functions
directly, but result is ugly-looking (albeit faster) protocol with some
~100+ liner functions :P

>    -- Aaron Watters

-Markus Stenberg

-- 
"...Deep Hack Mode--that mysterious and frightening state of
consciousness where Mortal Users fear to tread."
	- Matt Welsh



From pvelikho at cs.ucsd.edu  Fri Jul  9 11:10:50 1999
From: pvelikho at cs.ucsd.edu (Pavel Velikhov)
Date: Fri, 09 Jul 1999 08:10:50 -0700
Subject: PyApache and Apache 1.3.6 anyone ???
References: <19990708221333.B16851@hoffleit.org> <3785836D.3BE3EBC7@cs.ucsd.edu> <19990709112714.A2615@hoffleit.org>
Message-ID: <378610FA.2CFF1E54@cs.ucsd.edu>


Gregor Hoffleit wrote:
> 
> On Thu, Jul 08, 1999 at 10:06:53PM -0700, Pavel Velikho wrote:
> > Did you check the error log? If it is a configuration error it will
> > usually keep quiet about
> > it and just write it in the log file.
> > Hope this helps
> 
> Nope, there's no entry at all in the error log for all the silently
> dying starting attempts. I don't know how to debug apache with gdb,
> therefore I'm comletely lost where to look for hints about the
> failure.
> 
>         Gregor

Wow, never ran into *that* bad of a stuff. All that comes to mind is to
try to recompile apache again making sure all the libraries that python
uses are linked with, and maybe put in the debug flags, etc...

Sorry can't help much
Pavel



From neelk at brick.cswv.com  Thu Jul 22 21:46:26 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 22 Jul 1999 20:46:26 -0500
Subject: Stackless & String-processing
References: <000701bed335$8a812740$a72d2399@tim> <7n4nrl$sfe$1@cronkite.cc.uga.edu> <3795F4E4.8330755F@appliedbiometrics.com>
Message-ID: <7n8hhi$sp$1@brick.cswv.com>

In article <3795F4E4.8330755F at appliedbiometrics.com>,
Christian Tismer  <tismer at appliedbiometrics.com> wrote:
>
>I believe this thread would not have evolved if I had not
>begun the Stackless Python / continuation/coroutine/generator
>talk. Now I'm asking to excuse my limited understanding.
>
>Can you please be all a little more concrete?
>How would your patterns look in Python?
>Please show some small example code pieces.

I am working through the Hutton/Meijer paper, and (slowly!) 
implementing the parsers they describe. Basically, the important part
isn't the "monadic" part, but the "combinator" part. The idea is that
you can/should write primitive parsers, and then write higher-order
functions to weave those little parsers into a bigger parser. It's a
nice way to write recursive descent parsers without sinking into
spaghetti code, basically.

(This is kind of hard to do right in Python, because the lack of
lexical scoping and easy-to-use metaclasses makes it trickier to mimic
combinators they describe exactly. But it's still a fun project.)

One of the consequences of this approach is that outputting all the
parses for an ambiguous grammar as a stream of parse trees is very
straightforward. In Haskell or Gofer, you could rely on the laziness
of the language to avoid unnecessary computations, but in Python
it would be not-fun to wait for Python to compute the whole list. 

Here's a (simple, rough-draft) combinator I've written:

class Bind:
    def __init__(self, parser1, func):
        self.parser1 = parser1
        self.func = func # func(v :: AST) -> Parser; e.g. Result
    def __call__(self, input):
        parse_list = self.parser1(input)
        lst = []
        for ast, s in parse_list:
            foo = self.func(ast)(s)
            for elt in foo:
                lst.append( elt ) # <-- a generator would help a lot here!
        return lst

I would love to rewrite the __call__ method so that it returns a
generator that returns parses one at a time rather than returning a
list. Note that the combinator approach magnifies the benefits of
generators, since parsers are normally built through composition, and
each sub-parser can potentially generate a sequence of parses.
Laziness is a real win here. 

Also, it might make sense to add a xreadlines() method to file
objects, which returns the lines of a file one at a time, because then
you can write code in the natural style of:

for line in file.xreadlines():
    blah_blah_blah(line)

without having to read the whole file into memory. (It would
also likely eliminate about 80% of the push for assignment
expressions.)


Neel



From bwarsaw at cnri.reston.va.us  Wed Jul 21 10:56:35 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Wed, 21 Jul 1999 10:56:35 -0400 (EDT)
Subject: Python Mode for Alpha ?
References: <lawter-2107990058210001@client-151-197-119-26.bellatlantic.net>
Message-ID: <14229.57251.996871.502133@anthem.cnri.reston.va.us>

>>>>> "JL" == John Lawter <lawter at cse.buffalo.edu> writes:

    JL> Is there a more recent Python mode for Alpha than the 6.5 era
    JL> one ?  I have searched around the python.org site, but their
    JL> copy is missing.  Any help would be appreciated, as I don't
    JL> look forward to having to write my own.

Are you asking about python-mode.el for Emacs?  If so, the one at
http://www.python.org/emacs/python-mode/ will work for any modern
version of Emacs or XEmacs.  If not, I'm not sure I can help ya.

-Barry



From tim_one at email.msn.com  Sat Jul 24 00:23:32 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 24 Jul 1999 00:23:32 -0400
Subject: int(long(-sys.maxint-1)) fails on Linux
In-Reply-To: <37992994.F4A6A30@vision.arc.nasa.gov>
Message-ID: <000f01bed58c$4abb9e20$832d2399@tim>

[Chad Netzer]
> ...
> Yeah, I noticed that the test seemed to be missing, and that lea
> probably shouldn't set any condition codes (and as you point out,
> it would be very bad if it did).  I am curious as to how the x<<1
> is computed be lea (I don't understand the
> 0x0(,%edx,2),%eax notation...

Well, the notation is awful, but about the best you can expect from a
brainless disassembler (the CISC Intel instructions have gazillions (give or
take a bundle) of variations, and a disassembler generally pumps out the
first template that matches the bits it's looking at -- regardless of
whether a *human* would write it that way).

Anyway, the lea is simply using a scaled-index addressing mode, with a 0
base, a 0 offset, and with %edx treated as a "scale by 2" (shift left by 1)
scaled index.

Check out this link:

http://www.comp.it.bton.ac.uk/burks/burks/language/asm/artofasm/artof001.htm

Arguably the best book on assembler programming ever written!  The chapters
on x86 addressing modes and LEA will chase the dark clouds away.

to-be-replaced-by-deep-depression-but-them's-the-breaks-ly y'rs  - tim





From MHammond at skippinet.com.au  Sat Jul 10 04:54:12 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Sat, 10 Jul 1999 18:54:12 +1000
Subject: idle
References: <1tBh3.875$uZ.2396@news.flash.net>
Message-ID: <7m71kf$lp0$1@m2.c2.telstra-mm.net.au>

I believe the lastest version of IDLE can only be found via CVS -
http://www.python.org/download/cvs.html

Mark.

Tim Hamza wrote in message <1tBh3.875$uZ.2396 at news.flash.net>...
>Have there been any updates to idle.  I thought I saw something about a 0.5
>version, but I haven't been able to find anything at python.org.







From Nikolai.Kirsebom at p98.f112.n480.z2.fidonet.org  Thu Jul  1 07:18:23 1999
From: Nikolai.Kirsebom at p98.f112.n480.z2.fidonet.org (Nikolai Kirsebom)
Date: Thu, 01 Jul 1999 12:18:23 +0100
Subject: Python, ASP - Namespace (again)
Message-ID: <00001604@bossar.com.pl>

From: nikolai.kirsebom at siemens.no (Nikolai Kirsebom)

On Wed, 30 Jun 1999 09:00:20 +1000, "Mark Hammond"
<MHammond at skippinet.com.au> wrote:

>There is no obvious reason to me why this wont work.  Note that if you are
>not running win32all-125, then I recommend you upgrade - there have been
>some fixes (but none that adddress this, as far as I can tell).
>
>The best I can suggest is that you use the win32trace module and print
>statements to work out what the hell is going on.
>
>Eg, making your first line "import win32traceutil" will mean that all print
>statement go to a debugging collector.  You can view the print statements by
>starting Pythonwin on the same machine, and selecting "Tools/Trace Collector
>Debugging Output".
>
>Then use lots of print statements - specifically print dir() etc, to see
>exactly what the namespace is...
>
>Mark.
>

Mark, thanks for hint on using the trace utility.

The next time I checked the code written in my first message, it
worked ok !!!!

As I indicated in the message, I have one ASP file and two PY files.
Updating the ASP file (saving it) followed by a refresh in the browser
results in updated browser page content.  However, updating any of the
PY files does not result in any changes in the browser output.  Could
anyone answer when the PY modules are loaded.  I thought the PY
modules automatically are reloaded if the datestamp of the *.py file
is newer than the *.pyc file.

I'm using the statement:

from <module> import *

How (if possible) would I reload explicitly such a module ?


Nikolai



From nascheme at ucalgary.ca  Sat Jul 17 18:19:21 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 17 Jul 1999 22:19:21 GMT
Subject: Python and Boehm-Demers GC, I have code.
References: <000401bed023$d08d3360$06a22299@tim> <7mq2ol$hk8@ds2.acs.ucalgary.ca> <3790A898.DA7D61AF@appliedbiometrics.com>
Message-ID: <7mqvh9$ese@ds2.acs.ucalgary.ca>

On Sat, Jul 17, 1999 at 06:00:24PM +0200, Christian Tismer wrote:
> Neil Schemenauer wrote:
> ...
> > With some tuning maybe the gc version of Python will perform
> > better than the regular version.
> 
> I believe you will need to keep the RC as well, since at
> the moment, I think this gives more locality than GC,
> as Tim mentioned already.

It is my intention to leave the RCing alone.  My previous
ambition to remove it was misguided.  By tuning I meant doing
things like reducing how often the collector gets called.  Since
RC should take care of almost everything the collector should
very infrequently.  Then again, premature optimization is the
root of all programming evil.  I will concentrate on making
things work right now.

> If you can manage to make it rock solid, I would perhaps
> like to bundle it with stackless Python. Although, the
> collector which Guido proposed a while ago could perform
> even better, so I'm still a bit undecided.

I don't know about rock solid but it is running pretty well for
me now.  I fixed the problem with the tkinter callbacks and I can
now run Idle and Grail with no problems.  All regression tests
pass.  I will try to beat on it for a few days and see what what
breaks.  I haven't fixed all the extension modules yet.  There is
trouble if PyMem_* and malloc and friends are mixed.  I suspect
that many extension modules do this (including some of mine).  So
far I had to fix readline, regex, re, and some of the Python
internals.  I have sent Guido a patch.

Are you familiar with the (theoretically) copying garbage
collector for Python?  I can't find the name of the guy who wrote
it.  I think it would be interesting to integrate this collector
with your stackless Python.  That would remove the need for
keeping the separate GC stack (I think).  pybench shows Python
with this collector to be 68% slower than Python 1.5.1.  It
should be faster without the GC stack but it will probably still
be slower than regular Python.

I have a new version of the patch available at:


	http://www.acs.ucalgary.ca/~nascheme/python/gc.html


I will probably compile some Windows binaries (for people on
compiler challanged systems).  Hopefully people can try it out
and give me some feedback.


    Neil



From Per.Kistler at p98.f112.n480.z2.fidonet.org  Sun Jul  4 10:34:21 1999
From: Per.Kistler at p98.f112.n480.z2.fidonet.org (Per Kistler)
Date: Sun, 04 Jul 1999 15:34:21 +0100
Subject: Tkinter._test() problem
Message-ID: <000016be@bossar.com.pl>

From: Per Kistler <kistler at fnmail.com>

Hi All

I installed py1.5.2 with tcl/tk.8.1 on SuSE6.1 Linux.

After:

import Tkinter
Tkinter._test()

I get the following error, which I do not unerstand:

-----cut
>>python
Python 1.5.2 (#4, Jul  4 1999, 12:25:56)  [GCC egcs-2.91.66 19990314
(egcs-1.1.2  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import Tkinter
>>> Tkinter._test()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/per/lib/python1.5/lib-tk/Tkinter.py", line 1947, in _test
    root = Tk()
  File "/per/lib/python1.5/lib-tk/Tkinter.py", line 886, in __init__
    self.tk = _tkinter.create(screenName, baseName, className)
TclError: Can't find a usable init.tcl in the following directories:

This probably means that Tcl wasn't installed properly.
----cut

It seems, that python looks in some directories, which it does not
have. What do I have to do?

Thanks, Per.
-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net
------------------------------------------------------------



From andy at robanal.demon.co.uk  Mon Jul 26 15:16:52 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Mon, 26 Jul 1999 19:16:52 GMT
Subject: UK Python job opportunity (City)
Message-ID: <379fb3ff.1113270@news.demon.co.uk>

I know of a major financial firm looking for a Python developer in the
UK at the moment.  Sadly I'm doing two jobs already and can't take on
a third, even if it is 100% Python.

Anyone interested, send me an email saying briefly (who you are, what
you do, and what your availability is, and I'll put you in touch with
the right people.

Regards,

Andy Robinson





From fredrik at pythonware.com  Fri Jul  2 04:02:06 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 2 Jul 1999 08:02:06 GMT
Subject: embedding; return array
References: <377BAE23.4634E0A3@ncsa.uiuc.edu>
Message-ID: <00bc01bec461$2e725830$f29b12c2@secret.pythonware.com>

Randy Heiland wrote:
> I'd welcome any pointers into being able to return an (unsigned char)
> array from an embedded module.

how about:

    return Py_BuildValue("s#", buffer, buffer_size);

(gives you a string; you can use the array module to
process it further).

if you really want to return it as a python list, you
have to build the list yourself.  something like:

    PyObject* list;

    list = PyList_New(buffer_size);
    if (list)
        for (i = 0; i < buffer_size; i++) {
            PyObject* item;
            item = PyInt_FromLong(buffer[i]);
            if (item == NULL) {
                Py_DECREF(list);
                list = NULL;
                break;
            }
            PyList_SetItem(list, i, item);
        }
    }

    return list;

should do the trick.

</F>




From larsga at ifi.uio.no  Mon Jul 26 10:35:21 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 26 Jul 1999 16:35:21 +0200
Subject: How to use a variable form main script in a module.
References: <7nhmt5$su0@enews3.newsguy.com>
Message-ID: <wkg12bscly.fsf@ifi.uio.no>

* Peter Torstenson
|
| This is probaly a silly question from a newbie.
| In my main script I put a value for to a variable called "x".
| Now I want to use the "x" variable in a module.
| The module is imported into the main script. How to do this?

>From the way you describe it this sounds like something you shouldn't
do. Communication between different pieces of code through globabl
variables is in general frowned upon, and with good reason. It's good
for some things, but for those the variables tend to reside in a
module, and not in the global namespace.

Feel free to explain the reason why you want this if you want a
verdict on whether it's a good or bad idea.

[...wrong info in previous version of this post removed (global
doesn't work)...]

--Lars M.



From tgabriel at bellsouth.net  Fri Jul  9 11:44:49 1999
From: tgabriel at bellsouth.net (Xenophanes)
Date: Fri, 09 Jul 1999 11:44:49 -0400
Subject: Decent Tutorial
References: <000015af@bossar.com.pl>
Message-ID: <7m9tmg$6qg$28@rks1.urz.tu-dresden.de>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19990709/50b2693c/attachment.html>

From eugene.leitl at lrz.uni-muenchen.de  Wed Jul 21 14:00:47 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Wed, 21 Jul 1999 11:00:47 -0700 (PDT)
Subject: python vs. lisp (whereis quote?)
In-Reply-To: <lbzogh635zl.fsf@morgoth.uio.no>
References: <x88r9m3cyje.fsf@rpppc2.hns.com>
	<lbzk8rva3to.fsf@morgoth.uio.no>
	<x88zp0rbd1x.fsf@rpppc2.hns.com>
	<mtlncbhv94.fsf@astron.berkeley.edu>
	<lbzogh635zl.fsf@morgoth.uio.no>
Message-ID: <14230.2767.18575.943818@lrz.de>

Rolf Marvin B?e Lindgren writes:

 > some might try to do something on top of Guile, which I hope they won't.

Why? Because it's Scheme, or because it's Guile?



From victor at idaccr.org  Fri Jul  9 16:12:31 1999
From: victor at idaccr.org (Victor S. Miller)
Date: 09 Jul 1999 16:12:31 -0400
Subject: _tkinter problem
Message-ID: <ulpv21popc.fsf@runner.princeton.idaccr.org>

I had successfully built Python 1.5.2 with _tkinter using the system
installed versions of tk and tcl (4.1 and 7.5 respectively).  However,
I decided to download the new versions of tcl and tk (8.1 for both),
and build them.  They both built without problems, and ran through the
test sucessfully.  I then rebuilt Python with them (I'm running on
Solaris 2.6), which appeared to build successfully, however, when I
try Tkinter._test(), I get:

Python 1.5.2 (#5, Jul  9 1999, 15:36:18)  [GCC 2.7.2.2] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import _tkinter
>>> import Tkinter
>>> Tkinter._test()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/local/coop/lib/python1.5/lib-tk/Tkinter.py", line 1947, in _test
    root = Tk()
  File "/usr/local/coop/lib/python1.5/lib-tk/Tkinter.py", line 886, in __init__
    self.tk = _tkinter.create(screenName, baseName, className)
TclError: Can't find a usable init.tcl in the following directories: 
    



This probably means that Tcl wasn't installed properly.

>>> 


Now there's a perfectly good init.tcl in the directory
/usr/local/coop/lib/tcl8.1, which is where make install placed it.
Who is actually complaining about this, and how can I fix it?
-- 
Victor S. Miller     | " ... Meanwhile, those of us who can compute can hardly
victor at idaccr.org    | be expected to keep writing papers saying 'I can do the
CCR, Princeton, NJ   | following useless calculation in 2 seconds', and indeed
    08540 USA        | what editor would publish them?"  -- Oliver Atkin



From David.Ascher at p98.f112.n480.z2.fidonet.org  Fri Jul  2 01:30:55 1999
From: David.Ascher at p98.f112.n480.z2.fidonet.org (David Ascher)
Date: Fri, 02 Jul 1999 06:30:55 +0100
Subject: Debugging embedded pgm under WinNT
Message-ID: <00001690@bossar.com.pl>

From: David Ascher <da at ski.org>

On Fri, 2 Jul 1999, Gordon McMillan wrote:

> > WINNT/system32 right next to python15.dll, ran my program . . . and
> > now it couldn't find "multiarray" (my embedded Python code uses the
> > Numeric extensions).  I verified that sys.path was correct and
> > included the Numeric directories. Do I have to build the Numeric
> > extensions under Debug mode? 

FWIW, I have a version of compile.py (the thing which builds Numeric)
which, when debugged, will allow building multiple versions automatically,
both Release and Debug (thanks to MHammond).

--david





From effbot at hotmail.com  Fri Jul 30 13:55:02 1999
From: effbot at hotmail.com (F L)
Date: Fri, 30 Jul 1999 17:55:02 GMT
Subject: 1.5.2 for: else:
Message-ID: <19990730175502.89698.qmail@hotmail.com>

Skip wrote:
>Now we have to deal with the fredbot as well as the timbot.
>Anybody for a quick game of RoboWar?

note: the correct name is "effbot".

see
http://www.pythonware.com/people/fredrik/status.htm
for more info.

also see:
http://www.robocup.org

and:
http://www.aftonbladet.se/nyheter/9907/30/robocup.html
(in swedish, but you should be able to find the real-
player video links anyway)

and yes, we do apologise for that exception.  usually,
such messages should be caught before they end up on
the list.

Cheers,
the eff-bot administrators


_______________________________________________________________
Get Free Email and Do More On The Web. Visit http://www.msn.com



From kas at maps.magnetic-ink.dk  Fri Jul 16 11:43:21 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Fri, 16 Jul 1999 17:43:21 +0200
Subject: Snippet: bitcount()
Message-ID: <slrn7oukop.68v.kas@titan.magnetic-ink.dk>

  G'day mates.

Here's a snippet for the repository:

# The function bitcount() will count the number
# of bits in a number, iterating at most as many
# times as the number of bits.
#
def bitcount(x):
    n = 0
    while x:
        x = x & (x-1)
        n = n + 1
    # end while
    return n
# end def bitcount


Cheers,

  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From Richard.Jones at fulcrum.com.au  Tue Jul 27 02:53:19 1999
From: Richard.Jones at fulcrum.com.au (Richard Jones)
Date: Tue, 27 Jul 1999 06:53:19 GMT
Subject: Qt 
In-Reply-To: Message from Kikutani Makoto of 1999-Jul-27 5:32:48,
            <7njga0$361$1@ash.prod.itd.earthlink.net> 
References: <7njga0$361$1@ash.prod.itd.earthlink.net>
Message-ID: <199907270653.QAA01576@envy.fulcrum.com.au>

[Kikutani Makoto]
> There are kdebinding and qtbinding in:
>     ftp://ftp.kde.org/pub/kde/devel/
> 
> I guess qtbinding is obsolete and kdebinding should be used.
> Am I right ?
> I'm afraid that kdebinding is not updated recently, either.

   There's a fairly complete Qt/KDE binding maintained at:

      http://www.river-bank.demon.co.uk/software/

   The KDE developer site seems broken again (another facelift was necessary I 
suppose ;) and there doesn't seem to be any mention of this module on the Qt 
site. The author is working on Qt 2.0 support as well as splitting the KDE and 
Qt components into seperate modules. A friend of mine has recently made the 
module thread-safe and has submitted patches to the author of pykde.

   I'm using it and it's pretty darned good. I just want it to support 2.0 
right now :)


      Richard





From gmcm at hypernet.com  Wed Jul 14 09:40:41 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 14 Jul 1999 08:40:41 -0500
Subject: [Python-Dev] Interrupting a thread
Message-ID: <1280166671-10546045@hypernet.com>

[Mark wants to interrupt threads]

[Tim arrives with a bucket of cold water]
> > I don't think there's any portable way to do this.  Even
> > restricting the
> > scope to Windows, forget Python for a moment:  can you do
> > this reliably with
> > NT threads from C, availing yourself of every trick in the
> > SDK?  Not that I

[Mark is stubborn...] 
> Nope - not if I forget Python.  However, when I restrict myself _to_
> Python, I find this nice little ceval.c loop and nice little
> per-thread structures - even with nice-looking exception
> place-holders ;-)

[The iceman cometh...]
If you'd like your app to remain usable, you'd have to be certain that
this thread held no locks of any kind. If it did, it is likely that
said lock could never be released even if you found it!

If the thread is doing something, there are certainly places where you
can insert a "time to quit?" check. If it's hopelessly blocked, you
can ignore it or even cut it off from interacting with the rest of the
app. It'll consume some memory but no CPU and die when the OS cleans
up at the end of the process.

Note that the latter case (of hopelessly blocked) wouldn't be helped
by evil manipulations of his Python state.


- Gordon



From gmcm at hypernet.com  Thu Jul  1 00:14:42 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 30 Jun 1999 23:14:42 -0500
Subject: Mailbox, tell and read
In-Reply-To: <377AAE02.E8591727@prescod.net>
Message-ID: <1281323312-36476914@hypernet.com>

Paul Prescod writes:

> The mailbox message objects simulate a file interface by seeking
> through a mailbox. A read() returns the entire message. It does this
> by seeking to the start, reading to the message end, and returning
> the data. The problem is that message offsets are derived by
> file.tell() which I presume is in terms of bytes. But then it does a
> read() based on the offsets and read() is in terms of characters. 
> 
> These don't match up on Windows because of CR/LF pairs. I think that
> the right fix is to never use tell() and instead keep track of the
> location by counting characters. Does that make sense? -- 

How 'bout opening in binary, and munging line endings yourself just 
before returning the message? If you can't use tell(), you can't 
trust seek(), either.

- Gordon



From clarkeia at logica.com  Fri Jul 30 08:53:54 1999
From: clarkeia at logica.com (Ian Clarke)
Date: Fri, 30 Jul 1999 13:53:54 +0100
Subject: Python (Win) Database Advice needed
Message-ID: <7ns78n$mnu@romeo.logica.co.uk>

Hi, I am planning to write a python based web-database retrieval system.  I
am not very familiar with databases in general (although I have used SQL)
and was looking for some advice.  Basically this system will run on an NT
server and allow selected users to view and change data over the web.  I
need a mechanism for storing the data so that:
1) It can be accessed by several HTTP threads running at the same time
without risk of corruption
2) It can be migrated to other database formats such as Microsoft Access.
3) It is Open Source
4) It is simple to deploy

Thanks for any advice,

Ian.





From fredrik at pythonware.com  Fri Jul 30 10:19:22 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 30 Jul 1999 14:19:22 GMT
Subject: syntax
References: <37A1C94E.F54181DB@buf.fr>
Message-ID: <002801beda96$89097300$f29b12c2@secret.pythonware.com>

Xavier Bec <xav at buf.fr> wrote:
> Why are the double dots ":" necessary after a if, for or while
> instruction ?
> All the others instructions are terminated with a cariage return but not
> condition or loop ones. I know that a semi-column is facultative and
> only used if you have to put a second instruction on the same line. Why
> is not the same for if, etc.
> 
> if i == 10
> print i
> print "test"
> 
> or
> 
> if i == 10 : print i; print "test"
> 
> Any suggestion ?

readability.  think about it for a while, and then read:
http://www.deja.com/=dnc/getdoc.xp?AN=171314015

</F>




From coryalbrecht at hotmail.com  Fri Jul  2 15:23:58 1999
From: coryalbrecht at hotmail.com (Cory C. Albrecht)
Date: Fri, 02 Jul 1999 19:23:58 GMT
Subject: Compiling Python for Win32
References: <7lgjij$180bi_004@news.sentex.ca> <ZPQe3.4335$U6.23535@newsr2.twcny.rr.com>
Message-ID: <7lj3ke$1cm0s4_002@news.sentex.ca>

In article <ZPQe3.4335$U6.23535 at newsr2.twcny.rr.com>, "Darrell" <news at dorb.com> wrote:
>> What I would like is if somebody who has successfully built Python 1.5.2
>> using MSVC++ could do the "Export Makefile" command and send me the
..
>> will earn somebody my undying thanks. :-)

>Ok. I'm waiting for the undying thanks part.

And here you go. *undying thanks* :-)

>Hope this works for you.

Me too. Thanks alot, I appreciate it.

--
Cory C. Albrecht
http://www.sentex.ca/~cory/



From arcege at shore.net  Sun Jul 25 18:17:10 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Sun, 25 Jul 1999 22:17:10 GMT
Subject: Access dictionary in order?
References: <roy-2507991300370001@mc-as01-p61.med.nyu.edu> <14235.18248.924848.52652@dolphin.mojam.com>
Message-ID: <G1Mm3.1243$ui4.388782@news.shore.net>

Skip Montanaro <skip at mojam.com> wrote:

:     Roy> I've got a dictionary of items which I will sometimes want random access
:     Roy> to, but sometimes I'll want to be able to step through the items in a
:     Roy> fixed order as well.  Any good way to do that?

: You demonstrated the best way.  There is no need to save the key list unless
: the dictionary is very large though.  Just grab 'em and sort 'em when you
: need:

:     keylist = dict.keys()
:     keylist.sort()
:     for key in keylist: do_stuff(dict[key])

There are times when the entries cannot be sorted and the order must
still be preserved.

A good way to handle this is a modified UserDict subclass.

  from UserDict import UserDict
  class OrderedDict(UserDict):
    def __init__(self, dict=None):
      UserDict.__init__(self, dict)
      if dict and isinstance(dict, OrderedDict)
        self.ordered = dict.ordered[:]
      elif dict:
        self.ordered = dict.values()  # must be taken unordered
      else:
        self.ordered = []
    def __getitem__(self, index):
      if isinstance(index, type(0)):
        return self.ordered[index]
      else:
        return UserDict.__getitem__(self, index)
    def __setitem__(self, index):
      UserDict.__setitem__(self, index, value)
      self.ordered.append(value)

One problem to this mechanism is that numbers are not indices instead
of keys.  But you can do this in other ways to fill the application's
needs.

Remember that you aren't creating copies of the objects, just references
to the same object, so you should be creating only twice the memory
space as the dictionary itself (relatively speaking).

  -Arcege




From viofis at my-deja.com  Thu Jul  1 20:33:35 1999
From: viofis at my-deja.com (viofis at my-deja.com)
Date: Fri, 02 Jul 1999 00:33:35 GMT
Subject: please help with simple regexp
Message-ID: <7lh1d0$t67$1@nnrp1.deja.com>

I'm a beginner at Python, so this may be a stupid
question:

I have a string, for example, myString="the
Potato".
1. Convert it to all lowercase ("the potato")
2. Match it against a list of strings -
"potato,tomato,corn,etc...", extract "potato" (if
it matches with something on the list), then save
into myNewString.

Any help is greatly appreciated.

thanks,
Vadim



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From bmcd at es.co.nz  Sun Jul 11 00:46:49 1999
From: bmcd at es.co.nz (Ben Caradoc-Davies)
Date: 11 Jul 1999 04:46:49 GMT
Subject: Dreams.
References: <slrn7ofkrt.55r.sdossey@dolphin.openprojects.net>
Message-ID: <7m9toe$6qg$89@rks1.urz.tu-dresden.de>

Hmmm. I'm not sure what your background is, Scott, but you may not be aware of
some previous work --- the term you may be looking for is "continuation". The
use of continuations is the explicit aim of a number paradigms, particularly in
the functional language world and "artificial intelligence". Continuations can
be used to implement "backtracking", where, for example, a program explores
one branch of execution, reaches a dead end, and restores it's previous state
by falling back on a saved continuation.

A continuation can be described as a function representing the work yet to be
done.

Have you read "The Structure and Interpretation of Computer Programs" by
Abelson and Sussman? A real eye-opener, especially for someone with a physics
Fortran/C/Matlab background. Changed my world-view. I have heard this book
often recommended as one of the best. It is (used to be?) required reading for
MIT comp. sci. and e.e. undergraduates. The second edition came out recently.
The book uses the Scheme language
  http://www.swiss.ai.mit.edu/scheme-home.html

Check out Ken Dickey's intro to scheme
  ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/intro.txt
which has some examples of how to rewrite a function in a continuation passing
style.

Also look in on comp.lang.scheme, comp.lang.lisp and comp.lang.functional. Read
the functional programming faq.

One of the basic ideas behind functional (a.k.a. applicative) programming is
avoiding the use of assignment to get around the problems of saving state
which, as you have correctly deduced, causes a lot of problems. There has been
a large amount of work done on these problems since the seventies. Look for
resources on "closures," "continuations," "functional programming," and
"backtracking."

ObPython:
The funny thing is, coming from Scheme, I found that Guido had put the best of
Scheme into Python. Python's major advantages are clear syntax and semantics, a
well-defined foreign function interface, and (consequently) a large library and
numerous extensions.

"Alms for a poor *ex*-schemer!"
"Did you say *ex*-schemer?"
"Yes. One minute I was a schemer with a livelihood. Next, along comes Guido.
  *Cures* me! No so much as a by-your-leave!"

-- 
Ben Caradoc-Davies <bmcd at es.co.nz>



From thamza.nospam at geocities.com  Wed Jul 14 02:37:25 1999
From: thamza.nospam at geocities.com (Tim Hamza)
Date: Wed, 14 Jul 1999 06:37:25 GMT
Subject: togl widget?
Message-ID: <FeWi3.824$Dk6.3405@news.flash.net>

I am trying to track down the togl widget.  The link from the Starship
OpenGL page is dead.





From heiland at ncsa.uiuc.edu  Wed Jul  7 09:57:08 1999
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Wed, 07 Jul 1999 08:57:08 -0500
Subject: Py_BuildValue, PyArg_ParseTuple
Message-ID: <37835CB3.2B876BE3@ncsa.uiuc.edu>

Here's a very simple program where I want to get back out
(PyArg_ParseTuple) what I put in (Py_BuildValue) - however, I can't seem
to do that.  Can someone PLEASE enlighten me?

thanks,
--Randy

/* Py_BuildValue, PyArg_ParseTuple  usage */

#include "Python.h"

main(argc, argv)
    int argc;
    char **argv;
{
  unsigned char ba[3] = {2,4,8};  /* byte array */
  PyObject *pobj;
  char *cstr;
  int len;

  Py_Initialize();

  printf("Py_BuildValue of %d %d %d\n",ba[0],ba[1],ba[2]);
  pobj = Py_BuildValue("s#", ba,3);
  if (pobj == (PyObject *)NULL)
     printf("pobj is NULL\n");

  printf("\nPyArg_ParseTuple(pobj,...)\n" );
  PyArg_ParseTuple(pobj,"s#", &cstr,&len);
  printf("len=%d\n",len);
  printf("cstr=%s\n",cstr);

  printf("cstr[0]=%c\n",cstr[0]);
  printf("cstr[1]=%c\n",cstr[1]);
  printf("cstr[2]=%c\n",cstr[2]);

  printf("(int)cstr[0]=%d\n",(int)cstr[0]);
  printf("(int)cstr[1]=%d\n",(int)cstr[1]);
  printf("(int)cstr[2]=%d\n",(int)cstr[2]);
}





From chris at frostnet.advicom.net  Tue Jul 20 22:19:46 1999
From: chris at frostnet.advicom.net (Chris Frost)
Date: Wed, 21 Jul 1999 02:19:46 GMT
Subject: Merging two lists as a dictionary
Message-ID: <19990720211946.A12202@pooh.frostnet.net>

I have two lists and would like to merge them into a dictionary, with one
list forming the keywords and the other the values in the dictionary. How could
this be done? "map(None, keyword_list, value_list)" looked great, but it
returns a list of pairs (I'd like to be able to specify values according to
keywords...).

thanks for any pointers!
-- 
Chris Frost  |  <http://www.frostnet.advicom.net/chris/>
-------------+------------------------------------------
Public PGP Key:
   Email chris at frostnet.advicom.net with the subject "retrieve pgp key"
   or visit <http://www.frostnet.advicom.net/chris/about/pgp_key.phtml>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 256 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19990721/a669fc0e/attachment.sig>

From ivanlan at callware.com  Thu Jul 15 23:27:06 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 16 Jul 1999 03:27:06 GMT
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net> <FExMpA.Juv@midway.uchicago.edu>
Message-ID: <378EA68A.F2D8A7E@callware.com>

Hi All--

"Paul M." wrote (to mary jeanette):
> 

[lots of other good advice snipped]

> 
> If you're flustered by anything even remotely resembling a command
> window I believe (but am not sure) that the latest installer even puts
> a shortcut to IDLE (a nice tk driven interpretter/programming
> environment) in your Python folder on your Start>Programs>Python
> window.
> 

While the installer indeed puts a shortcut into your Python menu folder,
it fails to put one on your desktop.  It helps to have a shortcut on
your desktop.  Here's how, mary jeanette:

(I quote from _Teach Yourself Python in 24 Hours_ (forthcoming from
SAMS)):

``The first thing you need to do is to find where IDLE lives, and to do
that you need to know where Python lives; this should be simple, since
the Python directory should be in your path. On all my Windows systems,
Python is in "C:\Python", and you should open a Windows Explorer Window
in the Python directory. Within the Python directory, you should find
several folders, one of which is named "Tools"; enter that directory,
and that should contain another folder called "IDLE". Inside the IDLE
folder you should see an icon for "idle.py", and right next to it should
be "idle.pyw", which is the one we're looking for. Right click on the
idle.pyw icon, and select "Create Shortcut" on the popup menu; Windows
should create a whole new icon for you, entitled "Shortcut to idle.pyw".
Grab that icon and drop it on your desktop; you can change the name if
you wish, and you can even change the icon (although I happen to like
that cute little green python) to whatever you want. Once you're 
satisfied with the desktop location and appearance of your IDLE
shortcut, double-click on it to start it running.''

> It might help to know at least something about the OS you're using
> (e.g., to set your path to include the Python install directory so you
> can access the interpretter from any command line) - even if you do
> find it appalling.  However, the standard install makes it incredibly
> easy, even for the clueless.
> 

The installer does modify your autoexec.bat to include the Python path
on W95/98.  I seem to recall having to modify my path environment
variable on Windows NT, however, and it should.

> If all this is STILL too obscure, then yes, perhaps you should learn
> Perl, which I hear is known for its extreme readability and ease of
> use ;)
> 

As Christian Tismer says, "flea dirt."  And a more apt description of
Perl I have yet to encounter.

<unless-it-had-to-do-with-yellow-snow>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From donn at u.washington.edu  Mon Jul 19 13:18:27 1999
From: donn at u.washington.edu (Donn Cave)
Date: 19 Jul 1999 17:18:27 GMT
Subject: Killing a thread
References: <7mu6a5$h61$1@nnrp1.deja.com>
Message-ID: <7mvml3$u5k$1@nntp6.u.washington.edu>

catlee at my-deja.com writes:
| I'm writing a progrma where some user code is executed in a restricted
| environment running in a separate thread.  Now, if the user is
| malicious, s/he can just put 'while 1: 1+1'.  I'd like to be able to
| recover from this.  Is there any way I can kill the thread that is stuck
| in the infinite loop?

Python's thread support is evolving.  As a previous followup points out,
you shouldn't kill threads, and I guess that's fair enough - in terms
of normal program design, one should look for other ways to manage
threads.

But in default of any alternative, I suspect sometimes we have to violate
this commandment, and I think one of these days Python will provide the
tools to do what you have to do, not just what you want to do.  At the
moment though, it looks to me like you have to go outside Python's thread
support.  For example on BeOS (happens to be the only OS where I've played
with threads), I find I can kill a thread with os.kill(bad_thread_id, 21)
(SIGKILLTHR == 21)  The thread doesn't clean up normally, but it stops.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From Greg.McFarlane at p98.f112.n480.z2.fidonet.org  Sun Jul  4 08:49:12 1999
From: Greg.McFarlane at p98.f112.n480.z2.fidonet.org (Greg McFarlane)
Date: Sun, 04 Jul 1999 13:49:12 +0100
Subject: Upgrade or not ??
Message-ID: <000016c2@bossar.com.pl>

From: Greg McFarlane <gregm at iname.com>

Depending on exactly what you want to do, it may be possible in
Tkinter.  You can use something like this to find out when a widget
gets and loses focus:

    import Tkinter

    def gotFocus(event):
	print 'Focus in'
    def lostFocus(event):
	print 'Focus out'

    root = Tkinter.Tk()

    e = Tkinter.Entry()
    e.bind('<FocusIn>', gotFocus)
    e.bind('<FocusOut>', lostFocus)
    e.pack()

    root.mainloop()

Also, you can use this to set the focus to a widget.

    e.focus_set()

On 3 Jul, pierrette wrote:
> First a general question :
> Actually we're working with Python 1.5.1 and we're satisfied with it.
> Should we upgrade to 1.5.2 and if yes WHY?
> 
> No a Tkinter guestion :
> We developped a prototype of an application with python an tkinter, now
> we have to
> decide if we're gonna write the real app in python or something else !
> We now allready that in principle we could do it with python, the thing
> we're missing
> in tk is a 'on get focus' ,a 'on lost focus' event and a 'set focus to'
> command
> Pe: on lost focus of inputfield 1:
>             if inputfield1 = A:
>                 set focus to field2
>             elif inputfield1 = B:
>                 set focus to field 3
> 
> If those things are not available in tkinter what alternatives do we
> have ?
> - Changing the GUI (would be a shame but feasable), knowing that is has
> to be multi-platform WIN-LINUX? If so to what ?
> - Not write in python ? In this case we lose the work done for the
> prototype
> 
> Thanks for comments and remarks
> 
> Pierrette Serrand
> pserrand at worldnet.fr
> 
> 
> 
> 
> 

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at nms.otc.telstra.com.au)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From robin at jessikat.demon.co.uk  Sun Jul  4 18:56:03 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sun, 4 Jul 1999 23:56:03 +0100
Subject: Why have 300 articles just appeared in c.l.py?
References: <m33dz4oxxx.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <rDY7zDADa+f3Ew5l@jessikat.demon.co.uk>

In article <m33dz4oxxx.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson
<mwh21 at cam.ac.uk> writes
>I just had the slightly alarming experience of about 300 articles, all
>of which I had read, appearing unthreaded in this newsgroup.
>
>Did this happen to everyone, or is it some local snafu?
>
>TIA
>Michael
yes I just saw the same seems like a server bombed its date somewhere.
-- 
Robin Becker



From gmcm at hypernet.com  Fri Jul 30 11:56:57 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 30 Jul 1999 10:56:57 -0500
Subject: syntax
In-Reply-To: <37A1C94E.F54181DB@buf.fr>
Message-ID: <1278776075-25895447@hypernet.com>

Xavier Bec writes:
> 
> Why are the double dots ":" necessary after a if, for or while
> instruction ?
> All the others instructions are terminated with a cariage return but
> not condition or loop ones. I know that a semi-column is facultative
> and only used if you have to put a second instruction on the same
> line. Why is not the same for if, etc.

>From the standpoint of parsing, the colon is redundant to the 
following INDENT. As I understand the history, their usage is 
mandatory because of user testing done on ABC, which was one of a 
host of influences on the design of Python.

For some reason, the dominant hominid specie of planet Earth is 
uncomfortable without a certain level of redundancy. Some members 
thereof have taken this predeliction to ridiculous extremes, creating 
languages such as "Java".

Nanu-Nanu-ly y'rs

- Gordon



From stidolph at origin.ea.com  Fri Jul 23 01:31:22 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Fri, 23 Jul 1999 05:31:22 GMT
Subject: how do I find memory leaks?
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC26@molach.origin.ea.com>

The following is our Debug.py file.  We have a Globals.py file that has all
of our global variables, include one called 'debug' that is a class instance
of the Debug class.  Note that all base classes should have the following
code in them:

import Globals #for Globals.debug

class Name:

  def __del__(self):
    if __debug__: #we do this to toss in non-debug situations
      Globals.debug.LogDel(self)

  def __init__(self):
    if __debug__:
      Globals.debug.LogInit(self)

  def Destroy(self):
    if __debug__:
      Globals.debug.LogDestroy(self)
	#Eliminate pointers to other classes and notify any other classes
that point to it that
	#you are going away and need to have references eliminated.


Please notice that all base classes have these routines defined.  Classes
derived from them do NOT have to have this code - in fact, it will cause
warning messages.  As LogInit is called, we keep the stack trace for its
creation.  When LogDestroy is called it is noted, and when LogDel is called
the entry is removed.  I set up additional variables to control reporting in
debug mode.  The idea of this module is to track class instances and issue a
report at shutdown as to what (if anything) is left undeleted, or didn't
have the Destroy method called - or had the Destroy method called more than
once.  Output is to the standard output, which we have routed to
OutputDebugString so it goes to the MSDev Output debug window.  Any
questions?  Suggestions??

Thanks,

David Stidolph

P.S.  On our first use of this we found hundreds of memory leaks.  We solved
most within four hours.

####  And now for the Debug Module

#Debug Module
#
#This class is designed to track Python objects between their __init__() and
Destroy() method calls.
#There are different 'levels' to the tracking.  If we do it at all, we will
track the objects by the object
#id() using a dictionary.  If we have 'holdTraceback' active, then in
LogInit we get the current
#stack, translate it into a list of strings and use that list as the value
and the object reference as
#the key in the dictionary.
#
#As a dictionary we keep only one copy and don't allow for
LogInit/LogDestroy to be called more than
#once (have setting to throw exception if it occurs).  This means you put
the calls to LogInit and
#LogDestroy in the base class .. not the derived classes.  Proper class name
of the object is used
#by __class__.__name__.
#
#TODO:
#       Fix output for MSVC IDE to use absolute paths to Python files for
double-clicking
#       (Note that this works now for Windows 95, but not 98)
#       Add Python object for output in LogReport to allow output to file or
sending across net
#       Add print formatted output to wrap output (allow to send across
net).
#       Optimize use of stack trace to reduce memory requirements.

from sys import exc_info
from traceback import extract_stack

class Debug:
  """Only one instance allowed"""

  __refDebugCount = 0 # class variable

  def __del__(self):
    Debug.__refDebugCount = Debug.__refDebugCount - 1

  def __init__(self, trackObjects = 1, holdTraceback = 1,
raiseExceptionOnDuplicate = 0, displayInit = 0, displayDestroy = 0):
    if Debug.__refDebugCount != 0:
      raise "Only one copy of debug allowed"

    if __debug__:
      self.trackObjects = trackObjects
      self.holdTraceback = holdTraceback
      self.raiseExceptionOnDuplicate = raiseExceptionOnDuplicate
      self.displayInit = displayInit
      self.displayDestroy = displayDestroy
    else:
      self.trackObjects = 0               #if we don't have debug mode, then
it doesn't matter what
      self.holdTraceback = 0              #settings we use since nothing
SHOULD be calling us.
      self.raiseExceptionOnDuplicate = 0  #Just in case we ignore and set
for nothing.
      self.displayInit = 0
      self.displayDestroy = 0

    self.activeObjects = {}     #clear dictionary
    Debug.__refDebugCount = 1    #we are initialized

  def Destroy(self):
    Debug.__refDebugCount = 0
    self.activeObjects = None

  def LogDel(self,object):
    #We wrap the destruction of the key inside a try..except block in case
the LogDestroy is called without
    #calling the LogInit (rare) or LogDestroy is being called for a second
time (far more likely).
    try:
      objectID = id(object)
      value = self.activeObjects[objectID]
      if 0 == value[1]:
        if self.raiseExceptionOnDuplicate:
          raise "MEM: LogDel called on "+object.__class__.__name__+" without
calling Destroy()"
        else:
          print "MEM: LogDel called on "+object.__class__.__name__+" without
calling Destroy()"
          if self.holdTraceback:
            tb = value[2]
            for tuple in tb[:-2]: #-2 is to eliminate report of call to
LogInit()
              print str(tuple[0])+'('+str(tuple[1])+'): in
'+str(tuple[2])+','
              print '  '+str(tuple[3])
          print
      del self.activeObjects[objectID]
    except:
      if self.raiseExceptionOnDuplicate:
        raise "MEM: LogDel called twice on "+object.__class__.__name__+".
See if base class doing LogInit/LogDestroy/LogDel (only one should)"
      else:
        print
        print "MEM: LogDel called twice on "+object.__class__.__name__+".
See if base class doing LogInit/LogDestroy/LogDel (only one should)"
        print

  #Called from each base object Destroy() method.
  def LogDestroy(self,object):
    if self.trackObjects:
      if self.displayDestroy:
        print 'MEM: ',repr(object),'of class',object.__class__.__name__,'is
being Destroyed'
      #We wrap the destruction of the key inside a try..except block in case
the LogDestroy is called without
      #calling the LogInit (rare) or LogDestroy is being called for a second
time (far more likely).
      self.activeObjects[id(object)][1] = 1

  #Called from base object __init__(self) routines
  def LogInit(self,object):
    if self.trackObjects:
      if self.displayInit:
        #display that the object has been constructed and its __init__ has
been called
        print 'MEM: ',repr(object),'of class',object.__class__.__name__,'is
being initialized'

      #traceback of stack is generated using extract_stack() and used as the
value in the
      #dictionary with the object reference used as the key.
      if self.holdTraceback:
        #We check for the object already being in the dictionary ONLY if we
will throw an 
        #exception, silent otherwise
        if self.raiseExceptionOnDuplicate:
          if self.activeObjects.haskey(object):
            raise "Object has already called LogInit"
        tb = extract_stack()
        self.activeObjects[id(object)] = [object.__class__.__name__,0,tb]
      else:
        self.activeObjects[id(object)] = [object.__class__.__name__,0,1]

  #Called at program end to report on undestroyed Python objects.
  def LogReport(self):
    # For formatting that Visual Studio can recognize, we're adjusting
    # output to look like:
    # c:\test\gagob.cpp(236) : error C2065: 'Clear' : undeclared identifier
    if self.trackObjects:
      if len(self.activeObjects) > 0:
        print 
        print 'MEM: UnDestroyed Python Objects Report:'
        print '     %s python objects total.' % len(self.activeObjects)
        print
        print 'Summary:  These objects were not released.  Individual stack
traces follow'
        for objectID,value in self.activeObjects.items():
          print ' class ' + value[0]
        print
        print
        #Get Object reference and stack traceback for each undestroyed
Python object
        for objectID,value in self.activeObjects.items():
          className = value[0]
          destroyCalled = value[1]
          tb = value[2]
          if destroyCalled == 1:
            print 'MEM:  class ' + className + ' Destroy() called, but not
__del__.'
          else:
            print 'MEM:  class ' + className + ' Destroy() and __del__
methods not called.'
          if self.holdTraceback:
            for tuple in tb[:-2]: #-2 is to eliminate report of call to
LogInit()
              print str(tuple[0])+'('+str(tuple[1])+'): in
'+str(tuple[2])+','
              print '  '+str(tuple[3])
          print #blank line between
      else:
        print 'MEM: Congratulations!  No Python object leaks!'


def PrintStack():
  tb = extract_stack()
  for tuple in tb[:-2]: #-2 is to eliminate report of call to PrintStack()
    print str(tuple[0])+'('+str(tuple[1])+'): in '+str(tuple[2])+','
    print '  '+str(tuple[3])
    









-----Original Message-----
From: Matt Gushee [mailto:mgushee at havenrock.com]
Sent: Thursday, July 22, 1999 7:54 PM
To: python-list at cwi.nl
Subject: Re: how do I find memory leaks?


"Stidolph, David" <stidolph at origin.ea.com> writes:

> I posted a module called Debug to track down memory leaks.  It requires
> specific code be added to your __init__ and __del__ routines, as well as
> having a Destroy method we use to eliminate circular references.  If you
> want it I will post it again.

Please do!

Matt Gushee
Portland, Maine, USA



From greg.ewing at compaq.com  Wed Jul 21 23:02:25 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 22 Jul 1999 15:02:25 +1200
Subject: IDLE's PyShell
References: <000801bed335$8c233e80$a72d2399@tim>
Message-ID: <379689C1.B91E0D13@compaq.com>

Tim Peters wrote:
> 
> just-stop-opening-and-closing-the-shell<wink>-ly y'rs  - tim

Easier said than done when the "flatten" and "close"
gadgets are so close together up there in the corner
of the window. It's very easy to hit the wrong one
in the heat of battle.

By the way, is there any way to create a window that
doesn't have, for example, a close gadget? I think
I've seen windows like that in other Windows apps,
but I can't find any way of doing it in Tkinter.
I'd really like to expunge the close gadgets from
my modal dialogues -- they're so utterly useless
there...

Greg



From stephensong at iname.com  Sun Jul  4 08:11:55 1999
From: stephensong at iname.com (Gary Stephenson)
Date: Sun, 04 Jul 1999 22:11:55 +1000
Subject: Tuples cf. Expanded objects??
References: <377bbde1@news.xtml.co.uk> <7lh375$5ea$1@news.udel.edu> <377CAC46.508D502F@iname.com> <slrn7nprts.3a4.wtanksle@dolphin.openprojects.net>
Message-ID: <377F4F8B.10C82257@iname.com>


William Tanksley wrote:

> On Fri, 02 Jul 1999 22:10:46 +1000, Gary Stephenson wrote:
>
> An instance of an immutable type may be created in expanded form.
>

Yes, but can they also  be created in an non-expanded form?  If not, then
there is more of a relationship that your message implies - yes?

cheers,

gary




From Emile.van.Sebille at p98.f112.n480.z2.fidonet.org  Thu Jul  1 21:02:15 1999
From: Emile.van.Sebille at p98.f112.n480.z2.fidonet.org (Emile van Sebille)
Date: Fri, 02 Jul 1999 02:02:15 +0100
Subject: Alternatives to pickle/class nesting
Message-ID: <00001672@bossar.com.pl>

From: "Emile van Sebille" <emile at fenx.com>

Well, I've spent the last couple of hours trying to piece a small sample
together that illustrates my problem.  Good news... No success!  ;-)
This probably means I've got some squirrelly code in the application
somewhere, or that I can restructure towards my sample and get things
going.

Comments-still-welcome-but-I'm-out-of-my-pickle-ly yr's

--

Emile van Sebille
emile at fenx.com
-------------------




From dwolff567 at my-deja.com  Wed Jul 28 16:48:24 1999
From: dwolff567 at my-deja.com (dwolff567 at my-deja.com)
Date: Wed, 28 Jul 1999 20:48:24 GMT
Subject: [Tkinter] Inheriting from Toplevel
References: <7nl2h8$keo$1@nnrp1.deja.com> <001d01bed86f$39de44f0$f29b12c2@secret.pythonware.com>
Message-ID: <7nnqam$ect$1@nnrp1.deja.com>

I use grid only in dialog boxes which inherit from the dialog class
you posted on the pythonware site.

  Dave

In article <001d01bed86f$39de44f0$f29b12c2 at secret.pythonware.com>,
  "Fredrik Lundh" <fredrik at pythonware.com> wrote:
> dwolff567 at my-deja.com wrote:
> > I am having some fairly weird problems with my Tkinter app.  I am
> > spawning a new Toplevel window where I display a graph in a
> > canvas instance.  However, when I send the data to the widget to
> > be plotted, the whole thing freezes.
>
> I'd look for a widget that is created without a master, and
> displayed using a call to "grid"...
>
> if you don't use grid at all, you probably have to post more
> code...
>
> </F>
>
>


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From tim_one at email.msn.com  Tue Jul 20 21:08:24 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 20 Jul 1999 21:08:24 -0400
Subject: Cyclops 0.9.4
In-Reply-To: <3794F3A2.DF8BE8FA@compaq.com>
Message-ID: <001501bed315$8888e740$d0a22299@tim>

[Greg Ewing]
> Seems to me that Cyclops (or any other such thing
> implemented in pure Python) is incapable of finding
> what you *really* want to find -- which is groups
> of objects not reachable from anywhere.

1. That can't be done in C today either (Python has no internal way to find
all allocated objects, and there's no protocol to implement pointer-chasing
either).

2. Unless you're fighting a bug in the interpreter's reference-counting code
(very rare), those groups of unreachable objects necessarily contain cycles.
That's why Cyclops focuses on cycles.

3. They're really not what you're looking for either, in my experience.  All
forms of GC are mechanical approximations to what you really want:  that an
object go away when it *won't* be referenced again, not when it *can't* be
referenced again.  The latter is a conservative approximation to the former.

Most leaks I've seen are due to objects that are reachable, but where the
author didn't *expect* them to reachable.  Cyclops can help with that too,
by calling get_rootset() at the end and seeing which registered objects
still have a non-0 (adjusted) refcount that you were expecting to be
history.

That's the mode it's being used in with IDLE now:  things *are* still
reachable that nobody expects "should be" reachable.  What Cyclops can't do
(yet <wink>) is tell you from *where* they can be reached.  All it can say
for now is that they don't appear to be in cycles -- they're still reachable
"for some other" reason.

you-wanna-know-different-things-at-different-stages-ly y'rs  - tim





From jefftc at leland.Stanford.EDU  Tue Jul 20 02:21:15 1999
From: jefftc at leland.Stanford.EDU (Jeffrey Chang)
Date: Mon, 19 Jul 1999 23:21:15 -0700
Subject: Scientific Plotting?
In-Reply-To: <379392E7.EB0F29E7@ssec.wisc.edu>
References: <379392E7.EB0F29E7@ssec.wisc.edu>
Message-ID: <Pine.GSO.3.96.990719225603.8888A-100000@saga1.Stanford.EDU>

On Mon, 19 Jul 1999, Nick Bower wrote:
> People at my current work use MatLab or IDL for chomping data through
> data calculations and plotting the results in an environment that lets
> you get GUI's up and going very quickly (among other things).
> 
[...]
> But...What do people use for plotting?

I'm interested in this as well.  I looked at the different options a few
months ago and found Matlab still to be the most robust (didn't look at
IDL) for plotting scientific data.

Fortunately, Python interfaces very cleanly with Matlab.  With Andrew
Sterian's nice PyMat module
(http://www-personal.umich.edu/~asterian/Python), you can perform Matlab
operations on NumPy arrays and vice versa almost transparently.  The
biggest advantage of this is that you'd get easy access to Matlab
functions, like plotting, as well as the general scripting capabilities of
Python.

So, to answer your question, I'm using Matlab from python.

Jeff




From lvirden at cas.org  Fri Jul  2 07:56:10 1999
From: lvirden at cas.org (lvirden at cas.org)
Date: Fri, 02 Jul 1999 12:56:10 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001665@bossar.com.pl>

From: lvirden at cas.org


According to Mikael Lyngvig <mikael at pobox.com>:
:Well, TkInter is extremely slow, but not because of the Python part.
:Even the Tk widget tour, included with Tk, is so slow that you almost
:think "Please, draw the next character - you can do it." whenever a
:window with some text needs to be redrawn.

When I run the Tk test suite  on my SPARC, Tk cranks thru creating and
deleting windows as fast as the eye can track.

However, there are reports that the Windows version of Tk, because of its
use of the native window widgets, is noticably slower than the X
implementation.  Is that the platform to which you refer?

-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.



From gmcm at hypernet.com  Tue Jul 20 09:11:52 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 20 Jul 1999 08:11:52 -0500
Subject: [Q] __methods__ and __members__ special functions...
In-Reply-To: <7n1n7c$7ao$1@feed.teaser.fr>
Message-ID: <1279650000-9962641@hypernet.com>

Olivier Deckmyn wrote:
> 
> I want to know all the methods available to a instance.
...
> if I make the test with [].__methods__ it works as expected.
> But when I use my own object 
...
> this raises a NameError exception :(
> 
> Why ? Where is __methods__ declared ?
> Why does this work for any instance in the brower.py provided with
> pythonwin ?

__methods__ is an attribute only of types (objects implemented in C); 
essentially it's the table of names -> function pointers that the C 
code uses to tell Python about it's capabilities.  For instance 
objects, methods are function objects living in the object's 
__class__.__dict__ (or perhaps in any __dict__ all along the search 
path from instance to __class__ to __bases__ etc). Mark's browser 
(and dir) will try various things until the object responds.

As you poke deeper, you'll find there's a lot in Python that works 
this way. For example, the code "if x" involves a list of things that 
Python will try on "x" to determine whether x considers itself true 
or false.

- Gordon



From Neel.Krishnaswami at p98.f112.n480.z2.fidonet.org  Thu Jul  1 12:17:48 1999
From: Neel.Krishnaswami at p98.f112.n480.z2.fidonet.org (Neel Krishnaswami)
Date: Thu, 01 Jul 1999 17:17:48 +0100
Subject: [Q] Static type checker for Python
Message-ID: <0000164c@bossar.com.pl>

From: neelk at brick.cswv.com (Neel Krishnaswami)

In article <377ADAD7.EA105C50 at mojam.com>,
Skip Montanaro  <skip at mojam.com> wrote:
>> Can I make static type checker for Python?
>> or Is there any previous work about this?
>
>In theory, you could do some type inference and generate some warning
>messages, but since you don't declare variables to be of a particular
>type in Python, it's unlikely it would be very useful.  The same
>variable can point to many different types at runtime.

I'm about to go off on a tangent, so first I want to warn the original
poster that there aren't too many actual problems with dynamic typing
-- people have written huge and complex apps with no problems using
Smalltalk and Lisp, for example. 

Now, it is possible to teach Python to check the argument types of
functions at runtime, and I've written a little module called
TypeChecker to do just that. It's a wrapper class with callable
instances, and the instance checks the declaration you make in a
docstring against the actual types and classes of the arguments. 
For example:

def fact(n):
    """foo(n :: LongType) <-- here's the type declaration; it can be
			      either a type from the types module or a
			      class name.
    A factorial that only takes longs as arguments."""
    f = 1L
    for i in range(1L, n+1):
        f = f * i
    return f

f = TypeChecker(fact) 

>>> fact(3)
6L

>>> f(3L)
6L

>>>f(3)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/tmp/python-6-237", line 25, in __call__
  File "/usr/tmp/python-6-237", line 118, in check_args
TypeError: Argument 1 of fact wrong type, <type 'int'>, declared <type 'long
int'>

What it's doing is parsing the type declaration in the doc string to
generate a list of types and classes, and then the __call__ method of
the instance is comparing the declared types against the runtime
arguments it receives before running the function. I've always wanted
to abuse docstrings, and now I've gotten my chance. :) Though this is
a fairly mild abuse, considering that it doesn't block you from
putting other things into the docstring.

Obviously, this is not terribly useful. (It doesn't handle subclasses
correctly, f'ex, and can't yet be used on class methods.) But it is
interesting to see what Python static types might look like. (The
answer to my eyes is "fairly ugly;" real lexical scoping and
type/class unification seem to me to be prerequisites before static
typing will behave gracefully in Python.)

The actual code for TypeChecker is a little too long to put in the
post, so here's a URL:

    http://www.sff.net/people/neelk/free-software/TypeChecker.py

Neel



From hei at adtranzsig.de  Wed Jul 21 02:14:14 1999
From: hei at adtranzsig.de (Dirk-Ulrich Heise)
Date: Wed, 21 Jul 1999 08:14:14 +0200
Subject: Tcl/Python: Who was first?
Message-ID: <932552685.620225@news.adtranzsig.de>

I just saw a message called

README: Tcl-URL! - the weekly guide to Tcl resources

on comp.lang.tcl.announce. (Reading it for the first time
in my life)
Did Python steel this (good) idea from Tcl or was it the
other way round?


Dirk





From godzilla at netmeg.net  Fri Jul 23 16:13:36 1999
From: godzilla at netmeg.net (Les Schaffer)
Date: Fri, 23 Jul 1999 20:13:36 GMT
Subject: Python documentation available as ZIP files
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us> <14232.48512.648609.496768@lrz.de> <14232.48816.993092.9754@weyr.cnri.reston.va.us>
Message-ID: <m2ogh39l9w.fsf@netmeg.net>

>>>>> ">" == Fred L Drake <fdrake at cnri.reston.va.us> writes:

    >>   That would be my preference actually, but it just isn't
    >> common enough yet.


class IMHO  ( $0.02 ):
  
    """of course, one can ENCOURAGE "common enough" by supplying bzip2
    for something popular like documentation and watch your community
    GROW acceptance of the new compression standard"""

    smile = "  :-)  "

    def __init__(self, agree_wth_me):

       if agree_with_me == YEP:
          os.system('tar cvf dem_docs dem_docs.tar.bz2 --use_compress-program=bzip2')

	print self.__doc__

	print smile

        print   "Les Schaffer godzilla at netmeg.net"

	raise " WARNING: it must be friday afternoon!!!  "




From greg.ewing at compaq.com  Wed Jul 28 16:16:00 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 29 Jul 1999 08:16:00 +1200
Subject: Stackless & String-processing
Message-ID: <379F6500.F3230208@compaq.com>

Christian Tismer wrote:
> 
> def backtrack():
>     a = next_alternative(list)
>     b = next_alternative(otherlist)
>     c = next_alternative(listwhatever)
>     if specialcondition(a, b, c):
>         return 42
>     fail()

This might not do what you seem to think
it will do under Icon semantics. For each value
generated for a, you will get all the values of
b, etc., so you'll get the cross product of
the three lists. If you were expecting a parallel
iteration, that's not what you'll get.

This illustrates why I don't like this sort of
implicit generator syntax -- often it's not obvious
what it's going to do. Switching languages for a
moment, rewriting the above in Scheme it comes
out as:

(define (backtrack cont)
  (for-each-alternative list
    (lambda (a)
      (for-each-alternative otherlist
        (lambda (b)
          (for-each-alternative listwhatever
            (lambda (c)
              (if (specialcondition a b c)
                (cont 42)
                '()))))))))

which makes it perfectly clear that you have
three nested loops.

Currently there is no elegant way to write that
sort of thing in Python. Either you have to split
the thunks off into separate procedures, or put up
with a seriously crippled lambda syntax. Either
way, you get namespace problems. That's why I
suggested an extension, which would allow you to
write

def backtrack(cont):
  for_each_alternative(list) -> (a):
    for_each_alternative(otherlist) -> (b):
      for_each_alternative(listwhatever) -> (c):
        if specialcondition(a, b, c):
          cont(42)

which is actually one line shorter than the
generator version, and (to my way of thinking)
the control flow is clearer.

Greg



From nathan at islanddata.com  Mon Jul  5 21:12:59 1999
From: nathan at islanddata.com (Nathan Clegg)
Date: Mon, 05 Jul 1999 18:12:59 -0700 (PDT)
Subject: optimization
Message-ID: <XFMail.990705181259.nathan@islanddata.com>

I'm writing an application in python that receives data in a specified
format, but can't decide which method would be the most efficient.  The
data format is simple.  The data is a set of (key,value) pairs.  One-line
values are of the format:

Key: one line value

Multiline values appear like this:

.KEY
multiline value
multiline value
multiline value
multiline value
.KEY

where each value line is escaped if a period actually falls into the first
column.  I want to read these pairs into a dict.  The process for the
first case seems simple:

(key, value) = string.split(line, ':', 1)
dict[key] = value

But I'm not sure which approach I should take because of possible
repercussions.  Some of my initial thoughts, minus the nitpicky stuff
(like escaping periods and chomping lines):

1)
temp = ''
for line in file.readlines():
        if line[0] == '.':
                if temp:
                        temp = ''
                else:
                        temp = line
        else:
                if temp:
                        dict[temp] = dict[temp] + '\n' + line
                else:
                        (key, value) = string.split(line, ':', 1)
                        dict[key] = va lue

This approach is unappealing for two reasons.  First, I don't like the
idea of keeping track of too much state (temp).  I know it's not much, but
it just seems unnecessary.  Second, and more importantly, the multiline
case is likely to be much larger than the singleline case is numerous.  I
would rather optimize for it, and I know a = a + b is not the way to do
that.

2)
lines = file.readlines()
while lines:
        line = lines[0]; del lines[0]
        if line[0] == '.':
                index = lines.index(line)
                dict[line[1:]] = string.join(lines[:index], '\n')
                del lines[:index+1]
        else:
                (key, value) = string.split(line, ':', 1)
                dict[key] = va lue

This approach seems to handle the worst case better, by joining rather
than handling lines one by one.  However, I would much rather use a for
loop in thise case than a while.  I'm also worried about all the slicing
going on, though individual lines are likely to be 80 characters or less.

3)
As yet uncoded...reading the file into a single string rather than based
on lines, pulling out the multiline stuff based on a regular expression
search (something like \.(\w+)\n.*\.\1\n), but that sounds like it would
suffer too much from regex and string buffer copies.

Any advice or new methods?

Please forgive typos and syntax issues...the above really is pseudocode.


----------------------------------
Nathan Clegg
 nathan at islanddata.com





From fdrake at cnri.reston.va.us  Fri Jul 23 12:56:05 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 23 Jul 1999 16:56:05 GMT
Subject: Python documentation available as ZIP files
Message-ID: <14232.40613.53508.888049@weyr.cnri.reston.va.us>

  I've received enough hostile mail from Windows users not
understanding how to unpack .tgz archives (er, WinZip works for me...)
that I will be making most of the documentation packages available as
ZIP archives as well as .tgz archives.
  For the most part, the ZIP archives are approximately the same size
as the .tgz archives, but be aware that, even with maximal
compression, the HTML ZIP archive is substantially larger than the
.tgz equivalent.  This is likely caused by the ZIP format having two
directory entries for each file instead of one; with a lot of small
files, this can make a difference.
  The new files are now available on the master FTP and Web sites at
python.org; they should appear on mirrors over the coming week.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From tismer at appliedbiometrics.com  Thu Jul  1 17:03:05 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Thu, 1 Jul 1999 21:03:05 GMT
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain>
		<377B7E17.5BD7@nospam.com> <377bbf9b@news.xtml.co.uk>
Message-ID: <377BD789.C4A1D191@appliedbiometrics.com>


Nick wrote:
> 
> Don't want to put the cat amongst the pigeons but Lua
> (http://www.tecgraf.puc-rio.br/lua/), is a language specifically designed to
> be embedded and it runs about 50-100% faster than Python and is a lot more
> compact (30k or something). It has a lot of similarities syntactically as
> far as I can see.
> 
>  If I was to use an embedded scripting language in a game I would use Lua
> not Python. If I was to write tools for the game I would use Python.

I had a quick look, but not the time to check it completely.
Does it have coroutines or small threads to do very
quick context switches? This is something which
Python will be able to do very fast, soon. You can have
thousands of current contexts and switch them in no time.
This can simplify the implementation of games in a scale,
that the speed penalty might be compensated by that.
Well, we have to try...

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From skodela at photon.sindh.com  Wed Jul 28 09:51:28 1999
From: skodela at photon.sindh.com (skodela at photon.sindh.com)
Date: 28 Jul 1999 14:51:28 +0100
Subject: a little motivation, anyone?
References: <19990728052704.10608.00002260@ng-fq1.aol.com>
Message-ID: <m3pv1cyja7.fsf@photon.sindh.com>

I am not a professional programmer, but a hobbyist. You name the language 
I must have looked at it once. Though now I settled for a bit of Delphi under 
windows and Python under linux.

Found the python to be pretty good. Also my C knowledge is nonexisting. 
If you still want to see the programs [programlets I wrote/messed up]
well, I will send them to you and perhaps you can show me a bit of C.

Regards
sreekant



From gerrit.holl at pobox.com  Thu Jul  8 16:41:10 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 8 Jul 1999 20:41:10 GMT
Subject: Why are index() and count() only for mutable sequences?
In-Reply-To: <3784F778.6A6B4568@lemburg.com>; from M.-A. Lemburg on Thu, Jul 08, 1999 at 09:09:44PM +0200
References: <004701bec95f$186faa80$a0ee6dd1@BobAlexDesk> <3784F778.6A6B4568@lemburg.com>
Message-ID: <19990708224110.B29441@optiplex.palga.uucp>

On Thu, Jul 08, 1999 at 09:09:44PM +0200, M.-A. Lemburg wrote:
> Date: 	Thu, 08 Jul 1999 21:09:44 +0200
> From: "M.-A. Lemburg" <mal at lemburg.com>
> To: Bob Alexander <bobalex at ix.netcom.com>
> Cc: python-list at python.org
> Subject: Re: Why are index() and count() only for mutable sequences?
> 
> Bob Alexander wrote:
> > 
> > There is no modification of the sequence performed by these functions
> > -- it seems as though they could be available for all sequences. It
> > would certainly be useful for tuples as well as lists (strings, too,
> > for that matter, even though we have the more general string.find()).
> 
> You could try count(), index(), forall() and exists() from mxTools:
> 
> 	http://starship.skyport.net/~lemburg/mxTools.html
> 
> Albeit, they work with conditions (functions that return 1/0) to
> implement the testing process.
> 

Is this an anwser??
He asked: "why"...

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From pointal at lure.u-psud.fr  Tue Jul 20 03:18:19 1999
From: pointal at lure.u-psud.fr (Laurent POINTAL)
Date: Tue, 20 Jul 1999 07:18:19 GMT
Subject: Embedding python/multi-thread env
References: <379377AF.E8E7351D@cannontech.com>
Message-ID: <37941e27.85316128@news.u-psud.fr>

Depending what you really need. You must know that there is a global
lock between the different Python interpreters. If you write your own
C extensions (really easy with Python), you can release this global
lock during your long computations which dont interract with Python.

See Python/C API Reference Manual, chapter 8.1 "Thread State and the
Global Interpreter Lock".

A+

Laurent.

On Mon, 19 Jul 1999 14:08:31 -0500, Aaron Lauinger
<aaronl at cannontech.com> wrote:

>: I am looking for a suitable scripting language for a multi-threaded
>: server application.  I am wondering if python might be a fit.  The
>: server must be able to run multiple scripts concurrently (in separate
>: worker threads) and must be able to stop any of these scripts cleanly at
>: any time.  The scripting language must also allow writing C extensions -
>: I am believe python meets this req.  I am not so sure about the threaded
>: part.  If python cannot meet the threading requirement is there a
>: scripting language that can?
>: Tcl also seems like a candidate, but all of the Tcl docs i've found
>: avoids the topic of embedding multiple interpreters in a multi-threaded
>: application like the plague.  Does what i'm looking for even exist?
>: Any insights are much appreciated.
>: 
>: Aaron Lauinger
>: Cannon Technologies
>: http://www.cannontech.com
>: 

---
Laurent POINTAL - CNRS/LURE - Service Informatique Experiences
Tel/fax: 01 64 46 82 80 / 01 64 46 41 48
email  : pointal at lure.u-psud.fr  ou  lpointal at planete.net 



From janssen at parc.xerox.com  Thu Jul 29 15:37:33 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Thu, 29 Jul 1999 19:37:33 GMT
Subject: shelve portability problem
In-Reply-To: <001f01bed9f4$4453e8f0$f29b12c2@secret.pythonware.com>
References: <eZUl3.1080$wm4.135734@news.uswest.net> <7nirjd$n52$1@m2.c2.telstra-mm.net.au> <7nq6a7$vka$1@nnrp1.deja.com>
	<001f01bed9f4$4453e8f0$f29b12c2@secret.pythonware.com>
Message-ID: <grc_pxIB0KGWRO44oY@holmes.parc.xerox.com>

Excerpts from ext.python: 29-Jul-99 Re: shelve portability problem
"Fredrik Lundh"@pythonwa (527*)

> alright.  this might work:

> import shelve
> import gdbm

> def myopen(filename, flag="c"):
>     return Shelf(gdbm.open(filename, flag))

> db = myopen("odbfile")

This should work.  I use shelve with bsddb this way.

Bill



From delgado at Mathematik.Uni-Bielefeld.DE  Fri Jul  9 09:31:54 1999
From: delgado at Mathematik.Uni-Bielefeld.DE (Olaf Delgado)
Date: Fri, 9 Jul 1999 15:31:54 +0200
Subject: Binary pickle portability
In-Reply-To: <Pine.WNT.4.04.9907071517320.248-100000@rigoletto.ski.org>
References: <Pine.WNT.4.04.9907071517320.248-100000@rigoletto.ski.org>
Message-ID: <Pine.OSF.4.05.9907091524530.9120-100000@pinax.Mathematik.Uni-Bielefeld.DE>

On Wed, 7 Jul 1999, David Ascher wrote:

> There is a bug in pickling of NumPy arrays, which I expect will be fixed
> in the next release.  Binary pickles works across byte-order and byte-size
> platforms (when it's bug-free =).

Thank you for the information! It's a bit painful to have to convert the
arrays back and forth right now, but good to know that it's not my fault.

I am not sure I understand what you mean by byte-order and byte-size
platforms. What's a counterexample?

Olaf




From kikutani at sprintmail.com  Tue Jul 27 09:49:02 1999
From: kikutani at sprintmail.com (Kikutani Makoto)
Date: 27 Jul 1999 13:49:02 GMT
Subject: Qt
References: <7njga0$361$1@ash.prod.itd.earthlink.net> <199907270653.QAA01576@envy.fulcrum.com.au>
Message-ID: <7nkdce$t1c$1@oak.prod.itd.earthlink.net>

On Tue, 27 Jul 1999 06:53:19 GMT, Richard Jones <Richard.Jones at fulcrum.com.au> wrote:

>    There's a fairly complete Qt/KDE binding maintained at:
> 
>       http://www.river-bank.demon.co.uk/software/

Thanks a lot.

I fetched it and did make install.
But I got the following error at trying examples.

mako% python aclock.py                /usr/local/src/Lang/pyKDEsrc-0.8/examples
Traceback (innermost last):
  File "aclock.py", line 4, in ?
    from kde import *
  File "/usr/lib/python1.5/kde.py", line 7, in ?
    import libkdec
ImportError: /usr/lib/libkimgio.so.2: undefined symbol: png_set_strip_16

What's wrong ?
My Linux Box is Debian potato, and kdelibs2g is version 1.1.1-19990523.
Python is 1.5.2 (python-gtk 0.6.1-1)

-- 
Kikutani, Makoto  kikutani at sprintmail.com




From skip at mojam.com  Fri Jul 30 13:14:42 1999
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 30 Jul 1999 12:14:42 -0500 (CDT)
Subject: 1.5.2 for: else:
In-Reply-To: <9907300817450G.00576@quadra.teleo.net>
References: <00ad01beda66$d6bdb9b0$f29b12c2@secret.pythonware.com>
	<9907300817450G.00576@quadra.teleo.net>
Message-ID: <14241.56670.798968.417540@dolphin.mojam.com>

    Patrick> The truth is out. "Frederick Lundh" is a bot!

Now we have to deal with the fredbot as well as the timbot.  Anybody for a
quick game of RoboWar?

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758




From skip at mojam.com  Thu Jul 29 10:30:50 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 29 Jul 1999 09:30:50 -0500 (CDT)
Subject: Is "shelve" thread-safe on Windows?
In-Reply-To: <7npimc$l9b@romeo.logica.co.uk>
References: <7npimc$l9b@romeo.logica.co.uk>
Message-ID: <14240.25927.863885.288661@dolphin.mojam.com>

    Andrew> Can anyone suggest a simple way to gain the functionality of
    Andrew> "shelve" on a Windows platform, but which is safe for handling
    Andrew> concurrent accesses (like from CGI scripts)?

You might take a look at file locking.  On Unix there's the fcntl module.
Not sure if there is similar functionality on Windows.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From Use-Author-Address-Header at [127.1]  Tue Jul 20 21:02:01 1999
From: Use-Author-Address-Header at [127.1] (Michael Huster)
Date: Tue, 20 Jul 1999 18:02:01 PDT
Subject: Scientific Plotting?
Message-ID: <19990721010203.96101.qmail@hotmail.com>

I am also trying to find an adequate *object* based plotting package. I am 
writing some signal processing software and want to code plotting into the 
classes I am defining.

Some requirements:
1) Postscript output (for inclusion in high quality Word & Powerpoint 
documents) GIF, JPG, etc just do not have the resolution. They default to 72 
dpi rendering. You can overcome this on Linux using Image Magick's convert 
-dens 300. The resulting files are large.

2) Object oriented
3) Images

Our current assessment:

GIST: not enough flexibility in configuring the plots. It has the Postscript 
we need, though. We have it coded in now for some look at the data. But it 
is not the long term solution.

Plplot: uses stroke fonts, even in postscript output. We cannot get a 
uniform look with matlab produced plots. Bummer. Otherwise it looked good. I 
am afraid when blown up the fonts will look even worse. For example. To get 
a bold font you have to play with pen width and font size. Isn't 
"Helevetica-Bold" easier to specify? Also, it is at a developmental (& 
FORTRAN) deadend. No chance to change things.

Gnuplot: Runs gnuplot through a pipe. All of the data (upto many MB) also 
has to go through the pipe. Unsat. I want to plot efficiently.

plotlib: not enough high level wrappers. I do not want to have to calculate 
default axis size, where to put tick marks, etc.

Pgplot: looks like plplot.

Piddle & graphite: looks good, but not mature enough. We are keeping an eye 
on it. Probably will try it.

PyMat: I want to get away from needing Matlab licenses

I am kind of bummed. If we wrap intelligent python wrappers around 
something, we will post it.

Michael Huster
mhuster at hotmail.com


_______________________________________________________________
Get Free Email and Do More On The Web. Visit http://www.msn.com



From javanet at sinbad.net  Mon Jul 26 21:41:03 1999
From: javanet at sinbad.net (Arnaldo)
Date: Mon, 26 Jul 1999 17:41:03 -0800
Subject: Newbie question
Message-ID: <379d0ee6@news.gci.net>

I'm just trying out some examples from the book Learning Python and I run
into a
exercise that is not working for me.
It's on page 95 , exercise 4b.
This is the answer the book has
===========================================
l = [1, 2, 4, 8, 16, 32, 64]
x = 5
i = 0
while i < len(l):
    if 2 ** x == l[i]:
        print 'at index', i
        break
        i = i + 1
else:
     print x , 'not found'
============================================

 When I run this , python hangs. When I do Ctrl-c , this is what I get
Traceback (inntermost last):
    File "powera.py, line 5, in ?
        if 2 ** x  == l[i]:

What am I doing wrong?
Thanks I lot






From NewsReplys at aymara.tcp.co.uk  Fri Jul 23 17:25:24 1999
From: NewsReplys at aymara.tcp.co.uk (Chris Williamson)
Date: Fri, 23 Jul 1999 21:25:24 GMT
Subject: Memory leaks whilst embedding python
Message-ID: <8E0CE4469aymaratcpcouk@news.tcp.co.uk>

I'm experimenting with embedding python into an MFC prgram running under 
Windows NT.

I managed to build a debug version of the python dll ok, but when I tried 
running a simple script the MFC debugger complained of memory leaks.

Removing statements showed that the following leaked

void Button1_press() {

    	Py_Initialize();

}

Looking through the source I came to the conclusion that perhaps I should 
be doing the following:

void Button1_press() {

    	Py_Initialize();
    	Py_Finalize();

}

but this still leaked.

Further investigation revealed that at least one of the items which was 
leaking was some form of string object which was initialised whilst the 
python string module was loaded.


Can anyone cast any further light on this? 


    	    	    	Thanks,

    	    	    	    	Chris Williamson.



From news at dorb.com  Thu Jul  1 13:48:09 1999
From: news at dorb.com (Darrell)
Date: Thu, 1 Jul 1999 13:48:09 -0400
Subject: Name of function within the function?
References: <377B8C2F.80F44F9@americasm01.nt.com>
Message-ID: <ITNe3.4319$U6.23498@newsr2.twcny.rr.com>

import sys, string, os

def myDirExcept(howFarBack):
        """
        Get the directory of the module running at "howfarBack "
        in the stack frames
        This value is compiled into the .pyc file
        This can be a problem.
        This name will change depending on where python was run from
        when the pyc file was created
        Unix can store a relative path
        Windows stores an absolute path
        """
        try:
                raise ZeroDivisionError
        except ZeroDivisionError:
                f = sys.exc_info()[2].tb_frame

        for i in range(howFarBack):
                f = f.f_back

        t = f.f_code
        fname = t.co_filename
        p = string.split(fname, os.sep)
        fname = string.join(p[:-1], os.sep)
 print t.co_name, fname
        return fname


myDirExcept(0)

###################################

E:\ace\v2\code\comp>junk2.py
myDirExcept E:\ace\v2\code\comp


--
--Darrell
Dula, Debbie (EXCHANGE:RICH2:2C25) <violet at americasm01.nt.com> wrote in
message news:377B8C2F.80F44F9 at americasm01.nt.com...
> I've looked and looked, but can't find the answer to this ... I'm
> sure it's right in front of me ...
>
>
> I want the current function's name to be output in my error
> messages, but can't figure out how a function can get its own name
> (without already knowing it).  For example,
>
> def f1():
> print "Error in function", f1.__name__
>
> is not really useful for what I need as I might as well print "Error
> in function f1".  Is there any generic way I can get the current
> function's name?
>
> Thanks,
> Debbie





From aahz at netcom.com  Sat Jul 17 16:08:33 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 17 Jul 1999 20:08:33 GMT
Subject: I envy newbies
References: <000301bed08a$d63c68a0$27a22299@tim>
Message-ID: <7mqns1$315@dfw-ixnews16.ix.netcom.com>

In article <000301bed08a$d63c68a0$27a22299 at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>
>FAT32 and NTFS don't care.  FAT16 actually has room for 512 entries,
>although files/folders with "non-8.3 names" consume more than one entry.
>Details at e.g.
>
>    http://support.microsoft.com/support/kb/articles/q120/1/38.asp

Hmmm...  I'm suspicious; unless my memory is faulty, I *know* there were
filesystems with a maximum of 64 root entires.  Maybe it was just floppy
disks, though.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From gerrit.holl at pobox.com  Tue Jul 20 15:45:03 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Tue, 20 Jul 1999 19:45:03 GMT
Subject: Finding absolute pathname?
In-Reply-To: <7n23mm$ukr$1@nnrp1.deja.com>; from rdudfield@my-deja.com on Tue, Jul 20, 1999 at 03:13:40PM +0000
References: <7n23mm$ukr$1@nnrp1.deja.com>
Message-ID: <19990720214503.A30258@optiplex.palga.uucp>

On Tue, Jul 20, 1999 at 03:13:40PM +0000, rdudfield at my-deja.com wrote:
> How do I find the absolute pathname of the current directory?
> 
> eg /usr/doc, or c:\temp
> 
> os.curdir gives .
> 
> I need the absolute path name ( is that what you call it ? ).
> 
> Thanks in advance.
> 

os.getcwd() gets the Current Working Directory.

regards,
Gerrit.



From gerrit.holl at pobox.com  Wed Jul  7 04:05:32 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Wed, 7 Jul 1999 10:05:32 +0200
Subject: interaction between modules... globals?
Message-ID: <19990707100532.A951@optiplex.palga.uucp>

Hello,

I have a program which consists of several little modules, but I encounter a
problem with that. I have one module which returns a class instance, with
messages as members, either Dutch or English. Every module needs that instance,
but I don't want to pass it as an argument to every little function, the
modules should find out theirselves. But I _also_ don't want to make that
instance 5 times.

So I thougt: I need a sort of global between the modules. Is this possible?

I'm now pickling and unpickling it, but that _very_ slow, of course:
bash:gerrit at gerrit:~/project/python/discoverb>>> time ./discoverb.py </dev/null
Welke file zal ik de woorden uit halen? Traceback (innermost last):
  File "./discoverb.py", line 168, in ?
    run()
  File "./discoverb.py", line 145, in run
    file = raw_input(mes.whichfile + " ")
EOFError: EOF when reading a line

real    0m2.504s
user    0m2.060s
sys     0m0.190s

more than 2 seconds, it _has_ to be possible faster to find out a language!

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From fredrik at pythonware.com  Thu Jul 22 04:18:21 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Jul 1999 08:18:21 GMT
Subject: Tkinter question
References: <37965ED6.A4238A27@anisci.com> <379669B7.BF4EDB37@aw.sgi.com>
Message-ID: <00d801bed41b$c9203810$f29b12c2@secret.pythonware.com>

Gary Herron <gherron at aw.sgi.com> wrote:
> > I've got a tkinter canvas, and I want to get it's width.... I call
> > 
> > w = int( thecanvas.cget("width") )
> > 
> > this works fine.. except when I resize the window, this still returns
> > the width that the canvas was when it was created.. the canvas is
> > getting resized (it has a visible background color)..
> > 
> > any ideas?
> > 
> > joel
> > 
> > --
> > ----------------------------------------------------------------
> >                                                 joeld at anisci.com
> 
> Use
> 
>   thecanvas.winfo_width()
> 
> This works for any widget, not just canvas.
> 
> Here's a list of all the winfo_... methods, taken from Tkinter.py:

...

and here's the documentation:
http://www.pythonware.com/library/tkinter/introduction/basic-widget-methods.htm

</F>




From brian.pedersen at mail.danbbs.dk  Tue Jul 20 08:44:43 1999
From: brian.pedersen at mail.danbbs.dk (Brian Pedersen)
Date: Tue, 20 Jul 1999 12:44:43 GMT
Subject: python vs. lisp (whereis quote?)
References: <x88r9m3cyje.fsf@rpppc2.hns.com>
Message-ID: <37946F3B.ECBFE6D6@mail.danbbs.dk>

nbecker at fred.net wrote:
> 
> One feature of lisp (emacs lisp actually) that I think is pretty cool
> is the "hooks".  A hook is basically unevaluated code that will be run
> by eval later.
> 
> I suspect python lacks this capability.  As I understand it, the only
> similar capability is "eval".  But I think eval expects a string,
> which it will then parse.  While the capability is similar, having to
> parse a string could be inefficient.
> 
> Am I correct that what is missing is the ability to postpone
> evaluation using "quote"?

you could compile() your statement in advance, making the exec() a
little more efficient

--                                                             
Brian Pedersen, M.Sc.E.E.                         /\               
DSP Software and Algorithm Development       .~. /  \ .~.
URL: http://www.danbbs.dk/~kibria/brian/        ~    ~
NOTE: New e-mail address !



From nvsoft at worldnet.att.net  Mon Jul 12 14:11:09 1999
From: nvsoft at worldnet.att.net (val shkolnikov)
Date: Mon, 12 Jul 1999 14:11:09 -0400
Subject: SSL intro.
References: <7mce3s$4vf$1@news2.acs.oakland.edu>
Message-ID: <378A2FBC.A0760D28@worldnet.att.net>

Pls see  http://home.att.net/~nvsoft1/ssl_wrapper.html

"Jay R. Wren" wrote:

> I was wondering if anyone had some pointers on python and SSL.  I found a
> paper about SSLeay with python and it has the object reference, but it
> doesn't say how to actually install SSL for python.
>
> Thanks,
>
> Jay R. Wren




From dalke at bioreason.com  Sun Jul 11 17:16:13 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Sun, 11 Jul 1999 15:16:13 -0600
Subject: code coverage
References: <37883E0B.D956F029@bioreason.com>
Message-ID: <3789099D.E3336ED6@bioreason.com>

Hello,

  I improved the detection of exectuable lines in the code
coverage tool I mentioned yesterday.  Instead of getting
data from an __import__ of a module, I use the parser module
to compile the text of a file.  This gives me access to code
which would be executed during the import.

  Again, the results are at
     ftp://starship.python.net/pub/crew/dalke/trace.py

						Andrew
						dalke at bioreason.com



From ivnowa at hvision.nl  Wed Jul 14 05:03:11 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Wed, 14 Jul 1999 09:03:11 GMT
Subject: Recursive method
In-Reply-To: <378C214F.3CF9CDD3@eml.villa-bosch.de>
References: <378C214F.3CF9CDD3@eml.villa-bosch.de>
Message-ID: <199907140901.LAA22305@axil.hvision.nl>

On 14 Jul 99, Ralph Gauges wrote:

> Hans Nowak wrote: 
> > You have a nested function here. (I say function, because it does not
> > look like a method... it's missing the self argument!
> 
> Actualy it is a method, but JPython doesn't seem to have a
> self. I was missing that as well.

Hmm... I didn't know that... I always assumed that programs for 
regular Python and JPython would be compatible...?

> > So, to solve this, you could try a different approach... maybe
> > 
> > (you said it's a method, right?)
> > 
> > class Blahblah:
> >     ...
> >     def makeTree(self):
> >          ....
> >     def makenodes(self, C):
> >          ....
> 
> I tried that as well, but I got the same error message. Only
> after I moved the makenodes outside the class it worked.
> This might also be related to the missing 'self' in JPython
> since I can not call 'self.makenodes'.
> Since moving makenodes out of the class is actualy not what
> I want, I was wondering, if anybody has had a similar
> problem and found a solution.

I see the problem... not the solution, though, because I don't know 
much (anything, actually) about JPython. :^( I wonder, if JPython 
doesn't have 'self', how can you call another method from within a 
method? Things like this

def method1(self, a):
    self.methodb(a+1)

would not be possible. There must be a way around this... (looks 
expectantly at JPython guru(s))

Hans Nowak (ivnowa at hvision.nl)
Homepage: http://fly.to/zephyrfalcon



From aflinsch at njeb.att.com  Thu Jul 15 08:18:31 1999
From: aflinsch at njeb.att.com (Alex Flinsch)
Date: Thu, 15 Jul 1999 08:18:31 -0400
Subject: Terminating Python...
References: <BFC0587634C3D111AACB00805FA766FA02C1C652@exchny12.ny.smb.com> <378CFDDF.489CDCEF@compaq.com> <7mj8s9$op2@newsops.execpc.com> <004901bece61$41270b60$f29b12c2@secret.pythonware.com>
Message-ID: <378DD196.BAB576C6@njeb.att.com>


Fredrik Lundh wrote:

> in which way is typing "raise SystemExit<Return>" better
> than typing "<Control-D>" (which is documented in the
> Python Tutorial) ?
>

"raise SystemExit" could be embedded in some python code, while a <Control-D>
could not be.
"raise SystemExit" would work on any OS, no need to check if we are running
windows or something else.


The only advantage would be a single consistent method of exiting python.




From kas at maps.magnetic-ink.dk  Mon Jul 19 07:32:58 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Mon, 19 Jul 1999 13:32:58 +0200
Subject: [Q] Type checking...
References: <7mv000$sun$1@feed.teaser.fr>
Message-ID: <slrn7p637a.4v2.kas@titan.magnetic-ink.dk>

Olivier Deckmyn <olivier.deckmyn at mail.dotcom.fr> wrote:

> Abstract : [Q]How to check the type of an instance before(after?)
> sending it to a method...

One possible solution is:

=== snip ===

MAGIC="I'm in class A :-)"

class ClassA:
  def __init__(self):
    self.magic = MAGIC
      :
  # end def __init__
# end class Class A

def is_class_a(obj):
  return has_attr(obj,'magic') and get_attr(obbj,'magic') == MAGIC
# end def is_class_a

class ClassB:
  def myMethod(self,anInstance):
    if not is_class_a (anInstance):
      raise TypeError, ...
        :
    # end if
  # end def myMethod
# end class ClassB

=== snip ===


  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From tbryan at arlut.utexas.edu  Fri Jul 16 09:21:26 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Fri, 16 Jul 1999 08:21:26 -0500
Subject: Python 1.5.2 + Linux 2.0.36 == Bad Combo?
References: <slrn7ou0ru.3sp.kas@titan.magnetic-ink.dk>
Message-ID: <378F31D6.99BF0197@arlut.utexas.edu>

Klaus Alexander Seistrup wrote:
> 
>   G'day mates,
> 
> I'm running Python 1.5.2 on a Linux 2.0.36 kernel (RedHat 5.1 distro partly
> upgraded to v5.2) and have recently experienced malfunctioning and/or non-
> working python applications -- specifically:

Just to make sure....

Are you sure you're invoking the right Python?  Red Hat comes with 
Python, and if your created a separate installation in /usr/local or
something, then you need to make sure that all of your scripts are 
invoking the 1.5.2 interpreter and not the one that came with 
Red Hat 5.x.

If you want to blow away Red Hat's Python, you can get the 
Python RPMs that Oliver Andrich put together.  They're at
http://andrich.net/python/


-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From janssen at parc.xerox.com  Wed Jul 14 13:45:00 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Wed, 14 Jul 1999 17:45:00 GMT
Subject: problem compiling 1.5.2 with egcs-1.1.2 on Solaris 2.6
In-Reply-To: <99Jul14.091842pdt."3556"@watson.parc.xerox.com>
References: <99Jul14.091842pdt."3556"@watson.parc.xerox.com>
Message-ID: <orXAmQ0B0KGWJivABy@holmes.parc.xerox.com>

When I remove the "-ansi" switch to the compiler, everything compiles OK.

Bill



From fredrik at pythonware.com  Thu Jul  1 05:21:42 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 1 Jul 1999 09:21:42 GMT
Subject: Tkinter: why does an empty frame not shrink?
References: <19990625162938.39336@nms.otc.com.au>
Message-ID: <001701bec3a3$23b802a0$f29b12c2@secret.pythonware.com>

Greg McFarlane wrote:
> If you have several child widgets packed in a frame and you unpack the
> children one by one, the frame will gradually get smaller, except when
> the last child is unpacked, in which case the frame does not resize. 
> Does anyone know why the frame does not shrink to 0x0 when you unpack
> the last child?

when geometry propagation is on (as it is by default), the
packer sets the size of the geometry master to exactly fit
the child widgets.  if there are no children, the packer
simply keeps the current size.

> Any solution?

why not just insert an extra widget that you never delete?

from Tkinter import *

root = Tk()
root.geometry("400x400")

frame = Frame(background="red")
frame.pack(fill=X)

Frame(frame).pack()

for i in range(5):
    button = Button(frame, text="Unpack me")
    button.configure(command=button.pack_forget)
    button.pack()

mainloop()

</F>




From pserrand at worldnet.fr  Sat Jul  3 09:04:00 1999
From: pserrand at worldnet.fr (pierrette)
Date: Sat, 03 Jul 1999 15:04:00 +0200
Subject: Upgrade or not ??
Message-ID: <377E0A3B.CA3B258C@worldnet.fr>

First a general question :
Actually we're working with Python 1.5.1 and we're satisfied with it.
Should we upgrade to 1.5.2 and if yes WHY?

No a Tkinter guestion :
We developped a prototype of an application with python an tkinter, now
we have to
decide if we're gonna write the real app in python or something else !
We now allready that in principle we could do it with python, the thing
we're missing
in tk is a 'on get focus' ,a 'on lost focus' event and a 'set focus to'
command
Pe: on lost focus of inputfield 1:
            if inputfield1 = A:
                set focus to field2
            elif inputfield1 = B:
                set focus to field 3

If those things are not available in tkinter what alternatives do we
have ?
- Changing the GUI (would be a shame but feasable), knowing that is has
to be multi-platform WIN-LINUX? If so to what ?
- Not write in python ? In this case we lose the work done for the
prototype

Thanks for comments and remarks

Pierrette Serrand
pserrand at worldnet.fr






From eric at estinc.com  Wed Jul 28 17:25:04 1999
From: eric at estinc.com (Eric Lee Green)
Date: Wed, 28 Jul 1999 14:25:04 -0700
Subject: Problem with signals & exceptions...
References: <2BB05D8530326D0C.420888E8906528AF.C6BF325D12E9ED90@lp.airnews.net> <7nnbme$ns2$1@nntp6.u.washington.edu>
Message-ID: <DA4B3986E94F79C9.E3C10FB44F334D06.D3443001EBA1A2FA@lp.airnews.net>

Donn Cave wrote:
> 
> Eric Lee Green <e_l_green at hotmail.com> writes:
> | I'm having trouble catching my thrown exceptions under FreeBSD 3.2. I am
> | using the Python 1.5.2 that came with FreeBSD 3.2 (well, on disk 1 of
> | the <n>-disk set). The following program works properly under Red Hat
> | Linux version 6.0 running a self-compiled Python 1.5.2 (in place of the
> | 1.5.1 which comes with Red Hat 6.0, which does not properly handle
> | signals during socket i/o calls). How it works: it listens to port 6664.
> | When it gets a connection, it then sets an alarm handler and alarm and
> | then tries to read from the socket. If it does not get data within 10
> | seconds, it then times out, at which point the alarm handler throws an
> | exception and aborts the socket read.
> 
> I tried this exception-from-handler notion on Digital UNIX and BeOS,
> and it does seem to work the way I guess you expect, but I do find it
> weird. A signal handler executes outside the normal flow of control,
> so it's not obvious to me where its exceptions ought to surface.
> Probably it should be expected to work the same way on various platforms
> for any given version of Python, but would future versions of Python be
> guaranteed to work this way also?

That, of course, is the ten million dollar question. The operation of signals
in "C" is extremely well documented and described. Under Python, it's ???. 

I changed my code to set a global variable rather than throw an exception and
that works the way you'd expect. A socket.error exception is thrown, and I
catch it, check the global variable, if the global variable is set I print "got
a timeout!" else re-throw the exception. 

In the actual working code I scrapped the alarm signal altogether and instead
do a select() call with a timeout, which presumably is guaranteed to work the
same on all platforms (it better, the select call is a POSIX standard!). 

-E



From phd at emerald.netskate.ru  Fri Jul 30 08:53:59 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Fri, 30 Jul 1999 16:53:59 +0400 (MSD)
Subject: os.name
In-Reply-To: <V5ho3.4340$VL2.463886@news.direcpc.com>
Message-ID: <Pine.LNX.4.04.9907301653090.8130-100000@emerald.netskate.ru>

On Fri, 30 Jul 1999, Emile van Sebille wrote:
> >>> os.environ('OS')
> Traceback (innermost last):
>   File "<pyshell#8>", line 1, in ?
>     os.environ('OS')
> AttributeError: no __call__ method defined

   os.environ is a dictionary:

>>> os.environ['OS']

> --
> 
> Emile van Sebille
> emile at fenx.com
> -------------------

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From mal at lemburg.com  Thu Jul 29 07:35:30 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 29 Jul 1999 13:35:30 +0200
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com> <379f06e9@news.xtml.co.uk> <al8btcvudgi.fsf@sirppi.helsinki.fi>
Message-ID: <37A03C82.9491678@lemburg.com>

Markus Stenberg wrote:
> 
> "Nick Trout" <nick@#spam_me_not#videosystem.co.uk> writes:
> > >> Thus I've had great deal of interest in getting the speed up; however,
> > >> speeding up dynamic interpreted languages is nontrivial, especially
> > >> languages as dynamic as Python.
> > Are there any pieces of test code anywhere which could be used as a good
> > test for Pythons speed (encompasing all of its language features) so that if
> > anyone tinkered with the bytecode execution unit they could measure the
> > speed improvement. It would be useful if it also tested correctness to
> > ensure no mistakes?!
> 
> There is PyStone (comes along with distribution) and PyBench (which I
> haven't tried). I myself use a performance testing framework and ton of
> small examples to judge whether things speed up or not.

FYI, pybench can be downloaded from:

	http://starship.skyport.net/~lemburg/pybench-0.6.zip

There's currently no web-page for it... too busy, sorry. Usage should
be fairly obvious though and it's extendable too.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   155 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From gmcm at hypernet.com  Thu Jul  8 14:13:22 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 8 Jul 1999 13:13:22 -0500
Subject: win32 python as a dll?
In-Reply-To: <3784BD4B.7938B7BA@enews.nrl.navy.mil>
Message-ID: <1280668553-75882201@hypernet.com>

Scott Wolford wrote:

> Here's the smallest case that gives me problems:
> #include "Python.h"
> #include <stdio.h>
> int main(int argc, char **argv)
> {
> //  return Py_Main(argc, argv);
>   FILE *file = fopen(argv[0], "rt");
>   PyRun_AnyFile(file, argv[0]);
> }

Where's Py_Initialize? Py_Finalize?
 

- Gordon



From Jens.Calisti at eiskonzept.com  Wed Jul 14 10:31:18 1999
From: Jens.Calisti at eiskonzept.com (Jens Calisti)
Date: Wed, 14 Jul 1999 16:31:18 +0200
Subject: bytecode & embedding
Message-ID: <378c9c8c.0@hedwik.mik.net>

Hi all,

I am new to python so maybe someone can answer me this :

Is it possbile to exeute python-compiled-bytecode from a c-programm with
embedded python ?? (eval,exec ?)


Thanx,


Jens




From bucodi at worldnet.fr  Mon Jul 19 03:36:19 1999
From: bucodi at worldnet.fr (pserrand)
Date: Mon, 19 Jul 1999 09:36:19 +0200
Subject: designing question
Message-ID: <3792D572.7F223EC3@worldnet.fr>

I have the mission to develop a rather big database front-end
application. Since the application must be multi-platform LINUX -WIN it
will be developed in Python.

For infomatiion, this is what will be used :
Python of course
Tk
Pmw
Gadfly

I have two questions:
1. The database will hold abouth 50.000 lines in its biggest table, can
gadfly handle this without problems, what abouth acces time ??

2. The application will have abouth 50 different inputscreens wich acces
all to different tables.
I will create a class for every window wich inheret from a general
'input-screen'-class.
I wonder if a should place the db-access code (insert-update-delete) in
every specific window or if i should  create a big db class wich has all
possible db-acces in it ?
Ex :
Possibility 1 :
class window :
    def __init__
                ..
    def createwidget:
                ...
    def handlevents :
                ...
    def insertrecord :
            ...
    def deleterecord:
            ....
    def update record :
            ....
Possibility 2:
class window
    def __init__
                ..
    def createwidget:
                ...
    def handlevents :
            call here db().inserttable1()
                            db().updatetable1()
                            db().deletetable1()

class db
    def __init__:
            connect to db
    def inserttable1
        ..
    def updatetable1
        ...
    def deletetable1
        ...

Hope my question is clear and sorry for the bad english, i'm french :)

Pierrette




From fredrik at pythonware.com  Fri Jul 16 11:45:50 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 16 Jul 1999 15:45:50 GMT
Subject: Case mismatch on NT
References: <378CCB4A.BDAD6FC7@bigfoot.com> <5loghdcyr2.fsf@eric.cnri.reston.va.us> <378F3C9A.872455B0@bigfoot.com>
Message-ID: <001c01becfa2$48e74440$f29b12c2@secret.pythonware.com>

Hoon Yoon <hyoon at bigfoot.com> wrote:
>   If I use the second way, then I cannot use either portion of tk (and
> Dialog.py in Tk is pretty important)

afaik, it's obsolete.  new code should use
tkMessageDialog or tkFileDialog instead. see:

http://www.pythonware.com/library/tkinter/introduction/standard-dialogs.htm

if you have old code that relies on Dialog.py,
copy the file to the application directory, and
remove it from lib-tk...

</F>




From fredrik at pythonware.com  Tue Jul  6 09:30:16 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 6 Jul 1999 13:30:16 GMT
Subject: XML-RPC and firewall [solution]
References: <s77cf213.010@mail.conservation.state.mo.us>
Message-ID: <004701bec7b3$b085e170$f29b12c2@secret.pythonware.com>

Jeffrey Kunce wrote:
> I have been playing with XML-RPC on my intranet, but
> found that Fredrik's xmlrpclib.py wouldn't talk to external 
> servers through our firewall.
>
> Since urllib has proxy-server support built-in, I wrote a 
> urllib-based transport plugin for xmlrpclib.py (attached).
> Now I can speak XML-RPC with the rest of the world :-)

cool !

(I suppose I don't have to mention that I factored out the
Transport stuff just to make it easier to do things like this...)

...

just one question: does this generate the right content-type?
(the spec says text/xml; cannot find any code in your module
or urllib.py that deals with this).  but I might be missing some-
thing here.  anyway, my sample server doesn't really care about
this, but other servers might...

...

> If it checks out, maybe /F will include it in the distribution.

sure will !

</F>




From ottinger at oma.com  Thu Jul 29 14:43:40 1999
From: ottinger at oma.com (Tim Ottinger)
Date: Thu, 29 Jul 1999 18:43:40 GMT
Subject: Still Not Getting Comp.Object.Moderated?
Message-ID: <36B6392C.1D23C4C0@oma.com>

Several people have written me (in my capacity as
c.o.m software janitor) about problems in accessing
comp.object.moderated. 

We already have multiple reliable points of insertion to
the 'net. We're getting propagation internationally, and
sometimes with amazing speed. There's really no problem
at our end :-)

So the problem is at your end, or somewhere in the middle.
News propagation can be a tricky thing. :-(

Right now, the very best thing you can do is to gently,
kindly, politely, and patiently ask your ISP to carry 
the moderated group. You may point them to :
<URL:ftp://ftp.isc.org/pub/control/comp/comp.object.moderated>
to see the valid control messages and for comfirmation that this is a
valid newsgroup.

Quite often, you can reach them at news at YOUR_ISP or less
directly at support at YOUR_ISP. The ISP may not answer back 
the same day, so I reiterate, be patient. Some ISPs took
several weeks to add the group. Also remember the problem
may not even be their fault, but someone upstream from
them.

If your ISP cannot get a reliable feed, they may contact the
comp.object.maintainer at oma.com. Maybe we can work something
out.

In the meantime, there's DejaNews and NewsGuy, and Doc's
list of open news servers (but don't use servers without
asking -- there can be trouble there)

-- 
Your Janitor
Tim

--------- There is no craftite conspiracy -------------
Tim Ottinger         Object Mentor      OO Training and
ottinger at oma.com      www.oma.com          Mentoring
-------------------------------------------------------
We can interpret a bad temper as a sign of inferiority.
                                        -- Alfred Adler




From scott at chronis.pobox.com  Tue Jul 27 10:42:57 1999
From: scott at chronis.pobox.com (Scott Cotton)
Date: Tue, 27 Jul 1999 10:42:57 -0400
Subject: strptime: unpleasant surprise (py 1.5.2 on Linux)
In-Reply-To: <m3aesltwg8.fsf@havenrock.com>; from Matt Gushee on Sun, Jul 25, 1999 at 09:16:55AM +0900
References: <m3aesltwg8.fsf@havenrock.com>
Message-ID: <19990727104257.A30584@chronis.pobox.com>

It shouldn't segfault, however, the function does work if
you reverse the arguments:

Python 1.5.2 (#8, May 28 1999, 21:09:50)  [GCC 2.7.2.3] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import time
>>> time.strptime('%x', '31/7/99')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
ValueError: format mismatch
>>>

I'm not sure, their may be patches to fix the strptime
segfault on linux problem - check www.python.org.

scott


On Sun, Jul 25, 1999 at 09:16:55AM +0900, Matt Gushee wrote:
| Hi, Folks--
| 
| I'm working on some date-related functions, and wanted to see what
| kind of exception 'strptime' would raise if I fed it a European-style
| date string with a US format. But instead of an exception, I get:
| 
| >>> import time
| >>> time.strptime('31/7/99', '%x')  # US locale, wants D/M/Y
| Segmentation fault (core import)
| 
| >>> dumped DateTime                 # mxDateTime version
| >>> DateTime.strptime('31/7/99', '%x')
| Segmentation fault (core dumped)
| 
| Surely this isn't what we want, is it? Has anyone else noticed this
| behavior?
| 
| I'm running Python 1.5.2 on RedHat Linux 5.1 (2.0.34 kernel with glibc 
| 2.0.7).
| 
| Matt Gushee
| Portland, Maine, USA
| mgushee at havenrock.com
| 
| 
| -- 
| http://www.python.org/mailman/listinfo/python-list



From john.michelsen at gte.net  Sat Jul  3 13:06:28 1999
From: john.michelsen at gte.net (John Michelsen)
Date: Sat, 03 Jul 1999 17:06:28 GMT
Subject: Python GUIs: Summary and Conclusion
References: <99FA4A1A1280BEBD.0A1ABA62CC989479.4C2C6022F695BEAC@lp.airnews.net> <19990629092757.A4998@optiplex.palga.uucp> <46F9FB3A1847A014.2CBD270176138613.20F84A39E6ED95DA@lp.airnews.net> <b_Pe3.558$6M6.167478@news.shore.net> <395C236424AEEAA7.C2D70DEDE2306228.0C4BC0DFFEC49C2C@lp.airnews.net> <slrn7nprcu.3a4.wtanksle@dolphin.openprojects.net>
Message-ID: <oqrf3.352$Ux1.10175@dfiatx1-snr1.gtei.net>

>>Well, the only reason I haven't been there is that I've only worked on
>>one GUI app _ever_ - a 2D tile-based strategy game [1].  I did plan
>>support for console apps (curses, etc., using a client-server
>>architecture), but I didn't much fancy the idea of playing with a
>>command line interface :)
>
>I'm working on a 2D tile-based adventure game, which right now is text
>only.  I'm expanding it so that it can use graphics, but I need some
>libraries.  Let me know what you find!


I'm writing a Tkinter/PIL based client for the Worldforge project.
Check out:
http://www.worldforge.org/website/client/tkClient/

John





From MHammond at skippinet.com.au  Thu Jul  8 18:52:39 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 9 Jul 1999 08:52:39 +1000
Subject: Python/PythonWin and DDE?
References: <3784b800.0@news.dbtech.net>
Message-ID: <7m39vq$r5o$1@m2.c2.telstra-mm.net.au>

> Karl Putland wrote in message <3784b800.0 at news.dbtech.net>...
> Are there any docs about PythonWin and DDE?

No, although DejaNews may find some examples.

> Specifically can Python be used as a DDE client?

Yes.  Pythonwin acts as both a client as a server as it communicates between
processes.

Mark.







From tim_one at email.msn.com  Thu Jul  8 23:24:04 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 8 Jul 1999 23:24:04 -0400
Subject: Why are index() and count() only for mutable sequences?
In-Reply-To: <004701bec95f$186faa80$a0ee6dd1@BobAlexDesk>
Message-ID: <000201bec9ba$7f3da900$c49e2299@tim>

[Bob Alexander]
> There is no modification of the sequence performed by these functions
> -- it seems as though they could be available for all sequences. It
> would certainly be useful for tuples as well as lists (strings, too, for
> that matter, even though we have the more general string.find()).

I don't think there's a deep reason.  It was "just true" that immutable
objects (sequences or otherwise) didn't have methods when Python first came
out.  But when complex numbers were introduced, they came with a
.conjugate() method, and you can expect strings to grow methods too in 1.6
(Barry Warsaw has checked implementation code for that into the CVS tree,
disabled by default for now -- it's nice!).

As you say, strings have more flexible functions already, and there's reason
to suspect that Guido didn't expect anyone to use tuples of more than a few
elements each.

why-are-tabs-8-spaces<wink>-ly y'rs  - tim





From nick at #spam_me_not#videosystem.co.uk  Wed Jul 21 12:08:41 1999
From: nick at #spam_me_not#videosystem.co.uk (Nick Trout)
Date: Wed, 21 Jul 1999 17:08:41 +0100
Subject: Merging two lists as a dictionary
References: <19990720211946.A12202@pooh.frostnet.net> <19990721153539.S6170@xs4all.nl> <19990721084614.C15502@pooh.frostnet.net> <T%kl3.3903$VL2.416831@news.direcpc.com>
Message-ID: <3795eec9@news.xtml.co.uk>

I think it's because you have to have Tools->Options->send... News - plain
text selected, not HTML?

N

Emile van Sebille <emile at fenx.com> wrote in message
news:T%kl3.3903$VL2.416831 at news.direcpc.com...
>Can anyone tell me why all of Chris's posts are showing up as
>attachments at my end?  (Outlook Express 5, win95b)
>
>--
>
>Emile van Sebille
>emile at fenx.com
>-------------------
>
>
>Chris Frost <chris at frostnet.advicom.net> wrote in message
>news:19990721084614.C15502 at pooh.frostnet.net...
>
>





From genglerb at no_spam.uwplatt.edu  Mon Jul 19 18:55:59 1999
From: genglerb at no_spam.uwplatt.edu (Brett Gengler)
Date: Mon, 19 Jul 1999 17:55:59 -0500
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net> <378FE5F3.A8394D50@bellsouth.net> <3790FA96.9403AA17@ssec.wisc.edu>
Message-ID: <3793ACFF.1EC1AF7@no_spam.uwplatt.edu>

Nick Bower wrote:

> > Since you indicate you are a beginner, I would like to recommend to
> > you O'Reilly's Learning Python by Mark Lutz and David Ascher for
> > $23.96 plus shipping from Amazon.com.
>
> When I started out on Python maybe 1.5 years ago, there were only 2
> books available in Australia.  So thanks for mentioning this new
> "Learning Python" book - it looks more like what I would have wanted at
> the time.  But one thing - why the hell is there a rat on the cover???
> Anyone??? ;)
>
> > For a more advanced treatment O'Reilly's Programming Python is good.
>
> Have to strongly disagree there :)  This book is soooo long-winded.  I
> think the original poster said she had C++ experience, so would have
> been confronted with the same reality as me - I need to skim all
> this???  I had similar programming experience so found that this book
> just didn't cater to these needs.  I actually got up and going on the
> smaller web programming book before I bothered to buy Programming
> Python.
>
> > of these books are very good at presenting Python is an understandable
> > way.
>
> I would lean towards saying this is because Python is understandable,
> rather than Programming Python being overly creditable <:)  ie  The
> books don't need to be good when Python is as simple as it is imho.
>
> nick

I am half way done with reading(not skimming) Programming Python by Mark
Lutz .  I do have some C++  and Pascal experience, although both in a
limited form. (a couple programming books, a couple simple programs)
Although I started learning Python off the web site and had already learned
the majority of the features, I found that the book gave me a more concrete
look at the language along by supplying me with implementation ideas that I
think gave me a better feel .  Anyone who has a lot of experience with a
HLL like C++ may really not need this book (more power to you) but for the
most part it would seem like a good buy-it was for me.   I would have to
disagree that the book is long winded.  I just started reading it and it
has been very enjoyable.  It's been my experience that programming books
aren't written so you can just kick it back and read 10 chapters.   I
usually read a chapter and then devote time to goofing around and trying
out the things I just learned.  It's fun that way.   The basic functions of
the language ARE  simple --it's the implementation that can get complex
which is why I'm excited to start reading the rest of the book which deals
more on that.   I feel confident that most people could handle this book
from the beginner to the intermediate.  I'm not exactly sure if the
'learning python' book is really a necessary though.   I'd start with
'programming python.'  Between the web site and the mini-tutorial in the
'programming python' you should be just fine.





From ivanlan at callware.com  Sun Jul  4 21:02:16 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 5 Jul 1999 01:02:16 GMT
Subject: Please send more info
References: <7los49$nga$1@bgtnsc03.worldnet.att.net>
Message-ID: <37800418.BEF2A3FD@callware.com>

Hi Steve--

Steve Wagner wrote:
> 
> I have a lot of spare time and would like to learn python. Please send info
> on
> how to get started. Also do you know of any online tutorials? thanks in
> advance for your help.

Visit
	www.python.org

Download python, install it, run the tutorial.

And in a few months, buy my book;-)

<blatant-self-promotion>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From tim_one at email.msn.com  Sat Jul  3 14:37:56 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 3 Jul 1999 14:37:56 -0400
Subject: Upgrade or not ??
In-Reply-To: <377E0A3B.CA3B258C@worldnet.fr>
Message-ID: <000301bec583$2b83d400$769e2299@tim>

[Pierrette Serrand]
> First a general question :
> Actually we're working with Python 1.5.1 and we're satisfied with it.
> Should we upgrade to 1.5.2

Of course.

> and if yes WHY?

Carrots and sticks.  And why wouldn't you?  It's free!

More than a year passed between 1.5.1 and 1.5.2, spanning hundreds of little
improvements and bugfixes, along with a few major ones.  So, e.g., if you
find a bug in 1.5.1, the response will be (http://www.python.org/1.5/):

    Forget about reporting a bug until you've confirmed that the bug
    still occurs in the 1.5.2 release

IOW, 1.5.1 is no longer supported.  And while the dozen <wink> lines of code
you've already written for 1.5.1 are very likely to work fine under 1.5.2,
as time goes on more & more new useful modules and code snippets are likely
to require new-in-1.5.2 features.

> No a Tkinter guestion :

Not a Tk guy, so I'll skip to the end:

> ...
> - Not write in python ? In this case we lose the work done for the
> prototype

If you can get a prototype done in Python 10x faster than in anything else,
and get the first crucial rounds of redesign done at that speed too, you win
big overall even if you do end up recoding the refined design in some other
language.

IOW, be realistic here:  you lose the work done for a prototype no matter
what language you write it in.  That's why they call it a prototype instead
of a product <wink>.

would-you-rather-throw-away-500-lines-or-5000?-ly y'rs  - tim





From amk at mira.erols.com  Wed Jul  7 20:37:27 1999
From: amk at mira.erols.com (A.M. Kuchling)
Date: 8 Jul 1999 00:37:27 GMT
Subject: make or similar in Python
References: <378390DD.58CA6BD5@nevex.com>
Message-ID: <slrn7o7t3i.eo.amk@207-172-111-58.s58.tnt1.ann.va.dialup.rcn.com>

On Wed, 07 Jul 1999 13:39:42 -0400, Greg Wilson <gvwilson at nevex.com> wrote:
>Hi, folks.  Has anyone ever written a dependency management/update tool
>(like 'make') in Python?
>Greg (who'd really, really like to override some of the things that
>GNU make does)

It's not Python-related, but there's a tool called icmake which has a
C-like scripting syntax; do a Web search for it.  Unfortunately it
doesn't seem to be maintained any longer; the archive I found was
dated 1994.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
All this progress is marvelous... now if only it would stop!
    -- Allan Lamport



From mal at lemburg.com  Fri Jul 23 04:08:21 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 23 Jul 1999 10:08:21 +0200
Subject: mxCGIPython Project - Update
Message-ID: <379822F5.1D7EA7CD@lemburg.com>

Hi everybody,

Here' an update of the currently available list of pre-compiled
one-file easy-to-install Python 1.5.2 interpreters (ideally
suited for CGI scripting)...

cgipython-1.5.2-AIX-2-000771064C00-.gz
cgipython-1.5.2-FreeBSD-2.2.8-RELEASE-i386-i386.gz
cgipython-1.5.2-FreeBSD-3.0-RELEASE-i386-i386.gz
cgipython-1.5.2-FreeBSD-3.1-19990426-STABLE-i386-i386.gz
cgipython-1.5.2-FreeBSD-3.2-STABLE-i386-i386.gz
cgipython-1.5.2-HP-UX-A.09.03-9000_715-unknown.gz
cgipython-1.5.2-HP-UX-B.10.20-9000_782-unknown.gz
cgipython-1.5.2-IRIX-5.3-IP19-mips.gz
cgipython-1.5.2-IRIX-6.5-IP32-mips.gz
cgipython-1.5.2-IRIX64-6.5-IP30-mips.gz
cgipython-1.5.2-Linux-2.0.34-i586-RedHat-5.1.gz
cgipython-1.5.2-Linux-2.0.34-mips-unknown.gz
cgipython-1.5.2-Linux-2.0.35-i586-Slackware-3.6.gz
cgipython-1.5.2-Linux-2.0.35-i586-unknown.gz
cgipython-1.5.2-Linux-2.0.35-i686-unknown.gz
cgipython-1.5.2-Linux-2.0.36-i586-Debian-2.1.gz
cgipython-1.5.2-Linux-2.0.36-i586-unknown.gz
cgipython-1.5.2-Linux-2.2.7-i586-SuSE60.gz
cgipython-1.5.2-NetBSD-1.3.2-alpha.gz
cgipython-1.5.2-NetBSD-1.3.3-i386.gz
cgipython-1.5.2-OSF1-V4.0-alpha-alpha.gz
cgipython-1.5.2-OpenBSD-2.4-i386-Intel_Pentium_II.gz
cgipython-1.5.2-SunOS-4.1.4-sun4m.gz
cgipython-1.5.2-SunOS-5.5.1-sun4u-sparc-alternate.gz
cgipython-1.5.2-SunOS-5.5.1-sun4u-sparc.gz
cgipython-1.5.2-Winnt-4.0.zip

Some platforms still on the wish list:

* Solaris / Intel platforms
* BeOS
* MacOS
* MacOS X

The current place to look for contributed binaries is 

     ftp://starship.skyport.net/pub/cgipython/ 

This will probably change to starship.python.net in the near
future. 

The above binaries contain the Python 1.5.2 interpreter
plus the whole standard lib (except for Tkinter) in one
single file: this reduces installation time/problems and
startup time as well.

More information about the project and how to contribute
can be found here:

        http://starship.skyport.net/~lemburg/mxCGIPython.html

Cheers and thanks to everyone who contributed,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   162 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From Christian.Tismer at p98.f112.n480.z2.fidonet.org  Thu Jul  1 17:03:05 1999
From: Christian.Tismer at p98.f112.n480.z2.fidonet.org (Christian Tismer)
Date: Thu, 01 Jul 1999 22:03:05 +0100
Subject: Python suitable for a game engine?
Message-ID: <0000163d@bossar.com.pl>

From: Christian Tismer <tismer at appliedbiometrics.com>



Nick wrote:
> 
> Don't want to put the cat amongst the pigeons but Lua
> (http://www.tecgraf.puc-rio.br/lua/), is a language specifically designed to
> be embedded and it runs about 50-100% faster than Python and is a lot more
> compact (30k or something). It has a lot of similarities syntactically as
> far as I can see.
> 
>  If I was to use an embedded scripting language in a game I would use Lua
> not Python. If I was to write tools for the game I would use Python.

I had a quick look, but not the time to check it completely.
Does it have coroutines or small threads to do very
quick context switches? This is something which
Python will be able to do very fast, soon. You can have
thousands of current contexts and switch them in no time.
This can simplify the implementation of games in a scale,
that the speed penalty might be compensated by that.
Well, we have to try...

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From bkc at Murkworks.com  Mon Jul 26 21:38:43 1999
From: bkc at Murkworks.com (Brad Clements)
Date: Mon, 26 Jul 1999 21:38:43 -0400
Subject: Memory leaks whilst embedding python
References: <1279335510-28879097@hypernet.com> <7nclk9$sg1$1@news.clarkson.edu> <7nitl6$q2r$1@m2.c2.telstra-mm.net.au>
Message-ID: <7nj2kf$37b$1@news.clarkson.edu>

Heh, ain't that always the case.

On another note, config.h and friends really need a cleanup. The nested
ifdefs are very difficult to figure out.

What's the *best* way for me to address this? I'd hate to go through the
1.5.2 source cleaning this up, just to see all the work lost.

Or worse, *think* I'm cleaning it up, but breaking it for Unix folks.

I'd like to:

1. Cleanup config.h and get all configuration options there. Some are hiding
out in posix and socket module too
2. Encapsulate all memory alloc/free for all modules through a set of
"My_Malloc" type calls that could be #def as must malloc/free or something
else depending on the platform.
3. Module finalization.

Not having worked on group projects of this size, I'm not sure what the best
method is to address these and make sure the changes aren't lost (and that
the changes are acceptable)

One other thing ... The current config.h implies that a given compiler
defines the target environment, which isn't true. I"m using Borland C as a
cross compiler for Netware (MSC can do it too), so I have to separate the
run-time environment options from the compiler selection. Thats the first
thing I want to address.


--
Brad Clements,
bkc at murkworks.com
Mark Hammond <MHammond at skippinet.com.au> wrote in message
news:7nitl6$q2r$1 at m2.c2.telstra-mm.net.au...
> Brad Clements wrote in message <7nclk9$sg1$1 at news.clarkson.edu>...
>
>  Im sure the list is quite large.
> Unfortunately for you, it seems you are the only one for which it is a
> "giant problem"
>





From paul at prescod.net  Tue Jul 13 18:27:25 1999
From: paul at prescod.net (Paul Prescod)
Date: Tue, 13 Jul 1999 22:27:25 GMT
Subject: XML-RPC
References: <931899884snz@vision25.demon.co.uk>
Message-ID: <378BBD4D.9DA43FE5@prescod.net>

Phil Hunt wrote:
> 
> Does anyone know if an XML-RPC server exists for Python?
> 
> (XML-RPC is a protocol for making remote procedure calls, where
> the parameters are sent in XML. The XML-RPC spec is at www.xmlrpc.com
> There's a python XML-RPC client at www.pythonware.com/products/xmlrpc/
> but I'm not aware of a Python server).

I think that you want to look at Zope.

-- 
 Paul Prescod  - ISOGEN Consulting Engineer speaking for only himself
 http://itrc.uwaterloo.ca/~papresco

To me programming is more than an important practical art. It is
also a gigantic undertaking in the foundations of knowledge.
	- Grace Hopper



From tim_one at email.msn.com  Sat Jul 17 14:30:28 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Jul 1999 14:30:28 -0400
Subject: Snippet: bitcount()
In-Reply-To: <14224.16968.304071.174087@lrz.de>
Message-ID: <000001bed082$726b19a0$27a22299@tim>

This has been beaten beyond death so often I don't have the heart to do it
all again <wink>.  Some factoids:

+ Christian's post summarizes best efforts pre-1.5.2.  For popcount4 it's
missing the nbits function.  Here's an elegant one that's reasonably fast:

def nbits(x):   # x >= 0 assumed
    if x <= 2:
        return x
    sum = 1
    while x:
        shift = 1
        x = x2 = x >> 1
        while x2:
            x = x2
            sum = sum + shift
            shift = shift + shift
            x2 = x >> shift
    return sum

Jurjen N.E. Bos's implementation of the constructive reals (Python FTP
contrib) contains a marshal-based method that's quicker for very large
longs.

+ nbits *should* be a constant-time operation, but there's no way to do that
in Python today.

+ hex(long) takes time linear in the number of bits in 1.5.2.  It used to be
quadratic-time.  This makes old tradeoffs open to question.  The notion that
it's too memory-intensive is misguided (longs are immutable -- you can't
even add 1 to a long without consuming 2*nbits(long) bits, and hex(long) is
no worse wrt memory than adding 1 twice <wink>).

+ "long1 & long2" (both > 0) takes time proportional to
       min(nbits(long1), nbits(long2))
in 1.5.2.  It used to take time proportional to the max of those.  This
makes shift-and-mask methods more attractive than they used to be.

+ The format of Python longs is defined in the source distribution's
    Include/longintrepr.h
  marshal-based tricks exploit that more-than-less directly.

+ About once a year someone gets excited about providing a pile of fast,
basic longint ops (via extension module or adding methods to longs -- like
popcount, trailing zero count, parity, nbits, long <-> int vector
conversions, set/clear/test/invert bit number i, find closest 0/1 bit
before/after bit number i).  Unfortunately, nobody has persisted long enough
to write a line of code, perhaps because people take such delight in endless
argument about what to name things <0.1 wink>.

the-next-person-to-get-excited-should-keep-it-a-secret-until-the-code-
    is-released-ly y'rs  - tim





From victor at idaccr.org  Thu Jul 15 13:09:28 1999
From: victor at idaccr.org (Victor S. Miller)
Date: 15 Jul 1999 13:09:28 -0400
Subject: Translating Matlab into Python
Message-ID: <ul7lo196wn.fsf@harder.princeton.idaccr.org>

Does anyone know if there's a program out there to translate (or help
in translating) MATLAB code into Python (probably with NumPy extensions)?
-- 
Victor S. Miller     | " ... Meanwhile, those of us who can compute can hardly
victor at idaccr.org    | be expected to keep writing papers saying 'I can do the
CCR, Princeton, NJ   | following useless calculation in 2 seconds', and indeed
    08540 USA        | what editor would publish them?"  -- Oliver Atkin



From holger at phoenix-edv.netzservice.de  Wed Jul  7 08:36:46 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Wed, 7 Jul 1999 12:36:46 GMT
Subject: Win32: python-network-installation?
Message-ID: <378349DE.7ED68819@phoenix-edv.netzservice.de>

High,

we've got an intranet with several linux- and WinNT-servers and 
many Win32-clients (Win95/98). 
It must be possible to run on every machine my python-scripts,
but I don't like to install every time the newest python-version
on every client. So I thought about a chance to make the python.exe
and pythonw.exe available on every client by net. I only would have to
administrate one installation on one server and every client could
run the python-scripts, too.

For a first idea is to manage this by manipulate registry-entries 
on client by only executing a *.reg-file onto this machines.

Is this possible?
What do I have to do?

thanx,
Holger



From robert_j_roberts at rl.gov  Fri Jul 23 17:55:14 1999
From: robert_j_roberts at rl.gov (Roberts, Robert J)
Date: Fri, 23 Jul 1999 14:55:14 -0700
Subject: Dreaded Newbie Question
References: <K4tK+KAA26l3EwvK@epstein.co.uk>
Message-ID: <3798E4C1.FB952A5A@rl.gov>

Oh wow. I can actually answer one ;-)

I use:      fiSize = os.stat(pathin)[6]   #...where pathin is full path to
file.

...and:

stat (path) = Perform a stat system call on the given path. The return
value is a tuple of at least 10 integers giving the most important (and
portable) members of the stat structure, in the order st_mode, st_ino,
st_dev, st_nlink, st_uid, st_gid, st_size, st_atime, st_mtime, st_ctime.
More items may be added at  the end by some implementations. (On MS-DOS,
some items are filled with dummy values.)

...as found on /python/doc/lib/node105.html



Dani Epstein wrote:

> Hi!
>
> A newbie to Python (though currently coding in Delphi), I'm having a
> little difficulty tracking down stuff, although I have the tutorial and
> manuals. Where is the best place to get more info, and why is there no
> newsgroup for newbies like myself?
>
> For the moment though my grateful thanks goes to anyone who can tell me
> how to get a file's size, and where that info is to be found in the
> references.
>
> BTW, I'm running PythonWin 1.5.1
> --
>
> Dani Epstein




From larsga at ifi.uio.no  Tue Jul 20 16:19:31 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 20 Jul 1999 22:19:31 +0200
Subject: Jpython Question
References: <Pine.GSO.4.05.9907191525580.20433-100000@allegro>
Message-ID: <wkr9m3846k.fsf@ifi.uio.no>

* Ian Kezsbom
| 
| I was just wondering if jpython can instantiate java classes and
| invokes methods on them...any help would be welcome...

It can, and it's very easy too. Here's a snippet from an email I sent
earlier today:

C:\Mine dokumenter>jpython
JPython 1.1beta1 on java1.1.7B
Copyright (C) 1997-1999 Corporation for National Research Initiatives
>>> from java.util import Hashtable
>>> h=Hashtable()
>>> h.put("larsga at ifi.uio.no","Lars Marius Garshol")
>>> h.get("larsga at ifi.uio.no")
'Lars Marius Garshol'
>>>

It's actually as easy as this.

--Lars M.



From stadt at freemail.nl  Wed Jul 21 09:40:17 1999
From: stadt at freemail.nl (Richard van de Stadt)
Date: Wed, 21 Jul 1999 15:40:17 +0200
Subject: uu.py - get result of uuencode() in a string
Message-ID: <3795CDC1.40F5E36@freemail.nl>

Is it possible to get the output of the uuencode() function
into a string, instead of a file or stdout?

I think, usually one uses uuencode to get a file accross
via email.

In CyberChair (http://trese.cs.utwente.nl/CyberChiair)
I use smtplib to send email, and would like to include
the output of uuencode() without the overhead of first
writing it to a file and then reading that file back in.

The alternative, of course, is to just copy&paste the
relevant parts of the uuencode() function.

Richard.



From petrilli at trump.amber.org  Fri Jul  2 13:01:37 1999
From: petrilli at trump.amber.org (Christopher Petrilli)
Date: 02 Jul 1999 10:01:37 PDT
Subject: Socket programming.
References: <jevhfnnqbgz.fsf@apollo.orakel.ntnu.no>
Message-ID: <7lir9h$36@journal.concentric.net>

Kjetil ?degaard <kjetilod at stud.ntnu.no> wrote:
> My current implementation is threaded (using fork()); one thread for
> user interaction and one for server output.  It uses blocking
> sockets.  It works ok, but I've read some docs here and there (amongst
> them the Socket Programming HOWTO on www.python.org) and I get the
> feeling I should be using select() and non-blocking sockets.  Also,
> I'm not sure I want to use fork() -- it seems very low-level and
> depends on a Unix-like system.

First, fork() isn't really threading... if you want to do threading, then
you should use the threading module, if it' ssupported on your platform.

Second, for that kind of thign, take a look at asyncore and asynchat,
the first of which I've written documentation for that will be released
in the next itteration of the Python library reference.  asynchat
should be able to do what you want.

Chris
-- 
| Christopher Petrilli                      ``Television is bubble-gum for
| petrilli at amber.org                          the mind.''-Frank Lloyd Wright



From sjturner at ix.netcom.com  Wed Jul  7 11:25:08 1999
From: sjturner at ix.netcom.com (Stephen J. Turner)
Date: Wed, 07 Jul 1999 11:25:08 -0400
Subject: python core dump (SIGBUS) on Solaris
References: <3F8BA119B48AD11197470008C7A493CB98B443@HQOBGEX1> <37830BB2.94C4613D@uk.sun.com>
Message-ID: <37837154.7A6889A@ix.netcom.com>

Gary Pennington - UK Performance Centre wrote:
> I've done (a lot) more investigation since yesterday. It appears that the
> problem is being generated inside cPickle in the loads function. I altered the
> Cookie._debabelize function to trap the error more explicitly and found the
> following:-
> 
> Test Code :-
> 
>     str = _unquote(val)
>     try:
>         return loads(str)
>     except Exception:
>         return str
> 
> Python 1.5.2 (#1, Jul  6 1999, 16:50:14)  [GCC egcs-2.91.66 19990314
> (egcs-1.1.2  on sunos5
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import Cookie
> >>> C=Cookie.Cookie()
> >>> C.load("webJudgeUser=garyp;")
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "/usr/local/lib/python1.5/Cookie.py", line 544, in load
>     self.__ParseString(rawdata)
>   File "/usr/local/lib/python1.5/Cookie.py", line 572, in __ParseString
>     M.set(K, apply(self.net_setfunc, (V,)), V)
>   File "/usr/local/lib/python1.5/Cookie.py", line 342, in _debabelize
>     return loads(str)
> cPickle.BadPickleGet: path
> >>>
> 
> Hmmm, something going wrong in the cPickle loads function not being trapped in
> the exception handler unless you specifically state Exception.

There's a bug in cPickle that causes heap corruption if loads is passed
a string (as opposed to a pickled object) starting with 'g', 'h' or
'j'.  Try applying the attached patch to the Python 1.5.2
Modules/cPickle.c source file, recompile python, and then see if the
problem goes away.

Regards,
Stephen

--
Stephen J. Turner <sjturner at ix.netcom.com>
-------------- next part --------------
Index: cPickle.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/cPickle.c,v
retrieving revision 2.36
diff -u -r2.36 cPickle.c
--- cPickle.c	1999/06/15 14:36:59	2.36
+++ cPickle.c	1999/07/07 14:57:17
@@ -3017,6 +3017,7 @@
     PyObject *py_str = 0, *value = 0;
     int len;
     char *s;
+    int rc;
 
     if ((len = (*self->readline_func)(self, &s)) < 0) return -1;
     if (len < 2) return bad_readline();
@@ -3024,14 +3025,16 @@
     UNLESS (py_str = PyString_FromStringAndSize(s, len - 1)) return -1;
 
     value = PyDict_GetItem(self->memo, py_str);
-    Py_DECREF(py_str);
     if (! value) {
         PyErr_SetObject(BadPickleGet, py_str);
-        return -1;
-      }
+        rc = -1;
+    } else {
+      PDATA_APPEND(self->stack, value, -1);
+      rc = 0;
+    }
 
-    PDATA_APPEND(self->stack, value, -1);
-    return 0;
+    Py_DECREF(py_str);
+    return rc;
 }
 
 
@@ -3040,6 +3043,7 @@
     PyObject *py_key = 0, *value = 0;
     unsigned char key;
     char *s;
+    int rc;
 
     if ((*self->read_func)(self, &s, 1) < 0) return -1;
 
@@ -3047,14 +3051,16 @@
     UNLESS (py_key = PyInt_FromLong((long)key)) return -1;
     
     value = PyDict_GetItem(self->memo, py_key);
-    Py_DECREF(py_key);
     if (! value) {
         PyErr_SetObject(BadPickleGet, py_key);
-        return -1;
-      }
+        rc = -1;
+    } else {
+      PDATA_APPEND(self->stack, value, -1);
+      rc = 0;
+    }
 
-    PDATA_APPEND(self->stack, value, -1);
-    return 0;
+    Py_DECREF(py_key);
+    return rc;
 }
 
 
@@ -3063,6 +3069,7 @@
     PyObject *py_key = 0, *value = 0;
     unsigned char c, *s;
     long key;
+    int rc;
 
     if ((*self->read_func)(self, &s, 4) < 0) return -1;
 
@@ -3078,14 +3085,16 @@
     UNLESS (py_key = PyInt_FromLong((long)key)) return -1;
     
     value = PyDict_GetItem(self->memo, py_key);
-    Py_DECREF(py_key);
     if (! value) {
         PyErr_SetObject(BadPickleGet, py_key);
-        return -1;
-      }
+        rc = -1;
+    } else {
+      PDATA_APPEND(self->stack, value, -1);
+      rc = 0;
+    }
 
-    PDATA_APPEND(self->stack, value, -1);
-    return 0;
+    Py_DECREF(py_key);
+    return rc;
 }
 
 

From pduffin at mailserver.hursley.ibm.com  Wed Jul 28 04:31:18 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Wed, 28 Jul 1999 09:31:18 +0100
Subject: make or similar in Python
References: <378390DD.58CA6BD5@nevex.com>
Message-ID: <379EBFD6.5E51@mailserver.hursley.ibm.com>

Greg Wilson wrote:
> 
> Hi, folks.  Has anyone ever written a dependency management/update tool
> (like 'make') in Python?
> I'm not looking for something that reads makefiles, and happens to be
> written in Python; I'm looking
> for a Python module that users import, then provide data to, in order to
> manage dependencies.  (The
> answer I got from the Perl crowd was 'cool idea, but no'; I'm hoping
> I'll have more luck here.)
> Thanks,
> Greg (who'd really, really like to override some of the things that
> GNU make does)

If you just want a cross platform one then take a look at "bras"

	http://wsd.iitb.fhg.de/%7Ekir/brashome/

which is a pure Tcl solution.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From joyce at netcard.com  Thu Jul 22 19:50:26 1999
From: joyce at netcard.com (Joyce Stack)
Date: Thu, 22 Jul 1999 23:50:26 GMT
Subject: python + bash + new user
References: <3793BE52.43E023BB@netcard.com> <7n0f3d$1d7@dfw-ixnews10.ix.netcom.com> <3793C729.55DE42D0@netcard.com> <7n5nfq$j0$1@vvs.superst.iae.nl>
Message-ID: <3797AE3B.687785E5@netcard.com>

Carel,

I did exactly as you said and chmod's etc but it just says that it doesnt recognise
"./simple.py" as a command.???..what am i doing work?? anyother simple programs will
work...

Joyce


Carel Fellinger wrote:

> Joyce Stack <joyce at netcard.com> wrote:
> > Sorry about being vague...i meant the Bash prompt...i have gotten other simple
> > programs to work...changed mode and executed the programs ...but i dont know how to
> > run the program and pass a variable at the same time...
>
> Gerrit showed you how to pass a command-line-script to python, and I will
> show how to pass parameters to your script, you can ofcourse combine the two:)
>
>         $ cat simple.py
>         #!/usr/bin/python
>         from sys import argv
>         print "I'll show you my parameters:"
>         for p in argv:
>                 print p
>
>         $ ./simple.py one parameter and others
>         ./simple.py
>         one
>         parameter
>         and
>         others
>
> hope this helps
>
> --
> groetjes, carel




From tfb at tfeb.org  Thu Jul  1 15:41:13 1999
From: tfb at tfeb.org (Tim Bradshaw)
Date: 01 Jul 1999 20:41:13 +0100
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com>
	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com>
	<375FDCA7.62FF@mailserver.hursley.ibm.com>
	<7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com>
	<37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu>
	<37776FFA.2D85@mailserver.hursley.ibm.com>
	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it>
	<876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org>
	<377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov>
	<slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net>
	<wkzp1g2t2e.fsf@ifi.uio.no>
Message-ID: <ey3r9mskvja.fsf@lostwithiel.tfeb.org>

* Lars Marius Garshol wrote:

> Huh? Forth dates back to the 60s, whereas Scheme is from 1975, and I'm
> quite unsure of whether Chuck Moore knew Lisp at all.

That's OK, call/cc is trivially causality-violating so scheme can
happily be an influence on languages older than it is.

--tim



From rodisi01 at my-deja.com  Wed Jul 21 09:45:27 1999
From: rodisi01 at my-deja.com (Isidor)
Date: Wed, 21 Jul 1999 07:45:27 -0600
Subject: Tool for Newbz (& beyond?): ModuleWriter
Message-ID: <199907211249.HAA14644@cro.ots.ac.cr>

Hello Pythonians -

Here is a little tool I promised in a previous message....

Context/Justification:

I have just started working with Python. As I worked my way
through the tutorial, I fired up IDLE and started experimenting
with modules and some ideas of my own. Fairly quickly I had
hundreds, if not thousands of lines of attribute/method/data
testing, nascent function definitions, working functions, etc.
Invariably, I would do something that would send Python off into
never-never-land. Crash, boom. Ok, so I had saved the session as
"my long and involved session trying to use rfc822.txt." But of
course, when I opened it up in IDLE again, it was just dead text.
All those interwoven functions and stored values and imported
modules...all gone. Somehow I needed to rescue them from the
background noise of fiddling with slices and other such stuff. I
could probably do this rescue work with some extensive and
repetitive searching/copying/pasting, but I figured, hey, I'm
going to be doing this a lot, and hey, this is, among other
things, a scripting language, so....why not write a script to do
it.

So, if you have big, nasty, messy text files of IDLE sessions,
and you would like to rescue the juicy bits, this may be the tool
for you:

ModuleWriter.py

(This may be a bad name for this beast. I am considering
"ModuleMaker," "Modulizer," "module_o_matic," ;) "modOmatic,"
etc. Your suggestions will be happily considered.)


What it does:

ModuleWriter asks you (at a command prompt, no windows yet) for
an input file and an output file (and a sorting method). It then
scans your input file (a saved IDLE session...even one that is
currently open) and saves into various lists and dictionaries the
following things:

(a)    "import lines": lines that start out with "import" or
  "from [blah import blah]"
(b)    "CodeChunks": chunks of code that start with "def blah("
  (function definitions) or "class blah:".

Then it creates the output file which you specified. At the top
of the output file it puts some info, then the import lines, and
then the code chunks. You can choose which of the two existing
sorting methods to use for the code chunks:

(a)    'earliest': Sort the chunks by the earliest line number
  upon which a version of a given chunk appears.
(b)    'latest': Sort the chunks by the latest line
  number...etc...

For example:

If the code chunks in the source file are arranged like this:

def a():   #version 1
     ...
def a():   #ver 2
     ...
def b():
     ...
def c():
     ...
def a():   #ver 3
     ...

When sorted by 'earliest,' the output order will be: a, b, c.
Sorted by 'latest,' it will be: b, c, a.


NOTE 1: At present, the output includes only the *latest* version
of a function or class (as determined by the line number). In
other words, a func or class has a unique name, any code chunk
that has that same name will be considered a version of that func
or class. The version is identified by the line-number upon which
it appears. For any given func or class, ModuleWriter picks the
version with the highest line number. THIS IS A PROBLEM IF: you
had a nice working version of a function that you then fiddled
with and broke...such that later versions work more poorly than
earlier versions. Referring to the example above, only "version
3" of function "a" will appear in the output file. (See "What It
Will Be Able to Do With More Work" below for solutions to this
problem.)


NOTE 2: This tool does not attempt to produce an executable
module as its output. In fact, if you do try to execute the
output, you will find that nothing happens. This is not a missing
feature, this is a safety precaution. I don't know about you, but
I have a lot of pointless test functions floating around in my
IDLE sessions. I don't want them accidentally executed. So, after
running ModuleWriter, open up the output file, delete the
unwanted functions and classes, the repeated import statements,
etc., and then get rid of all the ">>> "s. The search and replace
dialog makes this last task incredibly easy.


GETTING IT:

I don't have web space where I can put this right now, so e-mail
me at rodisi01 at my-deja.com, and every day or every other day, I
will send it off to whoever has requested it. I know this is a
hassle, I'm sorry. If it looks like there is serious interest in
this thing, I will find some web space for it so you can click
and save to your heart's content. In fact, I thought I heard
someone in this list talk about offering web space for python
projects....;)...

ALTERNATIVELY:

If someone will point me to (or give me) some suggestions re.
posting code to the newsgroup, I could just do that. I'm worried
about line breaks and all that stuff. Would an ascii attachment
work in deja-world? (I've never had much access to newsgroups, so
this is all sort of new to me, and I have to use the free and
powerful (many many thanks!) but somewhat frustrating deja-news
interface.)


DISCLAIMER (and request for comments):

This is the first module I have written. It is clumsy and ugly.
It is not optimized. There's a lot of python I know nothing about
yet, etc. etc. However, it works. Since it's not the kind of
thing that you run a thousand times a day, its poor performance
is not a major problem for the time being. I find it very useful
even in its present, crude incarnation, and I suspect other
newbies may do so as well. So, I am releasing it as it is for
now, with plans to make it better as I go. Of course, comments
and suggestions will make the improvement process go much more
quickly! ;) ...Even something as simple as: hey, you should think
about using module x instead of module y for line q/ function z.


What It Will Be Able To Do With More Work (i.e, To Do):

1.   Easy:

     a)   Make it take arguments from the command line: e.g.,
          ModuleWriter(inputfile, outputfile, sortversion) .
     
     b)   Add a graphical interface. I am working on figuring out
          Tkinter now. As soon as I think I can pull it off, I'll
          add a Tkinter interface.

2.   Not as easy, but useful:

     Add support for documentation. I'm working on this already.
     What I mean by this is that ModuleWriter will find and
     manage lines from your IDLE session that start with things
     like: "##NoteLocal: ", "##NoteGlobal: ", "##FAQ: ",
     "##Example: ", etc. I need to figure out what it should do
     with such lines, but I think this will be really nice. Some
     should go in the requested output file...as notes about
     functions and classes. Some should go in a globally selected
     "faq" file or whatever, so that every session you do, you
     can add to a personal faq that eventually can be shared with
     others....etc etc etc.

3.   Harder but even more interesting:

     All the versions of code chunks get stored during file
     processing. With a nice little gui "dictionary browser"
     (something like the IDLE class browser?), you could view any
     of the different versions and select which ones you want in
     the output. Such a "dictionary browser" could also be used
     to browse the faq file created by ModuleWriter. etc.


Again, any suggestions, advice, etc. towards these ends will be
gratefully accepted.


Ok, I hope this is useful to someone out there, it sure is for
me!

Take care...

Isidor




From fredrik at pythonware.com  Wed Jul 14 11:44:14 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 14 Jul 1999 15:44:14 GMT
Subject: Recursive method
References: <1280168234-10452044@hypernet.com> <378CA422.B2E99CB7@eml.villa-bosch.de> <004901bece0b$46982210$f29b12c2@secret.pythonware.com> <378CACF0.373116E@eml.villa-bosch.de>
Message-ID: <008701bece0f$babde130$f29b12c2@secret.pythonware.com>

Ralph Gauges <ralph.gauges at eml.villa-bosch.de> wrote:
> > probably because there is no instance when the
> > class definition is executed...
> > 
> > ("class" is a *statement*, not a declaration, and
> > is executed when the script is run.  see:
> > http://www.python.org/doc/current/ref/class.html
> > for details.  "def" is also a statement, btw...)
> > 
> > </F>
> 
> This sounds right, but when I have a class that pops up a
> gui on my screen, as I do now, it doesn't seem to make much
> difference, wether I build the gui right in the class
> definition or in the __init__ method. Both is executed when
> I make an Instance

nope.  try running this script to see
what's going on:

...

class Foo:
    print "this is executed when the class statement is executed"
    def __init__(self):
        print "this is executed when a new instance is created"

foo = Foo()

...

next, try removing the "foo = Foo()" line
and run it again.

finally, try adding multiple copies of
"foo = Foo()" and run it.

get it?

</F>




From Lars.Marius.Garshol at p98.f112.n480.z2.fidonet.org  Thu Jul  1 14:17:02 1999
From: Lars.Marius.Garshol at p98.f112.n480.z2.fidonet.org (Lars Marius Garshol)
Date: Thu, 01 Jul 1999 19:17:02 +0100
Subject: [Q] dbase-file/clipper-index read-module
Message-ID: <00001610@bossar.com.pl>

From: Lars Marius Garshol <larsga at ifi.uio.no>


* Timo Rainio
| 
| anyone around who knows about a Python-module for dbase-file (.dbf)
| access?  support for clipper-style NTX-index structures would be
| great. I only want to _READ_ clipper-files here.

I have a module that can handle .dbf and .fpt files at

<URL: http://www.stud.ifi.uio.no/~larsga/download/python/ >

It can't handle NTX files, since I didn't need that when I wrote
it. :)

--Lars M.



From ivanlan at callware.com  Wed Jul 21 14:50:16 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 21 Jul 1999 18:50:16 GMT
Subject: pow()
Message-ID: <37961668.FDF78B51@callware.com>

Hi All--
I couldn't find an explanation for the three-argument version of pow()
on DejaNews, the Python 1.5.2 docs or the Python FAQ, so I'll ask now.

Can someone explain (in simple terms, for a bear of small brain) to me
what the modulo argument to __pow__ is for?

	__pow__(self,other[,modulo])

Thanks very much,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From Rob.Hooft at p98.f112.n480.z2.fidonet.org  Thu Jul  1 08:30:24 1999
From: Rob.Hooft at p98.f112.n480.z2.fidonet.org (Rob Hooft)
Date: Thu, 01 Jul 1999 13:30:24 +0100
Subject: Tkinter: why does an empty frame not shrink?
Message-ID: <000015fe@bossar.com.pl>

From: R.Hooft at EuroMail.com (Rob Hooft)

>>>>> "GM" == Greg McFarlane <gregm at iname.com> writes:

 GM> If you have several child widgets packed in a frame and you
 GM> unpack the children one by one, the frame will gradually get
 GM> smaller, except when the last child is unpacked, in which case
 GM> the frame does not resize.  Does anyone know why the frame does
 GM> not shrink to 0x0 when you unpack the last child?  Any soluton?

In one of my applications, I put the frame inside another frame, and
resize the empty container frame to 1x1 manually to solve this. Ugly,
but it works:

    def makeWidgets(self,fn=None):
        self.rightbarvar=variables.booleanvariable() # used in a menu entry
        self.rightbarvar.set(1)
        ....
        self.rightbarc=Tkinter.Frame(self)
       
self.rightbarc.pack(expand=Tkinter.NO,fill=Tkinter.Y,side=Tkinter.RIGHT)
        self.rightbar=Tkinter.Frame(self.rightbarc)
        self.rightbar.pack(expand=Tkinter.NO,fill=Tkinter.Y,side=Tkinter.RIGHT)
        ....

    def togglerightbar(self):
        if self.rightbarvar.get():
           
self.rightbar.pack(expand=Tkinter.NO,fill=Tkinter.Y,side=Tkinter.RIGHT)
        else:
            self.rightbar.pack_forget()
        self.rightbarc.configure(width=1,height=1)

Regards,

Rob Hooft
-- 
=====   R.Hooft at EuroMail.net   http://www.xs4all.nl/~hooft/rob/  =====
=====   R&D, Nonius BV, Delft  http://www.nonius.nl/             =====
===== PGPid 0xFA19277D ========================== Use Linux! =========



From nospam at bitbucket.com  Thu Jul  8 03:02:56 1999
From: nospam at bitbucket.com (Phil Mayes)
Date: Thu, 8 Jul 1999 00:02:56 -0700
Subject: Two Pythons talking to each other?
References: <199907080019.TAA22244@buffalo.fnal.gov> <199907080529.HAA19502@axil.hvision.nl>
Message-ID: <J2Yg3.61$cF5.4836@newsfeed.avtel.net>

Hans Nowak wrote in message <199907080529.HAA19502 at axil.hvision.nl>...

>On 7 Jul 99, Charles G Waldman wrote:
>> There's no reason, just because the two programs are running on the same
>> machine, that they can't talk to each other using network protocols
>>[snip]
>I just tested this, and it works. But now I have another problem &
>question (which may be stupid, but I know hardly anything about
>networking and sockets). My programs send commands to each other in
>IRC style (like "/say this") or just plain text (chat). For receiving
>these strings, I use a non-blocking socket connection which polls the
>server every 0.x seconds (the period may vary). However, when the
>strings are sent too fast, which is the case with the 127.0.0.1
>address, and sometimes with regular connections too, the other side
>will pick two strings up as one string.
>
>So my question is, how does one generally deal with this? Should I
>attach a newline (\n) (or maybe another separator character) after
>every string? Or are there other ways to guarantee that two commands
>will arrive at the other side as two strings?


I think your best bet is to look at asyncore.py and asynchat.py which
are in the std lib as of 1.5.2.  I've just prototyped a client/server
using these, with both running on one machine.  Oh yes, take a look at
the samples that came with the original async lib:
http://www.nightmare.com/software.html (Thanks, Sam)
--
Phil Mayes    pmayes AT olivebr DOT com








From salzr at certco.com  Wed Jul 28 11:23:43 1999
From: salzr at certco.com (Rich Salz)
Date: Wed, 28 Jul 1999 11:23:43 -0400
Subject: API question -- decref after adding to dictionary?
References: <rps64v$0$37nspbi$7gc@news.supernews.com> <379EBC8B.E36F020B@lemburg.com>
Message-ID: <rpu849$1$37nspbi$7gc@news.supernews.com>

>Yes it is needed. (Trust the code in dictobject.c, after all, that's
>what gets executed !)


Thanks.  I wasn't sure I was reading the code correctly.

The API documentation doesn't seem to say what the functions do to
object references.  Am I missing it, or do I really have to read the
implementation to figure it out?  If the latter, my interest in Python just
dropped a few points.  I mean, refcounts are about the most important
part of writing extensions... (And yes, I understand that this constrains
future implementations.  So be it.)
    /r$







From aahz at netcom.com  Mon Jul 19 21:44:50 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 20 Jul 1999 01:44:50 GMT
Subject: python + bash + new user
References: <3793BE52.43E023BB@netcard.com> <7n0f3d$1d7@dfw-ixnews10.ix.netcom.com> <3793C729.55DE42D0@netcard.com>
Message-ID: <7n0kai$3ov@dfw-ixnews19.ix.netcom.com>

In article <3793C729.55DE42D0 at netcard.com>,
Joyce Stack  <joyce at netcard.com> wrote:
>
>Sorry about being vague...i meant the Bash prompt...i have gotten other
>simple programs to work...changed mode and executed the programs ...but
>i dont know how to run the program and pass a variable at the same
>time...

Look up sys.argv -- don't forget to "import sys" in your program.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From schorsch at schorsch.com  Wed Jul  7 09:10:11 1999
From: schorsch at schorsch.com (Georg Mischler)
Date: Wed, 07 Jul 1999 13:10:11 GMT
Subject: VIM and block jumping
References: <3781347D.CB2FBAB3@starvision.com> <Pine.SOL.4.05.9907051548470.13052-100000@saturn.radsci.uci.edu>
Message-ID: <7lvjjb$c4d$1@nnrp1.deja.com>

Fuming Wang wrote:
>
> Hi, I have a small patch that can make vim do following indentation:
> def func():
>     print "hello world!"
>     if a == b and \
>        b == c or  \
>        e == f:
>         print "got it!"
>
> I can send you the patch if this is what you like.

This may not exactly be what the original poster asked for,
but I still think it's of relatively general interest.
Would you mind posting it here?


Thanks

-schorsch

--
Georg Mischler  --  simulation developper  --  schorsch at schorsch.com
+schorsch.com+  --  lighting design tools  --  http://www.schorsch.com/


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From seanr at infi.net  Fri Jul 30 15:24:38 1999
From: seanr at infi.net (Sean Robertson)
Date: Fri, 30 Jul 1999 15:24:38 -0400
Subject: Include statements with python scripting
Message-ID: <7nsu3t$kui$1@nw001t.infi.net>

How do you do an include statement in python?  I know that in Cold Fusion it
would be "<cfinclude template="blue_header_noad.html">" and in an shtml file
it would be "#include http://data.pilotonline.com/connect2/blue_header.html"
but I need to know how this is done in Python scripting.  This script is
being embedded in a template for a classifieds page.  I have something like
this so far:

<html><head><title><!--#var name--> Search Results</title></head>

<!--#if mainclass == Employment:-->
    <!--#include http://data.pilotonline.com/connect2/blue_header.html-->
<!--#else:-->
    <!--#include http://data.pilotonline.com/connect2/default_header.html-->

<font size="+1">Main classification: <!--#var mainclass--> .....

I just need to know If that is how #include works, or is there some other
way that this is supposed to be done.  Please let me know as soon as
possible.  I don't check the email address in the sig over the weekend, so
please send any emails to webolutionary at webolutionary.com

--
Sean Robertson
Multimedia Artist, PilotOnline.com
seanr at infi.net





From w_keller at gmx.de  Thu Jul 15 15:35:45 1999
From: w_keller at gmx.de (Wolfgang Keller)
Date: Thu, 15 Jul 1999 21:35:45 +0200
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net> <377B711A.E8B65E0B@cfdrc.com> <377BC620.F5553FF5@easystreet.com> <b%Pe3.4033$lr3.164497@news2.jacksonville.net> <14204.4556.177837.851763@lrz.de> <Pine.LNX.4.10.9907012309060.6263-100000@us2.mayo.edu>
Message-ID: <1duu1rx.148nojs68o9buN@pec-84.au1.m.uunet.de>

Travis Oliphant <olipt at mayo.edu> wrote:

> Right now it fits what I want to do very well and it works on
> systems with an X server (aren't X servers available for MAC's and
> Windows?)  

There's even a complete POSIX subsystem available that runs as an
application on top of the MacOS (MachTen from Tenon), but you'd better
not even consider calling a beast that runs on something like that a
'MacOS application', because no self-respecting Mac user would ever use
it unless you point a loaded shot gun at his head and threaten to kill
him if he doesn't. :-)

If you want to develop an application for the MacOS, no matter whether
it is 'cross platform' or not, you absolutely _have to_ respect certain
rules (most of them are mentioned in Inside Mac). The last one who tried
to 'transplant' the WinDos and Linux way-of-thinking onto the MacOS was
StarDivision, whose StarOffice became a real success on WinDos and got
standing ovations by the Linux crowd, but it was a _complete_ failure on
the MacOS, so badly that they never finished version 4.0 for the MacOS
and they didn't even try to port 5.0.

No offense intended, just to prevent you from putting a lot of work into
something comeplete useless.

Regards,

-- 
Wolfgang Keller               

Zu Risiken und Nebenwirkungen von Junkmail lesen Sie 
de.admin.net-abuse.mail und fragen sie Ihren Postmaster oder Provider



From tim_one at email.msn.com  Sat Jul 24 01:11:11 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 24 Jul 1999 01:11:11 -0400
Subject: int(long(-sys.maxint-1)) fails on Linux
In-Reply-To: <14233.17625.185090.895516@146.chicago-41-42rs.il.dial-access.att.net>
Message-ID: <001001bed592$f2850d20$832d2399@tim>

[Tim]
> Check out this link:
>     [xxx]
> Arguably the best book on assembler programming ever written!
> The chapters on x86 addressing modes and LEA will chase the dark
> clouds away.

[Skip Montanaro]
> Does the timbot have any other interesting URLs squirrelled away in its
> large backside cache?

Yes:  http://www.python.org.

Other than that, I don't save URLs -- links break too often for a bot to
waste cycles updating.  Search engines are humans' friends too!  Although
they're fickle.  A better link for that book is the author's own server, at

     http://webster.cs.ucr.edu/

or-at-least-likely-there-for-the-next-10-minutes<0.9-wink>-ly y'rs  - tim





From dkuhlman at netcom.com  Thu Jul  8 16:18:44 1999
From: dkuhlman at netcom.com (G. David Kuhlman)
Date: 8 Jul 1999 20:18:44 GMT
Subject: Two Pythons talking to each other?
References: <199907072203.AAA09601@axil.hvision.nl>
Message-ID: <7m3134$59@dfw-ixnews17.ix.netcom.com>

I'm not a CORBA fan myself, but CORBA seems appropriate for your
problem.  Take a look at:

    Fnorb -- http://www.dstc.edu.au/Products/Fnorb/

and 

    ILU -- ftp://ftp.parc.xerox.com/pub/ilu/ilu.html (your Web
        browser can read this page)

both of which support Python.

  - Dave


Hans Nowak (ivnowa at hvision.nl) wrote:
> Hello,
> 
> I am currently experimenting with networking and want to write some 
> games (where, obviously, two players compete over a network). Now I 
> want to test my programs offline. One way would be to start two 
> separate Python sessions (on the same machine), both running the same 
> game program, and let them somehow talk to each other like they would 
> over a network. But, what would be a good way to do this? I was 
> thinking about passing strings (commands etc) to a file, but two 
> programs cannot read from and write to the same file at the same 
> time. Does someone have any ideas about how to do this?
> 
> TIA,
> 
> --Hans Nowak (ivnowa at hvision.nl)
> Homepage: http://fly.to/zephyrfalcon



From bryanv at arlut.utexas.edu  Thu Jul 22 08:02:01 1999
From: bryanv at arlut.utexas.edu (Bryan VanDeVen)
Date: Thu, 22 Jul 1999 07:02:01 -0500
Subject: Repost: Re: Scientific Plotting?
References: <379392E7.EB0F29E7@ssec.wisc.edu> <7n5fhn$9e5$1@nnrp1.deja.com>
Message-ID: <37970839.769ED88C@arlut.utexas.edu>

mhuster at hotmail.com wrote:

> I am also trying to find an adequate *object* based plotting package. I
> am writing some signal processing software and want to code plotting
> into the classes I am defining.
> 
> Some requirements:
> 1) Postscript output (for inclusion in high quality Word & Powerpoint
> documents) GIF, JPG, etc just do not have the resolution. They default
> to 72 dpi rendering. You can overcome this on Linux using Image Magick's
> convert -dens 300. The resulting files are large.
> 
> 2) Object oriented
> 3) Images

I don't think it is quite what you are looking for, but you might check
out VTK at http://www.kitware.com/vtk.html  It's main focus is high
quality 3D data visualization but it contains an image pipeline as
well.  The main problem for you might be the PostScript support (I think
it is there but I am not sure how complete)  On the other hand, it does
build python bindings (and tcl bindings too) for itself when you build
it, which is nice.

-- 
Bryan Van de Ven
Applied Research Labs
University of Texas, Austin



From MHammond at skippinet.com.au  Thu Jul  8 03:58:34 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Thu, 8 Jul 1999 17:58:34 +1000
Subject: win32api RegisterWindowMessage, GetMessage?
References: <7lvs7i$fri$1@nnrp1.deja.com>
Message-ID: <7m1lj4$9en$1@m2.c2.telstra-mm.net.au>

nekotron at pobox.com wrote in message <7lvs7i$fri$1 at nnrp1.deja.com>...

>The problem I'm having is that I see no way with Python to be able to
>get this message.  I can use RegisterWindowMessage to create the message

Or just use win32con.WM_USER

>and pass this to the program so it knows where to Post to, but I can't
>figure out how to get this message.

This message would be processed as part of the "message loop".  Python.exe,
pythonw.exe etc dont have one.  Tkinter does have one (thus python.exe,
pythonw.exe get one when using tkinter!).  I have no idea how you would
process a custom message, but Im sure it could be done.  Pythonwin can
definately do it, but I wont bother explaining until I know you need it :-)

Mark.





From gmcm at hypernet.com  Fri Jul 30 09:25:58 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 30 Jul 1999 08:25:58 -0500
Subject: os.name
In-Reply-To: <7nr6of$nc6$1@nntp6.u.washington.edu>
Message-ID: <1278785134-25350540@hypernet.com>

Donn Cave writes:

> Holger Jannsen <holger at phoenix-edv.netzservice.de> writes:

> | It's because I need to differ between Win95, Win98 and WinNT....
> 
> Not clear, missing parts of speech, however, os.uname()...  Windows
> behavior unknown.

Windows gives  an AttributeError.

os.environ['OS'] gives "Windows_NT" on my NT box. Maybe some Win9xers 
could try it on theirs...

- Gordon



From johanw at easics.be  Thu Jul  1 06:09:33 1999
From: johanw at easics.be (Johan Wouters)
Date: Thu, 01 Jul 1999 12:09:33 +0200
Subject: Does the first argument of a method *has to be* self?
References: <19990630222121.A1121@optiplex.palga.uucp>
Message-ID: <377B3E5D.F49993CA@easics.be>

Gerrit Holl wrote:
> 
> Hello,
> 
> I never saw code where the first argument of a method wasn't self. Is it
> in theory possible?

As one thinks about it, the first argument being self (the instance)
is a handy feature. Why would you want to put it somewhere else?

As the documentation states "self" is just some conventional name.
Renaming self into something else is probably a bad idea as it
will make your code less readable for others (and YOU as time passes!).

Experience shows that having conventions is a powerful way to
make life and code maintenance easier.

Kind regards,
===================================================================
Johan Wouters             ===              Easics               ===
ASIC Designer             ===  System-on-Chip design services   ===
Tel: +32-16-395 616          ===================================
Fax: +32-16-395 619      Interleuvenlaan 86, B-3001 Leuven, BELGIUM
mailto:johanw at easics.be           http://www.easics.com



From python-list at teleo.net  Sun Jul 11 03:29:43 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Sun, 11 Jul 1999 00:29:43 -0700
Subject: Dreams.
References: <000401becb50$8d3d9c80$479e2299@tim>
Message-ID: <9907110100560B.08196@quadra.teleo.net>

On Sat, 10 Jul 1999, Tim Peters wrote:

> I expect you could get most of this with little effort and little cost under
> Unix via fork and a little IPC:  fork the process and try the speculative
> stuff in the child, while the parent waits for the child to tell it how
> things turned out.  If the child is happy, it tells the parent to die; if
> the child has a nightmare, it kills itself after telling the parent "it was
> all a bad dream", and the parent goes on as if nothing had happened. 

This is rich with possibilities. By extension, one could model the
entire human collective unconscious, through a selective rereading of
Carl Jung's and Erich Neumann's work on dreams and archetypes. Write a
Python class for each of the plots of the pertinent Greek tragedies
(Oedipus, Phaedra, etc.). Maybe throw in King Lear for the situation
where the parent process hangs and doesn't listen to its child. And
Hamlet, of course, for when the parent dies prematurely and the child
encounters an undecidable problem. 

message-queues-and-event-flags-are-
	pretty-much-Greek-to-me-ly y'rs  - patrick




From gmcm at hypernet.com  Sun Jul  4 08:43:49 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 4 Jul 1999 07:43:49 -0500
Subject: Tkinter._test() problem
In-Reply-To: <377F38AD.C6CDBA6D@fnmail.com>
Message-ID: <1281033721-53904975@hypernet.com>

Per Kistler writes:
> 
> I installed py1.5.2 with tcl/tk.8.1 on SuSE6.1 Linux.

That's your problem. Either drop back to tcl/tk 8.0, or may try 
Dieter Maurer's patches, posted June 1.


- Gordon



From skip at mojam.com  Fri Jul 30 12:51:32 1999
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 30 Jul 1999 11:51:32 -0500 (CDT)
Subject: Python as a parser (newbie)
In-Reply-To: <37A1D114.705957B9@yahoo.com>
References: <37A1D114.705957B9@yahoo.com>
Message-ID: <14241.54707.277544.772154@dolphin.mojam.com>

    Ryan> I am creating a program that needs to have scripting support.
    Ryan> I've looked at the faqs and cannot get the answer.  Can I use
    Ryan> Python's scripting ability to parse the input I need to give it,
    Ryan> and execute c/c++ functions realtime in my code?

Yes.  There are many ways to parse input.  Which way will depend largely
upon your input.  You can often get a long way by simply using string.split
and rummaging around ad hoc in the tokens it returns.  Other more complex
parsing tasks can be handled by regular expressions (the re module).  If the
mustard is still too thick to cut, more powerful parsing systems are
available as well, such as Aaron Watters' kwParsing package
(http://www.chordate.com/kwParsing/) and John Aycock's parser generator
(http://www.foretec.com/python/workshops/1998-11/proceedings/papers/aycock-little/aycock-little.html).

As for calling C/C++ functions realtime, "that's what tiggers do best".  You
can manually or mostly automatically write wrappers for function or class
libraries that make your C/C++ APIs callable from Python code.  For more
information, check out the Python Extending and Embedding documentation
(http://www.python.org/doc/current/ext/ext.html) and the SWIG web site
(http://www.swig.org/).

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From kas at maps.magnetic-ink.dk  Sat Jul 17 05:06:42 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Sat, 17 Jul 1999 11:06:42 +0200
Subject: Snippet: bitcount()
References: <000601bed02a$2235af20$06a22299@tim>
Message-ID: <slrn7p0hsv.e38.kas@titan.magnetic-ink.dk>

Tim Peters <tim_one at email.msn.com> wrote:

> > -- unless I decide to count several bits at a time, but
> > I wouldn't know how to construct such a thing.
>
> Oh, sure you do <wink>:
>
> count[0] = 0
> count[1] = 1
> count[2] = 1
> count[3] = 2
> ...
> count[255] = 8
>
> Now you can get the popcount of any one-byte int with one table-lookup.
> And the popcount of any N-byte int or long with N table lookups and N-1
> adds.  Put 2**16 entries in the table to cut the work in half.

I get your point, thanks, but I will spare this group for a posting of such
a 2**16 table, leave alone the full 32-bit version. ;-)

What I missed in my original posting, however, was to inform the forum that
the method fails for signed ints, like e.g. 0x8000000 (use longs instead),
and that it is best suited for, as you put it, "dealing with sparsely
populated large bit vectors", something my example with the two big twin
primes shows clearly.


Cheers,

  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From steven_crazyman at worldnet.att.net  Sun Jul  4 19:50:41 1999
From: steven_crazyman at worldnet.att.net (Steve Wagner)
Date: Sun, 4 Jul 1999 18:50:41 -0500
Subject: Please send more info
Message-ID: <7los49$nga$1@bgtnsc03.worldnet.att.net>

I have a lot of spare time and would like to learn python. Please send info
on
how to get started. Also do you know of any online tutorials? thanks in
advance for your help.






From ralph.gauges at eml.villa-bosch.de  Thu Jul 29 10:37:33 1999
From: ralph.gauges at eml.villa-bosch.de (Ralph Gauges)
Date: Thu, 29 Jul 1999 16:37:33 +0200
Subject: Use of variables form outside the class.
References: <Pine.LNX.4.04.9907291736120.31968-100000@emerald.netskate.ru>
Message-ID: <37A0672D.C709C35C@eml.villa-bosch.de>

Oleg Broytmann wrote:
> 
> On Thu, 29 Jul 1999, Peter Torstenson wrote:
> > I declare a variable "x" right in the beginning of a module.
> > Later I declare a class with function defs.
> > I managed to declare varibles inside the class and address them by putting
> > "self." in front of the variable.
> > How do I reach the "x" variable that is outside any class from inside the
> > class functions?
> 
> x = 1
> 
> class X
>    def set_x(self):
>       self.x = 2
> 
>    def set_global_x(self):
>       global x
>       x = 2
> 
> > Peter Thorstenson
> > The Newbie
> 
> Oleg.
> ----
>     Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
>            Programmers don't die, they just GOSUB without RETURN.


I guess I should read more carefully. I thought you were
defining the variable inside the class, but outside the
methods.
Sorry for the wrong answer. Olegs answer is the right one.

Ralph



From wtanksle at dolphin.openprojects.net  Fri Jul  2 20:00:32 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Sat, 03 Jul 1999 00:00:32 GMT
Subject: Slicing beyond list --  bug or featurette?
References: <377D234A.4C9BDAC0@cisco.com> <377D3B92.A5109CB9@callware.com>
Message-ID: <slrn7nqkkv.3df.wtanksle@dolphin.openprojects.net>

On Fri, 2 Jul 1999 22:22:10 GMT, Ivan Van Laningham wrote:
>Hi All--

>Pierre Fortin wrote:

>> Just ran into a situation which can be summarized as follows:

>[bobbitt]

>> list[2:50] returns [ 2,3,4 ]
>> list[20:] or list[20:30] returns []

>> Can I count on this, or should I code try/except for the day it's

>I don't presume to speak for Guido, but the way slicing is implemented
>it seems likely that you can count on it: list[:] is implemented as if

Furthermore, this makes fundamental sense -- an array reference is defined
to return the item stored at that location.  A slice is defined to return
a list containing the items within those locations.

An empty list is still a list; an item that doesn't exist, however, is
hard to return.

><but-they-don't-call-him-guido-svetz-for-nothing>-ly y'rs,
>Ivan;-)

We-have-VAYS-of-justifying-vhat-we-vant-ly y'rs,

-- 
-William "Billy" Tanksley



From janssen at parc.xerox.com  Wed Jul 28 18:12:07 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Wed, 28 Jul 1999 15:12:07 PDT
Subject: NameError on "buffer"?
Message-ID: <orbs0r4B0KGWRZV19A@holmes.parc.xerox.com>

With Python 1.5.2, when I import a module, I sometimes get a NameError
on "buffer", which is a new, and apparently undocumented built-in
function.  Why would that happen?  Is the compiler perhaps not updated? 
Here's an example stack trace:

  File "<string>", line 1, in ?
  File "multlangimpl.py", line 25, in ?
    import multlang, multlang__skel, ilu, sys
  File "multlang.py", line 13, in ?
    import iluRt ; _imported_modules['iluRt'] = iluRt;
_imported_modules['ilu'] = iluRt;
  File "/var/tmp/usr/tmp/@12642.1/src/runtime/python/iluRt.py", line 409, in ?
    import traceback
  File "/import/python-1.5.2/lib/python1.5/traceback.py", line 6, in ?
    import types
  File "/import/python-1.5.2/lib/python1.5/types.py", line 18, in ?
    BufferType = type(buffer(''))
NameError: buffer

Bill



From hank at prosysplus.com  Thu Jul  1 16:17:28 1999
From: hank at prosysplus.com (Hank Fay)
Date: Thu, 1 Jul 1999 16:17:28 -0400
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net> <377B711A.E8B65E0B@cfdrc.com> <377BC620.F5553FF5@easystreet.com>
Message-ID: <b%Pe3.4033$lr3.164497@news2.jacksonville.net>

Is it possible to use ActiveX controls in any of these GUI's?  Then you've
got a large choice of controls.

Hank

--
http://www.prosysplus.com
Al Christians <achrist at easystreet.com> wrote in message
news:377BC620.F5553FF5 at easystreet.com...
> I looked at Fox and put it aside because I need a grid control.
> wxPython seems to be the winner for Windows in that category, even
> though the grid is pretty minimal.
>
> Any other candidates with a grid?
>
> Al
>
>
>
>
> Lyle Johnson wrote:
> >
> > I've already worked up the binary distribution of FXPy for Windows; I
just
> > forgot to mention it in the previous post. The binary installer can be
found
> > at:
> >
> >     ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42-win32.exe
> >
> > I haven't worked up a binary distribution of the (C++) FOX distribution
> > itself, primarily because it changes most every day :) But that is in
the
> > plans as well.
> >
> > Thanks,
> >
> > Lyle
> >
> > Mikael Lyngvig wrote:
> >
> > > On Wed, 30 Jun 1999 09:07:51 -0500, Lyle Johnson <jlj at cfdrc.com>
> > > wrote:
> > >
> > > >Just to add fuel to the fire, we now have a Python interface to FOX,
a
> > > >cross-platform GUI toolkit:
> > > >
> > > >    http://www.cfdrc.com/pub/FOX/fox.html
> > > >
> > > >The latest version of the Python wrappers can be picked up at:
> > > >
> > > >    ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42.tar.gz
> > > >
> > > >I don't know how the speed of FXPy would compare with some of the
others
> > > >mentioned (Tk, wxWindows, Qt, GTK, ...) but it would be worthwhile
for
> > > >some independent opinions.
> > >
> > > I'm downloading it now.  The http URL is wrong, it should be
> > > http://www.cfdrc.com/FOX/fox.html.
> > >
> > > The presentation seems promising, but I'd like to suggest a binary
> > > distribution for PC/Intel computers.  I know this is not common for
> > > Unix distributions, but it's the norm for PCs.
> > >
> > > -- Mikael





From janssen at parc.xerox.com  Mon Jul 12 12:50:24 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Mon, 12 Jul 1999 16:50:24 GMT
Subject: Updated GNU info documentation available
In-Reply-To: <14218.3197.296612.826083@weyr.cnri.reston.va.us>
References: <14218.3197.296612.826083@weyr.cnri.reston.va.us>
Message-ID: <ErWVnEkB0KGWB5tZJ8@holmes.parc.xerox.com>

Excerpts from ext.python: 12-Jul-99 Updated GNU info documentat.. "Fred
L. Drake"@cnri.res (306*)

>   Thanks to Milan Zamazal, the GNU info format of the documentation
> has been updated to documentation release 1.5.2p1.
>   An archive is available at ftp://ftp.python.org/pub/python/doc/.
>   Thanks, Milan!

I'd also like to thank Milan for this; I use it every day!

Bill



From mark at chem.uwa.edu.au  Tue Jul 27 04:17:36 1999
From: mark at chem.uwa.edu.au (Mark C Favas)
Date: 27 Jul 99 08:17:36 GMT
Subject: IDLE's PyShell
References: <000801bed335$8c233e80$a72d2399@tim> <379689C1.B91E0D13@compaq.com>
Message-ID: <mark.933063456@declan>

Greg Ewing <greg.ewing at compaq.com> writes:

>By the way, is there any way to create a window that
>doesn't have, for example, a close gadget? I think
>I've seen windows like that in other Windows apps,
>but I can't find any way of doing it in Tkinter.

Try

f = Tk.Toplevel()
f.transient(parent)

to get a window without the close button - /F has some documentation on this at
http://www.pythonware.com/library/tkinter/introduction/x8886-style-methods.htm

Mark
--
Email  - mark at chem.uwa.edu.au      ,-_|\                           Mark C Favas
Phone  - +61 9 380 3482           /     \               Department of Chemistry
Fax    - +61 9 380 1005      ---> *_,-._/   The University of Western Australia
                                       v                               Nedlands
Loc    - 31.97 S, 115.81 E                               Western Australia 6009



From graham at sloth.math.uga.edu  Fri Jul 16 11:34:13 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 16 Jul 1999 15:34:13 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com>
Message-ID: <7mnjdl$ne$1@cronkite.cc.uga.edu>

Neel Krishnaswami (neelk at brick.cswv.com) wrote:
: I think that this weekend I'm going to try and put together a basic
: version of this parsing system in Python based on the ideas in this
: paper -- basically just write a simple AST class and the combinators
: they describe. Then I can add proper support for backtracking and lazy
: evaluation and all the stuff with serious hack value when Christian
: Tismer makes continuations available to us. :)

It seems to me that you might also be able to make the regular expression
module available through this interface. All you have to do is wrap the
regular expression up in a function/object that can be combined like the
other parsers. Indeed you will need "primitive parsers" for things like
numbers, and these are really just a special case of a "primitive"
regular expression parser.

graham

-- 
              This ain't no technological breakdown
                 Oh no, this is the road to hell



From kernr at mail.ncifcrf.gov  Mon Jul 26 19:29:54 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Mon, 26 Jul 1999 23:29:54 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <p51zdvn4c7.fsf@bidra241.bbn.hp.com> <14236.32616.980578.536614@dolphin.mojam.com>
Message-ID: <379cef2e.16150509@news.erols.com>

On Mon, 26 Jul 1999 10:46:45 -0500 (CDT), Skip Montanaro
<skip at mojam.com> wrote:

>I got a bit confused attributing this message.  I think we have too many
>Markus's running around loose... ;-)  I think I have the quoted text
>attributed correctly.  Apologies if not.

For completeness's sake, you got them switched.

[snip]

>Skip Montanaro	| http://www.mojam.com/
>skip at mojam.com  | http://www.musi-cal.com/~skip/
>847-475-3758

nobody's-perfect'ly yours,

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From phd at emerald.netskate.ru  Thu Jul 29 09:37:25 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Thu, 29 Jul 1999 17:37:25 +0400 (MSD)
Subject: Use of variables form outside the class.
In-Reply-To: <7npi9i$2j12@enews3.newsguy.com>
Message-ID: <Pine.LNX.4.04.9907291736120.31968-100000@emerald.netskate.ru>

On Thu, 29 Jul 1999, Peter Torstenson wrote:
> I declare a variable "x" right in the beginning of a module.
> Later I declare a class with function defs.
> I managed to declare varibles inside the class and address them by putting
> "self." in front of the variable.
> How do I reach the "x" variable that is outside any class from inside the
> class functions?

x = 1

class X
   def set_x(self):
      self.x = 2

   def set_global_x(self):
      global x
      x = 2

> Peter Thorstenson
> The Newbie

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From mikael at pobox.com  Fri Jul  2 11:27:52 1999
From: mikael at pobox.com (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 15:27:52 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <slrn7nai1j.8ll.zeitlin@seth.lpthe.jussieu.fr> <64F6457BE0585BB3.1CF2F67BC8A680BD.A19A9FEE23BC9E58@lp.airnews.net> <7li9cq$inp$1@srv38.cas.org>
Message-ID: <3DD7210548F6BDE5.98B81058C5D96DBC.4CEC932949BB7F0A@lp.airnews.net>

On 2 Jul 1999 11:56:10 GMT, lvirden at cas.org wrote:

>When I run the Tk test suite  on my SPARC, Tk cranks thru creating and
>deleting windows as fast as the eye can track.

Well, that's good.  But I think the key word here is "as fast as the
eye can track".  That's the core of the problem.  A good GUI library
should be creating and deleting windows many times faster than any
human eye can track.

That's definitely not the case with Tk.  So far I think I've counted
20 or 21 different, intermediate windows, when I resize my main window
after having maximized it.  

>However, there are reports that the Windows version of Tk, because of its
>use of the native window widgets, is noticably slower than the X
>implementation.  Is that the platform to which you refer?

Yes, that's the platform to which I refer.  But, no, I don't believe
one minute that the native window widgets are the cause of the
performance problems.

If Tk was using the native widgets, it should be impossible for it, no
matter how not-good the design is, to be slow. Even a batch file (the
pathetic script language that's part of Windows by default) should be
able to draw any given native widget fast enough, since its just a
question of parsing a few, simple parameters and tossing them to some
native code.


-- Mikael




From greg.ewing at compaq.com  Wed Jul 28 16:51:15 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 29 Jul 1999 08:51:15 +1200
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com>
Message-ID: <379F6D43.3531FE0@compaq.com>

Gordon McMillan wrote:
> 
> But finding a match in a list, and testing whether you
> fell off the end without finding one, is (without the else clause) a
> much messier proposal.

Not if you go about it the right way. I always
put such loops in a procedure of their own,
and use return:

  def find_the_holy_grail(stuff):
    for x in stuff:
      if is_holy_grail(x):
        return x
    return None

Not only does this not require a weird else
clause, it doesn't require break either. I
almost never use break - and when I do I usually
regret it later. I wouldn't mind if there weren't
any break - and without break, there would be no
need for the weird else.

Greg



From meowing at banet.net  Wed Jul 28 01:16:45 1999
From: meowing at banet.net (gja)
Date: 28 Jul 1999 05:16:45 +0000
Subject: threading question
References: <199907280430.XAA17615@buffalo.fnal.gov>
Message-ID: <mW.FFKEnx.7nr@whizzo.com.invalid>

Charles G Waldman wrote:

> When I run the trivial example attached below, which simply
> starts a daemon thread that does nothing, and then type "ps",
> I see *three* python processes.  I can understand two, but
> why three?

> Platform is Linux 2.2.10 with the most recent glibc (2.1.1),
> if that matters.

What you're seeing is a quirk of the Linux threading
implementation.  The extra process slot is used by the
threading library itself, to keep tabs on all your "real"
threads.

For a more details, see
<URL:http://pauillac.inria.fr/~xleroy/linuxthreads/faq.html>



From mikael at pobox.com  Fri Jul  2 13:57:00 1999
From: mikael at pobox.com (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 17:57:00 GMT
Subject: Python GUIs: Summary and Conclusion
References: <99FA4A1A1280BEBD.0A1ABA62CC989479.4C2C6022F695BEAC@lp.airnews.net> <19990629092757.A4998@optiplex.palga.uucp> <46F9FB3A1847A014.2CBD270176138613.20F84A39E6ED95DA@lp.airnews.net> <b_Pe3.558$6M6.167478@news.shore.net> <395C236424AEEAA7.C2D70DEDE2306228.0C4BC0DFFEC49C2C@lp.airnews.net> <slrn7nprcu.3a4.wtanksle@dolphin.openprojects.net>
Message-ID: <072D7417DC5C0BDA.0C10D861CC522F34.D6409B318AE170AF@lp.airnews.net>

Hi William,

On Fri, 02 Jul 1999 16:49:34 GMT, wtanksle at dolphin.openprojects.net
(William Tanksley) wrote:

>I'm working on a 2D tile-based adventure game, which right now is text
>only.  I'm expanding it so that it can use graphics, but I need some
>libraries.  Let me know what you find!

Well, I don't expect to find anything overly suitable for 2D
tile-based adventure games.  I stopped working on that project three
years ago.

However, it's kind of funny you write just now: I was considering to
resume that old project in Python using wxWindows earlier today.

I don't know if wxWindows is what you need, but that's probably where
I would be looking, if I was to resume my old project.

>>[1] Which I expect to change in the future, which is why I'm reviewing
>>portable GUIs (I don't do non-portable stuff, if there's anything I
>>can do to avoid it).
>
>That's right.

You said it :)  That pretty much leaves you with the choice between
TkInter, wxWindows, and FOX (from what I know).  I've looked, more or
less, briefly on all of them and decided to use wxWindows.  

For a 2D tile-based game you pretty much just need to be able to draw
tiles (square bitmaps) quickly.  I think wxWindows can do that, but
perhaps you should make some serious tests of those three, and the
others recommended by others.  

And let us know what you find :)


-- Mikael




From sreekantk at freenet.co.uk  Sun Jul 11 21:31:53 1999
From: sreekantk at freenet.co.uk (Sreekant Kodela)
Date: Mon, 12 Jul 1999 02:31:53 +0100
Subject: Help creating new user Please
Message-ID: <37894589.429E2F3E@freenet.co.uk>

Dear all

My problem is I am trying to add a new user through Python[using bog
standard adduser.] I can create a user
without any problem but when I give a command line option of '-p
passwordstr' to it, and later try to login
with that name and password it refuses saying incorrect password. I
heard that you have to give an encrypted
password on the command line. Is it true? If so how can I create a user
with particalar password through a
script.

Any ideas and help is much appreciated.

Thanks
sreekant




From cfelling at iae.nl  Wed Jul 21 20:09:30 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 22 Jul 1999 02:09:30 +0200
Subject: python + bash + new user
References: <3793BE52.43E023BB@netcard.com> <7n0f3d$1d7@dfw-ixnews10.ix.netcom.com> <3793C729.55DE42D0@netcard.com>
Message-ID: <7n5nfq$j0$1@vvs.superst.iae.nl>

Joyce Stack <joyce at netcard.com> wrote:
> Sorry about being vague...i meant the Bash prompt...i have gotten other simple
> programs to work...changed mode and executed the programs ...but i dont know how to
> run the program and pass a variable at the same time...

Gerrit showed you how to pass a command-line-script to python, and I will
show how to pass parameters to your script, you can ofcourse combine the two:)

	$ cat simple.py
	#!/usr/bin/python
	from sys import argv
	print "I'll show you my parameters:"
	for p in argv:
		print p

	$ ./simple.py one parameter and others
	./simple.py
	one
	parameter
	and
	others

hope this helps

-- 
groetjes, carel



From arcege at shore.net  Thu Jul 15 10:14:12 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 15 Jul 1999 14:14:12 GMT
Subject: bytecode & embedding
References: <378c9c8c.0@hedwik.mik.net>
Message-ID: <U0mj3.68$ui4.22553@news.shore.net>

Jens Calisti <Jens.Calisti at eiskonzept.com> wrote:
: Hi all,

: I am new to python so maybe someone can answer me this :

: Is it possbile to exeute python-compiled-bytecode from a c-programm with
: embedded python ?? (eval,exec ?)

I'm not sure how "supported" this is in Python 1.6 and beyond, but you
need to include <compile.h> and <eval.h> (after <Python.h>).

  #include <Python.h>
  #include <compile.h>
  #include <eval.h>

  ...

    PyObject *globals, *locals, *result;
    PyCodeObject *bytecode;

    result = PyEval_EvalCode(bytecode, globals, locals);

The globals and locals should be dictionaries (possibly taken from
PyModule_GetDict()).

  -Arcege



From c.evans at clear.net.nz  Sat Jul  3 18:24:25 1999
From: c.evans at clear.net.nz (Carey Evans)
Date: 04 Jul 1999 10:24:25 +1200
Subject: Python GUIs: Summary and Conclusion
References: <99FA4A1A1280BEBD.0A1ABA62CC989479.4C2C6022F695BEAC@lp.airnews.net> <MTOd3.139$6M6.44310@news.shore.net> <377845DD.23AD193A@compaq.com>
Message-ID: <87emipnzhi.fsf@psyche.evansnet>

Greg Ewing <greg.ewing at compaq.com> writes:

> I'd certainly be in favour of a native
> Tk interface, especially if it became the 
> "standard" one included in the distribution.
> Even if it wasn't any faster, I'd feel
> better while using it!

There's another place this has been done, that I haven't seen
mentioned.  Perl/Tk has no dependencies on Tcl - it's source is
generated by a collection of Perl scripts that work on the distributed
Tcl C source and convert it to XS, and all the Tcl parts have been
reimplemented in Perl.

It might be reasonable to do the same for Python.

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

			 print [<>]->[rand$.]



From tismer at appliedbiometrics.com  Tue Jul 13 11:54:16 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 13 Jul 1999 15:54:16 GMT
Subject: ANN: Stackless Python 0.3
References: <378A40AD.B4FDB79F@appliedbiometrics.com>
		<m37lo5kpd0.fsf@atrus.jesus.cam.ac.uk>
		<378AFD45.8793D507@appliedbiometrics.com> <TaIi3.1701$6M6.569174@news.shore.net>
Message-ID: <378B6128.DFA74F05@appliedbiometrics.com>


"Michael P. Reilly" wrote:
> 
> Christian Tismer <tismer at appliedbiometrics.com> wrote:
> 
> : Here is the patch.
> : It was one of those "=" instead of "==" bugs.
> 
> And newbies wonder why Python doesn't have assignment expressions. ;)

You say it. And by default, MSVC doesn't consider this worth a warning.

I was tempted, but decided to leave this remark to you :-)

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From tseaver at starbase.neosoft.com  Wed Jul 28 21:31:58 1999
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 29 Jul 1999 01:31:58 GMT
Subject: Problem with signals & exceptions...
References: <2BB05D8530326D0C.420888E8906528AF.C6BF325D12E9ED90@lp.airnews.net> <7nnbme$ns2$1@nntp6.u.washington.edu> <DA4B3986E94F79C9.E3C10FB44F334D06.D3443001EBA1A2FA@lp.airnews.net>
Message-ID: <8BA4CD0D8870A635.6A95A1686177812D.BB87D7873F982A86@lp.airnews.net>

In article <DA4B3986E94F79C9.E3C10FB44F334D06.D3443001EBA1A2FA at lp.airnews.net>,
Eric Lee Green  <eric at estinc.com> wrote:
>Donn Cave wrote:
>> 
>> Eric Lee Green <e_l_green at hotmail.com> writes:
>> | I'm having trouble catching my thrown exceptions under FreeBSD 3.2. I am
>> | using the Python 1.5.2 that came with FreeBSD 3.2 (well, on disk 1 of
>> | the <n>-disk set). The following program works properly under Red Hat
>> | Linux version 6.0 running a self-compiled Python 1.5.2 (in place of the
>> | 1.5.1 which comes with Red Hat 6.0, which does not properly handle
>> | signals during socket i/o calls). How it works: it listens to port 6664.
>> | When it gets a connection, it then sets an alarm handler and alarm and
>> | then tries to read from the socket. If it does not get data within 10
>> | seconds, it then times out, at which point the alarm handler throws an
>> | exception and aborts the socket read.
>> 
>> I tried this exception-from-handler notion on Digital UNIX and BeOS,
>> and it does seem to work the way I guess you expect, but I do find it
>> weird. A signal handler executes outside the normal flow of control,
>> so it's not obvious to me where its exceptions ought to surface.
>> Probably it should be expected to work the same way on various platforms
>> for any given version of Python, but would future versions of Python be
>> guaranteed to work this way also?
>
>That, of course, is the ten million dollar question. The operation of signals
>in "C" is extremely well documented and described. Under Python, it's ???. 

About the _only_ thing one can both portably and safely do in a signal handler
(in C) is to set a global of type sig_atomic_t.  Anything else leaves one open
to the vagaries of the libc / compiler implementor's tortured imagination.

>
>I changed my code to set a global variable rather than throw an exception and
>that works the way you'd expect. A socket.error exception is thrown, and I
>catch it, check the global variable, if the global variable is set I print "got
>a timeout!" else re-throw the exception. 
>
>In the actual working code I scrapped the alarm signal altogether and instead
>do a select() call with a timeout, which presumably is guaranteed to work the
>same on all platforms (it better, the select call is a POSIX standard!). 

If you think about it, signals are not even guaranteed to share the stack with
the code they interrupt, so how can poor python unwind said beast if you try
to throw an exception from one?  Oh, well, at least you ended up in the right
place :).
-- 
---------------------------------------------------------------
Tres Seaver           tseaver at palladion.com       713-523-6582
Palladion Software    http://www.palladion.com



From sprout at dok.org  Tue Jul 27 21:31:37 1999
From: sprout at dok.org (Chris Green)
Date: Wed, 28 Jul 1999 01:31:37 GMT
Subject: newbie namespace question
References: <m3u2qp9981.fsf@ghostplanet.dok.org> <7nlet7$pj2@dfw-ixnews10.ix.netcom.com>
Message-ID: <m3g1298sr7.fsf@ghostplanet.dok.org>

aahz at netcom.com (Aahz Maruch) writes:

> >
> >Ok. Excuse the bad example but when select_clist is called, how can I
> >access the previous self so that I can use self.grok_data().
> 
> If m_clist.connect() is *not* passing "self" to select_clist(), I don't
> think there's anything you can do.

Thanks - I figured out ( and messed up my followup to c.l.p ) the name 
space rules of how functions don't share the name space as global of
their enclosing functions. 

the proper solution for me seems to move everything into another
class, or use a global in the timecalc class. 

I found the answer on page 147 of programming python once I got home.

Cheers,
Chris
-- 
Chris Green <sprout at dok.org> <grapeape at uab.edu>
Laugh and the world laughs with you, snore and you sleep alone.



From ntaylor at newbridge.com  Thu Jul 22 15:59:38 1999
From: ntaylor at newbridge.com (Nick Taylor)
Date: Thu, 22 Jul 1999 15:59:38 -0400
Subject: Printable Python reference
Message-ID: <3797782A.FD39EF04@newbridge.com>

Hi,

I decided to take a look at Python today for the first time, and it
sounds very interesting.  Does anyone know of a printable (i.e. PDF,
PostScript, or plain-text) reference out there?  I prefer hard copy to
online docs.

thanks in advance,

Nick
http://www.undergrad.math.uwaterloo.ca/~n2taylor/




From Aahz.Maruch at p98.f112.n480.z2.fidonet.org  Thu Jul  1 21:07:28 1999
From: Aahz.Maruch at p98.f112.n480.z2.fidonet.org (Aahz Maruch)
Date: Fri, 02 Jul 1999 02:07:28 +0100
Subject: please help with simple regex
Message-ID: <00001648@bossar.com.pl>

From: aahz at netcom.com (Aahz Maruch)

In article <7lh1h6$t84$1 at nnrp1.deja.com>,  <viofis at my-deja.com> wrote:
>
>I have a string, for example, myString="the Potato".
>1. Convert it to all lowercase ("the potato")

import string
myString = string.lower(myString)

>2. Match it against a list of strings - "potato,tomato,corn,etc...",
>extract "potato" (if it matches with something on the list), then save
>into myNewString.

If you mean literally a list of strings, here's one approach:

stringList = [ 'potato', 'tomato', 'corn' ]
if myString in stringList :
	myNewString = myString

Of course, a better (more efficient) approach equivalent to above is

stringDict = { 'potato':None, 'tomato':None, 'corn':None }
if stringDict.has_key(myString) :
	myNewString = myString

(I'm pretty sure that's not what you want, but Python's regex is a much
smaller hammer than in Perl, and you should get used to trying out other
approaches.)
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From wolford at enews.nrl.navy.mil  Thu Jul  8 12:54:20 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Thu, 08 Jul 1999 12:54:20 -0400
Subject: Efficient string manipulation , Help please
References: <37856BF1.BD046003@freenet.co.uk>
Message-ID: <3784D7BC.3990816F@enews.nrl.navy.mil>

If you just want to remove the ' and " characters no matter what the user
intended, then take a look at the re regular expression package. Something
like:

import re
awkstr = 'awkward""st\'ring'
quote_re = re.compile('["\']')
print awkstr
awkstr = quote_re.sub('', awkstr)
print awkstr


Sreekant Kodela wrote:

> Dear all
>
> I seem to be in trouble here. I am trying to make some cgi scripts for
> my web site. I am using the postgresql as the backend db and a form
> based address book and news forum etc. The problem is everything is
> working ok as long as noone uses ' and " in the text. If they do, bingo
> , postgres refuses to accept. I tried using string library [mainly
> string.join(string.split('awkward'"'st'ring')). It works sometimes and
> doesn't some times.
>
> Any better and efficient way to deal with this situation !
>
> Any suggestions and ideas will be greatly appreciated.
>
> Thanks
> sreekant




From rgh at inmet.com  Tue Jul 20 17:57:39 1999
From: rgh at inmet.com (Randy Hudson)
Date: Tue, 20 Jul 1999 21:57:39 GMT
Subject: Jpython Question
References: <Pine.GSO.4.05.9907191525580.20433-100000@allegro>
Message-ID: <3794F0D3.ACAD7C6B@inmet.com>

Ian Kezsbom wrote:

> I was just wondering if jpython can instantiate java classes and invokes
> methods on them...any help would be welcome...

Yes. Furthermore, you can extend Java classes in JPython, and override and
add methods, or implement
Java interfaces. You can pass back objects of these JPython classes to Java methods
that accept arguments of the base type. It's a remarkably well-done two-way
integration.

-- Randy Hudson



From emile at fenx.com  Fri Jul 30 20:59:02 1999
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 30 Jul 1999 17:59:02 -0700
Subject: os.name
References: <1278785134-25350540@hypernet.com> <V5ho3.4340$VL2.463886@news.direcpc.com> <001e01beda8e$2111d2e0$f29b12c2@secret.pythonware.com>
Message-ID: <rWro3.4356$VL2.466946@news.direcpc.com>

<oops...>

Well, now both 95 and 98 say:

>>> import os
>>> os.environ['OS']
Traceback (innermost last):
  File "<pyshell#1>", line 1, in ?
    os.environ['OS']
  File "D:\Program Files\Python\Lib\os.py", line 247, in __getitem__
    return self.data[string.upper(key)]
KeyError: OS
>>>

I've-got-to-get-a-bigger-monitor-for-1600x1200-ly yr's

--

Emile van Sebille
emile at fenx.com
-------------------


Fredrik Lundh <fredrik at pythonware.com> wrote in message
news:001e01beda8e$2111d2e0$f29b12c2 at secret.pythonware.com...
> Emile van Sebille <emile at fenx.com> wrote:
> > >>> os.name
> > 'nt'
> > >>> os.environ('OS')
> > Traceback (innermost last):
> >   File "<pyshell#8>", line 1, in ?
> >     os.environ('OS')
> > AttributeError: no __call__ method defined
>
> Gordon asked you to try os.environ['OS'],
> not os.environ('OS') ;-)
>
> </F>
>





From thomas at xs4all.nl  Mon Jul  5 11:18:20 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Mon, 5 Jul 1999 17:18:20 +0200
Subject: CGI under NT
In-Reply-To: <3780ca11@news2.isys.net>; from dont@try.it on Mon, Jul 05, 1999 at 05:06:54PM +0200
References: <3780ca11@news2.isys.net>
Message-ID: <19990705171820.L26497@xs4all.nl>

On Mon, Jul 05, 1999 at 05:06:54PM +0200, dont at try.it wrote:


> Until now I wrote a script which takes alle parameters of the forms out of
> "QUERY_STRING". There is my problem. If I press the submit-button I get a
> Message that the script has problems to read the system-variable (because it
> is not there!), but if I call the script with the parameters directly
> (..../script.py?value=100&browser=netscape etc.) the script works.

> Can anyone helps me or give me a hint to more information on the web?

Use <FORM METHOD=GET ...> instead of METHOD=POST, or use a CGI module that
parses both POST and GET and returns them on an object. (Haven't worked with
CGI under Python yet, but I'm fairly certain there is such a module.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From hinsen at cnrs-orleans.fr  Wed Jul  7 04:16:09 1999
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 07 Jul 1999 10:16:09 +0200
Subject: Slicing beyond list --  bug or featurette?
References: <377D234A.4C9BDAC0@cisco.com>
Message-ID: <m3673wub7a.fsf@chinon.cnrs-orleans.fr>

Pierre Fortin <pfortin at cisco.com> writes:

> list[2:50] returns [ 2,3,4 ]
> list[20:] or list[20:30] returns []
>  
> Can I count on this, or should I code try/except for the day it's
> fixed?  This 'feature' reduces the error situations; but is it
> correct/intended?

Here's what the language reference manual says:

  Sequences also support slicing: a[i:j]selects all items with index k
  such that i <=k < j.

There is no mention of an error condition, and the description agrees
with the behaviour of the interpreter, so I'd consider it a feature.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------



From daniels at mindspring.com  Thu Jul  1 22:43:41 1999
From: daniels at mindspring.com (Alan Daniels)
Date: 2 Jul 1999 02:43:41 GMT
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain> <377B7E17.5BD7@nospam.com>
Message-ID: <slrn7noa2j.hg.daniels@localhost.localdomain>

On Thu, 01 Jul 1999 15:41:27 +0100, the infinitely wise Alex Maranda
(amaranda at nospam.com) spoke forth to us, saying...

[snip...]
>you use confusing terms; I understand what you're saying, but you
>shouldn't use the word 'thread' for anything else than its OS meaning.
>You could say the 'the bulk of processing'. You might very well end up
>using real Python threads; don't forget in this case to release/reaquire
>the global interpreter lock in your C++ callbacks, if you make blocking
>syscalls (unlikely in a game).

It was pretty late when I was writing the first post. :=)
You are correct, "the bulk of processing" is what I meant. NO threads
will be involved. The only justification I can see is better CPU
utilization on an SMP machine, and I don't happen to have one of those
lying around. Hence the engine will run in a single thread.

To keep things moving along, I'm considering dealing with the Python
scripts in a round-robin approach. Every frame, I allow X number of
scripts to handle an event, bumping that number upwards or downwards
to keep a consistent framerate.

>Unlikely. You won't know until you try. To my knowledge UnrealScript is
>eventually translated to C++.

I definitely do NOT want to turn this into a systems level project.
I'm a language junkie, but the internals of compilers don't interest
me much. I understand the reasons that Tim Sweeney made UnrealScript,
but to me it seemed kind of overkill. The world doesn't need another
scripting language.

I have the (VERY!) preliminary code at:
    http://www.mindspring.com/~daniels/engine/index.html
if you're interested, although there isn't much to look at.

-- 
=======================
Alan Daniels
daniels at mindspring.com
daniels at cc.gatech.edu



From robin at jessikat.demon.co.uk  Sun Jul 18 17:02:17 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sun, 18 Jul 1999 22:02:17 +0100
Subject: Python and Boehm-Demers GC, I have code.
References: <o+zLoAAp4ak3Ewjk@jessikat.demon.co.uk>
 <000101bed15c$c5685ca0$e7a22299@tim>
Message-ID: <TAPlnDAZDkk3EwmS@jessikat.demon.co.uk>

In article <000101bed15c$c5685ca0$e7a22299 at tim>, Tim Peters
<tim_one at email.msn.com> writes
>
...
>Not from me -- I don't even know how to spell Xoap.
>
>> Or the (I'm sure all the features are good) version of plumbo?
>
>It's Cyclops.py, and I just uploaded it to python.org's ftp contrib site.
>It should show up in the System directory later this week (send a complaint
>every 5 minutes  to ftpmaster at python.org until it does <wink>).
...
thanks
-- 
Ulysses-ly yrs Robin Becker



From greg.ewing at compaq.com  Tue Jul  6 17:58:10 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 07 Jul 1999 09:58:10 +1200
Subject: Inserting empty space into an existing file/pointer
References: <7ltn1m$nen$1@snatch.hq.eso.org>
Message-ID: <37827BF2.BAB43AAE@compaq.com>

NO_SPAMnpirzkal at hq.eso.org wrote:
> 
> I have the need to insert some empty space inside an already existing binary
> file. Is there a convenient, quick way of doin gthis with Python?

The simplest way is to read everything after the
insertion point into memory, and then write it back
further along in the file, e.g.:

f = open("myfile", "rb+")
f.seek(2047*1024)
data = f.read()
f.seek(2147*1024)
f.write(data)
f.close()

If there's too much data to fit in memory, you
could start at the end of the file and work
backwards, reading and writing suitable
sized chunks, but it would be messy to write.

Greg



From emile at fenx.com  Mon Jul 19 22:38:24 1999
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 19 Jul 1999 19:38:24 -0700
Subject: How to send text to a Windows program
References: <37932FF9.47E8FA2D@wiu.edu>
Message-ID: <AkRk3.3777$VL2.410197@news.direcpc.com>

Take a look at
http://x36.deja.com/getdoc.xp?AN=480302853&search=thread&CONTEXT=9324382
39.909574156&HIT_CONTEXT=932438107.906297455&HIT_NUM=1&hitnum=5

(found on dejanews by searching for 'python sendkeys' in all.

--

Emile van Sebille
emile at fenx.com
-------------------


Mark Clark <mumac at wiu.edu> wrote in message
news:37932FF9.47E8FA2D at wiu.edu...
> I am starting a Windows program with ShellExecute.  I would like to
> send some text to this program, similar to the way you would if you
used
> "SendKeys" in other programming languages.  Can someone give me an
> example of how this would be done?
> Thanks for your time.





From skodela at photon.sindh.com  Sun Jul 25 19:41:04 1999
From: skodela at photon.sindh.com (skodela at photon.sindh.com)
Date: 26 Jul 1999 00:41:04 +0100
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <0EAA6437F8435A58.6F61DA1F1117B965.CAF9F552BC9AB119@lp.airnews.net>
Message-ID: <m3d7xgcn73.fsf@photon.sindh.com>

PostgreSQL is on the surface looks slower than MySQL. Then again PG provides a lot 
more than the other. It is down to what one needs and also re: licences I think.

Thanks
sreekant



From phd at emerald.netskate.ru  Tue Jul 27 08:34:27 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Tue, 27 Jul 1999 16:34:27 +0400 (MSD)
Subject: 1.5.2 for: else:
In-Reply-To: <19990727142943.V4068@xs4all.nl>
Message-ID: <Pine.LNX.4.04.9907271632450.9849-100000@emerald.netskate.ru>

On Tue, 27 Jul 1999, Thomas Wouters wrote:
> http://www.python.org/doc/current/tut/node6.html#SECTION006200000000000000000
> 
> [..]
> 4.4 break and continue Statements, and else Clauses on Loops 
> 
> Loop statements may have an else clause; it is executed when the loop
> terminates through exhaustion of the list (with for) or when the condition
> becomes false (with while), but not when the loop is terminated by a break
> statement. 
> [..]
[skip]
> Else clauses get executed on normal loop exit, even if the loop was 'empty'.
> It does not get executed when you 'break' out of a loop.

   Aha, thanks. I expected "else" to be executed when "for" not executed on
empty list. Wrong assumption.

> -- 
> Thomas Wouters <thomas at xs4all.net>
> 
> Hi! I'm a .signature virus! copy me into your .signature file to help me spread!

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From fdrake at cnri.reston.va.us  Thu Jul  1 16:33:14 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Thu, 1 Jul 1999 20:33:14 GMT
Subject: print color strings?
In-Reply-To: <19990701213528.A8877@optiplex.palga.uucp>
References: <19990701213528.A8877@optiplex.palga.uucp>
Message-ID: <14203.53386.710688.87678@weyr.cnri.reston.va.us>

Gerrit Holl writes:
 > is there an easy way to print color strings in python?
 > 
 > A function like printcolor('brightyellow', 'red', 'spam'), that finds
 > out on what kind of terminal you are and prints it to stdout?

Gerrit,
  If you're doing this on Unix, take a look at the ncurses library;
there's a Python interface to that.  (I think Oliver Andrich did it;
searching DejaNews should find it fairly quickly since this keeps
coming up.)
  The standard curses module will be documented in the next release of 
the Library Reference.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From rjroyNoSpam at magma.ca  Sat Jul 31 00:17:52 1999
From: rjroyNoSpam at magma.ca (Robert J. Roy)
Date: Sat, 31 Jul 1999 04:17:52 GMT
Subject: Import a lot of modules, or, import 'foo'
References: <874silabe7.fsf@linuxcare.com>
Message-ID: <37a27933.264187381@news1.on.sympatico.ca>

On 30 Jul 1999 17:52:32 -0700, dwelton at linuxcare.com (David N. Welton)
wrote:

>
>So, basically, I'd like to do this:
>
> tests = ['fs', 'swap']
>
> for t in tests:
>     import t
>
> do more stuff on what promises to be a lot of modules
>
>I've looked around a bit, but can't find what I need.  Maybe I'm just
>going about this the wrong way, but... Hmm, now I want to know how to
>do this in any case:-)
>
>Thanks,
>-- 
>David N. Welton, Web Engineer, Linuxcare, Inc.
>415.354.4878 x241 tel, 415.701.7457 fax
>dwelton at linuxcare.com, http://www.linuxcare.com/
>Linuxcare. At the center of Linux.

you can use the builtin function __import__

See 
http://www.pythonware.com/people/fredrik/fyi/fyi06.htm
for more info



From schorsch at schorsch.com  Fri Jul  2 05:32:48 1999
From: schorsch at schorsch.com (Georg Mischler)
Date: Fri, 02 Jul 1999 09:32:48 GMT
Subject: Python GUIs: Portability or not?
References: <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net> <377B711A.E8B65E0B@cfdrc.com> <377BC620.F5553FF5@easystreet.com> <b%Pe3.4033$lr3.164497@news2.jacksonville.net> <14204.4556.177837.851763@lrz.de> <Pine.LNX.4.10.9907012309060.6263-100000@us2.mayo.edu> <5DEB4D7CBF56EC8D.37F29301ECE96FC0.D3C7C2F2D9068D65@lp.airnews.net>
Message-ID: <7li101$7fg$1@nnrp1.deja.com>

Mikael Lyngvig wrote:
> Travis Oliphant wrote:
>
>[...]
> >So, I looked in to PyGtk and PyGnome (with its Gnome Canvas) and I'm
> >impressed.  Right now it fits what I want to do very well and it
> >works on systems with an X server (aren't X servers available for
> >MAC's and Windows?)
>
> I don't know about the Mac, but I'm sure there are X servers available
> for NT.  However, not like they are available for Linux and other Unix
> boxes - you don't just download the X11 source, compile it, and that's
> it (unless a lot has changed the last few years).
>
> You'd probably have to buy a commercial X server for NT or, even
> worse, ask your users to do so.


Have a look at  http://www.microimages.com/www/html/freestuf/mix/.
This is a free X sever for both the mac and NT. It works pretty well
for what it is. But if you have a choice, and assuming that you plan
to work with X and non-X applications at the same time, then you
probably want a native GUI for your application.

There are other X servers around, but those who blend into the
host GUI (instead of creating their own X desktop within a window)
all seem to be commercial.

Now *this* would be a task for an open software project... ;)


Have fun!

-schorsch

--
Georg Mischler  --  simulation developper  --  schorsch at schorsch.com
+schorsch.com+  --  lighting design tools  --  http://www.schorsch.com/


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From pmoran at nas.nasa.gov  Mon Jul 12 19:26:05 1999
From: pmoran at nas.nasa.gov (Patrick J. Moran)
Date: 12 Jul 1999 23:26:05 GMT
Subject: bug with Irix configure
References: <7mdkm1$4c3$1@sun500.nas.nasa.gov> <378A6569.A5CEDC4E@bioreason.com>
Message-ID: <7mdtid$c2a$1@sun500.nas.nasa.gov>

In article <378A6569.A5CEDC4E at bioreason.com>, Andrew Dalke <dalke at bioreason.com> writes:
|> Patrick J. Moran <pmoran at nas.nasa.gov> said:
|> > I've encounted a bug with the configure on Irix.  I'm not sure
|> > where to report it, so to the newsgroup it goes.

|> ... helpful suggestions ...
|> 
|> So either the problem is limited to IRIX 6.4 or something else
|> is wrong.
|> 
|> 						Andrew
|> 						dalke at bioreason.com



I've learned a little more.  As you probably know, the compilers
check a license server each time they are executed, and print out
about a page worth of warning if they cannot find their license.
One can set the environment variable LM_LICENSE_FILE to tell the
compiler where to find the server.  Other than the annoying warning,
the compiler goes ahead as usual.  Or at least so I thought.
I've now tried running configure with and without the environment
variable set.  With the compiler able to find its license server,
configure figures out that it has size_t and so on, and doesn't
try to define replacements.  When the compiler can't find its
license, I get the problems I described previously.  Yuck.

Thanks for the response, sorry to bother you.

Pat

------------------------------------------------------------
Patrick J. Moran, Ph.D.      http://www.nas.nasa.gov/~pmoran
MRJ Technology Solutions           NASA Ames Research Center




From bradh at mediaone.net  Thu Jul  8 20:25:50 1999
From: bradh at mediaone.net (Brad Howes)
Date: 08 Jul 1999 20:25:50 -0400
Subject: interaction between modules... globals?
References: <19990707100532.A951@optiplex.palga.uucp>
Message-ID: <wjyg12yhdo1.fsf@bradh.ne.mediaone.net>

Gerrit Holl <gerrit.holl at pobox.com> writes:

> I have a program which consists of several little modules, but I
> encounter a problem with that. I have one module which returns a
> class instance, with messages as members, either Dutch or
> English. Every module needs that instance, but I don't want to pass
> it as an argument to every little function, the modules should find
> out theirselves. But I _also_ don't want to make that instance 5
> times.
> 
> So I thougt: I need a sort of global between the modules. Is this possible?

This is what I've done before. Plop this at the bottom of the file
that define your class (`Singleton' below) and just import this module
everywhere else.

import __main__
if not hasattr( __main__.__builtins__, 'gMe' ):
    __main__.__builtins__.gMe = Singleton()





From david at sundayta.co.uk  Mon Jul 19 05:05:43 1999
From: david at sundayta.co.uk (David Warnock)
Date: Mon, 19 Jul 1999 10:05:43 +0100
Subject: IDLE Keyboard Mappings
Message-ID: <3792EA67.B97AEB73@sundayta.co.uk>

Hi,

IDLE has different keyboard mappings for Unix and Windows eg Cut
Selection is Ctrl-w on Unix and Ctrl-x on Windows. Many commands on Unix
are ctrl-x ctrl-something eg Save, New Window etc.

I will be using IDLE on both Linux and Windows (mostly Linux) and would
like to be able to have a single keybaord mapping. 

Looking at the source code it appears I could simply change Bingings.py
from 

if sys.platform == 'win32':
    default_keydefs = windows_keydefs
else:
    default_keydefs = unix_keydefs

to simply hard code the defaults to my preferred settings (probably
windows as it uses less keystrokes) regardless of platform.

This does not seem a very good solution. Where is the correct place to
comment on IDLE? I would prefer a solution which allowed the default
mappings to be selected by a configuration file or which provided a menu
option to select them. As python is such a nice language I might even be
able to make the changes myself (despite having only just started with
python).

Thanks

Dave


-- 
David Warnock
Sundayta Ltd



From Andrew.Csillag at p98.f112.n480.z2.fidonet.org  Thu Jul  1 11:05:30 1999
From: Andrew.Csillag at p98.f112.n480.z2.fidonet.org (Andrew Csillag)
Date: Thu, 01 Jul 1999 16:05:30 +0100
Subject: coverage tool for Python?
Message-ID: <00001619@bossar.com.pl>

From: Andrew Csillag <andrew at starmedia.net>

Ovidiu Predescu wrote:
> 
> Hi,
> 
> Are there any coverage tools for Python?
> 
> Thanks,
> --
> Ovidiu Predescu <ovidiu at cup.hp.com>
> http://www.geocities.com/SiliconValley/Monitor/7464/
Theres pycover also at
ftp://ftp.python.org/pub/python/contrib/All/pycover-0.1.tar.gz which,
while not all that great, should give you some good stuff.
-- 
Writing perl programs is like pounding stakes into your feet:
You could if you had to, but would be better off if you didn't.



From tismer at appliedbiometrics.com  Thu Jul 15 10:31:49 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Thu, 15 Jul 1999 14:31:49 GMT
Subject: Meta-language for mxTextTools
References: <08bb01beceb7$334a58c0$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <378DF0D5.D679875@appliedbiometrics.com>


"Tony J. Ibbs (Tibs)" wrote:
> 
> I am aiming to use Marc-Andr? Lemburg's mxTextTools package for an
> application in my (hah!) spare time. Being me, the first thing I've set out
> to do is to provide the missing meta-language so that I can avoid typing in
> the tag tables as "bare" tuples.
> 
> What I've done so far is to produce an alpha version of the meta-language
> documentation, which can be found at
> 
>     http://www.tibsnjoan.demon.co.uk/mxtext/metalang.html
> 
> Translations of the examples from the mxTextTools distribution are also
> supplied.

I had a look. And after reading from your page

"""
The general notation: 

     <condition>:
         <suite>
     <statement>

(where <suite> is one or more statements) is exactly equivalent to: 

     <condition> F:<labelN> [T:next]
         <suite>
     <labelN>
     <statement>
"""
...

I got the impression: Why another language?
Can't we express the tagging tuples by Python programs?
Marc's tuples are nothing more than predefined actions,
conditions, and jumps to labels.

Given a restricted subset of Python, and a couple of known
functions which are to be used, the tags could be expressed
as a small number of Python functions, and
it would be a real application for the ByteCodeHacks
to turn these Python functions back into tagging tables.
We would go without any parsing since Python does it for us.

Whoops - I could make the tagging engine into a pluggable
interpreter for stackless Python, this is just another engine.

yes-I-have-to-duck-and-cover-ly y'rs - chris :-)

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From steve at renlabs.com  Tue Jul 13 15:31:44 1999
From: steve at renlabs.com (Steven Work)
Date: 13 Jul 1999 12:31:44 -0700
Subject: Embedded Python can't "import socket" -- "undefined symbol"
Message-ID: <87zp104a8v.fsf@solano.in.renlabs.com>

I am embedding Python in a C program (on Linux 2.2.x, Debian 2.1), and
trying to run a script via PyRun_SimpleFile().  "print 'hello, world'"
works, but "import socket" dies.  Can anyone help me find out why?

$ ./testfile
Traceback (innermost last):
  File "jukebox_xml.py", line 1, in ?
    import socket
ImportError: /usr/lib/python1.5/lib-dynload/socketmodule.so: 
	undefined symbol: _Py_NoneStruct
$ nm testfile | grep NoneStruct
0808ecfc D _Py_NoneStruct

And of course "import socket" works fine in /usr/bin/python.  
-- 
Steven Work
Renaissance Labs
steve at renlabs.com
360 647-1833



From eric.kidd at pobox.com  Sat Jul  3 15:15:50 1999
From: eric.kidd at pobox.com (Eric Kidd)
Date: Sat, 3 Jul 1999 15:15:50 -0400
Subject: Curses, Ncurses, Slang, etc?
In-Reply-To: <7lljta$bnj$1@news.rhein-zeitung.DE>; from olli@rhein-zeitung.de on Sat, Jul 03, 1999 at 06:14:02PM +0000
References: <19990703140217.D14374@moebius.dartmouth.edu> <7lljta$bnj$1@news.rhein-zeitung.DE>
Message-ID: <19990703151550.E14374@moebius.dartmouth.edu>

On Sat, Jul 03, 1999 at 06:14:02PM +0000, olli at rhein-zeitung.de wrote:
> I can only tell you that the ncurses module I wrote works for me on my
> RedHat 6 and Mandrake 6 systems. What problems do you have?

Nothing dramatic. I just need to set VERSION=1.5 and do something about
Setup.in:

  cp ./Setup.in Setup
  cp: ./Setup.in: No such file or directory
  make[1]: *** [Setup] Error 1
  make[1]: Leaving directory `...'
  make: *** [boot] Error 2

But I'm writing an end-user application and I didn't want to require anyone
to edit Makefiles.

> If support and so on is an issue, then I would use the slang bindings at
> the moment.

Slang bindings would be wonderful. Is there such a thing for Python, and if
so, where does it live?

Many thanks for your help!

Cheers,
Eric



From billtut at microsoft.com  Mon Jul 12 21:42:00 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Mon, 12 Jul 1999 18:42:00 -0700
Subject: Python and COM?
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F1D@RED-MSG-50>


> From: htrd90 at zepler.org [mailto:htrd90 at zepler.org]
> 
> 
> Bill Tutt <billtut at microsoft.com> wrote:
> 
> (posted an cc'ed to bill)
> 
> >(As an added bonus, this should actually work)
> >The only GUID from interp.py is the coclass one. Every other 
> GUID was just
> >generated.
> 
> I _think_ the python code would also need to be changed to allow
> QueryInterface for the IID of that dispinterface.
> 

Nope. Dispinterfaces are wierd in that they only tell you whats available
via the IDispatch interface. You have to do a QI for IDispatch, not the
dispinterface.

Bill



From Eugene.Leitl at p98.f112.n480.z2.fidonet.org  Thu Jul  1 07:14:07 1999
From: Eugene.Leitl at p98.f112.n480.z2.fidonet.org (Eugene Leitl)
Date: Thu, 01 Jul 1999 12:14:07 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <0000164b@bossar.com.pl>

From: Eugene Leitl <eugene.leitl at lrz.uni-muenchen.de>

Hank Fay writes:
 > Is it possible to use ActiveX controls in any of these GUI's?  Then you've
 > got a large choice of controls.

No system-dependencies here please. Particularly those relating to an 
obsolete system, doomed to die <0.02 wink>.




From mstenber at cc.Helsinki.FI  Thu Jul 29 02:44:32 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 29 Jul 1999 09:44:32 +0300
Subject: HPUX and threads?
References: <al8yag0x66w.fsf@sirppi.helsinki.fi> <7nnq4t$viq$1@towncrier.cc.monash.edu.au>
Message-ID: <al81zdsuf8v.fsf@sirppi.helsinki.fi>

jon at rdt.monash.edu.au (Jonathan Giddy) writes:
> Markus Stenberg <mstenber at cc.Helsinki.FI> writes:
> >Is HPUX-11 thread support tested/implemented/whatnot? 
> Implemented - yes, tested - not in your environment (until now).

Very well, let's see about it now ;-)

> >This at least does not look promising..
> ># /opt/python/bin/python
> >pthread_mutex_init: Invalid argument
> >Memory fault(coredump)
> 
> OK, this is probably due to an incorrect guess about which version of 
> pthreads is being used.
> 
> If you look at the beginning of the source file Python/thread_pthread.h
> you'll see a series of preprocessor tests, defining PY_PTHREAD_D4 and 
> friends.
> 
> It seems HPUX selects either Draft 4 or the final standard, based on whether
> _DECTHREADS_ is defined.  Try placing the following after the #endif
> 
> #if defined(PY_PTHREAD_D4)
> #error PY_PTHREAD_D4
> #elif defined(PY_PTHREAD_D6)
> #error PY_PTHREAD_D6
> #elif defined(PY_PTHREAD_D7)
> #error PY_PTHREAD_D7
> #else
> #error PY_PTHREAD_STD
> #endif

This resulted in PY_PTHREAD_STD as follows.

> make clean and recompile.  This will (hopefully) give you an error like:
> In file included from thread.c:146:
> thread_pthread.h:88: #error PY_PTHREAD_STD
> 
> which tells you which version was assumed.  Then, try undefining the
> assumed value, and replace it with the likely correct value.
>
> So, if you get PY_PTHREAD_STD, then replace the code above with:
> #undef PY_PTHREAD_STD
> #define PY_PTHREAD_D4

No luck with _this_, I'm afraid.. nor with D6 or D7. D6 and STD compiled,
but both resulted in coredump_ok binaries.

> then make clean; make
> 
> That will hopefully get you running.  A more permanent solution requires
> someone who knows the history of pthreads on HPUX.
> 
> Jon.

-Markus

-- 
"I don't truly see how _playing_ gods and demons is odder than
 _believing_ in them." [response to one of those anti-RPG comments]



From wtanksle at dolphin.openprojects.net  Fri Jul  2 20:25:51 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Sat, 03 Jul 1999 00:25:51 GMT
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> <wkzp1g2t2e.fsf@ifi.uio.no>  <14204.912.383867.589108@lrz.de> <slrn7noclt.429.cbbrowne@knuth.brownes.org>
Message-ID: <slrn7nqm4f.3df.wtanksle@dolphin.openprojects.net>

On Fri, 02 Jul 1999 03:33:03 GMT, Christopher B. Browne wrote:
>On Thu,  1 Jul 1999 17:12:06 -0700 (PDT), Eugene Leitl <eugene.leitl at lrz.uni-muenchen.de> posted:
>>Lars Marius Garshol writes:

>> > Huh? Forth dates back to the 60s, whereas Scheme is from 1975, and I'm
>> > quite unsure of whether Chuck Moore knew Lisp at all.

>>Chuck is mostly doing (quirky, but fascinating) hardware design 
>>nowadays. Check out http://www.ultratechnology.com/ to what he has
>>been up to recently.

>Anyone else that said the things Chuck Moore does would be considered
>a "net-kook."  He's had some odd-ball stuff such as:

Chuck would be a kook if he ever argued about what he does -- we need more
people like him, who are willing to explore without being driven to argue
with people about it.

>- Actually using a "chording" keyboard
>- Eliminating source code (OK)

He now thinks both were a bad idea -- a failed experiment.

>- A 20 bit microprocessor

This was driven by economics, not experiment -- he prefers 32 (or 33) bits
in real life ;-).  I think that given the choice, he'd make it 33 bits
internal and 32 bits external.

The extra bit, for those of you not following Chuck's work, provides a
carry bit for every number on the stack.  A very cool way to write
smaller, simpler code.

>- A conciously tail-recursive Forth

One of his newer things...  Time will tell.  It allows for writing code
without loops, but it's an implicit and required optimization, which is
not something Forth usually likes.

>- Designing chips by directly manipulating arrays, rather than using
>abstract languages.

To the best of my knowledge he never did this.  He wrote a VLSI CAD
program from scratch (without source code, see OK) so that he could design
his processor by drawing lines and connecting them.

>- The claim that code should *not* be written to be reusable, but rather
>should be tightened so as to "throw away everything that isn't being used."

Or more accurately, the design should not include things that aren't
required (and the requirements should be trimmed to include only what's
known to be needed right now).  You don't throw things out of working code
just for the heck of it, you know ;-).

>If it weren't that he's actually gotten truly useful results out of
>all of these things, he'd probably be considered "just another of the
>opinionated fools" that clutter up the Internet.

And most of the opinionated fools (ahem, like yours truly) don't have half
of the exploratory genius that he does.  Even if he does fail sometimes.

>Note that the final item on that list is really crucial; he minimizes
>things to the max, and the rather bizarre results are indeed direct
>results of that philosophy.

Yes, simple is beautiful.  Simple is also maintainable.  I think everyone
in these language groups has to agree with that; our languages are
different, but they're all based upon some conception of simplicity.

If I were more prone to cross-posting I'd toss this over to the Forth NG,
but we don't really need that :-).

>cbbrowne at hex.net - "What have you contributed to free software today?..."

I maintain Omega -- only LGPL, but it's only a game.

-- 
-William "Billy" Tanksley



From bobalex at ix.netcom.com  Thu Jul  8 12:29:46 1999
From: bobalex at ix.netcom.com (Bob Alexander)
Date: Thu, 8 Jul 1999 09:29:46 -0700
Subject: Why are index() and count() only for mutable sequences?
Message-ID: <004701bec95f$186faa80$a0ee6dd1@BobAlexDesk>

There is no modification of the sequence performed by these functions -- it seems as though they could be available for all sequences. It would certainly be useful for tuples as well as lists (strings, too, for that matter, even though we have the more general string.find()).

-- Bob
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19990708/7d69096c/attachment.html>

From cyberescue at p98.f112.n480.z2.fidonet.org  Thu Jul  1 19:35:36 1999
From: cyberescue at p98.f112.n480.z2.fidonet.org (cyberescue)
Date: Fri, 02 Jul 1999 00:35:36 +0100
Subject: newbie
Message-ID: <0000165c@bossar.com.pl>

From: "cyberescue" <ashfordv at usa.net>

I am new to programming. Where can I get a copy of Python?

mailto: ashfordv at usa.net






From guido at CNRI.Reston.VA.US  Thu Jul  8 09:09:28 1999
From: guido at CNRI.Reston.VA.US (Guido van Rossum)
Date: Thu, 08 Jul 1999 09:09:28 -0400
Subject: .pyo's without .pyc's?
In-Reply-To: Your message of "Thu, 08 Jul 1999 14:31:59 +0200."
             <37849A3F.28C0EF21@lemburg.com> 
References: <19990526140137.A66609@vislab.epa.gov> <374C6C6E.1D07326B@lemburg.com> <19990708074113.A61389@epa.gov>  
            <37849A3F.28C0EF21@lemburg.com> 
Message-ID: <199907081309.JAA01063@eric.cnri.reston.va.us>

> I've checked this and indeed you're right. exceptions.py is imported
> prior to all other modules at a time where the switch from .pyc to
> .pyo hasn't been done yet (this is done in _PyImport_Init(), yet
> exceptions.py is imported by _PyBuiltin_Init_2()):
[...]
> Still, I think you're right in that the .pyc -> .pyo switch should
> be done prior to importing *any* Python module, meaning that
> _PyImport_Init() should be called prior to importing anything.

Hmm, seems you're right.  As Linus says, all bugs will eventually be
found.  Seems like this would be the fix, right?  (_PyImport_Init() is 
pretty trivial so I don't doubt that it works.)

Index: pythonrun.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/pythonrun.c,v
retrieving revision 2.88
diff -c -r2.88 pythonrun.c
*** pythonrun.c	1999/04/07 18:32:51	2.88
--- pythonrun.c	1999/07/08 13:06:43
***************
*** 149,160 ****
  	PyDict_SetItemString(interp->sysdict, "modules",
  			     interp->modules);
  
  	/* phase 2 of builtins */
  	_PyBuiltin_Init_2(interp->builtins);
  	_PyImport_FixupExtension("__builtin__", "__builtin__");
  
- 	_PyImport_Init();
- 
  	initsigs(); /* Signal handling stuff, including initintr() */
  
  	initmain(); /* Module __main__ */
--- 149,160 ----
  	PyDict_SetItemString(interp->sysdict, "modules",
  			     interp->modules);
  
+ 	_PyImport_Init();
+ 
  	/* phase 2 of builtins */
  	_PyBuiltin_Init_2(interp->builtins);
  	_PyImport_FixupExtension("__builtin__", "__builtin__");
  
  	initsigs(); /* Signal handling stuff, including initintr() */
  
  	initmain(); /* Module __main__ */

--Guido van Rossum (home page: http://www.python.org/~guido/)



From martinb at talx.com.no.spam  Fri Jul 23 14:31:00 1999
From: martinb at talx.com.no.spam (Martin Bertolino)
Date: Fri, 23 Jul 1999 13:31:00 -0500
Subject: VC++ 6.0 exe file sizes
References: <o$h7sBAVZij3Ew7i@jessikat.demon.co.uk> <37981400.1980005@news.cswnet.com>
Message-ID: <7nacj7$ge1$1@newsin-1.starnet.net>

I can get the sources to compile the 1.5.2 without any problems, or any
changes.
Configuration: VC++6.0 + SP3 on NT4 Server + SP5.

Martin

James "Jimmy" Wilson <jiwils at acxiom.com> wrote in message
news:37981400.1980005 at news.cswnet.com...
> On Thu, 15 Jul 1999 19:20:05 +0100, Robin Becker
> <robin at jessikat.demon.co.uk> wrote:
>
> >Any VC++ gurus out there help with a problem I'm getting building the
> >python.exe file with VC++ 6.0 & win95 osr2. My hd has 4k clusters.
> >
> >No matter what optimisations etc I do the python.exe seems to come out
> >as 16384. I reckon it should be around 5k like the 1.5.2 distributed
> >version (which I think is built using VC++ 5.x).
> >
> >I see this particular problem, but what about all the other bits of exe
> >code etc?
> >--
> >Robin Becker
>
> You actually got it to compile?  Each time I tried, I got an error
> message close to "MASSIVE COMPILER ERROR".  I tried updating to VS SP3
> without any luck.  Did you happen to encounter this problem?
> --
> James "Jimmy" Wilson
> Software Engineer
> Acxiom Corporation
> ADN Development
> E-Mail: jiwils at acxiom.com





From ralph.gauges at eml.villa-bosch.de  Thu Jul 15 01:35:25 1999
From: ralph.gauges at eml.villa-bosch.de (Ralph Gauges)
Date: Thu, 15 Jul 1999 07:35:25 +0200
Subject: Recursive method
References: <1280168234-10452044@hypernet.com> <378CA422.B2E99CB7@eml.villa-bosch.de> <378CF9E0.C6DD4B63@compaq.com>
Message-ID: <378D731D.1CF72026@eml.villa-bosch.de>

> 
> Keep in mind that code in the class body is only
> executed *once*, not every time an instance is
> created.
> 

That was exactly the misconception I had, I thought this
code was executed each time, I create an instance. And I
didn't notice that this is wrong, because I only created one
instance, which seemed to do what it should (but actualy it
did nothing.)

Thanks

Ralph



From olevi at my-deja.com  Thu Jul 29 03:16:56 1999
From: olevi at my-deja.com (Ole-Kristian Villaboe)
Date: Thu, 29 Jul 1999 07:16:56 GMT
Subject: passing data to Fortran in Python
Message-ID: <7nov55$5g6$1@nnrp1.deja.com>

Is writing an extension module the only solution?

Maybe someone has already done so or know of someone who has?

Ole-Kr.  (olevi at esrf.fr)


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From darko.cvagic at arcus.tel.hr  Wed Jul 21 08:09:18 1999
From: darko.cvagic at arcus.tel.hr (Darko Cvagic)
Date: Wed, 21 Jul 1999 14:09:18 +0200
Subject: gadfly RDBMS ???
Message-ID: <3795B86E.86229F6@arcus.tel.hr>

What does it means that GADFLY is Relational ?
How can I make relations between tables in database?
Can I use KEYS



From William.Tanksley at p98.f112.n480.z2.fidonet.org  Fri Jul  2 20:00:32 1999
From: William.Tanksley at p98.f112.n480.z2.fidonet.org (William Tanksley)
Date: Sat, 03 Jul 1999 01:00:32 +0100
Subject: Slicing beyond list --  bug or featurette?
Message-ID: <00001698@bossar.com.pl>

From: wtanksle at dolphin.openprojects.net (William Tanksley)

On Fri, 2 Jul 1999 22:22:10 GMT, Ivan Van Laningham wrote:
>Hi All--

>Pierre Fortin wrote:

>> Just ran into a situation which can be summarized as follows:

>[bobbitt]

>> list[2:50] returns [ 2,3,4 ]
>> list[20:] or list[20:30] returns []

>> Can I count on this, or should I code try/except for the day it's

>I don't presume to speak for Guido, but the way slicing is implemented
>it seems likely that you can count on it: list[:] is implemented as if

Furthermore, this makes fundamental sense -- an array reference is defined
to return the item stored at that location.  A slice is defined to return
a list containing the items within those locations.

An empty list is still a list; an item that doesn't exist, however, is
hard to return.

><but-they-don't-call-him-guido-svetz-for-nothing>-ly y'rs,
>Ivan;-)

We-have-VAYS-of-justifying-vhat-we-vant-ly y'rs,

-- 
-William "Billy" Tanksley



From dont at try.it  Mon Jul  5 12:47:49 1999
From: dont at try.it (dont at try.it)
Date: Mon, 5 Jul 1999 18:47:49 +0200
Subject: CGI under NT
References: <3780ca11@news2.isys.net> <19990705171820.L26497@xs4all.nl>
Message-ID: <3780e363@news2.isys.net>

Hi Thomas,

I never used post as method, but I used get. If I am using post I have to
read the values out of an file which path is specified in CONTENT_FILE. It
is more to code if I use method=post.

Yes, I could use a module like cgi, but I don't want to do it. It must work
without it.

Do you know more about it?

Greetings
Moritz Schmitt

Thomas Wouters <thomas at xs4all.nl> schrieb in im Newsbeitrag:
19990705171820.L26497 at xs4all.nl...
> On Mon, Jul 05, 1999 at 05:06:54PM +0200, dont at try.it wrote:
>
>
> > Until now I wrote a script which takes alle parameters of the forms out
of
> > "QUERY_STRING". There is my problem. If I press the submit-button I get
a
> > Message that the script has problems to read the system-variable
(because it
> > is not there!), but if I call the script with the parameters directly
> > (..../script.py?value=100&browser=netscape etc.) the script works.
>
> > Can anyone helps me or give me a hint to more information on the web?
>
> Use <FORM METHOD=GET ...> instead of METHOD=POST, or use a CGI module that
> parses both POST and GET and returns them on an object. (Haven't worked
with
> CGI under Python yet, but I'm fairly certain there is such a module.)
>
> --
> Thomas Wouters <thomas at xs4all.net>
>
> Hi! I'm a .signature virus! copy me into your .signature file to help me
spread!
>
>





From movits at lockstar.com  Thu Jul  1 11:15:11 1999
From: movits at lockstar.com (Mordy Ovits)
Date: 01 Jul 1999 08:15:11 PDT
Subject: sms tests
References: <7l4th5$kub$1@nnrp1.deja.com>
Message-ID: <377BB09F.600887DA@lockstar.com>

It depend on how you plan on sending the SMS.  If you're doing it raw, it's
helpful to have a library for it already written.  If you're doing it via some
web gateway, it'll be a breeze in Python.
Also, if you have a good C library for SMS, you might consider SWIGing it.  See
www.swig.org
Mordy

marilena_kolia at my-deja.com wrote:
> 
> Hi,
> 
> I am trying to test the SMS performance and get some results about the
> time it takes a message to be transfered, but also the delays of each
> part of this procedure. I am planning to use Tcl programming. Does
> anybody know how to do it or have some suggestions about it?
> 
> Thank you
> 
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.

-- 
o Mordy Ovits
o Cryptographic Engineer
o LockStar Inc.
---------------------------------------------------------------------------
#!/usr/local/bin/python 
from sys import*;from string import*;a=argv;[s,p,q]=filter(lambda x:x[:1]!=
'-',a);d='-d'in a;e,n=atol(p,16),atol(q,16);l=(len(q)+1)/2;o,inb=l-d,l-1+d
while s:s=stdin.read(inb);s and map(stdout.write,map(lambda i,b=pow(reduce(
lambda x,y:(x<<8L)+y,map(ord,s)),e,n):chr(b>>8*i&255),range(o-1,-1,-1)))



From phxken at netzone.com  Fri Jul  2 12:43:36 1999
From: phxken at netzone.com (Farrell Kenimer)
Date: Fri, 02 Jul 1999 16:43:36 GMT
Subject: Tkinter just WON'T work right!
References: <7lhal1$7o$1@nnrp1.deja.com>
Message-ID: <377CEABB.12EFED5F@netzone.com>

joakal at my-deja.com wrote:

> Sorry to be so inspecific about my subject =) Also, I'm sure this
> question is asked frequently, but I couldn't find what I was looking for
> in any of the 1400 or so messages here (too many of them start with
> ``Help!'' or ``Python Newbie''). Anyway, here it is:
>
> I tried to use the Grail browser under Slackware 4.0 and Python 1.5.1. I
> got the error saying it's not configured for Tkinter, so I went into my
> /usr/lib/python1.5/config/Setup file and uncommented the appropriate
> stuff.
>
> I tried again, same problem. I read a bit more, and it said I have to
> rebuild Python to add Tkinter (don't remember having to do that before).
> So, argh, I tried to do that. Now I get an error when I use 'make'
> saying there's an error when it gets to sedscript.
>
> Can anyone help?
> Thanks in advance!
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.

Have just gotten grail.py to work on my COL2.2 Linux installation.
Spent hours trying to remake python for tk/tcl and then found out
that for COL2.2, it was already done.  tk8.0 and tc.8.0 are required.
grail package has two errors or had at the time I got it 2 weeks back.
Start python and see if you can import Tkinter.  We can talk in
e-mail as I am not a programmer and am only interested in grail
as a browser for another OS.

--
Farrell Kenimer of Phoenix, Arizona  USA


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19990702/4e110902/attachment.html>

From uyr at jhfsdf.com  Fri Jul 23 16:36:15 1999
From: uyr at jhfsdf.com (uyr at jhfsdf.com)
Date: Fri, 23 Jul 1999 14:36:15 -0600
Subject: None Under 18 Allowed 102375
Message-ID: <230799143615@jhfsdf.com>

Adults Only:

http://207.240.225.250




hXp(dcMN



From jbauer at rubic.com  Thu Jul 22 21:37:31 1999
From: jbauer at rubic.com (Jeff Bauer)
Date: Fri, 23 Jul 1999 01:37:31 GMT
Subject: python + bash + new user
References: <3793BE52.43E023BB@netcard.com>
		<7n0f3d$1d7@dfw-ixnews10.ix.netcom.com>
		<3793C729.55DE42D0@netcard.com>
		<7n5nfq$j0$1@vvs.superst.iae.nl> <3797AE3B.687785E5@netcard.com>
Message-ID: <3797C75B.2CFC9529@rubic.com>

Joyce Stack wrote:
> I did exactly as you said and chmod's etc but it just
> says that it doesnt recognise "./simple.py" as a command.???
> >         $ cat simple.py
> >         #!/usr/bin/python

Joyce,

Your python interpreter is not located in /usr/bin.  It may
be in /usr/local/bin.  On many systems you can use the 'which'
command to determine where the executable is, e.g.

  $ which python
  /usr/local/bin/python

Alternatively, you can use the following line which works on
many Unix systems:

  #!/usr/bin/env python
                ^---- note the space between env and python

symlink-it-there-when-your-sysadmin's-not-looking-ly yr's,

Jeff Bauer
Rubicon, Inc.



From johann at physics.berkeley.edu  Tue Jul 20 17:22:15 1999
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 20 Jul 1999 14:22:15 -0700
Subject: python vs. lisp (whereis quote?)
References: <x88r9m3cyje.fsf@rpppc2.hns.com> <lbzk8rva3to.fsf@morgoth.uio.no> <x88zp0rbd1x.fsf@rpppc2.hns.com>
Message-ID: <mtlncbhv94.fsf@astron.berkeley.edu>

nbecker  writes:
>>>>>> "Rolf" =3D=3D Rolf Marvin B=F8e Lindgren <roffe at morgoth.uio.no> wri=

Rolf> if eval expects a string, and if the string can contain code,
Rolf> then you have it, haven't you?

> Sort of - but it has to be parsed - so it may not be very efficient.

Just pass a function.  It's easy.

def hook_fcn (x):
  do_something_with (x)

register_hook (hook_fcn)

Or use a class, with some defined "hook" methods.

class Hook:
  def call (self):
    # override me in subclasses!
    return None

Emacs lisp is a bad ancient ugly dialect of lisp.  I'm waiting for the
day they port the whole thing over to either Scheme or Common Lisp.
The whole "quote things then eval later" seems much more complicated
than it has to be.

-- 
Johann Hibschman                           johann at physics.berkeley.edu



From emile at fenx.com  Fri Jul 30 21:03:57 1999
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 30 Jul 1999 18:03:57 -0700
Subject: Import a lot of modules, or, import 'foo'
References: <874silabe7.fsf@linuxcare.com>
Message-ID: <1%ro3.4357$VL2.467111@news.direcpc.com>

Hi David,

This seems to work for me:

>>> for t in tests:
 exec ('import %s' % t)


--

Emile van Sebille
emile at fenx.com
-------------------


David N. Welton <dwelton at linuxcare.com> wrote in message
news:874silabe7.fsf at linuxcare.com...
>
> So, basically, I'd like to do this:
>
>  tests = ['fs', 'swap']
>
>  for t in tests:
>      import t
>
>  do more stuff on what promises to be a lot of modules
>
> I've looked around a bit, but can't find what I need.  Maybe I'm just
> going about this the wrong way, but... Hmm, now I want to know how to
> do this in any case:-)
>
> Thanks,
> --
> David N. Welton, Web Engineer, Linuxcare, Inc.
> 415.354.4878 x241 tel, 415.701.7457 fax
> dwelton at linuxcare.com, http://www.linuxcare.com/
> Linuxcare. At the center of Linux.





From graham at sloth.math.uga.edu  Fri Jul 16 11:24:28 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 16 Jul 1999 15:24:28 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com>
Message-ID: <7mnirc$t2b$2@cronkite.cc.uga.edu>

Fernando Pereira (pereira at research.att.com) wrote:
: but, as far as I know, nothing beats
: the automata-theoretic optimization techniques for pattern matching and
: parsing, of which the typical modern compiler for regular expressions
: is just the best-known instance.

In what sense are you using the words "nothing beats". I am sure that
the automata-theoretic approach produces the fastest pattern matchers
and parsers. So on this sense "nothing beats them".  But the class of 
languages recognisable by automaton doesn't cover all the languages of
interest (in fact my guess is it covers very few ... context ...). So
in this sense lots of things "beat" the automata approach. And then
there are software engineering issues ...

graham

-- 
              This ain't no technological breakdown
                 Oh no, this is the road to hell



From clgonsal at kami.com  Fri Jul  9 07:04:51 1999
From: clgonsal at kami.com (C. Laurence Gonsalves)
Date: Fri, 09 Jul 1999 11:04:51 +0000
Subject: VIM and [Python] block jumping
References: <3781347D.CB2FBAB3@starvision.com> <Pine.SOL.4.05.9907051548470.13052-100000@saturn.radsci.uci.edu> <37815B80.B4B51F94@starvision.com>
Message-ID: <7m9tmn$6qg$33@rks1.urz.tu-dresden.de>

Angus MacKay wrote:
> 
> I am not looking for indentation but just code navigation.
> 
> when code uses identifiers for blocks (such as "()" in lisp, "{}" is
> most sane languages, "#if/#endif" for C preproc ...) vim can navigate
> blocks very easily (with [{/]} for "{}", [#/]# for "#if/#endif" ...)
> 
> with python this becomes harder. is there any way to do it? (I think I
> could write a function for it but I thought I'd see if it had already
> been done)

That's a neat idea. I just wrote this up, so it hasn't been tested a
lot, but it seems to work. The way I implemented this requires that you
have python support compiled into VIM. It's probably possible to
reimplement this as a pure-VIM script, but the python version is
probably orders of magnitude faster (not to mention being easier to
write). :-)

(I had once implemented the C-style [{ ]} motion keys as vim-scripts
before I realized that VIM already had built-in support for that. They
were quite slow.)

First, make a file called "vimMotion.py":

## begin vimMotion.py ##

import vim
import string
import re

# returns a tuple containing the indentation, and the line number where
we got
# that indentation. That can be different than lineNo if lineNo refers
to an
# empty (ie: only whitespace) line.
def getIndent(lineNo):
    for line in vim.current.buffer[lineNo:]:
        result = getIndentQuick(line)
        if result>=0:
            return (result,lineNo)
        lineNo = lineNo + 1
    return (0,len(vim.current.buffer)-1)
        
# grabs the initial whitespace and the stuff after it into two groups
indentRE = re.compile( r"([\t ]*)(.*)" )

# returns the indentation for a line, or -1 if it can't be determined
(ie:
# only whitespace)
def getIndentQuick(line):
    match = indentRE.match(line)
    if len(match.group(2)) == 0:
        # if the line is just whitespace, we don't know the indentation
        return -1
    else:
        # make sure we expand tabs like python does...
        return len(string.expandtabs(match.group(1)))
        
# if dir is positive, goes down, negative, goes up. Other than that, the
value
# is irrelevant
def blockMotion( dir ):
    lineNo0 = vim.current.window.cursor[0]-1
    indent0,lineNo = getIndent(lineNo0)
    newLine = lineNo0

    if indent0 == 0:
        # this is an optimization
        if dir>0:
            newLine = len(vim.current.buffer)-1
        else:
            newLine = 0
    elif dir>0:
        indent = indent0
        numLines = len(vim.current.buffer)
        while lineNo<numLines and indent >= indent0:
            lineNo = lineNo + 1
            indent,lineNo = getIndent(lineNo)
        newLine = lineNo
    else:
        indent = indent0
        lineNo = lineNo0
        while lineNo>0 and indent >= indent0:
            lineNo = lineNo - 1
            qIndent = getIndentQuick(vim.current.buffer[lineNo])
            if qIndent>=0:
                indent = qIndent
        newLine = lineNo
    # move the cursor
    vim.current.window.cursor = (newLine+1,0)

## end vimMotion.py ##

Then add the following three lines to your .vimrc:

pyfile vimMotion.py
nmap [{ :python blockMotion(-1)<CR>^
nmap ]} :python blockMotion(1)<CR>^

You should probably just execute these when editting a python file.

Note that python is a little bit funny as to where blocks can end,
because multiple blocks can all end at exactly the same position. That
means that doing ']}' can potentially (and often does) take you out of
several blocks. In a language with braces, ]} will always take you out
of one block at a time.

The other tricky part was figuring out the indentation of a line. The
two things that make this tricky are tabs (but string.expandtabs made
that pretty easy), and the fact that python seems to ignore
empty/whitespace-only lines when figuring out block levels. The "right"
thing to do seems to be to scan forward until we find a line with some
non-whitespace on it.

Let me know if you have any problems.

(note: this was posted to both comp.lang.python and comp.editors, since
people in both those groups might be interested in this)

-- 
  C. Laurence Gonsalves              "Any sufficiently advanced
  clgonsal at kami.com                   technology is indistinguishable
  http://www.cryogen.com/clgonsal/    from magic" -- Arthur C. Clarke



From skip at mojam.com  Thu Jul 29 16:48:55 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 29 Jul 1999 15:48:55 -0500 (CDT)
Subject: [Tutor] I want to learn to hack
In-Reply-To: <14240.47957.840790.39395@dolphin.mojam.com>
References: <19990721154610.B3663@optiplex.palga.uucp>
	<Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org>
	<14231.11726.916541.492153@nobot.cnri.reston.va.us>
	<slrn7q1dau.17t.esr@snark.thyrsus.com>
	<14240.47957.840790.39395@dolphin.mojam.com>
Message-ID: <14240.48630.244929.427876@dolphin.mojam.com>

    Skip> Eric, just think how much more they'll be hacking in Python
    Skip> instead of MSVC++, VB or Perl! ;-)

Ack!  Should have been:

    Eric, just think how much more *productively* they'll be hacking in
    Python instead of MSVC++, VB or Perl! ;-)

so-much-for-off-the-cuff-witicisms-ly y'rs,

S




From tismer at appliedbiometrics.com  Sat Jul  3 19:06:24 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Sat, 3 Jul 1999 23:06:24 GMT
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain>	<377B7E17.5BD7@nospam.com>
		<377bbf9b@news.xtml.co.uk>
		<377BD789.C4A1D191@appliedbiometrics.com> <377c870b@news.xtml.co.uk>
Message-ID: <377E9770.B9C7FA56@appliedbiometrics.com>


Nick wrote:
> 
> The language doesn't necessarily need to support it. You could implement the
> threaded system yourself in Lua, you then have more control over what goes
> on. Proper threads are relatively slow in Windoze (I believe) and would be a
> nightmare to debug? If Lua wasn't fast enough you could write a scheduler
> with priority assessment in C/C++ and attached Lua script callbacks to this.

I'm not discussing what I can write in C/C++ but what LUA has.

> Personally I wouldn't dream of useing proper threads for game logic/AI I
> would write a basic scheduler, it wouldn't be difficult. Grim Fandango uses
> Lua as its scripting language and I think they did something along these
> lines. I think Balders Gate is another game that is rumoured to use it.

Which doesn't help me very much. I'd like to figure out
why I should use LUA instead of Python. Please give me real facts.

At the moment, I see no reason to switch, but a big
advantage using Python, since I can extend it extremely.
Please let me understand your point.

Still unsatisfied -ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From mal at lemburg.com  Wed Jul 28 04:17:15 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Wed, 28 Jul 1999 10:17:15 +0200
Subject: API question -- decref after adding to dictionary?
References: <rps64v$0$37nspbi$7gc@news.supernews.com>
Message-ID: <379EBC8B.E36F020B@lemburg.com>

Rich Salz wrote:
> 
> I'm converting my own associative array into a python dictionary so I can
> callout to some python code.  I think the snippet below should be clear
> enough.
> 
> PyDictObject* newdict = PyDict_New();
> for (CKeyValList::iterator i(kvl); i; ++i) {
>         PyString* v = Pystring_FromString(i.value());
>         PyDict__SetItemString(p, i.key(), v);
>         Py_DECREF(v);
> }
> 
> Is the DECREF necessary?  Reading the source of dictobject.c makes me think
> it is, but the comments in object.h say otherwise.

Yes it is needed. (Trust the code in dictobject.c, after all, that's
what gets executed !)

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   156 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
Rich Salz wrote:



From nospamplease at inf.ufrgs.br  Wed Jul 28 16:05:00 1999
From: nospamplease at inf.ufrgs.br (Carlos A M dos Santos)
Date: Wed, 28 Jul 1999 17:05:00 -0300 (EST)
Subject: Building a shared libpython
Message-ID: <Pine.GSO.4.03.9907281701150.15012-100000@spica>

Hello

I'm installing Python 1.5.2 and noted that there is no configure option to
buld a shared libpython.  How could I do it?

Thanks


Casantos




From un4e at rz.uni-karlsruhe.de  Sat Jul 17 12:09:09 1999
From: un4e at rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 17 Jul 1999 16:09:09 GMT
Subject: Embedding: Defining packages?
Message-ID: <7mq9r5$irj$1@news.rz.uni-karlsruhe.de>

Hi,

I'm embedding Python in a rather complex C program. To represent the
complexity (and to keep the thing modular), I want to define several 
Python modules and group them in a package hierarchy:

         MyProgram
         +- Server
         |       +- Communications
         |       +- HardwareAccess
         +- Client
                 +- Communications
                 +- AbstractObjects

...etc. How do I do that? Is there a way to pass hierarchical module names
to Py_InitModule, such as

theModule = Py_InitModule(methodDefs, "MyProgram.Server.Communications");

My attempts so far haven't worked.

Thanks in advance,
         --Bjoern

PS I'm posting this via the newsgroup, as sending mail to python-list at cwi.nl
   doesn't seem to work. Has anything changed there that I'm not aware of?
-- 
--------------------------------/\--One OS to rule them all---Windows NT-----
Bjoern Giesler                 /  \  One OS to find them           
<un4e at rz.uni-karlsruhe.de>    / <> \  One OS to bring them all    
-----------------------------/______\--And in the Darkness bind them---------
    Thank you for your time, worship the Antichrist, and have a nice day.



From pmai at acm.org  Thu Jul  1 09:19:59 1999
From: pmai at acm.org (Pierre R. Mai)
Date: 01 Jul 1999 15:19:59 +0200
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <lwoghyexy5.fsf@copernico.parades.rm.cnr.it> <377B454C.6F59@mailserver.hursley.ibm.com>
Message-ID: <87aetgfqww.fsf@orion.dent.isdn.cs.tu-berlin.de>

Paul Duffin <pduffin at mailserver.hursley.ibm.com> writes:

> If they are statically scoped how does the following work.
> 
> Look at the following code.
> 
> 	(let ((a 2))
> 	  (let ((b 3))
> 	    (+ a b)))
> 
> When you expand the macro let you get something like.
> 
> 	((lambda (a)
> 	   ((lambda (b) 
> 	      (+ a b)) 3)) 2)
> 
> Which is equivalent (apart from side effects) of.
> 
> 	(defun inner (b)
> 	  (+ a b))
> 
> 	(defun outer (a)
> 	  (inner 3))
> 
> 	(outer 2)

They are _NOT_ equivalent, because the reference to a in inner is now
completely free, which it was not in the lambda formulation.  When you
change the nesting of functions/forms like this, you of course change
their meaning (this is basic Lambda-calculus).

> So what is the process which associates the use of "a" inside inner
> with the formal argument of outer. I 

In lexical scoping there is no such mechanism, since a variable has to 
be lexically apparent, for it to be accessible.  Since a is not
lexically apparent in inner, it is not accessible.  That is one of the 
important features of lexical scoping, since it ensures that I only
have to look at the lexically enclosing environment to understand
which variables are accessed.

Common Lisp allows dynamic scoping for certain, specially declared,
variables.  In CL you _could_ write the following as a somewhat
equivalent form of your TCL code below, but no Lisp programmer would
do it that way:

(defun inner (b)
  ;; Note that a is a free reference in here, so we rely on there
  ;; being a dynamic binding in effect for a when we are called.
  ;; This is clearly insane ;)
  (+ a b))

(defun outer (a)
  (declare (special a))
  (inner 3))

(outer 2)

=> 5

If you really wanted to do something this way in CL, then a would be
considered a global parameter to the behaviour of inner, and would be
declared prominently as such.  If that isn't the case (as in your
example, where arguably a is just a parameter of inner that you don't
care to pass explicitly), you would simply use closures/nested
functions as appropriate, or indeed pass the argument around, if you
want inner to be self-sufficient.  I.e. either do

a)

(defun outer (a)
  (flet ((inner (b) (+ a b)))
    (inner 3)))

b) 

(defun inner (a b)
  (+ a b))

(defun outer (a)
  (inner a 3))

or 

c)

(defvar *inner-frobnication-parameter* 0
  "This global parameter allows you to fine tune the frobnicator of
inner in a global fashion.  Please use with appropriate care, since
too high value here might result in the sudden destruction of the
whole of creation, which some would consider to be enormous lossage.")

(defun inner (b)
  "Frobnicates b via a general purpose frobnication-algorithm.  See
also *inner-frobnication-parameter*, which gives the total bias for
the algorithm.  See Knuth, Vol 1, P. 169 for a specification of the
algorithm and the parameters involved."
  (+ *inner-frobnication-parameter* b))

(defun outer (a)
  "Frobnicates 3 via inner, with a as the setting for the
*inner-frobnication-parameter*."
  (let ((*inner-frobnication-parameter* a))
    (inner 3)))

> The Tcl equivalent is
> 
> 	proc inner {b} {
> 	    upvar 1 a a;
> 	    expr {$a + $b}
> 	}
> 
> 	proc outer {a} {
> 	    inner 3
> 	}
> 
> 	outer 2

Even though I'd shoot every programmer that wrote the Lisp code I
first wrote above, it still seems to me to be a much better construct
than the corresponding upvar in Tcl:

a)  With Tcl it seems to me that the caller has no way of knowing
    which of the procedures he calls might simply start poking at his
    variables.  Worse:  Since upvar can be done for any n levels, he
    will have to inspect _all_ of the procedures ever called while he
    is active, to ensure that they are not using any of his internal
    variables.

    This to me seems clearly insane:  If I were maintaining the Tcl
    code above, and at some point decided that a was a misnomer and
    should be called first-of-two (or whatever else), I'd have to look 
    at all procedures ever called to ensure none was relying on my
    having a variable called a.  In Lisp the special declaration would 
    at least warn me about the fact that someone somewhere was relying 
    on me dynamically binding a.

b) The fact that upvar has to specify at which level the
   corresponding variable is to be found, makes apparent it's hackish
   nature.  Inner shouldn't have to know who exactly will provide it
   with the value for a.  In Tcl I can't call inner via a mediating
   function, like this:

   proc master {a} {
     middle 3
   }

   proc middle {c} {
     inner c
   }

   For this to work, I'd either have to modify inner (in which case
   the old call-chain via outer would cease to work), or I'd have to
   modify middle, to ensure middle passes along a.  Now imagine master 
   either calling inner directly or via middle (depending on some
   condition), and you get into serious trouble understanding and
   maintaining the various relationships and variable bindings
   involved.  This seems to me to take the problems of dynamic binding 
   to new levels (i.e. it somehow reminds me of assembler code poking 
   around in a caller's stack frame, to avoid copying values into new
   stack frames).

> As I have said before it is simply an explicit form of the Lisp
> mechanism which allows the let macro to be implemented using
> lambda. It is very rare that N is more than 1.

To repeat it: dynamic scoping is in no way necessary or even helpful
for implementing let via lambda.  LET is directly expressible via
LAMBDA in the lambda-calculus, which is lexically scoped.  The same
goes for Scheme, which doesn't have dynamic scoping either.  What you
are doing is performing invalid transformations on lambda-calculus
expressions, going on to claim that since you need dynamic scoping to
make those transformations work, that dynamic scoping is needed to
allow LET to be expressed in LAMBDA.  Had you any knowledge of Lambda
calculus, you wouldn't make these ridiculous claims.  BTW: The invalid
transformations you perform are common-place in Tcl, because Tcl (like
C) doesn't allow nested functions.

> > You are very mistaken.  Common Lisp macros are definitively not
> > another language.  They manipulate S-expressions, which are what Lisp
> 
> Lisp macros manipulate S-expressions but the macro definitions themselves
> are interpreted differently to Lisp expressions.

The macro-definitions are interpreted exactly in the same way as any
other Lisp expression, i.e. in

(defmacro blabla (a b)
  (my-expression))

(my-expression) is interpreted in exactly the same way as
(my-expression) would in any other context (with appropriate bindings
in place of course).

> A simple macro version of setq would convert from
> 	(setq symbol '(list))
> to
> 	(set 'symbol '(list))

It wouldn't do this, since set can only access "global variables",
whereas setq can access local variables.  Take note please that set
is deprecated, and that setq is not a macro, but a special form.  If
this doesn't mean anything to you, then please refer to the ANSI
Common Lisp standard (also available in a hypertext version, at
http://www.harlequin.com/books/HyperSpec), which would benefit this
discussion greatly.

> A macro setq is not interpreted the same way as set is because if it
> was an error would occur when the Lisp interpreter tried to get the
> value of the variable symbol before symbol was created.

This doesn't make any sense.  Please get your vocabulary right.  The
symbols in question are created at read-time, so there is no way that
either of the above forms could be executed in an environment where
the symbol in question didn't exist, unless you did some very evil
hacking with the package system, the consequences of which I wouldn't
claim to be well-defined.

Now that I've re-read your sentence for the xth time, it seems to me
that you are claiming that (setq symbol 5) is interpreted differently
than (set symbol 5), which is of course right.  That was the whole
point of defining a macro for setq, that transformed (setq symbol 5)
into (set 'symbol 5) and not into (set symbol 5).

But this is not inconsistent in my book, because that is exactly what
you asked for: You _wanted_ every (setq a b) form to be interpreted
as if (set 'a b) had been written.  So after macro-expansion, the
evaluation of (set 'a b) is totally consistent with the evaluation of
any other set form.

If you find this inconsistency troublesome, than you have to condemn
every kind of macro system, and Tcl, too, which allows the user to
change the behaviour of commands.

> I don't underestimate the power of Lisp macros I am just saying that I
> find the way they warp the otherwise simple Lisp syntax / semantics.
> 
> Lisp does not really need macros, they are really just a way of
> creating short cuts.

Firstly this sounds like the old argument against HLLs, i.e. we don't
need high-level languages, they are really just a way of creating
short cuts, which though true, totally misses the point of HLLs.

Secondly, since special forms (of which setq as well as quote and
lambda are prime examples) also warp the otherwise simple evaluation
rules of Lisp, it can be argued that Lisp effectively needs at least a
certain amount of warping, to be usable at all.  Even the pure lambda
calculus is warped in your eyes then.  Funny how all the warpiness of
Tcl doesn't grate on your nerves...

> Tcl does not need a macro system because it has a simple consistent
> syntax / semantics and all commands are treated equally and it is
> possible to change the behaviour of existing commands.

This is clearly no argument at all, because all the purported features
of Tcl you mention are clearly orthogonal to the existence of a macro
system.

> I would say that Tcl (without macro system) can do anything that Lisp
> (with macro system) can do.

Assembler can do anything either Lisp or Tcl can do.  This is no
argument against either Lisp or Tcl.  At the point where programming
language discussions turn to this Turing-machine-equivalence argument, 
it is clear that the discussion is fruitless.

end-of-line.

Regs, Pierre.

PS:  Should any further discussion seems sensical (it doesn't to me),
please heed the F'up to c.l.l, since neither Python nor Scheme are at
all relevant to this.

-- 
Pierre Mai <pmai at acm.org>         PGP and GPG keys at your nearest Keyserver
  "One smaller motivation which, in part, stems from altruism is Microsoft-
   bashing." [Microsoft memo, see http://www.opensource.org/halloween1.html]



From amitp at Xenon.Stanford.EDU  Fri Jul  9 14:32:31 1999
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 9 Jul 1999 18:32:31 GMT
Subject: Curses, Ncurses, Slang, etc?
References: <19990703140217.D14374@moebius.dartmouth.edu>
Message-ID: <7m5f7v$suo$1@nntp.Stanford.EDU>

 Eric Kidd  <eric.kidd at pobox.com> wrote:
| I'd like to write a Python application which uses either curses, ncurses,
| slang or a similar tty interface library. Unfortunately, the built-in
| curses module is hopeless broken on my RedHat 6.0 system (it doesn't export
| 'addch', for example).

I've been learning curses using Python & RedHat 6.0, and haven't had
any problems so far.  Could you show us a small program that results
in an error?

   - Amit





From phd at emerald.netskate.ru  Fri Jul 23 05:55:09 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Fri, 23 Jul 1999 13:55:09 +0400 (MSD)
Subject: "libmysqlclient.so.6" not found. (what does /sbin/ldconfig
 /usr/local/lib/mysql do ?)
In-Reply-To: <7n9c7f$ihd$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.04.9907231351440.30978-100000@emerald.netskate.ru>

Hi!

   You havn't mention your OS and version, looking for ldconfig I think it
is Linux.
   To permanently add /usr/local/lib/mysql to ld search directory path, add
it to /etc/ld.so.conf file (it is just plain text file - add yet another
line to the end of the file) and run ldconfig without parameters to rebuild
ld.so.cache.

On Fri, 23 Jul 1999 sweeting at neuronet.com.my wrote:
> When building Python from source, and with the MySQLmodule in the
> Modules directory, I got the following error :
> 
> Shared object "libmysqlclient.so.6" not found
> *** Error code 1
> 
> But libmysqlclient.so.6 does indeed exist in /usr/local/lib/mysql.
> 
> Fortunately, somebody had  come across this in a previous post/thread
> and the cure was to run " /sbin/ldconfig /usr/local/lib/mysql" before
> compiling. Sure enough this worked and now I have a fully functioning
> Python interpreter (and environment)  :)
> 
> However, upon rebooting the machine, if I try to run a Python program
> it will again complain that "libmysqlclient.so.6" can not be found
> (since the MySQLmodule is shared). Once again
> running " /sbin/ldconfig /usr/local/lib/mysql"  cures the problem but
> does this mean that I have to run it each and every time I restart the
> computer, as part of my startup scripts ? Is this error indicative of
> there being a larger problem or misconfiguration ?

   It is not misconfiguration - but ld should know where to find .so files.

> Thanking you in advance,
> 
> chas

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From martin.lemburg at bigfoot.de  Sun Jul 11 07:03:33 1999
From: martin.lemburg at bigfoot.de (martin.lemburg at bigfoot.de)
Date: Sun, 11 Jul 1999 13:03:33 +0200
Subject: python, tkInter -> tcl/tk problems
Message-ID: <7m9tok$6qg$94@rks1.urz.tu-dresden.de>

Hi,

I would be glad to get some help belonging to the use of tkInter with python 1.5.2!

The situation:

1. I already have tcl/tk 8.05 installed
2. tcl/tk is inside the path of WindowsNT
3. starting idle returns following error message: 
>Traceback (innermost last):
>  File "E:\programming\Python\Tools\idle\idle.py", line 3, in ?
>    PyShell.main()
>  File "E:\programming\Python\Tools\idle\PyShell.py", line 611, in main
>    root = Tk()
>  File "E:\programming\Python\Lib\lib-tk\Tkinter.py", line 893, in __init__
>    self.tk = _tkinter.create(screenName, baseName, className)
>TclError: Can't find a usable init.tcl in the following directories:
>    {} ./lib/tcl8.0 E:/programming/Python/tcl8.0/library E:/programming/Python/Tools/library

My tries to solve the problem:

1. I copied the lib/tcl8.0 directory from my tcl/tk installation into the python directory:
    <tcl/tk>/lib/tcl8.0    =>    <python>/tcl8.0/library
2. I copied the lib/tk8.0 directory from my tcl/tk installation into the python directory:
    <tcl/tk>/lib/tk8.0    =>    <python>/tk8.0/library

New situation:

1. starting idle returns following error message:
>Traceback (innermost last):
>  File "E:\programming\Python\Tools\idle\idle.py", line 3, in ?
>    PyShell.main()
>  File "E:\programming\Python\Tools\idle\PyShell.py", line 611, in main
>    root = Tk()
>  File "E:\programming\Python\Lib\lib-tk\Tkinter.py", line 893, in __init__
>    self.tk = _tkinter.create(screenName, baseName, className)
>TclError: Can't find a usable tk.tcl in the following directories:
>    . E:/programming/Python/tcl8.0/tk8.0 ./tk8.0 ./lib/tk8.0 ./library ./tk8.0/library
> ./library E:/programming/Python/tk8.0/library
>
>E:/programming/Python/tk8.0/library/tk.tcl: bad event type or keysym "MouseWheel"
>bad event type or keysym "MouseWheel"
>    while executing
>"bind Listbox <MouseWheel> {
>    %W yview scroll [expr - (%D / 120) * 4] units
>}"
>    (file "E:/programming/Python/tk8.0/library/listbox.tcl" line 179)
>    invoked from within
>"source [file join $tk_library listbox.tcl]"
>    (file "E:/programming/Python/tk8.0/library/tk.tcl" line 151)
>    invoked from within
>"source E:/programming/Python/tk8.0/library/tk.tcl"
>    ("uplevel" body line 1)
>    invoked from within
>"uplevel #0 [list source $tkfile]"

What know?

The tcl/tk-version delivered by python 1.5.2 is 8.0.5 too, so what goes wrong?
Why does python/tkInter looks for tcl/tk in non tcl/tk-standard directories?
What could I do to activate tkInter to use it with python 1.5.2?

I would be glad reading something from you - but it would be nice to email me your answers too!

Bye

Martin Lemburg

____________________________________________

Martin Lemburg
Bielfeldtstra?e 12 - D-22763 Hamburg
Tel./Ph.   +49 (40)  8803486
Fax.        +49 (441) 800 990 6462
eMail.    martin.lemburg at bigfoot.de





From tbryan at arlut.utexas.edu  Sat Jul  3 00:16:17 1999
From: tbryan at arlut.utexas.edu (Thomas A. Bryan)
Date: Sat, 03 Jul 1999 00:16:17 -0400
Subject: Getting to know Python
References: <377d57ae@pavlof.gci.net>
Message-ID: <377D8E90.A7DB085B@arlut.utexas.edu>

Arnaldo wrote:

> In the mean-time I'm eager to have with the interactive interpreter.
> How do I access the shell from the interpreter to enter commands like
> ,clear, ls, cp, etc.
> Is it posible, how do I do that?

Yes.  os.system() or os.popen()
Of course, many of these commands are available in the standard 
Python library.

For exmple, os.listdir() and shutil.copy() for ls and cp.

Python 1.5.2 (#1, Apr 18 1999, 16:03:16)  [GCC pgcc-2.91.60 19981201 
(egcs-1.1.1  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import os
>>> retval = os.system('ls /')
bin  boot  dev  etc  home  lib  lost+found  misc  mnt  opt  proc
  root  sbin  tmp  usr  var
>>> cmd = os.popen('ls /')
>>> output = cmd.readlines()
>>> cmd.close()
>>> output
['bin\012', 'boot\012', 'dev\012', 'etc\012', 'home\012', 'lib\012', 
'lost+found\012', 'misc\012', 'mnt\012', 'opt\012', 'proc\012', 
'root\012', 'sbin\012', 'tmp\012', 'usr\012', 'var\012']

If you're playing in the interpreter, learn to use the doc strings 
for the modules that have them:

>>> print os.system.__doc__
system(command) -> exit_status
Execute the command (a string) in a subshell.
>>> print os.popen.__doc__
popen(command [, mode='r' [, bufsize]]) -> pipe
Open a pipe to/from a command returning a file object.

To find out what's available, do a dir(os).

You may also want to read through/download the library reference from 
www.python.org

Here's a handy function...you'll probably want to modify it so that 
it writes to a file unless you have a lot of "history" on your terminal
window.

def get_docs(modulename):
    for el in dir(eval(modulename)):
        try:
            print el
            exec('print %s.%s.__doc__' % (modulename,el))
        except: #catch all exceptions
            print '%s has no __doc__' % el

#for example
import string
get_docs('string')



From chad at vision.arc.nasa.gov  Fri Jul 23 03:19:45 1999
From: chad at vision.arc.nasa.gov (Chad Netzer)
Date: Fri, 23 Jul 1999 00:19:45 -0700
Subject: int(long(-sys.maxint-1)) fails on Linux
References: <000001bed49f$c052e820$2c2d2399@tim> <3797D8A1.E6D95CCB@vision.arc.nasa.gov>
Message-ID: <37981790.24835500@vision.arc.nasa.gov>

Here I go, replying to my own posts again...

Chad Netzer wrote:

> C code:
>    if ((long)x < 0 && (sign > 0 || (x << 1) != 0))
>             goto overflow;
>     return (long)x * sign;
>
> not working:

deleted... It was actually working code, assembled with pgcc WITHOUT any optimization.
The bug I've describing only manifests itself when -O is selected (or even -O1 or -O2).
So I apologize if anyone looked at the code and couldn't find a problem; it didn't exist in
that version.

Here is some assembly of the same region, compiled with the -O1 flag, which DOES
exhibit the incorrect behavior.  Notice that it is quite different from the previous code,
and I think the bug is that it uses the "LEA" instruction improperly:

 <PyLong_AsLong+89>: test   %edx,%edx                            ; x < 0
 <PyLong_AsLong+91>: jge    0x806c22a <PyLong_AsLong+106>
 <PyLong_AsLong+93>: test   %ebx,%ebx                            ; sign > 0
 <PyLong_AsLong+95>: jg     0x806c231 <PyLong_AsLong+113> ; goto overflow
 <PyLong_AsLong+97>: lea    0x0(,%edx,2),%eax             ; ?? Where is (x << 1)?
 <PyLong_AsLong+104>: jne    0x806c231 <PyLong_AsLong+113> ; goto overflow
 <PyLong_AsLong+106>: mov    %edx,%eax
 <PyLong_AsLong+108>: imul   %ebx,%eax
 <PyLong_AsLong+111>: jmp    0x806c246 <PyLong_AsLong+134>

As you can see, the original code had add %eax, %eax instruction to shift
left by one, the optimized one doesn't.  I don't know what the heck it
was trying to do, and I don't know what condition codes "lea" sets, so
I can't know what it intended.  In any case, I speculate that line +104
was causing a jump to the overflow handler when it should not.

Anyway, I apologize for cluttering the newsgroup with assembly;  clearly the
problem is with pgcc w/ optimization turned on, which is a Mandrake 6.0 issue.

The lesson is NOT to trust pgcc with optimizations, and when compiling
Python 1.5.2, to turn them off (ie. make OPT=-g), at least when compiling
the longobject.c file.

Thanks again for all the help, it was at least a mildly interesting exercise to
track it down.  Sorry for the not double-checking the results of my last compilation
without optimizations.

Chad Netzer
chad at vision.arc.nasa.gov





From neelk at neelk  Sat Jul 10 03:53:46 1999
From: neelk at neelk (neelk at neelk)
Date: Sat, 10 Jul 99 07:53:46 GMT
Subject: Signature 0.1
Message-ID: <7m9tnh$6qg$59@rks1.urz.tu-dresden.de>

Signature.py is a module that makes reflecting on the function call
signatures of callable objects a lot easier. A common question in
Python is "How do I ask the interpreter what arguments a function
takes?" and while the information is obviously available, I haven't
found any simple, high-level interfaces to it.

That's why I wrote Signature -- it's useful enough that it's made it
into my personal library of frequently-used utilities.

You can find it at:

  http://www.sff.net/people/neelk/free-software/Signature.py

Caveats/TODO:

o I don't know if it will work on JPython -- I exploited a lot of
  undocumented object attributes to figure out what the call signature
  is, and I have no idea if the same data is available there.

o I haven't figured out how to look into C builtins yet -- anyone
  who can teach me how to do this from within Python will earn my
  undying gratitude. :)

Example of use:

    >>> def foo(x, y, z=-1.0, *args, **kw):
    ...     return (x+y)**z
    ...
    >>> f = Signature(foo)

    >>> print 'ordinary arglist:', f.ordinary_args()
    ordinary arglist: ('x', 'y', 'z')

    >>> print 'special_args:', f.special_args()
    special_args: {'keyword': 'kw', 'positional': 'args'}

    >>> print 'full arglist:', f.full_arglist()
    full arglist: ['x', 'y', 'z', 'args', 'kw']

    >>> print 'defaults:', f.defaults()
    defaults: {'z': -1.0}

    >>> print 'signature:', str(f)
    signature: foo(x, y, z=-1.0, *args, **kw)


Neel

<P><A HREF="http://www.sff.net/people/neelk/free-software/Signature.py">
Signature 0.1</A> - for reflecting on the function call signatures of
callable objects.  (09-Jul-99)

-- 
----------- comp.lang.python.announce (moderated) ----------
Article Submission Address:  python-announce at python.org
Python Language Home Page:   http://www.python.org/
Python Quick Help Index:     http://www.python.org/Help.html
------------------------------------------------------------



From bill at salome.rfa.org  Mon Jul 26 06:55:18 1999
From: bill at salome.rfa.org (Root Account)
Date: 26 Jul 1999 10:55:18 GMT
Subject: NT admin modules?
References: <379C0402.9DA8259D@hk.rfa.org>
Message-ID: <7nheqm$81n$1@salome.rfa.org>

I was thumbing through a Win32 Perl programming
book and found a lot of admin modules available
for the NT world - is there a similar assortment
for Python, or is someone working on porting these?

Thanks,
Bill
--
Bill Eldridge
Radio Free Asia
bill at hk.rfa.org



From William.Tanksley at p98.f112.n480.z2.fidonet.org  Sun Jul  4 17:35:43 1999
From: William.Tanksley at p98.f112.n480.z2.fidonet.org (William Tanksley)
Date: Sun, 04 Jul 1999 22:35:43 +0100
Subject: Tuples cf. Expanded objects??
Message-ID: <000017cf@bossar.com.pl>

From: wtanksle at dolphin.openprojects.net (William Tanksley)

On Sun, 04 Jul 1999 22:11:55 +1000, Gary Stephenson wrote:
>William Tanksley wrote:

>> On Fri, 02 Jul 1999 22:10:46 +1000, Gary Stephenson wrote:

>> An instance of an immutable type may be created in expanded form.

>Yes, but can they also  be created in an non-expanded form?  If not, then
>there is more of a relationship that your message implies - yes?

Of course they can -- that's how Eiffel works by default and how Python
works no matter what.

I wonder whether you know what "expanded" form means, though.  It means
that the object itself is stored in local storage, rather than a reference
to the object.  Of course, it's obvious that location of storage has
little to do with immutability of type (although it can take a lot of work
to store certain mutable types in local storage).

>gary

-- 
-William "Billy" Tanksley



From Juan.M..Gonzalez at p98.f112.n480.z2.fidonet.org  Fri Jul  2 11:22:13 1999
From: Juan.M..Gonzalez at p98.f112.n480.z2.fidonet.org (Juan M. Gonzalez)
Date: Fri, 02 Jul 1999 16:22:13 +0100
Subject: Waiting for an external process
Message-ID: <0000169b@bossar.com.pl>

From: "Juan M. Gonzalez" <jugonzal at globalserve.net>

Is there a better way to determine when an external process has finished
than using
win32process.CreateProcess / GetExitCodeProcess ?



From olivier.deckmyn at mail.dotcom.fr  Mon Jul 19 08:07:02 1999
From: olivier.deckmyn at mail.dotcom.fr (Olivier Deckmyn)
Date: Mon, 19 Jul 1999 14:07:02 +0200
Subject: [Q] Type checking...
References: <7mv000$sun$1@feed.teaser.fr> <slrn7p637a.4v2.kas@titan.magnetic-ink.dk>
Message-ID: <7mv4h7$c0$1@feed.teaser.fr>

Gloups !!!

There is no "nicer" way to know the type of an instance in Python
!!???!!!???

I can't make the "is" operator work as I would like to :(

Looks quite strange to me :-(

thx for your help anyway Klaus..

>
> MAGIC="I'm in class A :-)"
>
> class ClassA:
>   def __init__(self):
>     self.magic = MAGIC
>       :
>   # end def __init__
> # end class Class A
>






From aaron_watters at my-deja.com  Mon Jul 26 10:48:01 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Mon, 26 Jul 1999 14:48:01 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi>
Message-ID: <7nhseu$gf0$1@nnrp1.deja.com>

These points are all valid, but they can be deceptive.
I add some comments
(not meant to negate the valid observations):

In article <al81zdxkktf.fsf at sirppi.helsinki.fi>,
  Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
> Even in websites' case, I would worry; Python doesn't use multiCPU
very
> well, and with single CPU, it is fairly easy to saturate one with
> pure-Python code, especially if it is badly designed to boot.

Bad design aside.  If you have many CPU's for a web site you
might consider using plain old CGI scripts, since python would then
use all the cpu's.  In a single (multithreaded) process python will
only use one (thanks to the global interpreter lock).  Also, I
strongly feel CGI is more stable and reliable than other methods.

> Using a Gadfly database, _with_ pre-compiled queries: (very simple
> one-liners from table with one entry)
> select: 21.4182/sec [3.64s] (46.689ms/call)
> update: 14.3209/sec [3.98s] (69.828ms/call)

Note please that gadfly queries have a certain amount of fixed overhead.
You *might* find that queries over a table with one entry are not
much faster than queries over tables with 1000 entries, or maybe even
tables with 10000 entries that have appropriate indices.

However each gadfly query requires 10's of python function
calls regardless of how simple the query is.  But more complex queries
over more data don't necessarily require many more python function
calls (this was one of the design goals for the engine).

Also, I hope you have added the kjbuckets builtin...

On second thought 21 queries per second ain't too bad...
What are you comparing it to?  Updates are naturally slower
since they involve writes to a log file for recovery purposes.

> .. and so on. Even more distressing results lie in the speed of say,
> implementing protocols in Python (when we're talking _nearly_ realtime
> action).

Be careful here.  Protocol speed can be effected by a lot of stuff.
Some of it can be python and some of it can be other system overhead.
If a profiler or unix "top" says it's the python process and it's
implemented well, then you've got a point.  But until I see the
code I personally would withhold assignment of blame.

   -- Aaron Watters

===

less is more


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From cs at ctzen.com  Fri Jul  9 11:18:37 1999
From: cs at ctzen.com (chacs)
Date: Fri, 09 Jul 1999 15:18:37 GMT
Subject: Help! File Locking
References: <37856BF1.BD046003@freenet.co.uk>
Message-ID: <7m53s4$cmq$1@nnrp1.deja.com>

Hi All,

Need some help in file locking.

x1.py:
  import fcntl, FCNTL
  f = open('myfile', 'r+')
  fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
  input()
  x = str(int(f.readline()) + 1)
  f.seek(0)
  f.write(x)
  fcntl.flock(f.fileno(), FCNTL.LOCK_UN)
  f.close()

x2.py:
  import fcntl, FCNTL
  f = open('myfile', 'r+')
  fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
  x = str(int(f.readline()) + 1)
  f.seek(0)
  f.write(x)
  fcntl.flock(f.fileno(), FCNTL.LOCK_UN)
  f.close()

Both script are the same except x1 has an additional input() statement.

Basically, what the script does is to open a file, lock it and update
the first line to the next higher number.

My problem is (assume file content is "0"):

1. in shell 1, run x1.py, x1 got lock and block at input().
2. in shell 2, run x2.py, x2 block at flock().
3. in shell 1, input something to unblock input().
4. x1 update "0" to "1", unlock and exit.
5. x2 gets lock, BUT instead of getting the new file content (i.e. "1")
written by x1, it got the old content and update the file to "1" again
(instead of reading "1" and update to "2").

Any idea what is wrong ?

Thanx in advance.

C S Chang
cs at ctzen.com


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From joeld at anisci.com  Wed Jul 21 19:53:38 1999
From: joeld at anisci.com (Joel Davis)
Date: 21 Jul 1999 16:53:38 PDT
Subject: Tkinter question
Message-ID: <37965ED6.A4238A27@anisci.com>

I've got a tkinter canvas, and I want to get it's width.... I call

w = int( thecanvas.cget("width") )

this works fine.. except when I resize the window, this still returns
the width that the canvas was when it was created.. the canvas is
getting resized (it has a visible background color)..

any ideas?

joel

--
----------------------------------------------------------------
                                                joeld at anisci.com





From gduzan at gte.com  Thu Jul 15 10:21:33 1999
From: gduzan at gte.com (Gary Duzan)
Date: Thu, 15 Jul 1999 10:21:33 -0400
Subject: Problems using omniNames of omniORB2 from Fnorb Python ORB
References: <378DB759.DCE85A3E@eng.cam.ac.uk>
Message-ID: <378DEE6D.2317FE39@gte.com>

"D. Lopez-De-Ipina" wrote:
> 
> Hi,
> 
> First of all, apologies if you think this is not the forum to make
> questions about Fnorb, but I have posted my question into the Fnorb
> mailing list and I have not received any answer yet.

   There may be a problem with the list; I'm on it and I don't
recall seeing your message.
 
> [ ... ]
> 
> It turned out that the examples seemed to work perfectly as it was the
> case of my application, but if I left the server process running for a
> while (1-2 minutes) after a while crashed and 'said' the following:
> 
> [ ... ]
>
>   File "/home/dl231/python/Fnorb/orb/GIOPClientWorkerReactive.py", line
> 216, in
> close_connection
>     raise CORBA.TRANSIENT()
> Fnorb.orb.CORBA.TRANSIENT: Minor: 0 Completed: COMPLETED_NO
> 
> Please if anyboy has managed to interoperate omniORB objects with Fnorb
> ones in a safe way, could they please let me know and please explain
> what is going wrong? Any workaround for it?

   This is due to OmniORB's connection shutdown policy. After a
connection has been idle for a couple minutes (by default) it will
close the IIOP connection, which Fnorb sees as a CORBA.TRANSIENT
exception. It should be harmless, so you can to do something simple
like:

        while 1:
            try:
                boa._fnorb_mainloop()

            except CORBA.TRANSIENT:
                pass

            except KeyboardInterrupt:
                return 0


					Gary Duzan
					GTE Laboratories



From hyoon at bigfoot.com  Fri Jul 16 10:19:25 1999
From: hyoon at bigfoot.com (Hoon Yoon)
Date: Fri, 16 Jul 1999 10:19:25 -0400
Subject: Contractor need for integration prj (long)
Message-ID: <378F3F6D.E75E1395@bigfoot.com>

Hi,


  Our firm is looking to see if it is possible to wrap some of more
advanced lib 
in R like modreg.dll and specifically Loess function in there for both 
regression and prediction into MSVC++. We will probably take it to
either VB or 
Python for implementation issues (Guess which I prefer). Given that
locfit has 
many utitlities built in for data transformation, this would be a good
thing.
  If that's not possible, I would like to have dloess from Netlib
compiled under 
MSVC++, either linked by g77 or f2c or other fortran compiler.
  I am trying to make sure the final version is available to public. It
does not 
seem to be a major prb from our stand point, but that may change. If
developer 
wishes not to disclose his sources to public. That's fine with us. We,
however, 
need full sources, projects, obj, and exec files. Please submit bid with 
timeframe and qualification back to this e-mail.
  The objective is quickly implent Loess with VB or Python for
regression and 
prediction. I would prefer to use existing R dll's, because I can just
piggie 
back off of it everytime source changes or for new analytics; however,
quick 
timing is crucial. Any quick workable solution with high degree of
confidence 
would be chosen. I need to do multivariate regression with loess.
  We will also take this to Unix at some point.
  A pure Python version would be nice. I have a pure Gauss version. Only
prb with
it is that it's single variable only.
  We are Fortune 500 company.

  
S.Y.

REPLY to: hoon at bigfoot.com

R is available at: http://www.stat.cmu.edu/R/CRAN/
netLib w/ loess: http://www.wavelet.org/netlib/a/index.html

Gauss code: Pls don't just try translate this. This is single variable
only for
dependent variable. I providing it for reference. SAS also has Loess
built in.
/* LOESS - Local Regression and optional Robust Fitting and Symmetric
Errors
**
** (C) Copyright 1996  Aptech Systems, Inc.
** All Rights Reserved.
**
** This Software Product is PROPRIETARY SOURCE CODE OF APTECH
** SYSTEMS, INC.    This File Header must accompany all files using
** any portion, in whole or in part, of this Source Code.   In
** addition, the right to create such files is strictly limited by
** Section 2.A. of the GAUSS Applications License Agreement
** accompanying this Software Product.
**
** If you wish to distribute any portion of the proprietary Source
** Code, in whole or in part, you must first obtain written
** permission from Aptech Systems.
**
**-------------------**------------------**-------------------**-----------**
**-------------------**------------------**-------------------**-----------**
**
**> loess
**
**  Format:   { yhat,ys,xs } = loess(y,x)
**
**  Input:    y    Nx1 vector, dependent variable.
**
**            x    Nx1 vector, independent variable.
**
**  Output:   yhat    Nx1 vector, predicted depvar given indvars.
**
**            ys      numEvalx1 vector, ordinate values given abscissae
**                    values in xs.
**
**            xs      numEvalx1 vector, equally spaced abscissae values. 
**
**
**   Globals:
**
**   _loess_Span -- scalar, degree of smoothing, must be greater than
**                 2 / N.  Default = .67777.
**
**   _loess_NumEval -- scalar, number of points in ys and xs. 
**                     Default = 50.
**
**   _loess_Degree -- scalar, if 2, quadratic fit, otherwise linear.
**                    Default = 1.
**
**   _loess_WgtType -- scalar, type of weights.  Default weight type
**                 is Gaussian. If set to 1, robust, symmetric weights
**                 are used.
**
**   __ouput -- scalar, if 1 (default), iteration information and
results
**                 are printed, otherwise nothing is printed.  
**
**   Remarks:
**
**   Adapted from a program developed by Simon Jackson 
**     (acbsjack at gosnell.spc.uchicago.edu)
**
**   References:
**   William S. Cleveland. 1979. Robust Locally Weighted Regression and
**      Smoothing Scatterplots.  _JASA_. 74:829-836.
**   William S. Cleveland and Susan J. Devlin. 1988. Locally Weighted
**       Regression: An Approach to Regression Analysis by Local
**       Fitting. _JASA_. 83:596-609. [mainly on multivariate loess]
**   William S. Cleveland, Eric Grosse, and William M. Shyu. 1991.
**      Local Regression Models. in John M. Chambers and Trevor
**      J. Hastie (eds). _Statistical Models in S_. Pacific Grove:
**      Wadsworth.
**   Michael Friendly's routines in _SAS System for Statistical
**      Graphics, First Edition_. Cary, NC: SAS Institute.
**   Trevor J. Hastie and R.J. Tibshirani. 1990. _Generalized
**      Additive Models_. London: Chapman and Hall.
*/

#include loess.ext;


proc(3) = loess(y,x);

    local
data,n,q,iter,i,res,wts,b,yhat,dist,s,near,nx,ny,d,delta,ystar,
        xstar,u,x0,x1,x2,y1,y2,loind,upind,singlist,slope,dx,step,fixx,
        sing,flag2,mask,fmt,unit,flag,oldfmt;


    n = rows(x);
    unit = seqa(1,1,n);
    if floor(_loess_Span*n) < 2;        /* check for valid _loess_Span
*/
        if not trapchk(1);
            errorlog "Invalid value for _loess_Span";
            end;
        else;
            retp(error(0),error(0),error(0),error(0));
        endif;
    endif;

    if __output;
        if _loess_WgtType == 1;
            print "Gaussian weights";
        else;
            print "Symmetric weights";
        endif;
        if _loess_Degree == 1;
            print "Locally linear";
        else;
            print "Locally quadratic";
        endif;
        oldfmt = sysstate(19,0);
        format 4,2;
        print "_loess_Span = " _loess_Span;
        format 3,0;
        print "Number of evaluation points " _loess_NumEval;
        call sysstate(19,oldfmt);
    endif;

    data = sortc(x~y,1);    /* sort data first */
    x = data[.,1];
    y = data[.,2];

    yhat = zeros(n,1);
    delta = ones(n,1);      /* initialize robustness weights */
    res = y;        /* initialize residuals */
    s = ones(n,1);          /* initialize neighbor ranks */
    singlist = zeros(n,1);          /* initialize singularity
indicators  */

    xstar =
seqa(minc(x),(maxc(x)-minc(x))./(_loess_NumEval-1),_loess_NumEval);
    yhat = ones(n,1);

    iter = 1;
    do until iter > _loess_WgtType; /* iterations start 2 iterations for
                                      :: symmetric, 1 for Gaussian
                                      */
        if __output;
            print "Iteration "$+ftos(iter,"*.*lf",1,0);
        endif;
        i = 1;
        do until i > n;    /* loop over x values */
            flag2 = 0;      /* initialize singularity flag */
        loop:

            if _loess_Span <= 1;
                q = floor(_loess_Span.*n);
            else;
                q = n;
            endif;          /* number of nearest neighbors */
            x0 = x[i];      /* reference obs for this iteration */
            dist = abs(x-x0);       /* distance to other points */
            s = sortc(x~y~unit~dist,4);     /* identify sorted
distances  */
            s = s[1:q,.];
            nx = s[.,1];
            ny = s[.,2];    /* q nearest neighbors of x0 */
            d = maxc(s[.,4]);       /* distance to furthest nearest  */
            if _loess_Span > 1;
                d = d.*_loess_Span;
            endif;          /* p314, CGS */
            if d > 0;
                u = abs(nx-x0)./d;
                wts = (u .< 1).*((1-u^3)^3);       /* tri-cube
neighborhood
                                                    :: wts
                                                    */
                wts = delta[s[.,3]].*wts;           /* symmetric
robustness
                                                    :: weights
                                                    */
                if sumc(wts[2:q]) > .0001;

                    if _loess_Degree == 2;
                        b =
(ny.*wts)/((ones(rows(nx),1)~nx~nx.*nx).*wts);
                    else;
                        b = (ny.*wts)/((ones(rows(nx),1)~nx).*wts);
                    endif;
                        
                    if scalmiss(b);
                        if __output;
                            oldfmt = sysstate(19,0);
                            format 1,0;
                            print "Singularity obs " i;;
                            format 8,4;
                            print " x = " x0;;
                            print " y = " y[i];
                            call sysstate(19,oldfmt);
                        endif;
                        q = n;
                        if flag2 /= 1;
                            flag2 = 1;
                            goto loop;
                        endif;
                        singlist[i] = 1;
                    endif;
                    if _loess_Degree == 2;
                        yhat[i] = (1~x0~(x0^2))*b;          /* smoothed
value
                                                            :: - quad
                                                            */
                    else;
                        yhat[i] = (1~x0)*b;         /* smoothed value -
                                                    :: linear
                                                    */
                    endif;
                else;
                    yhat[i] = y[i];
                endif;
            else;
                yhat[i] = meanc(ny);
                if __output;
                    oldfmt = sysstate(19,0);
                    format 1,0;
                    print "x(0) = 0,  obs " i;;
                    format 8,4;
                    print " x = " x0;;
                    print " y = " y[i];
                    call sysstate(19,oldfmt);
                endif;
                singlist[i] = 1;
            endif;
            i = i + 1;
        endo;

        if sumc(singlist) /= 0;    /* singularity fixup - interpolation
*/

            if __output;
            print "Singularity fixup (interpolation) for "
sumc(singlist) ""\
                " obs";
            endif;
            fixx = indexcat(singlist,1);
            i = 1;
            do until i eq rows(fixx);
                if fixx[i] == 1 or fixx[i] == n;
                    yhat[i] = y[i];
                    i = i+1;
                    continue;
                endif;
                dx = x[fixx[i]]-x[fixx[i]-1];
                flag = 0;
                step = 0;
                do until flag == 1;
                    step = step+1;
                    if scalerr(indexcat(fixx,fixx[i]+step)) == 13;
                        flag = 1;
                    endif;
                endo;
                slope = (yhat[fixx[i]+step]-yhat[fixx[i]-1]);
                slope = slope./(x[fixx[i]+step]-x[fixx[i]-1]);
                yhat[fixx[i]] = yhat[fixx[i]-1]+(slope.*dx);
                i = i + 1;
            endo;
        endif;

        res = y - yhat;       /* residuals */
        if _loess_WgtType == 2;   /* get robustness weights */
            u = res./(6*median(abs(res)));
            delta = (abs(u) .< 1) .* ((1 - u.*u)^2);
        endif;
        iter = iter + 1;
    endo;

    if __output;
        if _loess_WgtType == 2;
            print "Obs    X      Y     Fitted Y    Weights  Residuals";
            mask = { 1 1 1 1 1 1 };
            let fmt[6,3] = "*.*lf " 3 0 "*.*lf" 8 3 "*.*lf" 8 3 "*.*lf"
8
                3 "*.*lf" 8 3 "*.*lf" 8 3 ;
            call printfm(seqa(1,1,n)~x~y~yhat~delta~res,mask,fmt);
        else;
            print "Obs    X      Y     Fitted Y    Residuals";
            mask = { 1 1 1 1 1 };
            let fmt[5,3] = "*.*lf " 3 0 "*.*lf" 8 3 "*.*lf" 8 3 "*.*lf"
8
                3 "*.*lf" 8 3 ;
            call printfm(seqa(1,1,n)~x~y~yhat~res,mask,fmt);
        endif;
    endif;

/* I N T E R P O L A T I O N */
    if _loess_NumEval == n or abs(_loess_NumEval-rows(unique(x,1))) < 2;
        xstar = unique(x,1);
        ystar = yhat[uniqindx(x,1)];
        goto done;
    endif;
    ystar = ones(_loess_NumEval,1);
    ystar[1] = yhat[1];
    ystar[_loess_NumEval] = yhat[n];         /* fix end points */
    i = 2;
    do until i == _loess_NumEval;
        d = x-xstar[i];
        upind = minindc(recode(d,d .< 0,maxc(d)+1));
        loind = maxindc(recode(d,d .> 0,minc(d)-1));
        x1 = x[loind];
        x2 = x[upind];
        y1 = yhat[loind];
        y2 = yhat[upind];
        ystar[i] = y1+((y2-y1)./(x2-x1).*(xstar[i]-x1));
        i = i+1;
    endo;
done:

    retp(yhat,ystar,xstar);
endp;
-------------- next part --------------
A non-text attachment was scrubbed...
Name: vcard.vcf
Type: text/x-vcard
Size: 202 bytes
Desc: Card for Hoon Yoon
URL: <http://mail.python.org/pipermail/python-list/attachments/19990716/33ab627f/attachment.vcf>

From arcege at shore.net  Mon Jul  5 21:32:05 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Tue, 06 Jul 1999 01:32:05 GMT
Subject: Python under Java?
References: <MrUJKLEB0KGWFViX4p@holmes.parc.xerox.com>
Message-ID: <p0dg3.895$6M6.293143@news.shore.net>

Bill Janssen <janssen at parc.xerox.com> wrote:
: Has anyone tried embedding Python under Java, using JNI?  When I try it
: (with JDK 1.2 on Solaris 2.6 and Python 1.5), I get SEGV inside
: Py_Initialize().

I don't think anyone has trying to _embed_ Python in Java.  But Python
has been _implimented_ with Java (instead of and in addition to C).
Look at JPython (http://www.jpython.org/).

  -Arcege




From flight at mathi.uni-heidelberg.de  Wed Jul 14 05:45:25 1999
From: flight at mathi.uni-heidelberg.de (Gregor Hoffleit)
Date: Wed, 14 Jul 1999 11:45:25 +0200
Subject: History of MAGIC number ?
Message-ID: <19990714114525.A739@hoffleit.org>

I'd like to submit a patch to the maintainer of the Linux file command
so that file recognizes Python .pyc and .pyo files.

.pyc and .pyo files are quiet easily recognizable on their magic
number, a long word at the start of the file. The format of the long
word is 32bit little endian, with the higher 16 bits set to 0x0a0d and 
the lower 16 bits set to a pyc version number (20121 at the moment
in Python 1.5.2).

Now if I want to write a robust test method for file I could either
test for 0x0a0d in position 2 (which is not very rigid), or I had to
list the explicit versions, like

  0  lelong   0x0a0d4e99    compiled Python code, magic version 20121

Now the problem is that this verion number has changed quite a few
times in Python's history, but I can't find a list of the ancient
values anywhere (Misc/HISTORY only lists the times changes took place, 
but not the values).

Btw, should I say "Python bytecode" or "compiled Python code" ?

	Gregor



From fdrake at cnri.reston.va.us  Fri Jul  9 17:28:17 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 9 Jul 1999 21:28:17 GMT
Subject: Grail won't work
In-Reply-To: <3786527C.6230C4D2@ssec.wisc.edu>
References: <3786527C.6230C4D2@ssec.wisc.edu>
Message-ID: <14214.26993.614685.151897@weyr.cnri.reston.va.us>

Nick Bower writes:
 > I just installed Grail 0.6 to take a look.  I have Py1.5 and Tk8 but it
 > won't run:

  Can either try it with Python 1.5.2.  I thought there was a patch on 
the Web site, but I don't see it now.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From Gordon.McMillan at p98.f112.n480.z2.fidonet.org  Sat Jul  3 22:43:49 1999
From: Gordon.McMillan at p98.f112.n480.z2.fidonet.org (Gordon McMillan)
Date: Sun, 04 Jul 1999 03:43:49 +0100
Subject: ordered threading
Message-ID: <000016c0@bossar.com.pl>

From: "Gordon McMillan" <gmcm at hypernet.com>

Robin Becker writes:

> In article <000a01bec583$30332b40$769e2299 at tim>, Tim Peters
> <tim_one at email.msn.com> writes

> >[Robin Becker]
> >> I have an unbounded number of worker threads which occasionally require
> >> access to a unique server thread. I can easily mutex access to the
> >> server thread, but is there any obvious/easy way to ensure that the
> >> blocked workers wake up in the right order?
> >
> >No threading system on earth allows specifying the order in which blocked
> >threads will wake up, although some systems (not Python's) allow you to
> >*influence* the order via assorted "priority" gimmicks.

> the workers are HTTP requests in separate threads from Zope I guess.
> I want to serialise access to the time/space hog. I guess it could
> run uniquely in the worker threads ie the locked resource would be
> access to the hog code. I'm not sure that would let me guarantee
> unlocking as there are ways for the worker threads to suicide.

If that's what you want, you don't really need to order the 
unblocking. Just give each thread an input queue protected by a 
Condition. Worker threads put requests on the Hog thread's input 
queue (including a ref to the return queue). If you find that some 
worker threads seem to get better access to the Hog's queue than 
others, a time.sleep(0.001) will probably fix it - since it will 
pretty much force a context switch.


- Gordon




From greg.ewing at compaq.com  Sun Jul 25 16:35:53 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Mon, 26 Jul 1999 08:35:53 +1200
Subject: Maintainability (was Re: Stackless & String-processing)
References: <001101bed3f2$5e8874a0$642d2399@tim>
			<7n675p$qsk@dfw-ixnews7.ix.netcom.com>
			<14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>
			<379798D9.FF3F811B@compaq.com> <14232.32203.792051.822513@weyr.cnri.reston.va.us> <37988C11.DB7D67C7@callware.com>
Message-ID: <379B7529.B3C4AD0F@compaq.com>

Ivan Van Laningham wrote:
> 
> Any reason it couldn't be done?

If I remember rightly, Guido's position last time
was that doing it the long-winded way wasn't enough
of a hardship to make it worth adding another language
construct. Not to mention another keyword, which as
we all know would cause the universe to collapse into
a quantum singularity if done prior to Python 2.0.

Greg



From alex at somewhere.round.here  Sun Jul 25 16:47:20 1999
From: alex at somewhere.round.here (Alex)
Date: 25 Jul 1999 16:47:20 -0400
Subject: simplebeginnerquestion(**args)
References: <379afff7.131877652@news.mindspring.com>
Message-ID: <etd4sisfodj.fsf@m2-225-11.mit.edu>

Python 1.5.1 (#1, Jan 31 1999, 18:55:21)  [GCC 2.8.1] on irix6
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> def keyword_args_demo(**args): print args
... 
>>> keyword_args_demo(first_argument=1,second_argument=2)
{'first_argument': 1, 'second_argument': 2}
>>> 

You have to explicitly say what each of the keyword arguments are
called.  If you don't want to name each argument, use a single asterix:

>>> def arbitrary_args_demo(*args): print args
... # Note the single asterix
... 
>>> arbitrary_args_demo(1,2,3,4)
(1, 2, 3, 4)

Alex.

-- 
I have tried to write the best I can; sometimes I have good luck and can
write better than I can.  --Ernest Hemingway



From aahz at netcom.com  Mon Jul  5 21:48:30 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 6 Jul 1999 01:48:30 GMT
Subject: What's ZOPE anyway?
References: <3scg3.1267$B95.91660@typ42b.nn.bcandid.com>
Message-ID: <7lrn9e$bot@dfw-ixnews15.ix.netcom.com>

In article <3scg3.1267$B95.91660 at typ42b.nn.bcandid.com>,
Jayr Magave <jmagave at amazon.com.br> wrote:
>
>I am sorry, but what's ZOPE? :-)

Basically, Zope is an architecture for creating web-based applications
through Python code.  Think of it as a competitor for Cold Fusion,
though it's capable of much more.  To see Zope in action, take a look at
http://www.searchbutton.com/ ; except for the actual search component,
almost everything is in Zope.  Do note that overall we're currently
using a rather limited subset of Zope, partly due to having started with
one of the predecessor pieces (Bobo).
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From ajung at sz-sb.de  Fri Jul 16 06:52:30 1999
From: ajung at sz-sb.de (Andreas Jung)
Date: Fri, 16 Jul 1999 10:52:30 GMT
Subject: sgmllib has problems with dots in tag names
Message-ID: <19990716125230.A14994@sz-sb.de>

The SGML parsers from the standard sgmllib and the XML sgmllib war both
unable to parse SGML tags with dots in the tag name like <TI.DOC>. The 
parsers callback functions only get the first part of the tag name (before
the dot) as argument (in this case 'TI'). Because the tags are valid SGML
tags this is a bit annoying. Ok, one could get a workaround by replacing
all dots in tags with an underscore however that's not a clean solution :-)

Any others ideas ?

Thanks,
Andreas




From mwh21 at cam.ac.uk  Mon Jul 26 08:35:11 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: Mon, 26 Jul 1999 13:35:11 +0100
Subject: mutating python VM opcodes
In-Reply-To: <14236.3528.398388.925726@lrz.de>
References: <14236.3528.398388.925726@lrz.de>
Message-ID: <Pine.LNX.4.10.9907261332080.17987-100000@atrus.jesus.cam.ac.uk>

On Mon, 26 Jul 1999, Eugene Leitl wrote:
> What's Python's reaction when presented with random bytecode sequence?
> Slight mutation of a valid bytecode sequence?
> 
> Core dump, error message, something else?
> 

Depends. If eval_code2 comes accross an invalid opcode it will throw a
SystemError exception. If, however, the opcode is valid but the argument
is corrupted, then a core dump is quite possible. A

LOAD_CONST 30000

will almost certainly provoke a segfault.

A "JUMP_FORWARD 30000" probably isn't too heathly either.

HTH
Michael




From skaller at maxtal.com.au  Tue Jul 27 13:12:26 1999
From: skaller at maxtal.com.au (John Max Skaller)
Date: Tue, 27 Jul 1999 17:12:26 GMT
Subject: unicode data tables
References: <mt2.0-21715-933088951@news.informatik.uni-bonn.de>
Message-ID: <379de856.1151461113@news.triode.net.au>

On Tue, 27 Jul 99 15:22:31 GMT, skaller at maxtal.com.au (John (Max) Skaller) wrote:

>I have posted the unicode (and other) encoding subdocument of
>interscript v1.0a9 to the Web. At
>
>	http://www.xenon.triode.net.au/~skaller/unicode/index.html

	Sigh. Sorry, that should be

	http://www.triode.net.au/~skaller/unicode/index.html

John Max Skaller                ph:61-2-96600850              
mailto:skaller at maxtal.com.au       10/1 Toxteth Rd 
http://www.maxtal.com.au/~skaller  Glebe 2037 NSW AUSTRALIA



From tim_one at email.msn.com  Sun Jul 25 23:57:37 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 25 Jul 1999 23:57:37 -0400
Subject: Cyclops 0.9.4
In-Reply-To: <379B809A.CB35D2D8@compaq.com>
Message-ID: <000001bed71b$00e74d40$492d2399@tim>

[Greg Ewing]
> ...
> I can look for certain kinds of object, though. If
> I open ten BlargWindows and then close them, I know
> there shouldn't be any BlargWindow instances left.
> If there are, I can take a closer look at them to
> find out what what's happening.

Perfectly reasonable.  Under Cyclops <wink>, register those 10 by hand, or
put a .register(self) call in BlargWindows.__init__.

> ...
> I'm not asking to be told about every possible way
> of reaching every live object, obviously!

That you weren't asking for that was obvious, but what you were asking for
wasn't.  Now it's clearer:

> What I mean is, once I've found (by whatever means) a live object that
> I know should be dead, I should be able to say "show me a path from the
> root set to this object". I don't need all paths, only one of them --
> somewhere along that path there will be a link that shouldn't be there,
> and it should be fairly easy to spot.

Sounds like a good idea to me too.

>> All objects ever allocated, or all objects whose refcounts haven't yet
>> fallen to 0?

> The latter.

>> Since "even ints are boxed" in Python, that's a major memory hit.

> Which is why I'm happy for it to be a debugging option,
> maybe requiring a different interpreter. Although if M&S
> is ever added, we're going to have to live with this
> overhead all the time -- or find a smarter way to handle
> it.

Guido suggested focussing on dicts; perhaps "containers" is sufficient
(being an object that contains a pointer -- doesn't seem any loss to exempt
ints, strings, floats, etc).

>> Let's say you have everything you want, and even more: ...

> Tim, this is expandio ad absurdum! I never asked for all that
> information.

Sure.  My point was that even with perfect & complete information, the hard
problems remain.

> All I said whas that it would be handy to have a way to discover the
> existence of dead objects taking up room in my heap.  I never expected
> the idea to be this controversial...

Na, *every* idea is *endlessly* controversial until someone actually
implements it; then it becomes an endless source of complaints <wink>.

You're absolutely right that Python has no way to tell you this now, with or
without Cyclops, and that it would be of some help.  "Purify"-like tools can
do that for C, and I was blue-skying because such tools have never been
*enough* of a help.  More info or better info is needed, but at this time we
can't get either.

dreamily y'rs  - tim





From donn at u.washington.edu  Thu Jul 29 00:43:27 1999
From: donn at u.washington.edu (Donn Cave)
Date: 29 Jul 1999 04:43:27 GMT
Subject: Problem with signals & exceptions...
References: <2BB05D8530326D0C.420888E8906528AF.C6BF325D12E9ED90@lp.airnews.net> <7nnbme$ns2$1@nntp6.u.washington.edu> <DA4B3986E94F79C9.E3C10FB44F334D06.D3443001EBA1A2FA@lp.airnews.net>
Message-ID: <7nom5f$1bms$1@nntp6.u.washington.edu>

Eric Lee Green <eric at estinc.com> writes:

| Donn Cave wrote:
|> 
|> Eric Lee Green <e_l_green at hotmail.com> writes:
|>| I'm having trouble catching my thrown exceptions under FreeBSD 3.2. I am
|>| using the Python 1.5.2 that came with FreeBSD 3.2 (well, on disk 1 of
|>| the <n>-disk set). The following program works properly under Red Hat
|>| Linux version 6.0 running a self-compiled Python 1.5.2 (in place of the
|>| 1.5.1 which comes with Red Hat 6.0, which does not properly handle
|>| signals during socket i/o calls). How it works: it listens to port 6664.
|>| When it gets a connection, it then sets an alarm handler and alarm and
|>| then tries to read from the socket. If it does not get data within 10
|>| seconds, it then times out, at which point the alarm handler throws an
|>| exception and aborts the socket read.
|> 
|> I tried this exception-from-handler notion on Digital UNIX and BeOS,
|> and it does seem to work the way I guess you expect, but I do find it
|> weird. A signal handler executes outside the normal flow of control,
|> so it's not obvious to me where its exceptions ought to surface.
|> Probably it should be expected to work the same way on various platforms
|> for any given version of Python, but would future versions of Python be
|> guaranteed to work this way also?
|
| That, of course, is the ten million dollar question. The operation of signals
| in "C" is extremely well documented and described. Under Python, it's ???. 

Well, it's not capricious, there are some limits to what can reasonably
be done with signals in an interpreter like Python.  In normal use the
main thing to remember is that signal handlers execute in the same thread,
and therefore start on instruction boundaries.  A C signal handler executes
on a machine instruction granularity, a Python signal handler executes on
a Python instruction granularity ... which could be a long time.  It doesn't
look like you've run into this here, though.  As another followup points
out, signal handlers are very limited.  Can you throw a C++ exception from
a handler, and catch it in the interrupted section?  Don't think so.

| I changed my code to set a global variable rather than throw an exception and
| that works the way you'd expect. A socket.error exception is thrown, and I
| catch it, check the global variable, if the global variable is set I print "go
| a timeout!" else re-throw the exception. 

That should be safe.

| In the actual working code I scrapped the alarm signal altogether and instead
| do a select() call with a timeout, which presumably is guaranteed to work the
| same on all platforms (it better, the select call is a POSIX standard!). 

Hm, don't think select() is POSIX 1003.1.  Occasionally it can be found
supported only for socket devices, another Berkeley-ism that POSIX takes
a characteristically dim view of.  But this application should be no problem.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From ScherBi at BAM.com  Tue Jul  6 15:14:43 1999
From: ScherBi at BAM.com (ScherBi at BAM.com)
Date: Tue, 6 Jul 1999 15:14:43 -0400 
Subject: python core dump (SIGBUS) on Solaris
Message-ID: <3F8BA119B48AD11197470008C7A493CB98B443@HQOBGEX1>

I have a bus error problem with this module too.  Some cookies cause the
error when trying to create an instance for reading cookies; eg:

ck = Cookie.Cookie(os.environ['HTTP_COOKIE'])

Not all cookies cause the error.  If I copy a suspect cookie string into a
variable in an interactive session and try to create an instance as above,
Python dumps core with a Bus Error.

I've been able to create a work-around for now by parsing the cookie string,
but just wanted to say that Mr. Pennington is not alone in this...

PS.
This is also on Solaris 2.6, compiled with gcc 2.8.1

Bill Scherer

-----Original Message-----
From: Gary Pennington - UK Performance Centre
[mailto:Gary.Pennington at uk.sun.com]
Sent: Tuesday, July 06, 1999 12:34 PM
To: python-list at python.org
Subject: python core dump (SIGBUS) on Solaris


Hi,

I've encountered a fairly unusual problem using python on Solaris.
Before I investigate further, has anyone else already resolved this one?

Env:
Solaris 2.6 (all latest relevant patches)
compiler egcs-2.91.66

Test case to reproduce core is pretty straightforward :-

python
>>> import Cookie
>>> C=Cookie.Cookie("webJudgeUser=garyp;")
>>> C

At this point when using python interactively, a sigbus is generated.

Here's some trace from gdb

foucalt:/home/garyp 82 $ gdb python
GNU gdb 4.18
Copyright 1998 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you
are
welcome to change it and/or distribute copies of it under certain
conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for
details.
This GDB was configured as "sparc-sun-solaris2.6"...
(gdb) run
Starting program: /usr/local/bin/python
[New LWP    2        ]
[New LWP    3        ]
Python 1.5.2 (#1, Jul  6 1999, 16:50:14)  [GCC egcs-2.91.66 19990314
(egcs-1.1.2  on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
[New LWP    4        ]
>>> import Cookie
>>> C=Cookie.Cookie("webJudgeUser=garyp;")
>>> C

Program received signal SIGBUS, Bus error.
0xef2c5ab8 in _smalloc () from /usr/lib/libc.so.1
(gdb) info threads
  7 Thread 3          0xef647c14 in _swtch () from
/usr/lib/libthread.so.1
  6 Thread 2 (LWP 2)  0xef2b98d0 in _signotifywait () from
/usr/lib/libc.so.1
* 5 Thread 1 (LWP 1)  0xef2c5ab8 in _smalloc () from /usr/lib/libc.so.1
  4 LWP    4          0xef2b973c in ___lwp_cond_wait () from
/usr/lib/libc.so.1
  3 LWP    3          0xef2b699c in door_restart () from
/usr/lib/libc.so.1
  2 LWP    2          0xef2b98d0 in _signotifywait () from
/usr/lib/libc.so.1
  1 LWP    1          0xef2c5ab8 in _smalloc () from /usr/lib/libc.so.1
(gdb) where
#0  0xef2c5ab8 in _smalloc () from /usr/lib/libc.so.1
#1  0xef2c5b00 in malloc () from /usr/lib/libc.so.1
#2  0x4c094 in PyString_FromString (str=0x9d698 "<stdin>")
    at stringobject.c:145
#3  0x4d8d8 in PyString_InternFromString (cp=0x9d698 "<stdin>")
    at stringobject.c:1098
#4  0x2a434 in jcompile (n=0x0, filename=0xe0e90 "", base=0xe0ed8)
    at compile.c:3423
#5  0x2a328 in PyNode_Compile (n=0x11d938, filename=0x9d698 "<stdin>")
    at compile.c:3382
#6  0x32f94 in run_node (n=0x11d938, filename=0x9d698 "<stdin>",
    globals=0xe5b30, locals=0xe5b30) at pythonrun.c:883
#7  0x322f8 in PyRun_InteractiveOne (fp=0xffffffff, filename=0x9d698
"<stdin>")
    at pythonrun.c:528
#8  0x32120 in PyRun_InteractiveLoop (fp=0xd7430, filename=0x9d698
"<stdin>")
    at pythonrun.c:472
#9  0x32038 in PyRun_AnyFile (fp=0xd7430, filename=0x9d698 "<stdin>")
    at pythonrun.c:449
#10 0x1dfc4 in Py_Main (argc=1, argv=0xefffe844) at main.c:287
#11 0x1d9e4 in main (argc=1, argv=0xefffe844) at python.c:12
(gdb)


You can see that the problem is being manifested in _smalloc.

I started thinking about what the problem could be and I think it may be
related to the mymalloc routines which are used in the pypcre routines.
I seem to recall encountering problems before when there are multiple
malloc functions being used.

Does an instant remedy suggest itself to anyone?

Investigations to date:-

Rebuilt python without using threads - MADE NO DIFFERENCE
Tried different combinations of regex and it seems to fail only when 31
bytes are required.


Yours,

Gary Pennington
Sun UK


-- 
http://www.python.org/mailman/listinfo/python-list



From graham at sloth.math.uga.edu  Mon Jul 19 12:15:57 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 19 Jul 1999 16:15:57 GMT
Subject: [Q] Type checking...go further...
References: <7mv9n8$md$1@feed.teaser.fr>
Message-ID: <7mvivt$cpj$3@cronkite.cc.uga.edu>

Olivier Deckmyn (olivier.deckmyn at mail.dotcom.fr) wrote:
: Thank you all for your answers !
: I finally use isinstance in the methods of my classes...
: 
: But to go further, is there a way in python to have typed methods parameters
: ?
: aka :

No there is no way. You cannot statically check types in Python. You
must use dynamic check as per "isinstance".

graham

-- 
            The girls is all salty, the boys is all sweet, 
                      the food ain't too shabby 
                    an' they piss in the streets
    In France, way down in France, way on down, way on down in France



From Michael.P..Reilly at p98.f112.n480.z2.fidonet.org  Thu Jul  1 11:37:34 1999
From: Michael.P..Reilly at p98.f112.n480.z2.fidonet.org (Michael P. Reilly)
Date: Thu, 01 Jul 1999 16:37:34 +0100
Subject: ioctl in python
Message-ID: <0000161c@bossar.com.pl>

From: "Michael P. Reilly" <arcege at shore.net>

Angus MacKay <amackay at starvision.com> wrote:
: hun?

: why is this expecting a 4 byte buffer? my C version uses an int*
: and it works no problem.

: as for 80 being a input I think you have it backwards (for linux-i386 at
least):
: 0x80046601  EXT2_IOC_GETFLAGS      int *
: 0x40046602  EXT2_IOC_SETFLAGS      const int *
: 0x80047601  EXT2_IOC_GETVERSION         int *
: 0x40047602  EXT2_IOC_SETVERSION         const int *

: cheers, Angus.

The four bytes is the sizeof(int *), but also (according to some linux
headers I was looking at) it is expecting a pointer to a long, not an
int.  A long is a 4 byte value (not to mention that an int is also a
long on most systems).

But, the ioctl call takes some 4 byte value, it could be cast from a
char, short, pointer or whatever; it is up to the called request to
interpret the value (int *, in this case) given.  In your original
(Python) program, you gave ioctl() a fairly short character string
(three bytes if I remember correctly), which would not be the size of a
long.

  -Arcege



From roy at popmail.med.nyu.edu  Fri Jul  9 08:26:45 1999
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Fri, 09 Jul 1999 08:26:45 -0400
Subject: Two Pythons talking to each other?
References: <199907080019.TAA22244@buffalo.fnal.gov> <199907080529.HAA19502@axil.hvision.nl> <3784486a.7993113@news.ricochet.net> <7m2lol$bnm$1@nntp6.u.washington.edu> <37859856.9981682@news.ricochet.net>
Message-ID: <roy-0907990826450001@qwerky.med.nyu.edu>

phil at ricochet.net (phil) wrote:
> Yes, if you are stated or you care about order then you probably
> need to sequence.

If you need to sequence datagrams, then you most likely might as well just
be using TCP instead of UDP.  It is extremely unlikely that any sequencing
you will do in user code will be faster/smaller/better/less-buggy/whatever
than the kernel code that already does it for you.

-- 
Roy Smith <roy at popmail.med.nyu.edu>
New York University School of Medicine




From Gerrit.Holl at p98.f112.n480.z2.fidonet.org  Fri Jul  2 12:16:28 1999
From: Gerrit.Holl at p98.f112.n480.z2.fidonet.org (Gerrit Holl)
Date: Fri, 02 Jul 1999 17:16:28 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001667@bossar.com.pl>

From: Gerrit Holl <gerrit at humbolt.nl.linux.org>

On Fri, Jul 02, 1999 at 11:56:10AM +0000, lvirden at cas.org wrote:
> From:	lvirden at cas.org
> Newsgroups: comp.lang.python
> Subject: Re: Python GUIs: Abandoning TkInter and welcoming wxPython?
> Date:	2 Jul 1999 11:56:10 GMT
> To:	python-list at python.org
> 
> According to Mikael Lyngvig <mikael at pobox.com>:
> :Well, TkInter is extremely slow, but not because of the Python part.
> :Even the Tk widget tour, included with Tk, is so slow that you almost
> :think "Please, draw the next character - you can do it." whenever a
> :window with some text needs to be redrawn.
> 
> When I run the Tk test suite  on my SPARC, Tk cranks thru creating and
> deleting windows as fast as the eye can track.
> 
> However, there are reports that the Windows version of Tk, because of its
> use of the native window widgets, is noticably slower than the X
> implementation.  Is that the platform to which you refer?

I think you should also say what kind of processor you have if you're
talking about this:

"Windows NT on a Pentium IV 500 MHZ runs over 10 times faster than a
Linux 386 30 MHZ" (is it?)

regards,
Gerrit.




From amk at mira.erols.com  Fri Jul  9 19:34:12 1999
From: amk at mira.erols.com (A.M. Kuchling)
Date: 9 Jul 1999 23:34:12 GMT
Subject: PyApache and Apache 1.3.6 anyone ???
References: <19990708221333.B16851@hoffleit.org> <3785836D.3BE3EBC7@cs.ucsd.edu> <19990709112714.A2615@hoffleit.org>
Message-ID: <slrn7od252.dl.amk@207-172-57-127.s127.tnt2.ann.va.dialup.rcn.com>

On Fri, 9 Jul 1999 11:27:14 +0200, 
	Gregor Hoffleit <flight at mathi.uni-heidelberg.de> wrote:
>On Thu, Jul 08, 1999 at 10:06:53PM -0700, Pavel Velikho wrote:
>Nope, there's no entry at all in the error log for all the silently
>dying starting attempts. I don't know how to debug apache with gdb,
>therefore I'm comletely lost where to look for hints about the
>failure.

This is probably because Apache is segfaulting for some reason,
something that's become very familiar since I started working on
mod_pcgi.  To debug an Apache binary with GDB, run "gdb httpd", and
then run Apache with the -X command-line argument.  -X disables
forking, so everything is handled in a single process; you may also
want to specify "-f <configfile".  Then you can set breakpoints, get
stack traces at segfaults, and all that sort of thing.  

	Running Apache under GDB is really slow, though; an
application such as Zope that makes lots of little HTTP requests will
run glacially slowly.  That shouldn't be a problem for getting
PyApache working.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
It is a rotten world / Artful politicians are its bane / Its saving grace is
the / Artlessness of the young / And the wonders of the sky.
    -- Epitaph, Ross Bay Cemetery, Victoria



From jam at newimage.com  Thu Jul  1 21:10:34 1999
From: jam at newimage.com (jam)
Date: Thu, 1 Jul 1999 21:10:34 -0400
Subject: please help with simple regex
In-Reply-To: <7lh1h6$t84$1@nnrp1.deja.com>; from viofis@my-deja.com on Fri, Jul 02, 1999 at 12:35:51AM +0000
References: <7lh1h6$t84$1@nnrp1.deja.com>
Message-ID: <19990701211034.C19548@toast.internal>

On Fri, Jul 02, 1999 at 12:35:51AM +0000, viofis at my-deja.com wrote:
> 
> I'm a beginner at Python, so this may be a stupid question:
> 
> I have a string, for example, myString="the Potato".
> 1. Convert it to all lowercase ("the potato")
> 2. Match it against a list of strings - "potato,tomato,corn,etc...",
> extract "potato" (if it matches with something on the list), then save
> into myNewString.
> 
> Any help is greatly appreciated.
> 
> thanks,
> Vadim
> 

hi.

I don't quite understand what it is you are trying to do. if you want to
take a string and find out if it contains some other string, you probably
don't need to worry about regular expressions at all.

what kind of problem are you trying to solve? do you 'need' to use a
specific solution for some reason?

I tried this:

[~] [9:02pm] [jam at toast-pts/6] % python
Python 1.5.2 (#1, Apr 18 1999, 16:03:16)  [GCC pgcc-2.91.60 19981201 (egcs-1.1.1  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import string
>>> buf = "The Carrot"
>>> validveggies = ("carrot", "corn", "spam")
>>> buf = string.lower(buf)
>>> for w in string.split(buf):
...     if w in validveggies:
...             print "valid veggie '%s' found." % (w)
... 
valid veggie 'carrot' found.

does that help?

regards,
Jeff
-- 
|| visit gfd <http://quark.newimage.com/> 
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From tavares at connix.com  Sun Jul 11 10:36:40 1999
From: tavares at connix.com (Chris Tavares)
Date: Sun, 11 Jul 1999 10:36:40 -0400
Subject: Dreams.
References: <7m9to0$6qg$74@rks1.urz.tu-dresden.de>
Message-ID: <3788ABF8.5EB01BD3@connix.com>


Scott Dossey wrote:

> Okay, I am new to the newsgroup, but I thought I'd post something
> interesting that I've just started working on, and get some feedback.  I've
> been toying around with the idea of implementing "dreams" into the Python
> source code.
>
> What are dreams, you say?  Look at the following python code fragment for an
> idea, first of all, of how they are put into code:
>
> #START PSEUDOCODE FRAGMENT
> import sys
> import dreams
>
> aDream=dreams.dream()
> try:
>         #Insert code here.
> except:
>         aDream.wakeup()
>         sys.exit(1)
>
> aDream.realize()
> #END PSEUDOCODE FRAGMENT
>
> The code, at the location denoted by "#Insert code here." would run, but if
> an exception was thrown, it will have been as if the code had "happened in a
> dream".  No variables will have been changed/deleted.  Nothing will have
> changed from before the "dream" in terms of Python's variable state.
> Ideally, (but this is much harder), all file changes, all things that are
> "possible" to return to their original state, would be returned as well.
> (You can't take back network traffic).
>
> Dream.wakeup() would restore to the state before the dream, and continue
> execution from that line forward.  Dream.realize() would commit to the
> dream, stop saving change-state, and delete the change state it had already
> recorded.
>
>

[... SNIP ...]

>
> I'll take feedback in terms of ideas, problems, feasibility statements,
> flames,  whatever.  Most of all, I want help!  I think this would add a very
> interesting and useful feature to Python (as perhaps, a compile time option).
> Guido, I know you're out there, your feedback would be most appreciated,
> along with the feedback of any other Python maintainers.
>
> -Scott Dossey
> sdossey at armored.net, sdossey at openprojects.net

Scott,

What you're describing is transactional programing, which is used very heavily in
big database work. Basically, you "start a transaction", and do a bunch of stuff.
If everything works, you "commit" the transaction and all the changes you did are
made permanent. If anything goes wrong, you "roll back" the transaction and any
state is restored to before the transaction started.

I've never heard of the idea of integrating this directly into a programming
language. You might want to check out some database books for ideas.

-Chris





From bwinton at tor.dhs.org  Fri Jul 30 21:41:02 1999
From: bwinton at tor.dhs.org (Blake Winton)
Date: Sat, 31 Jul 1999 01:41:02 GMT
Subject: IDLE mini-hack & ? (adding menu items, understanding *args, self)
References: <199907272152.QAA01078@cro.ots.ac.cr> <379F8524.40B89D81@compaq.com>
Message-ID: <slrn7q4lou.n8t.bwinton@tor.dhs.org>

On Thu, 29 Jul 1999 10:33:08 +1200, Greg Ewing wrote:
>Isidor wrote:
>> My guess is that insert is a method of text, and "insert" (first
>> argument) is the location of the insertion point
>That's right. To add text to the end of a text widget,
>you can use "end" instead of "insert", e.g.

Is there a list of these anywhere?

>> Can I attach zip to "self" somehow in funcx, e.g., line 3:
>> 'self.zip = "googoo"'?
>Yes, you can certainly do that. It's the usual way
>of assigning to an instance variable in Python.
>
>> Will it only work if self.zip is defined in the
>> __init__ func?
>No, there's no need to do that. You can add a new
>attribute to any instance of any class at any time
>just by assigning to it. There is no notion of
>"declaring" instance variables.

And you don't even need to be in the class definition!

>>> class test:
...     def spam( self ):
...             print self.eggs
...
>>> thistest = test()
>>> thistest.spam()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 3, in spam
AttributeError: eggs
>>> thistest.eggs = "I like eggs!"
>>> thistest.spam()
I like eggs!
>>>

So, if you added the requisite "self."s to your example, it would give
you similar behaviour.  (Especially if you called testy before testx.)

Later,
Blake.

-- 
One Will.  One Dream.  One Truth.  One Destiny.  One Love.



From kantel at mpiwg-berlin.mpg.de  Tue Jul 27 04:52:47 1999
From: kantel at mpiwg-berlin.mpg.de (Jörg Kantel)
Date: Tue, 27 Jul 1999 10:52:47 +0200
Subject: Editors for Macintosh
References: <379CB0C8.FA84F91C@callware.com>
Message-ID: <kantel-2707991052470001@demokritos.mpiwg-berlin.mpg.de>

In article <379CB0C8.FA84F91C at callware.com>, Ivan Van Laningham
<ivanlan at callware.com> wrote:

> Hi All--
> Are there any other good editors (i.e., Python-aware) besides Alpha out
> there for the Macintosh?  I don't want to _use_ one, I just want to know
> if there are any that people recommend.

I use BBEdit on the Mac and do to the fact that there is no syntax
colouring mode for python I've mapped the ending "py" to Tcl syntax
colouring.

BBEdit does it's job very well because I can save the Python-Files as
"Python Files" so they run by double click ;-) (Non Mac Users won't
understand that I suppose :o)

I don't use the IDE very often because I mostly deal with Tkinter on the
Mac too, but if you don't need that the IDE will be a very good choice
also.

J"org

-- 
J"org Kantel             Max-Planck-Institute for the History of Science
Computer-Department                           kantel at mpiwg-berlin.mpg.de
Wilhelmstr. 44   http://www.mpiwg-berlin.mpg.de/staff/kantel/kantel.html
D-10117 Berlin   fon: +4930-22667-220               fax: +4930-22667-299



From javanet at sinbad.net  Fri Jul  2 20:20:31 1999
From: javanet at sinbad.net (Arnaldo)
Date: Fri, 2 Jul 1999 16:20:31 -0800
Subject: Getting to know Python
Message-ID: <377d57ae@pavlof.gci.net>

I just downloaded and compiled python in a sunos4.1 OS. I'm starting to have
fun with it but I'm really new
to the language.
I'm planning to buy a book this weekend so I can learn more about python's
syntax and posibly implement some of my c-shell/perl scripts in python.
In the mean-time I'm eager to have with the interactive interpreter.
How do I access the shell from the interpreter to enter commands like
,clear, ls, cp, etc.
Is it posible, how do I do that?

thanks guys
ajr






From abhanja at cal.vsnl.net.in  Sun Jul 25 05:18:31 1999
From: abhanja at cal.vsnl.net.in (Arup Bhanja)
Date: Sun, 25 Jul 1999 09:18:31 GMT
Subject: Python info please
Message-ID: <379AD667.434E9A47@cal.vsnl.net.in>

Hi Pythoners,

Goede Morgon,

is there a list for python programming ? Can you give me the
subscription details?

Regards,

Arup Bhanja
Calcutta,INDIA.

Currently executing...
>>>myquotes=[]
>>>myquotes.append('I love monty python')
>>>myquotes
..done enjoy Python





From htrd90 at zepler.org  Mon Jul 12 17:49:44 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Mon, 12 Jul 1999 21:49:44 GMT
Subject: Python and COM?
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F0F@RED-MSG-50>
Message-ID: <378b5e41.10968341@news.freeserve.net>

Bill Tutt <billtut at microsoft.com> wrote:

(posted an cc'ed to bill)

>(As an added bonus, this should actually work)
>The only GUID from interp.py is the coclass one. Every other GUID was just
>generated.

I _think_ the python code would also need to be changed to allow
QueryInterface for the IID of that dispinterface.

>Hope the below helps.... The MIDL docs in MSDN are certainly less than
>useful. :(

Hmmm, yes. Ive checked on MSDN but can't find anything authoritative
on this, either way.

>// System idl include nonsense
>import "oaidl.idl";
>import "ocidl.idl";
>
>[
>	uuid(EB048AA4-C2D1-11d2-855D-00C04F797DBA),
>	helpstring("Python.Interepreter dispinterface")
>]
>dispinterface IPythonInterpreter
>{ properties: 
>methods:
>	[id(1)] HRESULT Eval([in] BSTR bstrExpression, [out, retval] VARIANT
>pvarReturnValue);
>	[id(2)] HRESULT Exec([in] BSTR bstrExpression);
>};
>
>[
>	uuid(60240FE4-C2D2-11d2-855D-00C04F797DBA),
>	version(1.0),
>	helpstring("Python.Interpreter Type Library")
>]
>library PythonInterpreterLib
>{
>	importlib("stdole32.tlb");
>	importlib("stdole2.tlb");
>
>	[
>		uuid(30BD3490-2632-11cf-AD5B-524153480001),
>		helpstring("Python.Interpreter COClass")
>	]
>	coclass PythonInterpreter
>	{
>		[default] dispinterface IPythonInterpreter;
>	};
>
>}


Toby Dickenson



From jbauer at rubic.com  Thu Jul  8 11:53:01 1999
From: jbauer at rubic.com (Jeff Bauer)
Date: Thu, 8 Jul 1999 15:53:01 GMT
Subject: os.kill on NT
References: <3784B57E.C0A57A@cern.ch>
Message-ID: <3784C95D.45E5E617@rubic.com>

Haimo,

I'm not sure if you were referring to me, but here's a 
reference to a post I made to the Python TODO list:

http://www.python.org/cgi-bin/todo.py?req=show&file=faq13.005.htp

Best regards,

Jeff Bauer
Rubicon, Inc.

---
"Haimo G. Zobernig" wrote:
> Some time ago (maybe not all that long) someone posted
> a small function implementing os.kill on Windows NT, using 
> the PythonWin extensions.



From tismer at appliedbiometrics.com  Sat Jul 17 07:24:55 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Sat, 17 Jul 1999 11:24:55 GMT
Subject: Snippet: bitcount()
References: <000601bed02a$2235af20$06a22299@tim> <slrn7p0hsv.e38.kas@titan.magnetic-ink.dk>
Message-ID: <37906807.79AE1344@appliedbiometrics.com>


Klaus Alexander Seistrup wrote:
> 
> Tim Peters <tim_one at email.msn.com> wrote:
> 
> > > -- unless I decide to count several bits at a time, but
> > > I wouldn't know how to construct such a thing.
> >
> > Oh, sure you do <wink>:
> >
> > count[0] = 0
> > count[1] = 1
> > count[2] = 1
> > count[3] = 2
> > ...
> > count[255] = 8
> >
> > Now you can get the popcount of any one-byte int with one table-lookup.
> > And the popcount of any N-byte int or long with N table lookups and N-1
> > adds.  Put 2**16 entries in the table to cut the work in half.
> 
> I get your point, thanks, but I will spare this group for a posting of such
> a 2**16 table, leave alone the full 32-bit version. ;-)

FYI, here is where Tim and I stopped a "race".
We both claimed to have the fastest version of bit counting.
Apparently, results were dependant from your CPU's cache size.
Mine had 512KB, Tims only 256KB, so on his machine he was faster,
and on mine, I was. We decided to stop there.

Here is Tim's most elegant last version which I like the best:

______________________________________________________________

_run2mask = {1: 0x5555555555555555L,
             2: 0x3333333333333333L,
             4: 0x0F0F0F0F0F0F0F0FL,
             8: 0x00FF00FF00FF00FFL}

def buildmask2(run, n):
    run2 = run + run
    k = (n + run2 - 1) / run2
    n = k * run2
    try:
        answer = _run2mask[run]
        k2 = 64 / run2
    except KeyError:
        answer = (1L << run) - 1
        k2 = 1
    while k > k2:
        k2 = k2 + k2
        if k >= k2:
            answer = answer | (answer << (run * k2))
        else:
            answer = answer | (answer << (run2 * (k - k2/2)))
            k2 = k
    if k2 > k:
        answer = answer >> (run2 * (k2 - k))
    return answer, n

def popcount4(x):
    lomask, n = buildmask2(1, nbits(x))
    x = x - ((x >> 1) & lomask)
    target = 2
    while n > target:
        lomask, n = buildmask2(target, n)
        x = (x & lomask) + ((x >> target) & lomask)
        target = target + target
        for i in range(1, target/2):
            if n <= target:
                break
            n = n >> 1
            n = (n + target - 1) / target * target
            x = (x & ((1L << n) - 1)) + (x >> n)
    return int(x)
______________________________________________________________

And here is my dirty, brutal hack which is at the "same" speed
for sufficiently inexact values of "same":
______________________________________________________________

import marshal, operator, array, string

def makeweighttrans() :
    weights = map(popcount, range(256))
    return string.join(map(chr, weights), '')

def popcount6(x, dumps = marshal.dumps,
                 add = operator.add,
                 trans = makeweighttrans(),
                 reduce = reduce,
                 translate = string.translate,
                 mkarray = array.array):
    digitcounts = mkarray("B", translate(dumps(x), trans))
    # overwrite type code & digit count
    digitcounts[0] = digitcounts[1] = digitcounts[2] = \
                     digitcounts[3] = digitcounts[4] = 0
    return reduce(add, digitcounts)
    
______________________________________________________________


So that' where you can get when trying to count bits.
I can't encourage you to try to outperform one of the
above and don't want to start another race (at least
I will not try). But if somebody comes up with a solution
which is, say, 30 percent faster for any distribution of
bits, and it uses standard Python only, then I will
give you a free website and pay your fee for the next
Python conference.

and-maybe-employ-you-in-my-company-ly y'rs - chris

p.s.: You can shorten the above by Mike Hudson's bytecodehacks,
but it will not get faster. Actually a tick slower under Win9X,
due to the way how VC++6.0 optimizes LOAD_FAST and LOAD_CONST
slightly different (they *are* technically equivalent).

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From Markus.Stenberg at p98.f112.n480.z2.fidonet.org  Sun Jul  4 20:19:28 1999
From: Markus.Stenberg at p98.f112.n480.z2.fidonet.org (Markus Stenberg)
Date: Mon, 05 Jul 1999 01:19:28 +0100
Subject: Is it just me, or..
Message-ID: <000017ca@bossar.com.pl>

From: Markus Stenberg <mstenber at cc.Helsinki.FI>

Is there a bug in the handling of pipes with the Python? Esp. in
non-blocking way.

Example:
	(r,w)=os.pipe()
	r = os.fdopen(r, "r", 0)
	w = os.fdopen(w, "w", 16384)
	w.write('foobarbaz')
	select.select([r],[],[]) # returns r, as it should
	r.read(1) # returns 'f'
	select.select([r],[],[]) # blocks!

Now select _blocks_, despite there being obviously some data that at least
should not have been buffered elsewhere. Also, fdopened stuff's reading
features are bit annoying, as doing read(bigNumber) results in blocking
call, as opposed to recv(bigNumber)'s semantics. Thus instead of pipes
(=fast) I am using TCP/UNIX sockets (=slow) but they seem to work better
for what I need. Am I missing something really obvious?

This is on Linux, if it matters.

-Markus

-- 
	This message was written in, or processed by a Linux system. 
	Linux, as in Linus' *NIX, not Lignux, as in RHS' arrogance.



From sebastien.tanguy at wanadoo.fr  Thu Jul 15 17:01:57 1999
From: sebastien.tanguy at wanadoo.fr (Sebastien Tanguy)
Date: 15 Jul 1999 23:01:57 +0200
Subject: Terminating Python...
References: <BFC0587634C3D111AACB00805FA766FA02C1C652@exchny12.ny.smb.com> <378CFDDF.489CDCEF@compaq.com> <7mj8s9$op2@newsops.execpc.com> <004901bece61$41270b60$f29b12c2@secret.pythonware.com> <378DD196.BAB576C6@njeb.att.com>
Message-ID: <87u2r5lj96.fsf@meryllon.death-gates.fr>

>>>>> "AF" == Alex Flinsch <aflinsch at njeb.att.com> writes:

AF> Fredrik Lundh wrote:

>> in which way is typing "raise SystemExit<Return>" better
>> than typing "<Control-D>" (which is documented in the
>> Python Tutorial) ?
>> 

AF> "raise SystemExit" could be embedded in some python code, while a
AF> <Control-D> could not be.  "raise SystemExit" would work on any
AF> OS, no need to check if we are running windows or something else.

and what about :

import sys
sys.exit(0)

Doesn't it work on every platform ?

greetings,
seb.
-- 
<<"regression testing" ? What's that ? If it compiles, it is good, 
  if it boots up it is perfect.>>		(Linus T.)



From arcege at shore.net  Thu Jul  8 14:02:55 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 08 Jul 1999 18:02:55 GMT
Subject: PyImport_ImportModule, PyModule_GetDict
References: <3784C57B.59DF0D33@ncsa.uiuc.edu> <hl5h3.1285$6M6.394545@news.shore.net>
Message-ID: <jJ5h3.1291$6M6.394545@news.shore.net>

Michael P. Reilly <arcege at shore.net> wrote:
: Randy Heiland <heiland at ncsa.uiuc.edu> wrote:
: : If I understand this correctly, the equivalent of doing this:

: :>>> import mymodule
: :>>> dir(mymodule)

: : is sort of this, in the C API:

: : PyObject *mod = PyImport_ImportModule("mymodule");
: :  PyObject *myDict = PyModule_GetDict(mod);

: : Q:  what is the C API equivalent of this:

: :>>> from mymodule import *
: :>>> dir()

Oops... forgot to initialize something.  It compiled and worked, but
would only work without optimized compilation. :)

: You will have to do something a little more:

:     PyObject *fromlist, *main, *globals;
:     PyObject *sysmod, *sysdict, *key, *val;
:     int pos;  /* used internally by PyDict_Next() */
^^^^^^^^^^^^^^
      int pos = 0;

:     /* tell __import__ to return the top-level module */
:     fromlist = Py_BuildValue("[s]", "*");
:     main = PyImport_ImportModule("__main__");
:     globals = PyModule_GetDict(main);

:     /* ImportModuleEx handles "from package.module import *" as well */
:     sysmod = PyImport_ImportModuleEx("sys", globals, globals, fromlist);

:     sysdict = PyModule_GetDict(sysmod);
:     /* perform standard name bindings */
:     while (PyDict_Next(sysdict, &pos, &key, &val) > 0) {
:       if (!PyString_Check(key) ||
:           (PyString_AS_STRING(key)[0] != '_'))
:         PyDict_SetItem(globals, key, val);
:     }

: If you look at the documentation for __import__, it does not perform the
: name binding in the globals dictionary (for some reason ;), so that must
: be performed manually.

:   -Arcege




From rceci at my-deja.com  Thu Jul  1 11:04:17 1999
From: rceci at my-deja.com (rceci at my-deja.com)
Date: Thu, 01 Jul 1999 16:04:17 +0100
Subject: 1.5.1 on IRIX
Message-ID: <00001615@bossar.com.pl>

From: rceci at my-deja.com

Hello,

I built the 1.5.1 distro on one of our irix6.4 systems ( n64-mips4
arch), using the -32 (-o32)  compiler.  This was installed in the
"normal" place - /usr/local/....  and ran fine on that machine.

Now I have moved this installation to a NFS mount, and tried running the
interpreter on another irix6.4 system with similar characterstics.  I am
setting PYTHONPATH to point to the location of new python runtime
libraries:

The libs originally where built/installed  -> /usr/local/lib/python1.5
re-installed as ${somedir}/lib/irix32, and set PYTHONPATH to
${somedir}/lib/irix32.

the interpreter itself is re-located in ${somedir}/bin/irix32/

I can run python in its new location on the old system, just fine:

    setenv PYTHONPATH ${somedir}/lib/irix32
    ${somedir}/bin/irix32/python -v

and get the normal debugging info when the modules are loaded, which
confirms they are loading out of my specified PYTHONPATH.

However, when I try running this installation on another one of our irix
systems, I get:

Could not find platform independent libraries <prefix>
Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
Python 1.5.1 (#3, May 21 1999, 16:24:24) [C] on irix646-32-mips2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
Contacting Registry ...
environment exported ...
-- Using registry version :  10.0

>>>

The messages about Registry, are coming from SWIGged module linked
against the interpreter.  I looks like that is working fine in this
installation on both systems.  It would appear that the warnings on the
second system are non-fatal.

Any ideas how to "fix" or suppress them?

Rob


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Charles.G.Waldman at p98.f112.n480.z2.fidonet.org  Fri Jul  2 14:08:28 1999
From: Charles.G.Waldman at p98.f112.n480.z2.fidonet.org (Charles G Waldman)
Date: Fri, 02 Jul 1999 19:08:28 +0100
Subject: C shared object files
Message-ID: <00001688@bossar.com.pl>

From: Charles G Waldman <cgw at fnal.gov>


I'm sure there are plenty of people on this list, including myself,
who'd be glad to help you.  As far as building an extension module
from your .c file, it would help to know what kind of platform you're
on.  If you're on a Unix platform with gcc, the command would look
something like:

gcc -I/path/to/python/includes -c myfile.c
gcc -shared -o mymodule.so myfile.o

As far as checking for presence of a module at runtime, the simplest
thing would be:

try:
	import module
except ImportError:
       #it doesn't exist...

Aren't there any local Python gurus at llnl.gov?





From MHammond at skippinet.com.au  Sat Jul 31 19:46:13 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Sun, 1 Aug 1999 09:46:13 +1000
Subject: Need help with TCP/IP client access from Windows
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90FF7@RED-MSG-50>
Message-ID: <7o01ss$cjh$1@m2.c2.telstra-mm.net.au>

Bill Tutt wrote in message
<4D0A23B3F74DD111ACCD00805F31D8100DB90FF7 at RED-MSG-50>...

>The correct way to go about doing these things is via the SHGetFolderPath
>API.
>(Calling SHGetSpecialFolderPath is also ok, but SHGetFolderPath is the
>preferred API entry point.)
>Docs at:
>http://msdn.microsoft.com/library/sdkdoc/shellcc/Shell/Functions/ShGetFolde
r
>Path.htm
>
>A description of all of the special folders is at:
>http://msdn.microsoft.com/library/sdkdoc/shellcc/shell/Functions/CSIDL.htm
>
>
>Ok so can I use these in Python yet?
>Unfortunately, not at the moment. If someone screams and yells I'm sure
Mark
>or I will write the necessary wrapping code.

Actually Bill, you can - althought it is well hidden.

There is a win32com.shell.shell extension that covers this.  It is part of
the COM extensions as all this cruft is heavily tied to COM - eg, creating
shortcuts, etc etc etc.

>>> from win32com.shell import shell, shellcon
>>> shell.SHGetSpecialFolderPath(0, shellcon.CSIDL_PERSONAL)
L'C:\\WINNT\\Profiles\\skip\\Personal'
>>>

SHGetFolderPath/Location arent exposed - I will add them - but as you
mention, the SHGetSpecial versions do work fine.

Mark.






From gmcm at hypernet.com  Thu Jul  1 14:01:02 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 1 Jul 1999 13:01:02 -0500
Subject: Mailbox, tell and read
In-Reply-To: <377B7742.F39CB0A4@prescod.net>
Message-ID: <1281273761-39459133@hypernet.com>

Paul Prescod wrote:

> Gordon McMillan wrote:
> > 
> > How 'bout opening in binary, and munging line endings yourself just
> > before returning the message? 
> 
> First, are there platforms other than DOS/Windows/Win32 where there
> is a difference between text and binary mode?

There's a theoretical difference on *nix - otherwise, they wouldn't 
have made up the distinction!

> I didn't want to munge
> things myself because I'm not sure if what is the "right" munge on
> the Mac, for instance. Can I just do a search and replace for CR/LF
> -> CR on all platforms, all of the time?

On Windows, Notepad (and the underlying default text widget) require
\r\n, but newer widgets / editors will be happy with \n. On *nix,
most tools will show a \r as a noise character. No idea how
forgiving Mac tools are.

AFAIK, we have
 *nix -> \n
 Windows -> \r\n
 Mac -> \r
 
However, since you're dealing with messages created who-knows-where 
and run through who-knows-what transformations, you may well have to 
munge no matter what.


- Gordon



From A.M..Kuchling at p98.f112.n480.z2.fidonet.org  Sun Jul  4 14:19:20 1999
From: A.M..Kuchling at p98.f112.n480.z2.fidonet.org (A.M. Kuchling)
Date: Sun, 04 Jul 1999 19:19:20 +0100
Subject: wddx wrapper?
Message-ID: <000017c9@bossar.com.pl>

From: amk at mira.erols.com (A.M. Kuchling)

On Thu, 1 Jul 1999 16:19:25 -0400, Hank Fay <hank at prosysplus.com> wrote:
>Does anyone know a reference to a Python wrapper for wddx (www.wddx.org)?

There's a WDDX module in the current XML-SIG's CVS tree (not in the
released code).  It's in xml.marshal.wddx, and the interface is much
like the marshal module:

>>> from xml.marshal import wddx
>>> print wddx.dumps( ('tuple', 1, 3.5, wddx.TRUE) )
<?xml version="1.0"?><!DOCTYPE wddxPacket SYSTEM "wddx_0090.dtd">
<wddxPacket version="0.9">
  <array length="4">
    <string>tuple</string>
    <number>1</number>
    <number>3.5</number>
    <boolean value="true"/>
  </array>
</wddxPacket>

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
If there's another world, he lives in bliss; / If there is none, he made the
best of this.
    -- Robert Burns, "Epitaph on William Muir"



From donn at u.washington.edu  Thu Jul  8 13:05:25 1999
From: donn at u.washington.edu (Donn Cave)
Date: 8 Jul 1999 17:05:25 GMT
Subject: Two Pythons talking to each other?
References: <199907080019.TAA22244@buffalo.fnal.gov> <199907080529.HAA19502@axil.hvision.nl> <3784486a.7993113@news.ricochet.net>
Message-ID: <7m2lol$bnm$1@nntp6.u.washington.edu>

phil at ricochet.net (phil) writes:
| On Thu, 8 Jul 1999 05:30:40 GMT, "Hans Nowak" <ivnowa at hvision.nl>
| wrote:
...
|> So my question is, how does one generally deal with this? Should I 
|> attach a newline (\n) (or maybe another separator character) after 
|> every string? Or are there other ways to guarantee that two commands 
|> will arrive at the other side as two strings?

| The primary protocols are TCP and UDP. UDP is non-connection oriented
| and will deliver a datagram. Sending a packet from one side to the
| other followed by another packet will not coalesce the packets into
| one. However, seems to me that you really want a session oriented
| solution. I suspect that the guys in the sockets group will tell you
| that a TCP connection is a stream, so all bets are off. Sending "hello
| wordl. " followed by sending "hello snailor" might result in receiving
| "hello wordl.hel" and "lo snailor".  The solution is to impose your
| own protocol on top of this using TCP as a delivery mechanism.
| Something as simple as <NL> delimiting  might do it for you. 

May be worth noting that application protocols can be and are also
designed on top of UDP, where among other things they may add some
kind of sequence monitor - datagram delivery isn't guaranteed, a
datagram may be discarded if its system buffer has filled up.

An alternative to separator data is a byte count prefix.  E.g.,

    def wpacket(self, data):
        count = struct.pack('!l', len(data))
        data = count + data
        while data:
            # write/send isn't guaranteed to send all.
            n = self.write(data)
            data = data[n:]

    def rpacket(self):
        count = self.read(4)
	data = ''
	while len(data) < count:
        	data = data + self.read(struct.unpack('!l', count)[0])
	return data

For later optimization, it might be nice to buffer the data instead of
reading it in little bits, or maybe define a minimum size packet that's
big enough most of the time, which could cut the reads about in half.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From htrd90 at zepler.org  Thu Jul  8 16:24:44 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Thu, 08 Jul 1999 20:24:44 GMT
Subject: XML-RPC and firewall [solution]
References: <s781d84b.054@mail.conservation.state.mo.us>
Message-ID: <378659ee.4241259@news.freeserve.net>

"Jeffrey Kunce" <kuncej at mail.conservation.state.mo.us> wrote:

>It bad content-type han't caused me any problems yet, though.

That would definitely cause problems for Zope's xml-rpc support.


Toby Dickenson



From wyc at mothra.lbl.gov  Fri Jul 16 21:21:51 1999
From: wyc at mothra.lbl.gov (William Y. Crutchfield)
Date: Sat, 17 Jul 1999 01:21:51 GMT
Subject: canvas, postscript, and fontmap
Message-ID: <378FD66B.726A9AE3@mothra.lbl.gov>

would some kind soul provide me an example of setting the fontmap in the
postscript method of canvas inTkinter?  I know what I want to do, I'm
just not sure what Tkinter is expecting fontmap to be, dictionary, list,
tuple?




From viofis at my-deja.com  Thu Jul  1 20:35:51 1999
From: viofis at my-deja.com (viofis at my-deja.com)
Date: Fri, 02 Jul 1999 01:35:51 +0100
Subject: please help with simple regex
Message-ID: <00001647@bossar.com.pl>

From: viofis at my-deja.com

I'm a beginner at Python, so this may be a stupid question:

I have a string, for example, myString="the Potato".
1. Convert it to all lowercase ("the potato")
2. Match it against a list of strings - "potato,tomato,corn,etc...",
extract "potato" (if it matches with something on the list), then save
into myNewString.

Any help is greatly appreciated.

thanks,
Vadim



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From joeld at anisci.com  Mon Jul 26 18:38:00 1999
From: joeld at anisci.com (Joel Davis)
Date: 26 Jul 1999 15:38:00 PDT
Subject: Python and Red Hat Linux 6
References: <379CD547.22A10A50@postoffice.pacbell.net>
Message-ID: <379CE4C8.4F7BAEF1@anisci.com>

that sound right.. did you set the file to be executable (chmod?)


Rob wrote:

> When programming in Perl or Python, I am using a header like the
> following on an executable file to cause the shell (bash or csh) to run
> the scripts as Perl/Python files:
> #!/usr/bin/perl   for Perl   and #!/usr/bin/python   for Python. I've
> checked and the compilers are located in the appropriate directories. I
> keep getting this when I run scripts:
> bash: <script name>: command not found. Does Linux support the '#!'
> notation in the file header, or do I always have to run the scripts by
> explicitly typing in 'perl' or 'python' before the script name?
> ~Rob~

--
----------------------------------------------------------------
                                                joeld at anisci.com





From jcw at equi4.com  Mon Jul  5 18:09:35 1999
From: jcw at equi4.com (Jean-Claude Wippler)
Date: Tue, 06 Jul 1999 00:09:35 +0200
Subject: ANN: Minotaur 0.1 - bridging Tcl, Python, and Perl
Message-ID: <37812D1F.F18721A4@equi4.com>

This is the first public announcement of a generalized extension which
lets you run Python scripts from Tcl, or Tcl from Perl, or any of the
other combinations.

Minotaur works by loading the appropriate shared library into another
script language's context.  So, for example, running Tcl from Python
means that the Python main program imports the Minotaur extension, which
in turn loads the Tcl libraries - thus Tcl becomes "embedded" into the
Python runtime environment (which is similar to embedding Tcl in a
normal C program).  As involved as that sounds, performance is already
surprisingly good, and the capability is turning out to be quite useful.

Minotaur 0.1 is an alpha level release, meaning: it works, aside from a
few quirks (see below), and bugs, but its API and implementation might
still change in incompatible ways in future releases.

An example, embedding a Tcl script in Python:
        import Minotaur
        Minotaur.initTcl()
        print Minotaur.Tcl("expr {1111 + 2222}")
        Minotaur.Tcl("""
            set result "Tcl version [info patchlevel]"
            Python "result = '$result'"
        """)
        print result

As you can see, a value is returned either directly (only if it's an
integer), or by using a reverse embedding, i.e. executing a Python
statement from the embedded Tcl system.  Also, this shows how tricky
quoting could get once you start mixing languages in this manner.  And
last but not least, note that the different language contexts are
separate: the two "result" variables used above are not related in any
way (one for Tcl, one for Python).  Also, Minotaur does not yet let you
share data, it merely lets you evaluate scripts (and copy things).

Now the gotcha's, quite a list of them right now:

- You need shared libraries to use Minotaur. Standard installations,
  especially of Perl and Python, do not always include these.  The
  Minotaur homepage has some info on rebuilds you might need to do.
- Tcl/Python can call Perl but to get results back, I had to introduce
  a secondary shared library, dynamically bound to the Perl runtime.
- There is currently no other interface than evaluating a single string.
  Passing int/string/whatever args efficiently is planned for later.
- The Minotaur extension has only been tested for Linux and Windows, but
  is built in a way which should also work on the Mac, and on all Unix
  systems, even those which do not support back-linking in shared libs. 
- Minotaur uses some global data; it's not yet re-entrant / thread-safe.
  As a matter of fact, Minotaur is still totally ignorant of threads.
- I have not yet succeeded in getting Minotaur working on the Mac, and
  ran into problems when trying to launch Tk from Python (Win/Linux).
- The current release does not include any Minotaur binaries, you may
  have to do quite a bit of tweaking to get this stuff to work for you.
  One of my plans is to substantially simplify deployment of all this.

And here's the fun part of things:

 - Minotaur uses Forth as low-level "super-glue" language.  This offers
   the ability to deal with all the inevitable "impedance mismatches"
   which will occur when crossing widely different software designs.
   Also, this can be done without compilation (and often portably).
 - Timing tests indicate that the overhead of calling through Minotaur
   is very low, when the necessary symbol lookups are done in advance.
   The C-based Forth in here uses pForth, and includes quite a bit of
   standard ("ANS") Forth for now, though this may change in the future.
 - The Minotaur extension is loadable from Perl, Python, and Tcl, 'cause
   it exports initialization calls for each of them.  This is essential,
   since the embedded language needs to reopen the extension to gain
   access to the same Forth context as the one which loaded it (this is
   how each embedded language can call back into its parent context).
 - Forth primitives have been defined which can open, lookup, and locate
   routines exported from shared libraries.  Due to the flexibility and
   performance of Forth, everything else needed to embed one language in
   the other is done in Forth, i.e. at run time.  Forth has no idea that
   it is being used to tie scripting languages together, for example.
 - Minotaur, and these shared library access functions have been tested
   on Linux, Windows, and Mac (partially).  The Forth code needed to
   implement such language bindings differs only in pathnames, so far.
 - But this is only the tip of the iceberg, really.  Proof-of-concept
   bindings have also been created for Java, MS COM, Lua, ICI, and PHP.

Minotaur is a spinoff of my long-term "TinyScript" project and is open
source, with a BSD-style license (do whatever you like with it, just
leave the copyright attributions intact, and don't sue me).  Minotaur
was written to see how far one could go in bringing scripting worlds
together, and to help me tie my own extensions into a range of languages
without having to write a whole range of interfaces.  It hasn't met this
goal yet, but my hope is that it will be of use to others and that a
number of language experts will extend and improve the mechanism so it
works well with all of them.

Minotaur has sort of a homepage here:
        http://mini.net/pub/ts2/minotaur.html
A mirror copy of my working area is being tracked here:
        http://mini.net/pub/ts2/  (a.k.a. ftp://mini.net/pub/ts2/)
And a snapshots of the whole area was recently created as:
        http://mini.net/pub/ts2-19990704.tar.gz

There is no documentation, release 0.1 is a "use the source Luke" kind
of project, but if you would like to try this out - or better still:
participate in whatever way you like - feel free to subscribe to the
Minotaur mailing list, by sending an email to minotaur-add at mini.net. 
There are a handful of subscribers now, with just a few outbursts of
discussions of where to take this, so far.  Which, at this stage, is
still a very open-ended question, IMO.

Comments, suggestions, praise, but especially offering to help with
whetever aspect of Minotaur you'd like to improve are most welcome. 
Language and platform bashing - yawn - will be redirected to /dev/null. 
If you wish to help lower language barriers, so that more code can be
re-used, more interfacing problems can be solved, and perhaps one day
even generic extensions can be written which serve multiple contexts,
please have a look at Minotaur and consider helping to make it happen.

One interesting option, would be to create a binding in SWIG which
generates the low-level glue to Forth, as well as language-specific
wrappers.  The key advantage is that this could let SWIG be used
entirely as a run-time tool, with no compiler/compilation in sight
(assuming the target library exists in shared library form, that is).

-- Jean-Claude Wippler <jcw at equi4.com>

P.S.  This announcement was posted to a range of newsgroups, but was not
      crossposted to avoid serving as flame bait for YAFLC (Yet Another
      Fruitless Language Comparison).  The list of newsgroups is:
		comp.lang.forth
		comp.lang.misc
		comp.lang.perl.moderated
		comp.lang.python
		comp.lang.tcl
      It's a lotta posting for such an early project, but that's because
      there seems to be no established "intra-lingual" newsgroup for
      this stuff.





From holger at phoenix-edv.netzservice.de  Wed Jul  7 09:50:40 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Wed, 7 Jul 1999 13:50:40 GMT
Subject: Win32: python-network-installation?
References: <378349DE.7ED68819@phoenix-edv.netzservice.de>
Message-ID: <37835B30.BFCC4A05@phoenix-edv.netzservice.de>

To stop some people burning their brains I've made some
of the job already:

I summarized all registry-entries from my installation
and changed paths inside to server-UNC-Paths (file appended!). 
I've made a test-server installation on my server 'as1' 
where the directory 'python' is opened for everybody inside
intranet.
To reach Python-dlls I copied the dlls '\WinNT\System32\PythonCOM15.dll'
and '\WINNT\System32\Python15.dll' to a directory '\\as1\python\dll'
onto the server.

Just tested that construction:
1)starting python-scripts via double-click works fine
2)starting python.exe or pythonw.exe with 
 '\\as1\python\pythonw.exe' works fine ,too 
3)only '\\as1\python\pythonwin\pythonwin.exe' occurs a
 'Microsoft Visual C++ Runtime Library'-RuntimeError:
 Program: \\as1\python\pythonwin\pythonwin.exe
 abnormal program termination

What do you think, does I forget any dll? Why does that error (3)
occur? (Perhaps I need that pythonwin.exe to check some
scripts when sitting at a client...?!)

What's about that registry-file? Do I really need all that entries?
What entries could I dismiss? 

Are there other (better) ways to get a network installation for Win32-Clients?

by,
Holger



--------------------ClientReg.reg------------
REGEDIT4

[HKEY_CLASSES_ROOT\.py]
@="Python.File"
"Content Type"="text/plain"

[HKEY_CLASSES_ROOT\.pyc]
@="Python.CompiledFile"

[HKEY_CLASSES_ROOT\.pyc]
@="Python.CompiledFile"

[HKEY_CLASSES_ROOT\.pyo]
@="Python.CompiledFile"

[HKEY_CLASSES_ROOT\.pyw]
@="Python.NoConFile"
"Content Type"="text/plain"

[HKEY_CLASSES_ROOT\Python.CompiledFile]
@="Compiled Python File"

[HKEY_CLASSES_ROOT\Python.CompiledFile\DefaultIcon]
@="\\\\as1\\Python\\pyc.ico"

[HKEY_CLASSES_ROOT\Python.CompiledFile\shell]

[HKEY_CLASSES_ROOT\Python.CompiledFile\shell\open]

[HKEY_CLASSES_ROOT\Python.CompiledFile\shell\open\command]
@="\\\\as1\\Python\\python.exe \"%1\" %*"

[HKEY_CLASSES_ROOT\Python.File]
@="Python File"

[HKEY_CLASSES_ROOT\Python.File\AutoRegister]
@="\\\\as1\\Python\\DLL\\PythonCOM15.dll"

[HKEY_CLASSES_ROOT\Python.File\DefaultIcon]
@="\\\\as1\\Python\\Py.ico"

[HKEY_CLASSES_ROOT\Python.File\shell]

[HKEY_CLASSES_ROOT\Python.File\shell\Edit]

[HKEY_CLASSES_ROOT\Python.File\shell\Edit\command]
@="\\\\as1\\Python\\Pythonwin\\pythonwin.exe /edit \"%1\""

[HKEY_CLASSES_ROOT\Python.File\shell\open]

[HKEY_CLASSES_ROOT\Python.File\shell\open\command]
@="\\\\as1\\Python\\python.exe \"%1\" %*"

[HKEY_CLASSES_ROOT\Python.NoConFile]
@="Python File (no console)"

[HKEY_CLASSES_ROOT\Python.NoConFile\DefaultIcon]
@="\\\\as1\\Python\\Py.ico"

[HKEY_CLASSES_ROOT\Python.NoConFile\shell]

[HKEY_CLASSES_ROOT\Python.NoConFile\shell\Edit]

[HKEY_CLASSES_ROOT\Python.NoConFile\shell\Edit\command]
@="\\\\as1\\Python\\Pythonwin\\pythonwin.exe /edit \"%1\""

[HKEY_CLASSES_ROOT\Python.NoConFile\shell\open]

[HKEY_CLASSES_ROOT\Python.NoConFile\shell\open\command]
@="\\\\as1\\Python\\pythonw.exe \"%1\" %*"


[HKEY_CURRENT_USER\Software\Python]

[HKEY_CURRENT_USER\Software\Python\Python for Win32]

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Editor]
"Tab Size"=dword:00000008
"Indent Size"=dword:00000008
"Use Tabs"=dword:00000001
"Auto Reload"=dword:00000001
"Backup Type"=dword:00000003
"Smart Tabs"=dword:00000000
"Default Font"="(-134217665, 0, 200, 0, 0, 0, 49, 'Courier')"

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Format]
"Syntax"="{7: (-1073741821, 0, 200, 0, 0), 6: (-1073741821, 0, 200, 0, 0), 5:
(-1073741821, 1, 200, 0, 8388608), 4: (-1073741821, 0, 200, 0, 8421376), 3: (-1073741821,
0, 200, 0, 128), 2: (-1073741821, 2, 200, 0, 32768), 1: (-1073741821, 1, 200, 0, 255)}"

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Interactive Window]
"left"=dword:0000001a
"top"=dword:00000043
"right"=dword:00000376
"bottom"=dword:00000176
"FormatTitle"="(-536870897, 0, 220, 0, 16711680, 184, 34, 'Arial')"
"FormatInput"="(-134217665, 0, 200, 0, 0, 0, 49, 'Courier')"
"FormatOutput"="(-134217665, 0, 200, 0, 8421376, 0, 49, 'Courier')"
"FormatOutputError"="(-134217665, 0, 200, 0, 255, 0, 49, 'Courier')"
"Show at startup"=dword:00000001

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Main Window]
"left"=dword:0000007b
"top"=dword:000000bf
"right"=dword:0000037b
"bottom"=dword:000002d8

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Python]

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Python Trace Collector]
"left"=dword:00000016
"top"=dword:00000016
"right"=dword:0000034c
"bottom"=dword:000001da

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Recent File List]

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Settings]
"Recent File List Size"=dword:0000000a

[HKEY_CURRENT_USER\Software\Python\Python for Win32\ToolbarDefault-Bar0]
"BarID"=dword:0000e801

[HKEY_CURRENT_USER\Software\Python\Python for Win32\ToolbarDefault-Bar1]
"BarID"=dword:0000e800
"XPos"=dword:00000071
"YPos"=dword:fffffffe
"MRUWidth"=dword:00000153
"Docking"=dword:00000001
"MRUDockID"=dword:0000e81b
"MRUDockLeftPos"=dword:00000071
"MRUDockTopPos"=dword:fffffffe
"MRUDockRightPos"=dword:000001d4
"MRUDockBottomPos"=dword:0000001e
"MRUFloatStyle"=dword:00002004
"MRUFloatXPos"=dword:80000000
"MRUFloatYPos"=dword:00000000

[HKEY_CURRENT_USER\Software\Python\Python for Win32\ToolbarDefault-Bar2]
"BarID"=dword:0000e81b
"Bars"=dword:00000003
"Bar#0"=dword:00000000
"Bar#1"=dword:0000e800
"Bar#2"=dword:00000000

[HKEY_CURRENT_USER\Software\Python\Python for Win32\ToolbarDefault-Summary]
"Bars"=dword:00000003
"ScreenCX"=dword:00000400
"ScreenCY"=dword:00000300

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Tools Menu]

[HKEY_CURRENT_USER\Software\Python\Python for Win32\Tools Menu\1]

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\.py]
@="Python.File"
"Content Type"="text/plain"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\.pyc]
@="Python.CompiledFile"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\.pyo]
@="Python.CompiledFile"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\.pyw]
@="Python.NoConFile"
"Content Type"="text/plain"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.CompiledFile]
@="Compiled Python File"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.CompiledFile\DefaultIcon]
@="\\\\as1\\Python\\pyc.ico"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.CompiledFile\shell]

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.CompiledFile\shell\open]

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.CompiledFile\shell\open\command]
@="\\\\as1\\Python\\python.exe \"%1\" %*"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File]
@="Python File"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File\AutoRegister]
@="\\\\as1\\python\\DLL\\PythonCOM15.dll"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File\DefaultIcon]
@="\\\\as1\\Python\\Py.ico"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File\shell]

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File\shell\Edit]

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File\shell\Edit\command]
@="\\\\as1\\Python\\Pythonwin\\pythonwin.exe /edit \"%1\""

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File\shell\open]

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File\shell\open\command]
@="\\\\as1\\Python\\python.exe \"%1\" %*"


[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.NoConFile]
@="Python File (no console)"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.NoConFile\DefaultIcon]
@="\\\\as1\\Python\\Py.ico"

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.NoConFile\shell]

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.NoConFile\shell\Edit]

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.NoConFile\shell\Edit\command]
@="\\\\as1\\Python\\Pythonwin\\pythonwin.exe /edit \"%1\""

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.NoConFile\shell\open]

[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.NoConFile\shell\open\command]
@="\\\\as1\\Python\\pythonw.exe \"%1\" %*"

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\App Paths\Python.exe]
@="\\\\as1\\Python\\Python.exe"

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\App Paths\Pythonwin.exe]
@="\\\\as1\\Python\\Pythonwin\\Pythonwin.exe"

[HKEY_LOCAL_MACHINE\SOFTWARE\Python]

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore]

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5]

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5\Dll]
@="\\\\as1\\python\\DLL\\Python15.dll"

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5\Help]

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5\Help\Main Python Documentation]
@="\\\\as1\\Python\\Doc\\index.html"

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5\Help\Python COM Reference]
@="\\\\as1\\Python\\win32com\\win32com.hlp"

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5\Help\Pythonwin Reference]
@="\\\\as1\\Python\\Pythonwin\\Help\\Pythonwin.hlp"

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5\Help\Win32 Extensions]
@="\\\\as1\\Python\\win32\\Help\\Python Win32 Extensions.hlp"

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5\InstallPath]
@="\\\\as1\\Python"

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5\InstallPath\InstallGroup]
@="Python 1.5"

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\Pythonwin]

[HKEY_LOCAL_MACHINE\SOFTWARE\Python\Pythonwin\Build]
@="123"
---------------End of ClientReg.reg------------



From arcege at shore.net  Wed Jul 21 20:30:00 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 22 Jul 1999 00:30:00 GMT
Subject: cin
References: <3srl3.1302$Gj8.191985152@newsa.telia.net>
Message-ID: <cCtl3.818$ui4.256788@news.shore.net>

Joacim L*wgren <fire at netlimit.com> wrote:
: Hi,

: I'm new to Python and I'm wondering what the C++ "cin" (or C
: "scanf()")equivalent in Python is. Doesn't seem there's a quick function for
: this.

Actually it is not a function, it is an operation (%).

>>> name, age = 'Michael', 30
>>> s = 'My name is %s and I am %d years old.'
>>> print s % (name, age)
My name is Michael and I am 30 years old.

Take a look at the Library Reference Manual, Section 2.1.5.1 "More String
Operations" (http://www.python.org/doc/current/lib/typesseq-strings.html).

  -Arcege





From brett.calcott at clear.net.nz  Thu Jul 15 08:15:14 1999
From: brett.calcott at clear.net.nz (Brett Calcott)
Date: Fri, 16 Jul 1999 00:15:14 +1200
Subject: sample code for embedding python
Message-ID: <932040933.554739@estelle.paradise.net.nz>

I'm new to python & am interested in embedding it in some C++ code
(windows).
I've a made a start but the doc seems a bit thin concerning the details.

can someone point me at some sample code (Doesn't need to be windows).
Preferably something bigger than the example given in demo directory, and
something smaller than the PythonWin stuff.







From david at sundayta.co.uk  Mon Jul 19 09:31:01 1999
From: david at sundayta.co.uk (David Warnock)
Date: Mon, 19 Jul 1999 14:31:01 +0100
Subject: IDLE Keyboard Mappings
References: <3792EA67.B97AEB73@sundayta.co.uk>
Message-ID: <37932895.19099366@sundayta.co.uk>

Oops,

I have now found and read some more documentation.

a) Sorry I did not cc Guido (maybe he won't notice my message that shows
I had not read all the docs:-)

b) I have now found and read extend.txt and so have seen that I did
understand what is happening correctly and that it is planned to change
in the future.

Dave

-- 
David Warnock
Sundayta Ltd



From wolford at enews.nrl.navy.mil  Wed Jul  7 17:07:44 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Wed, 07 Jul 1999 17:07:44 -0400
Subject: win32 python as a dll?
Message-ID: <3783C1A0.9F80D390@enews.nrl.navy.mil>

I'm having trouble embedding python into an application on win32. I've
done it successfully for the last few months by compiling python into a
dll using my own home-brew makefile. I don't want to keep doing it that
way though; I'd like to use the standard installed python.

Even with version 1.5.2, my program crashes in some mysterious place
after start-up, in NTDLL somewhere. I originally fixed the problem by
compiling python as a dll, then having python.exe use that dll. Why
isn't the standard win32 python distribution compiled as a dll? If it's
not, you also get problems dynamically loading dll's that use python,
because those dlls link with the static python.lib and so does the
application. Thus you have two of those thread context things (it's been
a while since I last tackled this problem).

What do you think about fixing the standard win32 distribution?

Thanks,
Scott




From mnot at pobox.com  Wed Jul 21 18:40:31 1999
From: mnot at pobox.com (Mark Nottingham)
Date: Wed, 21 Jul 1999 22:40:31 GMT
Subject: ISP CGI  Python Support
References: <BFql3.4593$2D1.40042@newsr2.san.rr.com>
Message-ID: <009d01bed3ca$0a48cbb0$0301a8c0@mnot.net>

I'd be thinking along the lines of:

* your customer wants it
* if you don't get it, they'll go elsewhere

To add to this,

* It's easier than perl to install
* It's stable
* It's being used in a variety of applications (pointer to python home page)
* Other ISPs do offer it (scare tactics are your friend)

You really shouldn't have to "convince" them, capitalism being the theory of
govt de jour.

Cheers,


----- Original Message -----
From: Bruce Duncan <bduncan1 at san.rr.com>
Newsgroups: comp.lang.python
To: <python-list at cwi.nl>
Sent: Thursday, July 22, 1999 7:07 AM
Subject: ISP CGI Python Support


> Greetings,
>
> I would like to see a short document that I can send to my Web hosting ISP
> to convince them to support CGI programming in Python.  Some ISP's do,
> but others don't.  Because this may be of common concern, perhaps such
> a Python "pitch" can be stored at python.org in addition to a list of
> providers
> that do support Python CGI.
>
> Comments,
>
> -bsd-
>
> bsd at san.rr.com
>
>
>
>
>




From krodgers at tdyryan.com  Fri Jul  2 14:40:16 1999
From: krodgers at tdyryan.com (krodgers at tdyryan.com)
Date: Fri, 02 Jul 1999 18:40:16 GMT
Subject: Debugging embedded pgm under WinNT
Message-ID: <7lj12e$jct$1@nnrp1.deja.com>

I have Python 1.5.2 embedded in an application on WinNT (4.0, SP4) using
Visual C++ 6.0 SP2.  When I compile the project in Debug mode, even
though it links correctly to the python15_d.lib file, it won't run
because it complains about "can't find python15_d.dll".  So I got the
source, compiled the python15_d.dll, put it in WINNT/system32 right next
to python15.dll, ran my program . . . and now it couldn't find
"multiarray" (my embedded Python code uses the Numeric extensions).  I
verified that sys.path was correct and included the Numeric directories.
Do I have to build the Numeric extensions under Debug mode?  Anybody
seen something like this before?  Thanks in advance . . .
--
Kevin Rodgers  krodgers at tdyryan.com  kmrodgers at my-deja.com


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Darrell at p98.f112.n480.z2.fidonet.org  Thu Jul  1 09:08:57 1999
From: Darrell at p98.f112.n480.z2.fidonet.org (Darrell)
Date: Thu, 01 Jul 1999 14:08:57 +0100
Subject: Compiling Python for Win32
Message-ID: <00001639@bossar.com.pl>

From: "Darrell" <news at dorb.com>

> What I would like is if somebody who has successfully built Python 1.5.2
> using MSVC++ could do the "Export Makefile" command and send me the
> result. A simple text list of the individual source files that make all
> of the Windows versions of the various python executables, libraries and
> DLLs, along with accompanying defines would also be welcome, and either
> will earn somebody my undying thanks. :-)
>
> Thanks in advance, eh!
>

Ok. I'm waiting for the undying thanks part.
Hope this works for you.
--
--Darrell


begin 666 makes.zip
M4$L#!!0````(`+:(X29V[[%WL 4```@7```)````8G-D9&(N;6%K[5A[;]LV
M$/^[!OP=KFI0Q, BY8$5G5 #=6PY<"L_8,5+.Q at P9)&V.4NBH$<28]UWWTFB
M9,G/95FZ#5A@!"3OP;OCW?THOH$NLWP>\%D(+7I/;>Y1'XPP(HS##76I;X:4
M0*_;^*Q!F]GT![@V`USA+DP#0J8R";QJY76G#=+):;-]4Y. at 7@=)JE9P4D]8
MX SNF'MUB1M,HSDR=S7#:-QHT.-@<7?&YA'NPE!CX%&+S1 at E,O+.S, at .F3N'
MD,,./3(JTGHMW+A:V3#@-1I0EAA2FZ+9$KQ]>X at M42P5+.RX]Z;-R(:9N8:U
MO069KSP"R\R<68&Y(?VPH"[XD>O&OB6!+0A/5T#HC"6T<$'!,?%T`'>+`QXO
M6-QQ3)> S5PJ0YO[0!]-Q[.I6M!2C'%R<LI,N"H[YE**]=6/N5X8#G at 0L*F-
MFR\XLV@`,]RV[)3I[S- at WTF<3K,\DE+*Z>/[=S5HK5S38=:9SMPEZ&SJF_Y*
MJNW7E]K]E[7%230<]H?0<('M/&P6E(]Y*^?ZALAYW)CPAV#2N\4X]D:Z7H_9
M=2/>)9FZD5U2T!P,ZI8MTT=:K71O];K#B)@-C6;=M])QNE&YO';'-.;MCVY;
MG6%='J.MG9X8A]3QSOR4JUIY`]=TSEQH1D'('>C&*1:@9!2VF)]*O@$-<VR#
M(?=XJ#5'0T,3MIQ+B3,-70<UB4=B06V<YINW(HE=62!R1F\5+KA[\>-V9AS0
ML1G[LB47L25-76OT0-VK/Z%+L+9("!34OSK[B'TOH/$.:0R%)0XGD4UE/OU5
MVLUU;[T[EQF9'B)[6^2RM_31.TBWV6'Y8_JS:*X)4NP]X!^;@<M#["@L" N"
MRLEIG+_(YRP)\PN43)%P<*\B0=^A*).,%;5_.J^369KU6!N3P;#_J:ZXW.9S
M#DJW!<K=%2 at W7T#YA8'2OP2E`Y(LCSNN94<$,T?,!\U\B+_$ZS';X,D)'O?#
M`K4%4J^E78]NTO%=IW=UF0XG.&[U[PR<M;W-S) ]:X&$KVA;FQ>(XYB;;"Z@
M)Q8D-;_A8W%W9QFN/#SLRW-TFB-AI./J0Y++U<JUT;RZ1'RUL*'3-&3)TJ2M
M-VZ,M4:^=?XH(V4*)L;UT*C#&/.E6M$[O<^H$Y%EF2I,%X3&**#^U66<?;"D
MODMM,9D3)D9H6N!Q;B<3Q"EBSP7%)/>FE[$%"VIGPMRFZY$9A6(218RDJV1J
M90PXM#PQV3K )"SCM#H@=SZ(IL$JP.ZG/J2]&11BVZ#@<?O4H6YHVJJ+?%@R
MZLXZ0OX87_ P3&N!B*MVKMZ_PZ!&X0Y^K"M4[<1GMDV-JQ84_.^9X4*5\F:G
M8+FC`RI"?IAV]3SN_>M/XFP.]*&,GNV6M3W1)HIE7K2SV*DE.#EM:>U)NZ-K
M-1P7MJ^EY9RMU>#CAP_Q4LZ4Y,:V4,PE.NP1[,KN'8>0BZ3WQN^)6Q/R!.02
M=Y"#.IZ#7(G^?SMNH;='D LYF+T\PG$$W>*8'K7COXMP1$"<(V"N at S!'7 at CF
M)O\HS$W^A[GO"G,K_&S;B7.BHO8 at G<M)^B&>@%K7^+DYO)5V`Z HO+T0*(H[
ML0*/G*H!]<(")(HN^E* F/>6;4C,+/].H)A#@6S]EOOQ>VR\FC871,O!(%>;
M3I,*0IT?(%,D6Y[W//G'Q^?)%Z6#J?\LZY\N7[;^J?(Y$/?Z$^W+K3;L-73L
M20-#/ Q=2"F']J5CW!JGXO6$4"]Y/NCTFOJHE3U&),MK8,X?%^Y,/W[(4>/G
MH!@L9EB=FR(B&0I9D5MV[',?OGW[<S>KI[P?'/@N5R$)J(4^R)*(;?RZ5 ,E
M'8B\5]!Z>7T17;\X[5==4GS "'$)>D%+Q+VL?@'I_6K#LJ==:?=>%?]^!X3B
MHX$LWB5?S(JC0<R3W>B/ADVMCBC735IX4.KG5OGRM=WK5;0X55$K7[5>B2Y0
MJOZ,-:VRDAU_`%!+`P04````" "VB.$FB(SYA^,%``"7& ``# ```%]T:VEN
M=&5R+FUA:^U8ZV_:2!#_7"3^AZD;54&Z0!ZZJK5JJ01,SJUY"(/2GI"0L1?8
MP_9:?B1!U_O?;]9>&_-.+FW5.QT?HO7.>W9F?IM]!6UJ!2QDTPB:Y(XXS"<!
M&%%L4P8WQ".!&1$;.NWZ)Q5:U"&_P+49X@[S8!PMJ!>1H&J'?KGT4FN!='+:
M:-U4)% 4D*1R"3^4C O.X)9Z5Y=H9A+/D+^M&D;]1H4.`XMY4SJ+T19%O:%/
M+#JEQ*XB[]2,G8AZ,X at 8[%9515UJIXGFRZ4--UZB&UM"?>(0#$&"UZ^/<";J
MI8*KFG=G.M3>\#=7LG*\(/.%Q6"9651+,#>D[^?$@R#V/!YDDN>"\&0)-IG2
MA!;-";@F'A:@-9Y_OF$QUS4]&QSJD2JT6 #DP71]A\@%+<5D)P=9FZZBK;KF
M0N(JE4<DH+#LL3"D$P==F#-JD1"F:'P]-#/8Y\:!4SF=9/4EI933A[=O*M!<
M>J9+K3.=>@O0Z20P at Z54.:@R]?X?*^1EU>]W^U#W at .X\>!JN'_E6%78-T0MH
MV&;WX;@SP&QVAKJN<';=X%:23R]VUA0T>CW%<JKD at 91+[8&NN-067WVCH016
MNDX-K;?=WLQR]NYPT-3Z2G6$[FH=L8Z(ZY\%*5>Y] JNR8QZT(C#B+G0YA47
MHF0<-6F02KX"%4MN at R$/NJ\VAGU#%>Z<2TD\=5T'.4E)XD%EE)>?O[03U[)T
MY+S^,IHS[^+7[1(YK&;S$-;]N>#^-'2UW@%YKXF$+L'**2%04/_B[ /.QI!P
M"VDF"\ZPR1_2;I9H8?H^]G-T at .?.>G->I?;D$-G?(F]EA#SXQU@<>E3+(PQE
MJ5_1))XJP!^=@L<BG$DTC JRM9-37O7(YRYL&A0HF2(1[%Y%@KY#42;)%;7>
MG2OV-.T5[*AQK]_]J-0\YK 9 at UJ[";7;*ZC=?(;:[Q1JW4NH:2!5JR/-LYS8
MQDH3W[U&NFS(HU[ 9H'I)E at 8C@:6,Z(Y<Q.D3E.]'MZDZUNM<W69+L>X;G9O
MC8PP^&U<[_6TCC; G9:_JX9\:XZT+^A;BQ7H(RY@;VY@)!8DDV(CQJ)+[B):
M^GCDE^<8-$/"4,?=^Z3PRZ5KHW%UJ4Q""_& I"E+ML8MO7YCK#2R726 8E*F
M8VQ<]PT%1E at UY9*N=3ZA6H2G1:HSW1!*XY $5Y>\#&%!`H\XXF-F4[%"[T*?
M,2?Y0+"SG9F at F/:=Z6=LX9PXF3!SR&IEQI'XB&-JI[OVQ,H8<&GYXB.RG+?G
MZ6HA%GG0'$)DZ?SA at ER\.\<?YBV,)^$RQ-DIWZ?#'6JVXT -RR$@+O$BTY$]
M%,4&DO<U%HIPC,*C,:TY(KBL7;U]@RF.H]TBV&AHP.6'N).!]S/4\*]O1G-Y
M=[FF/#@<6!#)?!+E@)$?3O?ZHSC ?<-MD[@^UC)JYE\V9L6\*4Z*C>"*$"'!
MR6E3;8U;FJY6<%UPKI(.A6RO`A_>O^=;.5-27MM"G$L,]>.XF5U^#J&FG=YE
M?S!FCNTGH*:X!1U3\QS43$S\%S 3,W(<-9&).HOC3,?!E1_ 8WSZ]P*L+1#6
M%2BK(<K:QU#6EK=Q=?SSX>KX?UQ]%*[B<48I[HQ6$%O<_%9HN\1_0O?!K6BU
MIP&N:+U#D"OZ/+&+54#DD/C13PS!^5S:"<)9P#\(AG/,J5I_YH'\Q?V7TZF$
M^-SKY6K3SZ014>=[R!15+=]_GOS#P_/DB]+A)'B6]T^77_?^J?(YXG>Z8_7S
M0.UWZCJ.MIXA7L4NI)1#_:P9`^-T]6YD$S]Y+]$Z#7W8+#S ))35/2!_4+DU
M`_Z0)?/G,(XU4^SD'5*B*@KED;OXB%<.^/KUT?>Z)SZ>''B1D"%)L87Q5"61
M;?[25H%:NA"=4,,PJJO[\.KI;;_J-<4'G!#WK^_HB;@2*A>07NTV/'ORM7KO
M7?7;QR 4'\UE\2;[W;PXFL>\\(WNL-]0%;P;M9D=<_C(N\5:O\=M8(.,OJ;"
ME?4KVPLQ%-:&0<::]MJVT16T;%K=`)WGF%T+_&]02P,$% ````@`MHCA)B;L
M2Q:I!0``Y!8```H```!P87)S97(N;6%K[5CIC]I&%/\<)/Z'%R>*%JEK[Z%&
MJ15+8<&L2,PAO&B3"@D9>X IML?RL0MJ^K_WC3TVYNYVD[25R@<TQSM^[\T[
MQO,*.M0.6<2F,33)`W%90$(PX\2A#&Z)3T(K)@YT._5/.K2H2WZ"&RO"%>9#
M8(41"64G"JJ5E^T62*_/&JW;F at 2:!I)4K>!$RVC@'.ZI?WV%*B;)#*D[NFG6
M;W7H,K"9/Z6S!/50E!D%Q*932AP9::=6XL;4GT',8)\@&27IW2:JKE:V(+Q$
M"%LL`^(2A"[!FS='Z5+14 at EDVW^P7.IL(2U$K"&7>+ZP!&PKMV<%UA;WXYSX
M$":^S\U+O5MBGJS (5.:[L5S`IZ%1P2HC7N=+]C,\RS?`9?Z1(86"X$L+2]P
MB5J24G9S>GS*-+=5]JR%Q 5J)XTO#?LLBNC$1?5S1FT2P105;YIEA8<@'#R-
MLTD>3U*V<[9\][8&S95O>=0^-ZB_`(-.0BM<2;4C`C/D?UL<#Z7!H#> N@]T
M[X'3:/.H=R*O9XK81\4.>XS&W3OT9'=H&!HG-TRN)9WZB;LAH-'O:[8KDR6I
M5CIWAN911\P&9D,+[6R<*=I,LP->Y<2]X5VS/=#D$8)M=\4X)EYP'F94U<HK
MN"$SZD,CB6+F08?'682<2=RD8<;Y"G0,M"V"PN2!WA at .3%V N9!2:^J&`6KJ
MD!1!;22"+E at Y*;#<%05EL(KGS+_\>3<XC at G9=O\FEDN.I6'H]2ZH!Q6D^Q*L
M(0F&DO at 7YQ^P!D:$:\B\F$/QF).X1&:3WZ3]9 _VVPN9.I-CV\'.]I:]9!D<
M)W#I"0DG5>0N7>](W 6 /SH%G\5876 at 4ESB5UV<\CI'.6S at T+.WD@H21!P6)
M_3V"<DXNJ/7+A>9,L^C''!GW![V/FN(SE\T8*)TF*/?7H-Q^!N57"DKO"I0V
M2+(\:ONVFS at 8/V+>;V3#ACKJAVP66E[:RZ+1G>V.:$'<Q-1M=Z^OLF&WJ=\,
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M\,+AO9N/XE".E9R<H% H2IRH!^5LWL!:+LP2O#YKZJUQJVWH-1R7`-2RM,W7
M:O#A_7N^5!"E8;'+Q*E$.3W5J_*KQK%.Y63WQ1_:I\;.$SJ5N'4<%_*<3I4J
M^/?W*;3W5*="$NHN3I&<:F?<L:>Q_'=;FB-ZFB?Z6AO[FO--^]KX'^AKX__[
MV at _H:RO\)MO?V$36/*6UB2PZW-Q$LJ8:\5")&I$@+C4[41J_7ZLKJL6>9I>C
M_T'MKJCPLOU[8<D?'+V:50OL at _U^(3:;IFF",M]#+DBV@^!Y_,OE\_C+W-$D
M?!;ZI_-OHG\J?]%?N[VQ_OE.'W3K!A:>OBG>>2ZEC$+_W#;OS+/\+<0A0?H2
MT.XVC&&S>%A(U]<=MW at HN+="_C"C\N<=7O^GF(X[/"(>2H%1@#OY[0Y?O_[%
M>].3G at ..?&6KD+K51CMD27B8OQC50,D&(OH5-$!>WS1+CT@'16\(/@)"W'"^
M(Q)QZ=(N(;L\;2%[XI7UX$WPVUL@!)_T9/FF^-U0G/1B$?!F;SAHZ!K>4CII
M*8\V"[N]>:O:4_55Q)P)J6U>HEZ(<K!1!G+2+-<VD/P)4$L#!!0````(`+:(
MX2;^[:04S 4``(\9```*````<'ET:&]N+FUA:]U8;6_:2!#^7"3^P]2-JB!=
M[(1<HQXJ4L&8B!8,PHG2GB(A8R^P%^.U_)(FNMY_O[%W;6SS5IJFU1V?UKLS
ML\_,SLXS["L84,MG`9N%T"'WQ&$>\<$((YLRN"0N\<V0V* /6A\UZ%*'_ 9M
M,\ 9YH+W&"Z8*]N!5ZV\['5!.CI6NY<U"9I-D*1J!3^:7 9.X(:ZYW7<8AK-
M47J@&4;K4 at .=@<7<&9U'N ]%P< C%IU18LLH.S,C)Z3N'$(&FPS):$G3.[AU
MM5*"\!(AE%3&Q"$(78+7KW?*)::E',B>>V\ZU"XAS4RL(.=T/K,(+#/UYQ',
MDO:7!7'!CUPW=B^);DYY^@@VF=%D+5P06)IX1("[Q5&/)RRV7)JN#0YUB0Q=
MY at -Y,)>>0QHY*_DP)\>GS%)?Y:5Y)\4&FWN=SPU'+ CHU,'M%XQ:)( 9;EQT
MR_2W0=AZ&L?3-)\DOG+\\/:B!BIS`X9[M3S/H59B7:KML,>!?Z^U.)'&X^$8
M6B[0C<=-@^)!K^7=T!"9C_O:[$LPT:\PCOIUO]^,Q?M&O$ORZ49.P8 Z&C4M
M1R8/I%H9&VK3M_B8FRY>JRU1C(6'UU>=WK at IWR*\GB[&(5EZ)SZ7JE9>09O,
MJ0MJ%(1L"8,XKP+4C,(.];GF*] PL4H"F9-C3;T>&YH`<RHE^%O]/C22$"0(
M:K<BR=")!%CJ?";)U\_>K"?#+B/E@!>QG,58U+[6TJ&Q=8-D78(5)*&0,__B
MY#W6O(#$._ H9E#8]"]IE\ $L<H^";8(W5L7IS*UI[N6O;7EC>'8(< M5"NK
M%2GV#_!'9^"R$$L%#<*<IG)T'*<ERBWO;.KG5E)#`N)60V)]@Z%44Z3Y9#0>
M?F at J+G/8G($RZ(!R<P[*Y2=0_J2@#.N@]$"2Y=N>:SF1C0DAOD<J#CMXMWKZ
M>9T/]8[6OK[DXXDZU(UA7Q-?@[9JX+#KK9^A9RUPY3-NV&6YU=M8W"Y/(#P+
MDCLI@#MP^O#[Z1]82=F:Z>ST0;%7Z*J5MJ&>UYO3P,*:2_C%3J8FW7[KTE@%
M at ZV?)2IE%B9&>VPTX18/OUKI]_2/:!3+_QVWR">$R3OBN\0YK\L.G4(4$%\,
MYS85HR_4#3S&G.0#R<1VYF+%M.]-+Q4+%L1)[6#U7(W,*$S-1]3FL_;42 at 5P
M:'GB(_,OB*;!8X#UJ&&)8JQ0U_+)DKBAZ31<%,'<;6S.:(QI7-]!02I<(.DU
M>N=O+S!F4;A!(;XCH.#FGADN&E)67+(P#=L?1"@W7_+-2ZL#%LO%;<_>Q/Z6
M[EX!4KY&2G!TW-&ZDVZOK]5PG$-6XY<LG:O!^W?OXJE,*#GE=:582E2V?;21
MLOPNTK!YJ_93*6-B\TAMO at K?QB6B$?@NVT^AF&3?YR48.9CZOYR!>+2V"X at X
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MLH[(AUQB9.#V_L&'KU^_L:,[Z,U at QU_Q!B1AM= /61(1CI^1:J#P@<A^!1V0
M5SUP^65ID^F"X1T at 1#?UC$A$@]<\`]ZHE9 =V$QO;49_O =9-=T3R7Q7^FPH
M]D8Q2WAC>#U6M29V#P-F1PX)THIO'9RZ^7ZNP!L-=(OO4RLV;B]$Q2A4BE14
M7,?##GP;ABTMRQ.A;8CB2"WPX@^)XHIB=\'%IA$SY=L;1IR<WHZ2A9,S^8U<
MG]9Y&YQ[+/GAA_$\KA1ZWSVN3;:YEN> U<>_4$L#!!0````(`+:(X2:YX_&P
MJ0T``+UU```,````<'ET:&]N,34N;6%K[5QK;QLW%OU<`_X/K!(4-K"1'&=;
MM$($U+'EK%I9,B0;21<&A-$,)4\]+\S#L;K=_[YWWB0O1=H)G>UBE0_!6.>2
MA^\AS[V<%^3"M>,P"5<I.:/WU LC&I-YFCEN2-[3@,962ATRN3CY=4C.78_^
MC;RS$O at E#$BT26_#X/7W72>)]O>^'9V3SLN#T_/WAQTR&)!.9W\/_AC45N05
M^> &;XZ!9IFMP?YB.)^?O!^224CL,%BYZPRX7, at WB:CMKESJ=,%V965>Z at 9K
MDH9$GE47\AI.SH!^?T\HQK=0#)1H1CT*5>B0[[[36!;9=YBBCH)[RW,=H;Q-
M)FW!F32_A1FQK;I6&V()J3_=TH#$61#DE2S:F4F\W!"'KMP"2V\I\2WH+ )L
M>?OG/]BA[UN!0SPWH%UR'L:$/EA^Y-$^DPO;V$5']E9M;;N^==?)LQP\H@&8
MQ\LP2=RE!T6X#5V;)F0%Y'S5K'A;,12]<K"LQU>G1 X>?OSAD)QM`LMW[5=C
M-[@C8W<96_&F<ZC,LBS]9V>8#ZO9;#HC)P%QI1WO)GR7HU$XG5=S`8B=\%.R
MF%Q!:TZNQ^-!;CZ>YRS%GT'F<1F<7EX.;*]+'^C^WL75>."[3O77;'XZB.WR
MN23BI]W6ELW-I]=79Z/9H'L#Q1U-JN>4^M&KN+3:WWM!WM&U&Y#3+$E#GUSD
M(RZ!E%EZYL9ERA=D"$-.,-C?.QF/2;^H=L%R>-,N$)Y7T$.]QL.3"8%1\<VK
MGV%M26AN7Y;D\&;AA;;E43]T,H]VP^7O';F=M4S2V+)3E8EM0X5BE44<6QL]
M5;X0AI'"8ND&5F*[KM(D3:BJM,M\B=.699FM5C0&G"JK;E,8JRK<LY)$GXMO
MI;?:(MGZ?$(_<C5YY.O5PR-RRN>>RN#2M>_45/,TAJ5T-%78.*Z=:LOB>(M 
MA=,X#D)MZX%5&"<*@Q6TG+8L*R^T]"5>P4*5>MHB at 5E*DU1O%UO^(XH6AW_0
M0&60!?HAM*:I%:]5[006R\SU'#=8A6JS:CBJ%H;"*MK$[OI64RH_=7U5(X%)
M&&GRB&":J9:7W,2S4GBY^FJK>QHG\%)2&4&?P59"62 P\:WXM<+$]:TU5:Z(
MKA^%L8JE-%#/H-+&42UD;J ?]IZ;I$'HJ'HI-]'G$P9KK9%ON:KVAX9-;I5+
M\Z/67-_YWE;#^BQ at 1E5=K;:Z#1U]K4,GR2)-EY=%TN>5W-NQ?O7Q-S&UG&+/
MJS*"74(:9JH&#^@G+9EF^&CKE)^H8(%6M70$(T.Y)A4&:7BG,K%C_<XI"A/W
M06^U@=> $L_)E 9)"N=&I46^-<R7@)@F2ALX&BGRB:U at K1]6,5U3?;4+JTC5
M#?%ZZ?IK?48A=+?6*H$-JJT?ZXF[#BS]BSOQX!"F;8 at DM._H(SC#+'#H@\JB
MV$CI^=(XC/1T:9Q!2U!_J9P$E9DVMTVBM4EO\P5$:Z#/!S8`>J/PC@;N'\JZ
MY><9NK1LU0Q/L^@12VBZB?1&]_8/1UW76:K@:"O\R0V*$:(@V&3WL%^'/8FL
M%%O.ATH3F)DZ$\\5"XQ,RCKM[[58)S^,$OCGKD at 0IH0^P': 2=M[>9 ?T,'.
MOW/<F$'JC*HJ;\VHPB49U2GSC,Y_.AHXJ_)8#X?_Q>5L^LN@%X1>N Y)[^*,
M]#Z\(;WW'TGOGR[I38]);T0ZW>[-*+"]S*&=^N_+4W@\(YW)V?#=]?OR^<-H
M\N:X?%S \]GTP[S\ZWH^7)R-%\./E]/9%?QT'C%=V#::?0O8;T!]'C+X39[ 
M$7^ @MJDT"R$*K!E\N]@E$)W'1]!G4(`KL?PZZ="KB at 4CBJM1XX>_G[T$^FM
M0E2P^MU!>J[0$$[+M+_W;G[ZYGBP3. \>T?+YBU^6IR/3][/V^*%LN$"R9H\
M%O-WL_F W, (V]\;CR:_0K:P`[DK\RQ_J#*]HW% O3?'^8 at D&;S9J\>UXU9/
M^0R*PM K_H SB>.M*\1R[JVH-DMNJ5?G$WJT?;*RM,X^<YWR5V=IUP;P:$<U
M5;[F5\\>G*,H;)OK,OA^\=0T0JZ4]3M'#Z_IZZ/\'S1FDBUA04VIW_]4:EC0
MOIX'C1[D^PT:I);7#R IS*S^MAD'27(I#OK=LF]AT]8?O?GQAYS5*>5=*$._
M`_]!8_?"+)7FDR\1P.I'A?&6V4]Z\']^HNIW&CVQZ9OINU^J_F/&$I:<1 M.
M;$(@(S,A3!"8$,Y(2R+&B4H8;.0D! E"$L)%"4DT:,4CA BR$<*%PPO"56D9
MD4B&\?(0MFB$(00QDA#".#%(1 49",&-`"0BHO0CP6O11X0$N0?!@M"#<%'B
MD1@(X at ZR$&0=C#>"#H)X*4>$61%'@O'RC<2 $VZD."O92 Q:L48"UC*-!&H%
M&AG(2C,2G!5E$,S*,3*P$6)$D)5@,-:<Q.70MG'+"2X(#%0#CQ-99* J+2^L
MB' CJ6"@%5,PIER)&@%%`JB2":*)!.?E$F3 "R42.%/.?R2.( ->%L$P*XB(
M*"^%('1;]RK*RPD?(LA('E*H%CL0R,L<"!8$#H0WT at 9&&E$#0ZV<@3%6I!!1
M49Y N"!,2/%(WDQ(C$ &@@PAXDB 0 :B]( ,!-$!X:+<@ U:H0%AHL0@,>#$
M!0G.RPI;#!0Y<%*"B#(B at AQ2I>6%`P1SD@%".;$`H8),@'!>(!!A[G O at L*Q
M7CX96E6//6 at +6WG64=PA+P_.AN>+\]%X> C/S(;]L#Q3U[\=DI_?OLU_:HR*
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MS, at H8:8KR0II,F)!9C-4X=-:H,.4K6YG>"A5LA]B9.5 PY2-FHA(>9W1;*<R
M,J7(*RJ8QKJSU#X17:N(FFU95E 529'8:G225E*MA+11<,WV)B,`BZ2B-FR8
MF)&6$;,H.YOM7DZU1MQ(TC9.SBCB$G91+S?<[(S<CLA%*=[HT*Z$?$S:ZON&
MJ]JZ!Q"IX#DP6M':[R"2<OX(LX.*<V=(> 5?A^GJ-IX2"37O13'/7/M at I-2<
M@\8T=^G>D? R?A_3G+G72,+8.).,O7$;-Y2$C?%/F:Y?X]V2T7*>+]/,M=],
M0LRYU,SRU at XYQ,IYZ at QQEJ?HQL\GXVP=@&87J-I_*')R?D6C35MY)3%AZPTS
M-EE:-V=!]\3/D.8<\P^CJ]-_#.>?_9&3ZK'@Z?P5OGDB:?7'[]8;P9UM&ZX/
M&I'WJ9?:MK7VTZ/@<)/_EX/BOEJ3-PKVEEGG8/F(]_6;W>BUH0*(-7B.\VBY
MCC8A""(I'YM at MJI,:(.,UGQE&^(V: (1"_$49M\<13B&2-G&:!A]9]01'IB.
MB?PP7;VM"N3S") -<1Z2 at BB;.!7C9-LJR0? &)VA;/@,XA5#:\Q.%RXR1\(M
MA.V8'<5MT ]B%N*!#%>ZZ,AME69!L]LO+DH)\:(0)K-CK V 0LQ";)393FXB
MJS M%W-E=B*W(5LBK1#,9;21I:_9YWK%;AF_S_>&:X+6$"47S6:T1:M8.)&1
M#9$SSY<'V$D9F\@[LPW+!.XA6B&FSS Q$Q*(F,5P0:/+0AELB#C;$$3#%2TC
M&#%?&]AHN'IE6"0F9,(EC;U7RL-8$5OV.>=ZH81-7OFE(E499_/30_*9=Y/@
M;[!95NWUZG7W^^[Q\K@\KS)?`A5J_>1#]%^P:HMM5=M^6"T9\T!:/)YJY!E>
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M6O$^SY<0<]O[_P!02P,$% ````@`MHCA)J =_J2R!0``0Q@```L```!P>71H
M;VYW+FUA:^U8ZV^;2!#_7$O^'Z:TJF+I`K9SK5I4I'-L'+GU2R96VE,DA&%M
M[P6SB$?BZ'K_^PVP8,"O6KGT'CH^+3NOW\S.S at R\@@$U/>:S>0`=<D]LYA(/
MM""T*(,KXA#/"(@%PT'KLPI=:I.?X-+P<8<YX#X&2^8\B);O5BLO>UT07I^U
MNU<U`10%!*%:P1>%,\$YW%#GHHE&9N$"V0>JIK6N5!@R,)DSIXL0+5'4ZKO$
MI'-*+!%YYT9H!]190,!@IR815:G##AJO5DH at 7B*(LLR$V 3A"_#FS6'&6+F0
MP]ES[@V;6B6PF8X-ZIS,5Q:"::0N/8)1DGY8$@>\T'$B#^,0YX1GCV"1.8UI
MP9+ RL!S`K06A3[:,-EJ93 at 6V-0A(G29!V1MK%R;R#DM^4C'9RC-,V?%E7$G
M1!J5X^[GEF/F^W1F(X EHR;Q88ZFBXX9WCX0^T_D;);FE9!0SM;OW]6 at Y;HV
M-6.U0NV0H at 3RJ6JB])E,1A-H.4!W'C'UBX>[E6TCC6<\VK/8 at Z\/KS%RPVF_
MKT3L?2VR$K\ZH5U0T!Z/%=,6R9I4*X/KOK*B%G^;:&W%,Y-U8JAXN?9%,>(>
M3:\[O8DBWB+:WI"O`[)RS[V$JUIY!9=D01UHAW[ 5C"(4LM'R3#H4"^1? 4J
MYE:)(?-YHK:G$TWE:.I"[$ZKWP<YCDB,H':;YAFZ$2-+ at Y&Q)@R-M]O9<%!+
M^02*:!H1FG9?;0U!WFLAI at NPP<0%<NI?G/^"U<\GD84DCBD6':&('O&%W4SW
MYKNZ2*W9(;*[EXPP!P9U1#;[K<2Q.R"'.!(KU<J&)$0N`CYT#@X+L&10/\B)
M2J_/HE1%OM6=1;T<)57$<>Y5Q.D[%*62//7U\63T29$<9K,%`VG0`>GF`J2K
M+R#]2D$:-4'J at 2"*MSW'M$,+DX*_C]NX[.!]ZPTOFLERV%$OIU?)6L?]SNA&
MP[>NNW5V&!1SB:2O:*;+<N3;B-\J;R H$^+;68*;-[JZ"QZQO,R:=<3/D##M
MX^Y#G&_Q7>:R-M37/]<_8 UF6[BRG +)VJBN5BZU]D53F?DF%FN2%(1X2^_V
M6U?:!@_;<?PHE:G0M<N)IL M)DRUTN\-/Z-6[!QWB<ID@^N\(YY#[(NFB"Y!
MZ!./+Q<6Y2MTS7<9L^,7[$.6O> 4P[HWW)3-7Q([U<-LLED989"J#ZF5[%HS
M,V7 I>GRE\S!J*[+0GW=((T/=7PP4'XX\Q]]+&WR0U)Z0:*.Z9$5<0+#EAV4
MPAL@[[D8&.BH9>#Q&>82>ZC<NWC_#@,9!KLDHLL&$B)RC6 I"UFARF(WNOS$
MX[OG,I=IY7/GY*+AQMLH"J5+7 25+[D"O#[KJ%V]V^NK-5SGL-62ZYKNU>"7
MCQ^CK8PI/OQMH8B+E\FC;2@=&PXU(2L9`']L"]*M$YH0GR2.:'E*$XHM_ =:
M4!:0PSS4OCO*\^]J51;O52O>KWK8KZP3^I7^=_0K_?GZE?Y_OWIZOWK$;ZD]
M#8M?D)-:%K^<L4H\;R+[Q UR+8R7N6=O8+IUH(6E(']0$\N*MFC^GCGR1^2=
MG-0([&[C<:8V>8UO`^K\"*DBT73=I\FOUT^3STO[,^])Z$^7+Z(_53YKF<.1
MKGZY5B?#5A\+R%CC_V(:0L*A?NEIU]I9]KO"(F[\[=X;MOO3SN870$S8M-'L
MV_[&\**_)W+T#R:J^W.\=-M"/"5RN9'A._Z]#=^^?>]$=-I'_(%/8QGBX)KH
MBRCP.$>_=FH@)0M^!R3T0=R,D>6?/;M4%Q0?`,%'EV=$PJ<II0')5%1"=NHX
MNG?(^^M=R,KJD5#F9\!G0W$TC%G2:Z/II*TJ\;A2J.SFZ<F;'[ZVNH2,OB7&
M:L5)ZP7.T%H;G?S.L8."@)NSVW%,.&^(;\7FK)E,6[E/Z,P:O]HG9LX_PA=]
MGR^[CR]MVF8)?J&9'\(>5_%"]2X:+IC^$U!+`P04````" "WB.$FWJS4\L8%
M``"8%@``" ```'IL:6(N;6%K[5A[;]I($/^[2'R'J5M50;K8>;15SZJE$C"(
MUCR$@]*>D)"Q%]C#]EI^)*'M??>;M=?&0"#)Y=J[DRY*HGW,>V?FM^L7T*5V
MR"(VBZ%)KHG+`A*"&2<.9= F/@FMF#C0Z]8_Z="B+OD%+JP(5Y@/7UTZE9TH
MJ%:>=UH at O3QJM-HU"30-)*E:P8G&*> 8KJA_?H;BI\D<:;NZ:=;;.O08V,R?
MT7F".BC*BP)BTQDECHRT,RMQ8^K/(6:P*T9&.7JOB6JKE2WUSU']!L.0N 1-
MEN#5JP-4J5BI9%['O[9<ZFS96 A8&UOB^<(2L*W<DQ586]PW"^)#F/@^=RR-
M:8EYN@*'S&BZ%R\(>!8>#* V'FN^8#//LWP'7.H3&5HL!')K>8%+U)*4<H#3
M0U-FF:>R9RTE+DZ[Q_'2<,"BB$Y=5+U at U"81S%#IIDM6N$_]GE,XFN;Y(V4[
M1[?OWM:@N?(MC]K'!O678-!I:(4KJ;977&;U7Q;&TV<X[ ^A[@.]\Z!IM'G$
M.]G6-T6NHV*'W423WB5&L3<R#(V3&R;7DD[]Q-T0T!@,--N5R2VI5KJ7AN91
M1\R&9D,+[6R<*=HLJSLCRDG[H\MF9ZC)8S2UTQ/CF'C!<9A152LOX(+,J0^-
M)(J9!UV>71%R)G&3AAGG"] QO;8("H>'>F,T-'5ARHF4^E(W#%#3<*06U,9I
MJ at 4K)S4K#T-!%ZSB!?-/W^RFQ7X1VX'?M..4V]$P]'H/U+WBTWT)U at 8)AI+X
M9\<?L-=%A&O((E@;7]MO3V3J3*4#V\'>;>Z&QYS$)3*;_KY%M.$KN0T.;>/?
MH>U="^X,Y'I=XHX#_M 9^"S&+D*CN,2GO#SB>8MTWM*A86DG%R1<W"M([-\A
M*.?D at EJ_GFC.+,MVK(G)8-C_J"D^<]F<@=)M at G)U#DK[,RB_45#Z9Z!T0)+E
M<<>WW<3!G!'S0:,8XB]W^O3D-7:J:QO7FR#UFOK%J)V-KSJ]\[-L.,%QLW]E
MXJP5;)V;'- at +7/^"NENLM#?FQ,[V`EIJ0UK+6SZ4E7O+>!6 at Z+,3=(KAQLC 
MU9LT3:N5"[-Q?J9-(QN[-,E"DBY-6D:];:XELNWC118IYY^8%T-3 at S%F0[5B
M='J?4"2"Q3*3ERT(@4E$PO,SGEJP)*%/7#&9.U2,T+(H8,Q-)P@]CCL7.Y9S
M;04Y6;0@;L[,7+(>64DL)DE"G6S5F=HY`0[M0$S6!W=MX]$YKELD/A2>1\DT
M6D78T]2;K.&"@H2 at 4-\.B4?\V')5'^FP'-2[2 at 3).6;@05CV`A%4[9R_>XL!
M3>)=<BP9%.SQX]K9Y-4("OX/K'BA2D4#4["*61BK".!QVJ>+B/<O/HI3V=\=
M\NU<5=[)1/&7R[=D8[GS2O#RJ*FW)JV.H==P7%)>RZHT7ZO!A_?O^5)!E.;$
M+A.G$CWS,!3E=XA#0.1DU[^?"$,3YQ% )"X4AT0\!8A2\?]N&$)?#P,1$E!W
M>9C /8A5/)P'T:H4[_\:7K7?<,QR!&AY`K at Z"%S.PX +?^=?:9#&X?A4/I%?
M"YSZ)^%K\C]\/1"^RL<7Q?B(L,=IRRLM/![05OCFN at O11"$]'--$8>U#-5&Y
MJ2X\9Z)&)(A+*">ZXX_!N*)M[*!<;O5/PKFBP<OVM\*+/[CM:M8X$ `'@T)L
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M-XH^)\WJ:\.*/P%02P$"% `4````" "VB.$F=N^Q=[ %```(%P``"0``````
M```!`" `MH$`````8G-D9&(N;6%K4$L!`A0`% ````@`MHCA)HB,^8?C!0``
MEQ@```P``````````0`@`+:!UP4``%]T:VEN=&5R+FUA:U!+`0(4`!0````(
M`+:(X28F[$L6J04``.06```*``````````$`( "V@>0+``!P87)S97(N;6%K
M4$L!`A0`% ````@`MHCA)O[MI!3,!0``CQD```H``````````0`@`+:!M1$`
M`'!Y=&AO;BYM86M02P$"% `4````" "VB.$FN>/QL*D-``"]=0``# ``````
M```!`" `MH&I%P``<'ET:&]N,34N;6%K4$L!`A0`% ````@`MHCA)J =_J2R
M!0``0Q@```L``````````0`@`+:!?"4``'!Y=&AO;G<N;6%K4$L!`A0`% ``
M``@`MXCA)MZLU/+&!0``F!8```@``````````0`@`+:!5RL``'IL:6(N;6%K
64$L%!@`````'``<`B@$``$,Q````````
`
end



From alan at littleford.net  Sat Jul 10 18:41:10 1999
From: alan at littleford.net (alan at littleford.net)
Date: Sat, 10 Jul 1999 22:41:10 GMT
Subject: Another reason assignment expressions would be nice (?)
Message-ID: <7m9to6$6qg$81@rks1.urz.tu-dresden.de>

I need to generate some large text files containing test cases for a
project. I decided to exploit the formatted print '%' and the exec()
function. The whole thing is wrapped up with a class (say foo)
instance which supplies useful functions (such as SomeFunction()) so I
can do things like

	a = foo()
	a.setTemplate("'A test line %d' % self.SomeFunction()")
	...
	a.run(100)

Calling run applies exec to the defined templates and so
the idea is the result of this would be hundred lines of: 

	A test line <somenumber>

(Of course the actual example would be more complicated, and class foo
provides loads of useful generator functions). Now here is the rub, I
might want two lines generated with a related value. Lets assume
SomeFunction() throws a random number, so I can't just call it again,
so what I would like to do is this:

	a = foo()
	a.setTemplate("'A test line %d' % 
		(self.i = self.SomeFunction()"))
	a.setTemplate("'Same value as above %d' % self.i")
	...
	a.run(100)

But of course I can't do this because '=' is a statement in the first
template. I could define foo to have some 'assignation' function:

def assign(var, val)
	var[0] = val
	return val

and then use
	
     "'A test line %d' % self.assign(self.var, self.SomeFunction())"
    " 'Same value as above %d' % self.var[0]"

but then the thing gets messy because I have to declare the mutable
array self.var somewhere before using it. All in all it seems I have
to jump through hoops because i = 3 doesn't 'return' the value 3.

Or am I missing something frightfully easy here ..

Tnx
Alan Littleford
alan at littleford.net




From Holger.Jannsen at p98.f112.n480.z2.fidonet.org  Fri Jul  2 05:31:28 1999
From: Holger.Jannsen at p98.f112.n480.z2.fidonet.org (Holger Jannsen)
Date: Fri, 02 Jul 1999 10:31:28 +0100
Subject: Why must path be in args in os.execv?
Message-ID: <0000169d@bossar.com.pl>

From: Holger Jannsen <holger at phoenix-edv.netzservice.de>

Thanks, Jeff.

That's a good answer.

Ciao,
Holger

Jeff Epler schrieb:
> 
> Holger,
> 
> On Unix, read the manpage for execve and friends.  You will see that the
> first argument is the path to the executable, and the subsequent arguments
> describe the arguments (including argv[0]) and optionally the environment.
> By convention, argv[0] is the basename of the executable when the
> executable is on the PATH, and the full path otherwise.  (Another unix
> convention is that when the first letter of argv[0] is "-", the shell is a
> login shell)
> 
> DOS doesn't really have execve.  It's possible that the imitation is less
> perfect than it could be.  Instead, it may do the equivalent of entering
> the string
>         string.join(argv, " ")
> in the shell, i.e., ignoring the path argument.
> 
> This looks like a shortcoming in the DOS (windows 9x, NT) implementation of
> exec*, in a way that makes it work differently from the Unix it attempts to
> imitate.  This wouldn't surprise me, since I recently discovered that NT's
> spawnl takes a unix-like array of arguments (const char *argv[]) but when
> any element contains a space, it is split into two arguments in the child
> process.  So, I had to break some code that had worked fine on Unix to
> generate argv, by quoting each argument (something that would *not* work on
> Unix) ..
> 
> Bad enough that Windows wants to copy Unix, they have to consistently get
> it subtly wrong too.
> 
> Jeff



From Felix.Puetsch at p98.f112.n480.z2.fidonet.org  Thu Jul  1 08:45:11 1999
From: Felix.Puetsch at p98.f112.n480.z2.fidonet.org (Felix Puetsch)
Date: Thu, 01 Jul 1999 13:45:11 +0100
Subject: Reflection in Python?
Message-ID: <0000160d@bossar.com.pl>

From: puetsch at gmx.de (Felix Puetsch)

Why do you post it 3 times? Is it such important?

-- 
Felix Puetsch



From Cory.C..Albrecht at p98.f112.n480.z2.fidonet.org  Thu Jul  1 16:37:39 1999
From: Cory.C..Albrecht at p98.f112.n480.z2.fidonet.org (Cory C. Albrecht)
Date: Thu, 01 Jul 1999 21:37:39 +0100
Subject: Compiling Python for Win32
Message-ID: <00001638@bossar.com.pl>

From: coryalbrecht at hotmail.com (Cory C. Albrecht)

Hello all,

I'm trying to compile Python myself, so I can embed it in my own 
programmes. I can compile my own C/C++ files properly, but then linking 
with python15.lib (or with python15.dll to get a fresh import library 
right then and there) produces many undefined references such as the 
following:

Error! E2028: _object near * near spam_system( _object near *, _object 
              near * ) is an undefined reference
Error! E2028: _object near * near SpamError is an undefined reference
Error! E2028: PyMethodDef near SpamMethods[] is an undefined reference
Error! E2044: exported symbol W?spam_system$n(pn$_object$$pn$1$)pn$1$ 
              not found
Error! E2044: exported symbol W?SpamError$npn$_object$$ not found
Error! E2044: exported symbol W?SpamMethods$n[]$PyMethodDef$$ not found

I'm doubting that this is caused by different name-mangling between 
compilers (I use Watcom C/C++ 11.0) since otherwise my other programmes 
would never be able to use the various standard Win32 API DLLs in my 
<windows>\system32 directory.

What I would like is if somebody who has successfully built Python 1.5.2 
using MSVC++ could do the "Export Makefile" command and send me the 
result. A simple text list of the individual source files that make all 
of the Windows versions of the various python executables, libraries and 
DLLs, along with accompanying defines would also be welcome, and either 
will earn somebody my undying thanks. :-)

Thanks in advance, eh!

--
Cory C. Albrecht
http://www.sentex.ca/~cory/



From MHammond at skippinet.com.au  Thu Jul  8 04:04:54 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Thu, 8 Jul 1999 18:04:54 +1000
Subject: Coredump: python 1.5.2
References: <19990708142614.A3595@longford.cs.monash.edu.au>
Message-ID: <7m1lv0$9t4$1@m2.c2.telstra-mm.net.au>

Toby J Sargeant wrote in message
<19990708142614.A3595 at longford.cs.monash.edu.au>...
>The following code causes a coredump under python 1.5.2

I dont suppose you wondered why the stack trace was so huge? :-)
>
>class Stuffed:
>   def __getattr__(self,key):
>      print self
>
>print Stuffed()

_Before_ the "print self" line (but not on the same line!) add a:
   print "in getattr for", key

And it should become obvious why it bombs.  Why Python's built-in recursion
check mechanism doesnt kick in on your platform and prevent the coredump is
something I can't possibly answer!

Mark.






From eric at estinc.com  Tue Jul 27 20:06:56 1999
From: eric at estinc.com (Eric Lee Green)
Date: Tue, 27 Jul 1999 17:06:56 -0700
Subject: Problem with signals & exceptions...
References: <1279007017-12004369@hypernet.com>
Message-ID: <5EE9FDEE57D4E297.04B6823ECEDE1867.06F21AA225B85CAB@lp.airnews.net>

Gordon McMillan wrote: 
> Eric Lee Green writes:
> > I'm having trouble catching my thrown exceptions under FreeBSD 3.2.
> > I am using the Python 1.5.2 that came with FreeBSD 3.2 (well, on
> > disk 1 of the <n>-disk set). The following program works properly
> > under Red Hat Linux version 6.0 running a self-compiled Python 1.5.2
> > (in place of the 1.5.1 which comes with Red Hat 6.0, which does not
> > properly handle signals during socket i/o calls). How it works: it
> > listens to port 6664. When it gets a connection, it then sets an
> > alarm handler and alarm and then tries to read from the socket. If
> > it does not get data within 10 seconds, it then times out, at which
> > point the alarm handler throws an exception and aborts the socket
> > read.
> 
> Why not avoid the pain of signals (which have enough gotchas to
> destroy portability) and use a select with a 10 sec timeout. If
> there's nothing in the readable list when it returns, bail out.

Actually, I already came to that conclusion. The folks who do this stuff,
though, still need to know that something's not right on FreeBSD. The signal is
caught, the exception is thrown... and gets thrown up all over the table in the
example program that I posted.

-E



From fredrik at pythonware.com  Thu Jul 15 11:12:20 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Jul 1999 15:12:20 GMT
Subject: httplib supporting HTTP/1.1 [Q]
References: <19990715094013.A29589@starship.python.net>
Message-ID: <006f01beced4$6f95cef0$f29b12c2@secret.pythonware.com>

Bruno Mattarollo <bmatt at server.python.net> wrote:
> I was wondering if someone was working on a module to
> support the HTTP/1.1 protocol...

http://www.lyra.org/greg/python/httplib.py

</F>




From olivier.deckmyn at mail.dotcom.fr  Tue Jul 20 07:38:20 1999
From: olivier.deckmyn at mail.dotcom.fr (Olivier Deckmyn)
Date: Tue, 20 Jul 1999 13:38:20 +0200
Subject: [Q] __methods__ and __members__ special functions...
Message-ID: <7n1n7c$7ao$1@feed.teaser.fr>

Hi all !

I want to know all the methods available to a instance.

Here is a copy of the official FAQ :
"""
4.60. Why doesn't dir() work on builtin types like files and lists?
It should have -- and it does starting with Python 1.5 (currently in
development -- see Questions 1.13 and 2.10).
Using 1.4, you can find out which methods a given object supports by looking
at its __methods__ attribute:


    >>> List = []
    >>> List.__methods__
    ['append', 'count', 'index', 'insert', 'remove', 'reverse', 'sort']
""""

if I make the test with [].__methods__ it works as expected.
But when I use my own object :

class ClassA:
    def __init__(self):
        pass

    def myMethod(self, a)
        return a*2

if I try __methods__ on a fresh instance :

a=ClassA()
a.__methods__

this raises a NameError exception :(

Why ? Where is __methods__ declared ?
Why does this work for any instance in the brower.py provided with pythonwin
?

Please help !

Olivier.








From cbbrowne at news.hex.net  Mon Jul 12 19:46:27 1999
From: cbbrowne at news.hex.net (Christopher Browne)
Date: Mon, 12 Jul 1999 23:46:27 GMT
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <377B454C.6F59@mailserver.hursley.ibm.com> <7liguf$58o$1@Starbase.NeoSoft.COM> <m0JTEFA0DTf3EwLV@ugglan.demon.co.uk> <slrn7nss0n.5u0.wtanksle@dolphin.openprojects.net> <378077E1.5E65@mailserver.hursley.ibm.com> <slrn7o58l7.4vr.wtanksle@dolphin.openprojects.net> <3783305C.33FA@mailserver.hursley.ibm.com> <slrn7o7d7o.6la.wtanksle@dolphin.openprojects.net> <3785B997.2001@mailserver.hursley.ibm.com>
  <w6btdmw4mm.fsf@wallace.nextel.no> <3789ABAB.1E71@mailserver.hursley.ibm.com> <slrn7okdnn.ep3.wtanksle@dolphin.openprojects.net>
Message-ID: <n7vi3.21348$AU3.605286@news2.giganews.com>

On Mon, 12 Jul 1999 18:41:59 GMT, William Tanksley
<wtanksle at dolphin.openprojects.net> wrote: 
>(although I am curious as to whether an indent/dedent based syntax might
>be amusing in Lisp).

That would provide opportunity for the implementation to be criticized
inanely by *absolutely everyone.*

Not only those that see Lisp as "too many parens," but also those that
knee-jerk into thinking that Python must be worthless due to
"indentation-sensitivity."

-- 
"I am a bomb technician. If you see me running, try to keep up..."
cbbrowne at ntlug.org- <http://www.ntlug.org/~cbbrowne/langlisp.html>



From kperacles at geocities.com  Thu Jul  8 10:32:51 1999
From: kperacles at geocities.com (Karl Putland)
Date: Thu, 8 Jul 1999 09:32:51 -0500
Subject: Python/PythonWin and DDE?
Message-ID: <3784b800.0@news.dbtech.net>

Are there any docs about PythonWin and DDE?

Specifically can Python be used as a DDE client?

Karl Putland
kperacles at geocities.com


From chad at vision.arc.nasa.gov  Fri Jul 23 22:48:52 1999
From: chad at vision.arc.nasa.gov (Chad Netzer)
Date: Fri, 23 Jul 1999 19:48:52 -0700
Subject: int(long(-sys.maxint-1)) fails on Linux
References: <000301bed4e6$44f499c0$392d2399@tim>
Message-ID: <37992994.F4A6A30@vision.arc.nasa.gov>

Tim Peters wrote:

> [Chad Netzer]
> > ...
> > Here is some assembly of the same region, compiled with the -O1
> > flag, which DOES exhibit the incorrect behavior.  Notice that it is
> > quite different from the previous code,
>
> Yes, but despite being wrong, it's much faster <wink>.

Hey that's what I thought too. :).

>
> So it's as if "x << 1" had never been computed (which the lea does, btw --
> it's tricky!) or tested against 0.

Yeah, I noticed that the test seemed to be missing, and that lea probably shouldn't
set any condition codes (and as you point out, it would be very bad if it did).  I
am curious as to how the x<<1 is computed be lea (I don't understand the
0x0(,%edx,2),%eax notation...


> Yup!  You should report it to "them", of course -- such things rarely fix
> themselves ...

Naturally; Since Oliver seems to have been able to pass it on to some PGCC personnel
for comment, I'll see that the Mandrake persons know about it (at the very least, I'll
post it to their support lists for others trying to use Python 1.5.2 w/ Mandrake 6)

Chad






From Richard.Jones at fulcrum.com.au  Wed Jul  7 21:17:33 1999
From: Richard.Jones at fulcrum.com.au (Richard Jones)
Date: Thu, 8 Jul 1999 01:17:33 GMT
Subject: GD/ImageMagick/PIL 
In-Reply-To: Message from Per Kistler of 1999-Jul-7 13:41:56,
            <37833D04.8367425A@fnmail.com> 
References: <37833D04.8367425A@fnmail.com>
Message-ID: <199907080117.LAA01597@envy.fulcrum.com.au>

[Per Kistler]
> I'm trying PIL right now, but was wondering, whether
> there is a Module around the libgd as well. 
> In Perl it's called GD. 
> And: Does PIL all, that ImageMagick does, or does one
> need both?

   Look in the Modules section on www.python.org for a link to the python gd 
module. I'd recommend that you stick with PIL (using PNGs) unless you _really_ 
need to use GIFs.


       Richard





From kikutani at sprintmail.com  Wed Jul 28 01:30:47 1999
From: kikutani at sprintmail.com (Kikutani Makoto)
Date: 28 Jul 1999 05:30:47 GMT
Subject: Qt
References: <7nkdce$t1c$1@oak.prod.itd.earthlink.net> <199907272335.JAA06521@envy.fulcrum.com.au>
Message-ID: <7nm4i7$t59$1@oak.prod.itd.earthlink.net>

On Tue, 27 Jul 1999 23:35:54 GMT, Richard Jones <Richard.Jones at fulcrum.com.au> wrote:

> > mako% python aclock.py                /usr/local/src/Lang/pyKDEsrc-0.8/exampl
> > es
> > Traceback (innermost last):
> >   File "aclock.py", line 4, in ?
> >     from kde import *
> >   File "/usr/lib/python1.5/kde.py", line 7, in ?
> >     import libkdec
> > ImportError: /usr/lib/libkimgio.so.2: undefined symbol: png_set_strip_16
> > 
> > What's wrong ?
> > My Linux Box is Debian potato, and kdelibs2g is version 1.1.1-19990523.
> > Python is 1.5.2 (python-gtk 0.6.1-1)
> 
>    Not sure - do you have the PNG library version that Qt requires? The example 
> works for me under redhat 6.0 with qt 1.44 and png 1.0.3 (not sure what the 
> actual requirements of Qt are though).

I have two PNG libraries (I don't know why).

  libpng2         1.0.2b-0.1     PNG library - runtime
  libpng0g        0.96-5         PNG library - runtime

I didn't know Qt requires PNG library, because qt1g(1.44-3) package
doesn't depend on libpng*.

Another related package is:
  kdelibs2g       1.1.1-19990523 KDE core libraries (runtime files)

Thank you.

-- 
Kikutani, Makoto  kikutani at sprintmail.com




From tim_one at email.msn.com  Thu Jul 15 16:17:31 1999
From: tim_one at email.msn.com (tim_one at email.msn.com)
Date: Thu, 15 Jul 1999 20:17:31 GMT
Subject: Getters and Setters
References: <000101becedb$1f4c3400$91a02299@tim> <m3emi9x07h.fsf@greebo.nodomain.de>
Message-ID: <7mlfkf$qdh$1@nnrp1.deja.com>

[about whether Tim's GetterSetter code created cycles]

[Bernhard Herzog <herzog at online.de]
> The cycle doesn't involve self, but self's dict will hold an
> indirect reference to itself. The instance object will be
> deleted, so it's __del__ method is called, but its __dict__
> won't be deleted:

Ah -- got it.  Thanks for the explanation.  self.__del__ is called,
though, so the cycles are easy to break:

def __del__(self):
    self.__dict__.clear()

even-tested-it-this-time<wink>-ly y'rs  - tim




Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From wtanksle at dolphin.openprojects.net  Fri Jul  2 12:58:37 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Fri, 02 Jul 1999 16:58:37 GMT
Subject: Tuples cf. Expanded objects??
References: <377bbde1@news.xtml.co.uk> <7lh375$5ea$1@news.udel.edu> <377CAC46.508D502F@iname.com>
Message-ID: <slrn7nprts.3a4.wtanksle@dolphin.openprojects.net>

On Fri, 02 Jul 1999 22:10:46 +1000, Gary Stephenson wrote:

>> There is probably something in the the FAQ.  BAsically, tuples are stored
>> more compactly and can be used as dictionary keys (because they ARE
>> readonly).

>How similar are Python's "Immutable Tuples" to Eiffel's "Expanded
>objects".  If one was to try to implement immutable tuples in another
>language, would an "expanded objects" implementation be the best, or even
>appropriate?

No similarity.  Expanded objects are objects which are handled by value
rather than reference.  Immutable types (which include tuples) are types
whose dimensions cannot change.

An instance of an immutable type may be created in expanded form.

>gary

-- 
-William "Billy" Tanksley



From aahz at netcom.com  Thu Jul  1 11:12:30 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 1 Jul 1999 15:12:30 GMT
Subject: Why must path be in args in os.execv?
References: <377A2590.D95D72FE@phoenix-edv.netzservice.de>
Message-ID: <7lg0gu$5ov@dfw-ixnews13.ix.netcom.com>

In article <377A2590.D95D72FE at phoenix-edv.netzservice.de>,
Holger Jannsen  <holger at phoenix-edv.netzservice.de> wrote:
>
>there's again something I don't understand. Hope you do.;-)
>
>I'd like to start a windows-executable with some arguments.
>Ok. 'os.execv' or anything like that is the thing.
>
>But if I don't repeat the path with the arguments, the execution fails.
>Better: Something happens, but not the correct execution.

If you look in a standard C-library docs, you'll see the same thing.
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From chad at vision.arc.nasa.gov  Thu Jul 22 04:54:40 1999
From: chad at vision.arc.nasa.gov (Chad Netzer)
Date: Thu, 22 Jul 1999 01:54:40 -0700
Subject: int(long(-sys.maxint-1)) fails on Linux
Message-ID: <3796DC4F.224795DA@vision.arc.nasa.gov>

Compiling python 1.5.2 on Linux Mandrake 6.0,  the "make test" step fails
because it gets an OverflowError during test_types.py on the following line:

int(long(-sys.maxint-1))

test_long.py fails also for the same reason.  I've tried this on another Linux
setup, using the default system python 1.5.1 compiled with just the straight
gcc (not egcs or pgcc) and got the same result.  Is this just the expected
behavior on Linux (in which case maybe the tests should be changed to reflect
this)  Can I ignore it?  (This doesn't overflow on my SGI, and I assume the
Pentium uses 2's-complement.  Perhaps this is a glibc issue?)  I should look
at the code, but I'm sure others must have come across this...

BTW, as a learning experience, I'm building all of Oliver Andrich's Python
packages for the Mandrake 6.0 architecture, and will make the results available
when I'm finished, unless Oliver beats me to the punch and obviates the need.
Our application depends on getting python installed easily, so I took the initiative.

Chad Netzer
chad at vision.arc.nasa.gov





From phd at emerald.netskate.ru  Tue Jul  6 08:06:04 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Tue, 6 Jul 1999 16:06:04 +0400 (MSD)
Subject: What's ZOPE anyway?
In-Reply-To: <3scg3.1267$B95.91660@typ42b.nn.bcandid.com>
Message-ID: <Pine.LNX.4.04.9907061557470.21542-100000@emerald.netskate.ru>

On Mon, 5 Jul 1999, Jayr Magave wrote:
> I went through www.zope.org and read everything I could, and found the
> following definition:
> 
> "Zope is an Open Source application server and portal toolkit used for
> building high-performance, dynamic Web sites. "
> 
> ... and the rest of the page goes on to describe ZOPE.
> 
> I am sorry, but what's ZOPE? :-)
> 
> Any code snipplet? Is this a language extension to Python?
> 
> Thanx,
> 
> Jair

   Zope is not a language or language extension. It is application server.
   Let look into it in a little more details.

   Zope consists of few components - all of them written in python.

   There is DocumentTemplates - powerful mechanism to separate data and
presenation.

   There is Object Database - OO database.

   There is Object Publisher - mechanism to publish python objects on the
web.

   And there is PCGI Server - a way to implement Persistent CGI. Persistent
CGI does not fork off HTTP server on every call, but lives in memory for a
while.

   All these components can be used independently. I, for example, use
ZTemplates in some projects, and I am very sorry I hadn't learned earlier -
ZTemplates are perfect!
   But when combined in one server - Zope - these components provides an
extremely powerful platform for creating, managing, programming web-sites.
Zope, e.g., provides so called "aquisition" - name searching mechanism,
which allows context-dependent rendering. And so on.

   There are some more, but these are basic Zope components.

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From kernr at mail.ncifcrf.gov  Tue Jul 20 21:43:48 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Wed, 21 Jul 1999 01:43:48 GMT
Subject: Scientific Plotting?
References: <19990721010203.96101.qmail@hotmail.com>
Message-ID: <37952384.14766416@news.erols.com>

On Tue, 20 Jul 1999 18:02:01 PDT, "Michael Huster"
<Use-Author-Address-Header@[127.1]> wrote:

>I am also trying to find an adequate *object* based plotting package. I am 
>writing some signal processing software and want to code plotting into the 
>classes I am defining.

[snip]

>Gnuplot: Runs gnuplot through a pipe. All of the data (upto many MB) also 
>has to go through the pipe. Unsat. I want to plot efficiently.

The Gnuplot module as updated by Michael Haggerty
(http://monsoon.harvard.edu/~mhagger/download/) puts the data into
temporary files.  Pipes are used for commands only.  If you haven't
used it recently, you might want to try it again.

[snip]

>Piddle & graphite: looks good, but not mature enough. We are keeping an eye 
>on it. Probably will try it.

True, they're not quite mature enough for a lot of work.  However,
since there don't seem to be many other contenders available for you,
banging on Graphite and PIDDLE to add the capabilities you need might
be faster.

>PyMat: I want to get away from needing Matlab licenses
>
>I am kind of bummed. If we wrap intelligent python wrappers around 
>something, we will post it.

Hope your problems get solved!

>Michael Huster
>mhuster at hotmail.com

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From news at dorb.com  Tue Jul 27 17:37:41 1999
From: news at dorb.com (Darrell)
Date: Tue, 27 Jul 1999 21:37:41 GMT
Subject: How to use a variable form main script in a module.
References: <7nhmt5$su0@enews3.newsguy.com> <wkg12bscly.fsf@ifi.uio.no>
Message-ID: <8E10B3B02dorb@209.242.64.100>

import sys
main= sys.modules['__main__']
main.setDebug(__name__)

larsga at ifi.uio.no (Lars Marius Garshol) wrote in 
<wkg12bscly.fsf at ifi.uio.no>:

>
>* Peter Torstenson
>|
>| This is probaly a silly question from a newbie.
>| In my main script I put a value for to a variable called "x".
>| Now I want to use the "x" variable in a module.
>| The module is imported into the main script. How to do this?
>
>From the way you describe it this sounds like something you shouldn't
>do. Communication between different pieces of code through globabl
>variables is in general frowned upon, and with good reason. It's good
>for some things, but for those the variables tend to reside in a
>module, and not in the global namespace.
>
>Feel free to explain the reason why you want this if you want a
>verdict on whether it's a good or bad idea.
>
>[...wrong info in previous version of this post removed (global
>doesn't work)...]
>
>--Lars M.
>




From da at ski.org  Sun Jul 11 13:16:38 1999
From: da at ski.org (David Ascher)
Date: Sun, 11 Jul 1999 10:16:38 -0700 (Pacific Daylight Time)
Subject: Binary pickle portability
In-Reply-To: <7m9tm6$6qg$20@rks1.urz.tu-dresden.de>
Message-ID: <Pine.WNT.4.05.9907111015400.94-100000@david.ski.org>

On Fri, 9 Jul 1999, Olaf Delgado wrote:

> On Wed, 7 Jul 1999, David Ascher wrote:
> 
> > There is a bug in pickling of NumPy arrays, which I expect will be fixed
> > in the next release.  Binary pickles works across byte-order and byte-size
> > platforms (when it's bug-free =).
> 
> Thank you for the information! It's a bit painful to have to convert the
> arrays back and forth right now, but good to know that it's not my fault.
> 
> I am not sure I understand what you mean by byte-order and byte-size
> platforms. What's a counterexample?

My statement was terribly worded.  What I meant was that binary pickles
can be transferred between platforms which have different byte ordering
(LSB vs MSB) and between platforms which have different word size (32 bits
vs 64 bits).

--david




From fredrik at pythonware.com  Thu Jul 15 07:20:41 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Jul 1999 11:20:41 GMT
Subject: Oneliners?
References: <378D99C4.BDACE15B@serop.abb.se>
Message-ID: <004601beceb4$15711440$f29b12c2@secret.pythonware.com>

Thomas S. Strinnhed <thstr at serop.abb.se> wrote:
> I heard there was a webserver in one line of Perl,
> does Python match that (or two or three...)?

how about:

    from asyncHTTPServer import *; HTTPServer(80, FileSource("/web")); mainloop()

(one line, three statements)

writing the asyncHTTPServer module is left as an exercise
(or you can wait for my Python/Tkinter book...)

</F>




From seanr at infi.net  Fri Jul 30 17:18:45 1999
From: seanr at infi.net (Sean Robertson)
Date: Fri, 30 Jul 1999 17:18:45 -0400
Subject: Include statements with python scripting
References: <7nsu3t$kui$1@nw001t.infi.net> <slrn7q44ck.8kv.wtanksle@dolphin.openprojects.net>
Message-ID: <7nt4pr$qin$1@nw001t.infi.net>

William Tanksley <wtanksle at dolphin.openprojects.net> wrote in message
news:slrn7q44ck.8kv.wtanksle at dolphin.openprojects.net...
> On Fri, 30 Jul 1999 15:24:38 -0400, Sean Robertson wrote:
> >How do you do an include statement in python?  I know that in Cold Fusion
it
> >would be "<cfinclude template="blue_header_noad.html">" and in an shtml
file
> >it would be "#include
http://data.pilotonline.com/connect2/blue_header.html"
> >but I need to know how this is done in Python scripting.  This script is
> >being embedded in a template for a classifieds page.  I have something
like
> >this so far:
>
> ><html><head><title><!--#var name--> Search Results</title></head>
>
> This isn't Python -- I'm sure you're trying to ask about Zope.  Remarkable
> product, written in Python (lovely plumage).  But not Python.  Try
> visiting www.zope.org for more info.
>
> >    <!--#include http://data.pilotonline.com/connect2/blue_header.html-->
>
> I wasn't aware of a #include in Zope; I've been using #var.  Generally
> speaking, it doesn't make any sense to include an entire HTML page inside
> another, so I think you might possibly be confused.

I was told that the program that uses this template uses python, but am
checking out zope nonetheless.  The reason I need to include and external
html fragment is that I want the page to have two seperate interfaces
depending on which part of the site the user is coming from.  If they are
searching CareerConnection, the mainclass variable will be Employers and the
results page needs to fit the interface of the CareerConnection site,
otherwise, they should get the old default interface.  I was using
"#include" only as a placeholder until I figured out how to do it in python.
If you know how, please let me know.





From rodh at hou.valmet.com  Thu Jul 15 10:08:56 1999
From: rodh at hou.valmet.com (Rod Haper)
Date: Thu, 15 Jul 1999 14:08:56 +0000
Subject: python?
Message-ID: <378DEB78.34747906@hou.valmet.com>

http://207.240.225.250



From fredrik at pythonware.com  Thu Jul  8 04:12:55 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 8 Jul 1999 08:12:55 GMT
Subject: Coredump: python 1.5.2
References: <19990708142614.A3595@longford.cs.monash.edu.au>
Message-ID: <009401bec919$af755420$f29b12c2@secret.pythonware.com>

Toby J Sargeant <tjs at longford.cs.monash.edu.au> wrote:
> The following code causes a coredump under python 1.5.2
> 
> class Stuffed:
>    def __getattr__(self,key):
>       print self
> 
> print Stuffed()

python looks for __str__ and/or __repr__ attributes when
you print an instance object, so that "print self" sends you
right back to __getattr__.

you run out of stack space, in other words.

</F>




From aahz at netcom.com  Thu Jul 22 19:22:35 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 22 Jul 1999 23:22:35 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
References: <001101bed3f2$5e8874a0$642d2399@tim> <14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net> <7n7bo8$204@dfw-ixnews19.ix.netcom.com> <37974758.FF2DAE55@callware.com>
Message-ID: <7n893r$h45@dfw-ixnews17.ix.netcom.com>

In article <37974758.FF2DAE55 at callware.com>,
Ivan Van Laningham  <ivanlan at callware.com> wrote:
>Aahz Maruch wrote:
>> 
>> I'm afraid I can't agree that there's *ever* a good time to use "from
>> foo import *" in a multi-programmer environment.  Too often, you have a
>> programmer new to a particular chunk of code trying to debug it; not
>> knowing exactly where a method/function call is coming from vastly
>> increases debugging time (particularly when there are hundreds of
>> modules and scripts).  If you want to use an unadorned class/function,
>> you can just take the trouble to explicitly list it as "from foo import
>> bar, baz".
>
>All true, except for perhaps Tkinter.  And my soon-to-be-released
>mayalib-1.2.  Both are specifically designed to use 'from xxx import *'
>and both carefully refrain from polluting your namespace.  But possibly
>you don't use Tkinter in your projects, Aahz?  (I think I can safely
>assume you don't use any Mayan dates. ...;-)

The point is that a programmer looking at a traceback has no way of
knowing from looking at a line of code in isolation where the
function/class comes from.  At least with "from foo import bar" a search
through the source will discover which module/package bar comes from.

Although polluting the namespace is an issue (and as Greg points out, if
someone subclasses, you definitely do have a greater likelihood for
namespace collisions), I'm trying to hammer on the idea that specifying
where things come from is a critical issue.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From amosg at gci.net  Sat Jul 31 03:28:42 1999
From: amosg at gci.net (Amos Gingerich)
Date: Fri, 30 Jul 1999 22:28:42 -0900
Subject: Mac Install
Message-ID: <37A2A5A9.48D17D9@gci.net>

I'm having a problem with the Mac install.  I downloaded the file Python
151.sit which expanded to 'Python 1.5.1 Installer'.  When I try to run
the installer file,  it attempts to run Stuffit.  I did a 'get info' on
the installer and see that it's kind is 'Stuffit document'.  Seems a
little weird to me.  Any ideas anyone?  Any help is appreciated.

Amos




From amk at mira.erols.com  Fri Jul  9 19:38:32 1999
From: amk at mira.erols.com (A.M. Kuchling)
Date: 9 Jul 1999 23:38:32 GMT
Subject: Python Code Repository?
References: <378645AA.5430A026@msfc.nasa.gov> <19990709163127.A1964@toast.internal>
Message-ID: <slrn7od2d6.dl.amk@207-172-57-127.s127.tnt2.ann.va.dialup.rcn.com>

On Fri, 9 Jul 1999 16:31:27 -0400, jam <jam at newimage.com> wrote:
>does anyone have any further suggestions or ideas on how we can get this
>rolling?

Michel Pelletier (the one at Digital Creations) was working on a
ZTrove at one point, I believe.  (For information on the whole idea of
Trove, see http://www.tuxedo.org/~esr/trove/).  I don't know what's
happened to ZTrove; anyone from DC care to comment?

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
    "If you had stayed with us, we could have given you life until death."
    "Don't I get that anyway?"
    -- Stheno and Lyta Hall, in SANDMAN #61: "The Kindly Ones:5"




From cgw at fnal.gov  Tue Jul 20 23:13:16 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Wed, 21 Jul 1999 03:13:16 GMT
Subject: Merging two lists as a dictionary
In-Reply-To: <19990720211946.A12202@pooh.frostnet.net>
References: <19990720211946.A12202@pooh.frostnet.net>
Message-ID: <14229.15052.805319.307262@buffalo.fnal.gov>

Chris Frost writes:
 > I have two lists and would like to merge them into a dictionary, with one
 > list forming the keywords and the other the values in the dictionary. How could
 > this be done? "map(None, keyword_list, value_list)" looked great, but it
 > returns a list of pairs (I'd like to be able to specify values according to
 > keywords...).

Unless I'm missing the point (which happens more than I'd like to
admit!)  I think what you're looking for is just:

dicp={}
for key,value in map(None,key_list,value_list):
    dict[key]=value




From billtut at microsoft.com  Tue Jul 13 22:43:35 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Wed, 14 Jul 1999 02:43:35 GMT
Subject: making a Windows NT service out of a python program
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F2E@RED-MSG-50>


> From: Mark Hammond [mailto:MHammond at skippinet.com.au]
> 
> 
> Andrew Cooke wrote in message <7mfot5$n30$1 at nnrp1.deja.com>...
> 
> >The worst thing about them is that they can be tricky to 
> debug.  If they
> >are running in the system account then they have a restricted
> >environment and the standard output channels send useful data to
> 
> he he - not with Python :-)  Just import win32traceutil at 
> the top of your
> service, and you can see all your print statements in Pythonwin :-)
> 

Heck, if you give it permission to interact with the desktop, you might
actually be able to use Mark's cool GUI debugger. :) (No I haven't tried
this...)

Bill



From esr at thyrsus.com  Fri Jul 30 00:55:36 1999
From: esr at thyrsus.com (Eric S. Raymond)
Date: Fri, 30 Jul 1999 04:55:36 GMT
Subject: Need help with TCP/IP client access from Windows
In-Reply-To: <NDBBIAPHEKGOBIHCBLIIOECECAAA.mcfletch@vrtelecom.com>; from Mike Fletcher on Thu, Jul 29, 1999 at 10:07:21PM -0400
References: <slrn7q1ej3.18d.esr@snark.thyrsus.com> <NDBBIAPHEKGOBIHCBLIIOECECAAA.mcfletch@vrtelecom.com>
Message-ID: <19990730005536.U7149@thyrsus.com>

Mike Fletcher <mcfletch at vrtelecom.com>:
> About the only way I know to do this at all reliably...
> 
> import os
> os.system( 'start %s'%url )
> 
> Which does whatever the user has defined as the proper thing to do with
> urls, normally, launch IE or Netscape.

Really?  That sounds perfect.  Just for my information, what is `start'?
-- 
		<a href="http://www.tuxedo.org/~esr">Eric S. Raymond</a>

The common argument that crime is caused by poverty is a kind of
slander on the poor.
	-- H. L. Mencken



From dborgelt at xxxxxxcom  Thu Jul  8 10:21:35 1999
From: dborgelt at xxxxxxcom (Dave Borgelt)
Date: Thu, 8 Jul 1999 09:21:35 -0500
Subject: Trouble with HTMLGen, HTMLtest
References: <e3UDcTMy#GA.342@nih2naae.prod2.compuserve.com> <37845270.BD3E08DD@starship.python.net>
Message-ID: <#dvnA1Uy#GA.131@nih2naaf.prod2.compuserve.com>

Sorry to be so slow here but how would I be 'in' that directory.  I added it
to my PythonPath so that the import works and the traceback shows that the
system knows about looking in that directory
(c:\programming\python\HTMLGen\filename.py).

What else would I need to do to get this to work?  To be 'in' that
directory?

Dave Borgelt

Berthold H?llmann <bhoel at server.python.net> wrote in message
news:37845270.BD3E08DD at starship.python.net...
> Dave Borgelt wrote:
> >
> > I am having a problem with HTMLGen that is driving me nuts.  It has to
do
> > with running the test code in HTMLtest.  (HTMLtest.test()).
> >
> > I expect the solution is simple but I can't see it.  The path to the
HTMLGen
> > files is in my PythonPath (hence the import works) but as you can see
the
> > test fails.
> >
> > -----------------------------
> >
> > >>> import HTMLtest
> > >>> HTMLtest.test()
> > Traceback (innermost last):
> >   File "<interactive input>", line 1, in ?
> >   File "c:\programming\python\HTMLGen\HTMLtest.py", line 26, in test
> >     'overview.html')      # manual top
> >   File "c:\programming\python\HTMLGen\HTMLtest.py", line 35, in overview
> >     doc = SeriesDocument('HTMLgen.rc')
> >   File "c:\programming\python\HTMLGen\HTMLgen.py", line 256, in __init__
> >     if resource: execfile(mpath(resource), self.__dict__)
> > IOError: [Errno 2] No such file or directory
> > >>>
> >
> > ------------------------------
> >
> > Any help will be appreciated.
> >
> > Dave Borgelt
>
> This only works in the directory where HTMLgen.rc (and probably other
> files) lie. This is usually the directory generated when unpacking
> HTMLge.
>
> Greetings
>
> Berthold
>
> --
> bhoel at starship.python.net / http://starship.python.net/crew/bhoel/
>         It is unlawful to use this email address for unsolicited ads
>         (USC Title 47 Sec.227). I will assess a US$500 charge for
>         reviewing and deleting each unsolicited ad.
>
>





From xav at buf.fr  Fri Jul 30 11:48:30 1999
From: xav at buf.fr (Xavier Bec)
Date: Fri, 30 Jul 1999 15:48:30 +0000
Subject: syntax
Message-ID: <37A1C94E.F54181DB@buf.fr>

Hello,

Why are the double dots ":" necessary after a if, for or while
instruction ?
All the others instructions are terminated with a cariage return but not
condition or loop ones. I know that a semi-column is facultative and
only used if you have to put a second instruction on the same line. Why
is not the same for if, etc.

if i == 10
	print i
	print "test"

or

if i == 10 : print i; print "test"

Any suggestion ?

-- 
Xavier Bec                         mailto:xav at buf.fr
BUF Compagnie                      http://www.buf.fr
3, rue Roquepine             tel: (33) 1 42 68 18 28
75008 Paris - FRANCE         fax: (33) 1 42 68 18 29



From pduffin at mailserver.hursley.ibm.com  Wed Jul 21 06:23:11 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Wed, 21 Jul 1999 11:23:11 +0100
Subject: Tcl/Python: Who was first?
References: <932552685.620225@news.adtranzsig.de>
Message-ID: <37959F8F.55D8@mailserver.hursley.ibm.com>

Dirk-Ulrich Heise wrote:
> 
> I just saw a message called
> 
> README: Tcl-URL! - the weekly guide to Tcl resources
> 
> on comp.lang.tcl.announce. (Reading it for the first time
> in my life)
> Did Python steel this (good) idea from Tcl or was it the
> other way round?
> 

I think that Tcl copied it from Python.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From kjetilod at stud.ntnu.no  Fri Jul  2 12:10:20 1999
From: kjetilod at stud.ntnu.no (Kjetil Ødegaard)
Date: 02 Jul 1999 18:10:20 +0200
Subject: Socket programming.
Message-ID: <jevhfnnqbgz.fsf@apollo.orakel.ntnu.no>

I'm trying to implement a simple irc module as the base for a simple
irc client.  However, I'm having some design problems, having never
done any socket programming before.

My current implementation is threaded (using fork()); one thread for
user interaction and one for server output.  It uses blocking
sockets.  It works ok, but I've read some docs here and there (amongst
them the Socket Programming HOWTO on www.python.org) and I get the
feeling I should be using select() and non-blocking sockets.  Also,
I'm not sure I want to use fork() -- it seems very low-level and
depends on a Unix-like system.

Advice would be appreciated.  If you think I'm being unclear or vague
-- you're probably right. ;)  I'm not really sure what I'm asking,
either.

-- 
Gravity brings me down.



From dave at see.sig.for.address  Tue Jul 13 15:07:53 1999
From: dave at see.sig.for.address (Dave)
Date: Tue, 13 Jul 1999 19:07:53 GMT
Subject: making a Windows NT service out of a python program
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F20@RED-MSG-50> <378B2AA8.58477120@appliedbiometrics.com> <378B2C50.B7C2134A@cern.ch> <7mfot5$n30$1@nnrp1.deja.com>
Message-ID: <378b7cd2.1801654@read.news.global.net.uk>

On Tue, 13 Jul 1999 16:18:59 GMT, Andrew Cooke
<andrew at andrewcooke.free-online.co.uk> wrote:

>Services are described quite well in the help files that come with
>VC/Developer Studio, if you're lucky enough to have that ;-)
>

If you dont then the entire MSDN library is available online at:

 http://msdn.microsoft.com/resources/libraries.asp

It really needs a fast connection if you are not to go crazy waiting
for it to download, but it is free and they have removed the
requirement for you to register before using it.

	Dave Kirby
--------------------------------------------------
All great ideas start as heresy and end as dogma.

dkirby@   <-figure this out, spambots!
bigfoot.   My opinions are my own,
com        but I'm willing to share.



From neelk at brick.cswv.com  Fri Jul  9 19:16:59 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 9 Jul 1999 18:16:59 -0500
Subject: VIM and [Python] block jumping
References: <3781347D.CB2FBAB3@starvision.com> <37815B80.B4B51F94@starvision.com> <3785D753.DB2FF1ED@kami.com> <3785DA06.974567F3@kami.com>
Message-ID: <7m9tn5$6qg$47@rks1.urz.tu-dresden.de>

In article <3785DA06.974567F3 at kami.com>,
C. Laurence Gonsalves <clgonsal at kami.com> wrote:
>Of course, right after I posted that, realized that there's a problem.
>The blockMotion function doesn't account for lines that are
>continuations of other lines. That means that triple-quoted string
>literals, lines with more {[('s that }])'s, and lines joined with \ can
>screw it up. Are there any other situations? I'll try to post a
>corrected version later. Some of those problems are pretty nasty though
>(triple-quoted strings in particular).

Being a faithful Emacs-ite, I've never used vim. But since you can
script vim with Python, it might be possible to use the tokenizer in
the standard library (module tokenize) to handle this cleanly. Likely
you'd need to extend it a bit to handle the syntax errors that show up
in half-written code gracefully, but taking vim that one step past
syntax-awareness and into structure-awareness would be a very slick
hack.


Neel



From dalke at bioreason.com  Sun Jul 18 19:20:30 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Sun, 18 Jul 1999 17:20:30 -0600
Subject: I envy newbies
References: <000301bed08a$d63c68a0$27a22299@tim> <7mqns1$315@dfw-ixnews16.ix.netcom.com>
Message-ID: <3792613E.482E85C0@bioreason.com>

Warning: thread shift approaching :)

Aahz (@netcom.com) said:
> I'm suspicious; unless my memory is faulty, I *know* there were
> filesystems with a maximum of 64 root entires.  Maybe it was just floppy
> disks, though.

For nostalgia's sake, I keep a copy of "The Peter Norton
Programmer's Guide to the IBM PC" - from back when the AT was
hot stuff.

In the section describing file layout formats it says that each
directory block on a floppy can hold 16 entries, and single sided
disks have 4 blocks (double sided has 7), so there are 64 possible
entries for the directory block of a single-sided (8 or 9 sectors)
floppy.  Subdirectories are treated differently.

If you have a hard disk, the size changes depending on the size
of the disk.  The smallest example he gives has a single sector
for the FAT, which supports up to 64 root directories.

So these are the limit you might be recalling?

						Andrew
						dalke at acm.org



From ovidiu at cup.hp.com  Mon Jul 19 20:01:38 1999
From: ovidiu at cup.hp.com (Ovidiu Predescu)
Date: Tue, 20 Jul 1999 00:01:38 +0000
Subject: Python and Boehm-Demers GC, I have code.
References: <7mo2iu$iqg@ds2.acs.ucalgary.ca>
Message-ID: <3793BC62.1D88476A@cup.hp.com>

Neil Schemenauer wrote:

> I have added the Boehm-Demers garbage collector to Python and it
> works well!

Great work! However, this is not enough, there are more things to do
;-/!

More than a year ago I've added support for the Boehm's collector to
Objective-C, an object oriented language based on C, and two libraries
written in Objective-C that use reference counting as a memory tracking
policy.

What I found is that I needed to introduce weak references in these
libraries, otherwise I would have objects alive without being used
anymore by the code. The reason were two places in these libraries where
objects were kept in a global table. Having references to objects from
this table prevented the objects from being collected, even if you drop
the references to the objects from the rest of your code. The
alternative is to require the programmer to remove the references from
these global tables when he is no longer interested in keeping the
objects alive, but this is error-prone and defeats the whole purpose of
automatic garbage collection.

I don't know exactly if there are any modules in Python that keep track
of objects in global variables but if this is the case you may need to
implement something like that.

One of the libraries I'm referring to is libFoundation and you can find
it at:

http://www.geocities.com/SiliconValley/Monitor/7464/libFoundation/

The interesting piece you may want to look at is GarbageCollector.m. It
defines a way to register observers for the finalization of other
objects. The registering is done programmatically, by interested objects
(like for example the NSNotificationCenter, which allows interested
objects to be informed when a particular event is posted by another
object).

The support for weak pointers in the Objective-C runtime was done using
the typed-memory facility of the Boehm's collector. Take a look at a
description of how it works on:

http://www.geocities.com/SiliconValley/Monitor/7464/Objective-C/

If you're interested in the sources of the runtime, get a recent
snapshot of EGCS and look in the libobjc directory.


Have fun,
-- 
Ovidiu Predescu <ovidiu at cup.hp.com>
http://andromeda.cup.hp.com/  (inside HP's firewall only)
http://www.geocities.com/SiliconValley/Monitor/7464/



From Hank.Fay at p98.f112.n480.z2.fidonet.org  Sat Jul  3 15:28:34 1999
From: Hank.Fay at p98.f112.n480.z2.fidonet.org (Hank Fay)
Date: Sat, 03 Jul 1999 20:28:34 +0100
Subject: wddx wrapper?
Message-ID: <000016bc@bossar.com.pl>

From: "Hank Fay" <hank at prosysplus.com>

Understood.  I'd like a python wrapper so I can escape th evil empire when
necessary. <s>

--
http://www.prosysplus.com
<Voynow>; Sasha <anonymous at web.remarq.com> wrote in message
news:930863445.26638 at www.remarq.com...
>
> If you're in a Windows environment you can use Python's COM
> support to to use the COM object they offer.
>
> Sasha Voynow
>
> sasha at ofoto.com
>
>
>
> **** Posted from RemarQ - http://www.remarq.com - Discussions Start Here
(tm) ****




From tfb at tfeb.org  Thu Jul  1 16:35:33 1999
From: tfb at tfeb.org (Tim Bradshaw)
Date: 01 Jul 1999 21:35:33 +0100
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com>
	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com>
	<375FDCA7.62FF@mailserver.hursley.ibm.com>
	<7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com>
	<37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu>
	<37776FFA.2D85@mailserver.hursley.ibm.com>
	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it>
	<876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org>
	<377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov>
	<slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net>
	<wkzp1g2t2e.fsf@ifi.uio.no>
	<slrn7nnfs3.19k.wtanksle@dolphin.openprojects.net>
Message-ID: <ey3pv2ckt0q.fsf@lostwithiel.tfeb.org>

[Removed all but comp.lang.lisp]
* William Tanksley wrote:

> Forth is very much like Lisp, with the subtraction of memory management
> and the addition of implicit parameter passing.  Forth is Lisp for people
> who prefer "as simple as possible" to "but no simpler."  (And it IS a very
> good language; I use it more than Scheme.)

For what it's worth there are ongoing forth-lisp links in the form of
HP's language for their advanced calculators, RPL, which stands for
Reverse Polish Lisp, although it is really closer to Forth with memory
management.  It's actually a delightful language to use considering
the limitations of the device it runs on. It is a little frustrating
to those used to Lisp sometimes though.

(Of course, that doesn't stop people clamouring for a C++ compiler for
the hp48!).

--tim



From MHammond at skippinet.com.au  Wed Jul 21 18:30:10 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Thu, 22 Jul 1999 08:30:10 +1000
Subject: IDLE's PyShell
References: <000801bed335$8c233e80$a72d2399@tim> <3795F12D.91B820BB@inrialpes.fr>
Message-ID: <7n5hig$4iq$1@m2.c2.telstra-mm.net.au>

Vladimir Marangozov wrote in message <3795F12D.91B820BB at inrialpes.fr>...

>function call, I would have submitted a patch. But I don't know very well
>the shell logic, so I'll let somebody else do it for me.

As long as you arent holding your breath, you should be fine :-)  OTOH,
maybe this is a good way to learn the shell logic.  No one in the world
understood the PyShell logic when released except the author (Guido).  The
rest of us had to read the source like you do.

>I get a huge output, containing the entire PyShell.py source...
>With all due respect to PyShell.py, its source shouldn't be there, isn't
it?

Why does this bother you?  This is an implementation detail of IDLE.  Its
downside is the memory usage, but the upside is that IDLE works.

Unless this is causing you a specific problem, then why worry?

Mark.






From rdudfield at my-deja.com  Tue Jul 20 11:13:40 1999
From: rdudfield at my-deja.com (rdudfield at my-deja.com)
Date: Tue, 20 Jul 1999 15:13:40 GMT
Subject: Finding absolute pathname?
Message-ID: <7n23mm$ukr$1@nnrp1.deja.com>

How do I find the absolute pathname of the current directory?

eg /usr/doc, or c:\temp

os.curdir gives .

I need the absolute path name ( is that what you call it ? ).

Thanks in advance.

--Rene.

Please don't reply to the email address.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From jhspies at futurenet.co.za  Tue Jul  6 12:23:36 1999
From: jhspies at futurenet.co.za (Johann Spies)
Date: Tue, 6 Jul 1999 18:23:36 +0200 (SAST)
Subject: AttributeError - a dumb question
In-Reply-To: <Pine.LNX.4.04.9907061943010.21542-100000@emerald.netskate.ru>
Message-ID: <Pine.LNX.3.96.990706182058.780A-100000@Johann>

Hallo Oleg,

Thanks for your reply.  Your answer helped me to back to the original
problem I had:

class F:
    def __init__(self):

        self.x = self.Krylys()

    def KryLys(self):
        l = [1,2,3,4,5]
        return l
    

l = F()

Now this assignment fails:

Traceback (innermost last):
  File "<stdin>", line 11, in ?
  File "<stdin>", line 4, in __init__
AttributeError: Krylys

I hope you can help me with this one.

Johann

 --------------------------------------------------------------------------
| Johann Spies                                 Windsorlaan 19              |
| jhspies at futurenet.co.za                3201 Pietermaritzburg             |
| Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
 --------------------------------------------------------------------------

     "Ask, and it shall be given you; seek, and ye shall
      find; knock, and it shall be opened unto you."        
                                  Matthew 7:7 




From init__self at my-deja.com  Mon Jul 26 14:16:27 1999
From: init__self at my-deja.com (Python Addict)
Date: Mon, 26 Jul 1999 18:16:27 GMT
Subject: Python and SlickEdit
References: <379CA3F9.A746217E@iRenaissance.com>
Message-ID: <%B1n3.667$QE1.16157@dfiatx1-snr1.gtei.net>

The SlickEdit people have been less than forthcoming in building support for
Python (and aren't really intrerested in anything but supporting Java, Perl
or C/C++ it seems). I've been hounding them for about a year for decent
Python support and finally gave up.

Supposedly the latest 4.0 patch will deal with triple quotes and single
quotes correctly but I haven't gotten around to installing it -- I've been
using VIM. With 4.0 up to the 'b' revision, when you try to define a color
extension map, it'll accept """ but then lock up if you go back to edit it.
You can edit the config text files by hand to remove it but it's a royal
pain.

The request for colorization and syntax support was met with a "build your
own, it's easy!" and canned instructions. If it's that easy, SlickEdit
should bloody well do it.

Bollocks to them, I say. I'll save my $59 yearly renewal money and send half
to Guido and half to the VIM folks :-)

Rowland Smith wrote in message <379CA3F9.A746217E at iRenaissance.com>...
|Does anyone know if there are any Python plug-ins for Visual SlickEdit
|and where they are?
|
|--
|Regards,
|Rowland
|
|:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
::::::::
|iRenaissance, Inc.
| -- Tools For Building Online Communities
|
| Rowland D. Smith
| Senior Software Developer
| http://www.iRenaissance.com
| http://www.CalendarCentral.com
| 919.743.0109
|:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
::::::::





From olipt at mayo.edu  Sun Jul 11 23:00:32 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Sun, 11 Jul 1999 22:00:32 -0500
Subject: Example Code: Bresenham Line Algorithm
In-Reply-To: <7m9tln$6qg$4@rks1.urz.tu-dresden.de>
References: <7m9tln$6qg$4@rks1.urz.tu-dresden.de>
Message-ID: <Pine.LNX.4.10.9907112154520.15581-100000@us2.mayo.edu>

Phillip,

I'm not sure what the Bresenham Line Algorithm does.  If you could send me
a brief description, I could evaluate whether or not I'd like to include
it in the package that I'm developing (Multipack), which builds on the
NumPy array extension to give Python the numerical convenience (and
eventually the graphical convenience) of common interactive data analysis
packages like MATLAB.

I'm posting this to the newsgroup to remind other individuals who would
like to contribute to such an effort that they can send their code to me
(Oliphant.Travis at altavista.net).  For people who are really interested in
this there is even a CVS server where development takes place.

Thanks, 

Travis Oliphant










From Andrew.Cooke at p98.f112.n480.z2.fidonet.org  Thu Jul  1 08:54:04 1999
From: Andrew.Cooke at p98.f112.n480.z2.fidonet.org (Andrew Cooke)
Date: Thu, 01 Jul 1999 13:54:04 +0100
Subject: theads & global namespaces
Message-ID: <0000160c@bossar.com.pl>

From: Andrew Cooke <andrew at andrewcooke.free-online.co.uk>

In article <377A5C88.32A1FAB6 at notes.cba.ufl.edu>,
  andersdt at notes.cba.ufl.edu wrote:
> Greetings all,
>
> I am trying to share globals between threads.  Is is possible for me
to
> share globals between threads that call functions from different
> modules?  Thus far I have had little success.  I ask this question
> because  I would prefer not to put all my code in one module.

Hi,

I may be teaching my granny to suck eggs here, so ignore this if it's
obvious, but in general you should not spin a thread while waiting for
another thread to do something (this doesn't fix your global problem,
it's just advice about multi-threaded programming).

All my multi-threaded experience is with java, but the classes in
Python are similar, so I'll have a go at explaining how it "should" be
done and someone can point out my errors...

(Incidentally, whether or not your program works as it is - assuming
the global access thing is fixed - will depend on the platform you
runit out, and the threads implementation used.  Using conditions,
it will run on any platform).

from threading import *

globalX = 0

def test():
  c = Condition()
  c.acquire() # get hold of the condition
  t = Thread(target=XChanger, argcs=c)
  t.start()
  c.wait() # this halts this thread and releases c
  print globalX
  c.release()

class XChanger:
  def run(cond): # cond is c, passed in from the args in Thread
    cond.acquire() # waits for c to be released (in c.wait() above)
    globalX = globalX + 1
    cond.notifyAll() # signal the other thread to restart when...
    cond.release() # ...we release this

print globalX # should print 0
test() # should print 1 inside test()
print globalX # should print 1

(This is only one way of doing things - it may not even be best, but
it shows what I'm thinking of).  Doug Lea has written a very good
book about threads, but it discusses Java.  Because the Python threading
module is similar, you may find it useful (especially if you already
know Java).

There's also a completely different approach to using threads, which
is based on a (fairly ancient) book by Hoare called Communicating
Sequential Processes.  It is more elegant, but not as popular (it's
used in occam, and there is a Java library to support it, but nothing
for Python that I know of - ask me if you're interested and I can ask
on a mailing list I am subscribed to).

Cheers,
Andrew


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From MHammond at skippinet.com.au  Sun Jul 11 09:54:52 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Sun, 11 Jul 1999 23:54:52 +1000
Subject: Python and COM?
References: <7m9to7$6qg$82@rks1.urz.tu-dresden.de>
Message-ID: <7mb8u4$es3$1@m2.c2.telstra-mm.net.au>

Karl Putland wrote in message <7m9to7$6qg$82 at rks1.urz.tu-dresden.de>...
> For a COM server implemented in python, can a type library be generated?

No, but nearly :-)  Python has the framework and even some
"proof-of-concept" code, but isnt done.

Feel free to offer to help - not a huge amount of work left given what has
been done already - the pycom-dev mailing archives have the details...

Mark.






From fredrik at pythonware.com  Mon Jul 26 13:28:13 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 17:28:13 GMT
Subject: Tkinter bug ?
References: <379C8F64.E48BF26@lri.fr> <001301bed786$e88a7490$f29b12c2@secret.pythonware.com> <379C938A.27F0EF@lri.fr>
Message-ID: <003901bed78c$3dc05240$f29b12c2@secret.pythonware.com>

Stephane Conversy <conversy at lri.fr> wrote:
> ok, but does it have something to do with the fact that in one case
> I do a from file1 import * and in the other import file1 ?

now that you mention it...

forget my previous answer (for the moment; you
might stumble upon that problem later).

...

if you'd added some "print photo" statements to File2.py
in your first example, you would have noticed that the
variable is None both before and after the call to
create_photo().  if you'd added similar statements to
File1.py, you would have noticed that create_photo
actually did set it to something else.

the reason for this is that all variables in Python contain 
references to objects, not the objects themselves.  the
assignment statement just changes a variable to refer to
a new value, it doesn't copy the object itself (nor does
this modify the old value).

and import is just a fancy assignment statement; the
from-import form copies object references from the
imported module, not the objects itself.

in other words, after the import, there are TWO "photo"
variables, one in each module.  and changing one to refer
to a new object doesn't automatically change the other
one.

> I'm a beginner in Python, and don't know exaclty how modules
> are implemented...

but in that case, you shouldn't be using from-import at all !

for more details, see:
http://www.pythonware.com/people/fredrik/fyi/fyi06.htm

note the fourth item under "Which Way Should I Use?" ;-)

</F>




From holger at phoenix-edv.netzservice.de  Mon Jul  5 07:00:50 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Mon, 5 Jul 1999 11:00:50 GMT
Subject: Win32: silent script
Message-ID: <37809062.6D463928@phoenix-edv.netzservice.de>

High,

I'd like to use a little python-script to start win-programs
with special parameters. So I've made some links up to that 
python-script in windows-menu. 
But I'd like to let the programs start without the 
rising dos-window, that occurs for a few centiseconds while
python-script is working. Or even let it start with a minimized
window outside the real screen or something like that....

How could I realize that?
Is there a startup-flag I could set when calling python.exe?
Other possibility: How could I set the 'minimized window' option
in that link by python-code?

thanx,
Holger



From jsight at mindspring.com  Sat Jul 10 00:15:11 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Sat, 10 Jul 1999 00:15:11 -0400
Subject: Python re module
References: <7m5qev$lue$1@nnrp1.deja.com>
Message-ID: <7m9tnf$6qg$57@rks1.urz.tu-dresden.de>

catlee at my-deja.com wrote:
> 
> Try the following:
> 
> import re
> re.search("((.)\\1+)","a")
> 
> Now, I know this isn't proper syntax (you shouldn't reference a group
> inside itself), but on my machine python hangs.
> 
> Is this a bug, or a feature? =)

Yes.

-- 
---------------
Jesse D. Sightler
http://www3.pair.com/jsight/

"Do not use a hatchet to remove a fly from your friend's forehead." 
      - Chinese Proverb



From tony at lsl.co.uk  Thu Jul 15 07:43:03 1999
From: tony at lsl.co.uk (Tony J. Ibbs (Tibs))
Date: Thu, 15 Jul 1999 11:43:03 GMT
Subject: Meta-language for mxTextTools
Message-ID: <08bb01beceb7$334a58c0$f0c809c0@lslp7o.lsl.co.uk>

I am aiming to use Marc-Andr? Lemburg's mxTextTools package for an
application in my (hah!) spare time. Being me, the first thing I've set out
to do is to provide the missing meta-language so that I can avoid typing in
the tag tables as "bare" tuples.

What I've done so far is to produce an alpha version of the meta-language
documentation, which can be found at

    http://www.tibsnjoan.demon.co.uk/mxtext/metalang.html

Translations of the examples from the mxTextTools distribution are also
supplied.

I still have to actually *implement* the thing (I expect to do this with
mxTextTools itself, with appropriate hand-bootstrapping), but since I'm
about to go off on holiday for a couple of weeks, it seems worth
"publishing" my thoughts so far and getting any comments.

Note that I am *not* claiming that this is a particularly sophisticated
animal, nor that it is necessarily the best beast that could be dreamt of -
just that it is something I can work with and should be easy to implement.

--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.demon.co.uk/
Feet first with 5 wheels...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)





From fdrake at cnri.reston.va.us  Fri Jul 23 13:47:09 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 23 Jul 1999 13:47:09 -0400 (EDT)
Subject: Python documentation available as ZIP files
In-Reply-To: <1279371240-26729947@hypernet.com>
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
	<1279371240-26729947@hypernet.com>
Message-ID: <14232.43677.109055.503791@weyr.cnri.reston.va.us>

Gordon McMillan writes:
 > Nah, it's because a zip file is (conceptually) a tar of a bunch of 
 > gzips, while a tgz is a gzip of a tar. Compression ratios get better 
 > on longer inputs. But "random" access is a lot easier on the former.

Gordon,
  Good point; with the larger files (PDF & PostScript), the files
would be large enough that the difference is trivial, but with lots of 
small files, the compression tables have to be rebuilt for each file.
This makes a lot of sense.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From mwh21 at cam.ac.uk  Wed Jul 28 02:43:14 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Jul 1999 07:43:14 +0100
Subject: String2List: odd function
References: <199907272152.QAA01065@cro.ots.ac.cr>
Message-ID: <m390819svx.fsf@atrus.jesus.cam.ac.uk>

"Isidor" <rodisi01 at my-deja.com> writes:

> When I read these messages, the first thing I thought was: "wow! 
> *that's* what exec and eval do?!" I had read about them in the 
> documentation, and I suspected that they were important, but I 
> couldn't figure out what the heck they really did. A few unsuccessful 
> experiments didn't help much either. Well, now I know a little more. 
> However, I suspect that this use is not among the common uses for 
> which exec and eval were designed. Would you be willing to reply with 
> some examples of how you use these functions normally? What are they 
> *really* used for? (As I collect more of these examples from others 
> and from my own banging around, they will go on a web page w. 
> credits, of course, to the example providers...and then someday...I 
> will join the PSA...and then...and then...)

This article:

http://x39.deja.com/[ST_rn=ps]/getdoc.xp?AN=453014340

contains a moderately long example of using `exec' I posted to the
list a while back.

Have fun!
Michael



From wolford at enews.nrl.navy.mil  Thu Jul  8 12:13:02 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Thu, 08 Jul 1999 12:13:02 -0400
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com> <3784BD4B.7938B7BA@enews.nrl.navy.mil>
Message-ID: <3784CE0E.6BD46E97@enews.nrl.navy.mil>

Sorry, forgot a line when I typed it in. I've added it below:

Scott Wolford wrote:

> Here's the smallest case that gives me problems:
> #include "Python.h"
> #include <stdio.h>
> int main(int argc, char **argv)
> {
> //  return Py_Main(argc, argv);
>   FILE *file = fopen(argv[0], "rt");

  Py_Initialize();

>   PyRun_AnyFile(file, argv[0]);
> }




From arcege at shore.net  Fri Jul  9 21:40:18 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Sat, 10 Jul 1999 01:40:18 GMT
Subject: Printing returned object in embedded app
References: <378694FC.D857F68A@srl.caltech.edu>
Message-ID: <6wxh3.1472$6M6.450919@news.shore.net>

Conrad Steenberg <conrad at srl.caltech.edu> wrote:
: I'm trying to construct an interactive front-end for Python, using GTK.

: My C program passes strings to PyRun_String, and obtains the returned object.
: Now in the normal interactive Python on a terminal, a command like
: 1+1
: would print out the value, but using my method it does not. Using
: print(1+1) works like a charm. However, when I try to have the returned object
: printed using eg. 
: PyObject_CallMethod(object,"print","") 
: I get an error, "AttributeError: print". Can anybody tell me what I'm doing
: wrong? Or how to get the result printed automatically like in interactive
: Python?

Try:
  PyObject_Print(object, stdout, Py_PRINT_RAW);

The first argument is the object to be printed, the second is the FILE
* to print to and the last is a flag (the only flag for now is
Py_PRINT_RAW).  It only outputs the object, no leading or trailing
whitespace.

Also, you could run "print object" from PyRun_SimpleString.

  -Arcege




From zero at ecom.net  Tue Jul 27 12:09:11 1999
From: zero at ecom.net (Edward B. Wilson II)
Date: Tue, 27 Jul 1999 09:09:11 -0700
Subject: MS-SQL Server:Connectivity Help...
Message-ID: <379DD9A7.589CE943@ecom.net>

I would like to introduce Python with a bang to my new work group.  Does
anyone have a driver to "insert" a single row into an MS-SQL server7.0
db.  It appears from the main Python page, one must compile their own
ODBC drivers with VC++ to get the most robust solution.  Can anybody
comment on this shortcoming.  Getting Python to work for MS_SQL/NT would
be a big entrance.

Thanx--

Ed





From greg.ewing at compaq.com  Tue Jul 20 17:37:27 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 21 Jul 1999 09:37:27 +1200
Subject: Tk: Temporarily change cursor?
References: <3793AADF.5EDBAFBC@compaq.com> <19990720145034.04299@nms.otc.telstra.com.au>
Message-ID: <3794EC17.64249239@compaq.com>

Greg McFarlane wrote:
> 
>     Blt: http://www.tcltk.com/blt/

I've just looked at the web page for Blt, and it says
this about the "busy" command:

  Manages an invisible "busy" window which prevents further
  user device (keyboard, mouse, button, etc.) interactions.

I don't like the sound of that. I don't want to lock
up the whole machine and prevent interaction with
windows belonging to other processes.

In an old news article I was referred to,
Fredrik Lundh <fredrik at pythonware.com> wrote:

> if you change the cursor for a toplevel (or the root window),
> that cursor will be used for all widgets that doesn't specify their own
> cursors. 

I tried things like that, but they didn't work.
Changing the root window's cursor doesn't affect
any other toplevels, even if they don't define their
own cursor. Also, when the pointer was over a subwidget
of a toplevel, changing the toplevel's cursor and then
doing an update didn't work reliably if the mouse
buton was down at the time!

I've since found a method that seems to work on Windows:
Use winfo_pointerxy and winfo_containing to find the
widget under the pointer, change its cursor, and
update. It'll only work if the pointer happens to be
over one of the application's windows at the time, but
if the user has just clicked on something, this will
probably be true. It'll do for now.

Thanks for the help,
Greg



From arcege at shore.net  Thu Jul  1 16:13:59 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 01 Jul 1999 20:13:59 GMT
Subject: Python GUIs: Summary and Conclusion
References: <99FA4A1A1280BEBD.0A1ABA62CC989479.4C2C6022F695BEAC@lp.airnews.net> <19990629092757.A4998@optiplex.palga.uucp> <46F9FB3A1847A014.2CBD270176138613.20F84A39E6ED95DA@lp.airnews.net>
Message-ID: <b_Pe3.558$6M6.167478@news.shore.net>

Mikael Lyngvig <mikael at pobox.com> wrote:
: On Tue, 29 Jun 1999 09:27:57 +0200, Gerrit Holl
: <gerrit.holl at pobox.com> wrote:

:>I think one could also make an application "multiple-interface". The user can
:>choose: svgalib (any support in python, by the way?), curses, commandline,
:>tk, qt/kde or some of the other stuff. The developer can put the interface
:>in a seperate class or even in a seperate module (I'll do with my discoverb
:>project (see signature)), so the user can exactly download what he want. I
:>believe KDE users would like KDE applications...

: I think that would be a good idea, but I don't think it is directly
: related to the issue at hand.  Such an API would certainly be slower
: than the actual implentation and the problem is, IMHO, that the actual
: implementation (of Tk) is too slow to be really interesting.

: As a matter of fact, I've always wished that I could make my apps so
: they took an option, say -ui:name, what would allow the user to
: specify which kind of user interface it had:

:    bui - batch user interface; Unix style command-line interface.
:    cui - console user interface; Unix curses interface.
:    gui - graphical user interface; X/Windows style interface.

: That way the user could use the application in any way he/she like,
: without having to resort to Win32 COM (automation server) APIs etc.

Been there, done that.  My pyirc client (alpha release) was designed
for just about any user-interface, and comes with three: dummy (no
input or output, just batch-mode)), curses and Tkinter.  You just can't
switch between them at runtime, ;)  I've done it for other apps too.

  -Arcege




From kernr at mail.ncifcrf.gov  Sat Jul 10 16:50:12 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Sat, 10 Jul 1999 20:50:12 GMT
Subject: Compiling Win32 Python Extensions with EGCS
Message-ID: <3787b19a.20772719@news.erols.com>

I have established a set of instructions and notes for compiling
Python extensions with the free EGCS compiler suite with the mingw32
target.  EGCS is the "experimental" development branch of the GCC
compiler.  The next version will be the official GCC-2.95.  Mumit Khan
has ported EGCS 1.1.2 (the current release) to Win32 systems.  See

    http://wwww.xraylith.wisc.edu/~khan/software/gnu-win32

for more information on this port.

If you just want the instructions and the necessary tools, go to my
web page:

    http://starship.python.net/crew/kernr/mingw32/Notes.html

Many thanks to Paul Sokolovsky for his groundbreaking work on porting
the Python headers to be used by EGCS/mingw32.

NOTE: The mingw32 target does not use Cygnus Solution's Cygwin
portability layer.  Programs compiled for the mingw32 target use the
system's C runtime DLL (for Python extensions, this is MSVCRT.DLL) and
not the Cygwin DLL.  Python extensions compiled as I describe do *not*
automatically fall under the GPL.

I, Robert Kern, may be reached at <kernr at ncifcrf.gov>.

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From gmcm at hypernet.com  Sat Jul 24 18:00:28 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 24 Jul 1999 17:00:28 -0500
Subject: Fork, Threading, and Select madness.  Please help!
In-Reply-To: <379A25CF.6935D6A9@mindspring.com>
Message-ID: <1279272671-32658737@hypernet.com>

Jesse D. Sightler wrote:

> Thanks for your thoughtful reply.  As it turned out, threading
> wasn't nearly as bad a solution as I thought it might be<0.5wink>. 
> I just downloaded Python 1.52 and recompiled with "./configure
> --with-threads" and everything worked out nicely on the FreeBSD box
> that my webserver uses.  

You're welcome. Wish all my advice came out that well!
 
> Of course, doing this on my linux box resulted in something that
> causes my kernel to blow up in flames a few hours after running the
> Python interpretter with any threaded program.  But that's ok, at
> least it works with my ISP.

If you compiled with gcc, you might try egcs. A recent thread on 
Linux threading problems arrived at that fix, but I'm not sure it's 
the same problem.

I'd never encountered the problem, but rebuilt with egcs anyway.


- Gordon



From gmcm at hypernet.com  Fri Jul  2 16:22:44 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 2 Jul 1999 15:22:44 -0500
Subject: Debugging embedded pgm under WinNT
In-Reply-To: <7lj12e$jct$1@nnrp1.deja.com>
Message-ID: <1281178906-45167517@hypernet.com>

Kevin Rodgers wrote:

> I have Python 1.5.2 embedded in an application on WinNT (4.0, SP4)
> using Visual C++ 6.0 SP2.  When I compile the project in Debug mode,
> even though it links correctly to the python15_d.lib file, it won't
> run because it complains about "can't find python15_d.dll".  So I
> got the source, compiled the python15_d.dll, put it in
> WINNT/system32 right next to python15.dll, ran my program . . . and
> now it couldn't find "multiarray" (my embedded Python code uses the
> Numeric extensions).  I verified that sys.path was correct and
> included the Numeric directories. Do I have to build the Numeric
> extensions under Debug mode? 

Yup. A pain in the butt, ain't it? 

Everything has to be running the same c runtime lib. However, you can 
compile python15.dll and your stuff in debug mode but using the 
release c runtime (Multithreaded DLL). This at least lets you debug 
your stuff and step into python15.dll. Other stuff will load, but 
won't have any symbols.  As I recall I had to tweak quite a few 
things in the build to get that to work, too.

- Gordon



From arcege at shore.net  Wed Jul  7 11:45:08 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Wed, 07 Jul 1999 15:45:08 GMT
Subject: Multiple program defined types -- access from C?
References: <1280771407-69691738@hypernet.com>
Message-ID: <8CKg3.1093$6M6.349834@news.shore.net>

Gordon McMillan <gmcm at hypernet.com> wrote:
: Nathan Froyd wrote:
:> I'm writing a program that needs a lot of program-defined types that
:> are also exposed as python classes to the user. The problem is, I
:> also need to access the innards of a lot of these classes from C.
:> Are there any elegant ways to do this? It doesn't seem as though
:> Python likes for class functions to be used from outside the C file
:> they're defined in.

: Python asks you to declare almost everything as file static to reduce 
: the chance of name clashes when loaded dynamically in Unix (or maybe 
: I should say non-Windows, where this is not a problem).

: The "proper" way to make a C API available to other modules is to use 
: a CObject (see Extending / Embedding section 1.12) to pass a void * 
: around. If you're not trying to provide a C API to other extensions, 
: you can probably get away with losing the "static", as long as you're 
: careful about how you name your functions.

I'll just amend to this.  Nathan mentioned that his data types are also
Python data types.  To access the data from other modules, it may be
more appropriate to access the C data through the Python API
(PyObject_GetItem and/or PyObject_GetAttrString) than to always use
PyCObject objects.  This lets handles the conversion internally.

The PyCObject type is very useful, mainly (IMO) because you can
associate a string with it to help identify the contained data
(something that is difficult in other P* languages ;).

One thing to NOT do is use the C functions from one Python module in
another module.  Sharing C functions within different C files built into
one Python C module is fine, but nasty things can occur if used outside
of that module (in a UNIX dynamic library context anyway :).

  -Arcege




From greg.ewing at compaq.com  Wed Jul 28 16:55:19 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 29 Jul 1999 08:55:19 +1200
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <slrn7psisn.da.wtanksle@dolphin.openprojects.net> <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com>
Message-ID: <379F6E37.777322BF@compaq.com>

Fredrik Lundh wrote:
> 
> what else did you expect from Guido?

I would have expected Guido to shriek in horror
and refuse to even consider such a semantically
ambiguous construct.

It seems quite out of character for both him
and Python.

Greg



From fredrik at pythonware.com  Fri Jul  9 05:59:57 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 9 Jul 1999 09:59:57 GMT
Subject: why is there no put in urllib ?
References: <Pine.LNX.4.05.9907082250040.11616-100000@ns.ac-copy.net>
Message-ID: <002801bec9f1$ce33b8b0$f29b12c2@secret.pythonware.com>

Joachim Schmitz <perry at ac-copy.net> wrote:
> may be this is a total dump question. But it would allow us Zopistas to
> upload our dtml pages with a simple python script to the Zope-server.

so what are you waiting for?  you have the
sources, don't you?  ;-)

(also see the XML-RPC firewall thread)

on the other hand, httplib supports PUT just
fine.  as long as there's no firewall in your way,
it's not much harder to use than urllib.

</F>




From achrist at easystreet.com  Thu Jul  1 15:48:48 1999
From: achrist at easystreet.com (Al Christians)
Date: Thu, 01 Jul 1999 12:48:48 -0700
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net> <377B711A.E8B65E0B@cfdrc.com>
Message-ID: <377BC620.F5553FF5@easystreet.com>

I looked at Fox and put it aside because I need a grid control.
wxPython seems to be the winner for Windows in that category, even
though the grid is pretty minimal.

Any other candidates with a grid?

Al




Lyle Johnson wrote:
> 
> I've already worked up the binary distribution of FXPy for Windows; I just
> forgot to mention it in the previous post. The binary installer can be found
> at:
> 
>     ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42-win32.exe
> 
> I haven't worked up a binary distribution of the (C++) FOX distribution
> itself, primarily because it changes most every day :) But that is in the
> plans as well.
> 
> Thanks,
> 
> Lyle
> 
> Mikael Lyngvig wrote:
> 
> > On Wed, 30 Jun 1999 09:07:51 -0500, Lyle Johnson <jlj at cfdrc.com>
> > wrote:
> >
> > >Just to add fuel to the fire, we now have a Python interface to FOX, a
> > >cross-platform GUI toolkit:
> > >
> > >    http://www.cfdrc.com/pub/FOX/fox.html
> > >
> > >The latest version of the Python wrappers can be picked up at:
> > >
> > >    ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42.tar.gz
> > >
> > >I don't know how the speed of FXPy would compare with some of the others
> > >mentioned (Tk, wxWindows, Qt, GTK, ...) but it would be worthwhile for
> > >some independent opinions.
> >
> > I'm downloading it now.  The http URL is wrong, it should be
> > http://www.cfdrc.com/FOX/fox.html.
> >
> > The presentation seems promising, but I'd like to suggest a binary
> > distribution for PC/Intel computers.  I know this is not common for
> > Unix distributions, but it's the norm for PCs.
> >
> > -- Mikael



From fredrik at pythonware.com  Fri Jul 30 09:19:17 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 30 Jul 1999 13:19:17 GMT
Subject: os.name
References: <1278785134-25350540@hypernet.com> <V5ho3.4340$VL2.463886@news.direcpc.com>
Message-ID: <001e01beda8e$2111d2e0$f29b12c2@secret.pythonware.com>

Emile van Sebille <emile at fenx.com> wrote:
> >>> os.name
> 'nt'
> >>> os.environ('OS')
> Traceback (innermost last):
>   File "<pyshell#8>", line 1, in ?
>     os.environ('OS')
> AttributeError: no __call__ method defined

Gordon asked you to try os.environ['OS'],
not os.environ('OS') ;-)

</F>




From tim_one at email.msn.com  Sat Jul 10 17:09:32 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 10 Jul 1999 17:09:32 -0400
Subject: VIM and [Python] block jumping
In-Reply-To: <7m5vtb$2b7$1@brick.cswv.com>
Message-ID: <000001becb18$821aca40$e69e2299@tim>

[C. Laurence Gonsalves]
> ...
> The blockMotion function doesn't account for lines that are
> continuations of other lines. That means that triple-quoted string
> literals, lines with more {[('s that }])'s, and lines joined with \ can
> screw it up. Are there any other situations?

No, those are the only ways a line can get continued in Python.  Note that
there are subtleties, like that a backslash preceding a newline doesn't
*always* mean continuation (it doesn't at the tail end of a comment!).

> I'll try to post a corrected version later. Some of those problems are
> pretty nasty though (triple-quoted strings in particular).

Also single-quoted strings continued via backslash.  Strings are the real
headache, because the special characters you're looking for lose their magic
when they're inside a string.

[Neel Krishnaswami]
> ...
> But since you can script vim with Python, it might be possible to use
> the tokenizer in the standard library (module tokenize) to handle this
> cleanly.  Likely you'd need to extend it a bit to handle the syntax
> errors that show up in half-written code gracefully ...

This can be made to work, but I found it too slow to bear in an editor --
"subjectively instantaneous" is the goal there.  Most of the tokens tokenize
delivers are simply irrelevant to this task, so most of the work that goes
into producing them is wasted.

The current CVS version of IDLE has a new module PyParse.py, which does the
minimum necessary to determine whether a line is part of a continuation, and
if so why.  It can do this 100x faster than tokenize, even in
non-pathological programs.  The next version of IDLE uses this to do Emacs
pymode-like "very smart" indentation, and the next version of Mark Hammond's
PythonWin also uses this module.

It's GUI-independent (straight self-contained Python), but as called from
AutoIndent.py (also shared between IDLE and PythonWin) takes major advantage
of the host system's colorization, to get a fast answer to the question "is
this random character in a string?".  The colorizer generally knows this
already, for every character in the program, and AutoIdent/PyParse use that
to find a good (close) place to begin reparsing the file.

If the host can't answer that question reliably, then PyParse needs to
reparse the entire file from its start in order to get a guaranteed-correct
answer, so that's what it does.  PyParse is fast enough that this is *still*
"subjectively instantaneous" for most files, but becomes a noticeable delay
near the end of long files (like e.g. Tkinter.py).

pyparse-is-incomprehensible-but-that's-why-it's-fast<wink>-ly y'rs  - tim





From tgm at math.uni-bonn.de  Fri Jul  9 08:48:00 1999
From: tgm at math.uni-bonn.de (Thomas Ackermann)
Date: Fri, 9 Jul 1999 12:48:00 GMT
Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
In-Reply-To: <19990708223856.A29441@optiplex.palga.uucp>; from Gerrit Holl on Thu, Jul 08, 1999 at 10:38:57PM +0200
References: <tot2m7.su.ln@final.math.uni-bonn.de> <19990708223856.A29441@optiplex.palga.uucp>
Message-ID: <7m9tm4$6qg$18@rks1.urz.tu-dresden.de>

On Thu, Jul 08, 1999 at 10:38:57PM +0200, Gerrit Holl wrote:
> On Thu, Jul 08, 1999 at 09:22:05PM +0200, Thomas Ackermann wrote:
> > I have not much time to experiment with those modules and would like
> > to hear opinions about to use what module     ;-)
> 
> Well,
> I used HTMLgen and I don't really understand why to use it. If you don't
> know HTML, you can't use it: you need to understand tags like <B>, <I>, <U>
> and such. I strongly recommend to have a look at HTML: if you look at it only
> one hour, you'll already understand the basics. Another hour and you understand
> frames and tables, and another hour to understand how it interacts with CGI.
> So you'll only need 3 hours, if not less, to learn web programming. When you're
> able to do that, you'll understand much more about HTMLgen (but won't need it,
> I think...)

Thanx - But i know HTML (and Java) itself already	;-)
What i do not know are the Python modules for it and wanted to get some
information from people who used several of the alternatives already.
But thanx anyway ...


> > So, what to use to allow for an Internet-Shop with pages created from
> > databases and allowing customers to purchase over net?
> > 
> 
> If you're doing such serious things, you should 
> 
>                                      **really**
> 
> learn web programming!!

Again, i want to use Python and a decent HTML/CGI module, i never said to
be completely unexperienced with WWW stuff ...

	Byebye,
-- 
Thomas Ackermann | Tel. +49-(0)228/631369|73-7773 | <tgm at math.uni-bonn.de>
             finger tgm at rhein.math.uni-bonn.de for public key
               GNU LINUX Python gtk pygtk MySQL FUDGE GURPS



From mstenber at cc.Helsinki.FI  Wed Jul  7 11:20:18 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 07 Jul 1999 18:20:18 +0300
Subject: Emacs: highlighting "self" tweak
References: <7luqpa$3s9$1@nnrp1.deja.com>
Message-ID: <al83dz032rx.fsf@myntti.helsinki.fi>

Preston Landers <mithy at my-deja.com> writes:
> I wish I knew of an easy way to make it a hook instead of a patch.  If
> someone can figure that out, I would *love* to hear from you.

.. from my own python-mode-hook (snippage) ..
  (setq python-font-lock-keywords
    ;; also color "self"as a 'type'
    (cons '("\\b\\(self\\)[ .,]?"
	    1 font-lock-type-face)
	  python-font-lock-keywords)
    )
.. from my own python-mode-hook (snippage) ..

-- 
UNIX sysadmin's day:
  gawk; talk; date; wine; grep; touch; unzip; touch; gasp; finger; gasp;
  mount; fsck; more; yes; gasp; umount; make clean; make mrproper; sleep



From tbryan at arlut.utexas.edu  Wed Jul 28 10:25:01 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Wed, 28 Jul 1999 09:25:01 -0500
Subject: Python and Red Hat Linux 6
References: <379CD547.22A10A50@postoffice.pacbell.net> <379CD5C2.5BA096A2@mail.danbbs.dk> <379d7fcf.52695@news.tiscalinet.it> <379D84E0.5F872B@arlut.utexas.edu> <Pine.LNX.4.10.9907271220420.19094-100000@atrus.jesus.cam.ac.uk> <379F122F.D89B8FD1@arlut.utexas.edu>
Message-ID: <379F12BD.68EEB65F@arlut.utexas.edu>

Thomas Bryan wrote:
> On my Red Hat 6.0 system, problems #1 and #2 report an identical
> error message.  I lost track of the thread, but I meant to suggest
> that if one has already checked #2, then he should also check #2.

ummm...oops....that should be "then he should also check #1"

> the-gremlins-have-been-modifying-my-PATH-ly yours

the-gremlins-have-been-modifying-my-news-posts-ly yours

-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From arcege at shore.net  Sun Jul  4 17:05:37 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Sun, 04 Jul 1999 21:05:37 GMT
Subject: Is it just me, or..
References: <al8btds8ehb.fsf@myntti.helsinki.fi>
Message-ID: <B0Qf3.810$6M6.258150@news.shore.net>

Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
: Is there a bug in the handling of pipes with the Python? Esp. in
: non-blocking way.

: Example:
:  (r,w)=os.pipe()
:  r = os.fdopen(r, "r", 0)
:  w = os.fdopen(w, "w", 16384)
:  w.write('foobarbaz')
:  select.select([r],[],[]) # returns r, as it should
:  r.read(1) # returns 'f'
:  select.select([r],[],[]) # blocks!

: Now select _blocks_, despite there being obviously some data that at least
: should not have been buffered elsewhere. Also, fdopened stuff's reading
: features are bit annoying, as doing read(bigNumber) results in blocking
: call, as opposed to recv(bigNumber)'s semantics. Thus instead of pipes
: (=fast) I am using TCP/UNIX sockets (=slow) but they seem to work better
: for what I need. Am I missing something really obvious?

: This is on Linux, if it matters.

At first glance, I'd say that it is the buffering in the fdopen(w, "w",
16384) call.  Put the bufsize to 0 or call the flush method.

  -Arcege




From salzr at certco.com  Mon Jul 26 14:27:37 1999
From: salzr at certco.com (Rich Salz)
Date: Mon, 26 Jul 1999 14:27:37 -0400
Subject: Embedding python/multi-thread env
References: <379377AF.E8E7351D@cannontech.com> <37941e27.85316128@news.u-psud.fr>
Message-ID: <933013656.237.57@news.remarQ.com>

Laurent POINTAL wrote in message <37941e27.85316128 at news.u-psud.fr>...
>Depending what you really need. You must know that there is a global
>lock between the different Python interpreters.

I believe that you cannot create multiple interepreters in a single
executable.
The python global lock is to prevent *threads* within the interpreter from
stomping on each other's critical sections.

Anyhow..  we have a similar need.  Since I am going to create an
interpreter, create threads for each worker, and have them jump into C code
to block waiting for the C++ code to get a client.  You might have to do
some tricky synchronization between the two "thread pools" but it's totally
workable.






From tim_one at email.msn.com  Mon Jul 26 00:58:28 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 26 Jul 1999 00:58:28 -0400
Subject: Maintainability (was Re: Stackless & String-processing)
In-Reply-To: <14232.32203.792051.822513@weyr.cnri.reston.va.us>
Message-ID: <000501bed723$7fee3f60$492d2399@tim>

[Fred L. Drake, "import whatever as something_else"]
> ...
>   As I said, I haven't noticed this come up for a while, so maybe I'm
> the only one, but I think this would encourage clarity and
> "localizing" the names of imported modules (adding a leading _ so when
> others "import *" from our modules their namespace isn't polluted too
> much).

New keywords are dead in the water, so maybe you'll find the attached
useful.  Make it fancier, e.g. _my_import(globals(),
something_else="whatever") if you like.  A version of this used to do that,
but with the functional interface it was too confusing for me.

go-ahead-import-my-day-ly y'rs  - tim

def _pvt_import(globs, modname, *items):
    """globs, modname, *items -> import into globs with leading "_".

    If *items is empty, set globs["_" + modname] to module modname.
    If *items is not empty, import each item similarly but don't
    import the module into globs.
    Leave names that already begin with an underscore as-is.

    # import math as _math
    >>> _pvt_import(globals(), "math")
    >>> round(_math.pi, 0)
    3.0

    # import math.sin as _sin and math.floor as _floor
    >>> _pvt_import(globals(), "math", "sin", "floor")
    >>> _floor(3.14)
    3.0
    """

    mod = __import__(modname, globals())
    if items:
        for name in items:
            xname = name
            if xname[0] != "_":
                xname = "_" + xname
            globs[xname] = getattr(mod, name)
    else:
        xname = modname
        if xname[0] != "_":
            xname = "_" + xname
        globs[xname] = mod





From MHammond at skippinet.com.au  Fri Jul  9 08:40:47 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 9 Jul 1999 22:40:47 +1000
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com> <37850F5B.F80CB0A7@enews.nrl.navy.mil> <7m3ajc$s3q$1@m2.c2.telstra-mm.net.au> <37859ef4.11675718@news.ricochet.net>
Message-ID: <7m4qgt$e17$1@m2.c2.telstra-mm.net.au>

phil wrote in message <37859ef4.11675718 at news.ricochet.net>...
>On Fri, 9 Jul 1999 09:03:03 +1000, "Mark Hammond"
><MHammond at skippinet.com.au> wrote:
>
>available but not the _d.dl. Couldn't there be two downloads, one with
>the optimised/non-debug and one without? I rather like the

There could of course - just a matter of "who"?  Guido does a fanstasic job
of putting together the Python binary set for Windows, even though Windows
is not his primary platform.

It is just asking too much for the existing people to do this for nothing.
This is the reason I am adding it as a registered user benefit - it is
simply too much of a burden to go _this_ far for free!

Mark.






From Michael.Hudson at p98.f112.n480.z2.fidonet.org  Thu Jul  1 21:17:47 1999
From: Michael.Hudson at p98.f112.n480.z2.fidonet.org (Michael Hudson)
Date: Fri, 02 Jul 1999 02:17:47 +0100
Subject: print color strings?
Message-ID: <00001645@bossar.com.pl>

From: Michael Hudson <mwh21 at cam.ac.uk>

Gerrit Holl <gerrit.holl at pobox.com> writes:

> Hello,
> 
> is there an easy way to print color strings in python?
> 
> A function like printcolor('brightyellow', 'red', 'spam'), that finds
> out on what kind of terminal you are and prints it to stdout?

That sort of finding out is what terminfo is for; there's no Python
interface that I know of (such a beast shouldn't be too hard to write,
but a bit pointless, I suspect). However there is the `tput' program;

os.system("tput setaf 1")

should turn your foreground colour red. 

Read the terminfo man pages for more codes.

You'd probably want to cache the escape codes rather than spinning a
new process each time you need one.

codes['redf'] = os.popen("tput setaf 1").read()
codes['redb'] = os.popen("tput setab 1").read()
codes['greenf']=os.popen("tput setaf 2").read()
...

and so on.

I'm certainly not an expert on such matters; reading man pages for
tput, terminfo and term and playing around will get you exactly where
I've got to.

All terribly unix-centric, of course.

HTH
Michael



From aahz at netcom.com  Wed Jul 28 19:28:56 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 28 Jul 1999 23:28:56 GMT
Subject: Newbie Problem
References: <7nndss$q1n$1@alexander.INS.CWRU.Edu>
Message-ID: <7no3no$6nd@dfw-ixnews10.ix.netcom.com>

In article <7nndss$q1n$1 at alexander.INS.CWRU.Edu>,  <rbl at hal.cwru.edu> wrote:
>
>Python 1.5.1 (#37, Apr 27 1998, 13:36:04)  [CW PPC w/GUSI w/MSL]
>Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>> import sys
>>>> import random
>>>> import string
>>>> 
>>>> def run(filename=''):
>... if filename=='':
>  File "<stdin>", line 2
>    if filename=='':
>     ^
>SyntaxError: invalid syntax

Python uses indentation for control structures; "def" is a control
structure.  Therefore, the "if" needs to be indented.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From gmcm at hypernet.com  Thu Jul  8 17:59:45 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 8 Jul 1999 16:59:45 -0500
Subject: "Wrapping" Functions in classes
In-Reply-To: <378508BC.4ECD2149@cisco.com>
Message-ID: <1280654972-76699269@hypernet.com>

Aaron Rhodes wrote:

> How do I wrap the methods in a class so that before
> they are called, another function gets called first.
> Is there a special __variable__ that I can use?
> This other function should be called for each of the
> methods in the class.

If you have the source distribution, look at Demo/metaclasses, 
specifically Trace.py.

You don't really need to use a metaclass though. Create a Proxy class 
which wraps your instances. Override __getattr__ to look up the 
attribute in the wrapped instance. If the type of the attribute is a 
types.Functiontype, wrap it in something like the TracingWrapper 
from the above mentioned Trace.py. Then the wrapped function's 
__call__ method is where you put your before / after code (and the 
apply, of course).

- Gordon



From tgm at math.uni-bonn.de  Thu Jul  8 15:22:05 1999
From: tgm at math.uni-bonn.de (Thomas Ackermann)
Date: Thu, 8 Jul 1999 21:22:05 +0200
Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
Message-ID: <tot2m7.su.ln@final.math.uni-bonn.de>

Hi!

I need to generate some hompages and cgi-script for database access,
forms and such and want to do it in Python, of course. As it is i have
no experience with it's WWW-related stuff ...

There are htmllib and cgi from the library, HTMLfmt and genCGI from
the "Internet Programming with Python" book, HTMLgen itself and
something just announced called "html", which seems to be a variant of
HTMLgen(?!)

I have not much time to experiment with those modules and would like
to hear opinions about to use what module     ;-)

Also, i am not exactly sure what Zope is - it seems to be something to
create WWW-Pages and allow access to databases and it has it's own
WebServer (a variant of Medusa) ...

So, what to use to allow for an Internet-Shop with pages created from
databases and allowing customers to purchase over net?

Can do Zope such things? Is it overkill at first? What can it do
beyond? (And yes, i read some documentation, but i do not even
understand what an "Application Server" is    ;-)

	  Thanx in advance and
	  Byebye,
-- 
  Thomas Ackermann | Tel. +49-(0)228/631369|73-7773 | <tgm at math.uni-bonn.de>
             finger tgm at rhein.math.uni-bonn.de for public key
               GNU LINUX Python gtk pygtk MySQL FUDGE GURPS



From gmcm at hypernet.com  Thu Jul 29 07:57:15 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 29 Jul 1999 06:57:15 -0500
Subject: any way to know when the program is exiting?
In-Reply-To: <199907290135.SAA22743@hpcll563.cup.hp.com>
Message-ID: <1278876857-19833523@hypernet.com>

Ovidiu Predescu wrote:
> 
> I was wondering if there is any way in a __del__ method to know when
> it's invoked on the way out, as a result of the program going away?
> 
> I have a __del__ method of a class that sends a message to a global
> variable. It happens that on the way out, that value in that
> variable is destroyed before the __del__ methods of my instances are
> invoked. The result is that I have a bogus value that doesn't
> respond to the usual methods anymore. How can I prevent sending
> messages to that object?

You rely on just what you described. Globals are nuked first, but 
they don't end up with random bogus values, they are set to the one 
and only None object.

  def __del__(self):
    if globalvar is None:
      # we're on the way out...

- Gordon



From bobalex at ix.netcom.com  Thu Jul  8 12:46:43 1999
From: bobalex at ix.netcom.com (Bob Alexander)
Date: Thu, 8 Jul 1999 09:46:43 -0700
Subject: Any *disadvantage* to using xrange() instead of range()?
Message-ID: <004d01bec961$77024240$a0ee6dd1@BobAlexDesk>

...even for iteration over a small sequence of integers?

My intuition would be that xrange() is the preferred way to iterate over a
numeric sequence, but the documentation seems to go out of its way to stress
that the advantage of xrange() over range() is minimal -- almost seems to be
discouraging us programmers from using it unless we have a "memory-starved"
machine.

Is there any reason why xrange() should not be the "idiomatic" way to
iterate over a sequence of integers?

-- Bob




From tim_one at email.msn.com  Sat Jul 10 17:26:31 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 10 Jul 1999 17:26:31 -0400
Subject: Longs
In-Reply-To: <378778A7.C6BAF5C3@callware.com>
Message-ID: <000601becb1a$e1b9f000$e69e2299@tim>

[Ivan Van Laningham]
> When were longs added to Python?

Before the Beginning, I think.  They were already in the first version I
saw, long before Python 1.0.  At first they followed a 1's-complement model,
and changing to 2's-comp (to match bounded ints) is the only significant
change I recall.

> Whose idea was it?

guido-created-the-integers-all-the-rest-is-the-folly-of-man-ly y'rs  - tim





From gmcm at hypernet.com  Fri Jul  2 09:44:55 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 2 Jul 1999 13:44:55 GMT
Subject: Mailbox, tell and read
In-Reply-To: <00d601bec461$b24c01b0$f29b12c2@secret.pythonware.com>
References: <00d601bec461$b24c01b0$f29b12c2@secret.pythonware.com>
Message-ID: <1281202769-43731732@hypernet.com>

Fredrik Lundh wrote:
> Gordon McMillan wrote:
> > > First, are there platforms other than DOS/Windows/Win32 where there
> > > is a difference between text and binary mode?
> > 
> > There's a theoretical difference on *nix - otherwise, they wouldn't 
> > have made up the distinction!
> 
> who are "they?"

Whatever AT&T employee it was. AFAIK, it predates any glimmer of an 
idea of a "PC".

> fwiw, the ANSI C rationale mentions Unix as the
> special case here... go figure ;-)

I have a vague recollection that it can make a difference on AIX. At 
least I remember fixing something by changing someone's "r" to "rb", 
but I've completely forgotten the context.

- Gordon



From greg.ewing at compaq.com  Sun Jul 25 18:20:42 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Mon, 26 Jul 1999 10:20:42 +1200
Subject: Extreme Newbie Question
References: <MPG.1202263db22c797c989681@news.globix.net>
Message-ID: <379B8DBA.3BAFD1F7@compaq.com>

Robert Zurer wrote:
> 
> In interactive mode how do you save the script that you've entered line
> by line to a file?

You don't. Interactive mode isn't really designed to
be used that way -- it's more for throwaway experimentation.

If you're developing a script, it's best to enter it
into a file right from the beginning. This is very
easy if you're using IDLE. Just open up a new window,
type in some code, save, and hit F5 to run it. Check the
output, tweak, rerun, repeat as necessary.

If you're not using IDLE, you can do much the same
thing with a text editor and a shell window side by
side. Enter your code in the text editor, and use
"python myscript.py" or whatever in the shell window
to run it.

Hope that helps,
Greg



From sweeting at neuronet.com.my  Fri Jul 23 05:21:55 1999
From: sweeting at neuronet.com.my (sweeting at neuronet.com.my)
Date: Fri, 23 Jul 1999 09:21:55 GMT
Subject: "libmysqlclient.so.6" not found. (what does /sbin/ldconfig /usr/local/lib/mysql do ?)
Message-ID: <7n9c7f$ihd$1@nnrp1.deja.com>

When building Python from source, and with the MySQLmodule in the
Modules directory, I got the following error :

Shared object "libmysqlclient.so.6" not found
*** Error code 1

But libmysqlclient.so.6 does indeed exist in /usr/local/lib/mysql.

Fortunately, somebody had  come across this in a previous post/thread
and the cure was to run " /sbin/ldconfig /usr/local/lib/mysql" before
compiling. Sure enough this worked and now I have a fully functioning
Python interpreter (and environment)  :)

However, upon rebooting the machine, if I try to run a Python program
it will again complain that "libmysqlclient.so.6" can not be found
(since the MySQLmodule is shared). Once again
running " /sbin/ldconfig /usr/local/lib/mysql"  cures the problem but
does this mean that I have to run it each and every time I restart the
computer, as part of my startup scripts ? Is this error indicative of
there being a larger problem or misconfiguration ?

Thanking you in advance,

chas


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From tismer at appliedbiometrics.com  Mon Jul 12 15:23:25 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Mon, 12 Jul 1999 19:23:25 GMT
Subject: ANN: Stackless Python 0.3
Message-ID: <378A40AD.B4FDB79F@appliedbiometrics.com>

ANNOUNCING:

               Stackless Python 0.3
           A Python Implementation Which
             Does Not Use The C Stack

What is it?
A plugin-replacement for core Python.
It should run any program which runs under Python 1.5.2 .
But it does not need space on the C stack.

Why did I write it?
Stackless Python was never written before (afaik), since it
was said to be impossible without major rewrites of core Python.
I am proving the controverse:
It is easy to write, just hard to think.

Recent changes:
Version 0.3 has been written while developing a continuation
module which is about to ship. A couple of enhancements and
changes were necessary to make continuations fly. Some drastic
changes to frame management have been made. Stackless Python
is now also able to keep track of recursive interpreter
incarnations and to detect/resolve possible collisions
between their frames. As a proof of concept, continuationsmodule.c
will ship in a few days.

Who needs it?
At the moment, this is only useful for C programmers
who want to try certain new ideas. Hardcore stuff.
It allows to modify the current execution state by
changing the frame stack chain without restictions,
and it allows for pluggable interpreters on a per-frame-basis.

The possibilities are for instance:

Continuations, with which we will build Coroutines and Generators

Restartable exceptions and Persistent execution state
are the next topic to think about. They are possible.

Stackless extension modules can be built. The new builtin
stackless "map" function is a small example for this.

Coroutines are able to run at the speed of
a single C function call, which makes them a considerable
alternative in certain algorithms.
This is no longer a speculation since I have working
coroutine prototypes, written with continuations.

Status of the project:
Stackless-ness has been implemented and tested with pystone.
pystone works correctly and is about 10 % slower than
with standard Python.

What I need at the moment is
- time to do the right design of coroutines on top of continuations
- your input, your testing, critics and hints.

Some still rough documentation is available at
http://www.pns.cc/stackless/stackless.htm

Source code and a VC++6.0 build for Windows can be found
from the document or directly from
ftp://ftp.pns.cc/pub/stackless_990712.zip
ftp://ftp.pns.cc/pub/stackless_990712_win32.zip

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From Randy.Heiland at p98.f112.n480.z2.fidonet.org  Thu Jul  1 04:06:27 1999
From: Randy.Heiland at p98.f112.n480.z2.fidonet.org (Randy Heiland)
Date: Thu, 01 Jul 1999 09:06:27 +0100
Subject: embedding; return array
Message-ID: <0000162b@bossar.com.pl>

From: Randy Heiland <heiland at ncsa.uiuc.edu>

I'd welcome any pointers into being able to return an (unsigned char)
array from an embedded module.

thanks,
Randy



From tim_one at email.msn.com  Thu Jul 22 03:07:50 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 22 Jul 1999 03:07:50 -0400
Subject: Python and Boehm-Demers GC, I have code.
In-Reply-To: <p5hfmymh9z.fsf@bidra241.bbn.hp.com>
Message-ID: <001501bed410$e91fe760$642d2399@tim>

[Markus Kohler]
> IMHO this is one of Pythons major design flaws.

[Tim]
>> What, specifically?

[Markus]
> That flaw is that loops are not implemented using closures.

Yet you later note too that functions in Python are unusually slow.  It's no
flaw for a language implementation to decline to build everything on its
least optimized constructs <wink>.

> Instead as far as I understand  oops in Python are implemented such that
> the loop counter is created and destroyed for each step of the loop.

Yes, "for" loops work this way in Python.

> Even if you optimize this heavily you will probably slower than
> with a design that "recycles" the loop variable.

The for/__getitem__ protocol requires passing a full-blown int object, and
int objects are immutable.  Trying to trick this is more trouble than it's
worth.

> You could compare this to tail recursive versus non-tail recursive
> functions.

Trust me on this one:  you don't want to get anywhere near this comparison
unless you want to see Guido running in the opposite direction <wink>.
Seriously, I agree it's a good comparison; Python was designed to favor
iteration over recursion, though, and for that reason too it's not like the
languages you appear to like better.

> Also closures make the implementation of efficient loops more
> difficult

Yet another killer reason to do it that way <wink>.

> they have the advantage that loops and other control structures can be
> better integrated into the OO concept. Smalltalk for example doesn't need
> Iterators for that reason.

I happen to like the CLU/Sather/Icon approach toward iterators
(semi-coroutines)better than Python's or Smalltalk's (or C++'s, or Perl's,
or ...); Python doesn't work that way under the covers either, but is
surprisingly *close* to being able to work that way.  So that way makes more
sense to me for Python; Python's implementation and its surface semantics
are very close in most areas.

>> Loop overhead is typically lareger in Python programs than in other
>> languages, but is also typically trivial all the same.

> Lately ee have had already some examples in this group where
> loop overhead was a major factor.

?  I don't recall that.  Last one I do recall, someone posted a pretty much
certifiably insane quadruply-nested numeric loop, recomputing range again at
each "for" level.  Several replies suggested floating the range calls out of
the loops to save time.  Christian Tismer quite properly pointed out that
this was pointless, since the loop overhead accounted for a tiny percent of
the runtime.  Someone else suggested an algorithmic change that chopped
orders of magnitude off the runtime.  An extreme outcome, but typical in
outline.

> I measured different kind of loops and even the more general
> while loop was faster in squeak than anything else in python.

Empty loops?  Or did they do something?  If empty, it's hard to care.  If
they did something, why do you think the difference is accounted for by the
loop mechanism rather than by the guts of what the loops did?  A specific
example would clarify matters.

> Functions calls seems also to be much slower in Python ...

Yes.  Even worse <wink>, when a (say) Scheme person comes to Python, one of
the surest ways to speed up their first programs is to encourage them to
rewrite their lambda-slinging code as explicit inline loops.  That is

    plus1 = map(lambda x: x+1, somelist)

runs much slower than

    plus1 = []
    for x in somelist:
        plus1.append(x+1)

> ...
> Another problem with loops could be reference counting.
> I'm wondering how often a reference count get's changed when
> running a loop ...

It really depends on the specifics of the loop body.  In general, each
instance of a name in a line of code causes at least one refcount operation
each time the line is executed.  There *is* a lot of refcount fiddling going
on, and nothing is ever exempted from it.  It's also the case that nothing
in Python is stack-allocated; every object may have indefinite extent,
there's no way to tell at compile-time, and so everything is heap-allocated
"just in case".  If you're looking for reasons why Python is slower than
many other languages, they're not hard to find <wink>.

darned-hard-to-change-though!-ly y'rs  - tim





From tratt at dcs.kcl.ac.uk  Sun Jul 11 09:12:24 1999
From: tratt at dcs.kcl.ac.uk (Laurence Tratt)
Date: Sun, 11 Jul 1999 13:12:24 GMT
Subject: database
References: <3787c551.1@ganymede.datanet.hu>
Message-ID: <9390dd1f49.tratt@tratt.freeuk.co.uk>

In message <3787c551.1 at ganymede.datanet.hu>
          "dzsi" <granadam at mail.datanet.hu> wrote:

> i have a database (comma separated text), and i would like to get some
> data out of it based on some query. let's say there are seven fields,
> f1...f7, and i am searching for entries in f1, and want to display f1..f7 
> for the field that matches the query. how do i go about doing that?

Probably the first thing is to make / get a library that handles your
database types. By the sounds of things, you're using a CSV file so I'm
bound to reccomend my own CSV library <wink>. It's at version 0.17 and is
available from:

  http://eh.org/~laurie/comp/python/csv/

Then you can do fun things like:

    import CSV
    csv = CSV.CSV()
    csv.load(<file name>, 1)
    for entry in csv:
        if entry["f1"] <matches my criteria>:
            print entry    # prints f1 ... f7


Laurie
-- 
http://eh.org/~laurie/comp/python/



From Michael.Hudson at p98.f112.n480.z2.fidonet.org  Sun Jul  4 14:26:08 1999
From: Michael.Hudson at p98.f112.n480.z2.fidonet.org (Michael Hudson)
Date: Sun, 04 Jul 1999 19:26:08 +0100
Subject: Another bytecodehacks!
Message-ID: <000016c4@bossar.com.pl>

From: Michael Hudson <mwh21 at cam.ac.uk>

It's been a while, but there's a new version of bytecodehacks on the
starship:

http://starship.python.net/crew/mwh/

Quite a lot has changed, especially under the hood. There's a
design-by-contract module, a function that will make another tail
recursive, a set of simple rationalizations/optimizations (originally
posted to this newsgroup by Corran Webster) and probably lots of stuff
I've forgotten about.

The documentation has been updated, but there are probably places
where it's out of date, or just plain wrong. If you find such a place
please do tell me!

The documentation contains what I think is quite a nice description of
one way of avoiding cycles in complex data structures (look at the
doumentation of fleshy).

Have fun, and thanks for reading!

Michael



From stidolph at origin.ea.com  Wed Jul 28 19:31:18 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Wed, 28 Jul 1999 23:31:18 GMT
Subject: 1.5.2 for: else:
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC5B@molach.origin.ea.com>

Please show your example using the else .. I'd like to see it.

-----Original Message-----
From: Vadim Chugunov [mailto:chega_ at yahoo.com]
Sent: Wednesday, July 28, 1999 4:58 PM
To: python-list at cwi.nl
Subject: Re: 1.5.2 for: else:


I would say that else: clause after a loop is really related to the if:
controlling the break,
rather than to the loop itself.

When I first saw Python syntax for loops I said: "A-ha!  So in Python I will
not
need a goto
in situations like this:
-----
for(Item* pitem=pseq->First(); pitem; pitem=pitem->Next())
    if (pitem->key==42)
        goto Found;

pitem = pseq->Insert(new Item());
Found:
    // use item pointed to by pitem
-----
In fact, I do not see any other good use for else: clause in a loop.
Maybe the docs should just explicitly say what this feature is good for ?

Vadim

William Tanksley <wtanksle at dolphin.openprojects.net> wrote in message
news:slrn7pum9t.2ro.wtanksle at dolphin.openprojects.net...
> On Wed, 28 Jul 1999 00:50:14 GMT, Fredrik Lundh wrote:
> >William Tanksley <wtanksle at dolphin.openprojects.net> wrote:
> >> The very worst part of the current else: behavior is that it changes
the
> >> meaning of else.  In other constructs, else is an alternate path to
take
> >> if the data being tested fails a single expression test.  In this
> >> contruct, else is a path taken if the code block belonging to the
previous
> >> test executes a certain instruction.
>
> >nope.  you've got it all backwards.  consider this:
>
> >    for an "if" statement, "else" is a path taken if the
> >    expression evaluates to false.
>
> Close enough -- I would say it's the path taken if the 'if' block isn't
> executed.
>
> >    for a "while" statement, "else" is a path taken if the
> >    expression evaluates to false.  or in other words,
> >    when the loop terminates by natural causes.
>
> In other words, it executes when you'd expect an 'else' to not execute --
> 'else' doesn't mean "natural causes".
>
> >    for a "for" statement, "else" is a path taken when
> >    there are no more elements to loop over.  or
> >    in other words, when the loop terminates by
> >    natural causes.
>
> In addition to the above carping, I have to add that the code following
> the for (or while) loop is what I'd expect to execute after the loop
> terminates.
>
> >not that complicated, was it?
>
> I didn't think it was either, but it seems that in spite of my initial
> understanding, I still managed to reverse it accidentally.
>
> >now, what you seem to have trouble with isn't
> >the "else"-clause -- it's the behaviour of "break":
>
> >"else" always works the same way.  and so does
> >"break".  what else did you expect from Guido?







From markusk at bidra241.bbn.hp.com  Tue Jul 27 07:50:17 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 27 Jul 1999 13:50:17 +0200
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com> <al86737ocwo.fsf@sirppi.helsinki.fi> <p5n1wjmmxc.fsf@bidra241.bbn.hp.com> <01d901bed782$b5c41740$f29b12c2@secret.pythonware.com>
Message-ID: <p5n1wimhvq.fsf@bidra241.bbn.hp.com>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> Markus Kohler <markusk at bidra241.bbn.hp.com> wrote:
> > My guess is that what calling speed makes slow on Python are all those
> > different kinds of ways to call the same function. 
> > Examples are default arguments and named arguments. 
> > IHMO these features may save some typing but they are not worth
> > sacrifying performance nor do they help increasing readability of the
> > code. 
> 
> really?  back in the 1.2 days, you had to write:
> 
>     button = Button(frame, {"text": "QUIT", "fg": "red", "command": frame.quit})
>     button.pack({"side": LEFT})
> 
> instead of:
> 
>     button = Button(frame, text="QUIT", fg="red", command=frame.quit)
>     button.pack(side=LEFT)
> 
> are you sure you find the first one being easier
> to read?

Not a big difference. But that's probaly a matter of taste. 

in Smalltalk the first call could look like:

Button  frame: aFrame text: 'QUIT' fg:'red' command: #quit

the name of the message would be frame:text:fg:command:
If you wanted  a default argument for fg for example you would
have to define a new message for example
frame:text:command:

If you get too many arguments you should change the design anyway  
 to be more OO and define a message 
frame:properties:

for example. 

> 
> on the other hand, 1.2 was much slower than the
> current releases.  go figure...
> 

I dont' doubt that. I'm just saying that there are implementation for
similiar languages how are significantly faster. 

Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From MHammond at skippinet.com.au  Thu Jul  8 03:40:16 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Thu, 8 Jul 1999 17:40:16 +1000
Subject: COM with Python?
References: <1280807192-67537824@hypernet.com>
Message-ID: <7m1kgr$7nt$1@m2.c2.telstra-mm.net.au>

Gordon McMillan wrote in message <1280807192-67537824 at hypernet.com>...
[Markus Stenberg]
> > >Spesifically, I'm looking for information on activation policies; how
to
> > >implement idispatch objects that are singletons for a single machine.
[Mark Hammond]
> > Your best bet is DejaNews - the singleton subject was done fairly
> > thourghly a while ago, and no significan changes have been made to
> > the COM package to make it any eaiser (alas)


> Hmm. I've read those posts, and never did find a resolution.

Well, what I meant was that all the threads Gordon mentions has done the
topic, as far as I can tell, pretty much to its conclusion with the current
state of win32com.  I agree the solutions, if applicable, still dont really
qualify as a "resolution".  Anything better will require some win32com
changes tho, and these havent been made (or even identified)

Ive never needed to implement a singleton, but my personal solution (from
someone else I cant recall in those threads) would be to implement the
functionality in a single Python object and have the COM objects delegate to
it.  Thus you dont really have a singleton COM object (as you are giving off
as many as necessary) but by keeping the functionality and state in a single
Python object you are can get the same result (for most cases!)

Mark.







From tre17 at pc142.cosc.canterbury.ac.nz  Tue Jul 20 20:40:25 1999
From: tre17 at pc142.cosc.canterbury.ac.nz (Timothy R Evans)
Date: 21 Jul 1999 12:40:25 +1200
Subject: tkinter call backs
References: <01bed202$58879280$e855c3d1@Cyberus.cyberus.ca>
Message-ID: <c754siy6dja.fsf@pc142.cosc.canterbury.ac.nz>

"Gordon Williams" <g_will at cyberus.ca> writes:

> I have a simple problem that I cant find the solution.  I have a menu item
> to open a file which has a command "openfile".  I would like to activate
> "tkFileDialog.askopenfilename" to get the name of the file.  This requires
> "parent=root" or something like that for the position of the message box.
> 
> How do I pass "root" into openfile?  
> 
> Thanks
> Gordon Williams
> 
> 
> 
> def openfile():
>     print "opening file"
>     fileName=tkFileDialog.askopenfilename(parent=root, title="Open")
>     if fileName:    #checks for cancel
>         print fileName  #name of file to open
>         file=open(fileName, 'r')
>         read='junk' #initialize read with something
>         while read != "":
>             read=file.readline()
>             print "number1",read[0:-1] # strips off line return character
> #       This has one more line than needed due to EOF line.
> #       Will use seek() to move to 2nd line
>         print "number of bytes ",file.tell()
>         file.close()

I get from this that you are wanting a Tkinter callback to execute
somthing like foo(bar) rather than just foo().  The Command class I
posted twice before does this:

class Command:
    def __init__(self, func, *args, **kw):
        self.func = func
        self.args = args
        self.kw = kw
                       
    def __call__(self, *args, **kw):
        args = self.args + args
        kw.update(self.kw)
        apply(self.func, args, kw)

Used as following:

...
Button(text='foo', command=Command(myfunction, myarg))
...

The args (including keywords) that are passed to the Command class are
passed on to myfunction when the callback is activated.

I'm wondering if this should be part of Tkinter in some way?

--
Tim Evans



From fredrik at pythonware.com  Thu Jul  1 02:55:52 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 1 Jul 1999 06:55:52 GMT
Subject: readline completion
References: <377AB209.5A5B9D32@starvision.com>
Message-ID: <013b01bec390$2d703d70$f29b12c2@secret.pythonware.com>

Angus MacKay wrote:
> how is readline completion done for python with the readline module?
> 
> I looked in the readlinemodule.c and saw that ESCESC was bound to
> completion but it never completes on anything for me.

http://www.python.org/doc/current/tut/node13.html
look under "key bindings"

</F>




From arcege at shore.net  Thu Jul  8 13:37:17 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 08 Jul 1999 17:37:17 GMT
Subject: PyImport_ImportModule, PyModule_GetDict
References: <3784C57B.59DF0D33@ncsa.uiuc.edu>
Message-ID: <hl5h3.1285$6M6.394545@news.shore.net>

Randy Heiland <heiland at ncsa.uiuc.edu> wrote:
: If I understand this correctly, the equivalent of doing this:

:>>> import mymodule
:>>> dir(mymodule)

: is sort of this, in the C API:

: PyObject *mod = PyImport_ImportModule("mymodule");
:  PyObject *myDict = PyModule_GetDict(mod);

: Q:  what is the C API equivalent of this:

:>>> from mymodule import *
:>>> dir()


You will have to do something a little more:

    PyObject *fromlist, *main, *globals;
    PyObject *sysmod, *sysdict, *key, *val;
    int pos;  /* used internally by PyDict_Next() */

    /* tell __import__ to return the top-level module */
    fromlist = Py_BuildValue("[s]", "*");
    main = PyImport_ImportModule("__main__");
    globals = PyModule_GetDict(main);

    /* ImportModuleEx handles "from package.module import *" as well */
    sysmod = PyImport_ImportModuleEx("sys", globals, globals, fromlist);

    sysdict = PyModule_GetDict(sysmod);
    /* perform standard name bindings */
    while (PyDict_Next(sysdict, &pos, &key, &val) > 0) {
      if (!PyString_Check(key) ||
          (PyString_AS_STRING(key)[0] != '_'))
        PyDict_SetItem(globals, key, val);
    }

If you look at the documentation for __import__, it does not perform the
name binding in the globals dictionary (for some reason ;), so that must
be performed manually.

  -Arcege




From mnot at pobox.com  Sat Jul 24 04:56:47 1999
From: mnot at pobox.com (Mark Nottingham)
Date: Sat, 24 Jul 1999 08:56:47 GMT
Subject: HTTP 1.1 anyone?
References: <37997331.0@pink.one.net.au>
Message-ID: <004301bed5b2$769dca60$0301a8c0@mnot.net>

Which part(s) of it? Are you looking for just one feature, or everything
(everything being a pretty tall order)?


----- Original Message -----
From: Arnold de Vos <adv at langdale.com.au>
Newsgroups: comp.lang.python
To: <python-list at cwi.nl>
Sent: Saturday, July 24, 1999 6:02 PM
Subject: HTTP 1.1 anyone?


> Is anyone working on an HTTP 1.1 implementation in Python?   I am
> considering creating variants of BaseHTTPServer and httplib to handle the
> 1.1 protocol.  But I would not like to re-invent the wheel.
>
> Thanks,
> Arnold
>
> ---
> Arnold deVos
> Langdale Consultants
> adv at langdale.com
>
>
>
>
>




From eugene.leitl at lrz.uni-muenchen.de  Fri Jul 23 15:07:44 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Fri, 23 Jul 1999 12:07:44 -0700 (PDT)
Subject: Python documentation available as ZIP files
In-Reply-To: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
Message-ID: <14232.48512.648609.496768@lrz.de>

Fred L. Drake writes:
 > 
 >   I've received enough hostile mail from Windows users not
 > understanding how to unpack .tgz archives (er, WinZip works for me...)

I think this really means lowering the standards. I wouldn't advocate
using a more efficient (but obscure) standard as bzip2 (at least not
yet), but everybody who is unable to unpack .tar.gz/.tgz archives (you
did include a pointer to WinZip in the documentation, didn't you?) is
perhaps not the right person to read Python documentation anyway.

 > that I will be making most of the documentation packages available as
 > ZIP archives as well as .tgz archives.



From pserrand at worldnet.fr  Sun Jul  4 12:54:50 1999
From: pserrand at worldnet.fr (pierrette)
Date: Sun, 04 Jul 1999 18:54:50 +0200
Subject: Upgrade or not ??
References: <377E0A3B.CA3B258C@worldnet.fr> <19990704224912.16776@nms.otc.com.au>
Message-ID: <377F91D9.D3A1E147@worldnet.fr>

Thanks, exactly what i needed.
Is there a way to find out what events a widget support ?
I browsed documentation on python.org but didn't find anything on FocusIn en
FocusOut. I there somewhere documentation on that ??

Pierrette

Greg McFarlane wrote:

> Depending on exactly what you want to do, it may be possible in
> Tkinter.  You can use something like this to find out when a widget
> gets and loses focus:
>
>     import Tkinter
>
>     def gotFocus(event):
>         print 'Focus in'
>     def lostFocus(event):
>         print 'Focus out'
>
>     root = Tkinter.Tk()
>
>     e = Tkinter.Entry()
>     e.bind('<FocusIn>', gotFocus)
>     e.bind('<FocusOut>', lostFocus)
>     e.pack()
>
>     root.mainloop()
>
> Also, you can use this to set the focus to a widget.
>
>     e.focus_set()
>
> On 3 Jul, pierrette wrote:
> > First a general question :
> > Actually we're working with Python 1.5.1 and we're satisfied with it.
> > Should we upgrade to 1.5.2 and if yes WHY?
> >
> > No a Tkinter guestion :
> > We developped a prototype of an application with python an tkinter, now
> > we have to
> > decide if we're gonna write the real app in python or something else !
> > We now allready that in principle we could do it with python, the thing
> > we're missing
> > in tk is a 'on get focus' ,a 'on lost focus' event and a 'set focus to'
> > command
> > Pe: on lost focus of inputfield 1:
> >             if inputfield1 = A:
> >                 set focus to field2
> >             elif inputfield1 = B:
> >                 set focus to field 3
> >
> > If those things are not available in tkinter what alternatives do we
> > have ?
> > - Changing the GUI (would be a shame but feasable), knowing that is has
> > to be multi-platform WIN-LINUX? If so to what ?
> > - Not write in python ? In this case we lose the work done for the
> > prototype
> >
> > Thanks for comments and remarks
> >
> > Pierrette Serrand
> > pserrand at worldnet.fr
> >
> >
> >
> >
> >
>
> --
> Greg McFarlane:    INMS Telstra Australia (gregm at nms.otc.telstra.com.au)
> Today's forecast:  Sunny, with occasional cloudy periods and a chance
>                    of precipitation in some areas.




From nascheme at ucalgary.ca  Thu Jul 15 18:25:25 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 15 Jul 1999 22:25:25 GMT
Subject: Getters and Setters
References: <000701bece99$6dbeede0$51a22299@tim> <m3aesy43jt.fsf@greebo.nodomain.de>
Message-ID: <7mln4l$c38@ds2.acs.ucalgary.ca>

Bernhard Herzog <herzog at online.de> wrote:
>So, putting the functions into the class would be better:
>
>              def getter(self, key=name[3:]):
>                  return getattr(self, key)
>              setattr(self.__class__, name, getter)
>              return getattr(self, name)

This is a very nice solution.  Thanks Bernhard.  It is a little
slower than Tim's method but avoids creating circular references
and doesn't create a function for each instance.  It also works
for class variables.


    Neil



From olivier.deckmyn at mail.dotcom.fr  Mon Jul 19 06:49:36 1999
From: olivier.deckmyn at mail.dotcom.fr (Olivier Deckmyn)
Date: Mon, 19 Jul 1999 12:49:36 +0200
Subject: [Q] Type checking...
Message-ID: <7mv000$sun$1@feed.teaser.fr>

Hi !

Abstract : [Q]How to check the type of an instance before(after?) sending it
to a method...

Here is my source :

class ClassA
...


class ClassB :
   def myMethod(self, anInstance)
...

and myMethod expects that anInstance is a ClassA !
How can I force my user to send only instances of this class (or subclasses)
?

Thx !









From mfein at aplcomm.jhuapl.edu  Wed Jul 14 13:36:21 1999
From: mfein at aplcomm.jhuapl.edu (Matt Feinstein)
Date: Wed, 14 Jul 1999 17:36:21 GMT
Subject: newbie questions about sys.path, wxpython, IDLE and pythonwin
References: <378B7A69.12874AAC@gnet.tn> <7mggvo$aug$1@m2.c2.telstra-mm.net.au>
Message-ID: <378cc9b6.22412677@aplnews>

On Wed, 14 Jul 1999 09:11:27 +1000, "Mark Hammond"
<MHammond at skippinet.com.au> wrote:

>MB wrote in message <378B7A69.12874AAC at gnet.tn>...
>>I need to know exactly how the sys.path is built. In my system I have no
>>pythonpath variable but the sys.path returns a list. I suppose it was
>>built by default. This list doesn't contain all my python
>>subdirectories. So I want to know when and how this list was built ?
>>(yes I examined the registry and found references to python directories
>>but they don't match my sys.path)
>
>Python uses a combination of the values you found in the registry, and some
>default values - things like ".\lib" are added always, regardless of the
>registry.  Remember, you have the same source to see how it works that I do.
>

 Well, with all due respect, this can't be the whole story--  after
installing the numeric extension, I found the numeric directory in
sys.path, but not in the registry list.  So, what -is- going on here?

--
Matt Feinstein
mfein at aplcomm.jhuapl.edu
Organizational Department of Repeated
and Unnecessary Redundancy



From jhspies at futurenet.co.za  Tue Jul  6 12:17:28 1999
From: jhspies at futurenet.co.za (Johann Spies)
Date: Tue, 6 Jul 1999 16:17:28 GMT
Subject: AttributeError - a dumb question
In-Reply-To: <199907061552.KAA23749@Starbase.NeoSoft.COM>
References: <199907061552.KAA23749@Starbase.NeoSoft.COM>
Message-ID: <Pine.LNX.3.96.990706181322.770A-100000@Johann>

Hallo Tres,

Thanks for your reply.

On Tue, 6 Jul 1999, you wrote:

> In article <Pine.LNX.3.96.990706171320.374A-100000 at Johann> you write:
> >Why does the following code produce an attribute error?
> >class F:
> >    def __init__(self):
> >
> >          x = self.Krylys()
> 
> If you want 'x' to be an attribute of the current instance, you
> need to say, 'self.x = self.Krylys()' here;  otherwise, you are
> assigning to a local variable of the __init__ method.

I have tried that.  Here is the new code with output (sorry for the type
in the previous example:

class F:
    def __init__(self):

        self.x = self.Krylys()

    def KryLys(self):
        l = [1,2,3,4,5]
        return l
    

l = F
print dir(l)
print hasattr(l, 'x')
print hasattr(l,'Krylys')
print l.x

output:

['KryLys', '__doc__', '__init__', '__module__']
0
0
Traceback (innermost last):
  File "<stdin>", line 15, in ?
AttributeError: x


> Because 'Krylys' <> 'Krypos'?

Stupid typing error on my side.

Johann



 --------------------------------------------------------------------------
| Johann Spies                                 Windsorlaan 19              |
| jhspies at futurenet.co.za                3201 Pietermaritzburg             |
| Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
 --------------------------------------------------------------------------

     "Ask, and it shall be given you; seek, and ye shall
      find; knock, and it shall be opened unto you."        
                                  Matthew 7:7 





From Martin.Casado at p98.f112.n480.z2.fidonet.org  Thu Jul  1 23:17:38 1999
From: Martin.Casado at p98.f112.n480.z2.fidonet.org (Martin Casado)
Date: Fri, 02 Jul 1999 04:17:38 +0100
Subject: C shared object files
Message-ID: <00001680@bossar.com.pl>

From: Martin Casado <casado at ogl.llnl.gov>


  Python Guru's,

   I am trying to get python to dynamically load a C object module, and 
first off I can't figure out how to compile myCollProggy.c into a 
shareable object file, and secondly I was curios if there is a way to check
at runtime if a module is on the system or not.  

  I have made arrangements with the proper authorities and there is a veritable
wealth of Karma points waiting for anyone who can help me :-)  Hehe thanks
very much in advance.

			-Martin 




From eugene.leitl at lrz.uni-muenchen.de  Sat Jul  3 22:50:58 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Sat,  3 Jul 1999 19:50:58 -0700 (PDT)
Subject: Language extensibility (was: Why is tcl broken?)
In-Reply-To: <wmcn1xenodh.fsf@polytechnique.fr>
References: <1283099377-77267484@hypernet.com>
	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com>
	<375FDCA7.62FF@mailserver.hursley.ibm.com>
	<7josm1$bgc$1@Starbase.NeoSoft.COM>
	<37628EBF.207F7A8F@iname.com>
	<37682DD3.43923243@compaq.com>
	<7ka2be$qja$1@nntp6.u.washington.edu>
	<37776FFA.2D85@mailserver.hursley.ibm.com>
	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it>
	<876748mcxg.fsf@home.ivm.de>
	<37788BB9.ABD@mailserver.hursley.ibm.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.ibm.com>
	<nkjzp1hizvb.fsf@tfeb.org>
	<377A3383.D453468@iname.com>
	<377aa228.31380423@news.jpl.nasa.gov>
	<slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net>
	<wkzp1g2t2e.fsf@ifi.uio.no>
	<ey3r9mskvja.fsf@lostwithiel.tfeb.org>
	<x27lojoums.fsf@todday.aiai.ed.ac.uk>
	<wmcn1xenodh.fsf@polytechnique.fr>
Message-ID: <14206.52213.119553.809177@lrz.de>

Jerome Kalifa writes:

 > > Well *Lisp* dates back to the late 50s.  Scheme may be from 75,
 > > but Lisp is a different matter.  And lambda calculus is, what,
 > > 1949?
 > > 
 > 
 > No, as a mathematical theory, it was developed before the war.

Church, A., The Calculi of Lambda Conversion, Princeton University
Press, Princeton, N.J., 1941 



From sjoerd at oratrix.nl  Thu Jul  8 04:47:57 1999
From: sjoerd at oratrix.nl (Sjoerd Mullender)
Date: Thu, 08 Jul 1999 10:47:57 +0200
Subject: Emacs: highlighting "self" tweak
In-Reply-To: Your message of 07 Jul 1999 23:12:33 +0200.
             <pcoiu7w17we.fsf@thoth.math.ntnu.no> 
References: <7luqpa$3s9$1@nnrp1.deja.com> <19990707133146.F3AD4301CF2@bireme.oratrix.nl> 
            <pcoiu7w17we.fsf@thoth.math.ntnu.no> 
Message-ID: <19990708084759.75830301CF2@bireme.oratrix.nl>

On Wed, Jul 7 1999 Harald Hanche-Olsen wrote:

> + Sjoerd Mullender <sjoerd at oratrix.nl>:
> 
> | You can use this piece of magic in your .emacs file (tested on XEmacs
> | 21.1):
> | 
> | (add-to-list 'after-load-alist
> | 	     (cons "python-mode"
> | 		   '((add-to-list (quote python-font-lock-keywords)
> | 				  (quote ("\\b\\(self\\)[ .,]?"
> | 					  1 font-lock-type-face))))))
> 
> which could be written more briefly as
> 
> (eval-after-load 
>  "python-mode"
>  '((add-to-list
>     'python-font-lock-keywords
>     '("\\b\\(self\\)[ .,]?" 1 font-lock-type-face))))

You learn something new every day!

But I think you have a couple of parens too many.  What about this:

(eval-after-load 
 "python-mode"
 '(add-to-list
    'python-font-lock-keywords
    '("\\b\\(self\\)[ .,]?" 1 font-lock-type-face)))

-- Sjoerd Mullender <sjoerd.mullender at oratrix.com>



From larsga at ifi.uio.no  Thu Jul 29 03:07:32 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 29 Jul 1999 09:07:32 +0200
Subject: Bit-twiddling
References: <000f01bed969$fbed9f00$71a22299@tim>
Message-ID: <wklnc0x7bf.fsf@ifi.uio.no>

* Tim Peters
| 
| Provided that seed was a long to begin with, this code almost works
| as is.  You just need to append "L" to the literals (for 69069 that
| isn't necessary, it just avoids the repeated expense of runtime
| widening; for 0xffffffff it *is* necessary on a 32-bit machine, else
| 0xffffffff == -1 gets widened to -1L and then the masking doesn't
| accomplish anything (i & -1L == i for all long i)).

Aha! I knew that 0xffffffff didn't work, but I had no idea that the
trick was to define it as a long. Thanks!
 
| The elements of mt are then all longs, but each one's *value* fits
| in a 32-bit C unsigned long: numerically, it computes the same
| values as would the C code.

Indeed it does. Thanks again! 

Based on this advice I took a stab at the rest of the algorithm and
now it actually works! Wheee! So if anyone wants a string of random
numbers with a period of 2**199937-1 and equi-distributed in 623
dimensions, just let me know. :-)
 
| if-you're-really-after-random-numbers-see-ivan-frohne-ly y'rs  - tim

I guessed that you would recognize the algorithm. :)

I wasn't really after random numbers as much as I was after some easy
diversion. This turned out to not be that easy for someone with a bit
(hah!) less than the necessary maths and C background. :(

Anyway, I see that Frohne uses the code I use above, except that his
69069 is L. But as you say that makes no difference. He doesn't use
the same code for number generation, though.

If I need more diversion I'll try to turn the C code into an extension
module as my first-ever extension module. Should be educational, at
least.

--Lars M.



From tim_one at email.msn.com  Tue Jul 20 01:06:38 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 20 Jul 1999 01:06:38 -0400
Subject: Stackless & String-processing
In-Reply-To: <7mvip2$cpj$2@cronkite.cc.uga.edu>
Message-ID: <001101bed26d$a5bc4580$139e2299@tim>

[Tim, on Graham & Fernando's combinator thread]
> Trust me on this one, fellows:  you each understand the other,
> and have for several (interesting) msgs already. ...

[Graham]
> Tim I don't think that we are agreeing.

Who said anything about agreeing?  You two obviously disagree, but
disagreement doesn't imply misunderstanding.  Heck, you've both repeated the
same unbudging positions so often even *I* understand what you're both
saying <wink>.

less-filling-tastes-great-ly y'rs  - tim





From heiland at ncsa.uiuc.edu  Wed Jul  7 17:24:12 1999
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Wed, 07 Jul 1999 16:24:12 -0500
Subject: embedding strategies; C++ server
Message-ID: <3783C57C.C4BDAED6@ncsa.uiuc.edu>

I have a scenario where I have a "black box" (C++) server that executes
"arbitrary" Python commands (strings) handed to it.  Consider the
following hypothetical situation:

client:  cmd = "init"
server:  Py_Initialize();

client:  cmd = "from geometry2D import *"
server:  PyRun_SimpleString(cmd);

client:  cmd = "square = Quad(0,0, 1,0, 1,1, 0,1)"
server:  PyRun_SimpleString(cmd);

client:  cmd = "square.GetPoints()"
server:  <detects that "GetPoints" is being invoked and
obtains/processes
              the actual points in C/C++>

Which embedding strategy lets me do this last step in the server? (any?)

thanks,
--Randy




From vit at lab136.cyber.kiev.ua  Mon Jul  5 12:27:22 1999
From: vit at lab136.cyber.kiev.ua (Vitaliy Perekupka)
Date: Mon, 05 Jul 1999 19:27:22 +0300
Subject: Problem with ZOPE under WinNT
Message-ID: <3780DCEA.B43FE208@lab136.cyber.kiev.ua>

I trying to use ZPublisher.
When I try to start using URL http://localhost/cgi-bin/pcgitest/getTime
(test that go with ZPublisher 3.0.0)
I get ""Temporarily Unavailable "".
In source code of generated HTML I see

No such file or directory
pcgi-wrapper-version 2.0a4

But all paths are correct (I guess).

When I test pcgi info file "pcgitest" with pcgifile.py 
it generate next error:

error attempting: 'from cgi_module_publisher import publish_module'

I don't know what to do.

Please answer to vit at lab136.cyber.kiev.ua



From pereira at research.att.com  Sun Jul 18 23:37:54 1999
From: pereira at research.att.com (Fernando Pereira)
Date: Mon, 19 Jul 1999 03:37:54 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com> <170719990020380651%pereira@research.att.com> <7mqaht$r27$2@cronkite.cc.uga.edu> <170719991835331004%pereira@research.att.com> <7mssne$m8r$1@cronkite.cc.uga.edu>
Message-ID: <180719992337546352%pereira@research.att.com>

In article <7mssne$m8r$1 at cronkite.cc.uga.edu>, Graham Matthews
<graham at sloth.math.uga.edu> wrote:
> You have totally misunderstood what I said. I said "you can have regular
> expression parsers as primitive parsers". There is no need for the compiler
> to "solve the halting problem". The programmer is free to use regular
> expression/automaton techniques where he/she wants. Such parsers take
> strings and return ASTs. They can then be combined with other parsers
> as if they were primitive parsers.
One last try. I understood your point from the beginning. However, you
did not understood *mine*. If you push the complex
pattern-matching/parsing machinery into black boxes, there's not much
for the monadic glue to do, is there? And even then, some
optimizations, those involving the various black boxes, will be
blocked. Here's a simple example. Assume that there are two black boxes
A and B, and that you use monadic glue to say "A or B". A and B may be
deterministic, but the result will not be, and exhaustive search will
be required in the glue. Sure, I can recognize such cases by hand and
create a new black box "A or B" which is optimized, but that is an
error-prone hand optimization. And one may not notice subtler
inefficiencies.

-- F



From gmcm at hypernet.com  Fri Jul  9 10:41:41 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 9 Jul 1999 09:41:41 -0500
Subject: COM with Python?
In-Reply-To: <7m1kgr$7nt$1@m2.c2.telstra-mm.net.au>
Message-ID: <1280594885-80315141@hypernet.com>

Mark Hammond wrote:

> Gordon McMillan wrote:

> > Hmm. I've read those posts, and never did find a resolution.
> 
> Well, what I meant was that all the threads Gordon mentions has done
> the topic, as far as I can tell, pretty much to its conclusion with
> the current state of win32com.  I agree the solutions, if
> applicable, still dont really qualify as a "resolution".  Anything
> better will require some win32com changes tho, and these havent been
> made (or even identified)
> 
> Ive never needed to implement a singleton, but my personal solution
> (from someone else I cant recall in those threads) would be to
> implement the functionality in a single Python object and have the
> COM objects delegate to it.  Thus you dont really have a singleton
> COM object (as you are giving off as many as necessary) but by
> keeping the functionality and state in a single Python object you
> are can get the same result (for most cases!)

Well, having just done some testing, let me put a resolution on this: 
the global Python object approach works (which surprised me!), and 
the Running Object Table approach, while it works, seems to have 
stability problems.

To explain: I hadn't realized that the COM plumbing would, when using 
CLSCTX_LOCAL_SERVER, reconnect to a running instance, (so I doubted 
that the global Python object approach would work). It does, and 
while I haven't left one running for 24 hrs yet, it appears to work 
flawlessly - the COM invoked pythonw process goes away as soon as all 
refs go to zero.

The ROT approach (using pythoncom.RegisterActiveObject()) is flaky. 
Sometimes the COM invoked pythonw goes away when it should, sometimes 
it hangs around. Worse, if you leave it running for long periods, it 
seems capable of freezing NT's UI (the machine appears to still be 
alive, but you can't _do_ anything except reboot). I'm more than 
willing to believe that this is NOT Mark's fault.

Attached is Andy's SpamServer, with some mods to allow all clients on 
the same machine to work with the same SpamServer (well, to the same 
list of SpamCans). Which will probably next lead to questions about 
threading models...

- Gordon

-------------- next part --------------
# copyright Andy Robinson 1997
# you may freely modify and reuse this under the same terms
# as the main python licence

# registration stuff modified and expanded by Harri Pasanen
# thanks also to Mark Hammond and Greg Stein

# modified 7/8/99 to demonstrate having all clients on a
# machine work with the same SpamServer data. 
# Gordon McMillan

from win32com.server.exception import COMException
import win32com.server.util
import win32com.client.dynamic
import sys
import pythoncom

_cans = None

class SpamServer:
    _reg_clsid_ = "{EC419091-35FC-11D3-9403-00609736B700}"
    _reg_desc_ = "Python SpamServer GPO"
    _reg_progid_ = "Python.SpamServerGPO"
    _reg_class_spec_ = "spamserver_GPO.SpamServer"
    _reg_clsctx_ = pythoncom.CLSCTX_LOCAL_SERVER
    _public_methods_ = ['AddCan',
                            'GetCanCount',
                            'GetCanAt',
                            'DeleteCan',
                            'GetDescriptionList',
                            'GetCrosstab']
    # you need to explicitly list the public methods for COM to work
    # spelling errors here lead to errors when you call CreateObject
    # from a client
    
    def __init__(self):
        "it keeps a collection of data - start with a sample object"
        global _cans
        if _cans is None:
            _cans = [SpamCan()]
        self.cans = _cans
        
    def GetCanCount(self):
        "return the number of objects held"
        return len(self.cans)
        
    def AddCan(self, contents = 'Spam', type = 'Can', qty = 3):
        "two uses, can just add a generic one, or one you created in a GUI"
        newOne = SpamCan()
        newOne.contents = contents
        newOne.type = type
        newOne.qty = qty
        self.cans.append(newOne) 			
        
    def DeleteCan(self,index):
        del self.cans[index]
        
    def GetCanAt(self, index):
        return win32com.server.util.wrap(self.cans[index])
        
    def GetDescriptionList(self):
            # for a list view
        return map(lambda x:x.GetDescription(), self.cans)
        
    def GetCrosstab(self):
            # some example data for yout to play with
        return [['e.g.','cans','bottles','bags'],
                ['spam',5,4,3],
                ['spinach',0,1,2],
                ['beer',12,4,2]]
        
class SpamCan:
    "just a simple 'packet' of data to play with"
    _public_methods_ = ['GetDescription']
    _public_attrs_ = ['contents','type','qty']
    
    def __init__(self):
        self.contents = 'spam'
        self.type = 'can'
        self.qty = 3
        
    def GetDescription(self):
        return '%d %ss of %s' %(self.qty, self.type, self.contents)
        
# each class needs a class ID.  You can get these by typing:
#	import pythoncom
#	g = pythoncom.CreateGuid()
#	print g
# in the interactive window, then pasting it into your code.
        
def RegisterSpam():
    # register them both - this must be called once for the
    # demo to work
    import win32com.server.register
    win32com.server.register.UseCommandLine(SpamServer)
    
def UnRegisterSpam():
    """ Unregister each server - use this before 
    deleting the server to keep your registr tidy"""
    
    print "Unregistering COM server..."
    from win32com.server.register import UnregisterServer
    UnregisterServer(SpamServer._reg_clsid_ ,
                   "Python.SpamServerGPO")
    print "SpamServer Class unregistered."

    # and finally some test code
    
def TestDirect():
    # does some stuff to verify the class works IN PYTHON
    s = SpamServer()
    print 'Created spam server'
    s.AddCan('Cheese', 'Sandwich', 2)
    print 'added 2 cheese sandwiches'
    s.AddCan('Chimay Rouge', 'Bottle', 6)
    print 'added some strong beer'
    print 'contains %d items:' % (s.GetCanCount())
    for item in s.GetDescriptionList():
        print '\t' + item
    s.DeleteCan(0)
    print 'ditched the beer, we are at work here!'
    print 'Tests passed'
    
def TestCOM():
    try:
        SpamSrv = win32com.client.dynamic.Dispatch("Python.SpamServerGPO")
        print 'Python.SpamServerGPO class created from COM'
    except:
        print "**** - The Python.SpamServer test server is not available"
        return
    print 'cans:', SpamSrv.GetCanCount()
    print 'Adding a cheese sandwich', SpamSrv.AddCan('Cheese', 'Sandwich', 2)
    print 'Adding some beer:', SpamSrv.AddCan('Chimay Rouge', 'Bottle', 6)
    print 'cans: ', SpamSrv.GetCanCount()
    descr = SpamSrv.GetDescriptionList()
    for item in descr:
        print item
    print "The Python.Spamserver COM Server worked OK."
    
    
if __name__=='__main__':
    """If they ran the whole script, register classes.
    Options also given to unregister and to test"""
    import sys
    if "/unreg" in sys.argv:
        UnRegisterSpam()
    elif "/test" in sys.argv:
        print "doing direct tests..."
        TestDirect()
        print "testing COM"
        TestCOM()
    else:
        RegisterSpam()


From gcollins99 at my-deja.com  Fri Jul  2 11:07:03 1999
From: gcollins99 at my-deja.com (gcollins99 at my-deja.com)
Date: Fri, 02 Jul 1999 15:07:03 GMT
Subject: RFC822 & Text of Mail Message
Message-ID: <7likid$e23$1@nnrp1.deja.com>

Greetings!

  Here is what I want to do but cannot find out how. I am using RFC822
to parse out the headers and store them somewhere which works just fine.
But I am also in need of getting the text of the incoming message. So
what module would I use to also capture the message text. I see no way
of doing this via rfc822 module since it's not designed to do such
things. WOuld I use mimetools? And if so, how?

Thanks!

-Glen Collins



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From greg.ewing at compaq.com  Sun Jul 25 17:24:42 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Mon, 26 Jul 1999 09:24:42 +1200
Subject: Cyclops 0.9.4
References: <000101bed4dc$3e51e280$392d2399@tim>
Message-ID: <379B809A.CB35D2D8@compaq.com>

Tim Peters wrote:
> 
> Nobody can sift thru hundreds of thousands of live objects by eye, which
> happens routinely in large long-running apps.

I can look for certain kinds of object, though. If
I open ten BlargWindows and then close them, I know
there shouldn't be any BlargWindow instances left.
If there are, I can take a closer look at them to
find out what what's happening.

> Well, 100,000 objects can often be reached in 500,000 ways via one direct
> link -- & it gets messier the longer the path length you consider.  You have
> to reduce the set of objects that are potentially interesting; Cyclops
> requires you to identify them in advance; I'm not sure how it could be
> easier to identify them later.

I'm not asking to be told about every possible way
of reaching every live object, obviously! What I mean
is, once I've found (by whatever means) a live object that
I know should be dead, I should be able to say "show me
a path from the root set to this object". I don't need all
paths, only one of them -- somewhere along that path there
will be a link that shouldn't be there, and it should be
fairly easy to spot.

That's what I meant when I said that cycles aren't what
I want to know about in that situation.

> All objects ever allocated, or all objects whose refcounts haven't yet
> fallen to 0?

The latter.

> Since
> "even ints are boxed" in Python, that's a major memory hit.

Which is why I'm happy for it to be a debugging option,
maybe requiring a different interpreter. Although if M&S
is ever added, we're going to have to live with this
overhead all the time -- or find a smarter way to handle
it.

> Let's say you have everything you
> want, and even more:  for every object ever allocated, Cyclops2 can give you
> a timestamped history of every change ever made to it, every path from which
> it could be reached at any given time, and similarly for every object
> reachable from it.

Tim, this is expandio ad absurdum! I never asked for all that
information. All I said whas that it would be handy to have a way
to discover the existence of dead objects taking up room in
my heap. I never expected the idea to be this controversial...

> I think you should try using it once <wink>.

You're right. I'll shut up now.

Greg



From fredrik at pythonware.com  Thu Jul 29 14:57:53 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 29 Jul 1999 18:57:53 GMT
Subject: shelve portability problem
References: <eZUl3.1080$wm4.135734@news.uswest.net> <7nirjd$n52$1@m2.c2.telstra-mm.net.au> <7nq6a7$vka$1@nnrp1.deja.com>
Message-ID: <001f01bed9f4$4453e8f0$f29b12c2@secret.pythonware.com>

<nadim_ghaznavi at my-deja.com> wrote:
> I'm having a problem moving a Python shelve ODB from Solaris to Linux.
> Here's some sample code to demo the problem:
> ...
> I know I can work around this by using pickle and a specific DB module,
> but does anyone know how I can get it working using the shelve module?

no, but let's find out.  hang on...

...

alright.  this might work:

import shelve
import gdbm

def myopen(filename, flag="c"):
    return Shelf(gdbm.open(filename, flag))

db = myopen("odbfile")

(untested)

</F>




From nick at #spam_me_not#videosystem.co.uk  Wed Jul 28 09:41:42 1999
From: nick at #spam_me_not#videosystem.co.uk (Nick Trout)
Date: Wed, 28 Jul 1999 14:41:42 +0100
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com>
Message-ID: <379f06e9@news.xtml.co.uk>

>> Thus I've had great deal of interest in getting the speed up; however,
>> speeding up dynamic interpreted languages is nontrivial, especially
>> languages as dynamic as Python.

Are there any pieces of test code anywhere which could be used as a good
test for Pythons speed (encompasing all of its language features) so that if
anyone tinkered with the bytecode execution unit they could measure the
speed improvement. It would be useful if it also tested correctness to
ensure no mistakes?!

N







From dan at control.com  Tue Jul 27 10:44:14 1999
From: dan at control.com (Dan L. Pierson)
Date: Tue, 27 Jul 1999 14:44:14 GMT
Subject: Python and SlickEdit
References: <379CA3F9.A746217E@iRenaissance.com> <7njrob$oq6$1@mailusr.wdf.sap-ag.de>
Message-ID: <379dc2b3.584259984@news.randori.com>

daniel.dittmar at sap-ag.de (Daniel Dittmar) wrote:

> Rowland Smith <rowland at iRenaissance.com> wrote:
> 
> >Does anyone know if there are any Python plug-ins for Visual SlickEdit
> >and where they are?
> 
> There's a Python module in the contributed area
> (ftp://ftp.slickedit.com/pub/usermacs/python.zip). 
> 
> It does code coloring, but no procedure outlining (at least not
> working). 

Wonder if that's the one I've submitted to them at least twice.  If
so, you're right that the procedure outlining is only partly working.
They never responded to my requests for help on it and I'm not a
wizard with their extension environment (mainly use Emacs, have for
many years, fire up Slick Edit every few months for some special use).
Slick-C isn't that bad as an extension language (IMO, not that good
either) but you really need to learn all of their internal extension
conventions too.

My hazy memory says that part of the problem was that all existing
examples of working procedure outlining and full tagging are
implemented as C dlls.  There was on critical function that the
comments in the sample file (i.e. the official documentation) claimed
would do what was needed but I couldn't find a Slick-C implementation
of that for any of their modes and mine didn't work for the usual
mysterious reasons.  This was what I was asking for help on...

BTW: I looked at their "Context Tagging" implementation for C++.
While its undocumented, it looks like someone *could* do something
similar for Python.  It all *seems* to be in Slick-C.  Be a lot of
work and you might run into something that had to be done in C/C++.

Dan Pierson, Control Technology Corporation
dan at control.com



From fdrake at cnri.reston.va.us  Mon Jul 12 11:40:45 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Mon, 12 Jul 1999 15:40:45 GMT
Subject: Updated GNU info documentation available
Message-ID: <14218.3197.296612.826083@weyr.cnri.reston.va.us>

  Thanks to Milan Zamazal, the GNU info format of the documentation
has been updated to documentation release 1.5.2p1.
  An archive is available at ftp://ftp.python.org/pub/python/doc/.
  Thanks, Milan!


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From fredrik at pythonware.com  Wed Jul 28 16:34:53 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Jul 1999 20:34:53 GMT
Subject: Building a shared libpython
References: <Pine.GSO.4.03.9907281701150.15012-100000@spica>
Message-ID: <001001bed938$b6b34330$f29b12c2@secret.pythonware.com>

Carlos A M dos Santos <nospamplease at inf.ufrgs.br> wrote:
> I'm installing Python 1.5.2 and noted that there is no configure option to
> buld a shared libpython.  How could I do it?

www.python.org => search => newsgroup search for
"shared libpython" brings up lots of hits.  start with
these:

    http://www.deja.com/=dnc/getdoc.xp?AN=488431463
    http://www.deja.com/=dnc/getdoc.xp?AN=398994871

</F>




From tbryan at arlut.utexas.edu  Tue Jul 27 04:04:38 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Tue, 27 Jul 1999 03:04:38 -0500
Subject: glob and directory
References: <379D57BE.53DA9B37@pk.highway.ne.jp>
Message-ID: <379D6816.9BD39ED1@arlut.utexas.edu>

too at pk.highway.ne.jp wrote:

> is there any way to determine
> if the given name is directory or not ?

os.path.isdir

If you're just globbing '*', then you could also use

os.listdir('.')

It'll save you from having to type "import glob". :)

-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From avv at quasar.ipa.nw.ru  Sat Jul  3 16:12:24 1999
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Sat, 03 Jul 1999 13:12:24 -0700
Subject: Curses, Ncurses, Slang, etc?
References: <19990703140217.D14374@moebius.dartmouth.edu>
Message-ID: <377E6EA7.DAD8657@quasar.ipa.nw.ru>

Eric Kidd wrote:

> Neither the ncurses module or the pwin module work out of the box on my
> Linux system. Are there any well-supported Python modules for
> curses/ncurses/slang which do work out of the box?

SLang itself is supported, and even a higher-level set of widgets, written
purely in Python, is underway, but slowly :-(.
You may pick up a binary for Red Hat 6.0 at:

ftp://starship.python.net/pub/crew/avv/pyslang/work/slang.so

or recompile it out-of-box from
ftp://starship.python.net/pub/crew/avv/pyslang/pyslang-0.2.0.tgz.

As usual, I remind that Red Hat, as well as other distributions, comes with
the Python wrapper for the 'snack' library, which provides higher-level
widgets than bare pyslang.

The documentation on pyslang is mostly one for the SLang itself, with the
modifications mentioned in my readme's.
Examples may give more information, as usual.

The principal advantage of pyslang I'm aware of is that you may use the same
script under win32 without any changes.

Next bug-fix release of pyslang is very close to completion.

Alexander





From mumac at wiu.edu  Tue Jul 20 10:09:55 1999
From: mumac at wiu.edu (Mark Clark)
Date: Tue, 20 Jul 1999 09:09:55 -0500
Subject: How to send text to a Windows program
References: <37932FF9.47E8FA2D@wiu.edu> <AkRk3.3777$VL2.410197@news.direcpc.com>
Message-ID: <37948333.77223E57@wiu.edu>

Thank you!

Emile van Sebille wrote:
> 
> Take a look at
> http://x36.deja.com/getdoc.xp?AN=480302853&search=thread&CONTEXT=9324382
> 39.909574156&HIT_CONTEXT=932438107.906297455&HIT_NUM=1&hitnum=5
> 
> (found on dejanews by searching for 'python sendkeys' in all.
> 
> --
> 
> Emile van Sebille
> emile at fenx.com
> -------------------
> 
> Mark Clark <mumac at wiu.edu> wrote in message
> news:37932FF9.47E8FA2D at wiu.edu...
> > I am starting a Windows program with ShellExecute.  I would like to
> > send some text to this program, similar to the way you would if you
> used
> > "SendKeys" in other programming languages.  Can someone give me an
> > example of how this would be done?
> > Thanks for your time.



From tim_one at email.msn.com  Tue Jul 13 00:03:43 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 13 Jul 1999 04:03:43 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
In-Reply-To: <14218.7197.219561.903989@weyr.cnri.reston.va.us>
References: <14218.7197.219561.903989@weyr.cnri.reston.va.us>
Message-ID: <000d01becce4$b296f540$31a02299@tim>

[Fred Drake, under process table duress]

>   Well, after long consideration and a quick check the Guido, it turns 
> out that Guido thinks they should be moved to the "Files and
> Directories" section.  Since I disagree, I've contacted several local
> bus companies to help in steering the course of future Python
> development.

Fred, a good whore goes where he's kicked.  Welcome to the stable <wink>.

a-python-pimp-won't-even-leave-you-a-dime-to-call-your-momma-ly y'rs  - tim





From wolford at enews.nrl.navy.mil  Thu Jul  8 16:51:39 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Thu, 08 Jul 1999 16:51:39 -0400
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com>
Message-ID: <37850F5B.F80CB0A7@enews.nrl.navy.mil>

I've finally got it. The answer below is basically it but it's even
trickier than that. You can't link an application with python.lib and
msvcrtd.lib; you *must* link with msvcrt.lib. Very subtle difference but
very important because when python.lib was linked, it was linked with
msvcrt.lib and the application can't then bring in msvcrtd.lib. I do think
this is a problem because python15_d.lib is provided but no python15_d.dll
is. I would think you'd provide (at the very least) python.lib linked with
msvcrt.lib, and python_d.lib linked with msvcrtd.lib (and the dlls), or
(to be really nice :) all four combinations. Don't get me wrong; I think
it's a great service to provide the pre-compiled package. I just want to
see it be the best it can be, so others can avoid this juicy trap. If
anyone wants the sample code I used (and compile/link instructions), just
email me.

Thanks alot for the help,
Scott

Gordon McMillan wrote:

> Probably the most common mistake of Windows embedders is to link to
> the wrong C runtime lib. Python uses Multithreaded DLL.
>
> But you might have done something more creative. Telling us how and
> when it bombs complete with messages and as small a code sample as
> possible will get you more and better answers.
>
> - Gordon




From cgw at fnal.gov  Tue Jul 13 18:29:31 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Tue, 13 Jul 1999 22:29:31 GMT
Subject: problem with SocketServer -- address already in use
In-Reply-To: <931899928snz@vision25.demon.co.uk>
References: <931899928snz@vision25.demon.co.uk>
Message-ID: <14219.48587.928540.252071@buffalo.fnal.gov>

Phil Hunt writes:
 > I'm trying to use SocketServer.py to write a TCP server. I ran
 > my program with ``python serv.py'', which went into an infinite
 > loop (which is what it should do). I exited my program with ^Z,
 > and tried to start it up again. I got an error message:
 > 
 > socket.error: (98, 'Address already in use')
 > 
 > which I presume is because something thinks the old invokation of
 > my program is still using port 7070. 

Did you actually kill the process after doing ^Z?  ^Z just suspends a process.

 > How do I clear the port, before I start up my server?

You don't.  The kernel clears the port, after a certain period of time 
has elapsed.  However, if the old process really is gone, and you want
to bind to the same address, without waiting for the kernel, use the 
SO_REUSEADDR flag to setsockopt:

self.socket.setsockopt(SOL_SOCKET,SO_REUSEADDR,1)

This must be done after a socket is created and before you call listen 
or bind.

I don't see an easy way to do this within the framework of SocketServer.py

It might be nice if there were an optional flag in SocketServer to
allow the user to specify this behaviour.




From bucodi at worldnet.fr  Fri Jul  2 06:21:02 1999
From: bucodi at worldnet.fr (pserrand)
Date: Fri, 02 Jul 1999 12:21:02 +0200
Subject: newbie
References: <7li14e$mts$1@Nntp1.mcs.net>
Message-ID: <377C928E.FEDA8941@worldnet.fr>

cyberescue wrote:

> I am new to programming. Where can I get a copy of Python?
>
> mailto: ashfordv at usa.net

www.python.org




From Jean-Claude.Wippler at p98.f112.n480.z2.fidonet.org  Fri Jul  2 14:41:25 1999
From: Jean-Claude.Wippler at p98.f112.n480.z2.fidonet.org (Jean-Claude Wippler)
Date: Fri, 02 Jul 1999 19:41:25 +0100
Subject: Why does shelve make such large files?
Message-ID: <00001671@bossar.com.pl>

From: Jean-Claude Wippler <jcw at equi4.com>

Johan Wouters wrote:
> 
> Gerrit Holl wrote:
> 
> > Ah, I understand.
> > So pickle is useful for very small datases, but when they're really
> > huge, one should use shelve. Isn't it?
> 
> Pickle will serialise your data into a certain format. This way you
> can "store" objects like lists as a whole. The serialising has the
> benefit that you can eg. put your object (be it a list or something
> else) through a pipe or socket and depickle it at the other side. This
> way you can have python processes that exchange data in a sort of
> native python format.
> 
> Shelve gives you alot more than just storing the data as a long
> sequence. Imagine the extra space being the administration that is
> needed for all the extra functionality (like tagging what data items
> are valid etc)
> 
> The overhead induced by a database management system can be enormous
> compared to the bare data you want to access, but this is mostly a
> space/time tradeoff: the more space you use the faster you can do
> things like searching and sorting etc etc.

Ahem, with all due respect to everyone... this is humbug.

Pickle serializes your data in one sweep to/from disk.  It's compact.

Shelve stores serialized pieces in a keyed-access form, and uses file
space allocation to support add/modify/delete.  It starts out with more
empty space, but as the amount of stored data grows, that space does not
expand linearly.  Consider shelves to have half their file space being
used, on *average* (plus or minus a factor 2 perhaps).

But the last claim is horrendously misleading, I'm afraid: data storage
is not a space/time tradeoff.  It's about throughput (I/O bottlenecks)
and overhead in managing the supported data and indexing schemes.  There
are order-of-magnitude performance differences in how several solutions
work, because of this.

Shudder.  The notion that a large database package, or a large datafile,
is faster, is so far from reality that it has to be corrected, even in
this Python-oriented newsgroup.  My apologies for the S/N ratio drop.

-- Jean-Claude

P.S.  Geloof niet alles wat je leest, Gerrit.



From tismer at appliedbiometrics.com  Tue Jul 13 04:21:13 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 13 Jul 1999 08:21:13 GMT
Subject: ANN: Stackless Python 0.3
References: <378A40AD.B4FDB79F@appliedbiometrics.com> <m37lo5kpd0.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <378AF6F9.5E25E045@appliedbiometrics.com>


Michael Hudson wrote:
> 
> Christian Tismer <tismer at appliedbiometrics.com> writes:
> 
> > ANNOUNCING:
> >
> >                Stackless Python 0.3
> >            A Python Implementation Which
> >              Does Not Use The C Stack
> >
> > What is it?
> > A plugin-replacement for core Python.
> > It should run any program which runs under Python 1.5.2 .
> > But it does not need space on the C stack.
> >
> 
> Very nice, but

Ouchh!

> >>> apply(l.sort,())
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> SystemError: error return without exception set
> >>> l
> [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
> 
> apply isn't totally hosed; it works for Python functions (I think) but
> not for builtins.

Does apply_orig behave well?

> This causes lots of make test to fail, though not test_builtin,
> helpfully.
> 
> I've had a poke through the source, but don't have the time just now
> to delve into its inner workings. I'll trace it with gdb sometime.

Thank you, I will look into it soon.
I should run all the tests, and not only pystone before I deliver.

> PS: To get it to compile I had to remove "static" from in front of
>     PyFrame_FiniChainFunc * PyFrame_FiniChain = NULL;
> in Objects/frameobject.c.

Gaah! Again! I also should not use MS VC++ but GCC I think.

thanks a lot - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From gregm at iname.com  Mon Jul  5 00:22:52 1999
From: gregm at iname.com (Greg McFarlane)
Date: Mon, 5 Jul 1999 04:22:52 GMT
Subject: Upgrade or not ??
In-Reply-To: <377F91D9.D3A1E147@worldnet.fr>; from pierrette on 04 Jul 1999 at 06:54:50PM
References: <377E0A3B.CA3B258C@worldnet.fr> <19990704224912.16776@nms.otc.com.au> <377F91D9.D3A1E147@worldnet.fr>
Message-ID: <19990705142252.44292@nms.otc.telstra.com.au>

Since Tkinter is just a wrapper around Tk, I find the best
documentation to be the Tk manual pages, which comes with the
Tk source distribution.

On 4 Jul, pierrette wrote:
> Thanks, exactly what i needed.
> Is there a way to find out what events a widget support ?
> I browsed documentation on python.org but didn't find anything on FocusIn en
> FocusOut. I there somewhere documentation on that ??
> 
> Pierrette
> 
> Greg McFarlane wrote:
> 
> > Depending on exactly what you want to do, it may be possible in
> > Tkinter.  You can use something like this to find out when a widget
> > gets and loses focus:
> >
> >     import Tkinter
> >
> >     def gotFocus(event):
> >         print 'Focus in'
> >     def lostFocus(event):
> >         print 'Focus out'
> >
> >     root = Tkinter.Tk()
> >
> >     e = Tkinter.Entry()
> >     e.bind('<FocusIn>', gotFocus)
> >     e.bind('<FocusOut>', lostFocus)
> >     e.pack()
> >
> >     root.mainloop()
> >
> > Also, you can use this to set the focus to a widget.
> >
> >     e.focus_set()
> >
> > On 3 Jul, pierrette wrote:
> > > First a general question :
> > > Actually we're working with Python 1.5.1 and we're satisfied with it.
> > > Should we upgrade to 1.5.2 and if yes WHY?
> > >
> > > No a Tkinter guestion :
> > > We developped a prototype of an application with python an tkinter, now
> > > we have to
> > > decide if we're gonna write the real app in python or something else !
> > > We now allready that in principle we could do it with python, the thing
> > > we're missing
> > > in tk is a 'on get focus' ,a 'on lost focus' event and a 'set focus to'
> > > command
> > > Pe: on lost focus of inputfield 1:
> > >             if inputfield1 = A:
> > >                 set focus to field2
> > >             elif inputfield1 = B:
> > >                 set focus to field 3
> > >
> > > If those things are not available in tkinter what alternatives do we
> > > have ?
> > > - Changing the GUI (would be a shame but feasable), knowing that is has
> > > to be multi-platform WIN-LINUX? If so to what ?
> > > - Not write in python ? In this case we lose the work done for the
> > > prototype
> > >
> > > Thanks for comments and remarks
> > >
> > > Pierrette Serrand
> > > pserrand at worldnet.fr
> > >
> > >
> > >
> > >
> > >
> >
> > --
> > Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
> > Today's forecast:  Sunny, with occasional cloudy periods and a chance
> >                    of precipitation in some areas.
> 
> 
> 

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From aahz at netcom.com  Thu Jul 22 00:37:13 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 22 Jul 1999 04:37:13 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
References: <001101bed3f2$5e8874a0$642d2399@tim>
Message-ID: <7n675p$qsk@dfw-ixnews7.ix.netcom.com>

In article <001101bed3f2$5e8874a0$642d2399 at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>
>That's fine for a one-man show.  "Maintainability" in my life cuts across
>programmers, and it's Not Good for each person to reinvent their own basic
>machinery -- even to have the ability to do so.  One reason I like Python is
>that it *doesn't* let the jerk down the hall invent their own control
>structures <0.5 wink>.

Yup.  Today I had to lay down the law to our programming team: use of
the construct "from foo import *" is now expressly, explicitly, and
completely forbidden.  "from foo import bar" is deprecated; "from foo
import foo" is acceptable; and "import foo" is strongly preferred.

I'd be in favor of getting "from foo import *" entirely out of the
language, but I suspect it's just too convenient for short hacks.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From bruce_dodson at bigfoot.com  Fri Jul 30 14:07:37 1999
From: bruce_dodson at bigfoot.com (Bruce Dodson)
Date: Fri, 30 Jul 1999 15:07:37 -0300
Subject: Python as a parser (newbie)
References: <37A1D114.705957B9@yahoo.com>
Message-ID: <pVlo3.42911$jl.30210756@newscontent-01.sprint.ca>

>From the sound of this, you don't really want to create a new parser.  You
just want to embed the Python scripting language in your application, and
export functionality from your application to the embedded python
interpreter.

I have done this with ArcView GIS.  It was quite painless.  Basically, just
look at the docs: there is a reference on the API for embedding, and a
tutorial on extending Python through modules written in C.  It is really
just a matter of of smearing those two ideas together.  You application
initializes the Python interpreter, then publishes some of its internal
routines as a Python module, and then whenever your application calls python
code, that Python code will be able to call back to your application through
that module.

It is one of Python's strengths that this is so easy.




From gmcm at hypernet.com  Mon Jul 19 09:23:36 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 19 Jul 1999 08:23:36 -0500
Subject: Python/Gadfly and SQL wildcards
In-Reply-To: <001d01bed1bb$0cecfd20$a95d38d4@default>
Message-ID: <1279735665-4809662@hypernet.com>

Colm Rafferty writes:

> Does anyone know how to implement the wildcard in gadfly SQL?
> e.g.
> Select * from table where fieldname = 'A*'
> or
> Select * from table where fieldname like 'A*'
> 
> * does not work!
> Though it does in other SQL languages

In ANSI SQL, that would be "fieldname like 'A%'"

Last I used Gadfly, it didn't support 'like', so you're stuck with

..fieldname >= 'A' and fieldname < 'B'

or the same using "between".

- Gordon



From smalleys at gte.net  Fri Jul 23 00:26:48 1999
From: smalleys at gte.net (Sue)
Date: Fri, 23 Jul 1999 04:26:48 GMT
Subject: Repost: Re: Scientific Plotting?
References: <379392E7.EB0F29E7@ssec.wisc.edu> <7n5fhn$9e5$1@nnrp1.deja.com> <7n76dn$rbc$1@nnrp1.deja.com> <x889088o5oc.fsf@rpppc2.hns.com>
Message-ID: <3797EF2E.A2D00D29@gte.net>

I've been using Grace to plot python-generated data for
three months.  No problems, does PS, GIF, Type 1 fonts,
NetCDF, etc.

Very nice package; builds on Linux, Sun and Alpha Unixes,
and a simple-to-install NT binary.

Sue.



From paul at digicool.com  Sat Jul 10 10:14:21 1999
From: paul at digicool.com (Paul Everitt)
Date: Sat, 10 Jul 1999 10:14:21 -0400
Subject: Python Code Repository?
References: <378645AA.5430A026@msfc.nasa.gov> <19990709163127.A1964@toast.internal> <slrn7od2d6.dl.amk@207-172-57-127.s127.tnt2.ann.va.dialup.rcn.com>
Message-ID: <3787553D.A4C1DB10@digicool.com>

"A.M. Kuchling" wrote:
> Michel Pelletier (the one at Digital Creations) was working on a
> ZTrove at one point, I believe.  (For information on the whole idea of
> Trove, see http://www.tuxedo.org/~esr/trove/).  I don't know what's
> happened to ZTrove; anyone from DC care to comment?

OK, lonnnggg post ahead.

Zope Product Library

  With the progress made on Zope2 and the Zope Portal Toolkit (PTK),
  we've made enough progress to begin construction of a "Freshmeat"
  package library for Zope software.  (Note: the jargon of "product",
  "package", "release", and "version" is awfully contentious.)

  Since this is effectively the same ground being covered here, I'd
  like to take a moment to describe what we are doing and see if folks
  here on the Python list find it interesting.  I'll also cross-post
  to the Zope list.  The desired outcome is for our experience and
  software to provide a boost to a real effort.  (Historical note: I
  was the shepherd for the original ill-fated "Locator-SIG".)

  Goals

    o Fast location of desired software

    o Keep central admin requirements *low* by distributing control

    o Allow package authors to administer their entries

    o Adhere to appropriate standards and initiatives

  Zope Product Library

    In our system, Authors create Products and provide Releases of
    those Products.  A Release is a version of a Product for a
    platform (if binary) or a source version in some packaging scheme
    (such as RPM).  Thus, a Release is the bits and a Product is the
    container for all information about those bits.

    The things capitalized in the above are "ZClasses".  For you
    non-Zopistas, a ZClass is like a Python class that you build
    through the web and store in the Zope object database.  It is an
    instance that can make instances.  It also provides a simple way
    to store "your kinds of things" in the Zope object database.

    The private, under-construction new zope.org site has a membership
    system where we give people a Zope home folder to stash their
    stuff.  They pick from a list of things they can add, such as
    uploaded images, HTML documents, etc.

    People with higher privilege -- such as an Author -- can also
    choose "Product" from the list of things they can add to their
    folder.  When they do so, the Product gets added to the site
    catalog.

    A Product is like a folder -- you can stick a home page, other
    docs, PDF files, etc. in it.  More importantly, inside a Product
    you can add a Release.  This will provide the kind of thing seen
    in Freshmeat.

    Thanks to acquisition, much of the information about a Release
    (author name and email, description, etc. ) comes from the Product
    and doesn't have to be repeated.

    The ZCatalog is awfully useful for this kind of system.  First, it
    is (almost) always up-to-date.  Whenever you add a Product or
    Release, or edit one, or delete one, the Zope transaction also
    updates the catalog.  Next, ZCatalog's indexing engine does both
    fielded and full-text searches.

    Another nice feature is that objects control how they are indexed,
    as the Catalog is calling methods on the site content to get at
    indexed content.  For instance, if you decided you'd like to untar
    the release and index the docstrings, go right ahead.  As long as
    you can implement a Python method to do it.  If you want the
    full-text to actually come from four properties about a Product,
    piece of cake.  If you want to create a Remote Product listing,
    where the content is retrieved by urllib for indexing, you can
    quickly do that as well.

    Once the content is in the ZCatalog, well, Bob's your uncle.  You
    can create multiple interactive search forms.  You can query the
    catalog to insert a tree on a page for browsing the contents.  You
    can create multiple presentations (by date like Slashdot and
    Freshmeat, by category like ODP, by author, by platform, etc.)

    More interestingly, you can get the Catalog to dish up requests as
    RDF.  Imagine browsing the package library using the tree control
    in Mozilla, which gets fed by RDF.  Or, imagine a command-line
    Python script that acts like rpmfind.

  What About Trove and ZTrove?

    This isn't Trove.  This *could* be Trove, but we are focused on
    getting something in place fast on zope.org.  We are solving our
    own problems first.

    Once we have made all our mistakes and learned from them, it would
    definately be time to talk about Trove and a Python package
    locator.

  ZCatalog

    The main advance that allows a high-end package system is the
    "ZCatalog",
    http://www.zope.org/Download/ZCatalog/ZCatalogTutorial.  This
    tutorial provides a good rundown of all the neat things about the
    catalog.

    Does the Catalog scale?  We think so, but that's a hard question.
    We have a *lot* of painful, painful experience in this area, based
    on a big consulting job we did a few years ago, and those lessons
    are reflected in the software.  Our indexing works very hard at
    keeping memory footprint moderately in check and dishing up
    incredibly fast searches.

    As a datapoint, we took a 15 Mb chunk of the rufus RPM database
    and loaded it into the indexer.  A combination fielded and
    full-text search took 0.004 seconds on a PII-350 (inside the
    indexer, formatting the response and returning it took a lot
    longer).

  CPAN and rufus.w3.org

    It appears to me that CPAN does the following functions:

      o Software repository for Perl code

      o Mirrored worldwide

      o Routes requests to a correct, available archive

    A related system that I know more about is rufus.w3.org's RPM
    database.  It adds a particularly useful feature that I *think*
    CPAN has: a command-line tool to query repositories and fetch
    packages for installation.  I unfortunately don't know much about
    the workflow of how package listings get into the repository, then
    how they get updated.

    I'll observe that the searchability for both of these systems is
    awfully weak.  For us in Zopeland, this is a major opportunity.

  The Mythical Python Locator

    So, what exactly *is* being looked for here?  A simple package
    registry?  Something integrated with dist-utils?  How much should
    this system scale?  What standards should it piggy-back?  Is it
    centralized or decentralized, both in implementation and
    administration?  What jargon do we agree to and what is the
    metadata we will choose?

    And most important, is anybody willing to actually work on it?

    In closing, though I'm skeptical that this will move from "brief
    flurry of activity" into accomplished project, I think this is an
    area where a definate disadvantage can be turned into a clear
    advantage.  The goal shouldn't be just to catch up.  The goal
    should be go farther, to show off and taunt, then to rub the
    others' noses in it. :^)



From olivier.deckmyn at mail.dotcom.fr  Mon Jul 19 09:35:35 1999
From: olivier.deckmyn at mail.dotcom.fr (Olivier Deckmyn)
Date: Mon, 19 Jul 1999 15:35:35 +0200
Subject: [Q] Type checking...go further...
Message-ID: <7mv9n8$md$1@feed.teaser.fr>

Thank you all for your answers !
I finally use isinstance in the methods of my classes...

But to go further, is there a way in python to have typed methods parameters
?
aka :

class MyClass:
    def myMethod(string myString):
        print myString

?

Thanx again all for your help !

Olivier,
happy python newbie.





From sdossey at dolphin.openprojects.net  Sat Jul 10 19:13:02 1999
From: sdossey at dolphin.openprojects.net (Scott Dossey)
Date: Sat, 10 Jul 1999 23:13:02 GMT
Subject: Dreams.
Message-ID: <slrn7ofkrt.55r.sdossey@dolphin.openprojects.net>

Okay, I am new to the newsgroup, but I thought I'd post something
interesting that I've just started working on, and get some feedback.  I've
been toying around with the idea of implementing "dreams" into the Python
source code.  

What are dreams, you say?  Look at the following python code fragment for an
idea, first of all, of how they are put into code:

#START PSEUDOCODE FRAGMENT
import sys
import dreams

aDream=dreams.dream()
try:
	#Insert code here.
except: 
	aDream.wakeup()
	sys.exit(1)

aDream.realize()
#END PSEUDOCODE FRAGMENT

The code, at the location denoted by "#Insert code here." would run, but if
an exception was thrown, it will have been as if the code had "happened in a
dream".  No variables will have been changed/deleted.  Nothing will have
changed from before the "dream" in terms of Python's variable state.
Ideally, (but this is much harder), all file changes, all things that are
"possible" to return to their original state, would be returned as well.
(You can't take back network traffic).  

Dream.wakeup() would restore to the state before the dream, and continue
execution from that line forward.  Dream.realize() would commit to the
dream, stop saving change-state, and delete the change state it had already
recorded. 

Before you start telling me about the numerous interaction problems this
will cause, and how impossible it will be to implement, think of what it
does to certain types of error handling!  You don't have to back out changes
when your change to a transactional database goes awry midway through the
change.  Even though functionality would be limited, (not all things could
or would be backed out), any automatic handling (even partial) of these
types of problems could be a huge blessing for certain types of programming.

Here's how I was thinking about implementing them (If you are not familiar
with python's C implementation this may be unfathomable to you):

The easiest way to do it that I can see is the following:
	      
	Step One:  Put a wrapper around each variable and function pointer
		in any PyObject and it's contained structures (IE,
		PyTypeObject), so that any reference to any object can be 
		"caught".  This would involve tweaks all over the python
		codebase to deal with the new way of accessing variables.
		I'll probably do this by changing the name and method of 
		access for everything inside a PyObject and letting the
		compiler find all the millions of references, then fixing these
		up to use the new wrappers.  It's too bad there are not access
		methods for all these things already, rather than being 
		indexed directly.

        Step Two:  Call pickling code inside those wrappers to save the
		state of the object before and after the calls.  Write the
		diffs out to the dream code so it can store it in a "dreamlog".
	        Deletions of objects will similarly "pickle" the objects and
 		store them in the dreamlog.  If a pickle of an object before
		it is referenced doesn't work, then
		"abort" the dream by restoring the state and
		throwing a "DreamException".
		
	Step Three:  Fix up more objects so they can be pickled, so that
		dreams are mode useful.  IE, make files pickleable, modules
		pickleable, text screen state and TK graphical screen state
		pickleable and restoreable, etc. 

	Step Four:  Major debugging, testing, and module usage documentation.
	
I'd probably include dreaming as a compile time option of python, since
support for it will cause some pretty significant code slowdown of python,
even when not "dreaming".

However, there are a couple of major problems with my approach:
1.	They break every C module known to man in a way that's not too hard
to fix, but will have to be done for every C module that is going to support
dreams.

2.  	Python will be slowed down.  Every call/variable reference 
	will have a little more overhead when not dreaming, A LOT when 
	dreaming.

3.	It's very invasive.

I'll take feedback in terms of ideas, problems, feasibility statements,
flames,  whatever.  Most of all, I want help!  I think this would add a very
interesting and useful feature to Python (as perhaps, a compile time option).
Guido, I know you're out there, your feedback would be most appreciated,
along with the feedback of any other Python maintainers.

-Scott Dossey
sdossey at armored.net, sdossey at openprojects.net




From cgw at fnal.gov  Mon Jul 26 17:47:15 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Mon, 26 Jul 1999 21:47:15 GMT
Subject: filter() + simple program not outputting
In-Reply-To: <379CC972.750BD42D@netcard.com>
References: <379CC972.750BD42D@netcard.com>
Message-ID: <14236.55139.408761.661465@buffalo.fnal.gov>

Joyce Stack writes:
 > Hello,
 > 
 > New to python but not to programming...can someone tell me why this
 > simple program is not outputting when i run from the BASH prompt....any
 > pointers would be helpful...thanking you in advance....
 > 
 > Joyce
 > 
 > 
 > #!/usr/bin/env python
 > 
 > def f(x):
 >         return x%2 != 0 and x%3 != 0
 >         filter(f, range(2, 25))
 >         f( )
 > 
 > 

Well, there are several reasons!  One is that you define a function
but never call it. 

You probably meant to say:

 > def f(x):
 >         return x%2 != 0 and x%3 != 0
 >
 > filter(f, range(2, 25))



having the "filter" inside the "def f" doesn't make much sense;
it will never be reached where it is (after a "return" statement).
I trust that you don't mean to be calling "f" recursively.

The final "f( )" makes no sense because f is defined as taking a
single argument.  Delete this line.

Finally, there's a difference between running in interactive mode and
running a python file as a script.  In interactive mode, values of
computed objects are printed by default.  In a script, they are not.
That is,  at the Python prompt,  entering "2+2" will result in "4"
being printed, but if you create a file "test.py" containing only the
line "2+2" and do "python test.py",  2+2 will be calculated but not
printed.   I believe that if you modify your example to something like:


 > def f(x):
 >         return x%2 != 0 and x%3 != 0
 >
 > print filter(f, range(2, 25))

you will get what you intended.

Hope this helps.




From gerrit.holl at pobox.com  Mon Jul  5 07:41:19 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Mon, 5 Jul 1999 13:41:19 +0200
Subject: php-like python scripting without apache access.
Message-ID: <19990705134119.A23384@optiplex.palga.uucp>

Hello all,

I want to use python like php, so that I have a document like:

print_header(title='discoverb 0.2')

discoverb is blablabla

print_footer()

and than, I want to have a header with a full-blown BASE tag, how do I do
that?

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From Gary.Stephenson at p98.f112.n480.z2.fidonet.org  Sat Jul  3 04:10:46 1999
From: Gary.Stephenson at p98.f112.n480.z2.fidonet.org (Gary Stephenson)
Date: Sat, 03 Jul 1999 09:10:46 +0100
Subject: Tuples cf. Expanded objects??
Message-ID: <00001668@bossar.com.pl>

From: Gary Stephenson <stephensong at iname.com>

>
> There is probably something in the the FAQ.  BAsically, tuples are stored
> more compactly and can be used as dictionary keys (because they ARE
> readonly).
>

How similar are Python's "Immutable Tuples"   to Eiffel's "Expanded objects". 
If
one
was to try to implement  immutable tuples in another language, would an
"expanded
objects" implementation be the best, or even appropriate?

many tias

gary



From dborgelt at xxxxxxcom  Thu Jul  8 16:15:29 1999
From: dborgelt at xxxxxxcom (Dave Borgelt)
Date: Thu, 8 Jul 1999 15:15:29 -0500
Subject: Trouble with HTMLGen, HTMLtest
References: <e3UDcTMy#GA.342@nih2naae.prod2.compuserve.com> <37845270.BD3E08DD@starship.python.net> <#dvnA1Uy#GA.131@nih2naaf.prod2.compuserve.com> <3784E2E7.DD86D47B@starship.python.net>
Message-ID: <#kFOp6Xy#GA.119@nih2naae.prod2.compuserve.com>

Thanks.  This solved the problem.

Dave

Berthold H?llmann <bhoel at server.python.net> wrote in message
news:3784E2E7.DD86D47B at starship.python.net...
> Dave Borgelt wrote:
> >
> > Sorry to be so slow here but how would I be 'in' that directory.  I
added it
> > to my PythonPath so that the import works and the traceback shows that
the
> > system knows about looking in that directory
> > (c:\programming\python\HTMLGen\filename.py).
> >
> > What else would I need to do to get this to work?  To be 'in' that
> > directory?
> >
> > Dave Borgelt
> >
> > Berthold H?llmann <bhoel at server.python.net> wrote in message
> > news:37845270.BD3E08DD at starship.python.net...
> > > Dave Borgelt wrote:
> > > >
> > > > I am having a problem with HTMLGen that is driving me nuts.  It has
to
> > do
> > > > with running the test code in HTMLtest.  (HTMLtest.test()).
> > > >
> > > > I expect the solution is simple but I can't see it.  The path to the
> > HTMLGen
> > > > files is in my PythonPath (hence the import works) but as you can
see
> > the
> > > > test fails.
> > > >
> > > > -----------------------------
> > > >
> > > > >>> import HTMLtest
> > > > >>> HTMLtest.test()
> > > > Traceback (innermost last):
> > > >   File "<interactive input>", line 1, in ?
> > > >   File "c:\programming\python\HTMLGen\HTMLtest.py", line 26, in test
> > > >     'overview.html')      # manual top
> > > >   File "c:\programming\python\HTMLGen\HTMLtest.py", line 35, in
overview
> > > >     doc = SeriesDocument('HTMLgen.rc')
> > > >   File "c:\programming\python\HTMLGen\HTMLgen.py", line 256, in
__init__
> > > >     if resource: execfile(mpath(resource), self.__dict__)
> > > > IOError: [Errno 2] No such file or directory
> > > > >>>
> > > >
> > > > ------------------------------
> > > >
> > > > Any help will be appreciated.
> > > >
> > > > Dave Borgelt
> > >
> > > This only works in the directory where HTMLgen.rc (and probably other
> > > files) lie. This is usually the directory generated when unpacking
> > > HTMLge.
> > >
> > > Greetings
> > >
> > > Berthold
>
> try:
>
> import os
> os.chdir("c:\programming\python\HTMLGen\")
> import HTMLtest
> HTMLtest.test()
>
> Good luck
>
> Berthold
> --
> bhoel at starship.python.net / http://starship.python.net/crew/bhoel/
>         It is unlawful to use this email address for unsolicited ads
>         (USC Title 47 Sec.227). I will assess a US$500 charge for
>         reviewing and deleting each unsolicited ad.
>
>





From rceci at my-deja.com  Fri Jul  2 02:27:07 1999
From: rceci at my-deja.com (rceci at my-deja.com)
Date: Fri, 02 Jul 1999 06:27:07 GMT
Subject: 1.5.1 on IRIX
References: <7lg019$fro$1@nnrp1.deja.com> <377BA02A.482D8FD8@bioreason.com>
Message-ID: <7lhm3k$3fs$1@nnrp1.deja.com>

Andrew,

Took your suggestions - moved the libs back under a python1.5 subdir,
and set PYTHONHOME.  Here is my results (this the output of my csh
script running the relocated python runtime ):

PYTHONPATH=/usr/mdi_dev//python/lib/irix32/python1.5
PYTHONHOME=/usr/mdi_dev//python/lib/irix32
python=/usr/mdi_dev//python/bin/irix32/python
Could not find platform independent libraries <prefix>
Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
# /usr/mdi_dev//python/lib/irix32/python1.5/exceptions.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/exceptions.py
import exceptions # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/exceptions.pyc
# /usr/mdi_dev//python/lib/irix32/python1.5/site.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/site.py
import site # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/site.pyc
# /usr/mdi_dev//python/lib/irix32/python1.5/os.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/os.py
import os # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/os.pyc
import posix # builtin
# /usr/mdi_dev//python/lib/irix32/python1.5/posixpath.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/posixpath.py
import posixpath # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/posixpath.pyc
# /usr/mdi_dev//python/lib/irix32/python1.5/stat.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/stat.py
import stat # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/stat.pyc
# /usr/mdi_dev//python/lib/irix32/python1.5/UserDict.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/UserDict.py
import UserDict # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/UserDict.pyc
Python 1.5.1 (#3, May 21 1999, 16:24:24) [C] on irix646-32-mips2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
# /usr/mdi_dev//python/lib/irix32/python1.5/string.pyc matches
/usr/mdi_dev//python/lib/irix32/python1.5/string.py
import string # precompiled from
/usr/mdi_dev//python/lib/irix32/python1.5/string.pyc
import strop # builtin
>>>

I have a PYTHONSTARTUP that simply imports os,sys,string.

It looks like python is loading all of its modules from the correct
place (*/usr/mdi_dev/python"), as opposed to where the interpreter was
told during buildtime ("/usr/local").

I am wondering - on this second machine that is causing the warnings to
be printed out, there is no python installation in /usr/local.  Does
python check its configured buildtime prefix/exec_prefix locations, EVEN
though PYTHONHOME and PYTHONPATH are setup so python is instructed to
get its runtimes from the new location?  That would be a bummer, because
we are
distributing binaries to our customers, and need to put the distro in
any arbitrary customer-specified location, and having these warnings pop
up would probably cause a lot of confusion.

Thanks

Rob




Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From bhoel at server.python.net  Thu Jul  8 17:03:10 1999
From: bhoel at server.python.net (Berthold =?iso-8859-1?Q?H=F6llmann?=)
Date: Thu, 08 Jul 1999 23:03:10 +0200
Subject: Python and Fortran code
References: <7m23i3$ktr$1@bignews.shef.ac.uk> <3784D688.A6541F39@enews.nrl.navy.mil>
Message-ID: <3785120E.A476F208@starship.python.net>

Scott Wolford wrote:
> 
> Look at a combination of the SWIG tool (I can't remember where it's at but it
> stands for Simplified Wrapper Interface Generator so you can search for it) and
> a C wrapper for you fortran function. Basically, make it so you can call the
> fortran from C. Then swig the C function and link the cstuff.o and fstuff.o with
> cstuff_wrap.o (from SWIG) into a shared library then you can follow choice 2
> from the other response to this post.
> 
> Scott
> 
> Jose L Gomez Dans wrote:
> 
> > Hi!
> >         I have a quetion on using python to do some simple processing on
> > some data I pre-process with some Fortran code. I wanted to be able to call
> > this subroutine from within python, as it is already "tried and tested" and
> > optimized. Basically, my subroutine looks like this:
> >
> >         subroutine ex1(exdata)
> >         dimension exdata(256,200)
> >         print*,'Doing some processing in FORTRAN'
> >  C      We said we were doing some processing, so we might as well do it :)
> >  C      Processing of exdata
> >         return
> >         end
> >
> >         This subroutine would need to get input from the user and would also
> > need to print stuff out for the user to see several parameters as the
> > software runs. I'd like to pass exdata as a Python object (NumPy array
> > object), and then recover it with the results. I wonder if I could somehow
> > do that compiling this subroutine as an object file and then calling it from
> > python (not that i know how to do it, but with scilab, it works :D).
> >
> >         If it's of any use, I am using Debian GNU/Linux and g77. And using
> > loads of complex numbers :))))
> >
> >         Thanks in advance!
> >         Jose
> >

For the Fortran wrapping you might take a look at cfortran.h avaible
from http://www-zeus.desy.de/~burow/ which is helpfull if you want
making the wrapping platform independent.

Greetings

Berthold
-- 
bhoel at starship.python.net / http://starship.python.net/crew/bhoel/
        It is unlawful to use this email address for unsolicited ads
        (USC Title 47 Sec.227). I will assess a US$500 charge for
        reviewing and deleting each unsolicited ad.



From robert_j_roberts at rl.gov  Thu Jul 22 20:33:12 1999
From: robert_j_roberts at rl.gov (Roberts, Robert J)
Date: Thu, 22 Jul 1999 17:33:12 -0700
Subject: Non-ODBC MS Sql Connect
Message-ID: <3797B848.74D04740@rl.gov>

How to do I make a non-odbc connection to a MS SQL Server?

Am running Python Windows NT/IIS.

Thanks,




From cjw at connection.com  Mon Jul 26 12:24:19 1999
From: cjw at connection.com (Colin J. Williams)
Date: Mon, 26 Jul 1999 12:24:19 -0400
Subject: Incompatibility.py
Message-ID: <379C8BB3.12D50F27@connection.com>

The following program gives different results under IDLE:
                                               ------------------------
''' To explore the Name given to a program, using DOS, PythonWin and IDLE.
'''
print 'The name is:',   __name__
                                                -----------------------
DOS command line returns: The name is: __main__
PythonWin returns:               The name is: __main__
IDLE returns:                        The name is: Incompatibility

Since test routines typically rely on the former, such routines, as for example
urllib.py,
fail under IDLE.

Colin W.







From quinn at lira.ugcs.caltech.edu  Fri Jul 16 19:26:17 1999
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 16 Jul 1999 23:26:17 GMT
Subject: propapating exceptions
Message-ID: <slrn7ovfss.4bc.quinn@lira.ugcs.caltech.edu>

I've lately found myself in the position of wanting to catch exceptions
based on their attributes, rather than on the exception itself.  So (in
mal's odbc module):

try:
	some sql command
except ODBC.MySQL.OperationalError, msg:
	if msg[1] == CR_SERVER_GONE_ERROR: # (snarfed out of errmsg.h)
		restart server, try again
	else:
		raise sys.exc_info()[0], sys.exc_info()[1]

It's a rather wordy way to pass the exception on without touching it,
and I still lose the traceback info.  Is there a better way to pass on
an exception as if it had never been caught?



From rurban at xarch.tu-graz.ac.at  Sun Jul 18 07:03:59 1999
From: rurban at xarch.tu-graz.ac.at (Reini Urban)
Date: Sun, 18 Jul 1999 11:03:59 GMT
Subject: Harlequin Is Back!
References: <%vti3.121$95.4053@newse2.tampabay.rr.com> <wkn1wzl9kt.fsf@ifi.uio.no> <71E04BBBB50245BC.87397A36A7EDBBA0.44F0816EBAA77500@lp.airnews.net> <378E57D8.A77A9F9C@nmia.o_r_g> <302747D6133FC474.B42081F80C6F00FB.672284C91DB0497B@lp.airnews.net>
Message-ID: <3791b247.269188472@judy.x-ray.local>

claird at Starbase.NeoSoft.COM (Cameron Laird) wrote:
>Coincidentally, I report today on several
>related projects of cross-language co-operation in <URL:
>http://www.sunworld.com/swol-07-1999/swol-07-regex.html#2>.

this article unfortunately has several broken links:

http://starbase.neosoft.com/~claird/comp.object.corba/COM.html
-> http://starbase.neosoft.com/~claird/comp.object.corba/CORBA.html

http://starbase.neosoft.com/~claird/comp.protocols.misc/COM.html 
-> ??
--                                         
Reini



From tseaver at Starbase.NeoSoft.COM  Tue Jul  6 02:00:12 1999
From: tseaver at Starbase.NeoSoft.COM (Tres Seaver)
Date: 6 Jul 1999 01:00:12 -0500
Subject: Win32: silent script
References: <37809062.6D463928@phoenix-edv.netzservice.de>
Message-ID: <7ls61c$3nu$1@Starbase.NeoSoft.COM>

In article <37809062.6D463928 at phoenix-edv.netzservice.de>,
Holger Jannsen  <holger at phoenix-edv.netzservice.de> wrote:
>High,
>
>I'd like to use a little python-script to start win-programs
>with special parameters. So I've made some links up to that 
>python-script in windows-menu. 
>But I'd like to let the programs start without the 
>rising dos-window, that occurs for a few centiseconds while
>python-script is working. Or even let it start with a minimized
>window outside the real screen or something like that....
>
>How could I realize that?
>Is there a startup-flag I could set when calling python.exe?
>Other possibility: How could I set the 'minimized window' option
>in that link by python-code?
>
>thanx,
>Holger

Use pythonw.exe.  Its entire purpose in life is to make the console
window unnecessary.

-- 
---------------------------------------------------------------
Tres Seaver           tseaver at palladion.com       713-523-6582
Palladion Software    http://www.palladion.com



From marlon at soda.CSUA.Berkeley.EDU  Thu Jul 22 17:48:16 1999
From: marlon at soda.CSUA.Berkeley.EDU (marlon)
Date: Thu, 22 Jul 1999 14:48:16 -0700
Subject: string.replace
Message-ID: <Pine.BSF.4.10.9907221446520.20655-100000@soda.CSUA.Berkeley.EDU>

can anyone tell me why this breaks?

Python 1.4 (Nov  4 1997)  [GCC 2.7.2.3]
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import sys, string
>>> line = 'cat "in" the hat'
>>> line = string.replace( line, '"', '')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
AttributeError: replace
>>> 


_




From bwarsaw at cnri.reston.va.us  Wed Jul 21 11:18:45 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Wed, 21 Jul 1999 11:18:45 -0400 (EDT)
Subject: Tcl/Python: Who was first?
References: <16E7930929E7C7E5.430278EC189FFB38.BB6793F4819F7C95@lp.airnews.net>
	<1279562730-15211949@hypernet.com>
Message-ID: <14229.58581.753416.424408@anthem.cnri.reston.va.us>

>>>>> "Gordo" == Gordon McMillan <gmcm at hypernet.com> writes:

    Gordo> Cameron wrote:

    >> ... Barry has Mailman and JPython and starship under con- trol,
    >> ...

    | But at such a cost
    | Now speaks only haiku
    | And eats living rodents

Chris pilots starship
Python.org is my curse
Mmmm!  Lemming curry!



From Ovidiu.Predescu at p98.f112.n480.z2.fidonet.org  Thu Jul  1 12:57:15 1999
From: Ovidiu.Predescu at p98.f112.n480.z2.fidonet.org (Ovidiu Predescu)
Date: Thu, 01 Jul 1999 17:57:15 +0100
Subject: coverage tool for Python?
Message-ID: <00001625@bossar.com.pl>

From: Ovidiu Predescu <ovidiu at cup.hp.com>

On Thu, 01 Jul 1999 11:05:30 -0400, Andrew Csillag <andrew at starmedia.net> 
wrote:

> Theres pycover also at
> ftp://ftp.python.org/pub/python/contrib/All/pycover-0.1.tar.gz which,
> while not all that great, should give you some good stuff.

Thanks a lot for the pointer, it is exactly what I'm looking for!

Greetings,
Ovidiu




From bgue at my-deja.com  Tue Jul  6 22:41:12 1999
From: bgue at my-deja.com (bgue at my-deja.com)
Date: Wed, 07 Jul 1999 02:41:12 GMT
Subject: Q (Tkinter):maximizing windows by default?
Message-ID: <7lueo7$93$1@nnrp1.deja.com>

Is it possible to have a Toplevel window maximized when it's created
(in W95)? An example would be appreciated...

Thanks,

Brian Gue
bgue at my-deja.com


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From pduffin at mailserver.hursley.ibm.com  Thu Jul  1 12:07:06 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Thu, 01 Jul 1999 17:07:06 +0100
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.i <lwoghyexy5.fsf@copernico.parades.rm.cnr.it> <7ldt5p$srh$1@dragon.flightlab.com> <lwhfnosvuz.fsf@copernico.parades.rm.cnr.it>
Message-ID: <377B922A.4A7B@mailserver.hursley.ibm.com>

Marco Antoniotti wrote:
> 
> jenglish at flightlab.com (Joe English) writes:
> 
> > Marco Antoniotti wrote:
> >
> > >BTW. Is TCL statically or dynamically scoped?
> >
> > Neither, really, since Tcl doesn't have nested procedures, lambda
> > expressions, "let/letrec", local bindings or the like.
> >
> > I'd describe Tcl as "namespace-scoped": there is an open-ended set
> > of top-level "namespace" environments, plus a distinguished "global"
> > environment.  In addition, each Tcl "proc" has its own local environment,
> > into which variables from other environments can be explicitly imported
> > via "variable", "global", explicit namespace reference, or "upvar".
> > "upvar" gives you something *sort* of like dynamic scope, but not
> > exactly.
> 
> Now, who was saying that Common Lisp macros are a source of
> inconsistencies? :)
> 

I was referring to the fact that CL macros 'complicate' the parsing /
evaluation rules of Lisp which otherwise would be very simple and
consistent.

I personally like Lisp a lot and take advantage of lots of short cuts
that macros provide but I think that there is a big learning curve to
climb if you want to write one yourself (which is why I referred to it
as another language). Tcl on the other hand allows me to use the same
commands that I use to write normal programs to extend the language by
adding new programming constructs.

Like macros / loathe macros is a subjective thing and probably changes
with your experience of them.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From kyliu at idiomtech.com  Sat Jul  3 15:51:42 1999
From: kyliu at idiomtech.com (Ken Liu)
Date: 3 Jul 1999 19:51:42 GMT
Subject: Problem with informixdb and python1.5.2
References: <7ljsbc$srn$1@news.harvard.net> <377E2D7B.BEBC065C@arlut.utexas.edu>
Message-ID: <7llpke$m2r$1@news.fas.harvard.edu>

Thomas Bryan <tbryan at arlut.utexas.edu> wrote:

> Do you now have both python 1.5.1 and python1.5.2 installed?

Yes. 

> Have you tried recompiling informixdb with the 1.5.2 header files?  
> I don't think that too much has changed between 1.5.1 and 1.5.2, but 
> it couldn't hurt to try.

That sounds like the thing to do.  In fact, I did change the PYTHONINCL and PYTHONLIB dirs
to point to the 1.5.2 directories and recompiled informixdb, but the same problem persists.  

so is the common experience that the modules don't work when you have multiple versions of
python installed?  Why is that?  Is there anyway for me to "unconfuse" the module so that
it would ignore the 1.5.1 installation (short of actually removing it...)?


thanks much

k



From fredrik at pythonware.com  Wed Jul 28 18:24:27 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Jul 1999 22:24:27 GMT
Subject: INT16 on a cray t3e
References: <379F7E80.ABD@mothra.lbl.gov>
Message-ID: <002701bed947$f5140d30$f29b12c2@secret.pythonware.com>

(mailed and posted)

Bill Crutchfield <wyc at mothra.lbl.gov> wrote:
> The Python Imaging Library (PIL) seems to require 16 bit integers. 
> Unfortunately, a cray t3e doesn't have anything shorter than a 32-bit
> integer.  
> 
> Has anyone solved this problem?

if I had a Cray, I would try changing the
tests in ImPlatform.h to look like:

#if SIZEOF_SHORT == 2
#define INT16 short
#elif SIZEOF_INT == 2
#define INT16 int
#else
#define INT16 short /* will have to do... */
#endif

worth trying, at least.  some operations will
go wrong, but most of the stuff (and every-
thing that is really important) should work
just fine anyway...

let me know if what I just said was true...

</F>




From gmcm at hypernet.com  Thu Jul 15 11:16:57 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 15 Jul 1999 10:16:57 -0500
Subject: getting values of global variables?
In-Reply-To: <378C5F81.E6F8B014@phoenix-edv.netzservice.de>
Message-ID: <1280074494-16090399@hypernet.com>

Holger Jannsen writes:
> 
> ok, sounds easy, but I get a little confused...;-(

That's because what's confusing you isn't what you think is confusing 
you!

> I'd like to parse over a file reading line per line.
> I want to write everything through a special filter
> which exchanges known tokens. This may be special
> dictionaries or global variables, too.
> The tokens have to be surrounded by '%'-charcter.
> 
> Getting a problem with the global varaibles. I couldn't
> get their values as string and havn't found the help
> in books yet. I think it's quite simple and I'm just a 
> "four blind mice".;-)
> 
> To check the filter I wrote:
> 
> -------------
> import string, os, re
> HEIN="Bl?d"
> 
> def egalizeTokens(aString):
>  "changes known tokens in a string"
>  try:
>   splitter=re.split('(%.*?%)',aString)
>  except:
>   return aString

Judging by the format of your input, a better regex would be 
'%[^ ]*?%' - which give you ' %1 %2 ' instead of a phony match.

>  for s in splitter:
>   print s
>   if len(splitter)==1:
>    aTok = checkForToken(s)
>   else:
>    ox = string.split(s, '%')
>    if len(ox)==1:
>     aTok = ox[0]
>    else:
>     aTok = checkForToken(ox[1])
>   if aTok==None:
>    splitter[splitter.index(s)] = s
>   else:			
>    splitter[splitter.index(s)] = aTok	
>  try:
>   aTok = string.joinfields(splitter,'')
>  except:
>   aTok = splitter[0] #only one element!
>  return aTok

Sorry, Holger, but that's a mess! Something like this will get you 
away from all those special cases:

 rslt = []
 for tok in splitter:
     if tok[0] != '%' or tok[-1]!= '%':
         rslt.append(tok)	#not a special token
     else:
         test = tok[1:-1]		#strip those '%'s
         rslt.append(checkForToken(test))
 return string.join(rslt)

> 
> def checkForToken(aString):
>  "checks a string for known tokens"
>  try: 
>   aTok = eval(aString)
>   if aTok==float(aString): #don't exchange numbers!
>    return None

Well, eval('HEIN') does indeed produce 'Bl?d'. But float('HEIN') 
raises an exception, which is caught by code that assumes it's "eval" 
that failed.

>   return aTok
>  except:
>   aTok = os.environ.get(aString)
>   if aTok==None:			
>    aTok = directories.get(aString)
>    if aTok==None:
>     aTok = files.get(aString)
>   return aTok

How about (leaving out directories and files, which you haven't 
included):

def checkForToken(aString):
 "checks a string for known tokens"
 print 'checking', aString
 try: 
  aTok = eval(aString)
 except:
  aTok = os.environ.get(aString)
  if aTok==None:			
   aTok = aString
 if type(aTok) != type(''):
     aTok = str(aTok)
 return aTok

> if __name__=='__main__':
>  print egalizeTokens("This is a test! %WINDIR% %1 %2 %HEIN%. That's
>  it") raw_input ("Press RETURN!")
> -------------------------------
> 
> Perhaps there's a much easier way to get what I want????
> (Still I am a newbie..;-)))
> 
> 
> 
> many thanx,
> Holger
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list

- Gordon



From aahz at netcom.com  Mon Jul 12 22:21:31 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 13 Jul 1999 02:21:31 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
References: <14217.63931.75257.566266@weyr.cnri.reston.va.us> <199907121445.HAA09860@netcom9.netcom.com> <14218.7197.219561.903989@weyr.cnri.reston.va.us> <wfwi3.2961$VL2.372747@news.direcpc.com>
Message-ID: <7me7rb$rfs@dfw-ixnews14.ix.netcom.com>

In article <wfwi3.2961$VL2.372747 at news.direcpc.com>,
Emile van Sebille <emile at fenx.com> wrote:
>
>Well, some trains don't care if you don't brake...

This train don't carry no gamblers, this train....
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From aahz at netcom.com  Tue Jul 20 18:21:14 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 20 Jul 1999 22:21:14 GMT
Subject: Python suitability questions
References: <0CCE781E44F2C872.F528377A62B41DCA.73CAA63AD6E79314@lp.airnews.net> <7n2a9p$trq$1@news.accu.uu.nl> <C8A42E1E0DB3D792.D5899D7736E26702.F072FD528818FB27@lp.airnews.net>
Message-ID: <7n2soq$228@dfw-ixnews17.ix.netcom.com>

In article <C8A42E1E0DB3D792.D5899D7736E26702.F072FD528818FB27 at lp.airnews.net>,
Jason Maskell <jmaskell at nospam.saskpublishers.sk.ca> wrote:
>
>Actually, not worried so much about indents and braces and all that, just
>thinking that I might have to have some ungodly source file organisation
>where the cards are in one or more text files, and then one or two other
>text files for various other things. I guess I'm just more interested in
>whether or not I can point the interpreter to an arbitrary chunk of code and
> let it do it's thing, or what exactly is involved. I'll do some more
>research. ;>

We put a lot of our code in a database.  We're starting to have second
thoughts about it, but it might work for you.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From ivanlan at callware.com  Thu Jul 29 17:23:42 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Thu, 29 Jul 1999 21:23:42 GMT
Subject: [Tutor] I want to learn to hack
References: <19990721154610.B3663@optiplex.palga.uucp>
		<Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org>
		<14231.11726.916541.492153@nobot.cnri.reston.va.us> <slrn7q1dau.17t.esr@snark.thyrsus.com>
Message-ID: <37A0C65E.8E2D52A@callware.com>

Hi All--

"Eric S. Raymond" wrote:
> 

[bobbitt]

> 
> I fear this is my fault.
> 
> Some time back, I published a page called "How To Become A Hacker"
> <URL:http://earthspace.net/~esr/faqs/hacker-howto.html">.  As a result,
> I get a couple dozen semi-illiterate email notes a week from teenagers
> wanting me to teach them everything I know.  Instantly.
> 
> (I tolerate this because once a month or so I get a letter from a kid
> who is well-spoken, intelligent, and thoughtful.  *Those* I help.)
> 
> Earlier this year I changed the document to recommend Python as a
> first language.  I still think this was a good idea, but I see it has
> had an unfortunate side-effect.
> --
>                 <a href="http://earthspace.net/~esr">Eric S. Raymond</a>

Oh, we'll forgive you.  We'll forgive *anyone* who recommends Python as
a first language!  But maybe you should change your document to point
people to the tutor list, for *really* new Python newbies, though? 

	http://www.python.org/psa/MailingLists.html

or maybe just

	mailto:tutor at python.org

might even be better.

<no-witticisms-from-those-under-deadline!>-ly y'rs,
Ivan 
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From wware-nospam at world.std.com  Sat Jul 31 23:36:47 1999
From: wware-nospam at world.std.com (Will Ware)
Date: Sun, 1 Aug 1999 03:36:47 GMT
Subject: Python microthreads
Message-ID: <FFropB.7tp@world.std.com>

I have been interested in using Python to program simulated
swarms of interacting agents. For this purpose, I have implemented
what I call "microthreads" in Python, modifying Python's virtual
machine, and maintaining a frame stack for each thread. I have
successfully run 1500 threads simultaneously. Source code and some
example usages are available at
ftp://ftp.std.com/pub/wware/uthread.tgz
for anybody interested in this. This is a pre-pre-thinking-about-
alpha-some-day release, and probably buggy, but the examples run
as expected.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com



From olipt at mayo.edu  Mon Jul 19 22:51:14 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Mon, 19 Jul 1999 21:51:14 -0500
Subject: Scientific Plotting?
In-Reply-To: <379392E7.EB0F29E7@ssec.wisc.edu>
References: <379392E7.EB0F29E7@ssec.wisc.edu>
Message-ID: <Pine.LNX.4.10.9907192140240.7296-200000@us2.mayo.edu>

> People at my current work use MatLab or IDL for chomping data through
> data calculations and plotting the results in an environment that lets
> you get GUI's up and going very quickly (among other things).  It's been
> a goal of mine that one day I could show them that this could be done
> with as little effort (with greater room for expansion) using Python.  
> Plus I'd like to put a strong case forward to our physics department for
> our up-and-coming new scientific programmers to learn something with a
> broader scope like Python than the now used IDL and recently abandoned
> Fortran-77.
> 

I use gist quite a bit. It is availale for Unix and integrated nicely with
NumPy.  I've attached some code which emulates the Matlab plot command
(mostly) using gist.  It is quite handy for interactive plotting.  I've
used gnuplot before too and it works well.  Look at the scientific
computing topic page at www.python.org there is a link to Janko Hauser's
page on plotting packages.  

The rapid GUI stuff will require using graphite on a Tk or Gtk or wxPython
canvas.  It is much, much more flexible than Matlab.

> 
> And lastly, can array operations (using NumPy) be done without for loops
> as in IDL/MatLab?  eg Multiplying 2 arrays is easy in these two packages
> because you don't need array subscripting and iteration.
> 

Yes, yes, yes.  The NumPy array is quite nice.  After using Matlab for
several years, I switched to Python and have really enjoyed the added
power and flexibility of Python with NumPy.

Oh, don't count out Fortan 77.  With the recently released PyFort and f2py
(check the matrix-sig) you can link Fortran codes into NumPy easily.  A
linked-in subroutine in Fortran can be used to get compiled speed and ease
of development.


-------------- next part --------------

import gist
import Numeric

_types = {'-':'solid','--':'dash',':':'dot','-.':'dashdot','-:':'dashdotdot'}
_corder = ['b','r','m','g','c','k','y']
_colors = {'k':'black','r':'red','b':'blue','m':'magenta','g':'green','y':'yellow','c':'cyan'}
_markers = { '+':'\2','.':'\1','*':'\3','o':'\4','x':'\5'}


def _find_and_set(dict, str, default):
    import string
    value = default
    for k in dict.keys():
        if string.find(str,k) >= 0:
            value = dict[k]
            break
    return value

def _parse_type_arg(thearg,nowplotting):
    indx = nowplotting % len(_corder)
    if type(thearg) is type(''):
        tomark = 1

        thetype = _find_and_set(_types,thearg,'none')
        thecolor = _find_and_set(_colors,thearg,_colors[_corder[indx]])
        themarker = _find_and_set(_markers,thearg,None)

        if (themarker == None):
            tomark = 0
            if thetype == 'none':
                thetype = 'solid'        

        return (thetype, thecolor, themarker, tomark)

    else:  # no string this time
        return ('solid',_colors[_corder[indx]],'Z',0)

global hold_on
hold_on = 0

def plot(x,y=None,*args):
    global hold_on
    if hold_on == 0:
        gist.fma()
    gist.animate(0)
    nargs = len(args)
    if nargs == 0:
        if y == None:
            y = x
            x = Numeric.arange(0,len(y))
        gist.plg(y,x,type='solid',color='blue',marks=0)
        return
    argpos = 0
    nowplotting = 0
    while 1:
        try:
            thearg = args[argpos]
        except:
            thearg = 0
        thetype,thecolor,themarker,tomark = _parse_type_arg(thearg,nowplotting)
        if themarker == 'Z':  # args[argpos] was data or non-existent.
            pass
        else:                 # args[argpos] was a string
            argpos = argpos + 1    
        gist.plg(y,x,type=thetype,color=thecolor,marker=themarker,marks=tomark)

        if argpos+1 >= nargs: break      # no more data
        nowplotting = nowplotting + 1
        x = args[argpos]
        y = args[argpos+1]
        argpos = argpos+2
    return

def imagesc(z):
    gist.fma()
    gist.animate(0)
    gist.pli(z)

def xlabel(text):
    gist.plt(text,0.40,0.40,justify='CC',opaque=1)





From nospam at bitbucket.com  Sat Jul  3 03:00:55 1999
From: nospam at bitbucket.com (Phil Mayes)
Date: Sat, 3 Jul 1999 00:00:55 -0700
Subject: Mailbox, tell and read
References: <377AAE02.E8591727@prescod.net>
Message-ID: <Pyif3.1021$dR.88004@newsfeed.avtel.net>

Paul Prescod wrote in message <377AAE02.E8591727 at prescod.net>...
>The mailbox message objects simulate a file interface by seeking through a
>mailbox. A read() returns the entire message. It does this by seeking to
>the start, reading to the message end, and returning the data. The problem
>is that message offsets are derived by file.tell() which I presume is in
>terms of bytes. But then it does a read() based on the offsets and read()
>is in terms of characters.
>
>These don't match up on Windows because of CR/LF pairs. I think that the
>right fix is to never use tell() and instead keep track of the location by
>counting characters. Does that make sense?


I am working with multiple mails in a single file, and found that using
seek/tell on the file in text mode didn't allow me to extract a single mail,
so I use binary, but I have to convert internally else quopri leaves =CRLF
inside the string.  Code fragment:
    # read the body from the mbx file with 'rb' to get the correct
    # amount of data in, then convert CRLF to LF else quopri fails.
    fmbx = open(mbxfile, 'rb')
    fmbx.seek(self._offbody)
    s = fmbx.read(self._offend - self._offbody)
    return string.replace(s, '\r\n', '\n') # runs in 2/3 the time of
    # ... joinfield/splitfield
--
Phil Mayes    pmayes AT olivebr DOT com







From aahz at netcom.com  Sun Jul  4 22:02:04 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 5 Jul 1999 02:02:04 GMT
Subject: Why have 300 articles just appeared in c.l.py?
References: <m33dz4oxxx.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <7lp3ms$ok5@dfw-ixnews11.ix.netcom.com>

In article <m33dz4oxxx.fsf at atrus.jesus.cam.ac.uk>,
Michael Hudson  <mwh21 at cam.ac.uk> wrote:
>
>I just had the slightly alarming experience of about 300 articles, all
>of which I had read, appearing unthreaded in this newsgroup.

If you looked at the From: line, there was an added suffix that included
"fidonet" to everyone's e-mail address.  From which I conclude that
someone's fidonet<->Usenet gateway barfed.  Killing all From: lines with
"fidonet" did the job.
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From kernr at mail.ncifcrf.gov  Mon Jul 19 18:58:32 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Mon, 19 Jul 1999 22:58:32 GMT
Subject: Scientific Plotting?
References: <379392E7.EB0F29E7@ssec.wisc.edu>
Message-ID: <3793a9df.3913871@news.erols.com>

On Mon, 19 Jul 1999 16:04:39 -0500, Nick Bower
<nick.bower at ssec.wisc.edu> wrote:

[snip]

>But...What do people use for plotting?  After looking at the Python
>Topic Guides, I've taken a quick look at the BLT and DISLIN pages, but
>both of these amazingly lack any screen shots to demonstrate their
>capabilities!!!

I use the gnuplot interface for quick, exploratory graphs.
(http://monsoon.harvard.edu/~mhagger/download/Gnuplot.html)
It can also produce Postscript hardcopies.  If you are used to gnuplot
proper, you can send it arbitrary commands through the interface.

Recently, I've been playing around with the alpha Graphite package
(http://www.strout.net/info/coding/python/graphite).  Graphite runs on
top of PIDDLE, a plotting API that currently has several backends
(e.g. PDF, Postscript, Tk, any image format available through PIL, and
more).  It's still quite rough but is pure Python; it's capabilities
expand as your needs do (if you have the time to figure out how to do
it).

>And lastly, can array operations (using NumPy) be done without for loops
>as in IDL/MatLab?  eg Multiplying 2 arrays is easy in these two packages
>because you don't need array subscripting and iteration.

That's exactly what NumPy was designed to do.  NumPy arrays overload
"*" to mean element-wise multiplication.  Matrix multiplication is
available through the "dot" function.

>Thanks muchly, Nick.

You're welcome.

Be sure to check out the Matrix and Plotting SIGs
(http://www.python.org/sigs/matrix-sig/ and
http://www.python.org/sigs/plot-sig/ respectively)

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From ovidiu at cup.hp.com  Thu Jul  1 12:57:15 1999
From: ovidiu at cup.hp.com (Ovidiu Predescu)
Date: Thu, 1 Jul 1999 16:57:15 GMT
Subject: coverage tool for Python? 
In-Reply-To: Your message of "Thu, 01 Jul 1999 11:05:30 EDT."
             <377B83BA.448D6525@starmedia.net> 
References: <377B83BA.448D6525@starmedia.net>
Message-ID: <199907011657.JAA12118@hpcll563.cup.hp.com>

On Thu, 01 Jul 1999 11:05:30 -0400, Andrew Csillag <andrew at starmedia.net> 
wrote:

> Theres pycover also at
> ftp://ftp.python.org/pub/python/contrib/All/pycover-0.1.tar.gz which,
> while not all that great, should give you some good stuff.

Thanks a lot for the pointer, it is exactly what I'm looking for!

Greetings,
Ovidiu





From rbl at hal.cwru.edu  Wed Jul 28 13:16:12 1999
From: rbl at hal.cwru.edu (rbl at hal.cwru.edu)
Date: 28 Jul 1999 17:16:12 GMT
Subject: Newbie Problem
Message-ID: <7nndss$q1n$1@alexander.INS.CWRU.Edu>

I'm new to Python, but quite excited about its possiblities.  I'm
running on a Mac G3, System 8.6.

There is some Python code I've downloaded that I'm trying to run.
If I just drop the .py file on the Python icon, I get some errors,
so I am trying to run it interactively for debugging.

Here is what happened:

Python 1.5.1 (#37, Apr 27 1998, 13:36:04)  [CW PPC w/GUSI w/MSL]
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import sys
>>> import random
>>> import string
>>> 
>>> def run(filename=''):
... if filename=='':
  File "<stdin>", line 2
    if filename=='':
     ^
SyntaxError: invalid syntax
>>> 

The original code segment is:

import sys
import random
import string

def run(filename=''):
	if filename=='':
		file = open( raw_input('Enter name of a textfile to read: '), 'r')
	else:
		file = open( filename, 'r')
	
	text = file.read()
	file.close()
	words = string.split(text)

===
I must be doing something stupid and/or naive.

Advice, please?

Thank you,
Robin Lake
rbl at po.cwru.edu




From g_will at cyberus.ca  Fri Jul 16 16:07:25 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Fri, 16 Jul 1999 20:07:25 GMT
Subject: Radio Buttons and Variables
References: <01becfaa$4fc73d80$be54c3d1@Cyberus.cyberus.ca> <001101becfc4$8e58b370$f29b12c2@secret.pythonware.com>
Message-ID: <01becfc6$d3445e60$6854c3d1@Cyberus.cyberus.ca>

Thanks for the solution. This has taken a lot of my time.

Do you know why (root) works and not just ().

Can you suggest an IDE that will not give me this problem?

Gordon Williams

Fredrik Lundh <fredrik at pythonware.com> wrote in article
<001101becfc4$8e58b370$f29b12c2 at secret.pythonware.com>...
> Gordon Williams <g_will at cyberus.ca> wrote:

> 
> the fact that you're running this under PythonWin is
> definitely the cause of the problem; the code looks
> fine, and I cannot repeat this problem using my
> favourite IDE...
> 
> > Any ideas to solve this?
> 
> try changing:
> 
>      foovar = IntVar()
> 
> to
> 
>      foovar = IntVar(root)
> 
> and see if that helps.
> 
> </F>
> 
> 



From pduffin at mailserver.hursley.ibm.com  Thu Jul  1 11:45:02 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Thu, 01 Jul 1999 16:45:02 +0100
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <lwoghyexy5.fsf@copernico.parades.rm.cnr.it> <377B454C.6F59@mailserver.hursley.ibm.com> <377B593B.F7F7504B@iname.com>
Message-ID: <377B8CFE.6113@mailserver.hursley.ibm.com>

Fernando Mato Mira wrote:
> 
> OK. Let's dispel Lisp mith #4 and then
> let's get this out of comp.lang.python
> [economics are different than other groups,
>  BEWARE of Netscape's scrollbar!]
> 
> Paul Duffin wrote:
> 
> > > > I am surprised at this because Tcl's [upvar] is simply an explicit form
> > > > of dynamic binding which I seem to remember is what Lisp uses. (That
> >
> > Mistake, I think I meant dynamic scoping.
> 
> >         (defun inner (b)
> >           (+ a b))
> >
> >         (defun outer (a)
> >           (inner 3))
> >
> >         (outer 2)
> 
> >
> > A simple macro version of setq would convert from
> >         (setq symbol '(list))
> > to
> >         (set 'symbol '(list))
> 
> All 3 the same instance of the first misconception.
> That's OLD, BROKEN Lisp!!

Please enlighten me as to what my misconception is. I am obviously not
a Lisp expert and I realise that there are plenty of different flavours
around but the above code works inside Emacs which says that it is
based on Common Lisp.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From nikolaosk at hotmail.com  Tue Jul 27 03:59:21 1999
From: nikolaosk at hotmail.com (Nikolaos Kantzelis)
Date: Tue, 27 Jul 1999 00:59:21 PDT
Subject: (no subject)
Message-ID: <19990727075921.64960.qmail@hotmail.com>

unsuscribe


______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com



From eugene.leitl at lrz.uni-muenchen.de  Thu Jul  1 20:12:06 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Thu,  1 Jul 1999 17:12:06 -0700 (PDT)
Subject: Language extensibility (was: Why is tcl broken?)
In-Reply-To: <wkzp1g2t2e.fsf@ifi.uio.no>
References: <1283099377-77267484@hypernet.com>
	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com>
	<375FDCA7.62FF@mailserver.hursley.ibm.com>
	<7josm1$bgc$1@Starbase.NeoSoft.COM>
	<37628EBF.207F7A8F@iname.com>
	<37682DD3.43923243@compaq.com>
	<7ka2be$qja$1@nntp6.u.washington.edu>
	<37776FFA.2D85@mailserver.hursley.ibm.com>
	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it>
	<876748mcxg.fsf@home.ivm.de>
	<37788BB9.ABD@mailserver.hursley.ibm.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.ibm.com>
	<nkjzp1hizvb.fsf@tfeb.org>
	<377A3383.D453468@iname.com>
	<377aa228.31380423@news.jpl.nasa.gov>
	<slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net>
	<wkzp1g2t2e.fsf@ifi.uio.no>
Message-ID: <14204.912.383867.589108@lrz.de>

Lars Marius Garshol writes:
 > 
 > Huh? Forth dates back to the 60s, whereas Scheme is from 1975, and I'm
 > quite unsure of whether Chuck Moore knew Lisp at all.

Chuck is mostly doing (quirky, but fascinating) hardware design 
nowadays. Check out http://www.ultratechnology.com/ to what he has
been up to recently.



From MHammond at skippinet.com.au  Tue Jul 13 23:07:02 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Wed, 14 Jul 1999 13:07:02 +1000
Subject: making a Windows NT service out of a python program
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F2E@RED-MSG-50>
Message-ID: <7mgupk$3sl$1@m2.c2.telstra-mm.net.au>

Bill Tutt wrote in message
<4D0A23B3F74DD111ACCD00805F31D8100DB90F2E at RED-MSG-50>...
>
>> >The worst thing about them is that they can be tricky to
>> debug.  If they
>> >are running in the system account then they have a restricted
>> >environment and the standard output channels send useful data to
>>
>> he he - not with Python :-)  Just import win32traceutil at
>> the top of your
>> service, and you can see all your print statements in Pythonwin :-)

Also, I meant to mention that Python services have a built-in "debug" mode,
where the service simply runs under Python.exe - this can be very handy, and
I use this most of the time when developing.

>Heck, if you give it permission to interact with the desktop, you might
>actually be able to use Mark's cool GUI debugger. :) (No I haven't tried
>this...)

Actually, I know someone who has.  Theoretically, it will work as long as
you restrict the debugging to a single thread.  The Pythonwin debugger is
not thread friendly (although I have been thinking about that, and decided
it could be.  Other alternative would be to build a "remote debugger"
interface, so you always could debug, regardless of desktop interaction -
dunno which way I will go (probably the former for now), but it wont be for
a while (or until I personally _really_ need it :)

Mark.





From andy at robanal.demon.co.uk  Wed Jul 21 17:22:22 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Wed, 21 Jul 1999 21:22:22 GMT
Subject: designing question
References: <1279735663-4809742@hypernet.com> <37932383.A9AA3B3B@worldnet.fr>
Message-ID: <3797392f.11013486@news.demon.co.uk>

pserrand <bucodi at worldnet.fr> wrote:

>> > 2. The application will have abouth 50 different inputscreens wich
>> > acces all to different tables. I will create a class for every
>> > window wich inheret from a general 'input-screen'-class. I wonder if
>> > a should place the db-access code (insert-update-delete) in every
>> > specific window or if i should  create a big db class wich has all
>> > possible db-acces in it ?
>>
>> Completely philosophical question. For flexibility, I favor the GUI
>> making no assumptions about how the data is stored. I would have the
>> GUI making calls like "app.addthingie(thingie)", where the app has an
>> "abstract" UI.
>
>I'm surprised that this is just a philosophical question :) I tought
>there would be some
>practical reasons to do it one or another way (p.e. maintenance of the
>application)
>The reason i posted this question is that i have very little experience
>with OOP. (actually i learned it with python <wink>) So i'm looking for
>some experience here
>
With 50 screens, it is going to be complex.  I'd suggest first
modelling the 'objects' your program deals with.  Thus, if it is an
invoicing program, you make objects which represent the invoices,
customers, payments etc in as natural a way as you can.    Build
Python classes and play with them at the interactive prompt; you can
do this faster in Python than in anything.   If the object model is
right, then building a natural GUI on top of it and a database behind
it are fairly straightforward tasks; and the program is easy to
extend.


Good luck,

Andy




From dalke at bioreason.com  Sun Jul 11 02:47:39 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Sun, 11 Jul 1999 00:47:39 -0600
Subject: code coverage
Message-ID: <7m9toh$6qg$92@rks1.urz.tu-dresden.de>

Hello,

  After the discussion a couple weeks back about code coverage
utilites for Python, I decided to look into them because I want
to ensure that our regression tests really do execute all (or at
least most) of the code in our packages.

  The two mentioned in the thread were:
 trace.py - http://www.musi-cal.com/~skip/python/trace.py
      by Skip Montanaro

 pycover - ftp://ftp.python.org/pub/python/contrib/All/pycover-0.1.tar.gz
      by Andrew Csillag


  Neither of these fit my needs, so I ended up highly modifying
trace.py.  For those interested, the new version is at
   ftp://starship.python.net/pub/crew/dalke/trace.py

I would appreciate people taking a look at it and testing it out;
It's in a pretty complete state, but needs review and suggestions
for improvements.


  The major features I added are:

     a more extensive command-line interface using getopt and
reporting a full --help message

     support for packages.  trace.py would use only basename of
the __file__ when generating a coverage log.  Thus, if there were
two submodules with the same name, one would overwrite the other.
Now the results are saved to the file named __name__ + '.covered'

     able to exclude coverage on a module based on its __file__
and/or on its __name__ (so you can exclude all the system modules
using sys.prefix).  pycover, for example, has built-in exclusion
of /usr/local, which isn't always appropriate for our installs.

     a big problem I had with both of the coverage programs was
they didn't do a very good job at telling me what executable lines
were not covered.  For example, one of my test scripts is:

=====
import daylight.Smiles

def main():
        mol = daylight.Smiles.smilin("COON")
        if len(mol.atoms) > 9:
                print "Hello" #pragma: NO COVER
                print "there"
        else:
                print "Small"
        print mol.cansmiles()

if __name__ == "__main__":
        main()
=====

when run through pycover, the un-executed lines start with a '!'

======
> import daylight.Smiles
  
> def main():
>       mol = daylight.Smiles.smilin("COON")
>       if len(mol.atoms) > 9:
!               print "Hello" #pragma: NO COVER
!               print "there"
!       else:
>               print "Small"
>       print mol.cansmiles()
  
> if __name__ == "__main__":
>       main()
======

As you see, the "else:" is marked as unexecuted, and it always will
be even if both branches of the if statment are taken, because else
generates no SET_LINENO instruction.  The basic problem is in
determining if a line should have been executed or not, and the
programs do it by looking for non-empty, non-comment ("""detection
of multiline quotes is also a problem
for these two programs""")

Now, nearly exery executable line, when compiled for the PVM, contains
code listing the line number.  So I wrote some code (based on dis.py)
to disassemble a module's code objects, look for the SET_LINENO,
and use those numbers to tell which executable lines haven't been
covered.

This trick doesn't work for everything, but is better than nothing.


So, if you want, please take a gander at the code, at
   ftp://starship.python.net/pub/crew/dalke/trace.py


						Andrew
						dalke at bioreason.com



From fredrik at pythonware.com  Mon Jul 19 13:08:32 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Jul 1999 17:08:32 GMT
Subject: tkinter call backs
References: <01bed202$58879280$e855c3d1@Cyberus.cyberus.ca>
Message-ID: <00ca01bed209$5543adf0$f29b12c2@secret.pythonware.com>

Gordon Williams <g_will at cyberus.ca> wrote:
> I have a simple problem that I cant find the solution.  I have a menu item
> to open a file which has a command "openfile".  I would like to activate
> "tkFileDialog.askopenfilename" to get the name of the file.  This requires
> "parent=root" or something like that for the position of the message box.
> 
> How do I pass "root" into openfile?  

after which he promptly answered his own
question:

>     fileName=tkFileDialog.askopenfilename(parent=root, title="Open")

what am I missing here?

</F>




From ajung at sz-sb.de  Sat Jul 17 11:26:28 1999
From: ajung at sz-sb.de (Andreas Jung)
Date: Sat, 17 Jul 1999 15:26:28 GMT
Subject: propapating exceptions
In-Reply-To: <378FDFC5.AE965089@arlut.utexas.edu>; from Thomas Bryan on Fri, Jul 16, 1999 at 08:43:33PM -0500
References: <slrn7ovfss.4bc.quinn@lira.ugcs.caltech.edu> <378FDFC5.AE965089@arlut.utexas.edu>
Message-ID: <19990717172628.A7182@sz-sb.de>

On Fri, Jul 16, 1999 at 08:43:33PM -0500, Thomas Bryan wrote:
> 
> I can't find the reference now, but I think that I read in 
> the Python docs that calling raise without an argument inside 
> an except, reraises the error.
> 
> That is, 
> 
> try:
>         some sql command
> except ODBC.MySQL.OperationalError, msg:
>         if msg[1] == CR_SERVER_GONE_ERROR: # (snarfed out of errmsg.h)
>                 restart server, try again
>         else:
>                 raise
> 
> Does that work?

This should work - raise without any arguments should reraise
the same exception without modifying the traceback - that's at
least the way it works on my system.

Cheers,
Andreas



From aahz at netcom.com  Mon Jul 12 10:45:55 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 12 Jul 1999 14:45:55 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787FB51.12FE0048@ssec.wisc.edu> <7m923i$dem@dfw-ixnews5.ix.netcom.com> <14217.63931.75257.566266@weyr.cnri.reston.va.us>
Message-ID: <7mcv33$iej@dfw-ixnews12.ix.netcom.com>

[posted & e-mailed]

In article <14217.63931.75257.566266 at weyr.cnri.reston.va.us>,
Fred L. Drake <fdrake at acm.org> wrote:
>Aahz Maruch writes:
>>
>> I think I understand how it happened: os.getcwd() is (probably
>> rightfully) documented there, so someone thought os.chdir() belonged
>> there, too.  AFAICT, that's the only mistake in that area of the docs,
>> though I could make a case that os.getcwd() should be moved (or at least
>> have a pointer in "F&D").
>
>  Are you saying that (either of) these are in the wrong place?  The
>notion of "current directory" is tied only to the entry in the process 
>table; should it be documented otherwise?
>  The "Process Parameters" section contains functions with provide any
>direct access to the process table entry, both queries and changes, so
>I think it really is the right place for these two functions.
>  I can easily add a textual reference to the right place within the
>"Files and Directories" section, but you'll need to explain very
>clearly why the descriptions should be moved.

It depends how you define "wrong place".  On the one hand, putting
os.chdir() and os.getcwd() in with the process parameters is a
*technically* correct decision, because they are associated with process
information.  OTOH, someone who is not familiar with the way processes
work will *NOT* look in that section; zie would expect to find the
information with all the other functions dealing with files and
directories -- after all, on the surface layer, these two functions deal
with directories, not processes.

Given that one of Python's goals is to be accessible to non-CS geeks,
os.chdir() and os.getcwd() *must* be documented in the section dealing
with files and directories.  As I said originally, I think a reasonable
compromise is to leave the process parameter docs alone and add a
pointer to that area in the files & directories docs (possibly with an
explanation of why they are documented somewhere else).
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From claird at Starbase.NeoSoft.COM  Sun Jul  4 16:13:00 1999
From: claird at Starbase.NeoSoft.COM (Cameron Laird)
Date: 4 Jul 1999 15:13:00 -0500
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <3779ED0F.2C67@mailserver.hursley.i <m3btdvhrlt.fsf@localhost.localdomain> <d2Of3.1233$U5.237262@ptah.visi.com>
Message-ID: <7lof8c$o9o$1@Starbase.NeoSoft.COM>

In article <d2Of3.1233$U5.237262 at ptah.visi.com>,
David Thornley <thornley at visi.com> wrote:
			.
			.
			.
>"Thinking Forth" is an excellent book to read.  So is Stroustrup's
>"Design and Evolution of C++".  I haven't seen a similar book about
>the Lisp way of thinking, not counting books specifically on learning
>Lisp and doing it well.  Then again, I find Common Lisp revolutionary
			.
			.
			.
Not *SICP*?
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From alex at somewhere.round.here  Sun Jul  4 20:11:04 1999
From: alex at somewhere.round.here (Alex)
Date: 04 Jul 1999 20:11:04 -0400
Subject: Why have 300 articles just appeared in c.l.py?
References: <m33dz4oxxx.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <etdhfnk2bxj.fsf@m2-225-12.mit.edu>

Same thing happened to me.
At last, a "me too!" post...
Alex.



From gmcm at hypernet.com  Wed Jul  7 23:56:39 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 7 Jul 1999 22:56:39 -0500
Subject: win32 python as a dll?
In-Reply-To: <3783C60C.E4311395@enews.nrl.navy.mil>
Message-ID: <1280719925-72790231@hypernet.com>

Scott Wolford writes:

> Now I'm a little confused. 

I was trying to figure out how to tell you that politely...

> Python does seem to be a dll (?); I found
> python15.dll in winnt/system32 directory. 

Correct. That's 99.999+% of python. python.exe does nothing but load 
it and call Py_Main.

> But I can't find
> python15_d.dll anywhere. Was that supposed to be installed? 

Nope, you have to build it.

> I think
> python15.dll bombs because it was compiled as optimized using the M$
> compiler and that optimizer has problems (at least w/ O2; I've been
> pretty successfull w/ O1). 

Probably not. It's been pretty thoroughly tested at this point.

> So I want to use python15_d but I can't
> find the dll.
> 
> Any help would be appreciated,

Probably the most common mistake of Windows embedders is to link to 
the wrong C runtime lib. Python uses Multithreaded DLL.

But you might have done something more creative. Telling us how and 
when it bombs complete with messages and as small a code sample as 
possible will get you more and better answers.


- Gordon



From claird at starbase.neosoft.com  Fri Jul 23 16:50:17 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 23 Jul 1999 20:50:17 GMT
Subject: Fw: ANNOUNCE: Mailing list for Australian Python Users
References: <007201bed42e$c4316c30$f29b12c2@secret.pythonware.com> <7n78nm$lic$1@news.accu.uu.nl> <379798BC.33FC0CA9@equi4.com> <19990723124741.A29309@optiplex.palga.uucp>
Message-ID: <1E4AC88B7E04EAB0.8140D4DF3103F8D3.0124F5E4F4834F96@lp.airnews.net>

In article <19990723124741.A29309 at optiplex.palga.uucp>,
Gerrit Holl  <gerrit.holl at pobox.com> wrote:
>On Fri, Jul 23, 1999 at 12:18:36AM +0200, Jean-Claude Wippler wrote:
>> Martijn Faassen wrote:
>> > 
>> > Lars Marius Garshol <larsga at ifi.uio.no> wrote:
>> > 
>> > > * Fredrik Lundh
>> > > |
>> > > | python-se, n?gon?  Link?ping?  :-)
>> > 
>> > > Tjah. Si Oslo, s? kan vi bli en del. :-)
>> > 
>> > Utrecht is veel beter denk ik, Python-NL en zo..
>> 
>> Doen.  Ik kom op de fiets.
>> 
>> > An-large-percentage-of-pythoneers-speaks-Dutch-anyway-ly yours,
>> 
>> De meerderheid, toch?
>> 
>
>Yep.
>
>Toch zie ik in nl.comp.programmeren bijna niemand die python gebruikt...
>
>Yes.
>
>But I don't see anyone using python in nl.comp.programmeren...
>
>groeten,
>regards,
>
>Gerrit Holl.

Dutch is more popular among Pythoneers than Python is
among the programming Dutch?!
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From mstenber at cc.Helsinki.FI  Sun Jul 25 07:53:00 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 25 Jul 1999 14:53:00 +0300
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net>
Message-ID: <al81zdxkktf.fsf@sirppi.helsinki.fi>

gward at cnri.reston.va.us (Greg Ward) writes:
> sreekant kodela <sreekantk at freenet.co.uk> wrote:
> If you're concerned about runtime performance -- ie. how long it takes
> to mangle strings and loop over lists and so forth -- I wouldn't worry
> about it.  Yes, Python is slow, but the Internet is so much slower that
> it hardly matters.

Even in websites' case, I would worry; Python doesn't use multiCPU very
well, and with single CPU, it is fairly easy to saturate one with
pure-Python code, especially if it is badly designed to boot.

For AMD K6-2 300mhz, your mileage may vary:

Normal null method call :  156332.1874/sec [0.19s] (6.3966us/call)
(=simple, def foo(): pass call)
Apply null method call  :   39783.6246/sec [0.75s] (25.136us/call)
(=simple, foo-call using apply)

normal get    :  114632.8877/sec [0.79s] (8.7235us/call)
normal set    :  108738.4606/sec [0.83s] (9.1964us/call)
normal getattr:  106416.8282/sec [0.85s] (9.3970us/call)
normal setattr:  109442.3212/sec [0.82s] (9.1372us/call)

Admittedly, getting approx. 100+k attribute accesses /null calls off
doesn't sound bad, but once you start poking around with dbs:

Using a Gadfly database, _with_ pre-compiled queries: (very simple
one-liners from table with one entry)
select: 21.4182/sec [3.64s] (46.689ms/call)
update: 14.3209/sec [3.98s] (69.828ms/call)

.. and so on. Even more distressing results lie in the speed of say,
implementing protocols in Python (when we're talking _nearly_ realtime
action). 

Thus I've had great deal of interest in getting the speed up; however,
speeding up dynamic interpreted languages is nontrivial, especially
languages as dynamic as Python.

My interest in Python's speed originally stemmed from need to write some
protocols and parts of a game engine in Python, but for instance LOS
calculation purposes, Python seems simply too slow.

Disclaimer: I'm still interested in hearing solutions to the speed
problem. Exhibit #2:

P/UDP:snd+rcv               :     992.0708/sec [0.95s] (1.0080ms/call)
(=Python-based UDP echo client+server; 1000 roundtrips, not bad, eh?)

Writing a simple TCP-ish connection-oriented protocol on top of UDP:
P/FakeTCP:snd+rcv      :      95.0250/sec [0.93s] (10.524ms/call)

Although network traffic caused by (simplex) connection-oriented protocol
is virtually same as the UDP case's, protocol slows down 90% thanks to the
Python's CPU use.

And compare that to the real TCP:
P/TCP:snd+rcv               :    2367.1947/sec [1.00s] (422.44us/call)

I am fairly certain that if I bothered to implement the protocol in C, I'd
get N times as much speed with same machine,

- Speed-does-matter-ly yours, Markus

P.S. This might surface as one example for 'why scripting languages don't
always cut it' in my master's thesis; if I have made some grievous
error(s?), corrections are welcome. 

P.P.S. Too bad Python doesn't have anything that creates _nice_ code, like
Stalin for Lisp ;-)

-- 
----- BEGIN GEEK CODE v3.0 BLOCK (*) -----
GCS d- s+: !a C+(++) UL+++$ P--- L+++ E++ W-- N+ W--- O---- M-- V- PS+++
Y+ PGP++>+++ t+@ 5++ X+ R* tv-- b+++ DI+++>++++ D---- G++ e++ h* r% z+
----- END GEEK CODE v3.0 BLOCK -----
* For spoilers, check http://www.geekcode.com



From MHammond at skippinet.com.au  Thu Jul  8 18:46:03 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 9 Jul 1999 08:46:03 +1000
Subject: os.kill on NT
References: <3784B57E.C0A57A@cern.ch> <3784C95D.45E5E617@rubic.com>
Message-ID: <7m39jh$qmu$1@m2.c2.telstra-mm.net.au>

Jeff Bauer wrote in message <3784C95D.45E5E617 at rubic.com>...
>Haimo,
>
>I'm not sure if you were referring to me, but here's a
>reference to a post I made to the Python TODO list:
>
>http://www.python.org/cgi-bin/todo.py?req=show&file=faq13.005.htp

I hadnt seen this before.  One style nit:  Im trying to get to the point
where "pywintypes" really is never imported directly.  My "recommended"
style is to catch "win32api.error" in that case, rather than
pywintypes.error.  They are indentical - ie, pywintypes.error is
win32api.error, and neither will go away, so really a tiny, tiny point.

Mark.






From tismer at appliedbiometrics.com  Sat Jul 17 12:00:24 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Sat, 17 Jul 1999 16:00:24 GMT
Subject: Python and Boehm-Demers GC, I have code.
References: <000401bed023$d08d3360$06a22299@tim> <7mq2ol$hk8@ds2.acs.ucalgary.ca>
Message-ID: <3790A898.DA7D61AF@appliedbiometrics.com>


Neil Schemenauer wrote:
...
> With some tuning maybe the gc version of Python will perform
> better than the regular version.

I believe you will need to keep the RC as well, since at
the moment, I think this gives more locality than GC,
as Tim mentioned already.
Early this year, I tried to speed up the VM, and one aspect
I tried was to delay DECREFS, doing them from time to
time in bulks. Although I could measure the extra overhead
away, Python became a little slower. It appears as a rule of
thumb to me: 

        RC is fine. Try to get rid of objects ASAP.

If you can manage to make it rock solid, I would perhaps
like to bundle it with stackless Python. Although, the
collector which Guido proposed a while ago could perform
even better, so I'm still a bit undecided.

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From greg.ewing at compaq.com  Thu Jul 22 17:35:56 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 23 Jul 1999 09:35:56 +1200
Subject: Cyclops 0.9.4
References: <001301bed410$e59eafe0$642d2399@tim>
Message-ID: <37978EBC.41BE8CE4@compaq.com>

Tim Peters wrote:
> 
> You want to know about objects still living that you don't expect to be
> living.  No system can answer that for you!  They can't know what you expect
> <wink>.

I'm quite happy to go through the list of living objects
and decide for myself which ones should be dead. What I
meant was, what I really want to know is how those objects
got reached. If there are cycles involved, it wouldn't
hurt to be told about them, but they're not the most
important thing.

> If what you want is a list of all reachable objects,

No, I'd like a list of *all* objects, reachable or not.
Cyclops could then do what it does now, but for the
isolated islands of garbage as well.

> All of those are also needed if Python is ever to move toward builtin
> portable mark-&-sweep (optional or not).  The advantage over "walking an
> (explicit) list" is no overhead (time or space) unless & until it's used.

If there were mark & sweep there wouldn't be unreachable
garbage, so in that case you are right -- there would be
no point in keeping a list. I'm only suggesting it as an
interim measure.

> then a Cyclops-like thingy could tell you not only that
> something unexpected is still alive, but also from where it can be
> reached

I don't understand why Cyclops couldn't be made to do
that now. It reached those objects somehow -- all it needs
to do is remember how!

I don't mean to sound ungrateful for Cyclops, by the way --
it's a great idea. All it needs is a little bit more help
from the Python core to make it even greater...

With-government-funding-I'm-sure-I-could-make-it-*really*-silly-oops-er-I-mean-useful,
Greg



From fredrik at pythonware.com  Tue Jul 13 15:19:13 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 13 Jul 1999 19:19:13 GMT
Subject: PIL+Tkinter query (I must be doing something _really_ stupid)
References: <al8673oe99o.fsf@myntti.helsinki.fi>
Message-ID: <012501becd64$981a0ba0$f29b12c2@secret.pythonware.com>

Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
> Ok, even with the topic, this is a warning: total clueness incoming ;)
> From Markus's adventures in PIL+Tkinter-land..
> 
> from Tkinter import *
> tk = Tk()
> Label(tk, text='foo').pack()
> tk.mainloop()
> 
> This works. Ok. (me, a Tcl/Tk guru, snicker)
> 
> Then, we add:
> import ImageTK
> and change Label to have
> Label(tk, image=ImageTk.PhotoImage(image)).pack()
> 
> The image is fine; image.save() produces the required output; it's RGB
> image with some map data, to be precise. HOWEVER: The label has correct
> size (size of the image) but no image data whatsoever. The same problem
> applies to all approaches I've tried (using Canvas and putting it on one,
> etc). 

see:
http://www.python.org/doc/FAQ.html#4.69

suggested change:

photo = ImageTk.PhotoImage(image)
w = Label(tk, image=photo)
w.photo = photo
w.pack()

</F>




From gcollins99 at my-deja.com  Fri Jul  2 11:07:03 1999
From: gcollins99 at my-deja.com (gcollins99 at my-deja.com)
Date: Fri, 02 Jul 1999 16:07:03 +0100
Subject: RFC822 & Text of Mail Message
Message-ID: <00001673@bossar.com.pl>

From: gcollins99 at my-deja.com

Greetings!

  Here is what I want to do but cannot find out how. I am using RFC822
to parse out the headers and store them somewhere which works just fine.
But I am also in need of getting the text of the incoming message. So
what module would I use to also capture the message text. I see no way
of doing this via rfc822 module since it's not designed to do such
things. WOuld I use mimetools? And if so, how?

Thanks!

-Glen Collins



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From ivnowa at hvision.nl  Thu Jul  8 01:30:40 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Thu, 8 Jul 1999 05:30:40 GMT
Subject: Two Pythons talking to each other?
In-Reply-To: <199907080019.TAA22244@buffalo.fnal.gov>
References: <199907080019.TAA22244@buffalo.fnal.gov>
Message-ID: <199907080529.HAA19502@axil.hvision.nl>

On 7 Jul 99, Charles G Waldman wrote:

> There's no reason, just because the two programs are running on the same
> machine, that they can't talk to each other using network protocols
> (TCP/IP, sockets, etc).  You can use the "loopback" address of 127.0.0.1
> to make all connections.  You really don't want to develop and test using
> a completely different communications scheme (e.g. pipes, files) then
> re-implement everything for the networked version, and debug it all over
> again.  Client and server can both run on the same box.

I just tested this, and it works. But now I have another problem & 
question (which may be stupid, but I know hardly anything about 
networking and sockets). My programs send commands to each other in 
IRC style (like "/say this") or just plain text (chat). For receiving 
these strings, I use a non-blocking socket connection which polls the 
server every 0.x seconds (the period may vary). However, when the 
strings are sent too fast, which is the case with the 127.0.0.1 
address, and sometimes with regular connections too, the other side 
will pick two strings up as one string.

So my question is, how does one generally deal with this? Should I 
attach a newline (\n) (or maybe another separator character) after 
every string? Or are there other ways to guarantee that two commands 
will arrive at the other side as two strings?

TIA,

--Hans Nowak (ivnowa at hvision.nl)
Homepage: http://fly.to/zephyrfalcon



From violet at americasm01.nt.com  Thu Jul  1 11:41:35 1999
From: violet at americasm01.nt.com (Dula, Debbie (EXCHANGE:RICH2:2C25))
Date: Thu, 01 Jul 1999 10:41:35 -0500
Subject: Name of function within the function?
Message-ID: <377B8C2F.80F44F9@americasm01.nt.com>

I've looked and looked, but can't find the answer to this ... I'm
sure it's right in front of me ...


I want the current function's name to be output in my error
messages, but can't figure out how a function can get its own name
(without already knowing it).  For example,

	def f1():
		print "Error in function", f1.__name__

is not really useful for what I need as I might as well print "Error
in function f1".  Is there any generic way I can get the current
function's name?

Thanks,
Debbie



From alan at littleford.net  Sat Jul 10 18:41:10 1999
From: alan at littleford.net (alan at littleford.net)
Date: Sat, 10 Jul 1999 22:41:10 GMT
Subject: Another reason assignment expressions would be nice (?)
Message-ID: <3787c452.1487712890@news.slip.net>

I need to generate some large text files containing test cases for a
project. I decided to exploit the formatted print '%' and the exec()
function. The whole thing is wrapped up with a class (say foo)
instance which supplies useful functions (such as SomeFunction()) so I
can do things like

	a = foo()
	a.setTemplate("'A test line %d' % self.SomeFunction()")
	...
	a.run(100)

Calling run applies exec to the defined templates and so
the idea is the result of this would be hundred lines of: 

	A test line <somenumber>

(Of course the actual example would be more complicated, and class foo
provides loads of useful generator functions). Now here is the rub, I
might want two lines generated with a related value. Lets assume
SomeFunction() throws a random number, so I can't just call it again,
so what I would like to do is this:

	a = foo()
	a.setTemplate("'A test line %d' % 
		(self.i = self.SomeFunction()"))
	a.setTemplate("'Same value as above %d' % self.i")
	...
	a.run(100)

But of course I can't do this because '=' is a statement in the first
template. I could define foo to have some 'assignation' function:

def assign(var, val)
	var[0] = val
	return val

and then use
	
     "'A test line %d' % self.assign(self.var, self.SomeFunction())"
    " 'Same value as above %d' % self.var[0]"

but then the thing gets messy because I have to declare the mutable
array self.var somewhere before using it. All in all it seems I have
to jump through hoops because i = 3 doesn't 'return' the value 3.

Or am I missing something frightfully easy here ..

Tnx
Alan Littleford
alan at littleford.net




From David.sparrman at delphi.se  Thu Jul  1 08:34:06 1999
From: David.sparrman at delphi.se (David.sparrman at delphi.se)
Date: Thu, 01 Jul 1999 13:34:06 +0100
Subject: Out of Office AutoReply: Daily Digest: comp.lang.python 1/2
Message-ID: <00001600@bossar.com.pl>

From: David.sparrman at delphi.se

Jag har f=F6r n=E4rvarande semster
=C5ter p=E5 kontoret m=E5ndag den 25 juli

/mvh David=20



From William.Tanksley at p98.f112.n480.z2.fidonet.org  Fri Jul  2 12:58:37 1999
From: William.Tanksley at p98.f112.n480.z2.fidonet.org (William Tanksley)
Date: Fri, 02 Jul 1999 17:58:37 +0100
Subject: Tuples cf. Expanded objects??
Message-ID: <0000167d@bossar.com.pl>

From: wtanksle at dolphin.openprojects.net (William Tanksley)

On Fri, 02 Jul 1999 22:10:46 +1000, Gary Stephenson wrote:

>> There is probably something in the the FAQ.  BAsically, tuples are stored
>> more compactly and can be used as dictionary keys (because they ARE
>> readonly).

>How similar are Python's "Immutable Tuples" to Eiffel's "Expanded
>objects".  If one was to try to implement immutable tuples in another
>language, would an "expanded objects" implementation be the best, or even
>appropriate?

No similarity.  Expanded objects are objects which are handled by value
rather than reference.  Immutable types (which include tuples) are types
whose dimensions cannot change.

An instance of an immutable type may be created in expanded form.

>gary

-- 
-William "Billy" Tanksley



From greg.ewing at compaq.com  Mon Jul 19 16:31:36 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 20 Jul 1999 08:31:36 +1200
Subject: Running 2 gui libs simultaneously
References: <1280251163-5463910@hypernet.com> <378BEC9C.271A69EB@compaq.com> <5lwvw1czzf.fsf@eric.cnri.reston.va.us>
Message-ID: <37938B28.E1E1C844@compaq.com>

Guido van Rossum wrote:
> 
> I don't think that's the problem, at least not on Unix.
> (On the other hand, I don't know if this perhaps is a real problem on
> Windows?)

You're probably right about Unix. But it would definitely
be a problem on the Mac, since there is just one event
queue per process.

I don't know enough about Windows internals to comment.

Greg



From greg.ewing at compaq.com  Tue Jul 13 21:49:16 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 14 Jul 1999 13:49:16 +1200
Subject: Running 2 gui libs simultaneously
References: <1280251163-5463910@hypernet.com>
Message-ID: <378BEC9C.271A69EB@compaq.com>

Gordon McMillan wrote:
> 
> You don't have to use Tk's mainloop. You can call update() in a loop.
> Perhaps wxWindows has something similar, or you can call Tk's update
> from within some idletask in wxWindows.

I don't fancy your chances of getting that to work.
The problem isn't just getting two main loops to
run at the same time -- it's making sure that each
event goes to the right UI manager.

You might be able to do it by reading events
yourself, figuring out somehow which UI they
are meant for, and somehow passing them on to
the relevant toolkit. Would be platform-dependent,
since you'd have to deal with events at a low
level.

Greg



From wolford at enews.nrl.navy.mil  Fri Jul  9 10:40:57 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Fri, 09 Jul 1999 10:40:57 -0400
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com> <37850F5B.F80CB0A7@enews.nrl.navy.mil> <7m3ajc$s3q$1@m2.c2.telstra-mm.net.au>
Message-ID: <7m9tmd$6qg$25@rks1.urz.tu-dresden.de>

Mark Hammond wrote:

> Scott Wolford wrote in message <37850F5B.F80CB0A7 at enews.nrl.navy.mil>...
>
> >msvcrt.lib and the application can't then bring in msvcrtd.lib. I do think
> >this is a problem because python15_d.lib is provided but no python15_d.dll
> >is. I would think you'd provide (at the very least) python.lib linked with
> >msvcrt.lib, and python_d.lib linked with msvcrtd.lib (and the dlls), or
> >(to be really nice :) all four combinations.
>
> The main reason we dont is size - we would need to ship the _d, and for
> people to really use it, also the .pdb.  These add too many megs.
>
> Further, almost _everyone_ who needs to use this _d version already has MSVC
> and already has the Python sources.  So building there own is trivial.

True. That's how I finally figured out my problem.

> Further, "all 4 combinations" doesnt work.  As you found out painfully, is
> it _critical_ that all Python systems share the same C runtime library.
> This is the whole point of the _d fiasco - to prevent the mixing and
> matching.

How about:
python.lib/dll and msvcrt.lib/dll
python.lib/dll and msvcrtd.lib/dll
python_d.lib/dll and msvcrt.lib/dll
python_d.lib/dll and msvcrtd.lib/dll

separated into two directories: msvcrt and msvcrtd

> We obviously need to document this process better - Im not sure where tho -
> what documentation did you use to get this far?

Not much, just figured it out myself mostly. I was only off by one character
after all: msvcrtd.lib instead of msvcrt.lib.

> It is also worth noting that Gordon pointed this out to you before this
> particular thread, and you still had the problem.  Im not sure that
> documenting it would be effective in all cases.

He said to use the Multithreaded DLL. Unfortunately, msvcrt.dll and msvcrtd.dll
are *both* the Multithreaded DLL; one is debug the other is not. It never
occured to me (well, it finally did) that it mattered whether msvcrt.dll or
msvcrtd.dll were used. I assumed that the semantics of the two DLLs were
essentially the same, just the implementation differed slightly. I'm pretty sure
(though I haven't tested it, maybe you'll know) that you can mix and match
optimized and debug code pretty easily under Unix.

> > Don't get me wrong; I think
> >it's a great service to provide the pre-compiled package. I just want to
>
> Well, FWIW, I will be making these _d versions (and all the win32 extensions
> _d versions) available in a seperate .zip file as a Registered User benefit
> (http://starship.python.net/crew/mhammond/registration).  I havent started
> this particular benefit yet tho.
>
> Mark.

I think I'll just build it myself from now on, though thanks for the service. I
usually have trouble compiling windows software but Python was smooth as smooth
can be.





From olli at rhein-zeitung.de  Sat Jul  3 14:14:02 1999
From: olli at rhein-zeitung.de (olli at rhein-zeitung.de)
Date: Sat, 03 Jul 1999 19:14:02 +0100
Subject: Curses, Ncurses, Slang, etc?
Message-ID: <000016ae@bossar.com.pl>

From: olli at rhein-zeitung.de

Hi,

I can only tell you that the ncurses module I wrote works for me on my RedHat 6
and Mandrake 6
systems. What problems do you have?

If support and so on is an issue, then I would use the slang bindings at the
moment.

Bye, Oliver

Eric Kidd <eric.kidd at pobox.com> wrote:
> I'd like to write a Python application which uses either curses, ncurses,
> slang or a similar tty interface library. Unfortunately, the built-in
> curses module is hopeless broken on my RedHat 6.0 system (it doesn't export
> 'addch', for example).

> Neither the ncurses module or the pwin module work out of the box on my
> Linux system. Are there any well-supported Python modules for
> curses/ncurses/slang which do work out of the box?

> Thank you for any help you can provide!

> Cheers,
> Eric




-- 
Oliver Andrich, RZ-Online, Schlossstrasse Str. 42, D-56068 Koblenz
Telefon: +49-261-39210-40 Fax: -33 |       http://rhein-zeitung.de 
                                   |            http://andrich.net
GMC d- s++:++ a- C++ UL++++ P-- L+++ E--- W++ N+ w-- O- M-- 
V- PS++ PE Y+ PGP+ t 5+ X R+ tv+ b++ DI++ D- G e h- r z+



From phd at emerald.netskate.ru  Tue Jul  6 11:45:29 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Tue, 6 Jul 1999 19:45:29 +0400 (MSD)
Subject: AttributeError - a dumb question
In-Reply-To: <Pine.LNX.3.96.990706171320.374A-100000@Johann>
Message-ID: <Pine.LNX.4.04.9907061943010.21542-100000@emerald.netskate.ru>

On Tue, 6 Jul 1999, Johann Spies wrote:
> Why does the following code produce an attribute error?
> class F:
>     def __init__(self):
>           x = self.Krylys()

   This will create local var x. I am pretty sure you wanted to write it

           self.x = self.Krylys()

>     def KryLys(self):
>         l = [1,2,3,4,5]
>         return l
>     
> 
> l = F

   You've created reference to the class definition, instead of creating
class instance:

l = F()

> Johann
> 
> 
>  --------------------------------------------------------------------------
> | Johann Spies                                 Windsorlaan 19              |
> | jhspies at futurenet.co.za                3201 Pietermaritzburg             |
> | Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
>  --------------------------------------------------------------------------
> 
>      "Ask, and it shall be given you; seek, and ye shall
>       find; knock, and it shall be opened unto you."        
>                                   Matthew 7:7 

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From aahz at netcom.com  Thu Jul 22 20:07:07 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 23 Jul 1999 00:07:07 GMT
Subject: Dreaded Newbie Question
References: <K4tK+KAA26l3EwvK@epstein.co.uk>
Message-ID: <7n8bnb$eni@dfw-ixnews5.ix.netcom.com>

In article <K4tK+KAA26l3EwvK at epstein.co.uk>,
Dani Epstein  <spamkiller at epstein.co.uk> wrote:
>
>A newbie to Python (though currently coding in Delphi), I'm having a
>little difficulty tracking down stuff, although I have the tutorial and
>manuals. Where is the best place to get more info, and why is there no
>newsgroup for newbies like myself?

There's a mailing list.  Look on www.python.org for the "tutor list".

>For the moment though my grateful thanks goes to anyone who can tell me
>how to get a file's size, and where that info is to be found in the
>references.

Should be os.stat().
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From skip at mojam.com  Tue Jul 27 15:19:06 1999
From: skip at mojam.com (Skip Montanaro)
Date: Tue, 27 Jul 1999 14:19:06 -0500 (CDT)
Subject: Python Performance
In-Reply-To: <14238.335.458728.140169@lrz.de>
References: <3794C642.AB6A3E5D@freenet.co.uk>
	<7n5f0c$snu$1@ffx2nh5.news.uu.net>
	<al81zdxkktf.fsf@sirppi.helsinki.fi>
	<p51zdvn4c7.fsf@bidra241.bbn.hp.com>
	<14236.32616.980578.536614@dolphin.mojam.com>
	<379D89BC.DA91BA8C@hp.com>
	<14237.60407.130426.461846@dolphin.mojam.com>
	<14238.335.458728.140169@lrz.de>
Message-ID: <14238.1372.334823.378192@dolphin.mojam.com>

    Eugene> Skip Montanaro writes:
    >> True.  I was thinking about the situation I'm in (small developer, a few
    >> machines to maintain/upgrade/replace).

    Eugene> What's wrong with recoding the hot spots in C? Python is
    primarily a rapid prototyping/scripting/glue language, it was never
    meant as a sole implementation language for applications.

There's nothing wrong with recoding hot spots in C.  Many people do it all
the time.  Practically speaking, however, many Python users don't have the
technical depth or other resource (VC++, for instance) to write chunks of
their applications in C or C++.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From robin at jessikat.demon.co.uk  Sun Jul 18 06:36:25 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sun, 18 Jul 1999 11:36:25 +0100
Subject: Python and Boehm-Demers GC, I have code.
References: <7mr0n1$dmk@ds2.acs.ucalgary.ca>
 <000e01bed0fa$04316520$36a02299@tim>
Message-ID: <o+zLoAAp4ak3Ewjk@jessikat.demon.co.uk>

In article <000e01bed0fa$04316520$36a02299 at tim>, Tim Peters
<tim_one at email.msn.com> writes
>[Tim, speculates on why pystone/pybench aren't much affected by BDW w/
> refcounting left on]
>
...
>So screw it <wink>.  Find someone with an important program that's leaking
>memory and see whether this fixes *that*.  This (not speed) is BDW's real
>strength anyway.  Chalk up a few victories there, and selling will be much
>easier.
>
>A cautionary tale:  Over the past month I've been poking away at tracking
>leaks in IDLE.  After building a capable-enough cycle-finding tool (a
>feature-bloated but much faster offshoot of Lars's Plumbo.py), I found
>cycles all over the place.  As things turned out, though, hardly any of them
>could be considered trash!  The way functions point to their module globals,
>which in turn point to imported modules, some of which eventually point into
>sys, from whose sys.modules everthing can be reached ... an isolated cycle
>was very rare indeed.
>
>It came as a real surprise to me that "real GC" wouldn't have helped.  This
>may be a particularly nasty problem for GUI programs, so I'm not too quick
>to generalize.  OTOH, don't be too quick to discount it either <wink>.
>
...
very interesting; I'm spending a bit of time on Zope at the moment. This
if anything will be a long running thing which shouldn't leak memory.
Any chance of an analysis? Or the (I'm sure all the features are good)
version of plumbo?
-- 
features-to-one-are-bugs-to-others-ly yrs
Robin Becker



From un4e at rz.uni-karlsruhe.de  Tue Jul 20 17:18:03 1999
From: un4e at rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 20 Jul 1999 21:18:03 GMT
Subject: Embedding: Creating / Deleting CObjects?!
Message-ID: <7n2p2b$bmt$1@news.rz.uni-karlsruhe.de>

Hi,

I'm truly baffled. I'm trying to define a backend for my embedded Python
system, and the corresponding C routines seem to work, but the program
crashes somewhere in the Python library routines, dances the Fandango all
over my internal object representation, and whatnot. 

I'm posting a couple of fragments of my code, just so that somebody can tell
me whether this is the canonical way of handling things (this side of Python
is not exactly well documented...) 

/* Content destructor function */
static void
PLPythonDestructor(proplist_t obj)
{
  PLRelease(obj);
}

/* CObject Creator */
static PyObject *
PLPythonNewInstance(PyObject *self, PyObject *args)
{
  proplist_t plClassName, plArgs;
  instance_t instance;
  char *className;
  PyObject *pyArgs, *retval;

  printf("NewInstanceVar\n");

  if(!PyArg_ParseTuple(args, "sO", &className, &pyArgs))
    return NULL;

  plArgs = PLPy2PLObject(pyArgs); /* NULL is OK here */
  plClassName = PLMakeString(className);
  instance = PLObjectNewInstance(plClassName, plArgs); /* [My own braindead
							  class system.
							  Works, though.] */
  if(plClassName)
    PLRelease(plClassName);
  if(plArgs)
    PLRelease(plArgs);

  if(!instance)
    {
      PyErr_SetString(PyExc_RuntimeError, "Couldn't create new instance");
      return NULL;
    }
  retval = PyCObject_FromVoidPtr((void *)instance,
				 PLPythonDestructor);
  Py_INCREF(retval);
  return retval;
}

/* CObject Destructor */
static PyObject *
PLPythonDeleteInstance(PyObject *self, PyObject *args)
{
  instance_t instance;
  PyObject *instanceObject;

  printf("DeleteInstance\n");

  if(!PyArg_ParseTuple(args, "O", &instanceObject))
    return NULL;

  /*
  instance = (instance_t) PyCObject_AsVoidPtr(instanceObject);
  PLObjectDeleteInstance(instance);
  */

  Py_XDECREF(instanceObject);
  return Py_None;
}

If anything about the above code remains unclear, please ask. I WANT to get
this stuff to work!

TIA,
		--Bjoern
-- 
--------------------------------/\--One OS to rule them all---Windows NT-----
Bjoern Giesler                 /  \  One OS to find them           
<un4e at rz.uni-karlsruhe.de>    / <> \  One OS to bring them all    
-----------------------------/______\--And in the Darkness bind them---------
    Thank you for your time, worship the Antichrist, and have a nice day.



From kuncej at mail.conservation.state.mo.us  Thu Jul  8 10:22:50 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Thu, 08 Jul 1999 09:22:50 -0500
Subject: PIL anim gif
Message-ID: <s7846df1.032@mail.conservation.state.mo.us>

I recently had to create some animated GIFs.
I used PIL to create the individual frames.
Then I used a 3rd-party command-line utility
to create the animations:
   WhirlGif 2.01
   http://www.msg.net/utility/whirlgif/ 
I built it on both NT and AIX with no problem.
I could supply NT executable and sample python
scripts on request.

My only complaint is that PIL can't produce
compressed GIFs. Maybe the new version of
PIL (due this weekend?) will have more
GIF animation support.

  --Jeff

>>> Per Kistler <kistler at fnmail.com> 07/07/99 09:31AM >>>
Hi All

How to make an animated gif from several Image
objects in PIL? According to the docu, it seems,
that it can only anim gifs which are such already.

Regards, Per.
-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net 
------------------------------------------------------------




From gerrit.holl at pobox.com  Fri Jul  2 06:31:52 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 2 Jul 1999 12:31:52 +0200
Subject: Why does shelve make such large files?
In-Reply-To: <377BF01D.16F8AA33@cup.hp.com>; from Ovidiu Predescu on Thu, Jul 01, 1999 at 10:47:57PM +0000
References: <19990701211450.A6827@optiplex.palga.uucp> <377BF01D.16F8AA33@cup.hp.com>
Message-ID: <19990702123152.B9975@optiplex.palga.uucp>

On Thu, Jul 01, 1999 at 10:47:57PM +0000, Ovidiu Predescu wrote:
> Gerrit Holl wrote:
> 
> > Is it really necesarry for shelve to make such large files?
> > Have a look at this:
> > /tmp> python
> > Python 1.5.2 (#1, Apr 18 1999, 00:16:12)  [GCC 2.7.2.3] on linux2
> > Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> > >>> import shelve
> > >>> d = shelve.open('database')
> > >>> d['key'] = 'value'
> > >>>
> > /tmp> ls -l database
> > -rw-rw-r--   1 gerrit   gerrit      16384 Jul  1 21:13 database
> >                                     ^^^^^
> > 
> > 16 KB for only one key!!?
> > pickle seems to make _much_ smaller files!
> > 
> > Why is this?
> 
> The shelve module uses DBM which is like a small database that allows        <
> you to store objects and search it for a given key. DBM allows you to        <
> store millions of objects and search for them later without requiring        <
> you to load all the file in memory first.                                   <
> 

Interesting...

> The pickle module on the other hand is serializing the objects with the
> purpose of deserializing them _all_ from the file later. Pickle does not
> offer you any way to search for data based on a key, you have to do this
> yourself after the objects have been created from the file. This is
> opposed to the way shelve handles this, all the key accesses and
> insertions in a shelve object are actually reads or writes to or from
> the DBM file.
> 
> And to answer your question, DBM is creating these big files because of
> the way it manages the database. The data in the database file could
> have gaps as a result of multiple insertions and deletions. Pickle's
> data in files is a simple representation of the objects that were
> written and there is no way to update the file other than rewriting it
> entirely.
> 

Ah, I understand.
So pickle is useful for very small datases, but when they're really huge, one
should use shelve. Isn't it?

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From mwh21 at cam.ac.uk  Thu Jul  1 19:17:47 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 02 Jul 1999 00:17:47 +0100
Subject: print color strings?
References: <19990701213528.A8877@optiplex.palga.uucp>
Message-ID: <m3hfnoc63o.fsf@atrus.jesus.cam.ac.uk>

Gerrit Holl <gerrit.holl at pobox.com> writes:

> Hello,
> 
> is there an easy way to print color strings in python?
> 
> A function like printcolor('brightyellow', 'red', 'spam'), that finds
> out on what kind of terminal you are and prints it to stdout?

That sort of finding out is what terminfo is for; there's no Python
interface that I know of (such a beast shouldn't be too hard to write,
but a bit pointless, I suspect). However there is the `tput' program;

os.system("tput setaf 1")

should turn your foreground colour red. 

Read the terminfo man pages for more codes.

You'd probably want to cache the escape codes rather than spinning a
new process each time you need one.

codes['redf'] = os.popen("tput setaf 1").read()
codes['redb'] = os.popen("tput setab 1").read()
codes['greenf']=os.popen("tput setaf 2").read()
...

and so on.

I'm certainly not an expert on such matters; reading man pages for
tput, terminfo and term and playing around will get you exactly where
I've got to.

All terribly unix-centric, of course.

HTH
Michael




From kernr at mail.ncifcrf.gov  Sun Jul 11 12:58:07 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Sun, 11 Jul 1999 16:58:07 GMT
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787D8DA.3774AA7D@ssec.wisc.edu> <7m8p7t$o7s@dfw-ixnews10.ix.netcom.com> <7m9to7$6qg$83@rks1.urz.tu-dresden.de>
Message-ID: <3788cc5e.8912779@news.erols.com>

On Sat, 10 Jul 1999 21:02:58 -0500, Nick Bower
<nick.bower at ssec.wisc.edu> wrote:

>yeah, that's what i originally thought.  sure enough there were some updates in the
>os documentation, but not os.chdir.  it's something that would be really helpfull
>if anyone linked to the documentation sees this.

It's in Section 6.1.1 "Process Parameters" lib/os-procinfo.html.  It
might be a good idea, though, to place it with os.mkdir and os.rmdir
in Section 6.1.4 "Files and Directories" lib/os-file-dir.html.

If you think so, e-mail the maintainer at python-docs at python.org; he
takes suggestions very well.

>nick

[snip]

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From svoynow at yahoo.com  Tue Jul 20 18:34:34 1999
From: svoynow at yahoo.com (Sasha Voynow)
Date: Tue, 20 Jul 1999 15:34:34 -0700
Subject: printing with Python/Windows
Message-ID: <FP6l3.3374$K3.72023@nuq-read.news.verio.net>

What's the simplest way to print a file (in this case an image)
from Python on WinNT. I'm not really versed in the win32api
or MFC. But as a stopgap measure I've got to get this to work on a win32
platform. I scoured the Pywin / win32 extensions docs, but to no avail.

Thanks,

Sasha Voynow





From Rick at tiac.net  Thu Jul 29 19:19:59 1999
From: Rick at tiac.net (Rick at tiac.net)
Date: 29 Jul 1999 18:19:59 -0500
Subject: I need to speak with a webmaster  1090
Message-ID: <37a0d38f@news.mdc.net>

If you own a small business you owe it yourself to take a 1 minute stop at http://www.mdc.net/~rick/index2.html otherwise you never know what might haave been. Is less than 1 minute to much to ask to improve your business 95%? Also if your a webmaster this is just great for you to pick up some serious cash as well. The choice is yours, be the future or history, choose wisely.
wfykdhpbupsotmcdodh




From olivier.deckmyn at mail.dotcom.fr  Tue Jul 20 03:00:30 1999
From: olivier.deckmyn at mail.dotcom.fr (Olivier Deckmyn)
Date: Tue, 20 Jul 1999 09:00:30 +0200
Subject: [Q] Accessors...
References: <7mvbkj$pl$1@feed.teaser.fr> <37939CC2.BD5A1AC9@compaq.com>
Message-ID: <7n16ue$66u$1@feed.teaser.fr>

Thanx A LOT greg :)

I had started to code such a system after reading the "Getters and Setters"
thread in this newsgroup (1999/07/15)

Have a nice day.

Greg Ewing <greg.ewing at compaq.com> a ?crit dans le message :
37939CC2.BD5A1AC9 at compaq.com...
> Olivier Deckmyn wrote:
> >
> > This way, because it exists two methods get_XXX and set_XXX, a special
> > proporty XXX would be implicitly created, and would make the user call
> > "automatically" the get/set methods.
>
> Yes, there is a way. If a class has a method
> called __getattr__, it is automatically called
> whenever you try to get the value of an attribute
> which doesn't exist. So you can do something like
>
>   class AutoAccess:
>
>     def __getattr__(self, name, *args, **kw):
>       return apply(getattr(self, "get_"+name), args, kw)
>
> then any class which inherits from AutoAccess will
> have any attempt to get a non-existent XXX attribute
> redirected to a get_XXX method. (Note that the above
> implementation is a bit simplistic -- it will enter an
> infinite loop if there is no get_XXX method defined
> in the class.)
>
> There is a corresponding feature for setting an
> attribute: if there is a __setattr__ method, it gets
> called on every attempt to set the value of an
> attribute, so you can do
>
>     def __setattr__(self, name, value):
>       try:
>         setter = getattr(self, "set_"+name)
>       except AttributeError:
>         setter = None
>       if setter:
>         setter(value)
>       else:
>         self.__dict__[name] = value
>
> This is more complicated because the __setattr__
> method, if present, is called *before* checking
> whether the attribute exists (by necessity, since
> setting an attribute which doesn't already exist
> is perfectly legal). Hence the code to catch the
> case where there is no set_XXX method and set the
> attribute directly.
>
> As you can see, what you are asking for can be
> done, but it is expensive -- attribute access
> through these methods will be considerably slower
> than just calling get_XXX and set_XXX methods
> directly. You'll have to decide whether it's
> worth the price in your application.
>
> Warning: The above code is for illustration
> purposes only -- I haven't tested it!
>
> Greg





From Richard.Jones at fulcrum.com.au  Tue Jul 20 21:15:15 1999
From: Richard.Jones at fulcrum.com.au (Richard Jones)
Date: Wed, 21 Jul 1999 01:15:15 GMT
Subject: where is gfplus? 
In-Reply-To: Message from Greg McFarlane of 1999-Jul-21 10:6:17,
            <19990721100617.07994@nms.otc.telstra.com.au> 
References: <19990721100617.07994@nms.otc.telstra.com.au>
Message-ID: <199907210115.LAA31985@envy.fulcrum.com.au>

[Greg McFarlane]
> I have been trying to get through to the gfplus site
> <http://shamrock.med.nyu.edu/~jeff/gfplus/> but it seems to be down. 
> Does anyone know of anywhere else I can get gfplus (an interactive
> gadfly shell)?  Or can anyone send it too me (it can't be that big:).
> 
> Also, is there a gui interface to gadfly?  Something that lists all
> the tables, all the columns in a table, all the data in a table,
> indexes, etc.

   DBList.py is a Gtk-based GUI I wrote that does table query and view - pretty 
simple stuff I wrote while learning Gtk. I'll be writing a Qt version some time 
soon I suppose since I find this kind of thing handy. DBList.py is available 
through the gadfly contributions page:

     http://www.chordate.com/gfcontrib.html
     http://www.deja.com/getdoc.xp?AN=406461470

   I've put gfplus 1.0 (I don't seem to have 1.1 anywhere) at:

     http://redback.spyda.net/~richard/gfplus

   I've also emailed Jeff to see what's up...


       Richard





From bucodi at worldnet.fr  Fri Jul  2 06:22:52 1999
From: bucodi at worldnet.fr (pserrand)
Date: Fri, 02 Jul 1999 12:22:52 +0200
Subject: test - ignore
References: <377C7879.428A3E19@worldnet.fr>
Message-ID: <377C92FC.9F3E6A08@worldnet.fr>

pserrand wrote:

> test with python

reply test




From lvirden at cas.org  Fri Jul  2 08:04:54 1999
From: lvirden at cas.org (lvirden at cas.org)
Date: Fri, 02 Jul 1999 13:04:54 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001666@bossar.com.pl>

From: lvirden at cas.org


According to Vadim Zeitlin <zeitlin at dptmaths.ens-cachan.fr>:
:
: The sad truth is that macs are not very common nowadays. I would like to help
:develop wxMac, but I don't have one (nor at home nor anywhere near me in the

I wonder why various industry rags have proclaimed during the past
10 months about Apple's G3 machines outselling other vendor
machines.  ...


-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.



From k_ganesan at yahoo.com  Thu Jul  1 03:32:03 1999
From: k_ganesan at yahoo.com (Karthik Ganesan)
Date: Thu, 1 Jul 1999 00:32:03 -0700
Subject: idle- setting breakpoints
Message-ID: <930814504.1858126659@news.earthlink.net>

Hi,

To set a breakpoint under idle:
-have a "Python Shell" (the one which opens up when you fire up idle) open
-Select Debug->Debugger, a window should pop open having the GUI for the debugger
Now you can set breakpoint in your code.

If you dont have this debug window open, you get a bell from the function
PyShellEditorWindow.set_breakpoint_here(self, event=none) as the attribute
self.flist.pyshell.interp.debugger is <None>.

It seems that the function PyShell.toggle_debugger() (called when you select the above
menu option) gives this attribute a valid value.

-karthik





From mal at lemburg.com  Thu Jul 22 09:17:02 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 22 Jul 1999 15:17:02 +0200
Subject: ISP CGI  Python Support
References: <BFql3.4593$2D1.40042@newsr2.san.rr.com>
Message-ID: <379719CE.BC3A2B53@lemburg.com>

Bruce Duncan wrote:
> 
> Greetings,
> 
> I would like to see a short document that I can send to my Web hosting ISP
> to convince them to support CGI programming in Python.  Some ISP's do,
> but others don't.  Because this may be of common concern, perhaps such
> a Python "pitch" can be stored at python.org in addition to a list of
> providers
> that do support Python CGI.

How about telling them how easy it is to install Python at their
site... you already know the link to the mxCGIPython project.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   162 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org  Fri Jul  2 13:57:00 1999
From: Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 18:57:00 +0100
Subject: Python GUIs: Summary and Conclusion
Message-ID: <00001682@bossar.com.pl>

From: mikael at pobox.com (Mikael Lyngvig)

Hi William,

On Fri, 02 Jul 1999 16:49:34 GMT, wtanksle at dolphin.openprojects.net
(William Tanksley) wrote:

>I'm working on a 2D tile-based adventure game, which right now is text
>only.  I'm expanding it so that it can use graphics, but I need some
>libraries.  Let me know what you find!

Well, I don't expect to find anything overly suitable for 2D
tile-based adventure games.  I stopped working on that project three
years ago.

However, it's kind of funny you write just now: I was considering to
resume that old project in Python using wxWindows earlier today.

I don't know if wxWindows is what you need, but that's probably where
I would be looking, if I was to resume my old project.

>>[1] Which I expect to change in the future, which is why I'm reviewing
>>portable GUIs (I don't do non-portable stuff, if there's anything I
>>can do to avoid it).
>
>That's right.

You said it :)  That pretty much leaves you with the choice between
TkInter, wxWindows, and FOX (from what I know).  I've looked, more or
less, briefly on all of them and decided to use wxWindows.  

For a 2D tile-based game you pretty much just need to be able to draw
tiles (square bitmaps) quickly.  I think wxWindows can do that, but
perhaps you should make some serious tests of those three, and the
others recommended by others.  

And let us know what you find :)


-- Mikael



From ralph.gauges at eml.villa-bosch.de  Wed Jul 14 11:29:52 1999
From: ralph.gauges at eml.villa-bosch.de (Ralph Gauges)
Date: Wed, 14 Jul 1999 17:29:52 +0200
Subject: Recursive method
References: <1280168234-10452044@hypernet.com> <378CA422.B2E99CB7@eml.villa-bosch.de> <004901bece0b$46982210$f29b12c2@secret.pythonware.com>
Message-ID: <378CACF0.373116E@eml.villa-bosch.de>

> 
> probably because there is no instance when the
> class definition is executed...
> 
> ("class" is a *statement*, not a declaration, and
> is executed when the script is run.  see:
> http://www.python.org/doc/current/ref/class.html
> for details.  "def" is also a statement, btw...)
> 
> </F>

This sounds right, but when I have a class that pops up a
gui on my screen, as I do now, it doesn't seem to make much
difference, wether I build the gui right in the class
definition or in the __init__ method. Both is executed when
I make an Instance, the only difference I could see is that
one time I had a self and the other time I didn't.
Maybe I am missing the point here, but I still don't get it
completely.

Ralph



From greg.ewing at compaq.com  Mon Jul 19 18:46:55 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 20 Jul 1999 10:46:55 +1200
Subject: Tk: Temporarily change cursor?
Message-ID: <3793AADF.5EDBAFBC@compaq.com>

Is there an easy way to temporarily change the
cursor in Tk while a piece of code runs? Or
even a hard way that always works?

There doesn't seem to be any way to directly
change the cursor currently displayed -- you
have to change the cursor of a particular widget.

I've had some success with changing the cursor
of a widget and then doing an update, but it
only works reliably if I pick exactly the right
widget, which is difficult, especially if the
spot in the code where I want to do this is
far removed from the user action which
initiated it.

This is such a commonly done thing, surely
there must be some provision for it in Tk?

Windows, Python 1.5.2 and Tk 8.0, by the way.

Thanks for any help,
Greg



From martinp at mincom.com  Fri Jul 30 03:06:36 1999
From: martinp at mincom.com (Martin Pool)
Date: Fri, 30 Jul 1999 17:06:36 +1000
Subject: Need help with TCP/IP client access from Windows
References: <slrn7q1ej3.18d.esr@snark.thyrsus.com> <NDBBIAPHEKGOBIHCBLIIOECECAAA.mcfletch@vrtelecom.com> <19990730005536.U7149@thyrsus.com>
Message-ID: <37A14EFC.B72AAA67@mincom.com>

"Eric S. Raymond" wrote:
> 
> Mike Fletcher <mcfletch at vrtelecom.com>:
> > About the only way I know to do this at all reliably...
> >
> > import os
> > os.system( 'start %s'%url )
> >
> > Which does whatever the user has defined as the proper thing to do with
> > urls, normally, launch IE or Netscape.
> 
> Really?  That sounds perfect.  Just for my information, what is `start'?

To a unix person, you could say that it forks into another task, 
uses a heuristic to work out how to handle the arguments, and 
executes that program.  It's a bit bizarre.

The help from NT4WSsp5 is below.

-- 
 /\\\  Mincom | Martin Pool          | martinp at mincom.com
// \\\        | Software Engineer    | Phone: +61 7 3303-3333
\\ ///        | Mincom Ltd.          | 
 \///         | Teneriffe, Brisbane  | Speaking for myself only



> help start
Starts a separate window to run a specified program or command.

START ["title"] [/Dpath] [/I] [/MIN] [/MAX] [/SEPARATE | /SHARED]
      [/LOW | /NORMAL | /HIGH | /REALTIME] [/WAIT] [/B]
[command/program] [param
eters]

    "title"     Title to display in  window title bar.
    path        Starting directory
    I           The new environment will be the original environment
passed
                to the cmd.exe and not the current environment.
    MIN         Start window minimized
    MAX         Start window maximized
    SEPARATE    Start 16-bit Windows program in separate memory space
    SHARED      Start 16-bit Windows program in shared memory space
    LOW         Start application in the IDLE priority class
    NORMAL      Start application in the NORMAL priority class
    HIGH        Start application in the HIGH priority class
    REALTIME    Start application in the REALTIME priority class
    WAIT        Start application and wait for it to terminate
    B           Start application without creating a new window. The
                application has ^C handling ignored. Unless the
application
                enables ^C processing, ^Break is the only way to
interrupt the
                application
    command/program
                If it is an internal cmd command or a batch file then
                the command processor is run with the /K switch to
cmd.exe.
                This means that the window will remain after the command
                has been run.

                If it is not an internal cmd command or batch file then
                it is a program and will run as either a windowed
application
                or a console application.

    parameters  These are the parameters passed to the command/program


If Command Extensions are enabled, external command invocation
through the command line or the START command changes as follows:

non-executable files may be invoked through their file association just
    by typing the name of the file as a command.  (e.g.  WORD.DOC would
    launch the application associated with the .DOC file extension).
    See the ASSOC and FTYPE commands for how to create these
    associations from within a command script.

When executing an application that is a 32-bit GUI application, CMD.EXE
    does not wait for the application to terminate before returning to
    the command prompt.  This new behavior does NOT occur if executing
    within a command script.

When executing a command line whose first token is CMD without an
    extension or path qualifier, then replaces CMD with the value of the
    COMSPEC variable, thus avoiding picking up random versions of
    CMD.EXE when you least expect them.

When executing a command line whose first token does NOT contain an
    extension, then CMD.EXE uses the value of the PATHEXT
    environment variable to determine which extensions to look for
    and in what order.  The default value for the PATHEXT variable
    is:

        .COM;.EXE;.BAT;.CMD

    Notice the syntax is the same as the PATH variable, with
    semicolons separating the different elements.

When executing a command, if there is no match on any extension, then
looks to see if the name, without any extension, matches a directory
name
and if it does, the START command launches the Explorer on that path.
If done from the command line, it is the equivalent to doing a CD /D
to that path.



From Fernando.Mato.Mira at p98.f112.n480.z2.fidonet.org  Thu Jul  1 12:23:52 1999
From: Fernando.Mato.Mira at p98.f112.n480.z2.fidonet.org (Fernando Mato Mira)
Date: Thu, 01 Jul 1999 17:23:52 +0100
Subject: Reflection in Python?
Message-ID: <0000160a@bossar.com.pl>

From: Fernando Mato Mira <matomira at iname.com>

If you are into reflection, you should really take a look at CLOS,
before setting things in stone in Python.

You can find my stuff in
http://ligwww.epfl.ch/matomira/en/publications.html (the last book was
finally released in '96 as "Advances in Metalevel Architectures and
Reflection"). For DB interface stuff, there was
paper in LUV'95 I think.





From greg.ewing at compaq.com  Mon Jul 19 16:57:07 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 20 Jul 1999 08:57:07 +1200
Subject: Strange side effect from reading a list of numbers from argv
References: <O3sk3.6047$Sg6.52046@news1.rdc1.sdca.home.com> <slrn7p4lit.d2e.jepler@craie.inetnebr.com>
Message-ID: <37939123.DA1AE2AD@compaq.com>

Jeff Epler wrote:
> 
> It's come up before that "string <
> int" doesn't cause an exception, but I don't remeber what the good reason
> for it was.

The reason is so that the built-in sort method
will work on a list of differently-typed objects.
Whether this is a "good" reason is a matter
of opinion...

Greg



From roy at popmail.med.nyu.edu  Fri Jul 30 13:18:28 1999
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Fri, 30 Jul 1999 13:18:28 -0400
Subject: Recursive constructors?
Message-ID: <roy-3007991318280001@qwerky.med.nyu.edu>

Is it legal for a constructor function to call itself recursively?  I.e.:

class foo:
   def __init__(self, x):
      if (whatever):
         self.bar = foo(y)

-- 
Roy Smith <roy at popmail.med.nyu.edu>
New York University School of Medicine




From tfb at tfeb.org  Fri Jul  2 16:27:49 1999
From: tfb at tfeb.org (Tim Bradshaw)
Date: 02 Jul 1999 21:27:49 +0100
Subject: Language extensibility
References: <1283099377-77267484@hypernet.com>
	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it>
	<876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org>
	<377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov>
	<slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net>
	<wkzp1g2t2e.fsf@ifi.uio.no> <ey3r9mskvja.fsf@lostwithiel.tfeb.org>
	<x27lojoums.fsf@todday.aiai.ed.ac.uk>
	<gZ6f3.650$6M6.197790@news.shore.net>
Message-ID: <ey3oghulrui.fsf@lostwithiel.tfeb.org>

* Michael P Reilly wrote:

> Supposedly, Lisp was created in 1953, a year after Fortran, and is said
> to be the second higher-level (non-assembly) language written.

No, I think it's 1958 -- the 40th anniversary conference was last
year, after all!

(A good year for programming languages and guitars both).

--tim




From amitp at Xenon.Stanford.EDU  Thu Jul 22 17:37:05 1999
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 22 Jul 1999 21:37:05 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
References: <001101bed3f2$5e8874a0$642d2399@tim> <7n675p$qsk@dfw-ixnews7.ix.netcom.com> <14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>
Message-ID: <7n82u1$jid$1@nntp.Stanford.EDU>

 Skip Montanaro  <skip at mojam.com> wrote:
| 
| What Tim said about programmers not inventing control structures is somewhat
| different than what Aahz put his foot down about.  There's nothing wrong
| with "from foo import *" when it's used in the right context.  Some modules
| in the standard distribution were designed specifically to support this
| (types and Tkinter come to mind).  Used willy-nilly, you can get into
| trouble with it, one-man show or not.
| 

Here's a gotcha with import *:

import re
from string import *  # Great whenever you're doing string processing
re.compile('abc')




:-(

I think the system modules should define all internal variables with a
'_' prefix! :-)

    - Amit




From janssen at parc.xerox.com  Wed Jul 28 18:53:29 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Wed, 28 Jul 1999 15:53:29 PDT
Subject: NameError on "buffer"?
In-Reply-To: <orbs0r4B0KGWRZV19A@holmes.parc.xerox.com>
References: <orbs0r4B0KGWRZV19A@holmes.parc.xerox.com>
Message-ID: <ErbsbdIB0KGWFZV3MM@holmes.parc.xerox.com>

The problem seems to have been that I was using the Python libraries
from 1.5, but the bin directory for Python 1.5.2 was on my PATH
environment variable.  Wild!  This seems to be an (unintended)
consequence of the code in Modules/getpath.c.  So the rule is, implement
Py_SetProgramName when embedding Python.

Bill



From gherron at aw.sgi.com  Thu Jul 22 12:06:38 1999
From: gherron at aw.sgi.com (Gary Herron)
Date: Thu, 22 Jul 1999 16:06:38 GMT
Subject: Is anybody aware of a tree GUI widget for tkinter or Pmw?
References: <3796E213.9F326412@eng.cam.ac.uk>
Message-ID: <3797418E.841FD680@aw.sgi.com>

"D. Lopez-De-Ipina" wrote:
> 
> Hi,
> 
> Could anybody please tell me if there exits any TREE control widget for
> tkinter or Pmw? I know that Pmw provides a rather rich set of
> 'megawidgets' but the tree control has not been implemented yet as far
> as I know.
> 
> Thanks a lot,
> 
> Diego

Tkinter does not have such a widget.  Pmw has a start on one in its Demo
directory (.../Pmw/Pmw_0_8/contrib/PmwTreeBrowse.py), but it fails even
to run, much less do something useful.

On the other hand, wxPython (=Python+wxWindows) has a nice tree widget,
and of course there is always Pythonwin which exposes tree MFC CTreeCtrl
widget.

-- 
Dr. Gary Herron <gherron at aw.sgi.com>
206-287-5616
Alias | Wavefront
1218 3rd Ave, Suite 800, Seattle WA 98101



From jrwren at ice.acs.oakland.edu  Mon Jul 12 10:53:50 1999
From: jrwren at ice.acs.oakland.edu (Jay R. Wren)
Date: Mon, 12 Jul 1999 10:53:50 -0400
Subject: SSL intro.
Message-ID: <7mce3s$4vf$1@news2.acs.oakland.edu>

I was wondering if anyone had some pointers on python and SSL.  I found a
paper about SSLeay with python and it has the object reference, but it
doesn't say how to actually install SSL for python.

Thanks,

Jay R. Wren





From fredrik at pythonware.com  Wed Jul 28 15:00:38 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Jul 1999 19:00:38 GMT
Subject: global and import
References: <XFMail.990728113603.nathan@islanddata.com>
Message-ID: <000901bed92b$7bdb3fe0$f29b12c2@secret.pythonware.com>

Nathan Clegg <nathan at islanddata.com> wrote:
> I'm having a little trouble understanding how global and import interact. 

quick answer: they don't.

"global" doesn't have anything to do with global
variables in the usual sense; it's just a way to
tell Python to look in the *module* namespace
instead of in the *local* namespace when you're
inside a function or a method.

> Currenly I have a script that includes a couple of global variables that
> allow for a small amount of communication between multiple threads and
> signal handlers.  I would like to move some of the functions to separate
> modules for several reasons, but need to maintain just the two global
> variables between them.

put these two variables in a common module,
and access them explicitly from the others. and
just forget about the global statement.

consider this:

common.py:

    mode = 0 # default

onemodule.py:

    import common

    def myfunc(self):
        print "look ma, no global"
        common.mode = 1
        mylib.dosomething()
        common.mode = 0

twomodule.py:

    import common

    def new_great_taste(argh):
         compute(common.mode, argh)

etc.

</F>




From thamza.nospam at geocities.com  Sat Jul 10 02:10:05 1999
From: thamza.nospam at geocities.com (Tim Hamza)
Date: Sat, 10 Jul 1999 06:10:05 GMT
Subject: idle
Message-ID: <1tBh3.875$uZ.2396@news.flash.net>

Have there been any updates to idle.  I thought I saw something about a 0.5
version, but I haven't been able to find anything at python.org.





From alex at somewhere.round.here  Fri Jul  2 22:38:38 1999
From: alex at somewhere.round.here (Alex)
Date: 02 Jul 1999 22:38:38 -0400
Subject: Problem embedding C++ code
References: <7ldqfs$nbj$1@nnrp1.deja.com>
Message-ID: <etdiu8231ap.fsf@m2-225-12.mit.edu>

I had this problem a while back.  It went away when we made everything
that gets included a shared library.  That included the C++ library, I
think...  it was on linux, and I think we had to include libg++.so
instead of libg++.a.  

HTH, and good luck.
Alex.

> CC -c -KPIC link-wrap.c -I/home/mdeshpan/SW/include/python1.5
> "link-wrap.c", line 13: Warning: String literal converted to char* in
> formal argument 1 in call to Py_BuildValue(char*, ...).
> 1 Warning(s) detected.
> CC -G link-wrap.o -lapic -lpthread -o link.so
> daffy:~/Swig$python
> Python 1.5.2 (#3, Jun 29 1999, 11:17:23) [C] on sunos5
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import link
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> ImportError: ld.so.1: python: fatal: relocation error: file
> ./libapic.so: symbol _ex_rethrow_q: referenced symbol not found

This is a gratuitous signature, included so that the news server won't
complain that I've got more included text than new text.  Snappy, huh?



From arcege at shore.net  Tue Jul 13 10:37:39 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Tue, 13 Jul 1999 14:37:39 GMT
Subject: ANN: Stackless Python 0.3
References: <378A40AD.B4FDB79F@appliedbiometrics.com> <m37lo5kpd0.fsf@atrus.jesus.cam.ac.uk> <378AFD45.8793D507@appliedbiometrics.com>
Message-ID: <TaIi3.1701$6M6.569174@news.shore.net>

Christian Tismer <tismer at appliedbiometrics.com> wrote:

: Here is the patch.
: It was one of those "=" instead of "==" bugs.

And newbies wonder why Python doesn't have assignment expressions. ;)

  -Arcege




From roffe at morgoth.uio.no  Wed Jul 21 14:44:44 1999
From: roffe at morgoth.uio.no (Rolf Marvin Bøe Lindgren)
Date: 21 Jul 1999 20:44:44 +0200
Subject: Python Mode for Alpha ?
References: <lawter-2107990058210001@client-151-197-119-26.bellatlantic.net> <14229.57251.996871.502133@anthem.cnri.reston.va.us> <14229.58766.234709.815774@weyr.cnri.reston.va.us> <o9vhbevwgh.fsf@panix3.panix.com> <14229.61952.83320.458319@weyr.cnri.reston.va.us>
Message-ID: <lbzr9m1etb7.fsf@morgoth.uio.no>

[ Fred L Drake

|   I've no idea; I've not used a Macintosh as a development environment 
| for at least four years!

well shame to you!


:)

-- 
Rolf Lindgren                                        http://www.uio.no/~roffe/
roffe at tag.uio.no



From jlj at cfdrc.com  Thu Jul  1 09:46:02 1999
From: jlj at cfdrc.com (Lyle Johnson)
Date: Thu, 01 Jul 1999 08:46:02 -0500
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net>
Message-ID: <377B711A.E8B65E0B@cfdrc.com>

I've already worked up the binary distribution of FXPy for Windows; I just
forgot to mention it in the previous post. The binary installer can be found
at:

    ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42-win32.exe

I haven't worked up a binary distribution of the (C++) FOX distribution
itself, primarily because it changes most every day :) But that is in the
plans as well.

Thanks,

Lyle

Mikael Lyngvig wrote:

> On Wed, 30 Jun 1999 09:07:51 -0500, Lyle Johnson <jlj at cfdrc.com>
> wrote:
>
> >Just to add fuel to the fire, we now have a Python interface to FOX, a
> >cross-platform GUI toolkit:
> >
> >    http://www.cfdrc.com/pub/FOX/fox.html
> >
> >The latest version of the Python wrappers can be picked up at:
> >
> >    ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42.tar.gz
> >
> >I don't know how the speed of FXPy would compare with some of the others
> >mentioned (Tk, wxWindows, Qt, GTK, ...) but it would be worthwhile for
> >some independent opinions.
>
> I'm downloading it now.  The http URL is wrong, it should be
> http://www.cfdrc.com/FOX/fox.html.
>
> The presentation seems promising, but I'd like to suggest a binary
> distribution for PC/Intel computers.  I know this is not common for
> Unix distributions, but it's the norm for PCs.
>
> -- Mikael




From salzr at certco.com  Tue Jul 27 16:37:36 1999
From: salzr at certco.com (Rich Salz)
Date: Tue, 27 Jul 1999 16:37:36 -0400
Subject: API question -- decref after adding to dictionary?
Message-ID: <rps64v$0$37nspbi$7gc@news.supernews.com>

I'm converting my own associative array into a python dictionary so I can
callout to some python code.  I think the snippet below should be clear
enough.

PyDictObject* newdict = PyDict_New();
for (CKeyValList::iterator i(kvl); i; ++i) {
        PyString* v = Pystring_FromString(i.value());
        PyDict__SetItemString(p, i.key(), v);
        Py_DECREF(v);
}

Is the DECREF necessary?  Reading the source of dictobject.c makes me think
it is, but the comments in object.h say otherwise.

Replies to me will be summarized to the list.  Thanks.
    /r$






From heiland at ncsa.uiuc.edu  Wed Jul  7 09:23:38 1999
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Wed, 07 Jul 1999 08:23:38 -0500
Subject: Python array --> C
References: <1280771409-69691649@hypernet.com>
Message-ID: <378354DA.D45DF723@ncsa.uiuc.edu>

Gordon McMillan wrote:

> Randy Heiland writes:
>
> > In my continuing quest to better understand embedding/extending, I
> > have the following question...
> >
> > in Demo/embed, I have modified demo.c so that the static module
> > xyzzy_foo(self, args) now does the following (instead of returning
> > 42L): {
> >         unsigned char retval[5] = {0,2,4,8,16};
> >         return Py_BuildValue("s#", retval,5);
> > }
> > i.e., it returns an array of unsigned chars.
> >
> > From main(), I would like to invoke xyzzy.foo() and get this array
> > of unsigned chars into a C array. How do I do this?  I assume I'll
> > need to use PyRun_String(...) somehow - true?
>
> Not at all. You'll want to use PyString methods to get its size and
> char *, etc. You can use python as a lib without ever involving the
> interpreter. See the C API reference manual (and be prepared to grep
> the objects and modules subdirectories).
>
> - Gordon

Can you please be more specific (sorry, I need a lot of hand-holding).
Exactly what PyString method
would I call in main() to (1) invoke the xyzzy.foo() method and (2) get
the resulting byte array into a C array?

--Randy




From larsga at ifi.uio.no  Thu Jul 22 02:48:30 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 22 Jul 1999 08:48:30 +0200
Subject: python vs. lisp (whereis quote?)
References: <x88r9m3cyje.fsf@rpppc2.hns.com> <lbzk8rva3to.fsf@morgoth.uio.no> <x88zp0rbd1x.fsf@rpppc2.hns.com> <mtlncbhv94.fsf@astron.berkeley.edu> <lbzogh635zl.fsf@morgoth.uio.no> <14230.2767.18575.943818@lrz.de> <lbzvhbdeu5f.fsf@morgoth.uio.no>
Message-ID: <wkvhbd6uyp.fsf@ifi.uio.no>

* Eugene Leitl
|
| Why? Because it's Scheme, or because it's Guile?

* Rolf Marvin B?e Lindgren
| 
| because too many of those involved in Guile don't understand the
| language. 

Another thing is that Scheme doesn't have a standardized package
system[1], is slower than Common Lisp and also lacks a lot of things
that are built in to Common Lisp.

And, besides, you can get Schemes written in Common Lisp.

--Lars M.

[1] This is really important in Emacs where people will be mixing
    modules from authors all over the world in the same environment.
    If two of them use the same name for different variables chaos
    follows.



From wolford at enews.nrl.navy.mil  Fri Jul  9 10:40:57 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Fri, 09 Jul 1999 10:40:57 -0400
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com> <37850F5B.F80CB0A7@enews.nrl.navy.mil> <7m3ajc$s3q$1@m2.c2.telstra-mm.net.au>
Message-ID: <378609F9.99338FD@enews.nrl.navy.mil>

Mark Hammond wrote:

> Scott Wolford wrote in message <37850F5B.F80CB0A7 at enews.nrl.navy.mil>...
>
> >msvcrt.lib and the application can't then bring in msvcrtd.lib. I do think
> >this is a problem because python15_d.lib is provided but no python15_d.dll
> >is. I would think you'd provide (at the very least) python.lib linked with
> >msvcrt.lib, and python_d.lib linked with msvcrtd.lib (and the dlls), or
> >(to be really nice :) all four combinations.
>
> The main reason we dont is size - we would need to ship the _d, and for
> people to really use it, also the .pdb.  These add too many megs.
>
> Further, almost _everyone_ who needs to use this _d version already has MSVC
> and already has the Python sources.  So building there own is trivial.

True. That's how I finally figured out my problem.

> Further, "all 4 combinations" doesnt work.  As you found out painfully, is
> it _critical_ that all Python systems share the same C runtime library.
> This is the whole point of the _d fiasco - to prevent the mixing and
> matching.

How about:
python.lib/dll and msvcrt.lib/dll
python.lib/dll and msvcrtd.lib/dll
python_d.lib/dll and msvcrt.lib/dll
python_d.lib/dll and msvcrtd.lib/dll

separated into two directories: msvcrt and msvcrtd

> We obviously need to document this process better - Im not sure where tho -
> what documentation did you use to get this far?

Not much, just figured it out myself mostly. I was only off by one character
after all: msvcrtd.lib instead of msvcrt.lib.

> It is also worth noting that Gordon pointed this out to you before this
> particular thread, and you still had the problem.  Im not sure that
> documenting it would be effective in all cases.

He said to use the Multithreaded DLL. Unfortunately, msvcrt.dll and msvcrtd.dll
are *both* the Multithreaded DLL; one is debug the other is not. It never
occured to me (well, it finally did) that it mattered whether msvcrt.dll or
msvcrtd.dll were used. I assumed that the semantics of the two DLLs were
essentially the same, just the implementation differed slightly. I'm pretty sure
(though I haven't tested it, maybe you'll know) that you can mix and match
optimized and debug code pretty easily under Unix.

> > Don't get me wrong; I think
> >it's a great service to provide the pre-compiled package. I just want to
>
> Well, FWIW, I will be making these _d versions (and all the win32 extensions
> _d versions) available in a seperate .zip file as a Registered User benefit
> (http://starship.python.net/crew/mhammond/registration).  I havent started
> this particular benefit yet tho.
>
> Mark.

I think I'll just build it myself from now on, though thanks for the service. I
usually have trouble compiling windows software but Python was smooth as smooth
can be.





From tim_one at email.msn.com  Wed Jul 21 00:57:30 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 21 Jul 1999 00:57:30 -0400
Subject: Python and Boehm-Demers GC, I have code.
In-Reply-To: <p5k8rvmymj.fsf@bidra241.bbn.hp.com>
Message-ID: <000601bed335$893916e0$a72d2399@tim>

[Tim]
>     for i in xrange(10000000):
>         pass
> [and RC vs GC implications]

[Markus Kohler]
> True.
> IMHO this is one of Pythons major design flaws.

What, specifically?

> There is no need to create a new object for each step of the loop.
> Smalltalk has loops at least as flexible as Python and it doesn't
> have this problem.

Sorry, I've missed what "the problem" is thought to be here.  Loop overhead
is typically lareger in Python programs than in other languages, but is also
typically trivial all the same.

> Loops in squeak (www.squeak.org) are already 2 to 3 times faster
> than in Python.

All possible loops?  All loops you've personally timed?  One loop you timed
in a moment of boredom <wink>?  An empty loop that counts to 10000000?  I
don't doubt that Squeak is faster in many areas and for many reasons, but
I've been running Python for most of the decade and reducing loop overhead
is about the last thing I'd bother to aim at.

> ...
> Is there anything that would prevent someone to reimplement loops
> such that only one variable is allocated ?

In the loop above, there are actually two int objects created per iteration:
the one explicitly created by xrange and bound to "i", and a hidden "loop
counter" int object used to drive the for/__getitem__ protocol.  The latter
can be (& has been, experimentally) reimplemented via an internal "mutable
int" type created for that purpose, but it didn't buy enough to be worth
adopting.  Most loop bodies contain more than "pass" <wink>, and the more
work they do the less the fixed overhead matters.

unboxing-*all*-ints-now-that-would-buy-something-ly y'rs  - tim





From mfletch at vrtelecom.com  Fri Jul  2 13:29:17 1999
From: mfletch at vrtelecom.com (Mike C. Fletcher)
Date: Fri, 02 Jul 1999 17:29:17 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net><3773c3a0.102825445@news.oh.verio.com><xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net><xJRc3.292$Of3.67094@news.shore.net><7laovo$c4s$1@Starbase.NeoSoft.COM><377A24B7.AD6FAA1A@cfdrc.com><3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net><377B711A.E8B65E0B@cfdrc.com><377BC620.F5553FF5@easystreet.com> <377BF583.6846FBF9@quasar.ipa.nw.ru>
Message-ID: <NF6f3.6498$_A4.93414@news2.rdc1.on.home.com>

None of the demonstrations I've seen are using a "grid" method.  There is
one which looks similar, the "header" demo, but that is a set of "list"
widgets standing next to each other and a sitting below a header widget.  To
create a real grid, you would need to do some (likely C++) programming.  You
would probably also want to make such thing allow for editing in-field, and
have the capability of updating directly from an application table to the
on-screen representation.  It should be possible to create such a thing in
Python by dropping data into a matrix of entry fields in response to
scrolling/moving etc., but a grid object generally wants a high-performance
solution with all sorts of funky options.

A "rich" text editor is evident in the Fox IDE, FXEditor, a separate project
from the Core Fox libraries, though I have not yet downloaded that project
to test it.  Something similar should likely become part of a base Fox
distribution if it is to become a widely deployed Python GUI.

Fox has a basic GUI editor under development (it is a separate project
again), though my tests with it so far have been unsuccessful, as the
application exits silently about every-other time I click on a widget.

Neither tkinter nor Fox has any perceptible "slowness" on my system, so I
cannot comment in that direction.

I'm continuing my experiments, will continue to report as time permits.

Enjoy,
Mike


Alexander V. Voinov <avv at quasar.ipa.nw.ru> wrote in message
news:377BF583.6846FBF9 at quasar.ipa.nw.ru...
...
> Did you look at the demos for the underlying toolkit? Some of them
presents a
> 'table' which is probably less rich than that of wxWindows, but it does
exist.
>
> What it really lacks is a 'richtext'/html widget, sort of which is present
and
> actively developed in wxWindows. As I understand it's just about to become
> available in wxPython.
...





From stidolph at origin.ea.com  Wed Jul 28 17:25:57 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Wed, 28 Jul 1999 21:25:57 GMT
Subject: 1.5.2 for: else:
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC54@molach.origin.ea.com>

As I am understanding for .. else, the following

for window in self.windows:
  self.ProcessWindow(window)
else:
  print 'windows processed'

is the same as

for window is self.windows:
  self.ProcessWindow(window)
print 'windows processed'

and the only case where the else statement has any value over simply adding
code after the for loop would be if a break statement were executed - in
which case the else would NOT be called.  Personally, I would rather have
the else execute if none of the code within the for loop were executed.  It
would save a variable assignment/test as in:

executed = 0
for window in self.windows:
  executed = 1
  self.ProcessWindow(window)
if not executed:
  self.HandleProcessWindowNotCalled()

This forces the variable to be set numerous times (needless).

I'm with those that see else being an alternate execution (like the if)
rather than some sort of continuation.  I'm also with those that favor "if
it isn't broke, don't fix it".

-----Original Message-----
From: Gordon McMillan [mailto:gmcm at hypernet.com]
Sent: Wednesday, July 28, 1999 4:38 PM
To: python-list at cwi.nl
Subject: Re: 1.5.2 for: else:


William Tanksley wrote:
> On Tue, 27 Jul 1999 21:43:00 -0500, Gordon McMillan wrote:

[snip Billy's silly error and his entirely insufficient embarassment 
<wink>]

> However, you're wrong about the best way to detect a 2 in the list
> using this style.  To steal the same example:
> 
> try:
>  for x in [1,2,3]:
>   print x
>   if x == 2: raise neverMind
>  print "modern else clause here"
> except neverMind: print "found two"
> else:
>  print "another place to put the modern else clause"
> 
> Wow, TMTOWTDI.  Oops, did I say the wrong thing?
> 
> Of course, by demonstrating both ways to do it, I'm making the code
> more complex; but you can see that this solution solves the problem
> quite elegantly.

WIth an extra class, and 3 or 4 more lines of code, not to mention 
treading on the toes of all who think exceptions should only be used 
for exceptional conditions (or maybe it _is_ exceptional when 
something you wrote works <1e3 wink>).

> The place to use MY definition of 'else', OTOH, is when you're
> building a structure during the iteration, but the result of an
> empty loop is different than the initial value the variable has to
> have when going through the loop.
> 
> Here you go.  This example might convert a Python list into a Lisp
> tree-list -- but to imitate Lisp (okay, it's a toy problem, I'm
> cheating) an empty list produces NIL ~= None.
> 
> var = Node(None,None)
> for x in pyList:
>   var.car = x
>   var.cdr = Node()
> else:
>   var = None

Not a Lisper, but isn't what you're doing better done thusly:

def test1(lst):
 var = None
 for x in lst:
  var = (x, var)
 return var

print `test1([1,2,3])`

print `test1([])`
 
Produces:
(3, (2, (1, None)))
None
 
> >insufferably-sanctimoniously-y'rs
> 
> >- Gordon
> 
> unstoppably-confidantly y'rs
> 
> -- 
> -William "Billy" Tanksley

so-self-righteous-without-the-winks-I-might-almost-be-mistaken
 -for-that-other-GM-ly y'rs

- Gordon





From holger at phoenix-edv.netzservice.de  Tue Jul  6 05:14:20 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Tue, 6 Jul 1999 09:14:20 GMT
Subject: Name of function within the function?
References: <377B8C2F.80F44F9@americasm01.nt.com> <377BB3FF.53A1CB62@starmedia.net>
Message-ID: <3781C8EC.144B8C21@phoenix-edv.netzservice.de>

BTW: How do I get 'the instruction string of the code object' 
(reference manual s.17)?
Or: With sys.exc_info()[2].tb_lasti I get the index to some list, but 
where could I get the list?
 
Why is  sys.exc_info()[2].tb_frame.f_lasti much higher that ..tb_lasti?
Which stack/list is meant here?

I'd like to get mostly 7 informations about an exception:
1)what kind of error (sys.exc_type?)
2)which script (sys.exc_info()[2].tb_frame.f_code?)
3)which line in script causes error? (sys.exc_info()[2].tb_lasti?)
4)which instruction causes error (???)
5)variables (ok, easy to get, I think...)
6)name of function (sys.exc_info()[2].tb_frame.f_code.co_name)
7)name of calling function  (sys.exc_info()[2].tb_frame.f_back.f_code.co_name)

Why are sys.exc_type and sys.exc_value are so different to sys.exc_info[:2]???

many questions, I know...;-)
But YOU'll solve that! You're kind, aren't you???;-)

Ciao,
Holger



From dan at control.com  Thu Jul 22 11:29:53 1999
From: dan at control.com (Dan L. Pierson)
Date: Thu, 22 Jul 1999 15:29:53 GMT
Subject: Embedding question
References: <816010E2456BD111A48700805FBBE2EEFDF4EF@ex-quebec-u1.baan.com>
Message-ID: <379729cf.151903937@news.randori.com>

Gaetan Corneau <Gaetan_Corneau at baan.com> wrote:

> can Python be put in a Windows NT program's DLL? The "main" program should
> know nothing about the Python interpreter. it is a program that is almost
> entirely made of "plug ins".

Yes.  The plug-in's initialization function would have to initialize
Python and the plug-in's termination function would have to terminate
it (somewhat less important).

Dan Pierson, Control Technology Corporation
dan at control.com



From moshez at math.huji.ac.il  Thu Jul  8 06:12:24 1999
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Thu, 8 Jul 1999 13:12:24 +0300
Subject: Xlib wrapper for Python
Message-ID: <Pine.SUN.3.95-heb-2.07.990708130942.17745A-100000@sunset.ma.huji.ac.il>

Hi. 

I saw
http://www.uni-klu.ac.at/unihome/header/tutorial/python--1.4/
X-extension/Xlib-functions-and-variables.html

And started wondering where do I get this Xlib wrapper?

Thanks in advance.
--
Moshe Zadka <mzadka at geocities.com>. 
#!/usr/bin/tail -1
Just another tail hacker.




From catlee at my-deja.com  Mon Jul 19 17:27:07 1999
From: catlee at my-deja.com (catlee at my-deja.com)
Date: Mon, 19 Jul 1999 21:27:07 GMT
Subject: Killing a thread
References: <7mu6a5$h61$1@nnrp1.deja.com> <7mvml3$u5k$1@nntp6.u.washington.edu> <ePJk3.4$Z11.6082@news-dal.corridex.com> <7n04dm$8ca$1@nnrp1.deja.com>
Message-ID: <7n056p$8nj$1@nnrp1.deja.com>

Whoops!  Never mind.  That previous program didn't work at all...


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From cgw at fnal.gov  Tue Jul  6 17:10:17 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Tue, 6 Jul 1999 21:10:17 GMT
Subject: Inserting empty space into an existing file/pointer
In-Reply-To: <7ltn1m$nen$1@snatch.hq.eso.org>
References: <7ltn1m$nen$1@snatch.hq.eso.org>
Message-ID: <14210.28857.177175.975158@buffalo.fnal.gov>

It's not really a Python question, it's a matter of what operations
are supported by the underlying filesystem. And there's no filesystem
that I know of that will let you "insert" stuff into a file.  (I'm
interested to see if any of the gurus here tell me I'm wrong).


You could however do something like this:

import os
filename = "whatever"
infile = open(filename,'r')
outfile = open(filename+".tmp",'w')

outfile.write(infile.read(initial_count))
outfile.write('\000' * n_zeros)
outfile.write(infile.read(final_count))
outfile.close()

os.rename(filename, filename+".old")
os.rename(filename+".tmp", filename)

Note that this might be somewhat memory-intensive if "initial_count"
or "final_count" are large.  Breaking the I/O into blocks is left as
an exercize.




From jbauer at rubic.com  Thu Jul  8 21:32:52 1999
From: jbauer at rubic.com (Jeff Bauer)
Date: Fri, 9 Jul 1999 01:32:52 GMT
Subject: Problem with ZOPE under WinNT
References: <3780DCEA.B43FE208@lab136.cyber.kiev.ua>
Message-ID: <37855144.39146878@rubic.com>

> When I test pcgi info file "pcgitest" with pcgifile.py
> it generate next error:
> 
> error attempting: 'from cgi_module_publisher import publish_module'

Vitaliy,

I haven't heard of many people using the pcgi module with NT.
Nice to know someone's using it.  Just comment out the lines
in pcgifile.py that are attempting to import from
cgi_module_publisher -- it's been removed from the
latest version:

  http://starship.python.net/crew/jbauer/persistcgi/pcgifile/pcgifile.py

Sorry for the confusion.

-Jeff Bauer

P.S.  You'll get quicker and better responses by contacting
the Zope list directly.



From neelk at brick.cswv.com  Thu Jul 15 20:50:06 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 15 Jul 1999 19:50:06 -0500
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu>
Message-ID: <7mlvju$6v3$1@brick.cswv.com>

In article <7mkrrt$a81$1 at cronkite.cc.uga.edu>,
Graham Matthews <graham at sloth.math.uga.edu> wrote:
>
>[Parser Combinators] 
>
>They are something like what Neel is talking about (as the name
>"parser combinators" suggests you write small "string parsers" as
>functions, and then have "parser combinators" that stitch these
>together). Note that in this parser combinator approach there is no
>distinction between the code for "the pattern" and the code for "the
>processing" -- they are all written in the same language
>(Clean/Haskell in my case). 

Thanks for the tip -- it led me to do some digging on the web and I
found a nice description of this in a paper called "Monadic Parser
Combinators" by Graham Hutton and Erik Meijer. Its URL is:

  http://www.cs.nott.ac.uk/Department/Staff/gmh/monparsing.ps

>At first I found the idea of having the two types of code (the
>pattern and the processing) in the same language (rather than in 2
>distinct languages as in the traditional way) a bit difficult. But I
>perservered and I am glad I did. I found that once you reach a
>certain level of familiarity with combinators you can write new
>pattern and processing code very quickly and very reliably. The
>trick is to get the combinators right.

I think that this weekend I'm going to try and put together a basic
version of this parsing system in Python based on the ideas in this
paper -- basically just write a simple AST class and the combinators
they describe. Then I can add proper support for backtracking and lazy
evaluation and all the stuff with serious hack value when Christian
Tismer makes continuations available to us. :)

This really seems like a clean and straightforward way of doing
things.  My experience with recursive descent parsers has been pretty
bad, but now I suspect that this is because I wasn't organizing my
work in a systematic enough fashion. 


Neel



From fredrik at pythonware.com  Fri Jul 30 13:54:28 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 30 Jul 1999 17:54:28 GMT
Subject: Recursive constructors?
References: <roy-3007991318280001@qwerky.med.nyu.edu>
Message-ID: <002901bedab5$09e4d500$f29b12c2@secret.pythonware.com>

Roy Smith <roy at popmail.med.nyu.edu> wrote:
> Is it legal for a constructor function to call itself recursively?  I.e.:
> 
> class foo:
>    def __init__(self, x):
>       if (whatever):
>          self.bar = foo(y)

why not just try it? ;-)

(hint: the class statement will finish before
you get around to actually create an in-
stance of the class, and at that time, 'foo'
is already in the GLOBAL namespace).

</F>




From 71351.356 at compuserve.com  Tue Jul 27 20:32:20 1999
From: 71351.356 at compuserve.com (T. C. Mits)
Date: Tue, 27 Jul 1999 20:32:20 -0400
Subject: make or similar in Python
References: <378390DD.58CA6BD5@nevex.com>
Message-ID: <7nlj9o$mj7$1@ssauraac-i-1.production.compuserve.com>

Howdy:
    Funny, I'm searching for the exact same thing.

    My idea is to use XML as a repository for the dependencies and other
required configuration properties.  A Python script will query the db and
generate dynamic makefiles.   A pain is that this has to work on NT and Unix
and perhaps even by web based via Servelets.


Greg Wilson <gvwilson at nevex.com> wrote in message
news:378390DD.58CA6BD5 at nevex.com...
> Hi, folks.  Has anyone ever written a dependency management/update tool
> (like 'make') in Python?
> I'm not looking for something that reads makefiles, and happens to be
> written in Python; I'm looking
> for a Python module that users import, then provide data to, in order to
> manage dependencies.  (The
> answer I got from the Perl crowd was 'cool idea, but no'; I'm hoping
> I'll have more luck here.)
> Thanks,
> Greg (who'd really, really like to override some of the things that
> GNU make does)
>
>





From news at dorb.com  Tue Jul 20 12:59:46 1999
From: news at dorb.com (Darrell)
Date: Tue, 20 Jul 1999 12:59:46 -0400
Subject: Python suitability questions
References: <0CCE781E44F2C872.F528377A62B41DCA.73CAA63AD6E79314@lp.airnews.net>
Message-ID: <NW1l3.1051$u74.20294@newsr1.twcny.rr.com>

Jason:

> And of course I need some sort of internal messaging system
What do you mean by this ?

>The scripting language has to be comfortable being completely
> embedded in my C++ framework, and having really no system access at all.
You can embed python but it can still access your system.
There is a "Restricted execution" mode

>How flexible can I lay out my source, is it as stiff as say, C?
Lay things out almost anyway you can imagine.
My experience was that years of C++ had limited my imagination.
It took me quite a while to stop writing C++ in Python.

--Darrell








From tim_one at email.msn.com  Sat Jul  3 02:03:15 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 3 Jul 1999 02:03:15 -0400
Subject: How do you make clickable Python/Win9x scripts?
In-Reply-To: <377D8ED0.49BEF319@postoffice.pacbell.net>
Message-ID: <000301bec519$bdd41640$d69e2299@tim>

[Rob]
> Is there a way to make a Python script file a clickable .BAT file in
> Win9x?

This kind of thing is the best I've ever come across:

rem=''' Here's a cheery greeting you can't get rid of, so use it
@echo off
:any other bat cmds you'd like to do first
:and then execute python on this file:
%PATH_TO_YOUR_PYTHON_DIR%python %0
goto endofpython
'''

name = raw_input("What's your name? ")
print "Hi,", name

rem = '''
:endofpython
:any other bat cmds you'd like to do afterwards
:'''

The batcrap is hidden from Python in multiline triple-quoted strings
assigned to "rem", the initial lines of which W9X treat as comments.  Note
the use of label notation (:''') to hide the final string closer from the
bat processor.  BTW, you're usually better off starting bat comments via ":"
than "rem"!  To understand why, try sticking

: <hi>

in one bat file and

rem <hi>

in another <wink>.

command.com-leaves-the-teensiest-bit-to-be-desired-ly y'rs  - tim





From amaranda at nospam.com  Fri Jul  2 06:48:36 1999
From: amaranda at nospam.com (Alex Maranda)
Date: Fri, 02 Jul 1999 11:48:36 +0100
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain> <377B7E17.5BD7@nospam.com> <slrn7noa2j.hg.daniels@localhost.localdomain>
Message-ID: <377C9904.41AB@nospam.com>

Alan Daniels wrote:
[snip]
> will be involved. The only justification I can see is better CPU
> utilization on an SMP machine, and I don't happen to have one of those
> lying around. Hence the engine will run in a single thread.
This justification doesn't fly on SMP machines precisely because of the
global interpreter lock. This was debated to death in previous threads.
All we can do is wait till the interpreter is fully re-entrant and the
need for the global lock goes away...
BTW (completely off topic), I wonder what's the current thinking on this
- the Linux kernel went for finer-grained locks.

> I have the (VERY!) preliminary code at:
>     http://www.mindspring.com/~daniels/engine/index.html
> if you're interested, although there isn't much to look at.
I'm downloading it now. You can use negative version numbers;
SmallEiffel is now at version -0.78. They'll stop when they'll reach 0.0
:-)

Cheers,
-- 
Alex Maranda         mailto: amaranda at spider dot com
Spider Software Ltd. Tel: +44 (0)131 4757036
Edinburgh, UK        http://members.xoom.com/Alex_Maranda
STREAMS based communications protocols for embedded systems



From MHammond at skippinet.com.au  Fri Jul  9 10:04:57 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Sat, 10 Jul 1999 00:04:57 +1000
Subject: COM with Python?
In-Reply-To: <1280594885-80315141@hypernet.com>
Message-ID: <002c01beca14$07f4c130$0801a8c0@bobcat>

Gordon McMillan wrote:
> Well, having just done some testing, let me put a resolution on this:
> the global Python object approach works (which surprised me!), and
> the Running Object Table approach, while it works, seems to have
> stability problems.

Excellent!
>
> To explain: I hadn't realized that the COM plumbing would, when using

he he - I did, so I knew it would :-)

> CLSCTX_LOCAL_SERVER, reconnect to a running instance, (so I doubted
> that the global Python object approach would work). It does, and
> while I haven't left one running for 24 hrs yet, it appears to work
> flawlessly - the COM invoked pythonw process goes away as soon as all
> refs go to zero.

With build 125, here is an even better trick to impress your friends.
Install Python on 2 NT or DCOM capable 95/98 machines.  One one machine,
say:

x = win32com.client.DispatchEx("Python.Interpreter", "other_machine_name")

And bang away.  Looking at the remote machine show pythonw.exe living
happily, and correctly going away when the remote client deletes x.  [Even
better, we could substitute a proper NT service in place of pythonw.exe,
but that is a different thread, hopefully not for some time :-]

> The ROT approach (using pythoncom.RegisterActiveObject()) is flaky.
> Sometimes the COM invoked pythonw goes away when it should, sometimes

My guess is that this is because we dont handle the concept of "weak
references" that (IIRC) the ROT has.  Ive never really understood it, but I
wouldnt be surprised to find it tied up with IExternalConnection that
AXControls (and Active Scripting) attempt to use to implement "weak
references".  Python can handle this interface, but the framework doesnt -
ie, you must explicitely declare you support this interface and implement
it.  Ive seen my objects be QI'd for this interface, but never seen any
difference in functionality regardless of what I do (ie, implement it,
ignore it, implement it strangely, etc!)

> it hangs around. Worse, if you leave it running for long periods, it
> seems capable of freezing NT's UI (the machine appears to still be

Eeek!

> alive, but you can't _do_ anything except reboot). I'm more than
> willing to believe that this is NOT Mark's fault.

Whew :-)  Although it would seem clear we are screwing _something_ up -
many other apps use the ROT without this problem.  But Ive never seen a
good example of using the ROT that does anything we dont...

Thanks...

Mark.




From ze5yr at my-deja.com  Fri Jul 30 19:00:22 1999
From: ze5yr at my-deja.com (ze5yr)
Date: Fri, 30 Jul 1999 23:00:22 GMT
Subject: 1.5.2 for: else:
References: <1278996512-12636218@hypernet.com> <slrn7pun42.2ro.wtanksle@dolphin.openprojects.net> <379FC039.73C436FC@mincom.com>
Message-ID: <7ntaq2$6mh$1@nnrp1.deja.com>

In article <379FC039.73C436FC at mincom.com>,
  Martin Pool <martinp at mincom.com> wrote:
> I doubt if I'll often use for: else or try: else, and perhaps
> they could have clearer names,

How about for: then?

but-what-about-all-my-variables-named-'then'-ly yrs, Cliff


--
Cliff Crawford                           -><-
http://www.people.cornell.edu/pages/cjc26
                            empty-handed I go,
)O(       and behold the spade is in my hands


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From kauer at pheno.physics.wisc.edu  Wed Jul 21 02:17:26 1999
From: kauer at pheno.physics.wisc.edu (Nikolas Kauer)
Date: Wed, 21 Jul 1999 01:17:26 -0500 (CDT)
Subject: ImportError: Unresolved symbol
Message-ID: <199907210617.BAA26862@pheno.physics.wisc.edu>

When I run a Python CGI script (first line #!/patjh/bin/python), I get the 
following error in the Apache error log:

[Tue Jul 20 16:58:55 1999] [error] [client ...] Premature end of script headers: /usr/share/test/apache_1.3.6/cgi-bin/db_
driver.cgi
Traceback (innermost last):
  File "/usr/share/test/apache_1.3.6/cgi-bin/db_driver.cgi", line 24, in ?
    from cgi_namespace import *
  File "/usr/share/test/python/lib/python1.5/lib-dynload/cgi_namespace.py", line 8, in ?
    from cgi_state import State
  File "/usr/share/test/python/lib/python1.5/lib-dynload/cgi_state.py", line 1, in ?
    import types, DateTime, string
  File "/usr/share/test/python/lib/python1.5/lib-dynload/DateTime.py", line 8, in ?
    from mxDateTime import *
ImportError: Unresolved symbol in /usr/share/test/python/lib/python1.5/lib-dynload/mxDateTime.so: timegm

When I start an interactive Python session, and do "import DateTime", or execute the Python script on 
the command line (like tcsh> ./db_driver.cgi) no error messages appear, everything seems to work fine.  
The Web server runs on Digital Unix 4.0E.

The only solution I can currently think of would be to try to statically link mxDateTime.so ...

Any hints, anybody?

Thanks!



From phd at emerald.netskate.ru  Wed Jul 14 04:58:36 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Wed, 14 Jul 1999 12:58:36 +0400 (MSD)
Subject: Embedded Python can't "import socket" -- "undefined symbol"
In-Reply-To: <87zp104a8v.fsf@solano.in.renlabs.com>
Message-ID: <Pine.LNX.4.04.9907141255020.8520-100000@emerald.netskate.ru>

On 13 Jul 1999, Steven Work wrote:
> I am embedding Python in a C program (on Linux 2.2.x, Debian 2.1), and
> trying to run a script via PyRun_SimpleFile().  "print 'hello, world'"
> works, but "import socket" dies.  Can anyone help me find out why?
> 
> $ ./testfile
> Traceback (innermost last):
>   File "jukebox_xml.py", line 1, in ?
>     import socket
> ImportError: /usr/lib/python1.5/lib-dynload/socketmodule.so: 
> 	undefined symbol: _Py_NoneStruct
> $ nm testfile | grep NoneStruct
> 0808ecfc D _Py_NoneStruct
> 
> And of course "import socket" works fine in /usr/bin/python.  

   _Py_None_Struct (and other symbols) should be exported from your main
program to link successfully. You haven't mention your OS and compiler, but
I think it is ELF-based UNIX, and probably gcc, right?
   Add -Wl,-E to compiler flags. This means "pass -E to ld linker", which
in turn means "export all symbols from main linked module".
   "man ld" for detalils on -E flag.

> -- 
> Steven Work
> Renaissance Labs
> steve at renlabs.com
> 360 647-1833

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From tim_one at email.msn.com  Thu Jul 22 03:07:43 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 22 Jul 1999 03:07:43 -0400
Subject: Cyclops 0.9.4
In-Reply-To: <37967D4C.9D9434ED@compaq.com>
Message-ID: <001301bed410$e59eafe0$642d2399@tim>

Starting at the end:

[Greg Ewing]
> So, again, cycles aren't really what you want to know about!

You want to know about objects still living that you don't expect to be
living.  No system can answer that for you!  They can't know what you expect
<wink>.

Cycles have a special status in Python, so they're often a *useful* thing to
look for.  For different apps, and for different times within a single app,
you may want to look for other things.

If you have specific objects in mind that you expect will be dead at time T,
you can register them with Cyclops before time T, and at time T Cyclops will
tell you whether they probably would have been dead had you not registered
them <wink>.

If you have no idea which objects to look for, you're not going to get help
short of divine revelation.  An object living after you expect it to die is
an error in program logic, and as hard to track down as any other kind of
programmer flub; heck, harder than most, because usually an error of
omission (you neglected to lose all the references) -- there's no line of
code "to blame".

> ...
> I think the Python core needs an enhancement to allow
> walking a list of all allocated objects. Cyclops in
> conjuction with that would be very useful. Maybe a compile
> time or run time option, since it would slow down
> allocation/deallocation a little bit.

If what you want is a list of all reachable objects, Python needs three
different things:

1) A little internal cleanup to make it easy to get at "the root set" (the
minimal set of objects from which it's guaranteed that all live objects are
reachable).  The lack of this in the core is why Cyclops (& Plumbo) makes
you "register" "interesting" objects.

2) A protocol for pointer-chasing, and changing extension modules to play
along with it.  The lack of this in the core is why Cyclops has a funky set
of methods for installing type-specific pointer-chasing functions.

3) Likely a notion of "mark bits", to identify objects already seen.
Cyclops fakes these via internal dicts keying off object id.

All of those are also needed if Python is ever to move toward builtin
portable mark-&-sweep (optional or not).  The advantage over "walking an
(explicit) list" is no overhead (time or space) unless & until it's used.
I'd like to see Python move this direction for several reasons (one of which
is indeed that then a Cyclops-like thingy could tell you not only that
something unexpected is still alive, but also from where it can be
reached -- chasing that chain back would be a real help).

in-the-meantime-it's-a-gloriously-fun-game<wink>-ly y'rs  - tim





From e_l_green at hotmail.com  Tue Jul 27 18:38:29 1999
From: e_l_green at hotmail.com (Eric Lee Green)
Date: Tue, 27 Jul 1999 15:38:29 -0700
Subject: Problem with signals & exceptions...
Message-ID: <2BB05D8530326D0C.420888E8906528AF.C6BF325D12E9ED90@lp.airnews.net>

I'm having trouble catching my thrown exceptions under FreeBSD 3.2. I am
using the Python 1.5.2 that came with FreeBSD 3.2 (well, on disk 1 of
the <n>-disk set). The following program works properly under Red Hat
Linux version 6.0 running a self-compiled Python 1.5.2 (in place of the
1.5.1 which comes with Red Hat 6.0, which does not properly handle
signals during socket i/o calls). How it works: it listens to port 6664.
When it gets a connection, it then sets an alarm handler and alarm and
then tries to read from the socket. If it does not get data within 10
seconds, it then times out, at which point the alarm handler throws an
exception and aborts the socket read.  

The test:
  In one shell window, type:
     python testalrm.py
  in the other shell window type:
     telnet localhost 6664
type NOTHING just let it time out...

The output in the first window:

bash-2.03$ python testalarm.py
Got alarm!
READ_TIMEOUT
Traceback (innermost last):
  File "testalarm.py", line 30, in ?
    raise IOError, why
IOError: READ_TIMEOUT
bash-2.03$ 

The program:
----------------------------testalarm.py------------------

# A routine to test the 'alarm' signal handling.

import socket
import signal

READ_TIMEOUT="READ_TIMEOUT"

def alarmhandler(signum,stackframe):
    print "Got alarm!"
    raise IOError,READ_TIMEOUT


s=socket.socket(socket.AF_INET,socket.SOCK_STREAM)
addr=("",6664)
s.bind(addr)
s.listen(5)

connection,address=s.accept()

oldhandler=signal.signal(signal.SIGALRM,alarmhandler)

signal.alarm(10)  # wait 10 seconds for an alarm!
try:
        str=connection.recv(4096) # get a string
except IOError, why:
        print why
        if why == READ_TIMEOUT :
                print "why = read timeout"
        else:
                raise IOError, why
signal.alarm(0)
print str

----------------end program-------------------------------------



From Neel.Krishnaswami at p98.f112.n480.z2.fidonet.org  Thu Jul  1 13:06:15 1999
From: Neel.Krishnaswami at p98.f112.n480.z2.fidonet.org (Neel Krishnaswami)
Date: Thu, 01 Jul 1999 18:06:15 +0100
Subject: total idiot question: +=, .=, etc...
Message-ID: <0000164e@bossar.com.pl>

From: neelk at brick.cswv.com (Neel Krishnaswami)

In article <000d01bec11f$8f966e20$e19e2299 at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>[Neel Krishnaswami]
>
>> With the clear way of writing it, the NewArticle class would
>> mysteriously fail to do the right thing, because the __cmp__ method
>> would look in the Article class dictionary without bothering to check
>> NewArticle's.
>
>The trap is hard to avoid, though:  if a direct instance of Article is
>"self", and a direct instance of NewArticle is "other", even keying off
>self.__class__.etc is a surprise (well, at least to "other"!).  In part
>you're taking a single-dispatch approach to a problem with multiple-dispatch
>headaches.  

<light dawns> 

Of course! I had noticed I wasn't having this problem in Dylan, and 
it somehow failed to occur to me that the only significant difference
between Dylan and Python (Dylan's multiple-dispatch object model)
might have something to do with it. :)


Neel



From ntaylor at newbridge.com  Fri Jul 23 09:29:40 1999
From: ntaylor at newbridge.com (Nick Taylor)
Date: Fri, 23 Jul 1999 09:29:40 -0400
Subject: Printable Python reference
References: <3797782A.FD39EF04@newbridge.com> <m31ze0kwmc.fsf@havenrock.com> <slrn7pfqsb.f9h.bwinton@tor.dhs.org>
Message-ID: <37986E44.955390FB@newbridge.com>

Blake Winton wrote:

> .tar.gz files are no problem with Windows.  WinZip handles them better
> (IMHO) than "tar -xvzf".  I'm forced to use NT during the day, and it's
> not my favourite, but there are a few things it does really well, and
> uncompressing stuff is one of them.

Yeah, although WinZip can load .tars but seems to have trouble uncompressing
them.  Good thing I've got the highly useful MKS Toolkit on my NT box.

Thanks for the info.  I look forward to working with Python.




From ta16 at cornell.edu  Mon Jul 19 11:08:04 1999
From: ta16 at cornell.edu (Tom Adelman)
Date: Mon, 19 Jul 1999 11:08:04 -0400
Subject: debugging MSVC DLLs
Message-ID: <37933F54.4AE54D39@cornell.edu>

I'm wondering if someone can help me with this (or should I direct this
question to a Microsoft group?):

What I'm doing now:
    I write classes in MSVC++, compile them as a DLL, and SWIG them.  I
then run python.exe from within the Microsoft Developer Studio, and then
within Python load this DLL.  This way, I can use the MS debugger to
look at what's going on within my classes while I manipulate them in
Python.
    The problem with this is that I can only debug the DLL whose project
I used to start Python; I don't know how to, for example, write a new
DLL and switch to debugging the new one.

What I want to do:
    I would rather run Python from within Windows, load various DLLs,
and debug whichever DLL I need, or a new one that I write, while
remaining in the same Python execution.

Does anyone know if it's possible to do this?  Can you tell me how?

Thanks for any help,
Tom Adelman
ta16 at cornell.edu







From mlongval at geocities.com  Tue Jul 27 23:45:19 1999
From: mlongval at geocities.com (Michael Longval MD)
Date: Tue, 27 Jul 1999 23:45:19 -0400
Subject: Am stupid and now going crazy...
Message-ID: <Pine.LNX.4.10.9907272336360.4460-100000@vagabond.longval.ca>

Hello all,

Here is the problem:

I upgraded to Mandrake6.0 (RedHat6.0) on my ThinkPad 380D.

I want to use the same configuration as I had under RH5.2, which included
MySQL and the Python module MySQLmodule.so wich was loaded dynamically.

I have tried the following:

1) Using the old MySQLmodule.so ---> doesn't work... why exactly I'm not
sure.

2) Tried to compile the standalone module MySQLmodule.so as describe in
the docs, does not work.... complicated error message.

3) Tried to compile a whole version of Python 1.5.2 with MySQLmodule
included in the source tree.  Couldnot find the mysqlclient.so module.

So here I am quietly thinking maybe I should throw the whole thing out the
window..... :-)

Does anybody have working modules:

	MySQLmodule.so
	mysqlclient.so

Or even perhaps a version of Python 1.5.1 or 1.5.2 with MySQL support
included?


Thanks very much.

Micheal J. Longval, MD
mlongval at geocities.com





From gmcm at hypernet.com  Tue Jul  6 13:37:05 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 6 Jul 1999 12:37:05 -0500
Subject: AttributeError - a dumb question
In-Reply-To: <Pine.LNX.3.96.990706182058.780A-100000@Johann>
References: <Pine.LNX.4.04.9907061943010.21542-100000@emerald.netskate.ru>
Message-ID: <1280843419-65357321@hypernet.com>

> Hallo Oleg,
> 
> Thanks for your reply.  Your answer helped me to back to the
> original problem I had:

Look at the case of the "L" in Krylys and KryLys.

Now kick yourself, hard!

> class F:
>     def __init__(self):
> 
>         self.x = self.Krylys()
> 
>     def KryLys(self):
>         l = [1,2,3,4,5]
>         return l
> 
> 
> l = F()
> 
> Now this assignment fails:
> 
> Traceback (innermost last):
>   File "<stdin>", line 11, in ?
>   File "<stdin>", line 4, in __init__
> AttributeError: Krylys
> 
> I hope you can help me with this one.
> 
> Johann
> 
>  -------------------------------------------------------------------
>  -------
> | Johann Spies                                 Windsorlaan 19              |
> | jhspies at futurenet.co.za                3201 Pietermaritzburg             |
> | Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
>  -------------------------------------------------------------------
>  -------
> 
>      "Ask, and it shall be given you; seek, and ye shall
>       find; knock, and it shall be opened unto you."        
>                                   Matthew 7:7 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list

- Gordon



From ivanlan at callware.com  Thu Jul 29 18:31:39 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Thu, 29 Jul 1999 22:31:39 GMT
Subject: Need help with TCP/IP client access from Windows
Message-ID: <37A0D64B.3FBD73D5@callware.com>

Hi All--

Gordon "Mr. Spaceman" McMillan wrote:
> 
> Eric Raymond writes:
> 

[snippet]

> However, this worked fine for me:
> 
> browse='"C:\\Program Files\\Netscape\\Communicator\\Program\\netscape"
> %s'
> 

Won't work for me--I almost never install NS in the assumed default
path.  In fact, the only things I install in the default paths (or even
on the c: drive) are Microsoft goobies, which hate it if you try to
confuse them.  Plus I have some systems that have no NetScape on them,
'cause I have to use IE5 for work.  Bleaugh.

> Note the double quotes around the nowhere-but-Windows <snicker> path
> with spaces in it, and that argv[1] is expected to be a URL (same for
> IE). Also note that the path doesn't go thru the c runtime lib, so
> those wonderful backslashes can't be avoided.
> 

Hey, I got spaces in my directories on Linux!  Had'm for years.  WFFM!

<there's-always-someone-out-to-puncture-your-balloon>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From gmcm at hypernet.com  Sat Jul 10 00:07:20 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 9 Jul 1999 23:07:20 -0500
Subject: _tkinter problem
In-Reply-To: <ulpv21popc.fsf@runner.princeton.idaccr.org>
Message-ID: <1280546566-83222563@hypernet.com>

Use Tk 8.0, or try Dieter Maurer patches, posted early June.

Victor S. Miller wrote:

> I had successfully built Python 1.5.2 with _tkinter using the system
> installed versions of tk and tcl (4.1 and 7.5 respectively). 
> However, I decided to download the new versions of tcl and tk (8.1
> for both), and build them.  They both built without problems, and
> ran through the test sucessfully.  I then rebuilt Python with them
> (I'm running on Solaris 2.6), which appeared to build successfully,
> however, when I try Tkinter._test(), I get:
> 
> Python 1.5.2 (#5, Jul  9 1999, 15:36:18)  [GCC 2.7.2.2] on sunos5
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import _tkinter
> >>> import Tkinter
> >>> Tkinter._test()
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "/usr/local/coop/lib/python1.5/lib-tk/Tkinter.py", line 1947,
>   in _test
>     root = Tk()
>   File "/usr/local/coop/lib/python1.5/lib-tk/Tkinter.py", line 886,
>   in __init__
>     self.tk = _tkinter.create(screenName, baseName, className)
> TclError: Can't find a usable init.tcl in the following directories:
> 
> 
> 
> 
> 
> This probably means that Tcl wasn't installed properly.
> 
> >>> 
> 
> 
> Now there's a perfectly good init.tcl in the directory
> /usr/local/coop/lib/tcl8.1, which is where make install placed it.
> Who is actually complaining about this, and how can I fix it? --
> Victor S. Miller     | " ... Meanwhile, those of us who can compute
> can hardly victor at idaccr.org    | be expected to keep writing papers
> saying 'I can do the CCR, Princeton, NJ   | following useless
> calculation in 2 seconds', and indeed
>     08540 USA        | what editor would publish them?"  -- Oliver
>     Atkin
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list

- Gordon



From roy at popmail.med.nyu.edu  Fri Jul  9 08:26:45 1999
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Fri, 09 Jul 1999 08:26:45 -0400
Subject: Two Pythons talking to each other?
References: <199907080019.TAA22244@buffalo.fnal.gov> <199907080529.HAA19502@axil.hvision.nl> <3784486a.7993113@news.ricochet.net> <7m2lol$bnm$1@nntp6.u.washington.edu> <37859856.9981682@news.ricochet.net>
Message-ID: <7m9tm1$6qg$15@rks1.urz.tu-dresden.de>

phil at ricochet.net (phil) wrote:
> Yes, if you are stated or you care about order then you probably
> need to sequence.

If you need to sequence datagrams, then you most likely might as well just
be using TCP instead of UDP.  It is extremely unlikely that any sequencing
you will do in user code will be faster/smaller/better/less-buggy/whatever
than the kernel code that already does it for you.

-- 
Roy Smith <roy at popmail.med.nyu.edu>
New York University School of Medicine




From larsga at ifi.uio.no  Thu Jul  1 13:11:53 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 01 Jul 1999 19:11:53 +0200
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net>
Message-ID: <wkzp1g2t2e.fsf@ifi.uio.no>

* William Tanksley
| 
| (Yes, I know -- the ancestry goes the other way around: Lisp ->
| Scheme -> Forth.)

Huh? Forth dates back to the 60s, whereas Scheme is from 1975, and I'm
quite unsure of whether Chuck Moore knew Lisp at all.

--Lars M.



From vaton at postoffice.pacbell.net  Sat Jul  3 00:17:20 1999
From: vaton at postoffice.pacbell.net (Rob)
Date: Fri, 02 Jul 1999 21:17:20 -0700
Subject: How do you make clickable Python/Win9x scripts?
Message-ID: <377D8ED0.49BEF319@postoffice.pacbell.net>

Is there a way to make a Python script file a clickable .BAT file in
Win9x?
It's possible with Perl, and looks something like this:
@rem = '
@echo off
C:\perl\perl5\bin\perl -x -S %0 %1%
goto endofperl
@rem ';
#!.\perl\bin\perl -w

print "Hello World!\n";
__END__
:endofperl

The above Perl file would be named world.bat.





From mhuster at hotmail.com  Wed Jul 21 17:54:05 1999
From: mhuster at hotmail.com (mhuster at hotmail.com)
Date: Wed, 21 Jul 1999 21:54:05 GMT
Subject: Repost: Re: Scientific Plotting?
References: <379392E7.EB0F29E7@ssec.wisc.edu>
Message-ID: <7n5fhn$9e5$1@nnrp1.deja.com>

Repost:
My apologies for mangling my return address on the post. I wanted to
avoid more spam by the mangling. My e-mail address was included at the
bottom. Hmmm. If only I could count the number of spam e-mails I will
get from this repost,...

Original reply:

I am also trying to find an adequate *object* based plotting package. I
am writing some signal processing software and want to code plotting
into the classes I am defining.

Some requirements:
1) Postscript output (for inclusion in high quality Word & Powerpoint
documents) GIF, JPG, etc just do not have the resolution. They default
to 72 dpi rendering. You can overcome this on Linux using Image Magick's
convert -dens 300. The resulting files are large.

2) Object oriented
3) Images

Our current assessment:

GIST: not enough flexibility in configuring the plots. It has the
Postscript we need, though. We have it coded in now for some look at the
data. But it is not the long term solution.

Plplot: uses stroke fonts, even in postscript output. We cannot get a
uniform look with matlab produced plots. Bummer. Otherwise it looked
good. I am afraid when blown up the fonts will look even worse. For
example. To get a bold font you have to play with pen width and font
size. Isn't "Helevetica-Bold" easier to specify? Also, it is at a
developmental (& FORTRAN) deadend. No chance to change things.

Gnuplot: Runs gnuplot through a pipe. All of the data (upto many MB)
also has to go through the pipe. Unsat. I want to plot efficiently.

plotlib: not enough high level wrappers. I do not want to have to
calculate default axis size, where to put tick marks, etc.

Pgplot: looks like plplot.

Piddle & graphite: looks good, but not mature enough. We are keeping an
eye on it. Probably will try it.

PyMat: I want to get away from needing Matlab licenses

I am kind of bummed. If we wrap intelligent python wrappers around
something, we will post it.

Michael Huster
mhuster at hotmail.com



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From g_will at cyberus.ca  Mon Jul 19 12:18:58 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 19 Jul 1999 16:18:58 GMT
Subject: tkinter call backs
Message-ID: <01bed202$58879280$e855c3d1@Cyberus.cyberus.ca>

I have a simple problem that I cant find the solution.  I have a menu item
to open a file which has a command "openfile".  I would like to activate
"tkFileDialog.askopenfilename" to get the name of the file.  This requires
"parent=root" or something like that for the position of the message box.

How do I pass "root" into openfile?  

Thanks
Gordon Williams



def openfile():
    print "opening file"
    fileName=tkFileDialog.askopenfilename(parent=root, title="Open")
    if fileName:    #checks for cancel
        print fileName  #name of file to open
        file=open(fileName, 'r')
        read='junk' #initialize read with something
        while read != "":
            read=file.readline()
            print "number1",read[0:-1] # strips off line return character
#       This has one more line than needed due to EOF line.
#       Will use seek() to move to 2nd line
        print "number of bytes ",file.tell()
        file.close()



From janssen at parc.xerox.com  Wed Jul 14 12:18:36 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Wed, 14 Jul 1999 09:18:36 PDT
Subject: problem compiling 1.5.2 with egcs-1.1.2 on Solaris 2.6
Message-ID: <99Jul14.091842pdt."3556"@watson.parc.xerox.com>

I get the following error when compiling Modules/signalmodule.c on
Solaris 2.6 with egcs 1.1.2.  It works fine with the Sun SunPro compiler.

Bill

gcc -ansi  -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -c ./signalmodule.c
./signalmodule.c:104: `SIGMAX' undeclared here (not in a function)
./signalmodule.c:104: size of array `Handlers' has non-integer type
./signalmodule.c: In function `signal_signal':
./signalmodule.c:230: `SIGMAX' undeclared (first use in this function)
./signalmodule.c:230: (Each undeclared identifier is reported only once
./signalmodule.c:230: for each function it appears in.)
./signalmodule.c: In function `signal_getsignal':
./signalmodule.c:281: `SIGMAX' undeclared (first use in this function)
./signalmodule.c: In function `initsignal':
./signalmodule.c:367: `SIGMAX' undeclared (first use in this function)
./signalmodule.c:381: storage size of `act' isn't known
./signalmodule.c: In function `finisignal':
./signalmodule.c:592: `SIGMAX' undeclared (first use in this function)
./signalmodule.c: In function `PyErr_CheckSignals':
./signalmodule.c:628: `SIGMAX' undeclared (first use in this function)
*** Error code 1
make: Fatal error: Command failed for target `signalmodule.o'




From holger at phoenix-edv.netzservice.de  Fri Jul 16 07:37:42 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Fri, 16 Jul 1999 11:37:42 GMT
Subject: getting values of global variables?
References: <1280074494-16090399@hypernet.com>
Message-ID: <378F1986.36A8E3F2@phoenix-edv.netzservice.de>

Hugh Gordon,

remember me to spent an cup of ice, when we meet us;-))

Gordon McMillan wrote:
> 
> Holger Jannsen writes:
> >
> > ok, sounds easy, but I get a little confused...;-(
> 
> That's because what's confusing you isn't what you think is confusing
> you!

Ah, yeah!!!!

> > ...
> >   splitter=re.split('(%.*?%)',aString)
> >  except:
> >   return aString
> 
> Judging by the format of your input, a better regex would be
> '%[^ ]*?%' - which give you ' %1 %2 ' instead of a phony match.

Thanx for involving, but [sorry;-)] you've gorgotten that there could be 
spaces within that tokens, too. Check it out by setting a 
environment-variable like "set HEIN BLOED=ON". You would have missed
that token in common with execution of your expression.
That's a real problem, because there could be spaces in keys of
possible other directories, too. E.g.:
aDict={}
aDict["my house"]="very nice"
Ok, I could restrict the tokens to be without any spaces, but it's
not really the thing i like... (see below!)

> >...[better not show again;-))]
> Sorry, Holger, but that's a mess! Something like this will get you
> away from all those special cases:
> 
>  rslt = []
>  for tok in splitter:
>      if tok[0] != '%' or tok[-1]!= '%':
>          rslt.append(tok)       #not a special token
>      else:
>          test = tok[1:-1]               #strip those '%'s
>          rslt.append(checkForToken(test))
>  return string.join(rslt)

Ok, I see, but the right way is:
return string.joinfields(rslt, '')

> > def checkForToken(aString):
> >  "checks a string for known tokens"
> >  try:
> >   aTok = eval(aString)
> >   if aTok==float(aString): #don't exchange numbers!
> >    return None
> 
> Well, eval('HEIN') does indeed produce 'Bl?d'. But float('HEIN')
> raises an exception, which is caught by code that assumes it's "eval"
> that failed.

Ooouch! That's one of the 'four blind mice', isn't it?

> >...
> How about (leaving out directories and files, which you haven't
> included):
> 
> def checkForToken(aString):
>  "checks a string for known tokens"
>  print 'checking', aString
>  try:
>   aTok = eval(aString)
>  except:
>   aTok = os.environ.get(aString)
>   if aTok==None:
>    aTok = aString
>  if type(aTok) != type(''):
>      aTok = str(aTok)
>  return aTok

The directories- and file-dictionaries do I get from another class of
that module. They are not important to understand the difficulties of that problem. 
Thy just act like os.environ (see below).

I think my problem is to get that values of the dict.keys with spaces 
inside or appending, so e.g. "%HEIN BLOED%" or "%HEIN %".
They are not possible as global variables inside python-code, but
even in all dictionaries e.g. like os.environ!

And there is another difficulty problem: 
I declared, that a standalone string also have to be tested, so e.g. 
egalizeTokens("WINDIR") must return "c:\WinNT". It's the same
as egalizeTokles("%WINDIR%")!
Therefore I put
	if len(splitter)==1:
		return checkForToken(aString)                
inside that function egalizeTokens.
OK. But what to do with several strings, that could be interpreted
as numbers??? So I often have to interprete version-strings like '1-2-22'.
eval('1-2-22') = -24!!!;-(
How could I forbid that, but let %HEIN%=3 an allowed token? Because of that
I'd written that code with 'if float(...)'!

Do ya' understand my problem?

Ciao,
Holger



From jlj at cfdrc.com  Fri Jul  2 13:51:06 1999
From: jlj at cfdrc.com (Lyle Johnson)
Date: Fri, 02 Jul 1999 12:51:06 -0500
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <377a6bb0.10820509@news.oh.verio.com> <377BDEF7.C34A6088@compaq.com>
Message-ID: <377CFC0A.56AFF110@cfdrc.com>

Yeah, I got it wrong the first time. It should read:

    http://www.cfdrc.com/FOX/fox.html

I confused it with the ftp URL, which does need the "pub" in there.

Greg Ewing wrote:

> Kevin Dahlhausen wrote:
>
> > Lyle Johnson <jlj at cfdrc.com> wrote:
>
> > >Just to add fuel to the fire, we now have a Python interface to FOX, a
> > >cross-platform GUI toolkit:
>
> > >    http://www.cfdrc.com/pub/FOX/fox.html
>
> This link doesn't work for me. Has
> it been moved somewhere else?
>
> Greg




From jcw at equi4.com  Thu Jul 22 18:18:36 1999
From: jcw at equi4.com (Jean-Claude Wippler)
Date: Fri, 23 Jul 1999 00:18:36 +0200
Subject: Fw: ANNOUNCE: Mailing list for Australian Python Users
References: <007201bed42e$c4316c30$f29b12c2@secret.pythonware.com> <wkogh56j2o.fsf@ifi.uio.no> <7n78nm$lic$1@news.accu.uu.nl>
Message-ID: <379798BC.33FC0CA9@equi4.com>

Martijn Faassen wrote:
> 
> Lars Marius Garshol <larsga at ifi.uio.no> wrote:
> 
> > * Fredrik Lundh
> > |
> > | python-se, n?gon?  Link?ping?  :-)
> 
> > Tjah. Si Oslo, s? kan vi bli en del. :-)
> 
> Utrecht is veel beter denk ik, Python-NL en zo..

Doen.  Ik kom op de fiets.

> An-large-percentage-of-pythoneers-speaks-Dutch-anyway-ly yours,

De meerderheid, toch?

-- Jean-Claude



From Joshua.Rosen at p98.f112.n480.z2.fidonet.org  Thu Jul  1 13:31:17 1999
From: Joshua.Rosen at p98.f112.n480.z2.fidonet.org (Joshua Rosen)
Date: Thu, 01 Jul 1999 18:31:17 +0100
Subject: scope-resolution
Message-ID: <0000164d@bossar.com.pl>

From: Joshua Rosen <rozzin at geekspace.com>

Eckh--apparently, I had a dead-brain day, and took you much too
seriously.
Of course, there's got to be someone out there who needed to here that
stuff....

I imagine that you were using typing as a metaphor for scoping, and,
now that I see it, I can appreciate it.

It would be nice to have some non-mandatory scope-resolution
capabilities beyond self-referencing in instance-methods and `global'
statements.

What would be a different way of getting the effect of:

	ClassX.foo(ClassY())
?

Have bound and unbound methods be different creatures, perhaps....

No, never mind, I suppose--thinking about it, so long as you can
assure `this variable is a class variable', then everything should be
fine. Actually, just adding the instance's namespace to the list of
namespaces searched, in methods, shouldn't be too bad to implement,
and probably wouldn't break any of the existing code, eh? [function
locals, instance locals, globals]? 

		-Rozzin.



From Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org  Thu Jul  1 23:47:15 1999
From: Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 04:47:15 +0100
Subject: win32com.client -- beginner question
Message-ID: <00001652@bossar.com.pl>

From: mikael at toolwood.com (Mikael Lyngvig)

On Wed, 30 Jun 1999 08:43:06 +0200, Axel Schmitz-Tewes
<uzs59g at uni-bonn.de> wrote:

>Hello,
>
>I ' ll try to use the comclient ability of python. In the Doku I have
>found
>the example:
>
>o = win32com.client.Dispatch("Excel.Application")
>o.Visible = 1
>o.Workbooks.Add() # for office 97 ? 95 a bit different!
>o.Cells(1,1).Value = "Hello"
>
>Because I have installed Excel95 I have given it a try and that's the
>result:
>
>Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
>Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>> import win32com.client
>>>> o= win32com.client.Dispatch("Excel.Application")
>>>> o.Workbooks.Add()
>Traceback (innermost last):
>  File "<stdin>", line 1, in ?
>  File "E:\Programme\Python\win32com\client\dynamic.py", line 400, in
>__getattr_
>_
>    raise AttributeError, "%s.%s" % (self._username_, attr)
>AttributeError: Excel.Application.Workbooks
>
>But Workbooks.Add is a available method as I can test with VBScript.
>Is my installation corrupt or do I do some silly mistake or perhaps some
>more serious?

Did you remember to run the makepy.py script in
Python\win32com\client?  You need to run that one, select the COM
interface you want to use, e. g. "Microsoft Excel 5.0 Object Library
(1.0)" (or similar).  makepy.py then generates the necessary Python
definitions, in a file in the Python\win32com\gen_py directory (named
according to the COM interface's GUID.

After that you should be able to access Excel without too much
trouble.

There's a introduction to the Win32COM stuff, which includes a brief
description of makepy.py, in the QuickStartClientCom.html file
included in the Win32COM installation.


-- Mikael



From greg.ewing at compaq.com  Thu Jul  8 16:39:18 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 09 Jul 1999 08:39:18 +1200
Subject: Two Pythons talking to each other?
References: <14212.51977.601241.772478@buffalo.fnal.gov> <Pine.SGI.3.95.990708120053.9140Q-100000@hppc.fnal.gov>
Message-ID: <37850C76.E873B866@compaq.com>

"petravick at FNAL.GOV" wrote:
> 
> AF_INET will run about the whole protocol stack, AF_UNIX
> is usually implemented to be more like a pipe...

On BSD systems, pipes used to be implemented as a
pair of unix-domain sockets. Don't know if it's
still done that way.

Greg



From pereira at research.att.com  Sat Jul 17 18:35:33 1999
From: pereira at research.att.com (Fernando Pereira)
Date: Sat, 17 Jul 1999 22:35:33 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com> <170719990020380651%pereira@research.att.com> <7mqaht$r27$2@cronkite.cc.uga.edu>
Message-ID: <170719991835331004%pereira@research.att.com>

In article <7mqaht$r27$2 at cronkite.cc.uga.edu>, Graham Matthews
<graham at sloth.math.uga.edu> wrote:

> <graham at sloth.math.uga.edu> wrote:
> : > But the class of 
> : > languages recognisable by automaton doesn't cover all the languages of
> : > interest (in fact my guess is it covers very few ... context ...).
> Fernando Pereira (pereira at research.att.com) wrote:
> : In what sense of "very few"? Most programming languages are designed to
> : have an LR(k) or LL(k) backbone, and regular lexical structure 
> 
> So the regular expressions help you with the regular lexical structure.
> But they don't help with the LR(n) bit. They also don't help with context
> dependencies. That leaves out a lot of languages.
By "automaton" you seem to have misunderstood "finite-state automaton."
I was referring more generally to determinizable classes of automata,
eg. determinizable PDAs. Hence my reference to LR and LL constructions,
which effectively attempt to determinize PDAs (and fail with conflicts
if they cannot). 
> 
> <graham at sloth.math.uga.edu> wrote:
> : > And then
> : > there are software engineering issues ...
> Fernando Pereira (pereira at research.att.com) wrote:
> : Agreed. However, more traditional lexer- and parser-generators had
> : sensible answers to many of those issues, that might not be as elegant
> : as monadic approaches, but that paid serious attention to optimization
> : in the regular and CF backbones. 
> 
> I don't really see why using parser combinators means that you cannot
> optimise. You are not precluded from using regular expression parsers
> as primitive parsers, nor are you precluded from using LALR(n) parsers
> as primitive parsers. Hence you can optimise those parts of your
> parsing with automaton if you want. So I don't really understand your
> criticism.
The monadic approach mixes up the search structure for pattern matching
or parsing (what is elsewhere represented as an automaton) with other
data processing. Unless your optimizer can solve the halting problem,
it will not be able to determinize otherwise determinizable control
structures because it cannot figure out the data dependencies.
> 
> Also as I said parser combinators are really more of a sequencing 
> technique -- one that applies nicely to parsing.
> 
Sure, but so what? Automata are too a sequencing technique, usable for
many other things besides pattern matching and parsing. The fundamental
difference is that automata-theoretic approaches separate neatly the
representation of sequences from what one wants to do with the parts of
those sequences, allowing optimizations that are difficult to obtain
within general-purpose programs.
> Fernando Pereira (pereira at research.att.com) wrote:
> : By the way, I find it faintly amusing
> : to see the functional programming community rediscovering and dressing
> : in categorial garb a simple idea that was a main motivator for the
> : development of logic programming in the early 70s (see
> : <http://akpublic.research.att.com:9000/~pereira/bib.html#Pereira+Warren-
> : 80:DCGs> for a survey). 
> 
> Yes combinators are much like definite clause grammars in prolog. But if
> I remember correctly from my use of DCG's, they gave you a fixed set of
> combinators?
It's easy enough to define one's own combinators, but not as elegantly
as in a higher-order functional language. 
>Moreover they were limited to string parsing. 
Not true. There's nothing in DCGs that ties them to strings. The
underlying programming pattern (often called "difference lists") is in
fact a general accumulator pattern that has been used in many other
applications, eg. code generation.
> Fernando Pereira (pereira at research.att.com) wrote:
> : It *is* a good idea for rapid prototyping, but
> : interspersing pattern matching/parsing with arbitrary computations
> : (even purely functional ones) effectively blocks optimization, which is
>                                 -------------------------------
> 
> Can you explain why? I don't see it, since you can have reg expr parsers
> and other such fast parsers as primitive parsers that you combine ...
Either you put all the interesting stuff in separate primitive parsers,
in which case I'm not sure what's the point of the monadic approach, or
you combine parsing and other data processing more tightly, which takes
advantage of the monadic approach to achieve greater conciseness, but
then it is hard to optimize the parser because the other data
processing is in the way of the necessary analyses. Focusing on
determinization (a crucial part of regular and LR constructions), you
need to determine whether two alternative computation paths that
consume the same input sequence can be merged into a single path. In
general, this is undecidable since it requires figuring out whether two
arbitrary functions (those associated to those two paths) produce the
same output. Now, it might be interesting to see whether compiler
optimization and code rewriting techniques could be used in trying to
recognize possible otimizations, but the general point stands that
tangling pattern-matching or parsing with general computation makes
optimization very difficult.

-- F

-- 
-- F



From tbryan at arlut.utexas.edu  Fri Jul 16 21:43:33 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Fri, 16 Jul 1999 20:43:33 -0500
Subject: propapating exceptions
References: <slrn7ovfss.4bc.quinn@lira.ugcs.caltech.edu>
Message-ID: <378FDFC5.AE965089@arlut.utexas.edu>

Quinn Dunkan wrote:
> 
> I've lately found myself in the position of wanting to catch exceptions
> based on their attributes, rather than on the exception itself.  So (in
> mal's odbc module):
> 
> try:
>         some sql command
> except ODBC.MySQL.OperationalError, msg:
>         if msg[1] == CR_SERVER_GONE_ERROR: # (snarfed out of errmsg.h)
>                 restart server, try again
>         else:
>                 raise sys.exc_info()[0], sys.exc_info()[1]
> 
> It's a rather wordy way to pass the exception on without touching it,
> and I still lose the traceback info.  Is there a better way to pass on
> an exception as if it had never been caught?


I can't find the reference now, but I think that I read in 
the Python docs that calling raise without an argument inside 
an except, reraises the error.

That is, 

try:
        some sql command
except ODBC.MySQL.OperationalError, msg:
        if msg[1] == CR_SERVER_GONE_ERROR: # (snarfed out of errmsg.h)
                restart server, try again
        else:
                raise

Does that work?

-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From perry at ac-copy.net  Thu Jul  8 16:55:27 1999
From: perry at ac-copy.net (Joachim Schmitz)
Date: Thu, 8 Jul 1999 20:55:27 GMT
Subject: why is there no put in urllib ?
Message-ID: <Pine.LNX.4.05.9907082250040.11616-100000@ns.ac-copy.net>

hi,

may be this is a total dump question. But it would allow us Zopistas to
upload our dtml pages with a simple python script to the Zope-server.
 

Gru?

Joachim Schmitz

WWW-Consultant

email: js at ac-copy.net
tel: +49-241-89491-0
fax: +49-241-89491-29





From BobAlex at uppercase.xerox.com  Thu Jul  8 17:59:08 1999
From: BobAlex at uppercase.xerox.com (Alexander, Bob)
Date: Thu, 8 Jul 1999 14:59:08 -0700 
Subject: Regular expression (re) anomaly with group names and alternatives
Message-ID: <AF223B12F1FDD2119253006097C4FCEC0AD159@cats.pdri.xerox.com>

It would be nice (IMO) if group names could be duplicated in different
alternatives (sections separated by the | operator) of a regular expression,
with the name taking on the value in the alternative that matches (if any).
However, that doesn't work -- is that a bug or by design?

In the following Python session, the pattern is such that the name A can be
assigned in either of the two alternitives. Since only one alternative can
succeed, it seems that A should have the value of whichever alternative
succeeded. In this example, the first alternative succeeds, which should
assign 'a' to A. However, the presence of A in the second alternative, which
didn't even take part in this match, causes A to be set to None.

>>> import re
>>> m = re.match('(?P<A>.)(?P<B>.)|(?P<A>x)', 'ab')
>>> m
<re.MatchObject instance at 7fe110>
>>> m.groupdict()
{'B': 'b', 'A': None}

My vote: it's a bug. This  would be a useful behavior in some cases, such as
this one where I discovered this anomaly:

	r'("(?P<reference>[^"\n]*)(?P<completed>"?)|(?P<reference>.*))'

After the match succeeds, I could simply use group('reference'), rather than
write some additional code to analyze the situation.

Bob



From bwinton at tor.dhs.org  Fri Jul 23 00:02:50 1999
From: bwinton at tor.dhs.org (Blake Winton)
Date: Fri, 23 Jul 1999 04:02:50 GMT
Subject: Printable Python reference
References: <3797782A.FD39EF04@newbridge.com> <m31ze0kwmc.fsf@havenrock.com>
Message-ID: <slrn7pfqsb.f9h.bwinton@tor.dhs.org>

On 23 Jul 1999 10:01:15 +0900, Matt Gushee <mgushee at havenrock.com> wrote:
>Nick Taylor <ntaylor at newbridge.com> writes:
>> I decided to take a look at Python today for the first time, and it
>> sounds very interesting.  
>It is!

I'll second that...

>> Does anyone know of a printable (i.e. PDF,
>> PostScript, or plain-text) reference out there?
>Try ftp://ftp.python.org/pub/python/doc . You'll find PS, PDF, and
>LaTeX documents. They're packaged as .tar.gz archives, though, which
>might be a problem if you run Windoze. Maybe there are ZIP versions
>somewhere.

I'll only do this once, I promise.

.tar.gz files are no problem with Windows.  WinZip handles them better
(IMHO) than "tar -xvzf".  I'm forced to use NT during the day, and it's
not my favourite, but there are a few things it does really well, and
uncompressing stuff is one of them.  (The other two are COM and ASP.
COM because there's nothing like it in the Unix world.  (Don't even
mention CORBA.  The benefit of COM is that _everybody_ uses it, and so
it's useful.  The disadvantage of CORBA is that almost no-one uses it,
and so it's useless.)  ASP because you can use any scripting language
you want (provided it has a couple of hooks).  PHP is decent, but I
don't see why I should have to learn yet another scripting language when
I already know VBScript, JavaScript, Perl, Bash, Python, etc...)

Anyways, that was my rant.  I'll stop now.

Later,
Blake.

-- 
One Will.  One Dream.  One Truth.  One Destiny.  One Love.



From MHammond at skippinet.com.au  Wed Jul 21 18:09:44 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Thu, 22 Jul 1999 08:09:44 +1000
Subject: Spreading the word about Python; ?s and ideas
References: <199907211307.IAA14980@cro.ots.ac.cr>
Message-ID: <7n5gc7$2up$1@m2.c2.telstra-mm.net.au>

Isidor wrote in message <199907211307.IAA14980 at cro.ots.ac.cr>...

> To reiterate (more concisely):

> There may be a lot of office scripters out there, and they are
> probably more than ready to hear the word of python. Targeting
> them through marketing and educational materials might be an easy
> way to get a lot of people quickly ramped up to more serious

Absolutely.  This is an excellent idea,  A tutorial aimed at this segment
would be excellent.  You could introduce COM very early (and very vaguely).
The point is they end up doing all the same things they can do with Word etc
very early on.

...

[snip even more good ideas]

> Ok, that's the end of my rant. I hope this is a useful catalyst
> for discussion.

Problem is that quite a few of these ideas have previously been identified.
Python is not short of good ideas, just short of people to implement them.

The best and incredibly useful thing you personally could do for Python is
to step up and start doing these.  Your Office/VBA background would make you
uniquely placed to be able to write the tutorial you mention - and you
should do it before you forget exactly what you struggled with :-)

So take some first steps, then post your results.  These are always better
than posting good ideas, and could then generate some real discussion.  Post
a first-start to something, and people may well give you some excellent
comments and feedback.

The _excellent_ thing about that is that all these smart people give you
excellent feedback, you adopt the feedback and your tool/documentation moves
from quite ordinary to damn good.  The great thing is you still get all the
credit, but really everyone else made it what it is :-)

Mark.






From arnold at dstc.edu.au  Tue Jul 20 23:23:51 1999
From: arnold at dstc.edu.au (David Arnold)
Date: Wed, 21 Jul 1999 03:23:51 GMT
Subject: Merging two lists as a dictionary 
In-Reply-To: Your message of "Tue, 20 Jul 1999 21:19:46 EST."
             <19990720211946.A12202@pooh.frostnet.net> 
References: <19990720211946.A12202@pooh.frostnet.net>
Message-ID: <199907210323.NAA12503@piglet.dstc.edu.au>

-->"Chris" == Chris Frost <chris at frostnet.advicom.net> writes:

  Chris> I have two lists and would like to merge them into a
  Chris> dictionary, with one list forming the keywords and the other
  Chris> the values in the dictionary. How c= ould this be done?
  Chris> "map(None, keyword_list, value_list)" looked great, but it
  Chris> returns a list of pairs (I'd like to be able to specify
  Chris> values according to keywords...).

it's ugly, but ...

reduce(lambda d,i,setitem=operator.__setitem__: setitem(d,i[0],i[1]) or d, map(None, keyword_list, value_list), {})




d



From janssen at parc.xerox.com  Tue Jul 20 14:34:31 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Tue, 20 Jul 1999 18:34:31 GMT
Subject: Python suitability questions
In-Reply-To: <7n2a9p$trq$1@news.accu.uu.nl>
References: <0CCE781E44F2C872.F528377A62B41DCA.73CAA63AD6E79314@lp.airnews.net>
	<7n2a9p$trq$1@news.accu.uu.nl>
Message-ID: <grZA4rsB0KGW1DZnsu@holmes.parc.xerox.com>

Excerpts from ext.python: 20-Jul-99 Re: Python suitability ques..
Martijn Faassen at vet.uu.n (1737)

> There are various ways to integrate Python with C or C++ code. You can
> hand code things of course, and I've heard Python's extending/embedding
> API is quite nice.

On this topic, you can also use ILU to integrate Python with C or C++,
in somewhat the same fashion as SWIG.  You describe the API with ISL,
implement the true side with Python, and call the surrogate-side stubs
from C or C++.  ILU handles the gluing of the Python to the C or C++.

Another advantage of doing this with ILU is that you only have to do it
once.  After that, you can load the same Python implementation into C,
C++, and Python (of course), with ILU 2.0alpha14.  With the next ILU
release, you'll also be able to load that module into languages with
threaded runtimes, such as Java and Common Lisp.  And you can do it
without having to use the Python API to C, or the Java API to C, or ...
-- your Python-side programming is in Python, the ILU code does all the
C or C++ glue.

There's an example of this in ILUSRC/examples/multlang/, and the next
release will also have examples of loading Python into Java and Common
Lisp via this mechanism.

Bill



From news at dorb.com  Thu Jul  1 13:38:17 1999
From: news at dorb.com (Darrell)
Date: Thu, 1 Jul 1999 13:38:17 -0400
Subject: getopt and Non-Unixes?
References: <19990701174416.A27152@optiplex.palga.uucp>
Message-ID: <sKNe3.4318$U6.23620@newsr2.twcny.rr.com>

You end up using -v -d -f or -vdf this last one will throw Dos people for a
loop.
I supply the help message if the command line isn't understood.

def runIt():
 try:
  opts, args = getopt.getopt(sys.argv[1:], "l:hdpf:o:")
 except getopt.error, msg:
  sys.stderr.write("Error: %s\n" % str(msg))
  print help_str
  sys.exit(1)


--
--Darrell
Gerrit Holl <gerrit.holl at pobox.com> wrote in message
news:19990701174416.A27152 at optiplex.palga.uucp...
> Hello all,
>
> I was wondering how getopt worked on dos/windows (and other operating
systems).
> While in dos, the default way to handle arguments is "/v /d /f" (isn't
it?).
>
> I don't have access to any micro$oft machine (it doesn't make me
unhappy...),
> but I just wanted to know how my program works on Dos/windows/macintosch
and
> other non-Unix systems.
>
> regards,
>
> Gerrit.
>
>





From hschroeder at gmx.net  Fri Jul 30 05:58:49 1999
From: hschroeder at gmx.net (Henning Schroeder)
Date: Fri, 30 Jul 1999 11:58:49 +0200
Subject: How to set text color using pyKDE 0.8?
References: <7npb63$2fdn@enews3.newsguy.com>
Message-ID: <p0trn7.2h.ln@odin.henning-schroeder.de>

Peter Torstenson <p.t at iname.com> wrote:
> I'm using pyKDE 0.8 binding for python.
> I made a QLabel on the main widget. Now I'd like to change the color of the
> text inside this QLabel.
> I managed to change the backgroundcolor and the font. But I can't find any
> info on how to change the color of the text, or the foregroundcolor.
This is a little bit tricky. Every widget has a palette() function which
returns a QPalette. This palette consists of three color-groups which can
be accessed through normal(), disabled() and disabled() and can be changed
trought its corresponding set-functions (setNormal(), ...). According to
the state of the widget one of these color-groups is shown. You can
find more about setting up color-groups in the QT docs (e.g qcolorgroup.html).

Hope that helps
  Henning



----------------------------
#!/usr/bin/python

from kde import *

import sys
app = QApplication(sys.argv)
label = QLabel("Hello world!")

pal = label.palette()
pal.setNormal(QColorGroup(red, green, blue, black, gray, yellow, white))
label.setPalette(QPalette(pal))

label.show()
app.setMainWidget(label)
app.exec_loop()
----------------------------



From Robin.Becker at p98.f112.n480.z2.fidonet.org  Sat Jul  3 23:35:17 1999
From: Robin.Becker at p98.f112.n480.z2.fidonet.org (Robin Becker)
Date: Sun, 04 Jul 1999 04:35:17 +0100
Subject: ordered threading
Message-ID: <000016bb@bossar.com.pl>

From: Robin Becker <robin at jessikat.demon.co.uk>

In article <000a01bec583$30332b40$769e2299 at tim>, Tim Peters
<tim_one at email.msn.com> writes
>[Robin Becker]
>> I have an unbounded number of worker threads which occasionally require
>> access to a unique server thread. I can easily mutex access to the
>> server thread, but is there any obvious/easy way to ensure that the
>> blocked workers wake up in the right order?
>
>No threading system on earth allows specifying the order in which blocked
>threads will wake up, although some systems (not Python's) allow you to
>*influence* the order via assorted "priority" gimmicks.
I was being naive.
>
>You could easily enough build an OrderedLock class whose acquire and release
yes I guess that's what I'll try and do
>methods implement whatever scheme you want, hiding the unordered nature of
>an internal mutex under a layer of serial-number cruft.  But then why not do
>these things serially in a single thread to begin with?
the workers are HTTP requests in separate threads from Zope I guess. I
want to serialise access to the time/space hog. I guess it could run
uniquely in the worker threads ie the locked resource would be access to
the hog code. I'm not sure that would let me guarantee unlocking as
there are ways for the worker threads to suicide.
>
>If you have frequent rigid ordering requirements, threads are likely the
>wrong approach.
>
>> Also if blocked can a thread be killed in any way?
>
>No, not short of killing the whole program.  A thread blocked on a condition
>that will never become true is a symptom of confused application design, not
>of a weakness in the thread model <wink>.
>
>expand-the-condition-to-include-"or-i-should-die-now"-ly y'rs  - tim
>
>
>
>

-- 
Robin Becker



From phd at emerald.netskate.ru  Tue Jul  6 12:33:55 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Tue, 6 Jul 1999 20:33:55 +0400 (MSD)
Subject: mxCGIPython question (was: Build the Python library into the
 binary)
In-Reply-To: <37821C75.405CD9C5@lemburg.com>
Message-ID: <Pine.LNX.4.04.9907062032400.23992-100000@emerald.netskate.ru>

On Tue, 6 Jul 1999, M.-A. Lemburg wrote:
> The same problem seems to exist on Unix platforms. For some
> reason, the symbols defined in cgipython are not accessible
> to the dynamic linker on Unix. Don't know why, because 'nm cgipython'
> clearly shows them as being defined...

   I am not sure, but I think on most systems symbols should be exported
explicitly with -E linker flag (for gcc it is -Wl,-E option)

> Cheers,
> -- 
> Marc-Andre Lemburg
> ______________________________________________________________________
> Y2000:                                                   178 days left
> Business:                                      http://www.lemburg.com/
> Python Pages:                           http://www.lemburg.com/python/

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From arcege at shore.net  Fri Jul  2 13:50:04 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Fri, 02 Jul 1999 17:50:04 GMT
Subject: Language extensibility
References: <1283099377-77267484@hypernet.com>   <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> <wkzp1g2t2e.fsf@ifi.uio.no> <ey3r9mskvja.fsf@lostwithiel.tfeb.org> <x27lojoums.fsf@todday.aiai.ed.ac.uk>
Message-ID: <gZ6f3.650$6M6.197790@news.shore.net>

In comp.lang.python Jeff Dalton <jeff at todday.aiai.ed.ac.uk> wrote:
: Tim Bradshaw <tfb at tfeb.org> writes:

:> * Lars Marius Garshol wrote:
:> 
:> > Huh? Forth dates back to the 60s, whereas Scheme is from 1975, and I'm
:> > quite unsure of whether Chuck Moore knew Lisp at all.

: Well *Lisp* dates back to the late 50s.  Scheme may be from 75,
: but Lisp is a different matter.  And lambda calculus is, what,
: 1949?

Supposedly, Lisp was created in 1953, a year after Fortran, and is said
to be the second higher-level (non-assembly) language written.

But then... I wasn't around at that time, so what do I know. ;)

  -Arcege




From conrad at vanir.srl.caltech.edu  Sat Jul 10 14:24:33 1999
From: conrad at vanir.srl.caltech.edu (Conrad Steenberg)
Date: 10 Jul 1999 18:24:33 GMT
Subject: Printing returned object in embedded app
References: <378694FC.D857F68A@srl.caltech.edu> <6wxh3.1472$6M6.450919@news.shore.net>
Message-ID: <slrn7oep57.a57.conrad@vanir.srl.caltech.edu>

On Sat, 10 Jul 1999 01:40:18 GMT, Michael P. Reilly <arcege at shore.net> wrote:
>
>Try:
>  PyObject_Print(object, stdout, Py_PRINT_RAW);

Thanks! 

Conrad
-- 



From mstenber at cc.Helsinki.FI  Fri Jul 30 07:23:17 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 30 Jul 1999 14:23:17 +0300
Subject: NT select.select?
Message-ID: <al8zp0epeje.fsf@sirppi.helsinki.fi>

Is there some low limit on maximum number of sockets you can have in the
Python-NT's select call? A program that happens to work perfectly on Linux
seems to die on NT around 64(?) sockets to the 'too many file descriptors
in call' error.

Any portable ways to bypass it?

-Markus

-- 
"He who fights with monsters should look to it that he himself does
not become a monster. And when you gaze long into an abyss the abyss
also gazes into you."
                    - Friedrich Nietzsche, _Beyond Good and Evil_  



From graham at sloth.math.uga.edu  Fri Jul 23 10:24:16 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 23 Jul 1999 14:24:16 GMT
Subject: Stackless & String-processing
References: <001101bed3f2$5e8874a0$642d2399@tim>
Message-ID: <7n9tug$d6f$2@cronkite.cc.uga.edu>

[Tim]
: > If it isn't clear yet <wink>, efficiency isn't driving Graham
: > here.  He's found a cool new approach, and is having fun with it.
[Graham Matthews]
: > Efficiency drives all my programs *when it's a necessity*.
[Tim]
: Ack, I'm speaking English here.  If efficiency isn't your *primary* concern,
: then efficiency isn't "driving" you.  Dictionary.  Look up.  

No Tim I suggest you look up the dictionary. The dictionary definition of
"driven" does not preclude one being driven by more than one concern (and
indeed driven by more than one concern each of equal importance). For
example I am driven by efficiency, correctness,  and expressiveness in 
roughly equal importance. So efficiency is driving me. Perhaps the motto
here is that he who lives by the dictionary shall die by the dictionary.

At this point I am signing off this thread. It is no longer remotely
relevant to the group. I refer all questions about combinators to
Tim, who, although he has never written a combinator based parser,
appears to know everything about them. Oh and he can look up words in
dictionary too, although perhaps not understand the definition he finds.

graham

-- 
               Black velvet in that little boy's smile
                 Black velvet and a soft southern sky
            A new religion that'll bring you to your knees
                     Black velvet, if you please



From jh at cddk.dk  Wed Jul 21 06:23:27 1999
From: jh at cddk.dk (Jesper Hertel)
Date: Wed, 21 Jul 1999 12:23:27 +0200
Subject: printing with Python/Windows
References: <FP6l3.3374$K3.72023@nuq-read.news.verio.net>
Message-ID: <7n4738$ab6$1@news1.tele.dk>

Hi Sasha,

Maybe wxPython (http://alldunn.com/wxPython/) is something? There is an
example of printing (and even previewing) in the package. And because
wxPython is platform independent, you can even reuse your code on other
platforms.


Jesper Hertel

Sasha Voynow wrote in message ...
>What's the simplest way to print a file (in this case an image)
>from Python on WinNT. I'm not really versed in the win32api
>or MFC. But as a stopgap measure I've got to get this to work on a win32
>platform. I scoured the Pywin / win32 extensions docs, but to no avail.
>
>Thanks,
>
>Sasha Voynow
>
>





From philen at earn.aa.ans.net  Fri Jul  9 00:35:30 1999
From: philen at earn.aa.ans.net (Phillip Lenhardt)
Date: Fri, 09 Jul 1999 04:35:30 GMT
Subject: Example Code: Bresenham Line Algorithm
Message-ID: <7m9tln$6qg$4@rks1.urz.tu-dresden.de>

If there is some python algorithm repository this
should go in, let me know. Otherwise, enjoy!

def Brensenham_line(x,y,x2,y2):
    """Brensenham line algorithm"""
    steep = 0
    coords = []
    dx = abs(x2 - x)
    if (x2 - x) > 0: sx = 1
    else: sx = -1
    dy = abs(y2 - y)
    if (y2 - y) > 0: sy = 1
    else: sy = -1
    if dy > dx:
        steep = 1
        x,y = y,x
        dx,dy = dy,dx
        sx,sy = sy,sx
    d = (2 * dy) - dx
    for i in range(0,dx):
        if steep: coords.append((y,x))
        else: coords.append((x,y))
        while d >= 0:
            y = y + sy
            d = d - (2 * dx)
        x = x + sx
        d = d + (2 * dy)
    coords.append((x2,y2))
    return coords



From jam at newimage.com  Tue Jul 13 14:03:17 1999
From: jam at newimage.com (jam)
Date: Tue, 13 Jul 1999 14:03:17 -0400
Subject: MySQL Module
In-Reply-To: <19990713122901.A2538@pooh.frostnet.net>; from Chris Frost on Tue, Jul 13, 1999 at 05:29:01PM +0000
References: <19990713122901.A2538@pooh.frostnet.net>
Message-ID: <19990713140317.D11617@toast.internal>

On Tue, Jul 13, 1999 at 05:29:01PM +0000, Chris Frost wrote:
> 
> Are than any modules for connecting to a MySQL sever? I didn't see
> anything in the contributed section of python.org. (If there aren't, what
> other databases do people think are better, why?)
> 
> thanks!
> -- 
> Chris Frost  |  <http://www.frostnet.advicom.net/chris/>

hmm.. I really like postgresql, and it comes with a python interface already
made that's really easy to deal with. have a look
<http://www.postgresql.org/>.

regards,
J
-- 
|| visit gfd <http://quark.newimage.com/> 
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 210 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19990713/fe4e924a/attachment.sig>

From roger.burnham at gte.net  Wed Jul 21 15:43:46 1999
From: roger.burnham at gte.net (Roger Burnham)
Date: Wed, 21 Jul 1999 19:43:46 GMT
Subject: printing with Python/Windows
References: <FP6l3.3374$K3.72023@nuq-read.news.verio.net> <37952079.44083903@news.gte.net> <932533928.18948@www.remarq.com>
Message-ID: <379620b8.10675324@news.gte.net>

On Tue, 20 Jul 1999 21:12:03 -0800, Voynow, Sasha <sasha at tunes.com> wrote:

>
>Thanks for the help.
>
>Once I started digging into solving this problem my mind
>was boggled by what
>seemed to be a whole  lot of extra overhead involved in
>what was essentially
>a simple scripting task (i.e. doing some batch image
>processing with PIL and
>sending the results to a printer). But I guess that's life
>under Windows...
>I'll have to take a look at this script and see if I can
>modify it for
>"unattended" execution, as that is what I'm really after.
>

Okay, done.  See my web site:

http://starship.python.net/crew/roger/

for DisplayImage.py.  From the doc:


Image Display and Printing demo.

Demonstrates a number of techniques for displaying and printing images using
the PythonWin MFC framework.  This method of printing is suitable for
"automatic" printing, and so, uses the default printer settings.

As this was ripped out of a larger app, I left some other stuff in as well:
    *) The pixel value under the cursor is displayed in the title bar.
    *) You can draw up to four circles on the image using the mouse.
    *) After four circles are drawn, you can drag them to new locations.
    *) You can delete a circle with the delete or backspace key.
    *) Right click on the image to get an options menu (print only for now).

Usage:
    from the windows command line:
        PythonWin.exe /app path/DisplayImage.py [-p -m mag] [imagePath]
        where:
            path:      The directory containing this script.
            -p:        If present, print the image and exit.
            -m mag:    Only used if -p present, the magnification to print at,
                       defaults to 3.
            imagePath: The image to open and display.
        You may need to supply the PythonWin.exe path as well.
        If you use the -p option, you must supply the imagePath as well.
        If no imagePath is given, an open file dialog is used to get the image.

    from the Python/PythonWin command prompt (or, within a script):
        import DisplayImage
        DisplayImage.displayImage() # Use file dialog to get and display an
                                    # image
        DisplayImage.displayImage(imgPath) # display given image
        DisplayImage.printImage(imgPath, mag) # print image at mag.
        
Requires:  PIL (Python Imaging Library)
           PythonWin (Build 125 or later)



Cheers,
 

Roger Burnham   
Cambridge Research & Instrumentation   
rburnham at cri-inc.com   
http://www.cri-inc.com/   
http://starship.python.net/crew/roger/   
PGP Key: http://www.nai.com/default_pgp.asp   
PGP Fingerprint: 5372 729A 9557 5F36 177F  084A 6C64 BE27 0BC4 CF2D   



From greg.ewing at compaq.com  Wed Jul 14 17:15:11 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 15 Jul 1999 09:15:11 +1200
Subject: Terminating Python...
References: <BFC0587634C3D111AACB00805FA766FA02C1C652@exchny12.ny.smb.com>
Message-ID: <378CFDDF.489CDCEF@compaq.com>

"Velasquez, Carlos" wrote:
> 
> Hello,
> 
> I just downloaded and built Python v1.51 for the Sun Solaris platform.
> One question is, how do I exit python?

Ctrl-D at the interpreter prompt should do the trick.
Or if it's running code, Ctrl-C to get its attention
and then Ctrl-D.

Greg



From pserrand at p98.f112.n480.z2.fidonet.org  Fri Jul  2 10:22:52 1999
From: pserrand at p98.f112.n480.z2.fidonet.org (pserrand)
Date: Fri, 02 Jul 1999 15:22:52 +0100
Subject: test - ignore
Message-ID: <00001660@bossar.com.pl>

From: pserrand <bucodi at worldnet.fr>

pserrand wrote:

> test with python

reply test



From tim_one at email.msn.com  Sat Jul 17 00:34:29 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Jul 1999 04:34:29 GMT
Subject: I envy newbies
In-Reply-To: <378F2304.ACC1AF7E@callware.com>
References: <378F2304.ACC1AF7E@callware.com>
Message-ID: <000001bed00d$a9a70480$d2a22299@tim>

[Ivan]
> ...
> Oh, poo, Timmy.  Everything else under the M$ sun mucks with your
> autoexec.bat file, why shouldn't Python?  "Backups are for p'takh!"
> --Klingon Programmer Aphorism;-)

In 5 years of running 95 and NT, only one program (a virus scanner) has ever
had the audacity to change my autoexec.bat.  This isn't DOS!  Well, OK, 95
is.  But 95 has other mechanisms too, and if you install software that edits
yours it must be old enough to vote <wink>.

Changing that file is dangerous, in part because bat files have secret
line-length limitations, the installer can't possibly know where in the path
to place a new directory (potential shadowing), and an uninstaller would
have to be smart to undo its changes (e.g. blindly restoring the file it
started from would be wrong).

So modern a-go-go installations do the safe thing instead:  wildly overwrite
every DLL in the System directory <wink>.

> Besides, I don't think IDLE will work if you don't have Tcl in your
> path,

Poor Guido recently wrote FixTk.py (in Lib/lib-tk/) to try to guess where
Tcl lives under Windows (the Tcl/Tk installer doesn't touch autoexec.bat
either); it's invoked automatically as soon as IDLE tries to import Tkinter.

> and if you have to add that, why not add Python?

Because "24 Hours" is in the title of your book, and it can take a newbie
twice that long just to recover from a botched edit of this critical shared
system file <0.7 wink>.

> Well, obviously I should go forth and re-format my hard drive and re-
> install W98 and then re-install Python. ...  But I really don't feel
> like a little root canal today, so I'll take your word for it that my
> autoexec.bat file is left virgin and pristine by Python's installer.

The installation script is in (the source distribution's)
PCbuild/python15.wse -- although reading that will increase your
appreciation for root canals <wink>.  If you want to check it out, just edit
your autoexec.bat then reinstall Python over your current installation.

> ...
> I still maintain that the installer *ought* to modify your path; really,
> that should be safe enough,

Afraid not -- and that's why nothing else (written this decade) does it
anymore either.

making-newbies-do-it-by-hand-cleverly-shifts-the-blame-for-the-
    problems-onto-them-ly y'rs  - tim





From hanson at europa.com  Sat Jul 10 01:13:58 1999
From: hanson at europa.com (hanson at europa.com)
Date: Sat, 10 Jul 1999 05:13:58 GMT
Subject: Python Code Repository?
Message-ID: <199907100513.WAA11086@fine.otime.com>

I looked at Troves web site and reviewing the mailing list archives.. I
think that trove is not off the ground yet. 

Using CPAN is not fun. I find it's organization is not intuitive (about intuitive as perl itself). In the python world I find that not having a organized central repository not as bad as it seems. I tend to find new modules by skimming the mailing list and reading www.python.org & freshmeat. Lately, though there is a lot of redundancy on the mailing list where interested parties are inquiring about previously discussed modules. 

One thing that I have found useful using python & the contributed modules
is that they reside where the author has control, so that he/she may provide hints & documentation and divulge the uses & enchancements to his/her module. Also, most of the modules have an interdependency between python and products. 

To the point, I don't think that a CPAN like structure is as useful for
python as it sounds, but I feel that there is something else we can do to
help the python community be more efficient in proliferating modules &
extensions to the language. Where a repository for modules would be
useful is when the author has not provided packages such as RPMs or full
installation & usage instructions.

A perl script that I use (very) frequently called cvsweb.
cvsweb allows you to view your cvs logs & differences. 
We could rewrite and enhance this tool allow for an intuitive query architecture for viewing source & changelogs of python
and it's extensions. This could easily written as a Zope product.

The next idea would be something in python that would make it easy to
manage packages & dependencies. RPM is a example of a good architecture that allows the user to easily install & upgrade his/her tools without losing all of the flexibility of self-compiled packages. If I had a python interface to the redhat rpm's it would do a world of good for trying out new things. 

Notice that there are probably only two  or three installers for windows & theyall know how to deinstall what they install.
If we wrote something up for windows that could recognize these formats,
it would be almost the same as a package manager. (of course you would
need some way of wading through all that microsoft piddle)

I know that linuxconf (redhat) was written mostly in python. (am I wrong?)
what is the status of pythoneers supporting or extending this?
 
On a sidenote one of the best things about python is that most every
module & extension comes with a demo and test facility. This is
very helpful when trying out new things. Thanks :>

,Andrew Hanson


PS. my shift key was in use by my cat.









From mikael at toolwood.com  Thu Jul  1 23:28:45 1999
From: mikael at toolwood.com (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 03:28:45 GMT
Subject: Python GUIs: Summary and Conclusion
References: <99FA4A1A1280BEBD.0A1ABA62CC989479.4C2C6022F695BEAC@lp.airnews.net> <19990629092757.A4998@optiplex.palga.uucp> <46F9FB3A1847A014.2CBD270176138613.20F84A39E6ED95DA@lp.airnews.net> <b_Pe3.558$6M6.167478@news.shore.net>
Message-ID: <395C236424AEEAA7.C2D70DEDE2306228.0C4BC0DFFEC49C2C@lp.airnews.net>

On Thu, 01 Jul 1999 20:13:59 GMT, "Michael P. Reilly"
<arcege at shore.net> wrote:

>Been there, done that.  My pyirc client (alpha release) was designed
>for just about any user-interface, and comes with three: dummy (no
>input or output, just batch-mode)), curses and Tkinter.  You just can't
>switch between them at runtime, ;)  I've done it for other apps too.

Well, the only reason I haven't been there is that I've only worked on
one GUI app _ever_ - a 2D tile-based strategy game [1].  I did plan
support for console apps (curses, etc., using a client-server
architecture), but I didn't much fancy the idea of playing with a
command line interface :)

Anyway, it could be fun to have wxBatch and wxConsole variants of
wxWindows.  wxBatch would then, supposedly, map a wxCheckBox into a
boolean option, a wxTextCtrl into a file name or a string option, and
so on.  (I don't have any good ideas for wxPostScriptDC and
wxLayoutConstraints, though).  wxConsole should be fairly trivial,
except for the pixel-fiddling parts.

Now, there's a task!

[1] Which I expect to change in the future, which is why I'm reviewing
portable GUIs (I don't do non-portable stuff, if there's anything I
can do to avoid it).


-- Mikael




From wang at rjka.com  Wed Jul 28 11:06:42 1999
From: wang at rjka.com (Huayin Wang)
Date: Wed, 28 Jul 1999 10:06:42 -0500
Subject: A Zope build question: where is sha.py
Message-ID: <199907281506.KAA04703@ws89.rjka.com>

when compiling zope2.0 I got the following error message:
....
  File "../zope/Zope-2.0.0b1-src/zpasswd.py", line 90, in ?
    import sys, string, sha, binascii, whrandom, getopt, getpass, os
ImportError: No module named sha
....

I think it is a valid complain because I can't find sha.py anywhere on my system.
Please help.





From robert_j_roberts at rl.gov  Fri Jul 23 14:29:08 1999
From: robert_j_roberts at rl.gov (Roberts, Robert J)
Date: Fri, 23 Jul 1999 11:29:08 -0700
Subject: Non-ODBC MS Sql Connect
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F98@RED-MSG-50>
Message-ID: <3798B472.68DEF409@rl.gov>

Thanks for the quick reply.

I tried it, however, and get:

  File "C:\Web\http\isearch\testNonOdbc01.py", line 44, in Main
    import win32com.client
  File "C:\Program Files\Python\win32com\client\__init__.py", line 8, in ?
    import dynamic, CLSIDToClass, gencache, pythoncom
  File "C:\Program Files\Python\win32com\client\dynamic.py", line 22, in ?
    import pythoncom
ImportError: No module named pythoncom

I have pythoncom documentation in: c:\python\win32com\html

...but no pythoncom.py

I re-downloaded the latest win32all and reinstalled, but still no
pythoncom.py.

Am I being brain dead or am I missing something?

Bill Tutt wrote:

> You can always use ADO from Python on NT.
>
> Quick example:
> import win32com.client
> import win32com.client.makepy
> import win32com.client.genpy
> import win32com.client.gencache
> adolib = win32com.client.gencache.EnsureModule(
>         '{00000200-0000-0010-8000-00AA006D2EA4}', # TypeLib GUID for ADO
>         win32con.LANG_NEUTRAL, # LCID
>         2, # Major version #
>         0, # Minor version #
>         win32com.client.genpy.GeneratorProgress()) # Silent progress
> indicator.
> # ALL loaded typelibrary constants...
> k = win32com.client.constants
> cxnSQLServer = adolib.Connection()
> cxnSQLServer.CursorLocation = k.adUseNone
> cxnSQLServer.Provider = "SQLOLEDB"
> cxnSQLServer.Open("Server=MachineName;Database=master;UID=user;PWD=apple")
>
> rs = adolib.Recordset()
> rs.Open("select * from sysobjects", cxnSQLServer, k.adOpenForwardOnly,
>         k.adLockReadOnly)
> # etc.....
>
> Bill
>
> > -----Original Message-----
> > From: Roberts, Robert J [mailto:robert_j_roberts at rl.gov]
> > Sent: Thursday, July 22, 1999 5:33 PM
> > To: python-list at cwi.nl
> > Subject: Non-ODBC MS Sql Connect
> >
> >
> > How to do I make a non-odbc connection to a MS SQL Server?
> >
> > Am running Python Windows NT/IIS.
> >
> > Thanks,
> >
> >




From jwpolley at collins.rockwell.com  Wed Jul 14 14:32:44 1999
From: jwpolley at collins.rockwell.com (Jonathan Polley)
Date: Wed, 14 Jul 1999 13:32:44 -0500
Subject: Redirect stdout to file
References: <816010E2456BD111A48700805FBBE2EEFDF47B@ex-quebec-u1.baan.com>
Message-ID: <378CD7CC.D6B4B6CC@collins.rockwell.com>

I've made some changes to IDLE (thank you Guido, thank you, thank you, thank
you) that allows me to log printed output to a file (as well as journal user
entered scripts).  The restriction is that you MUST be using IDLE.  I can email
the changes if you are interested.

Jonathan Polley
jwpolley at collins.rockwell.com

Gaetan Corneau wrote:

> Hello,
>
> I'm in Barneveld Holland right now (hey, maybe I'll go visit the famous
> Stichting Mathematisch Centrum ;) and I would like to redirect output from
> the print statement to a text file. I think a saw a message about that some
> time ago, but I can't find it right now.
>
> Anyone could help?
>
> Thanks,
> ______________________________________________________
>    Gaetan Corneau
>    Software Developer (System integration Team)
>    BaaN  Supply Chain Solutions
>    E-mail: Gaetan_Corneau at baan.com
>    Compuserve: Gaetan_Corneau at compuserve.com
>    ICQ Number: 7395494
>    Tel: (418) 266-8252
> ______________________________________________________
> "Profanity is the one language all programmers know best"




From eugene.leitl at lrz.uni-muenchen.de  Tue Jul 27 14:58:23 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Tue, 27 Jul 1999 11:58:23 -0700 (PDT)
Subject: Python Performance
In-Reply-To: <14237.60407.130426.461846@dolphin.mojam.com>
References: <3794C642.AB6A3E5D@freenet.co.uk>
	<7n5f0c$snu$1@ffx2nh5.news.uu.net>
	<al81zdxkktf.fsf@sirppi.helsinki.fi>
	<p51zdvn4c7.fsf@bidra241.bbn.hp.com>
	<14236.32616.980578.536614@dolphin.mojam.com>
	<379D89BC.DA91BA8C@hp.com>
	<14237.60407.130426.461846@dolphin.mojam.com>
Message-ID: <14238.335.458728.140169@lrz.de>

Skip Montanaro writes:

 > True.  I was thinking about the situation I'm in (small developer, a few
 > machines to maintain/upgrade/replace).

What's wrong with recoding the hot spots in C? Python is primarily a
rapid prototyping/scripting/glue language, it was never meant as a
sole implementation language for applications.



From joyce at netcard.com  Mon Jul 26 16:48:00 1999
From: joyce at netcard.com (Joyce Stack)
Date: Mon, 26 Jul 1999 20:48:00 GMT
Subject: filter() + simple program not outputting
Message-ID: <379CC972.750BD42D@netcard.com>

Hello,

New to python but not to programming...can someone tell me why this
simple program is not outputting when i run from the BASH prompt....any
pointers would be helpful...thanking you in advance....

Joyce




#!/usr/bin/env python

def f(x):
        return x%2 != 0 and x%3 != 0
        filter(f, range(2, 25))
        f( )


-------------- next part --------------
A non-text attachment was scrubbed...
Name: joyce.vcf
Type: text/x-vcard
Size: 214 bytes
Desc: Card for Joyce Stack 
URL: <http://mail.python.org/pipermail/python-list/attachments/19990726/6e4e5440/attachment.vcf>

From ajs at pigpond.com  Wed Jul 28 02:19:33 1999
From: ajs at pigpond.com (Andrew Snare)
Date: 28 Jul 1999 16:19:33 +1000
Subject: Python and Red Hat Linux 6
References: <XFMail.990727125533.nathan@islanddata.com> <IrbXbQsB0KGWFZV7RK@holmes.parc.xerox.com>
Message-ID: <m3zp0hthxm.fsf@taita.pigpond.com>

>>>>> "Bill" == Bill Janssen <janssen at parc.xerox.com> writes:

  >> You have to give #! an absolute pathname for it to work.  env is
  >> (somewhat) reliably at /usr/bin and is a program used exclusively
  >> for running other programs in certain environments.  In this
  >> case, it is used because it will search your path for python
  >> rather than requiring it to be explicit.

  Bill> Of course, it only works if python is on the PATH of the
  Bill> program executing the script.  Personally, I don't find the
  Bill> /usr/bin/env hack useful.

It used to be more useful when python tended to be in /usr/local/bin
rather than /usr/bin which is now the case on most Linux boxes. The
env hack also has the disadvantage that it doesn't let you pass
through the -O option to the python interpreter (at least under
Linux). This is the case on various other flavours of unix as well,
which is why perl reparses the first line of the perl script to get
the options instead of relying on what it was given as argv.

This topic seems to come up rather regularly. As always, I strongly
suggest people read "man perlrun" and "man tclsh" to get a good
understanding of the pro's and con's of the various methods. Maybe
this is for the FAQ?

 - Andrew thought-it-was-already-in-there Snare
-- 
#!/usr/bin/env python
print(lambda s:s+"("+`s`+")")\
('#!/usr/bin/env python\012print(lambda s:s+"("+`s`+")")\\\012')
print(lambda x:x%`x`)('print(lambda x:x%%`x`)(%s)')



From cbbrowne at news.brownes.org  Thu Jul  1 23:33:03 1999
From: cbbrowne at news.brownes.org (Christopher B. Browne)
Date: Fri, 02 Jul 1999 03:33:03 GMT
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> <wkzp1g2t2e.fsf@ifi.uio.no>  <14204.912.383867.589108@lrz.de>
Message-ID: <slrn7noclt.429.cbbrowne@knuth.brownes.org>

On Thu,  1 Jul 1999 17:12:06 -0700 (PDT), Eugene Leitl <eugene.leitl at lrz.uni-muenchen.de> posted:
>Lars Marius Garshol writes:
> > 
> > Huh? Forth dates back to the 60s, whereas Scheme is from 1975, and I'm
> > quite unsure of whether Chuck Moore knew Lisp at all.
>
>Chuck is mostly doing (quirky, but fascinating) hardware design 
>nowadays. Check out http://www.ultratechnology.com/ to what he has
>been up to recently.

Anyone else that said the things Chuck Moore does would be considered
a "net-kook."  He's had some odd-ball stuff such as:
- Actually using a "chording" keyboard
- Eliminating source code (OK)
- A 20 bit microprocessor
- A conciously tail-recursive Forth
- Designing chips by directly manipulating arrays, rather than using
abstract languages.
- The claim that code should *not* be written to be reusable, but rather
should be tightened so as to "throw away everything that isn't being used."

If it weren't that he's actually gotten truly useful results out of
all of these things, he'd probably be considered "just another of the
opinionated fools" that clutter up the Internet.

Note that the final item on that list is really crucial; he minimizes
things to the max, and the rather bizarre results are indeed direct
results of that philosophy.
-- 
Those who do not understand Unix are condemned to reinvent it, poorly. 	
-- Henry Spencer          <http://www.hex.net/~cbbrowne/lsf.html>
cbbrowne at hex.net - "What have you contributed to free software today?..."



From kernr at mail.ncifcrf.gov  Wed Jul 21 17:48:04 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Wed, 21 Jul 1999 21:48:04 GMT
Subject: Scientific Plotting?
References: <19990721010203.96101.qmail@hotmail.com> <3795DF85.320EADBE@ssec.wisc.edu> <Pine.LNX.4.10.9907211108100.11968-100000@us2.mayo.edu> <37961CA3.186124E7@ssec.wisc.edu>
Message-ID: <37963fea.2365369@news.erols.com>

On Wed, 21 Jul 1999 14:16:51 -0500, Nick Bower
<nick.bower at ssec.wisc.edu> wrote:

>> I totally agree with you.  I'm encouraged by the efforts made and
>> the progress of PIDDLE and Graphite.  You should look in to them
>> if you want to help out to make python more plotting-friendly.
>
>Sorry to ask a silly question, but where can I find this Graphite?  I've
>been searching the python site with no success.

Graphite:  http://www.strout.net/info/coding/python/graphite
PIDDLE:    http://www.strout.net/info/coding/python/piddle

>cheers, nick
>
>-- 
>
>Nick Bower

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From bwarsaw at cnri.reston.va.us  Fri Jul 23 21:49:34 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Fri, 23 Jul 1999 21:49:34 -0400 (EDT)
Subject: Xemacs Python Mode.
References: <199907231949.PAA07568@python.org>
	<3798CF69.1FC04C23@jps.net>
	<3798DE2D.C008286D@aw.sgi.com>
	<3798E523.7AF3DE05@jps.net>
	<14232.63872.409563.299007@227.chicago-26-27rs.il.dial-access.att.net>
Message-ID: <14233.7086.70469.968277@anthem.cnri.reston.va.us>

>>>>> "SM" == Skip Montanaro <skip at mojam.com> writes:

    SM> In fact, maybe Emacs' tabify command will just do the right
    SM> thing if you have tab-width set properly.  Try C-h f tabify
    SM> RET.

M-x tabify should do it unless you have different indentation levels
(some with equiv 4 spaces, some with equiv 8 spaces).  In that case,
tho' I haven't tried it, I'm guessing Gordon's tabcleaner.py should do 
the trick nicely.

ftp://ftp.python.org/pub/python/contrib/All/tabcleaner.py

-Barry



From fredrik at pythonware.com  Fri Jul 16 15:51:10 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 16 Jul 1999 19:51:10 GMT
Subject: Radio Buttons and Variables
References: <01becfaa$4fc73d80$be54c3d1@Cyberus.cyberus.ca>
Message-ID: <001101becfc4$8e58b370$f29b12c2@secret.pythonware.com>

Gordon Williams <g_will at cyberus.ca> wrote:
> I'm having trouble getting radio buttons and variables working properly. 
> The code works properly the first time it is run from Pythonwin but the
> second time the variable does not update properly.  This means that
> foovar.get()=0 regardless of the radio button pressed.
> 
> Each time the script is run, print foovar indicates a new variable (eg
> PY_VAR0 then PY_VAR1 etc).  Could this be the problem?  

the fact that you're running this under PythonWin is
definitely the cause of the problem; the code looks
fine, and I cannot repeat this problem using my
favourite IDE...

> Any ideas to solve this?

try changing:

     foovar = IntVar()

to

     foovar = IntVar(root)

and see if that helps.

</F>




From Haimo.Zobernig at cern.ch  Tue Jul 13 08:08:48 1999
From: Haimo.Zobernig at cern.ch (Haimo G. Zobernig)
Date: Tue, 13 Jul 1999 14:08:48 +0200
Subject: making a Windows NT service out of a python program
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F20@RED-MSG-50> <378B2AA8.58477120@appliedbiometrics.com>
Message-ID: <378B2C50.B7C2134A@cern.ch>

Christian Tismer wrote:
> 
> Bill Tutt wrote:
> >
> > Take a look at the stuff in win32\demos\service\ of your friendly latest
> > Win32 extension install tree.
> >
> > It shows you how to wrap Python code into an NT service quite easily.
> >
> > The code in win32\lib\win32serivceutil.py appears to do almost everything
> > you would ever want to do with a service. Very cool stuff...
> 
> This is all great stuff if you have understood it already.
> For me, it is still unclear what makes up a service, and
> what I need at the minimum to turn a Python program into
> an NT service.
> 
> I need nothing more than to run Medusa as a service. Did
> anybody do this before? Do I need special calls at all?
> 
> thanks - chris
> 
> --
> Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
> Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
> Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
> 10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
> PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
>      we're tired of banana software - shipped green, ripens at home

I'm in exactly the same situation - no idea what really distinguishes a
"service"
from any old program, or how programs behave differently when run as a service.
Anyway, thanks for Bill's pointer, I will be trying out the pipeTestService.py 
example and see if I can adapt that for my purpose.

Thanks, Haimo
 
 -- 
Haimo G. Zobernig                                  tel   +41 22 7677361
Univ. of Wisconsin                                 fax   +41 22 7678370
c/o CERN Bld. 32 R-A02
1211 Geneve 23, Switzerland                email Haimo.Zobernig at cern.ch



From roffe at morgoth.uio.no  Tue Jul 20 08:44:19 1999
From: roffe at morgoth.uio.no (Rolf Marvin Bøe Lindgren)
Date: 20 Jul 1999 14:44:19 +0200
Subject: python vs. lisp (whereis quote?)
References: <x88r9m3cyje.fsf@rpppc2.hns.com>
Message-ID: <lbzk8rva3to.fsf@morgoth.uio.no>

[ nbecker

| One feature of lisp (emacs lisp actually) that I think is pretty cool
| is the "hooks".  A hook is basically unevaluated code that will be run 
| by eval later.
|
| I suspect python lacks this capability.  As I understand it, the only
| similar capability is "eval".  But I think eval expects a string,
| which it will then parse.  While the capability is similar, having to
| parse a string could be inefficient.
|
| Am I correct that what is missing is the ability to postpone
| evaluation using "quote"?

in Lisp, there is no syntactic difference between code and data, and
that is what makes the hook mechanism possible the way it is implemented
in Emacs.  if Python had a way of representing code as data, Python
could have had the same support for hooks. 

if eval expects a string, and if the string can contain code, then you
have it, haven't you?

-- 
Rolf Lindgren                                        http://www.uio.no/~roffe/
roffe at tag.uio.no



From arcege at shore.net  Fri Jul  9 21:40:18 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Sat, 10 Jul 1999 01:40:18 GMT
Subject: Printing returned object in embedded app
References: <378694FC.D857F68A@srl.caltech.edu>
Message-ID: <7m9tna$6qg$53@rks1.urz.tu-dresden.de>

Conrad Steenberg <conrad at srl.caltech.edu> wrote:
: I'm trying to construct an interactive front-end for Python, using GTK.

: My C program passes strings to PyRun_String, and obtains the returned object.
: Now in the normal interactive Python on a terminal, a command like
: 1+1
: would print out the value, but using my method it does not. Using
: print(1+1) works like a charm. However, when I try to have the returned object
: printed using eg. 
: PyObject_CallMethod(object,"print","") 
: I get an error, "AttributeError: print". Can anybody tell me what I'm doing
: wrong? Or how to get the result printed automatically like in interactive
: Python?

Try:
  PyObject_Print(object, stdout, Py_PRINT_RAW);

The first argument is the object to be printed, the second is the FILE
* to print to and the last is a flag (the only flag for now is
Py_PRINT_RAW).  It only outputs the object, no leading or trailing
whitespace.

Also, you could run "print object" from PyRun_SimpleString.

  -Arcege




From robin at jessikat.demon.co.uk  Thu Jul  8 05:08:08 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 8 Jul 1999 10:08:08 +0100
Subject: Coredump: python 1.5.2
References: <19990708142614.A3595@longford.cs.monash.edu.au>
Message-ID: <YLo3xDA4pGh3Ewue@jessikat.demon.co.uk>

In article <19990708142614.A3595 at longford.cs.monash.edu.au>, Toby J
Sargeant <tjs at longford.cs.monash.edu.au> writes
>The following code causes a coredump under python 1.5.2
>
>class Stuffed:
>   def __getattr__(self,key):
>      print self
>
>print Stuffed()
>
...
this is actually a recursive call of print.
-- 
Robin Becker



From howjoe at yahoo.com  Thu Jul  8 15:21:16 1999
From: howjoe at yahoo.com (How Joe)
Date: Thu, 8 Jul 1999 19:21:16 GMT
Subject: jobs for python programmers?
Message-ID: <19990708192116.14108.rocketmail@web125.yahoomail.com>

python pros,

Is there any websites or places advertise
jobs for python programmers?

Thanks.
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com




From clgonsal at kami.com  Tue Jul 13 17:53:23 1999
From: clgonsal at kami.com (C. Laurence Gonsalves)
Date: Tue, 13 Jul 1999 21:53:23 +0000
Subject: VIM and [Python] block jumping
References: <000001becb18$821aca40$e69e2299@tim>
Message-ID: <378BB553.F30A0C16@kami.com>

Tim Peters wrote:
> 
> [C. Laurence Gonsalves]
> > ...
> > The blockMotion function doesn't account for lines that are
> > continuations of other lines. That means that triple-quoted string
> > literals, lines with more {[('s that }])'s, and lines joined with \ can
> > screw it up. Are there any other situations?
> 
> No, those are the only ways a line can get continued in Python.  Note that
> there are subtleties, like that a backslash preceding a newline doesn't
> *always* mean continuation (it doesn't at the tail end of a comment!).

I was meaning to test some of those weird cases.

> Also single-quoted strings continued via backslash.  Strings are the real
> headache, because the special characters you're looking for lose their magic
> when they're inside a string.

Yup, but I'd thought of a simple way to determine if I'm in a string
though...

> The current CVS version of IDLE has a new module PyParse.py, which does the
> minimum necessary to determine whether a line is part of a continuation, and
> if so why.  It can do this 100x faster than tokenize, even in
> non-pathological programs.  The next version of IDLE uses this to do Emacs
> pymode-like "very smart" indentation, and the next version of Mark Hammond's
> PythonWin also uses this module.
> 
> It's GUI-independent (straight self-contained Python), but as called from
> AutoIndent.py (also shared between IDLE and PythonWin) takes major advantage
> of the host system's colorization, to get a fast answer to the question "is
> this random character in a string?".  The colorizer generally knows this
> already, for every character in the program, and AutoIdent/PyParse use that
> to find a good (close) place to begin reparsing the file.

That's exactly what I was thinking of doing. It's possible to ask Vim
what syntax highlighting a character is getting. That would make it easy
to determine if a \ is in a string, for example.

Is PyParse available any way other than from IDLE CVS?

Thanks.

-- 
  C. Laurence Gonsalves   clgonsal at kami.com  
http://www.cryogen.com/clgonsal/

  "Any sufficiently advanced technology is indistinguishable from
magic." 
                                                           -- Arthur C.
Clarke



From steve at renlabs.com  Wed Jul 14 14:10:06 1999
From: steve at renlabs.com (Steven Work)
Date: 14 Jul 1999 11:10:06 -0700
Subject: Embedding Python in C
References: <7mi084$r80$1@kopp.stud.ntnu.no>
Message-ID: <87iu7n3xxd.fsf@solano.in.renlabs.com>

(Followup narrowed to comp.lang.python)

"Thomas R?berg" <thomar at ifi.ntnu.no> writes:

> I am writing a program in C/C++ and want to embed Python. The one problem I
> encounter is that when I try to import a module using
> Pyrun_SimpleString("import mymodule\n"), this module is not found. I have
> specified the correct paths at the correct places in my Makefile (at least I
> think I have). Anyone had problem with this??

I just fought an embedding battle yesterday with Debian 2.1 --
trouble with the Debian python package (linking against
/usr/lib/python1.5/config/libpython1.5.a) but all smooth if I build
and install Python myself (linking against my
/usr/local/lib/python1.5/config/libpython1.5.a).  I haven't tried to
figure out why the Debian python package gave me fits.
-- 
Steven Work
Renaissance Labs
steve at renlabs.com
360 647-1833



From fdrake at cnri.reston.va.us  Mon Jul 12 12:56:12 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Mon, 12 Jul 1999 16:56:12 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
In-Reply-To: <7md2ip$19q2$1@nntp6.u.washington.edu>
References: <3787D36A.815B7C29@ssec.wisc.edu>
	<3787D8DA.3774AA7D@ssec.wisc.edu>
	<7m8p7t$o7s@dfw-ixnews10.ix.netcom.com>
	<3787FB51.12FE0048@ssec.wisc.edu>
	<7m923i$dem@dfw-ixnews5.ix.netcom.com>
	<14217.63931.75257.566266@weyr.cnri.reston.va.us>
	<7md2ip$19q2$1@nntp6.u.washington.edu>
Message-ID: <14218.7724.879301.831590@weyr.cnri.reston.va.us>

Donn Cave writes:
 > For what it's worth, me too:  current working directory is a process
 > setting.  It's relevant to files and directories, to be sure, but whoever
 > put chdir() and getcwd() in process parameters knew what he was doing.

  I am vindicated!
  Does this mean you'll make a contribution to my "Don't Brake for
Dutchmen" campaign?  ;)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From poinot at onera.fr  Wed Jul 28 10:20:15 1999
From: poinot at onera.fr (Marc POINOT)
Date: Wed, 28 Jul 1999 16:20:15 +0200
Subject: beep
References: <379F0751.71214802@lri.fr> <379F1B1A.2199B575@inrialpes.fr> <379F0EF9.54CDDBEC@lri.fr>
Message-ID: <379F119E.432EA0BA@onera.fr>

Stephane Conversy wrote:

> Vladimir Marangozov wrote:
>
> > Perhaps printing the standard escape sequence '\a' (Alert) would do?
> >
> >
>
> is it portable ? (ie Mac OS)
>

As a matter of fact, <Klung> is much more heavy than <Bip>.

Marcvs [alias Don't know about <beep>]





From guido at CNRI.Reston.VA.US  Mon Jul 12 17:07:58 1999
From: guido at CNRI.Reston.VA.US (Guido van Rossum)
Date: Mon, 12 Jul 1999 17:07:58 -0400
Subject: Which starship?
Message-ID: <199907122107.RAA09523@eric.cnri.reston.va.us>

There are two starship sites: starship.skyport.net (old) and
starship.python.net (new).

I just completed a rough conversion of most pages on the Python
website (www.python.org) to point to the new site whenever possible.

I did the following: for each link to a page on the old starship, if
the corresponding page on the new starship exists, or if the page on
the old starship doesn't exist, I changed the link to point to the
page on the new starship.

I haven't looked at all the links (there were hundreds).  If you have
pages on both starships and you want the *old* starship to be the
"official" link, please mail the details (which page on www.python.org
has the link, and where it points to) to webmaster at python.org.

I couldn't fix some areas of the site, especially the "topics" area.
I hope Andrew Kuchling will convert these if necessary.  I also didn't 
change some pages that are only of historic importance, such as
conference proceedings and the Tim Peters Archives.

--Guido van Rossum (home page: http://www.python.org/~guido/)



From Pierre.Fortin at p98.f112.n480.z2.fidonet.org  Fri Jul  2 16:38:34 1999
From: Pierre.Fortin at p98.f112.n480.z2.fidonet.org (Pierre Fortin)
Date: Fri, 02 Jul 1999 21:38:34 +0100
Subject: Slicing beyond list --  bug or featurette?
Message-ID: <00001692@bossar.com.pl>

From: Pierre Fortin <pfortin at cisco.com>

Just ran into a situation which can be summarized as follows:

list = [ 0,1,2,3,4 ]

list[<0..4>] returns the appropriate item

list[m] where m is >= len(list) expectedly fails

list[2:50] returns [ 2,3,4 ]
list[20:] or list[20:30] returns []
 
Can I count on this, or should I code try/except for the day it's
fixed?  This 'feature' reduces the error situations; but is it
correct/intended?

Curious,
Pierre
--
Linux -- The C(r)ash-Free Operating System



From nick at NOSPAMvideosystem.co.uk  Fri Jul  9 13:17:07 1999
From: nick at NOSPAMvideosystem.co.uk (Nick Trout)
Date: Fri, 9 Jul 1999 18:17:07 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <000015d7@bossar.com.pl>
Message-ID: <37862ccf@news.xtml.co.uk>

wx runs on more platforms than Windows and X. If you really want to get rid
of Tcl/Tk why not implement Tk in wxWindows. This would give people the
flexiblity to use either interface without the bulk of Tcl/Tk and still
retain portability. You would then only have to include wxWindows with
Python and an interface for Tk written in wxWindows.

Speaking as a new user I find both Tcl/Tk and wxWindows useful. It would be
nice if the Tcl bit were removed as it seems bulky and redundant and a Tk
interface would be faster (sorry, just reitering!). Tk is great for quick
knocked up interfaces and wxWindows seems better for writing larger more
"windowsie" projects!

Haven't had a chance to look at FOX yet.

N


Alexander V. Voinov <Alexander.V..Voinov at p98.f112.n480.z2.fidonet.org> wrote
in message news:000015d7 at bossar.com.pl...
>From: "Alexander V. Voinov" <avv at quasar.ipa.nw.ru>
>
>Lyle Johnson wrote:
>
>> Just to add fuel to the fire, we now have a Python interface to FOX, a
>> cross-platform GUI toolkit:
>>
>>     http://www.cfdrc.com/pub/FOX/fox.html
>>
>> The latest version of the Python wrappers can be picked up at:
>>
>>     ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42.tar.gz
>>
>> I don't know how the speed of FXPy would compare with some of the others
>> mentioned (Tk, wxWindows, Qt, GTK, ...) but it would be worthwhile for
>> some independent opinions.
>
>It looks great. It seems "intuitively obvious" (for me, of course :-) both
>in programming and in using (I mean keyboard handling, at least under
>Windows). Yes, the principal question, I ask always when see a new GUI is:
>
>IS IT POSSIBLE TO DO EVERYTHING WITHOUT MOUSE?
>
>Gtk still fails to answer this positively. Qt-2.0 is very close to the
>ideal, as I can see.
>
>The only thing surprised me a bit: I failed to switch to a non-US keyboard
>layout while in text entry field (window.py example). I run Windows NT 4.0,
>SP4. I.e. the toolbar indicator said I'm in the required mode, but the
>letters I entered were just those ABC-xyz... I can proceed without this in
>getting started and evaluation, but it's necessary for any application I
>would deliver to users.
>
>Alexander
>





From gmcm at hypernet.com  Sat Jul 10 20:30:04 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 10 Jul 1999 19:30:04 -0500
Subject: Simple Qu: Changing directory
In-Reply-To: <3787D36A.815B7C29@ssec.wisc.edu>
Message-ID: <1280473245-87635258@hypernet.com>

Nick Bower wrote:

> I feel really dumb asking this question, but how do I change the
> current working directory?  For browsing the file system I mean.
> 
> I've found exec commands, but these would surely be device dependent
> and there doesn't seem to be anything obvious in the os or os.path
> documentation.  i looked through the tutorials too.

os.chdir()

- Gordon



From tim_one at email.msn.com  Sun Jul 11 02:13:11 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 11 Jul 1999 02:13:11 -0400
Subject: Dreams.
In-Reply-To: <slrn7og8m8.17k.bmcd@ripley.sulaco.net>
Message-ID: <001201becb64$74714120$479e2299@tim>

[Ben Caradoc-Davies]
> Hmmm. I'm not sure what your background is, Scott, but you may
> not be aware of some previous work --- the term you may be looking
> for is "continuation". ...

I don't think so.  Continuations capture control-flow state but Scott is
aiming to dream away <wink> changes in data state.  Bindings aren't saved or
restored by continuations:

(define cont #f)
(define (test)
  (let ((i 0))
    (call/cc
     (lambda (k)
       (set! cont k)))
    (set! i (+ i 1))
    (print i) (newline)))

Then:

> (test)
1
> (cont)
2
> (cont)
3
> (cont)
4
>

Indeed, if bindings *were* restored, invoking a saved continuation couldn't
do anything other than repeat what it did the first time it ran <wink>.

Fully agreed that SICP is a great book, and functional languages (of which
Scheme is not one) are worth learning too, but Scott has a harder problem in
mind than continuations alone can solve.

you-just-don't-see-many-functional-databases-these-days<wink>-ly y'rs  - tim





From MHammond at skippinet.com.au  Thu Jul  1 18:56:49 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 2 Jul 1999 08:56:49 +1000
Subject: Newbie has trouble with in/output
References: <7lglb4$ksu$1@news.vossnet.de>
Message-ID: <7lgrl7$qev$1@m2.c2.telstra-mm.net.au>

Florian W. wrote in message <7lglb4$ksu$1 at news.vossnet.de>...
>1.) Being in a pipe:
>
>I won't to filter the output of the compiler (e.g. only errors and warning
>are reported, no logo etc.)
>using the pipe-mechanismen :
>
>cl.exe blabla | python filter_compiler_output.py
>
>How do I catch the input and how do print the output ?

You simply read() to get the input, and print to write output!!

>2.) Catch the output of an application:

Check out the win32pipe module in the win32all package.  Documentation
(sort-of :) is in the help file, or searching dejanews for that module name
will return plenty of hits...

Mark.






From skip at mojam.com  Wed Jul 28 12:43:41 1999
From: skip at mojam.com (Skip Montanaro)
Date: Wed, 28 Jul 1999 11:43:41 -0500 (CDT)
Subject: A Zope build question: where is sha.py
In-Reply-To: <199907281506.KAA04703@ws89.rjka.com>
References: <199907281506.KAA04703@ws89.rjka.com>
Message-ID: <14239.12896.351951.595436@dolphin.mojam.com>

    Huayin> when compiling zope2.0 I got the following error message:
    Huayin> ....
    Huayin>   File "../zope/Zope-2.0.0b1-src/zpasswd.py", line 90, in ?
    Huayin>     import sys, string, sha, binascii, whrandom, getopt, getpass, os
    Huayin> ImportError: No module named sha
    Huayin> ....

"sha" is part of the Wayne's World package. ;-)

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From gmcm at hypernet.com  Sun Jul 25 10:29:06 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 25 Jul 1999 09:29:06 -0500
Subject: Win32 Installer - new release
Message-ID: <1279213353-36226633@hypernet.com>

 A new release (01a) of the Win32 Installer package is available on
starship:

 http://starship.python.net/crew/gmcm/install.html

This release incorporates bug fixes and some enhancements from (and 
inspired by) Robin Dunn.

 - A generated install script is now "safe"
 - File atimes, mtimes copied if possible
 - Install packages can be built from any directory
 - Ability to gather all dependencies in a directory, so you can use 
   WinZip or WIse or whatever to distribute the results.

What is it?
 Briefly, it is a sort of compilerless Freeze - a way of distributing 
Python apps with all the required Python support included.

 Python-style license.

Contact: gmcm at hypernet.com

<P><A HREF="http://starship.python.net/crew/gmcm/install.html">
Win32 Installer release 01a</A> - A compiler-less way of 
distributing Python apps on Windows.  (24-Jul-99)



From Christopher.Petrilli at p98.f112.n480.z2.fidonet.org  Thu Jul  1 23:01:37 1999
From: Christopher.Petrilli at p98.f112.n480.z2.fidonet.org (Christopher Petrilli)
Date: Fri, 02 Jul 1999 04:01:37 +0100
Subject: Socket programming.
Message-ID: <0000167e@bossar.com.pl>

From: Christopher Petrilli <petrilli at trump.amber.org>

Kjetil =D8degaard <kjetilod at stud.ntnu.no> wrote:
> My current implementation is threaded (using fork()); one thread for
> user interaction and one for server output.  It uses blocking
> sockets.  It works ok, but I've read some docs here and there (amongst
> them the Socket Programming HOWTO on www.python.org) and I get the
> feeling I should be using select() and non-blocking sockets.  Also,
> I'm not sure I want to use fork() -- it seems very low-level and
> depends on a Unix-like system.

First, fork() isn't really threading... if you want to do threading, then
you should use the threading module, if it' ssupported on your platform.

Second, for that kind of thign, take a look at asyncore and asynchat,
the first of which I've written documentation for that will be released
in the next itteration of the Python library reference.  asynchat
should be able to do what you want.

Chris
--=20
| Christopher Petrilli                      ``Television is bubble-gum for
| petrilli at amber.org                          the mind.''-Frank Lloyd Wrig=
ht



From ivanlan at callware.com  Tue Jul 27 15:31:34 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 27 Jul 1999 19:31:34 GMT
Subject: Serial port IO question
References: <379DF55E.EAFCBBF7@hti.com>
Message-ID: <379E0916.F54A0DDB@callware.com>

Hi All--

David Stokes wrote:
> 

[snip]

>     def get_ups_data_p(self):
>         upsdata = open( upsdev, 'r+');      # open port to ups <<==Shouldn't this be "rb+"
>         upsdata.write("\015")                   # send ^M to ask for
> status

Take care,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From see-sig at wavedivision.com  Thu Jul 22 17:23:39 1999
From: see-sig at wavedivision.com (Rick Robino)
Date: 22 Jul 1999 21:23:39 GMT
Subject: Python and the Web
References: <3793dd72.0@hoh.olympus.net> <AFCEC557B580B99F.7F0096BF9FDA5C47.0D15AF43E5EFE822@lp.airnews.net>
Message-ID: <7n824r$q04$1@news.gstis.net>

Cameron Laird <claird at starbase.neosoft.com> wrote:

> In article <3793dd72.0 at hoh.olympus.net>,
> Michael Novello <nospam at this.address> wrote:
>>Can anyone tell me where to find information on using python to grab form
>>data and create dynamic web pages?
> 			.
> 			.
> 			.
> <URL:http://starbase.neosoft.com/~claird/comp.lang.python/web_python.html>
> operates at a higher level, but each of its references details a specific
> approach to your question.

I didn't see this listed on Cameron's page or the topics page, but I would
recommend HTMLgen:

http://starship.python.net/crew/friedrich/HTMLgen/html/main.html

IMHO, this package is absolutely fabulous and if not directly usable
out of the box, the classes, functions and excellent documentation in
here should make it a snap for you to derive your own classes, as well
as providing a entire framework to support your customizations. Check
out the other conributed web-related modules too.

Sorry if this link was mentioned before, and a big, BIG, thanks to
Robin Friedrich.

--Rick 
                                ^^^^^^^^^^^^^^
Rick Robino                                  mailto:rrobino at wavedivision.com



From Richard.Jones at fulcrum.com.au  Thu Jul 29 19:41:27 1999
From: Richard.Jones at fulcrum.com.au (Richard Jones)
Date: Thu, 29 Jul 1999 23:41:27 GMT
Subject: How to set text color using pyKDE 0.8? 
In-Reply-To: Message from "Peter Torstenson" of 1999-Jul-29 7:42:39,
            <7npb63$2fdn@enews3.newsguy.com> 
References: <7npb63$2fdn@enews3.newsguy.com>
Message-ID: <199907292341.JAA13112@envy.fulcrum.com.au>

["Peter Torstenson"]
> I'm using pyKDE 0.8 binding for python.
> I made a QLabel on the main widget. Now I'd like to change the color of the
> text inside this QLabel.
> I managed to change the backgroundcolor and the font. But I can't find any
> info on how to change the color of the text, or the foregroundcolor.

   I've just done some poking around... Qt 2.0 looks like it has really good 
support for color control (presumably as part of the introduction of styles), 
but of course we need pyKDE 0.9 for that...

   I even tried supplying a QPainter through drawContents, but of course a 
setBrush()'ed QColor gets nuked when the drawContents does a begin()...

   A friend says he did it once, but he can't remember offhand how...


       Richard





From lazrnerd at ufl.edu  Wed Jul 21 09:03:52 1999
From: lazrnerd at ufl.edu (Craig Schardt)
Date: Wed, 21 Jul 1999 13:03:52 GMT
Subject: Scientific Plotting?
References: <19990721010203.96101.qmail@hotmail.com> <37952384.14766416@news.erols.com>
Message-ID: <3795c243.764244896@news.ufl.edu>

On Wed, 21 Jul 1999 01:43:48 GMT, kernr at mail.ncifcrf.gov (Robert Kern)
wrote:

>On Tue, 20 Jul 1999 18:02:01 PDT, "Michael Huster"
>
>[snip]
>
>>Gnuplot: Runs gnuplot through a pipe. All of the data (upto many MB) also 
>>has to go through the pipe. Unsat. I want to plot efficiently.
>
>The Gnuplot module as updated by Michael Haggerty
>(http://monsoon.harvard.edu/~mhagger/download/) puts the data into
>temporary files.  Pipes are used for commands only.  If you haven't
>used it recently, you might want to try it again.
>

I have modified Michael Haggerty's Gnuplot.py to work under Windows.
The Windows version of Gnuplot doesn't support pipes but there is a
helper application, pgnuplot, which works around this problem and
provides a pipe interface. The combination works well on NT and 98
with PythonWin.

If anyone is interested in these programs, please let me know.

-craig.
lazrnerd at ufl.edu



From Terry.Reedy at p98.f112.n480.z2.fidonet.org  Thu Jul  1 21:04:37 1999
From: Terry.Reedy at p98.f112.n480.z2.fidonet.org (Terry Reedy)
Date: Fri, 02 Jul 1999 02:04:37 +0100
Subject: Tuples & lists
Message-ID: <00001649@bossar.com.pl>

From: tjreedy at udel.edu (Terry Reedy)

In article <377bbde1 at news.xtml.co.uk>, nick at NOSPAMvideosystem.co.uk says...

>Why both having tuples and lists? As far as I can see there're the same, but
>tuples are read only..? Have I missed something in the tutorial?

There is probably something in the the FAQ.  BAsically, tuples are stored 
more compactly and can be used as dictionary keys (because they ARE 
readonly). 

TJR



From shillcoc at mry.rbd.com  Fri Jul  9 12:43:59 1999
From: shillcoc at mry.rbd.com (Scott Shillcock)
Date: Fri, 09 Jul 1999 09:43:59 -0700
Subject: console show in Python?
Message-ID: <7m9tmi$6qg$30@rks1.urz.tu-dresden.de>

Hello,

I am moving to Python from Tcl/Tk.
Is there something similar to the Tcl/Tk 'console show' command in
Python.

I am using Tkinter for my gui and once I call mainloop() I no longer
have access to the Python command line. I want to be able to interact
with my gui from the command line to test new code.

>From Tcl/Tk I could turn on and off the interactive command line with
'console show' and 'console hide'. That is all I want to do with Python.
Can anyone help me with this or point me in the right direction.

Thanks,
Scott Shillcock



From ralf.doering at prakinf.tu-ilmenau.de  Thu Jul 29 01:56:44 1999
From: ralf.doering at prakinf.tu-ilmenau.de (Ralf Doering)
Date: 29 Jul 1999 07:56:44 +0200
Subject: Newbie Problem
References: <7nndss$q1n$1@alexander.INS.CWRU.Edu>
Message-ID: <xzvhb49exv.fsf@prakinf.tu-ilmenau.de>

rbl at hal.cwru.edu writes:

> 
> There is some Python code I've downloaded that I'm trying to run.
> If I just drop the .py file on the Python icon, I get some errors,
> so I am trying to run it interactively for debugging.
> 
> Here is what happened:
> 
> Python 1.5.1 (#37, Apr 27 1998, 13:36:04)  [CW PPC w/GUSI w/MSL]
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import sys
> >>> import random
> >>> import string
> >>> 
> >>> def run(filename=''):
> ... if filename=='':
>   File "<stdin>", line 2
>     if filename=='':
>      ^
> SyntaxError: invalid syntax
> >>> 
> 
> The original code segment is:

[...]

Just indent the "if" statement.

Ralf
-- 
No sig -- no fun



From olipt at mayo.edu  Fri Jul  2 14:14:42 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Fri, 2 Jul 1999 13:14:42 -0500
Subject: GUI's etc: PyGtk on Windows
Message-ID: <Pine.LNX.4.10.9907021307420.21537-100000@us2.mayo.edu>

Previously I mentioned that Gtk works on any system with an X server,
which is true.

I also want to point out that Gtk is a cross-platform solution
because there is a port of the GDK (the underlying windowing calls that
Gtk uses) for Win32.  You can find a link at http://www.gtk.org

>From what I've read it looks like Gtk was intentionally written on top of
GDK so that a port of GDK would make Gtk available.  The person who wrote
the GDK port to Win32 did it to get the GIMP to run on Windows.
There are screenshots of the result available on the site.

So if someone develops mainly for X and just wants to make sure that it is
not impossible to run on Windows, you can use PyGTK.  In fact a great
program using GTK can be a strong motivator to get the ports of GDK going.

Just some thoughts,

Travis




From heiland at ncsa.uiuc.edu  Thu Jul  1 14:06:27 1999
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Thu, 01 Jul 1999 13:06:27 -0500
Subject: embedding; return array
Message-ID: <377BAE23.4634E0A3@ncsa.uiuc.edu>

I'd welcome any pointers into being able to return an (unsigned char)
array from an embedded module.

thanks,
Randy




From fredrik at pythonware.com  Sun Jul 18 11:15:14 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 18 Jul 1999 15:15:14 GMT
Subject: I envy newbies
References: <000101bed082$747b0de0$27a22299@tim>
Message-ID: <005901bed130$57228cb0$f29b12c2@secret.pythonware.com>

Tim Peters <tim_one at email.msn.com> wrote:
> > Usually, these things modify the path by simply adding a line like this:
> >
> > PATH=L:\BC5\BIN;%PATH%
> 
> That's cute:  they rely on a bug in the PATH cmd <wink>?  The "=" should be
> a blank, or "set " should preface this, although as given it works by
> accident.

it's an old quirk in command.com (and friends).  it has
always treated "=" as a kind whitespace.  try:

    echo=>q
    copy=q==qq
    type====qq
    more<====q
    echo===(etc)

probably something they inherited from CP/M...

python=spam.py doesn't really work, though...

</F>

    "Using all of Bill's money (as $1 bills), you could cover
    the king-sized mattress with bills 232,347,457.26 bills
    deep. That means Bill would have a jump of 997,199
    inches or 15.73 miles to get from the bed to the floor
    each morning, not counting the thickness of the bed
    itself." -- from the Washington Post




From mgushee at havenrock.com  Wed Jul 21 16:40:14 1999
From: mgushee at havenrock.com (Matt Gushee)
Date: 22 Jul 1999 05:40:14 +0900
Subject: Need feedback on Tkinter project.
Message-ID: <m3zp0pspn5.fsf@havenrock.com>

[ I posted this last night, but it hasn't appeared in the
newsgroup. I'm wondering if my newsreader is just hiding my own
posts. Anyway, I apologize if this appears twice. ]

Greetings, Pythoneers--

I would like your comments on a project of mine.

What I've done so far is a game -- a Tkinter-based solitaire game
called Pegboard which more-or-less works and can be enjoyed as is, if
you're into that sort of thing.

But what I'm really after is to use the game as the basis of a Tkinter 
tutorial. I don't mean for it to be all-inclusive or to replace
Frederik Lundh's excellent work, but it does cover some points that
Lundh doesn't cover in any depth (or at least, hadn't gotten around to 
last I knew), particularly the Canvas and Menu widgets.

Your comments on the code itself are most welcome, but I'm especially
interested in your comments on my comments -- and docstrings -- and
places that should have comments/docstrings but don't. That will help
me understand what to write when I begin the actual tutorial.

Pegboard is available (in .tar.gz and .zip packages) from:

	http://www.havenrock.com/softlab/pegboard/index.html

By the way, you need to have the tkFont module.

I look forward to hearing from you!

Matt Gushee
Portland, Maine, USA
mgushee at havenrock.com



From lawter at cse.buffalo.edu  Wed Jul 21 01:00:15 1999
From: lawter at cse.buffalo.edu (John Lawter)
Date: Wed, 21 Jul 1999 05:00:15 GMT
Subject: Python Mode for Alpha ?
Message-ID: <lawter-2107990058210001@client-151-197-119-26.bellatlantic.net>

Is there a more recent Python mode for Alpha than the 6.5 era one ?  I
have searched around the python.org site, but their copy is missing.  Any
help would be appreciated, as I don't look forward to having to write my
own.



From Gordon.McMillan at p98.f112.n480.z2.fidonet.org  Fri Jul  2 06:22:44 1999
From: Gordon.McMillan at p98.f112.n480.z2.fidonet.org (Gordon McMillan)
Date: Fri, 02 Jul 1999 11:22:44 +0100
Subject: execfile(), import and namespaces
Message-ID: <0000168e@bossar.com.pl>

From: "Gordon McMillan" <gmcm at hypernet.com>

Nikolas Kauer wrote:

> We use a modification of "Listing One" of Andrew Kuchling's "A CGI
> Framework in Python" article in Web Techniques (February 1998) for
> our dynamic Web site. [...]  The object files would be loaded as
> modules with 'import'.
> 
> Problem: In the following code snipet environment variables are
> passed to the executed Python file via the corresponding global
> namespace argument of the execfile() function; whereas the 'import'
> statement does not allow for such arguments.

But __import__ does. Check the library docs under built-in-functions.

- Gordon




From Greg.Ewing at p98.f112.n480.z2.fidonet.org  Fri Jul  2 17:34:48 1999
From: Greg.Ewing at p98.f112.n480.z2.fidonet.org (Greg Ewing)
Date: Fri, 02 Jul 1999 22:34:48 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <0000163f@bossar.com.pl>

From: Greg Ewing <greg.ewing at compaq.com>

Kevin Dahlhausen wrote:

> Lyle Johnson <jlj at cfdrc.com> wrote:

> >Just to add fuel to the fire, we now have a Python interface to FOX, a
> >cross-platform GUI toolkit:

> >    http://www.cfdrc.com/pub/FOX/fox.html

This link doesn't work for me. Has
it been moved somewhere else?

Greg



From mithy at my-deja.com  Tue Jul  6 17:20:23 1999
From: mithy at my-deja.com (Preston Landers)
Date: Tue, 06 Jul 1999 21:20:23 GMT
Subject: ++Python - Emacs functions to expand shorthand
Message-ID: <7ltrua$pkr$1@nnrp1.deja.com>

 Hello Everyone,

I'm pleased to release a small set of Emacs functions I wrote to
assist lazy Pythoneers like myself.

It hooks into the standard python-mode and gives you two new commands;

py-shorthand-line-expand   -- bound to C-z -- works only on the
current line.

py-shorthand-buffer-expand -- bound to C-c C-z -- works on the whole
buffer.

These keys expand shorthand expressions.  A couple of examples:

foo += bar  --->  foo = foo + bar
++foo.bar   --->  foo.bar = foo.bar + 1

I know what you're thinking, and yes, I do have entirely too much free
time on my hands.  I also think that maybe Python 2 should incorporate
this syntactic sugar, but until such time, I give you these Emacs
macros to ease your carpal tunnel syndrome. [1]

I've tested them with FSF Emacs 20.3.2 and Xemacs 20.4.  Nothing has
exploded yet.

I do realize that C-z is normally bound to suspend-or-iconify, but I
personally never use that and C-z just seemed convenient.  *Feel free*
to change the key bindings at the end of the lisp file that follows.

The "homepage" for this absurdity will soon be somewhere on
http://askpreston.com/projects/ though there's nothing there at the
moment.

Comments, criticisms, job offers, flames, and live nude girls to:

---Preston Landers <preston at askpreston.com>

[1] Naturally, Emacs itself is known to be a major factor in this
tragic affliction.  Press Control-Meta-Hyper-Shift-H to learn more.

py-shorthand.el follows.  Installation instructions contained therein.

---cut-here-------------------------------------------

;; py-shorthand.el  ---   version 1.0 -- July 5th, 1999
;;
;; Some functions to expand certain shorthand expressions into legal
;; Python for the terminally lazy.
;;   For instance, 'foo += bar' is expanded to 'foo = foo + bar'
;;
;; Copyleft (C) 1999 Preston Landers <preston at askpreston.com>
;; Distributed under the GNU General Public License
;;
;; By default, expanding the current line is bound to C-z (which is
;; normally reserved for iconify / suspend, which I never use.)  Feel
;; free to change it (it's at the bottom of this file.)  Expanding
;; shorthand in the entire buffer is bound to C-c C-z which isn't
;; otherwise bound to anything in my Xemacs 20.4.
;;
;; To install, just make sure this file is in your lisp load path
;; somewhere, and put the following in your .emacs file:
;;
;; (load "py-shorthand")
;;
;; If you don't know what a lisp load path is, make a directory in your
;; home directory called lisp and copy this file to that directory, then
;; put this in your .emacs file instead:
;;
;; (setq load-path (append load-path (list "~/lisp")))
;; (load "py-shorthand")
;;
;; Known bugs: 'foo .= bar' is expanded to 'foo = foo . bar' but should
;; be 'foo = foo + bar'.  Suggestions?
;;
;; An exact list of the types of shorthand expressions recognized
;; follows (with identifying regular expressions.)
;;
;; foo.bar +=
;; \\(\\s-*\\)\\([A-Za-z_0-9\\.\\-]+\\)\\(\\s-*\\)\\([-\\+\\.\\*\\\]\\)=
;; ...becomes...
;; foo.bar = foo.bar +
;; \\1\\2\\3=\\3\\2 \\4

;; ++foo.bar (or --foo.bar)
;; \\(\\s-*\\)\\(\\+\\|-\\)\\2\\([A-Za-z_\\-0-9\\.]+\\)
;; ...becomes...
;; foo.bar = foo.bar + 1
;; \\1\\3 = \\3 \\2 1

;; foo.bar++ (or foo.bar--)
;; \\(\\s-*\\)\\([A-Za-z_\\-0-9\\.]+\\)\\(\\+\\|-\\)\\3
;; ...becomes...
;; foo.bar = foo.bar + 1
;; \\1\\2 = \\2 \\3 1

(defvar py-shorthand-statement-regexp
  "\\(\\s-*\\)\\([A-Za-z_0-9\\.-]+\\)\\(\\s-*\\)\\([-\\+\\.\\*\\\]\\)="
  "Regexp for recognizing foo += bar (or -= *= \=).")

(defvar py-shorthand-statement-replace
  "\\1\\2\\3=\\3\\2 \\4"
  "Replacement string for foo += bar; becomes foo = foo + bar.")

(defvar py-shorthand-precrement-regexp
  "\\(\\s-*\\)\\(\\+\\|-\\)\\2\\([A-Za-z_\\-0-9\\.]+\\)"
  "Regexp for recognizing --foo or ++foo.")

(defvar py-shorthand-precrement-replace
  "\\1\\3 = \\3 \\2 1"
  "Replacement string for ++foo; becomes foo = foo + 1.")

(defvar py-shorthand-postcrement-regexp
  "\\(\\s-*\\)\\([A-Za-z_\\-0-9\\.]+\\)\\(\\+\\|-\\)\\3"
  "Regexp for recognizing foo++ or foo--.")

(defvar py-shorthand-postcrement-replace
  "\\1\\2 = \\2 \\3 1"
  "Replacement string for foo++; becomes foo = foo + 1.")

(defun py-shorthand-current-end-point ()
  "Internal function to find current end of line."
  (save-excursion
    (end-of-line)
    (point)
    ))

(defun py-shorthand-line-expand ()
  "Expands shorthand notation such as += on the current line into valid
Python."
  (interactive)
  (save-excursion
    (let ((end-pos (save-excursion (end-of-line) (point))))
      (beginning-of-line)
      (while (re-search-forward py-shorthand-statement-regexp
(py-shorthand-current-end-point) t 1)
        (replace-match py-shorthand-statement-replace)
        (beginning-of-line))
      (beginning-of-line)
      (while (re-search-forward py-shorthand-precrement-regexp
(py-shorthand-current-end-point) t 1)
        (replace-match py-shorthand-precrement-replace)
        (beginning-of-line))
      (beginning-of-line)
      (while (re-search-forward py-shorthand-postcrement-regexp
(py-shorthand-current-end-point) t 1)
        (replace-match py-shorthand-postcrement-replace)
        (beginning-of-line)))))

(defun py-shorthand-buffer-expand ()
  "Expands shorthand notation such as += in the entire buffer into valid
Python."
  (interactive)
  (save-excursion
    (goto-char (point-min))
    (while (re-search-forward py-shorthand-statement-regexp nil t)
      (replace-match py-shorthand-statement-replace))
    (goto-char (point-min))
    (while (re-search-forward py-shorthand-precrement-regexp nil t)
      (replace-match py-shorthand-precrement-replace))
    (goto-char (point-min))
    (while (re-search-forward py-shorthand-postcrement-regexp nil t)
      (replace-match py-shorthand-postcrement-replace))))

(add-hook 'python-mode-hook
	  (lambda ()
	     (define-key py-mode-map "\C-z" 'py-shorthand-line-expand)
	     (define-key py-mode-map "\C-c\C-z" 'py-shorthand-buffer-expand)
	     ))


--
||  Preston Landers <preston at askpreston.com>  ||
||           http://askpreston.com            ||


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From adv at langdale.com.au  Sat Jul 24 04:02:58 1999
From: adv at langdale.com.au (Arnold de Vos)
Date: Sat, 24 Jul 1999 18:02:58 +1000
Subject: HTTP 1.1 anyone?
Message-ID: <37997331.0@pink.one.net.au>

Is anyone working on an HTTP 1.1 implementation in Python?   I am
considering creating variants of BaseHTTPServer and httplib to handle the
1.1 protocol.  But I would not like to re-invent the wheel.

Thanks,
Arnold

---
Arnold deVos
Langdale Consultants
adv at langdale.com





From dborgelt at xxxxxxcom  Wed Jul  7 18:05:34 1999
From: dborgelt at xxxxxxcom (Dave Borgelt)
Date: Wed, 7 Jul 1999 17:05:34 -0500
Subject: Trouble with HTMLGen, HTMLtest
Message-ID: <e3UDcTMy#GA.342@nih2naae.prod2.compuserve.com>

I am having a problem with HTMLGen that is driving me nuts.  It has to do
with running the test code in HTMLtest.  (HTMLtest.test()).

I expect the solution is simple but I can't see it.  The path to the HTMLGen
files is in my PythonPath (hence the import works) but as you can see the
test fails.

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

>>> import HTMLtest
>>> HTMLtest.test()
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "c:\programming\python\HTMLGen\HTMLtest.py", line 26, in test
    'overview.html')      # manual top
  File "c:\programming\python\HTMLGen\HTMLtest.py", line 35, in overview
    doc = SeriesDocument('HTMLgen.rc')
  File "c:\programming\python\HTMLGen\HTMLgen.py", line 256, in __init__
    if resource: execfile(mpath(resource), self.__dict__)
IOError: [Errno 2] No such file or directory
>>>

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


Any help will be appreciated.

Dave Borgelt

dborgelt at XXXcsiXXX.com







From Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org  Fri Jul  2 11:27:52 1999
From: Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 16:27:52 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001674@bossar.com.pl>

From: mikael at pobox.com (Mikael Lyngvig)

On 2 Jul 1999 11:56:10 GMT, lvirden at cas.org wrote:

>When I run the Tk test suite  on my SPARC, Tk cranks thru creating and
>deleting windows as fast as the eye can track.

Well, that's good.  But I think the key word here is "as fast as the
eye can track".  That's the core of the problem.  A good GUI library
should be creating and deleting windows many times faster than any
human eye can track.

That's definitely not the case with Tk.  So far I think I've counted
20 or 21 different, intermediate windows, when I resize my main window
after having maximized it.  

>However, there are reports that the Windows version of Tk, because of its
>use of the native window widgets, is noticably slower than the X
>implementation.  Is that the platform to which you refer?

Yes, that's the platform to which I refer.  But, no, I don't believe
one minute that the native window widgets are the cause of the
performance problems.

If Tk was using the native widgets, it should be impossible for it, no
matter how not-good the design is, to be slow. Even a batch file (the
pathetic script language that's part of Windows by default) should be
able to draw any given native widget fast enough, since its just a
question of parsing a few, simple parameters and tossing them to some
native code.


-- Mikael



From aahz at netcom.com  Mon Jul 19 19:24:02 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 19 Jul 1999 23:24:02 GMT
Subject: SMTP problem
References: <hrMk3.1158$Gc6.192154624@newsb.telia.net> <3793A296.7A0FAD18@callware.com>
Message-ID: <7n0c2i$op3@dfw-ixnews11.ix.netcom.com>

In article <3793A296.7A0FAD18 at callware.com>,
Ivan Van Laningham  <ivanlan at callware.com> wrote:
>
>2)  Since you're on Windows, and windows doesn't have any native SMTP
>processing abilities, you'll need to replace 'localhost' in the example
>program with the name of your mail server.  

Well, zie *could* be running Python scripts on a machine running
Exchange configured to speak SMTP -- but that would be Very Stupid.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From phil at ricochet.net  Fri Jul  9 02:53:21 1999
From: phil at ricochet.net (phil)
Date: Fri, 09 Jul 1999 06:53:21 GMT
Subject: Two Pythons talking to each other?
References: <199907080019.TAA22244@buffalo.fnal.gov> <199907080529.HAA19502@axil.hvision.nl> <3784486a.7993113@news.ricochet.net> <7m2lol$bnm$1@nntp6.u.washington.edu>
Message-ID: <7m9tls$6qg$9@rks1.urz.tu-dresden.de>

On 8 Jul 1999 17:05:25 GMT, donn at u.washington.edu (Donn Cave) wrote:

>phil at ricochet.net (phil) writes:
>| On Thu, 8 Jul 1999 05:30:40 GMT, "Hans Nowak" <ivnowa at hvision.nl>
>| wrote:
>...
>May be worth noting that application protocols can be and are also
>designed on top of UDP, where among other things they may add some
>kind of sequence monitor - datagram delivery isn't guaranteed, a
>datagram may be discarded if its system buffer has filled up.

Yes, if you are stated or you care about order  then  you probably
need  to sequence. Not only can datagrams not turn up, but they can
turn up out of order. Rare on a LAN, sure.

phil.







From Thomas.Bryan at p98.f112.n480.z2.fidonet.org  Sat Jul  3 00:56:20 1999
From: Thomas.Bryan at p98.f112.n480.z2.fidonet.org (Thomas Bryan)
Date: Sat, 03 Jul 1999 05:56:20 +0100
Subject: Dumb beginner question; what is IDLE?
Message-ID: <000016a9@bossar.com.pl>

From: Thomas Bryan <tbryan at arlut.utexas.edu>

Andy L Krietemeyer wrote:
> 
> I've seen several references to IDLE.  Is this an IDE for python? Or
> what is it?  Thanks,  Andy.

>From http://www.python.org/1.5/NEWS-152.txt
"""
I've added IDLE: my Integrated DeveLopment Environment for Python.
Requires Tcl/Tk (and Tkinter).  Works on Windows and Unix (and should
work on Macintosh, but I haven't been able to test it there; it does
depend on new features in 1.5.2....This is very much a work in
progress.  I'd like to hear how people like it compared to PTUI (or
any other IDE they are familiar with).
"""

-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From Voynow  Wed Jul 21 01:12:03 1999
From: Voynow (Voynow)
Date: Tue, 20 Jul 1999 21:12:03 -0800
Subject: printing with Python/Windows
References: <FP6l3.3374$K3.72023@nuq-read.news.verio.net> <37952079.44083903@news.gte.net>
Message-ID: <932533928.18948@www.remarq.com>

Thanks for the help.

Once I started digging into solving this problem my mind
was boggled by what
seemed to be a whole  lot of extra overhead involved in
what was essentially
a simple scripting task (i.e. doing some batch image
processing with PIL and
sending the results to a printer). But I guess that's life
under Windows...
I'll have to take a look at this script and see if I can
modify it for
"unattended" execution, as that is what I'm really after.

Sasha




* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From samschul at pacbell.net  Fri Jul 23 20:07:19 1999
From: samschul at pacbell.net (Sam Schulenburg)
Date: Sat, 24 Jul 1999 00:07:19 -0000
Subject: Python based ASPI SCSI test environmrent
Message-ID: <NBdm3.949$Fl.41523@typhoon-sf.snfc21.pbi.net>

 The following link will provide a public domain ASPI SCSI test environment
 that works under Windows 95/98, NT.
 This environment functions under the python scripting language, which is
 also public domain.

To use the SCSI python environment, one must have python allready installed
 on their system with both ASPI and tcl/tk; which is included with the
Python
 disribution ; working. Although tcl/tk and Python's Tkinter interface to
 tcl/tk is not necessary to use the ASPI functions. They will work under
 Python in console mode with the following command from the python prompt::

 >>> from aspi import *
 http://starship.python.net/crew/schulenburg/

 Samuel W. Schulenburg
 samschul at pacbell.net







From kauer at pheno.physics.wisc.edu  Wed Jul 21 17:55:07 1999
From: kauer at pheno.physics.wisc.edu (Nikolas Kauer)
Date: Wed, 21 Jul 1999 16:55:07 -0500 (CDT)
Subject: ImportError: Unresolved symbol
Message-ID: <199907212155.QAA19956@egret.physics.wisc.edu>

Fredrik and Marc, thanks for your replies!  
I added 'SetEnv LD_LIBRARY_PATH /local/shlib' to Apache's httpd.conf.  
Problem occurs in exactly the same way ...
I then looked at all environment variables and figured that only 
LD_LIBRARY_PATH and _RLD_ARGS=-ignore_unresolved could potentially 
be involved.  So next try with 
PassEnv LD_LIBRARY_PATH _RLD_ARGS
in httpd.conf

Success!  Error disappears ...
Problem: Python scripts that use some functions of the DateTime module don't work.  
They do work on Linux, so Fredrik seems to be right that timegm is missing on 
Digital Unix.

Attempt to build mxDateTime.so without timegm by commenting out the appropriate line 
in the Setup file in subdir mxDateTime.  (The DU C compiler fails like before, but 
gcc does the job.)  Put new mxDateTime.so in place and try again ...

Success!  Everything is working fine, now.

Again, thanks for your help.



From tseaver at starbase.neosoft.com  Fri Jul 30 17:43:01 1999
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 30 Jul 1999 21:43:01 GMT
Subject: Need help with TCP/IP client access from Windows
References: <slrn7q1ej3.18d.esr@snark.thyrsus.com> <1278839420-22085362@hypernet.com> <19990730005211.T7149@thyrsus.com> <7nsf2g$ifp@dfw-ixnews11.ix.netcom.com>
Message-ID: <AD3E221BA5854AD1.8DC43033B1ED11E9.1D246CE91AB40FE7@lp.airnews.net>

In article <7nsf2g$ifp at dfw-ixnews11.ix.netcom.com>,
Aahz Maruch <aahz at netcom.com> wrote:
>In article <19990730005211.T7149 at thyrsus.com>,
>Eric S. Raymond <esr at thyrsus.com> wrote:
>>Gordon McMillan <gmcm at hypernet.com>:
>>>
>>> Probably better just to have a default file, and let the user 
>>> override that if need be. Which may well never happen.
>>
>>OK, but where to put it?  I don't know what the conventions are.
>
>I'd stick it in the directory your script is in.  That's pretty standard
>for Windoze.  In fact, I'd make it "<base script name>.ini".
>--
>                      --- Aahz (@netcom.com)
>
>Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
>Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)

Well, that's the "well-behaved" version -- the _really_ standard thing is to
put it in %WINDIR%, to preserve the system's fossil record  of every app you
ever ran.

In all seriousness:  DON'T write to the registry;  DO confine your application
to a single, monlithic directory structure; DON'T worry about multi-user unless
you KNOW the client has a rigidly-enforced set of conventions defining what
"home directory" means.

Best,

Tres.
-- 
---------------------------------------------------------------
Tres Seaver           tseaver at palladion.com       713-523-6582
Palladion Software    http://www.palladion.com



From daniel.dittmar at sap-ag.de  Tue Jul 27 04:48:18 1999
From: daniel.dittmar at sap-ag.de (Daniel Dittmar)
Date: Tue, 27 Jul 1999 08:48:18 GMT
Subject: Python and SlickEdit
References: <379CA3F9.A746217E@iRenaissance.com>
Message-ID: <7njrob$oq6$1@mailusr.wdf.sap-ag.de>

Rowland Smith <rowland at iRenaissance.com> wrote:

>Does anyone know if there are any Python plug-ins for Visual SlickEdit
>and where they are?

There's a Python module in the contributed area
(ftp://ftp.slickedit.com/pub/usermacs/python.zip). 

It does code coloring, but no procedure outlining (at least not
working). 

Daniel Dittmar





From phil at ricochet.net  Fri Jul  9 02:34:25 1999
From: phil at ricochet.net (phil)
Date: Fri, 09 Jul 1999 06:34:25 GMT
Subject: Two Pythons talking to each other?
References: <199907072203.AAA09601@axil.hvision.nl> <378465E9.7762BE9A@fnmail.com> <14212.51977.601241.772478@buffalo.fnal.gov> <3784CD0C.24437EFC@fnmail.com> <roy-0807991239500001@qwerky.med.nyu.edu>
Message-ID: <7m9tlo$6qg$6@rks1.urz.tu-dresden.de>

On Thu, 08 Jul 1999 12:39:50 -0400, roy at popmail.med.nyu.edu (Roy
Smith) wrote:

>Per Kistler <kistler at fnmail.com> wrote:
>> But does NT and Mac have loopback IP addresses?
>
>The Mac certainly does.
>
NTshould be fine as well. I usu. test stuff  using 127.0.0.1/localhost
on NT.

phil.




From tim_one at email.msn.com  Thu Jul 22 03:07:48 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 22 Jul 1999 03:07:48 -0400
Subject: IDLE's PyShell
In-Reply-To: <3795F12D.91B820BB@inrialpes.fr>
Message-ID: <001401bed410$e7963a20$642d2399@tim>

[Vladimir Marangozov]
> ...
> Actually, I really don't understand very well how the PyShell works.

Then you understand it a lot better than I do <0.7 wink>.

> Here's another little exercise that causes some weird behavior. If you can
> reproduce it, then it's most probably a bug unrelated to my
> personal install.
>
> 1) start a brand new idle.py (the PyShell welcomes you to Pythonland)
> 2) press RETURN a couple of times (just to warm the linecache ;-)
> 3) See the contents of the cache dict by typing "PyShell.linecache.cache"
> 4) Now provoke a name error by typing, say, "bonk"
>
> we get the usual Traceback ... NameError: bonk.
> Everything seems to be perfect, until...
>
> 5) See again the contents of the cache with PyShell.linecache.cache
>
> I get a huge output, containing the entire PyShell.py source...

Yes, that happens for me too.

> With all due respect to PyShell.py, its source shouldn't be
> there, isn't it?

I don't know!  linecache.py predates IDLE's existence, and every time I've
bumped into it it seemed like an optimization that created more problems
than it solved.  So if I were to pursue this, I'd look harder at tossing it
than fixing it.

But then I'm not sure why it's there either.  I expect that in your case,
it's actually traceback.py that's stuffing all of PyShell.py into the
linecache, as part of formatting the NameError exception, snaking thru
InteractiveInterpreter.showtraceback.  That last is hiding from you that
PyShell.py is in the true NameError traceback (you see an edited version of
it by the time it gets to the shell window).

now-you-understand-it-better-than-i-do<wink>-ly y'rs  - tim





From kistler at fnmail.com  Wed Jul  7 10:31:32 1999
From: kistler at fnmail.com (Per Kistler)
Date: Wed, 07 Jul 1999 16:31:32 +0200
Subject: PIL anim gif
Message-ID: <378364C4.E18D7C1E@fnmail.com>

Hi All

How to make an animated gif from several Image
objects in PIL? According to the docu, it seems,
that it can only anim gifs which are such already.

Regards, Per.
-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net
------------------------------------------------------------



From tim_one at email.msn.com  Sun Jul 18 20:57:16 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 18 Jul 1999 20:57:16 -0400
Subject: How fast can we multiply?
In-Reply-To: <37922175.FEEBE203@appliedbiometrics.com>
Message-ID: <000201bed181$a5bf0be0$4ba02299@tim>

[Christian Tismer, hiding from continuations <wink>]
> ...
> Version 0.2 is now 4 times faster at 100000 bits.

Factor of 5.7 on my machine.

> ...
> fastlong.py
> ...
> It is a well known fact that by some factorization, integer
> multiplication's complexity can be brought from O(n**2) down to
> O(n**1.5).

The exponent is actually log2(3) ~= 1.58.

> The method is sometimes called Karatsuba algorithm.
> ...
> I think to remember that the GMP library uses this algorithm,
> but I didn't look it up and tried my own straightforward way.

Right, GMP does do this.  There's a much faster method (O(n)) based on
Fourier transforms, but it's much harder to code and I've only seen it used
in David Bailey's MPFUN package (a highly optimized Fortran pkg for
arbitrary-but-fixed precision float arithmetic + the std elementary
functions).

> The reason why I don't aim to modify the C implementation any
> longer is: It will not be reasonably faster in C. The real
> working horse is still the basic internal long multiplication
> (sometimes called "highschool algorithm") which is fast for small
> long integers. The only effect of a C port would be to move the
> break_even point to something smaller.

Tee hee.  I wonder how much of Python we could speed up by throwing out C
code <wink>?  Integer exponentiation can also be done faster in Python, and
much faster with a faster multiply to build on.  Seem to recall that
repr(dict) is/was quadratic-time in len(dict), while easy to make
linear-time if coded in Python.  C is so clumsy you just settle for the
simplest thing to *spell*.

> ...
> A function to obtain the bit length of a long would
> be most helpful.

The lack of this has gotten unbearable, eh?

> Some kind of split function as well.

That one is much weaker; unlike bit length, the obvious ways to split and
join already have the right O() behavior.

> ...
> (1) Aho/Hopcroft/Ullmann:
>     "The Design and Analysis of Computer Algorithms"

Knuth Vol 2 is a better reference for fast integer mult, although it goes
off the deep end on theoretical refinements of no practical value ("what if
the number of bits is bigger than the number of electrons in the
universe?").

> ...
>     def __mul__(self, other):
>         """Karatsuba alogrithm
>         (a+bX)(c+dX)=(X^2+X)(bd)+(X)(a-b)(d-c)+(X+1)(ac)
>         """
>         val1 = self.val
>         val2 = fastlong(other).val
>         if not val1 and val2:
>             return 0

The test only succeeds when val1 == 0 and val2 != 0 -- not what you
intended.  OTOH, it's debatable whether programs multiply by 0 often enough
to pay back the cost of testing for it!

>         bitcount = max(nbits(val1), nbits(val2))

min is a better choice here; see below.

>         if bitcount < self.break_even:
>             return val1 * val2
>         shift = bitcount / 2
>         (hi1, lo1) = split_num(val1, shift)
>         (hi2, lo2) = split_num(val2, shift)

Need a fancier strategy when nbits(val1) and nbits(val2) are far apart.
This strategy takes strong advantage of the special-case-for-0 test above
<wink>, but can easily run 100s of times slower than the builtin mult; e.g.,
change your test case to lambda x:x*5.

This is all the more reason to write in Python, though!  Much clumsier to
handle imbalance in C.  In the meantime, using "min" instead of "max" to
compute bitcount greatly speeds the worst cases, and doesn't hurt the best
ones.

> ...
> def nbits(x):
>     """not clean, rather fast, inaccurate"""
>     s=marshal.dumps(x)
>     return abs(struct.unpack("i", s[1:5])[0])*15 -7

I've been using the following since 1.5.2 came out.  Cleaner and accurate.
The one above is substantially faster, but it's now a small-factor thing
rather than an O() difference:

def nbits(n, hex=hex, long=long, len=len,
             d2bits={'0':0,
                     '1':1,
                     '2':2, '3':2,
                     '4':3, '5':3, '6':3, '7':3,
                     '8':4, '9':4, 'A':4, 'B':4,
                     'C':4, 'D':4, 'E':4, 'F':4}):
    """nbits(n): return number of bits in n.

    n must be >= 0; 0->0, 1->1, 2->2, 3->2, 4->3, 5->3, ...

    In 1.5.1 and before this takes time quadratic in the number
    of bits.  In 1.5.2, linear.  "Should be" constant time, but
    that requires getting at the internal representation.
    """

    if n < 0:
        raise ValueError("nbits arg must be >= 0")
    # hex representation is ashex[2:-1] (leading 0x and trailing L).
    ashex = hex(long(n))
    num_hex_digits = len(ashex) - 3
    return (num_hex_digits - 1)*4 + d2bits[ashex[2]]

> ...
> def timing(func, args, n=1, **keywords) :
> 	import time
> 	time=time.time

Not time.clock?

> ...
> # the end ----------------------------------------------------

looks-more-like-the-beginning-to-me<wink>-ly y'rs  - tim





From roy at popmail.med.nyu.edu  Thu Jul  8 13:40:41 1999
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Thu, 08 Jul 1999 13:40:41 -0400
Subject: Two Pythons talking to each other?
References: <14212.51977.601241.772478@buffalo.fnal.gov> <Pine.SGI.3.95.990708120053.9140Q-100000@hppc.fnal.gov>
Message-ID: <roy-0807991340410001@qwerky.med.nyu.edu>

petravick at fnal.gov wrote:
> AF_UNIX is likely to me more effcient if that matters.
> 
> AF_INET will run about the whole protocol stack, AF_UNIX
> is usually implemented to be more like a pipe...

Almost certainly such a small difference it couldn't possibly matter.

I just tried an experiment.  I copied a 53 Mbyte file from one directory
to another on my Digital Unix alpha.  I did it twice, once using "cp", the
other time using "ftp" connected to localhost.  The results:

cp:  0.015 sec user CPU time, 4.088 sec system CPU time
ftp: 0.084 sec user CPU time, 3.300 sec system CPU time

Of course, I'm only showing the client side CPU time.  It's reasonable to
expect that the server side CPU time was similar, so at first blush, I'd
guess the ftp really took a total of about 3.6 second, or about 1.5x the
CPU time of the "cp" command.

But, in both cases, real time was 10x the CPU time, indicating that I/O
speed totally swamped everything else.

I don't know what the python job in question was supposed to be doing, but
it's almost inconvievable that anything you do in python to the data won't
totally swamp the CPU time spent doing the TCP/IP protocol processing in
the kernel.

-- 
Roy Smith <roy at popmail.med.nyu.edu>
New York University School of Medicine




From emile at fenx.com  Wed Jul 21 10:40:59 1999
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 21 Jul 1999 07:40:59 -0700
Subject: Merging two lists as a dictionary
References: <19990720211946.A12202@pooh.frostnet.net> <19990721153539.S6170@xs4all.nl> <19990721084614.C15502@pooh.frostnet.net>
Message-ID: <T%kl3.3903$VL2.416831@news.direcpc.com>

Can anyone tell me why all of Chris's posts are showing up as
attachments at my end?  (Outlook Express 5, win95b)

--

Emile van Sebille
emile at fenx.com
-------------------


Chris Frost <chris at frostnet.advicom.net> wrote in message
news:19990721084614.C15502 at pooh.frostnet.net...





From skip at mojam.com  Thu Jul 29 10:07:06 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 29 Jul 1999 09:07:06 -0500 (CDT)
Subject: Python Performance
In-Reply-To: <al8btcvudgi.fsf@sirppi.helsinki.fi>
References: <3794C642.AB6A3E5D@freenet.co.uk>
	<7n5f0c$snu$1@ffx2nh5.news.uu.net>
	<al81zdxkktf.fsf@sirppi.helsinki.fi>
	<379C3704.A6F8984D@lemburg.com>
	<379f06e9@news.xtml.co.uk>
	<al8btcvudgi.fsf@sirppi.helsinki.fi>
Message-ID: <14240.24411.661891.318094@dolphin.mojam.com>

    Markus> There is PyStone (comes along with distribution) and PyBench
    Markus> (which I haven't tried). I myself use a performance testing
    Markus> framework and ton of small examples to judge whether things
    Markus> speed up or not.

PyStone, while stable, is not a very complete test of Python's capabilities.
PyBench is useful, though it does a lot of repetitious operations and would
be easy prey for a little variable lifetime analysis.  Markus, if you have a
test suite you use and would be willing to release to others, I would
encourage you to do so.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758




From gmcm at hypernet.com  Sat Jul 24 09:25:10 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 24 Jul 1999 08:25:10 -0500
Subject: Python documentation available as ZIP files
In-Reply-To: <b3c1a02649.tratt@tratt.freeuk.co.uk>
Message-ID: <1279303589-30799085@hypernet.com>

Laurence Tratt writes:

...
> Admittedly, as a non-Windows user, WinZip has always worked when
> I've tried it on .tar.gz files, but maybe I'm not using it correctly
> <wink>.

Most likely, you're not downloading properly <wink>. You're supposed 
to use a browser, and let it choose a sensible name like 
RockyAndBullWinkle_tar.gz.

> If you want to preach the benefits of .tar.gz I'd reccomend
> preaching about .tar.bz2... And that way you can try educating lots
> of Unix users as well as the Windows gang <wink>.

Tee hee. Use it on my Linux box. Someone pointed me to a Windows 
port. The readme says "can't run in a 16 bit subsystem", but the port 
*is* an MSDOS executable (not a 32 bit console executable). It did 
manage to decompress, but mangled file names beyond recognition.

- Gordon



From gvwilson at nevex.com  Tue Jul 20 11:27:15 1999
From: gvwilson at nevex.com (gvwilson at nevex.com)
Date: Tue, 20 Jul 1999 11:27:15 -0400 (EDT)
Subject: (un)directed graph type in Python?
Message-ID: <Pine.LNX.4.10.9907201126270.22309-100000@akbar.nevex.com>

Hi.  Does someone have a Python module for representing large directed and
undirected graphs in Python that is memory-safe?  My first efforst are all
pretty simple, and all leak memory because of circular references...

Thanks,
Greg




From arcege at shore.net  Wed Jul  7 20:50:08 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 08 Jul 1999 00:50:08 GMT
Subject: Extension Writing doc update
Message-ID: <4BSg3.1175$6M6.373488@news.shore.net>

I've updated the document I created a couple of weeks ago describing
how to write your own Python type objects in C.

I've added sections describing:
* How to write numeric callback methods (__add__, __invert__, etc.)
* How to write sequence callback methods (__len__, __add__, __getitem__)
* How to write mapping callback methods (__len__, __getitem__)
* How to add doc strings to your Python/C types
* A (for now) terse explaination on how to make your own coerce
function.

As before, the verbage isn't fleshed out.  I wanted to get the content
out there, more than the long winded descriptions.

The USL is http://starship.python.net/~arcege/extwriting/pyext.html. My
email address is arcege at shore.net.

  -Arcege




From arcege at shore.net  Thu Jul  8 13:50:48 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 08 Jul 1999 17:50:48 GMT
Subject: Any *disadvantage* to using xrange() instead of range()?
References: <004d01bec961$77024240$a0ee6dd1@BobAlexDesk>
Message-ID: <Yx5h3.1288$6M6.394545@news.shore.net>

Bob Alexander <bobalex at ix.netcom.com> wrote:
: ...even for iteration over a small sequence of integers?

: My intuition would be that xrange() is the preferred way to iterate over a
: numeric sequence, but the documentation seems to go out of its way to stress
: that the advantage of xrange() over range() is minimal -- almost seems to be
: discouraging us programmers from using it unless we have a "memory-starved"
: machine.

: Is there any reason why xrange() should not be the "idiomatic" way to
: iterate over a sequence of integers?

: -- Bob

The range() function has been shown to be faster for small numbers (<100?).
But when the number gets larger, and especially within a nested loop,
the range() function can be far less efficient.

I've never really delved into the the code to find out _exactly_ why it
is faster for smaller numbers, but it makes sense.  For-loops are
optimized for sequences, not sequencable objects (like a xrange
object).  Once the sequence size gets too large, the creation time
become a factor, and slows the loop down on average (not each iteration,
but the entire execution time of the loop).

  -Arcege




From tim_one at email.msn.com  Thu Jul 15 03:00:22 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 15 Jul 1999 03:00:22 -0400
Subject: Stackless & String-processing
In-Reply-To: <7mj5f4$5pt$1@brick.cswv.com>
Message-ID: <000601bece8f$b549a420$51a22299@tim>

[Neel Krishnaswami]
> ...
> I've been looking at Icon, and it occurs to me that if coroutines and
> generators were available at the Python level, it might yield a way of
> doing string processing that is more "Pythonic" than regexps.

Yes, and you can find much more about this in the very early days of the
String SIG archives.

> Regexps are nice because when you have a pattern that they can
> directly represent, then you can simply specify a pattern and then you
> don't have to worry about the tedious bookkeeping of looping over the
> string and keeping track of the state variable.
>
> However, when you need to match a pattern that a regexp can't match,
> then suddenly you need to break the string into tokens with regexps
> and then loop over the pieces and keep track of a bunch of state
> variables that don't necessarily correspond to the pieces you are
> actually interested in.
>
> This is unpleasant, because a) clumsy bookkeeping is bad, and b)
> there's two different sorts of syntax needed to do basically
> similar tasks.

The latter is one of the arguments Griswold gave in the paper that
introduced Icon, contrasting Icon's uniform approach to SNOBOL4's distinct
pattern and procedural sublanguages.

I don't think he'd make the same argument today!  Icon is an idiosyncratic
delight, but most string pattern matching was in fact easier (to write,
modify, or read) in SNOBOL4.  Once you start using Icon for complex pattern
matching, you'll soon discover that it's very hard to go back to old code
and disentangle "the pattern" from "the processing" -- *because* everything
looks the same, and it's all intermixed.  The distinct sublanguages in
SNOBOL4 enforced a clean separation; and that was more often an aid than a
burden.

Have noted before that writing string matching code in Icon feels a lot like
writing in an assembly language for SNOBOL4.  Of course the latter didn't
have Icon's power in other areas (e.g. it's at least possible to program
structural pattern-matchers in Icon, using generators to match e.g. tree
patterns; SNOBOL4's patterns started & ended with strings).

> If we could compose generators just like functions, then the bookkeeping
> can be abstracted away and the same approach will work for arbitrarily
> complicated parsing tasks.

The real advantage of regexps is speed, & that's probably while they'll
always be much more popular.  SNOBOL4 and Icon didn't define "bal" builtins
because you couldn't code that pattern yourself <wink>.  bal is beyond a
regexp's abilities, but it's still a simple kind of pattern, and just runs
too darned slow if implemented via the recursive definition as run with the
general backtracking machinery.

> So I think it would be nice if these lovely toys were available at
> the Python level. Is this a possibility?

I've been bugging Guido about generators since '91, but for the billion and
one uses other than string matching.  Anything is possible -- except that
I'll ever stop bugging him <wink>.

> (I'd love to be able to define coroutines in Python like so:
>
> def evens(z):
>     for elt in z:
>         if z % 2 == 0:
>             suspend(z)

How do you intend to resume it?

> It would probably require some rethinking of Python's iteration
> protocol though.)

That's not a hangup; Guido already knows how to do it cleanly.  Like any Big
New Feature there are real questions about costs vs benefits, and so far
this stuff has been widely viewed as "too esoteric".  I think it's natural
as breathing, to the point that a *non*-resumable function strikes me as
never inhaling <wink>.

For now, there's a working implementation of a generator protocol (somewhat
more flexible than Icon's) in the source distribution, under
Demo/threads/Generator.py.  I also posted a general (thread-based) coroutine
implementation a bit over 5 years ago.  Building these on Christian's
stackless Python instead should run much faster.

BTW, generators are much easier to implement than coroutines -- the former
don't require threads or stacklessness or continuations under the covers,
just some relatively straightforward changes to Python's current
implementation (Steven Majewski noted this 5 years ago).  This is why
generators work in all ports of Icon, but coroutines are an optional feature
that's supported only if a platform guru has taken the time to write the
platform-dependent context-switching assembly code Icon coexps require.

degeneratoredly y'rs  - tim





From roger.burnham at gte.net  Tue Jul 20 21:27:06 1999
From: roger.burnham at gte.net (Roger Burnham)
Date: Wed, 21 Jul 1999 01:27:06 GMT
Subject: printing with Python/Windows
References: <FP6l3.3374$K3.72023@nuq-read.news.verio.net>
Message-ID: <37952079.44083903@news.gte.net>

On Tue, 20 Jul 1999 15:34:34 -0700, "Sasha Voynow" <svoynow at yahoo.com> wrote:

>What's the simplest way to print a file (in this case an image)
>from Python on WinNT. I'm not really versed in the win32api
>or MFC. But as a stopgap measure I've got to get this to work on a win32
>platform. I scoured the Pywin / win32 extensions docs, but to no avail.
>
>Thanks,
>
>Sasha Voynow
>
>

Being MFC, its not exactly simple, but I've posted a script that will allow you to
select an image with a file dialog, display it, and then print it at a specified
magnification (uses the print item in the standard PythonWin file menu).

This was taken from a larger app, but shows how to print textual info after the
image and how to handle mutliple pages.  No checking for the image spilling off
the right edge or bottom of the page.

It requires PIL and at least PythonWin build 125 (I believe).  The script must be
run from PythonWin to work...

You can get the script at

http://starship.python.net/crew/roger/

Cheers,

Roger Burnham   
Cambridge Research & Instrumentation   
rburnham at cri-inc.com   
http://www.cri-inc.com/   
http://starship.python.net/crew/roger/   
PGP Key: http://www.nai.com/default_pgp.asp   
PGP Fingerprint: 5372 729A 9557 5F36 177F  084A 6C64 BE27 0BC4 CF2D   



From kas at maps.magnetic-ink.dk  Fri Jul 16 10:17:34 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Fri, 16 Jul 1999 16:17:34 +0200
Subject: Python 1.5.2 + Linux 2.0.36 == Bad Combo?
References: <slrn7ou0ru.3sp.kas@titan.magnetic-ink.dk> <378F31D6.99BF0197@arlut.utexas.edu>
Message-ID: <slrn7oufnu.5ng.kas@titan.magnetic-ink.dk>

Thomas Bryan <tbryan at arlut.utexas.edu> wrote:

> > I'm running Python 1.5.2 on a Linux 2.0.36 kernel (RedHat 5.1 distro
> > partly upgraded to v5.2) and have recently experienced malfunctioning
> > and/or non-working python applications -- specifically:
> 
> Just to make sure....
> 
> Are you sure you're invoking the right Python?  Red Hat comes with 
> Python, and if your created a separate installation in /usr/local or
> something, then you need to make sure that all of your scripts are 
> invoking the 1.5.2 interpreter and not the one that came with 
> Red Hat 5.x.

I used the --prefix=/usr opetion when I configured Python and the previous
version is not present anymore anyway.  Also, the "make check" passed when
I installed the latest version and I don't experience other peculiarities
with Python than the three mentioned -- which may, and may not, be related.

> If you want to blow away Red Hat's Python, you can get the 
> Python RPMs that Oliver Andrich put together.  They're at
> http://andrich.net/python/

Considering that Preston is able to reproduce the Kernel Oops using the
same kernel and python version as I do, I belive it's something deeper
than the python installation.  But thanks anyway, I'll take a look at the
URL. :-)


  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From parkw at better.net  Mon Jul 19 18:53:56 1999
From: parkw at better.net (William Park)
Date: Mon, 19 Jul 1999 18:53:56 -0400
Subject: Scientific Plotting?
In-Reply-To: <379392E7.EB0F29E7@ssec.wisc.edu>; from Nick Bower on Mon, Jul 19, 1999 at 04:04:39PM -0500
References: <379392E7.EB0F29E7@ssec.wisc.edu>
Message-ID: <19990719185356.C314@better.net>

On Mon, Jul 19, 1999 at 04:04:39PM -0500, Nick Bower wrote:
> People at my current work use MatLab or IDL for chomping data through
> data calculations and plotting the results in an environment that lets
> you get GUI's up and going very quickly (among other things).  It's been
> a goal of mine that one day I could show them that this could be done
> with as little effort (with greater room for expansion) using Python.  
> Plus I'd like to put a strong case forward to our physics department for
> our up-and-coming new scientific programmers to learn something with a
> broader scope like Python than the now used IDL and recently abandoned
> Fortran-77.
> 
> But...What do people use for plotting?  After looking at the Python
> Topic Guides, I've taken a quick look at the BLT and DISLIN pages, but
> both of these amazingly lack any screen shots to demonstrate their
> capabilities!!!

Hi Nick,

I use PiCTeX for plotting.  It's more to do with my habit, not at all
related to actual need or performance :).

> 
> And lastly, can array operations (using NumPy) be done without for loops
> as in IDL/MatLab?  eg Multiplying 2 arrays is easy in these two packages
> because you don't need array subscripting and iteration.

Well, YES.  That's what 'class object' is all about.  You can define
"matrix" class, and then do something like
    A = matrix(...)
    B = matrix(...)
    C = A * B

William

> 
> Thanks muchly, Nick.
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From greg.ewing at compaq.com  Fri Jul 16 00:34:46 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 16 Jul 1999 16:34:46 +1200
Subject: Getters and Setters
References: <000701bece99$6dbeede0$51a22299@tim>
Message-ID: <378EB666.BBEC521D@compaq.com>

Tim Peters wrote:
> 
> Python goes one better:  you can write an explicit method if needed, but it
> doesn't encourage bad practice by automatically supplying methods that
> expose every data attribute, crying "use me! abuse me!" <wink>.

There's something to be said for designing a
language so that the same syntax is used for
accessing an attribute whether it's directly
exposed or got at through a method.

I once designed a class system for Scheme
in which all instance variable access from
outside the object was through methods. For
instance variables declared as public, access
methods were automatically created. (There was
also a read-only category which got getters
only.)

The beauty of this was that I could start off
defining something as a directly accessible 
instance variable, and then later change my mind,
and not have to change any of the code which
used the object.

The moral of this is that it's a good idea to
always use accessor methods from outside the
object, and thus having a way of generating
them automatically would be useful.

To avoid inadvertently exposing everything,
they should only be generated for attributes
somehow marked as public or read-only.

I think that instead of being completely
automatic it should be something you explicitly
request when defining the class, e.g.

  class Foo:
    blarg = 0
    parrot = ""

  define_public(Foo, 'blarg')
  define_read_only(Foo, 'parrot')

Implementations of these two functions left
as an exercise...

Greg



From 801d at my-deja.com  Wed Jul  7 16:33:40 1999
From: 801d at my-deja.com (Scott McClellan)
Date: Wed, 07 Jul 1999 20:33:40 GMT
Subject: calldll & PyArg_ParseTuple strings
References: <000016b8@bossar.com.pl>
Message-ID: <7m0dj3$no1$1@nnrp1.deja.com>

Thanks, Gordon, this works for me now. I still have a few more questions
regarding the PyArg_ParseTuple input format strings used in
calldll.call_foreign_function:

1- If I want to pass a long int by value from Python to the C/C++ DLL,
do I still use 'l' in the format string?  All of the examples I've seen
deal with parameters of type long*, rather than type long.

2- How do I handle DLL functions which expect parameters of type
unsigned char*, i.e. addresses of buffers into which the function copies
a fixed amount of binary data (may include '\000' byte values). I've had
some luck passing the name of a Python string and using 's' or 's#' in
the format string, but some of my results appear truncated, and the DLL
functions do not return a length associated with the data returned in
the buffer (required for 's#'?).  Also, since Python strings are
supposed to be immutable, having the DLL stuff them with data doesn't
seem to be a good idea.

Any help would be appreciated.

Regards,
Scott

In article <000016b8 at bossar.com.pl>,
  "Gordon McMillan" <Gordon.McMillan at p98.f112.n480.z2.fidonet.org>
wrote:
> From: "Gordon McMillan" <gmcm at hypernet.com>
>
> Scott McClellan writes:
> >
> > I'm attempting to use Sam Rushing's calldll module to call a C
> > function in a dll, with the following prototype:
> >
> > long __declspec(dllexport)func(long *x, long *y, long *z)
> >
> > I expect the function to give certain values back in x, y, and z,
> > but so far have been unable to make that happen. Let's say that I
> > have three long integers in Python that I initialize as follows:
> >
> > x=y=z=0L
>
> A Python int is a C long. You won't find a Python long in C.
>
> > My question is, in the following call from Python,
> >
> > call_foreign_function(pFunc, 'lll', 'l', (<?>,<?>,<?>))
> >
> > what should the tuple (<?>,<?>,<?>) contain?
>
> params = calldll.membuf(12)
> params.write('\000'*12)
> px = params.address()
> call_foreign_function(pFunc, 'lll', 'l', (px, px+4, px+8))
> rslt = params.read(0,12)
> x,y,z = struct.unpack('lll', rslt)
>
> Actually, Sam's got a slicker way of doing it, but I don't understand
> how it works, so I'll stick with the dumb-but-effective method.
>
> > If I pass (x,y,z), the
> > Python interpreter (Idle in my case) crashes, and this is
> > understandable, since it results in the contents of x, y, and z
> > being used as addresses to write to.  I've tried
> > (id(x),id(y),id(z)), since id() is supposed to return the address of
> > a Python object - Python didn't crash as before, but nothing was
> > copied into x, y, and z, either.
>
> id returns the address as a string. You wouldn't want this anyway.
> The Python int 0 is not a slot - it IS the one and only int object
> with value 0. If the foreign function were to return a 3 into the
> Python int (which it wouldn't - it would clobber something else in
> the struct) you would then have
>
> >> 3 + 0
> 6
>
> (The most commonly used ints are static singletons - try
>  sys.getrefcount(0))
>
> - Gordon
>
>


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gherron at aw.sgi.com  Thu Jul 22 20:52:21 1999
From: gherron at aw.sgi.com (Gary Herron)
Date: Fri, 23 Jul 1999 00:52:21 GMT
Subject: Dreaded Newbie Question
References: <K4tK+KAA26l3EwvK@epstein.co.uk>
Message-ID: <3797BCC5.4FF75D39@aw.sgi.com>

Dani Epstein wrote:
> 
> Hi!
> 
> A newbie to Python (though currently coding in Delphi), I'm having a
> little difficulty tracking down stuff, although I have the tutorial and
> manuals. Where is the best place to get more info, and why is there no
> newsgroup for newbies like myself?
> 
> For the moment though my grateful thanks goes to anyone who can tell me
> how to get a file's size, and where that info is to be found in the
> references.
> 
> BTW, I'm running PythonWin 1.5.1
> --
> 
> Dani Epstein

Use os.stat to get a tuple of information about the file.  The item at
index 6 (symbolically stat.ST_SIZE) is the size in bytes.  The value is
an integer (or a long integer) depending on the particular os.

>>> import os, stat
>>> os.stat('.login')
(33188, 30804640L, 33554464, 1, 27117, 20, 620L, 932451656, 797807295,
925402414)
>>> os.stat('.login')[stat.ST_SIZE]
620L

Documentation is in the "Module os" section.

-- 
Dr. Gary Herron <gherron at aw.sgi.com>
206-287-5616
Alias | Wavefront
1218 3rd Ave, Suite 800, Seattle WA 98101



From larsga at ifi.uio.no  Tue Jul 20 17:37:02 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 20 Jul 1999 23:37:02 +0200
Subject: python vs. lisp (whereis quote?)
References: <x88r9m3cyje.fsf@rpppc2.hns.com> <37946F3B.ECBFE6D6@mail.danbbs.dk> <etdn1wrgjms.fsf@m2-225-11.mit.edu>
Message-ID: <wkpv1n80ld.fsf@ifi.uio.no>

* alex at somewhere.round.here
|
| Perhaps you can define a do-nothing function that you can then
| change behind the module's back:

You can use functions to emulate hooks and indeed Emacs Lisp does use
this already. However, this only goes part of the way towards the
functionality that quote provides. Python has no obvious equivalent of
things like:

  `(html
     (head
       (title "My web page"))
   
     (body :attrs ((lang "en"))
       (h1 "My web page")
       ,@(produce-page)
       (hr)
       (address "blah blah blah")))

or

  (sql-select 
    `(name age
      :from person
      :where (= id ,id))
      :order-by age)

This and performance are the two things I'm least happy with in Python
right now, I think.

--Lars M.



From donn at u.washington.edu  Wed Jul 28 16:50:34 1999
From: donn at u.washington.edu (Donn Cave)
Date: 28 Jul 1999 20:50:34 GMT
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <slrn7psisn.da.wtanksle@dolphin.openprojects.net> <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com> <slrn7pum9t.2ro.wtanksle@dolphin.openprojects.net>
Message-ID: <7nnqeq$g2a$1@nntp6.u.washington.edu>

wtanksle at dolphin.openprojects.net (William Tanksley) writes:
...
| try:
|    ...
| except something:
|    ...
| else:
|    ...
|
| The else block here executes iff the except blocks don't execute.

Beg to differ - it executes iff the try block executes to completion.
Whether an except block executes or not is immaterial.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From tim_one at email.msn.com  Tue Jul 20 20:20:14 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 20 Jul 1999 20:20:14 -0400
Subject: Stackless & String-processing
In-Reply-To: <7n27vi$mm4$2@cronkite.cc.uga.edu>
Message-ID: <001401bed30e$cdb13c20$d0a22299@tim>

[Tim]
> Who said anything about agreeing?  You two obviously disagree, but
> disagreement doesn't imply misunderstanding.  Heck, you've both
> repeated the same unbudging positions so often even *I* understand
> what you're both saying <wink>.

[Graham Matthews]
> : What is the point of your post Tim?

'Twas but a gentle attempt to break the thread out of the loop it appears to
be retracing now for the third time.

> Getting your jollies from adding useless material to a thread?

Sometimes, yes.  Don't think it applies in this case, though.

> :
> : + other equally useful contributions.

My sig?  There was nothing else of mine in the msg.

sarcasm-doesn't-work-on-imaginary-targets<wink>-ly y'rs  - tim





From tim_one at email.msn.com  Sat Jul 17 17:03:52 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Jul 1999 17:03:52 -0400
Subject: Python and Boehm-Demers GC, I have code.
In-Reply-To: <7mq2ol$hk8@ds2.acs.ucalgary.ca>
Message-ID: <000401bed097$e02f7c00$27a22299@tim>

[Neil Schemenauer]
> ...
> The collection of reference cycles is coming for free.  I expected there
> to be some speed tradeoff.

pybench and pystone have small working sets (they're primarily CPU-speed
tests); so there's very little reachable for BDW to trace, so BDW doesn't
have much to do, so runs fast when it's invoked; and with refcounting on,
most pystone/pybench needs are handled by internal free lists that don't
invoke malloc at all after initial setup, so BDW doesn't even get called
often.  IOW, BDW likely isn't doing much of anything in these tests besides
some initial mallocs.

It's curious that list operations appeared to take a major hit in pybench.
This may be related to "[GC_realloc] is very likely to allocate a new
object, unless MERGE_SIZES is defined in gc_priv.h" from the BDW README.
Python uses realloc over & over when growing lists.

> So far the only problem I see is sorting out the mallocs (ie. extension
> modules may use their own malloc).  This should be done anyhow in order
for
> alternate Python malloc implemetations to be used.

I don't think that's going to be happen:  Python doesn't ask anyone to
change anything about how they like to handle their memory today, and Guido
has pronounced that a Major Feature on multiple occasions.  Vladimir
Marangozov's PyMalloc (see his Starship page) caters to that, letting Python
use its own malloc without anyone else getting involved.

> The other problem is objects that expect not to be collected even
> though Python has no references to them (Tkinter callbacks).  I
> don't think this problem is unsolvable.  Things are much better
> than I expected.

On one particular platform, and that platform's version of threads (or are
you running without threads?), yes -- looks great so far <wink>.

> With some tuning maybe the gc version of Python will perform
> better than the regular version.

With or without refcounting?  As before, I think refcounting's contribution
to Python's current performance is vastly under-appreciated.

[about Tkinter callbacks]
> This is really nasty.  I am trying to allocate this data with
> regular malloc.  If I use malloc and free then this data should
> not get collected and the GC should realize that the Python
> functions still have references to them.

This won't work.  A pointer *must* be reachable from *Python's* "root set",
else BDW won't know it exists.  From the BDW README again:

    Note that pointers inside memory allocated by the standard "malloc"
    are not seen by the garbage collector.  Thus objects pointed to only
    from such a region may be prematurely deallocated.  It is thus
    suggested that the standard "malloc" be used only for memory regions,
    such as I/O buffers, that are guaranteed not to contain pointers to
    garbage collectable memory.

> So far I am not having much luck with this approach.

That's only because it's a hopeless dead end <wink>.  As I suggested before,

    ... or _tkinter.c has to maintain a list of allocated
    PythonCmd_ClientData thingies (until PythonCmdDelete is called
    back from Tcl), ...

IOW, this is certainly solvable, but has to be approached in a way that will
work <wink>.  Let BDW do the malloc for these (else the contained CO pointer
will be invisible to it).  The ClientData struct has to remain reachable
from _tkinter.c, though, until Tcl tells PythonCmdDelete it's got no more
use for it.

Another possibility is to use "standard" malloc/free for ClientData structs,
but arrange for Tkinter.py to hold on to a reference to the callable object
(so BDW can find it from Python's root set).  It already holds on to the
*name* of the synthesized Tcl cmd, in a list bound to data attr
_tclCommands.  Change the list to a dict mapping the name to its associated
CO, and fiddle all uses of _tclCommands accordingly.  That's probably the
fastest way out of this jam (&, indeed, should also run faster than the
list.remove() approach used today).

btw-keep-in-mind-that-software-doesn't-work<wink>-ly y'rs  - tim





From jam at newimage.com  Fri Jul  9 16:31:27 1999
From: jam at newimage.com (jam)
Date: Fri, 9 Jul 1999 16:31:27 -0400
Subject: Python Code Repository?
In-Reply-To: <378645AA.5430A026@msfc.nasa.gov>; from Drew McDowell on Fri, Jul 09, 1999 at 06:55:03PM +0000
References: <378645AA.5430A026@msfc.nasa.gov>
Message-ID: <19990709163127.A1964@toast.internal>

On Fri, Jul 09, 1999 at 06:55:03PM +0000, Drew McDowell wrote:
> 
> After seeing the "Brensenham line algorithm" post above, I'm curious
> if there is a public "python code repository" (something akin to CPAN)
> out there?  Or does Starship assume that role?
> 
> -Drew
> 

hi.

I don't think anything as nice as CPAN exists for python-- it's the one
'BadThing' I've heard when I suggested using python in projects. let's make
one. ;) I know enough about python and cgi scripting and sql databases to come
up with something pretty quick, and all we'd need is submissions. there
might already be some examples of code snippets out on the net in various
places.. we'd have to collect them and insert them into the database, but
that's a pretty easy process as well. I'll volunteer quark for the testbed.
does anyone have any further suggestions or ideas on how we can get this
rolling?

regards,
Jeff
-- 
|| visit gfd <http://quark.newimage.com/> 
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From Gary.Pennington at uk.sun.com  Thu Jul  8 07:54:14 1999
From: Gary.Pennington at uk.sun.com (Gary Pennington - UK Performance Centre)
Date: Thu, 08 Jul 1999 12:54:14 +0100
Subject: python core dump (SIGBUS) on Solaris
References: <3F8BA119B48AD11197470008C7A493CB98B443@HQOBGEX1> <37830BB2.94C4613D@uk.sun.com> <37837154.7A6889A@ix.netcom.com> <slrn7o7nb9.eo.amk@207-172-111-58.s58.tnt1.ann.va.dialup.rcn.com>
Message-ID: <37849166.DD658F5@uk.sun.com>

I agree, I think that is a bad design decision.

It would be more efficient to assume that you are processing a string and
only try to unpickle if the string is detected as a pickle. The tricky bit
is detecting that it is a pickled object in the string. Does a pickled
string start with some kind of magic sequence?

Pseudo-Code

if string contains pickle
    return loads(string)
return string

Gary


"A.M. Kuchling" wrote:

> On Wed, 07 Jul 1999 11:25:08 -0400,
>         Stephen J. Turner <sjturner at ix.netcom.com> wrote:
> >> >>> C.load("webJudgeUser=garyp;")
> ...
> >There's a bug in cPickle that causes heap corruption if loads is passed
> >a string (as opposed to a pickled object) starting with 'g', 'h' or
> >'j'.  Try applying the attached patch to the Python 1.5.2
>
>         Wait a sec... why is it automatically trying to unpickle the
> value of the cookie, when the value obviously wasn't pickled in the
> first place.  While the cPickle core-dumping bug should definitely be
> fixed, cookie.py also shouldn't try to unpickle things by default.
>
> --
> A.M. Kuchling                   http://starship.python.net/crew/amk/
> If Abaton's the unearthly reflection of all that we are, Britain's in far
> worse shape than I thought.
>     -- John Constantine in HELLBLAZER #93




From janssen at parc.xerox.com  Fri Jul  9 16:14:43 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Fri, 9 Jul 1999 13:14:43 PDT
Subject: patch to Modules/posixmodule.c for Java JVM compatibility
Message-ID: <99Jul9.131447pdt."3569"@watson.parc.xerox.com>

Guido,

I've been experimenting with loading Python into a Java VM.  This is
part of expanded ILU capabilities for the next ILU release -- one will
be able to build class libraries in C, C++, or Python that can then be
loaded into any of the other ILU languages.  I've found a small
problem with the Modules/posixmodule.c code, which assumes that the
"environ" space provided by the POSIX environment is writable.  I'm
not sure of what the POSIX standard says about this, but I find that
it's not true with Java (JDK 1.2 on Solaris 2.6), and Python will
segfault when loaded via a native library into the JVM.  By doing a
copy when forming the os.environ dictionary, this can be avoided.
Here's a patch, which will apply to 1.5.2.  Enjoy.

Bill

*** 1.1	1999/07/06 02:21:15
--- Modules/posixmodule.c	1999/07/09 20:03:06
***************
*** 278,284 ****
--- 278,287 ----
  convertenviron()
  {
  	PyObject *d;
+ 	char buf[1024];
  	char **e;
+ 	char *copy, *p;
+ 
  	d = PyDict_New();
  	if (d == NULL)
  		return NULL;
***************
*** 287,302 ****
  	/* XXX This part ignores errors */
  	for (e = environ; *e != NULL; e++) {
  		PyObject *v;
! 		char *p = strchr(*e, '=');
  		if (p == NULL)
  			continue;
  		v = PyString_FromString(p+1);
  		if (v == NULL)
  			continue;
  		*p = '\0';
! 		(void) PyDict_SetItemString(d, *e, v);
  		*p = '=';
  		Py_DECREF(v);
  	}
  #if defined(PYOS_OS2)
      {
--- 290,314 ----
  	/* XXX This part ignores errors */
  	for (e = environ; *e != NULL; e++) {
  		PyObject *v;
! 		if (strlen(*e) < sizeof(buf)) {
! 		    copy = buf;
! 		    strcpy (copy, *e);
! 		} else {
! 		    copy = malloc(strlen(*e) + 1);
! 		    strcpy (copy, *e);
! 		}
! 		p = strchr(copy, '=');
  		if (p == NULL)
  			continue;
  		v = PyString_FromString(p+1);
  		if (v == NULL)
  			continue;
  		*p = '\0';
! 		(void) PyDict_SetItemString(d, copy, v);
  		*p = '=';
  		Py_DECREF(v);
+ 		if (copy != buf)
+ 		    free(copy);
  	}
  #if defined(PYOS_OS2)
      {



From fredrik at pythonware.com  Mon Jul 19 08:03:03 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Jul 1999 12:03:03 GMT
Subject: [Q] Type checking...
References: <7mv000$sun$1@feed.teaser.fr>
Message-ID: <001b01bed1de$a8610cb0$f29b12c2@secret.pythonware.com>

Olivier Deckmyn <olivier.deckmyn at mail.dotcom.fr> wrote:
> Abstract : [Q]How to check the type of an instance before(after?) sending it
> to a method...
> 
> Here is my source :
> 
> class ClassA
> ...
> 
> 
> class ClassB :
>    def myMethod(self, anInstance)
> ...
>
> and myMethod expects that anInstance is a ClassA !
> How can I force my user to send only instances of this
> class (or subclasses)?

assert isinstance(anInstance, ClassA)

</F>




From jlj at cfdrc.com  Thu Jul  1 09:43:05 1999
From: jlj at cfdrc.com (Lyle Johnson)
Date: Thu, 01 Jul 1999 08:43:05 -0500
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <377a6bb0.10820509@news.oh.verio.com>
Message-ID: <377B7069.2DBEA496@cfdrc.com>

Kevin,

Right now each instance has its own copy of the message table, and this is
set up in the class constructor. But I would prefer that the message table
data belonged to the class itself (as it is in the C++ FOX, static class
data).

Lyle

Kevin Dahlhausen wrote:

> Lyle Johnson <jlj at cfdrc.com> wrote:
>
> >Just to add fuel to the fire, we now have a Python interface to FOX, a
> >cross-platform GUI toolkit:
> >
> >    http://www.cfdrc.com/pub/FOX/fox.html
>
> How did you get around the requirement of  preprocessor
> table definitions for message handlers?
>
> (I'm downloading it now....)




From ralph.gauges at eml.villa-bosch.de  Wed Jul 14 10:52:18 1999
From: ralph.gauges at eml.villa-bosch.de (Ralph Gauges)
Date: Wed, 14 Jul 1999 16:52:18 +0200
Subject: Recursive method
References: <1280168234-10452044@hypernet.com>
Message-ID: <378CA422.B2E99CB7@eml.villa-bosch.de>

You are absolutely right. So far I had some methods in my
class and made the whole gui right inside the class
definition. This must have created some problems with
namespaces and 'self'. Now I moved the whole gui creation
inside the __init__ method and now 'self' works.
Now I have one more question: Why is there no 'self' right
inside the class definition?

class Test:
	def TestMethod(self):
		print "ABC"

	

	TestMethod()            # <- this doesn't work, because
TestMethod expects an argument (self)
	TestMethod(self)        # <- this doesn't work, because
there is no 'self'
	self.TestMethod()       # <- this doesn't work, because
there is no 'self'

	__init__(self):
		self.TestMethod() # <- now this works just fine


Thanks a lot for the help and I hope, somebody can come up
with an answer to this last question.

Ralph



From markus_kohler at hp.com  Tue Jul 27 06:28:12 1999
From: markus_kohler at hp.com (Markus Kohler)
Date: Tue, 27 Jul 1999 12:28:12 +0200
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk>
		<7n5f0c$snu$1@ffx2nh5.news.uu.net>
		<al81zdxkktf.fsf@sirppi.helsinki.fi>
		<p51zdvn4c7.fsf@bidra241.bbn.hp.com> <14236.32616.980578.536614@dolphin.mojam.com>
Message-ID: <379D89BC.DA91BA8C@hp.com>

Skip Montanaro wrote:
> 
> I got a bit confused attributing this message.  I think we have too many
> Markus's running around loose... ;-)  I think I have the quoted text
> attributed correctly.  Apologies if not.
> 
>     Stenberg> Yes that is much too slow. Good OO style requires a lot of
>     Stenberg> function calls.  I'm sure Visualworks a commercial Smalltalk
>     Stenberg> implementation can do millions of calls on this machine. Ok
>     Stenberg> it's a JIT and it's commercial but's still Smalltalk and
>     Stenberg> therefore at least as dynamic as python. I guess even Squeak
>     Stenberg> (www.squeak.org) would be faster. If you would sent us code, I
>     Stenberg> could prove it.
> 

That's me (Markus Kohler). 

> Better yet, since the Python source code is available, you could prove it by
> speeding up the interpreter.  Many people have optimized different parts of
> Python over the lifetime of the software.  It's significantly faster than it
> was a couple years ago.
> 
> I think one of the impediments to improving the speed of the language has
> been the dramatic performance improvement in hardware.  There's not a lot of
> incentive to pour lots of time into speeding up method calls when for a few
> hundred bucks you can simply drop in a faster processor, disk or networking
> card.

Also it's true that hardware still improves, it's also true that if I would try
sell
an application to a customer with thousands of users and I would tell him to buy
a 
new machine because I use a slow interpreter, they would probably kill me. 
There are also applications where every second you save is worth a lot of many. 
Production testing of IC's comes to my mind ...

> 
>     Kohler> P.S. This might surface as one example for 'why scripting
>     Kohler> languages don't always cut it' in my master's thesis; if I have
>     Kohler> made some grievous error(s?), corrections are welcome.
> 
Even it was not me saying this I would still like to add a comment. 

> I don't recall anyone ever saying scripting languages would always cut it.

Agreed, but they should cut as much as possible shouldn't they ?

> 
>     Kohler> P.P.S. Too bad Python doesn't have anything that creates _nice_
>     Kohler> code, like Stalin for Lisp ;-)
> 
> There is a substantially shorter history to Python than to Lisp.  Give it
> time.  In addition, Lisp has been knocking around the academic world for
> 40-odd years.  With that many graduate students looking for a thesis topic,
> it was likely that some of those theses would focus on performance
> enhancement.

Yes there's hope that Python get's faster. I'm just want to avoid that people 
repeat history by just saying "by new hardware if it's to slow". 


Markus

-- 
Markus Kohler  mailto:markus_kohler at hp.com



From phil at ricochet.net  Fri Jul  9 02:53:21 1999
From: phil at ricochet.net (phil)
Date: Fri, 09 Jul 1999 06:53:21 GMT
Subject: Two Pythons talking to each other?
References: <199907080019.TAA22244@buffalo.fnal.gov> <199907080529.HAA19502@axil.hvision.nl> <3784486a.7993113@news.ricochet.net> <7m2lol$bnm$1@nntp6.u.washington.edu>
Message-ID: <37859856.9981682@news.ricochet.net>

On 8 Jul 1999 17:05:25 GMT, donn at u.washington.edu (Donn Cave) wrote:

>phil at ricochet.net (phil) writes:
>| On Thu, 8 Jul 1999 05:30:40 GMT, "Hans Nowak" <ivnowa at hvision.nl>
>| wrote:
>...
>May be worth noting that application protocols can be and are also
>designed on top of UDP, where among other things they may add some
>kind of sequence monitor - datagram delivery isn't guaranteed, a
>datagram may be discarded if its system buffer has filled up.

Yes, if you are stated or you care about order  then  you probably
need  to sequence. Not only can datagrams not turn up, but they can
turn up out of order. Rare on a LAN, sure.

phil.







From fredrik at pythonware.com  Wed Jul 21 13:03:12 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Jul 1999 17:03:12 GMT
Subject: Merging two lists as a dictionary
References: <19990720211946.A12202@pooh.frostnet.net> <19990721153539.S6170@xs4all.nl> <19990721084614.C15502@pooh.frostnet.net> <T%kl3.3903$VL2.416831@news.direcpc.com>
Message-ID: <000d01bed39a$eb327750$f29b12c2@secret.pythonware.com>

Emile van Sebille <emile at fenx.com> wrote:
> Can anyone tell me why all of Chris's posts are showing up as
> attachments at my end?  (Outlook Express 5, win95b)

he's posting "multipart/signed" messages, by some reason.
looks like outlook express cannot deal with that.

(you can read his posts using ctrl-F3, btw...)

</F>




From Fredrik.Lundh at p98.f112.n480.z2.fidonet.org  Fri Jul  2 04:05:47 1999
From: Fredrik.Lundh at p98.f112.n480.z2.fidonet.org (Fredrik Lundh)
Date: Fri, 02 Jul 1999 09:05:47 +0100
Subject: Mailbox, tell and read
Message-ID: <00001658@bossar.com.pl>

From: "Fredrik Lundh" <fredrik at pythonware.com>

Gordon McMillan wrote:
> > First, are there platforms other than DOS/Windows/Win32 where there
> > is a difference between text and binary mode?
> 
> There's a theoretical difference on *nix - otherwise, they wouldn't 
> have made up the distinction!

who are "they?"

fwiw, the ANSI C rationale mentions Unix as the
special case here... go figure ;-)

</F>



From cfelling at iae.nl  Thu Jul 22 21:15:39 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 23 Jul 1999 03:15:39 +0200
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have code.)
References: <000601bed335$893916e0$a72d2399@tim> <p5hfmymh9z.fsf@bidra241.bbn.hp.com> <Pine.LNX.4.10.9907211223270.11985-100000@atrus.jesus.cam.ac.uk> <p5n1woncm0.fsf@bidra241.bbn.hp.com>
Message-ID: <7n8fnr$2al$1@vvs.superst.iae.nl>

Markus Kohler <markusk at bidra241.bbn.hp.com> wrote:
> Michael Hudson <mwh21 at cam.ac.uk> writes:
>> On 21 Jul 1999, Markus Kohler wrote:
>> > Lately ee have had already some examples in this group where loop
>> > overhead was a major factor. 
>> 
>> Really? When? Sorry, I don't want to sound disbelieving, but I don't
>> remember that.

> It was about reading a file and put the words into some hash table. 
> As far as I can remember to loop overhead was a dominating factor. 
> And Python was much slower than perl for example. 

Well, if I remember correctly, eventually the perl version was slower
then the python optimized version:) The point was that reading bytes is
*much* slower then reading large chunks of data. So it had to do with
disk IO and not looping. Atleast that's what I remember from that thread.
(I checked with deja-news and low and behold I'm right! that is if we
 refer to the same thread titled 'Speed of python' way back in feb:)

-- 
groetjes, carel



From pduffin at mailserver.hursley.ibm.com  Thu Jul 29 06:22:54 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Thu, 29 Jul 1999 11:22:54 +0100
Subject: Why not standard ? (Re: Building a shared libpython)
References: <Pine.GSO.4.03.9907281701150.15012-100000@spica> <001001bed938$b6b34330$f29b12c2@secret.pythonware.com>
Message-ID: <37A02B7E.4083@mailserver.hursley.ibm.com>

Fredrik Lundh wrote:
> 
> Carlos A M dos Santos <nospamplease at inf.ufrgs.br> wrote:
> > I'm installing Python 1.5.2 and noted that there is no configure option to
> > buld a shared libpython.  How could I do it?
> 
> www.python.org => search => newsgroup search for
> "shared libpython" brings up lots of hits.  start with
> these:
> 
>     http://www.deja.com/=dnc/getdoc.xp?AN=488431463
>     http://www.deja.com/=dnc/getdoc.xp?AN=398994871
> 

Why is this not available with the standard distribution ?
Could someone who knows explain why, or someone who can add it to the
standard distribution.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From casantos at inf.ufrgs.br  Wed Jul 28 16:16:49 1999
From: casantos at inf.ufrgs.br (Carlos A M dos Santos)
Date: Wed, 28 Jul 1999 17:16:49 -0300 (EST)
Subject: Building a shared libpython
In-Reply-To: <Pine.GSO.4.03.9907281701150.15012-100000@spica>
Message-ID: <Pine.GSO.4.03.9907281713420.15012-100000@spica>

I wrote:

> I'm installing Python 1.5.2 and noted that there is no configure option to
> buld a shared libpython.  How could I do it?

Sorry, I forgot to mention that I'm using an UltraSPARC with Solaris
2.5.1, gcc 2.8.1 and gnu binutils 2.9.1.


Casantos at inf dot ufrgs dot br




From gmcm at hypernet.com  Thu Jul 29 07:57:15 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 29 Jul 1999 06:57:15 -0500
Subject: global myfunc; def myfunc():
In-Reply-To: <ffQn3.40339$jl.29112404@newscontent-01.sprint.ca>
Message-ID: <1278876856-19833602@hypernet.com>

Bruce Dodson wrote:

> I have noticed that, when using python as an embedded language in
> another program, using a temporary dictionary for a private
> namespace, and using __main__'s dictionary for my global namespace,
> I can say:
> 
> global myfunc
> def myfunc():
>   ...
> 
> to bind myfunc into the global namespace, which allows it to persist
> across calls.  But I want to make sure that is "intended behavior"
> and not just some implementation quirk, so I can decide whether to
> document that as a feature.

It may be an unusual use, but there's nothing quirky about the 
mechanism being used. You're safe.

- Gordon



From fdrake at cnri.reston.va.us  Wed Jul 21 12:14:56 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Wed, 21 Jul 1999 16:14:56 GMT
Subject: Python Mode for Alpha ?
In-Reply-To: <o9vhbevwgh.fsf@panix3.panix.com>
References: <lawter-2107990058210001@client-151-197-119-26.bellatlantic.net>
	<14229.57251.996871.502133@anthem.cnri.reston.va.us>
	<14229.58766.234709.815774@weyr.cnri.reston.va.us>
	<o9vhbevwgh.fsf@panix3.panix.com>
Message-ID: <14229.61952.83320.458319@weyr.cnri.reston.va.us>

>>>>> On Wed, 21 Jul 1999 15:21:50 GMT, Fred L Drake said:
 > No, he's asking about an editing mode for Alpha, a Mac editor.  I
 > remember having one when I used a Macintosh, but that was almost 5
 > years ago.  I have no idea where to get such a beast these days.

Josef Sachs writes:
 > <URL: ftp://ftp.python.org/pub/python/mac/old/AlphaPythonMode.sit.hqx>?
 > Did that stop working with some recent Alpha release?

  I've no idea; I've not used a Macintosh as a development environment 
for at least four years!


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From joeld at anisci.com  Tue Jul 13 16:48:35 1999
From: joeld at anisci.com (Joel Davis)
Date: 13 Jul 1999 13:48:35 PDT
Subject: Wrapping C++ w/inheritance
Message-ID: <378BA78E.B6C19463@anisci.com>

A few weeks ago I posted here asking about wrapping C++ classes in
python, in a way that lets you write derived classes in Python that have
virtual methods which get called from c++..

I got pointers to lots of good info, but I couldn't find a concise
example. I got a little sample working, and figured I would post it in
case other people are trying to figure out how to do the same thing.

Anyways, if anyone is interested, I put it up at
http://www.flashnet.net/~joeldav/pywrap/

joel
--
----------------------------------------------------------------
                                                joeld at anisci.com





From robin at jessikat.demon.co.uk  Fri Jul  2 13:03:34 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 2 Jul 1999 18:03:34 +0100
Subject: ordered threading
Message-ID: <TbLVhCAmDPf3EwPU@jessikat.demon.co.uk>

I have an unbounded number of worker threads which occasionally require
access to a unique server thread. I can easily mutex access to the
server thread, but is there any obvious/easy way to ensure that the
blocked workers wake up in the right order?

Also if blocked can a thread be killed in any way?
-- 
Robin Becker



From Scott.McClellan at p98.f112.n480.z2.fidonet.org  Sat Jul  3 16:12:29 1999
From: Scott.McClellan at p98.f112.n480.z2.fidonet.org (Scott McClellan)
Date: Sat, 03 Jul 1999 21:12:29 +0100
Subject: calldll & PyArg_ParseTuple strings
Message-ID: <000016b5@bossar.com.pl>

From: Scott McClellan <801d at my-deja.com>

Hello,

I'm attempting to use Sam Rushing's calldll module to call a C function
in a dll, with the following prototype:

long __declspec(dllexport)func(long *x, long *y, long *z)

I expect the function to give certain values back in x, y, and z, but so
far have been unable to make that happen. Let's say that I have three
long integers in Python that I initialize as follows:

x=y=z=0L

My question is, in the following call from Python,

call_foreign_function(pFunc, 'lll', 'l', (<?>,<?>,<?>))

what should the tuple (<?>,<?>,<?>) contain?  If I pass (x,y,z),
the Python interpreter (Idle in my case) crashes, and this is
understandable, since it results in the contents of x, y, and z being
used as addresses to write to.  I've tried (id(x),id(y),id(z)), since
id() is supposed to return the address of a Python object - Python
didn't crash as before, but nothing was copied into x, y, and z, either.

Can anyone point me in the right direction?

Thanks,
Scott McClellan


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From ivanlan at callware.com  Sat Jul 10 12:45:27 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sat, 10 Jul 1999 16:45:27 GMT
Subject: Longs
Message-ID: <7m9tno$6qg$66@rks1.urz.tu-dresden.de>

Hi All--
When were longs added to Python?  Whose idea was it?

Thanks,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From phd at emerald.netskate.ru  Fri Jul  9 05:54:12 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Fri, 9 Jul 1999 13:54:12 +0400 (MSD)
Subject: PyApache and Apache 1.3.6 anyone ???
In-Reply-To: <19990708221333.B16851@hoffleit.org>
Message-ID: <Pine.LNX.4.04.9907091351510.657-100000@emerald.netskate.ru>

On Thu, 8 Jul 1999, Gregor Hoffleit wrote:
> Is anybody successfully running PyApache on an Apache 1.3.6 ?
> Preferably on a Linux system ? When I build PyApache and load it from
> httpd.conf, /usr/sbin/apache just doesn't start up. It doesn't give
> any error or what, it just starts up and quits immediately.
> 
> Any experiences ?

   I run PyApache on Linux and Solaris. I had some problems
compiling PyApache on Linux, and finally I got an advice to add -Wl,-E to
gcc compilation flags.
   After this, all things went well.

> Thanks in advance,
> 
> 	Gregor

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From htrd90 at zepler.org  Fri Jul  9 17:22:25 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Fri, 09 Jul 1999 21:22:25 GMT
Subject: win32 python as a dll?
References: <1280583297-81012085@hypernet.com> <37863B4F.FB1780F3@enews.nrl.navy.mil>
Message-ID: <7m9tn3$6qg$44@rks1.urz.tu-dresden.de>

Scott Wolford <wolford at enews.nrl.navy.mil> wrote:

>I guess I didn't pose my question properly. Under Unix, can you not link with a
>debug foo.so at link-time, then at run-time load an optimized version, as long
>as the code used to generate the two is roughly the same? It seems to work for
>me. Why can't you do the same thing under Windows?

msvcrt.dll and msvcrtd.dll are not even roughly the same. There are
differences in the data structures behind malloc, fopen, etc. They are
given different names to prevent them getting mixed up.

If optimisation level was the only difference then you would not need
separate .lib files.



Toby Dickenson



From kernr at mail.ncifcrf.gov  Sat Jul 10 16:50:12 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Sat, 10 Jul 1999 20:50:12 GMT
Subject: Compiling Win32 Python Extensions with EGCS
Message-ID: <7m9tnr$6qg$70@rks1.urz.tu-dresden.de>

I have established a set of instructions and notes for compiling
Python extensions with the free EGCS compiler suite with the mingw32
target.  EGCS is the "experimental" development branch of the GCC
compiler.  The next version will be the official GCC-2.95.  Mumit Khan
has ported EGCS 1.1.2 (the current release) to Win32 systems.  See

    http://wwww.xraylith.wisc.edu/~khan/software/gnu-win32

for more information on this port.

If you just want the instructions and the necessary tools, go to my
web page:

    http://starship.python.net/crew/kernr/mingw32/Notes.html

Many thanks to Paul Sokolovsky for his groundbreaking work on porting
the Python headers to be used by EGCS/mingw32.

NOTE: The mingw32 target does not use Cygnus Solution's Cygwin
portability layer.  Programs compiled for the mingw32 target use the
system's C runtime DLL (for Python extensions, this is MSVCRT.DLL) and
not the Cygwin DLL.  Python extensions compiled as I describe do *not*
automatically fall under the GPL.

I, Robert Kern, may be reached at <kernr at ncifcrf.gov>.

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From claird at starbase.neosoft.com  Sun Jul 18 23:13:05 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Jul 1999 03:13:05 GMT
Subject: what is python?
References: <7mt81n$s6e$1@bgtnsc03.worldnet.att.net>
Message-ID: <7970018B1A94450D.D0FA47F7C6A2DEE5.9D19F01B85820D25@lp.airnews.net>

In article <7mt81n$s6e$1 at bgtnsc03.worldnet.att.net>,
Paul Johnston <johnston.p at worldnet.att.net> wrote:
>Hello people.
>
>I'm a relatively new programmer with a pretty strong java background.
>Although I've never used python, I keep hearing more and more about it, and
>I can't help but wonder what I'm missing.  I want to learn it, but I think
>the fastest way is to put it to work right away.  Therefore, I guess I'm
>trying to figure out what python excels in and then I'll allocate a
>problem-domain (in my own work) to apply it to -- hopefully this will be the
>most pragmatic approach.
Mr. Bower has quite rightly pointed you to python.org as the one
reliable place to start in researching such a question.  I'll add,
though, that this was precisely the subject of our <URL:http://
www.sunworld.com/sunworldonline/swol-02-1998/swol-02-python.html>.
>
>So what are people using python for?  What types of problems is python
Darn near everything.  Shockingly close to everything.
>better suited relative to, say, java?  I do alot of server-side programming
Hmph.  If you *really* need portability, Python's ahead of Java,
in several senses.  Lars Marius Garshol, by the way, has a few
wise words to say on their relative performance.

People use Python when they could use Java 'cause they get
things done quicker.  People who favor Python--I'm one of them--
find it *inviting* to program in Python.
>& xml stuff.
Server-side programming and XML--great subjects.  While there
are many ways to go about server-side Python, we occasionally
argue about whether Zope <URL:http://
starbase.neosoft.com/~claird/comp.lang.python/web_python.html#Zope>
will turn out to be the "killer app" for Python.  Its benefits
in this domain are dramatic enough to inspire such speculation.

It's hard to beat Java for richness of XML libraries.  Python
comes closer than any other language, with only a couple of
possible exceptions.  Programming Python remains more *fun*
than Java, even when XMLing.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From wolfgang.grafen at de.bosch.com  Tue Jul 27 08:48:15 1999
From: wolfgang.grafen at de.bosch.com (Wolfgang Grafen)
Date: Tue, 27 Jul 1999 14:48:15 +0200
Subject: Access dictionary in order?
References: <roy-2507991300370001@mc-as01-p61.med.nyu.edu>
Message-ID: <379DAA8F.1E34513C@de.bosch.com>

Roy Smith wrote:
> 
> I've got a dictionary of items which I will sometimes want random access
> to, but sometimes I'll want to be able to step through the items in a
> fixed order as well.  Any good way to do that?
> 
> The best I've come up with so far is to keep a parallel list of the
> dictionary keys in sorted order and use that to step through when I need
> to:
> 
>    for key in keylist:
>       do_stuff (dict[key])
> 
> which will work, but seems kind of clumsy.  Anything smoother I can do?

Look here:
Probably seqdict is best for you. Look at

http://www.germany.net/teilnehmer/100,366919/Python/Modules/Modules.html

from seqdict import seqdict
dict = seqdict()
.
.
.
for key,value in dict.items():
  do_stuff(key,value)

I think this works for you. But seqdict and mseqdict brings much, much
more
new operations on sequential dictionaries. Tell me what your think.

Wolfgang
-------------- next part --------------
A non-text attachment was scrubbed...
Name: wolfgang.grafen.vcf
Type: text/x-vcard
Size: 329 bytes
Desc: Card for Wolfgang Grafen
URL: <http://mail.python.org/pipermail/python-list/attachments/19990727/62e7d6f9/attachment.vcf>

From kperacles at geocities.com  Fri Jul 23 11:22:06 1999
From: kperacles at geocities.com (Karl Putland)
Date: Fri, 23 Jul 1999 09:22:06 -0600
Subject: COM bites - help please
References: <1279389887-25608327@hypernet.com>
Message-ID: <xR%l3.20$CP5.3484@news.uswest.net>

Gordon McMillan <gmcm at hypernet.com> wrote in message news:1279389887-25608327 at hypernet.com...
> Karl Putland is frustrated:
> 
> > I've got a couple of pyCOM things that are kicking my @$$! 
> 
> Bill Tutt got most of 'em...
> 
> > 2. A computed attribute in __getattr__ does not want to be accessed
> > through com.  It is in the _public_attrs_  list but produces the
> > following error when attempting to access it.
> 
> Both you and the COM support are playing attribute games, and not 
> agreeing on how the game is played. Can you expose Get methods to 
> COM,  and play your games there? That way, COM will find something 
> real (a method object), and you can still do lazy evalutaion.
> 

Kind of works this way already.  field.value was just aliased to field.read.  No shift key and it was easier to read.


> BTW, I find it very helpful to use a separate (wrapper) object:
> 
>  class RealWork:
>    ...
> 
>  class COMRealWorkWrap:
>    _public_methods_ = [ ... ]
>    def __init__(self, real_obj):
>      ...
> 
> That also means I can put all my Unicode conversion in one place.

Hmmm... I might have to try this.  Does your method require that you mirror all of the functions and attributes of the RealWork in COMRealWorkWrap???  Right now I'm just using a procedure to add the attributes to the objects.  And I tried subclassing the root object that I wanted to expose.

class GMcomConnector(gmc.GMConnector):
    """COM wrapper of gmc.GMConnector"""
    _public_methods_ = ['Init', 'GMTable','cleanup']
    _public_attrs_ = ['DefaultDir','User']
    _reg_verprogid_ = "Python.GMConnector.1"
    _reg_progid_ = "Python.GMConnector"
    _reg_desc_ = "Python - GoldMine Connection Manager"
    _reg_clsid_ = "{8DAF3841-3E9B-11D3-A4DC-820D6CB34D6B}"
    _reg_class_spec_ = "gmCOM.GMcomConnector"

    def GMTable(self, table_name):
        """ Override the base to wrap the Table in a COM wrapper"""
        if debug:
            print "getting table=gmc.GMConnector.GMTable(self, TableName)", self, table_name
            print '\n', dir(table_name)
        try:
            # Another Unicode string conversion
            TableName=str(table_name)
            table=gmc.GMConnector.GMTable(self, TableName)
        except:
            import traceback
            traceback.print_exc()
            raw_input("Hit return to exit...")

        
        # set pythoncom specific attributes
        if debug: print "Wrapping Table"
        gm_com_table_prep(table)

        # turn table.fields into a collection
        table.fields=win32com.server.util.Collection(table.fields)

        # Return the COM wrapped table
        return win32com.server.util.wrap(table)
    def Init(self,
             DefaultDir='d:\\goldmine\\demo',
             User='MASTER',
             Password='ACCESS'):
        
        self.DefaultDir = DefaultDir
        self.User = User
        self.Password = Password


def gm_com_table_prep(table_inst):
    """Preppare for COM wrapper of gmtasble.GMTable

    Need to see if we can generate the COM specific attributes on a per
    instance basis.

    GMcomTable is not pulicly accesable.  GMTables are only created by
    the factory in GMcomConnector.  They get wrapped in the COM interface
    there

    """

    # Changes the date format to a string from a date tuple
    table_inst.COM = 1
    
    table_inst._public_methods_ = ['append',
                        'bottom',
                        'close',
                        'delete',
                        'filter',
                        'goto',
                        'is_sql',
                        'new',
                        'open',
                        'range',
                        'read',
                        'replace',
                        'search',
                        'seek',
                        'set_order',
                        'skip',
                        'tables',
                        'top',
                        'unlock']

    table_inst._public_attrs_ = ['name',
                                 'handle',
                                 'records',
                                 'rec_no',
                                 'fields',
                                 'fieldnames']
    if debug: print "\nLooping over fields to set _public_attrs_ for table", table_inst.name
    for name in table_inst.fieldnames:
        table_inst._public_attrs_.append(string.lower(name))
    if debug: print "\nWrapping Fields"
    tmpfields=[]
    for field in table_inst.fields:
        field._public_methods_ = ['read',
                                  'write',
                                  'append']
        field._public_attrs_ = ['name',
                                'size',
                                'type'
                                'value']
        name = string.lower(field.name)
        field = win32com.server.util.wrap(field)
        tmpfields.append(field)
        setattr(table_inst, name, field)
    table_inst.fields = tmpfields


> 
> - Gordon
> 
> 

Thanks

Karl Putland
kperacles at geocities.com





From MHammond at skippinet.com.au  Thu Jul  8 03:56:03 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Thu, 8 Jul 1999 17:56:03 +1000
Subject: Win32: python-network-installation?
References: <378349DE.7ED68819@phoenix-edv.netzservice.de> <37835B30.BFCC4A05@phoenix-edv.netzservice.de>
Message-ID: <7m1led$97e$1@m2.c2.telstra-mm.net.au>

Holger Jannsen wrote in message
<37835B30.BFCC4A05 at phoenix-edv.netzservice.de>...

>To reach Python-dlls I copied the dlls '\WinNT\System32\PythonCOM15.dll'
>and '\WINNT\System32\Python15.dll' to a directory '\\as1\python\dll'
>onto the server.

Dont forget pywintypes15.dll.

>3)only '\\as1\python\pythonwin\pythonwin.exe' occurs a
> 'Microsoft Visual C++ Runtime Library'-RuntimeError:
> Program: \\as1\python\pythonwin\pythonwin.exe
> abnormal program termination
>
>What do you think, does I forget any dll? Why does that error (3)
>occur? (Perhaps I need that pythonwin.exe to check some
>scripts when sitting at a client...?!)

That error normally means 2 Python DLLs were found.  My guess is either
pythoncom15.dll or pywintypes15.dll.

The more I think about it, the problem is almost certainly the fact that
win32api.pyd, win32ui.pyd etc all use pywintypes15.dll - not as a Python
module, but as a regular DLL.  So assuming you still have pywintypes15.dll
in your system directory, the DLL load process will be:

* Python does "import win32api" - part of the NT process finds and loads
pywintypes15.dll, probably from the system directory.
* Python does "import pywintypes" - Python searches for Pywintypes, consults
the registry and finds it listed somewhere else.  Python loads this DLL.

Things then go boom very quickly.

I would put all the DLLs (not necessarily .pyds, but python15.dll,
pywintypes15.dll, pythoncom15.dll) and all the .EXEs (python.exe,
pythonw.exe and pythonwin.exe) into the same directory.  Then seek out and
destroy all other copies of these files from your machine.

>Are there other (better) ways to get a network installation for
Win32-Clients?

Nope, but feel free to help :-)

>[HKEY_CURRENT_USER\Software\Python\Python for Win32]
ALL of these HKCU\Software\Python\Python for Win32" entries should go.  the
installer doesnt create them - Pythonwin does when it shuts down for the
first time.  Unless there are specific preferences you want to default for
these users, dont add them.

>[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File\AutoRegister]
>@="\\\\as1\\python\\DLL\\PythonCOM15.dll"

Dunno where this came from - the installer didnt add it.

>[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5]

You are missing all the important ones - PythonPath and Modules keys are
critical for the win32 extensions to work.

Note the installer sources are on my starship page (in the "buildtools"
archive).

Also note you can become a registered user :-)
http://starship.python.net/crew/mhammond/registration

Mark.






From Ryan_Phillips at yahoo.com  Fri Jul 30 12:21:40 1999
From: Ryan_Phillips at yahoo.com (Ryan Phillips)
Date: Fri, 30 Jul 1999 09:21:40 -0700
Subject: Python as a parser (newbie)
Message-ID: <37A1D114.705957B9@yahoo.com>

I am creating a program that needs to have scripting support.  I've
looked at the faqs and cannot get the answer.  Can I use Python's
scripting ability to parse the input I need to give it, and execute
c/c++ functions realtime in my code?

Any help would be appreciated.

Thanks,
Ryan Phillips



From fredrik at pythonware.com  Wed Jul 28 11:20:11 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Jul 1999 15:20:11 GMT
Subject: A Zope build question: where is sha.py
References: <199907281506.KAA04703@ws89.rjka.com>
Message-ID: <002801bed90c$b01c9250$f29b12c2@secret.pythonware.com>

Huayin Wang <wang at rjka.com> wrote:
> when compiling zope2.0 I got the following error message:
> ....
>   File "../zope/Zope-2.0.0b1-src/zpasswd.py", line 90, in ?
>     import sys, string, sha, binascii, whrandom, getopt, getpass, os
> ImportError: No module named sha
> ....
> 
> I think it is a valid complain because I can't find sha.py anywhere on my system.
> Please help.

it probably works better if you send zope questions 
to one of the zope mailing lists:

    http://www.zope.org/Community/

(some hints, though: are you using Python 1.5.2?
have you configured Python to include the sha
module -- it's on by default, but you maybe you're
using old configuration files...).

</F>





From mcg at halcyon.com  Wed Jul 28 12:52:34 1999
From: mcg at halcyon.com (Michael C. Gilbert)
Date: Wed, 28 Jul 1999 16:52:34 GMT
Subject: python based webmail code?
In-Reply-To: <14239.11258.615882.396388@weyr.cnri.reston.va.us>
References: <rpu849$1$37nspbi$7gc@news.supernews.com>
 <rps64v$0$37nspbi$7gc@news.supernews.com>	<379EBC8B.E36F020B@lemburg.com>
 <rpu849$1$37nspbi$7gc@news.supernews.com>
            <14239.11258.615882.396388@weyr.cnri.reston.va.us>
Message-ID: <v03102849b3c4e56ef47e@[208.215.245.132]>

Hi --

I'm looking for a moderately scalable web based email system and I would
prefer to find one in Python, maybe with a SQL back end.

Any leads?

-- Michael





From fdrake at cnri.reston.va.us  Fri Jul 23 15:12:48 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 23 Jul 1999 15:12:48 -0400 (EDT)
Subject: Python documentation available as ZIP files
In-Reply-To: <14232.48512.648609.496768@lrz.de>
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
	<14232.48512.648609.496768@lrz.de>
Message-ID: <14232.48816.993092.9754@weyr.cnri.reston.va.us>

Eugene Leitl writes:
 > I think this really means lowering the standards. I wouldn't advocate
 > using a more efficient (but obscure) standard as bzip2 (at least not

  That would be my preference actually, but it just isn't common
enough yet.

 > yet), but everybody who is unable to unpack .tar.gz/.tgz archives (you
 > did include a pointer to WinZip in the documentation, didn't you?) is

  It's right there on the downloads page!  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From fredrik at pythonware.com  Thu Jul 29 11:26:38 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 29 Jul 1999 15:26:38 GMT
Subject: Is "shelve" thread-safe on Windows?
Message-ID: <001d01bed9d6$c1ee4800$f29b12c2@secret.pythonware.com>

I wrote:
>
>     http://www.python.org/doc/current/lib/msvcrt-files.html
> 
>     msvcrt.locking(file.fileno(), mode, bytes)
> 
> if anyone knows how to use the "mode" argument,
> please let us know...

okay, okay.  Snipped from the Microsoft
docs:

    The locking function locks or unlocks nbytes
    bytes of the file specified by handle. Locking
    bytes in a file prevents access to those bytes
    by other processes. All locking or unlocking
    begins at the current position of the file
    pointer and proceeds for the next nbytes
    bytes. It is possible to lock bytes past end
    of file.

    mode must be one of the following manifest
    constants:

    _LK_LOCK = 1 # lock the file region

    Locks the specified bytes. If the bytes cannot
    be locked, the program immediately tries again
    after 1 second. If, after 10 attempts, the bytes
    cannot be locked, the constant returns an error.

    _LK_NBLCK = 2 # non-blocking lock

    Locks the specified bytes. If the bytes cannot
    be locked, the constant returns an error.

    _LK_NBRLCK = 4 # non-blocking lock for writing)

    Same as _LK_NBLCK.

    _LK_RLCK = 3 # lock for writing

    Same as _LK_LOCK.

    _LK_UNLCK = 0 # unlock the file region

    Unlocks the specified bytes, which must have
    been previously locked.

    Multiple regions of a file that do not overlap
    can be locked. A region being unlocked must
    have been previously locked. locking does
    not merge adjacent regions; if two locked
    regions are adjacent, each region must be
    unlocked separately. Regions should be
    locked only briefly and should be unlocked
    before closing a file or exiting the program.

I'm sure Fred Drake would love if someone could take
this, rephrase it (so it doesn't look as it's cut right out
of Microsoft's copyrighted docs), and submit patches
to the library reference.  don't forget to emphasize
that you must use seek (or possibly os.seek(file.fileno())?)
to go to the beginning of the region before you issue
the locking command.

Extra bonus points if you tweak msvcrt so it exports
the various mode flags.

And even more bonus points if you add docstrings
to msvcrt.

Supermega many bonus points for anyone who
implements and documents a locking abstraction
which works on all major unixes *and* windows.

</F>




From greg.ewing at compaq.com  Wed Jul 21 23:44:02 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 22 Jul 1999 15:44:02 +1200
Subject: gadfly RDBMS ???
References: <Pine.LNX.4.04.9907211629470.12211-100000@emerald.netskate.ru> <3795C21F.A0F13648@onera.fr> <3795EA7F.4A452490@arcus.tel.hr>
Message-ID: <37969382.C25210F4@compaq.com>

Darko Cvagic wrote:
> 
> Thanks, I think that I'm exchange terms like RELATION and LINK
> So, Gadfly IS RELATIONAL, but I can't make something like:
>         CREATE LINK order_customer
>         FOREIGN KEY order(cno)
>         REFERENCES customer
>         ON DELETE RESTRICT
> 
> I can't do it in gadfly, am I wright?
> What is name for construction like this?
> I was thought that it is RELATIONAL , I was wrong!

I think you're confused about the meaning of the term
"relational" in this context. In relational database
terminology, "table" and "relation" are synonymous.

"Relation" is a mathematical term for a set of tuples.
The tables in a relational database can be thought of
as mathematical relations. Each row (tuple) of the
table "relates together" the data items that it 
contains, hence the name. Relations have some nice
mathematical properties which make them good for
building a database out of, and there is a large
body of theory on the subject.

Just to be confusing, a relation in a relational
database isn't necessarily the same thing as a
relationship in an entity-relationship model.
A relation (table) could be used to represent either
entities or relationships in an ER model.

Furthermore, although there are relationships
(in the non-technical sense) between different
tables, as far as the relational model is concerned,
they are implicit. You don't have to explicitly
declare "links" between tables. Some relational
database systems allow you to do so if you want,
for efficiency reasons, but they don't require
you to. If they did, they wouldn't qualify as
relational.

The same goes for indexes, key declarations,
integrity constraints, etc. etc. They're optional
extras to the basic relational model which
Gadfly doesn't provide -- but that doesn't mean
Gadfly isn't relational!

Hope that helps,
Greg



From tim_one at email.msn.com  Wed Jul 28 22:27:52 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 28 Jul 1999 22:27:52 -0400
Subject: HPUX and threads?
In-Reply-To: <al8yag0x66w.fsf@sirppi.helsinki.fi>
Message-ID: <000b01bed969$f5b62120$71a22299@tim>

[Markus Stenberg]
> Is HPUX-11 thread support tested/implemented/whatnot?

You may want to ask on the Thread-SIG if you don't get an answer here.

> This at least does not look promising..
>
> # /opt/python/bin/python
> pthread_mutex_init: Invalid argument
> Memory fault(coredump)
>
> (also tests failed with the same message, obviously)

Are you running pthreads?  Python is sure trying to <wink>.

The msg is coming from line 263 of Python/thread_pthread.h.  It means the
platform's

		status = pthread_mutex_init(&lock->mut,
					    pthread_mutexattr_default);

returned a non-zero (failure) status.  "Invalid argument" is what the
platform perror() produced.  A coredump after an error here isn't
surprising.

This is likely the first call made to any pthreads function, and is so
vanilla that I'd guess pthreads itself doesn't work on your platform (does
it from C?), or it's some non-std pthreads variant that thread_pthread.h
doesn't know about.

not-that-insight-implies-a-cure-ly y'rs  - tim





From ivanlan at callware.com  Wed Jul  7 22:51:48 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Thu, 8 Jul 1999 02:51:48 GMT
Subject: [PSA MEMBERS] Using (physical) units and quantities
References: <19990707181324.5617.qmail@lannert.rz.uni-duesseldorf.de> <3783D5AD.70D95451@mail.ncifcrf.gov> <3783E395.699436D1@quasar.ipa.nw.ru>
Message-ID: <37841244.2F9184CE@callware.com>

Hi All--

"Alexander V. Voinov" wrote:
> 
> Robert Kern wrote:
> 
> > Although UNITS.py is a small convenience module, you might want to
> > rename it so that those of us unfortunate enough to use case-insensitive
> > Windows can use modules that import both.
> 
> Please explain me this statement: Are there any versions of Windows with
> case-censitive filesystems? FAT32? Is it possible to switch this on and off
> for the existing versions: NT 4.0, 95, 98?
> 

I use all of them.  I have used all file systems on all of them.  They
all suck.

Change the name of your module, make life easier for everyone who has to
suffer under the yoke of the Borg collective.

<or-is-it-change-the-name-of-arkansas?>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From gerrit.holl at pobox.com  Thu Jul  1 15:35:28 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 1 Jul 1999 21:35:28 +0200
Subject: print color strings?
Message-ID: <19990701213528.A8877@optiplex.palga.uucp>

Hello,

is there an easy way to print color strings in python?

A function like printcolor('brightyellow', 'red', 'spam'), that finds
out on what kind of terminal you are and prints it to stdout?

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From claird at starbase.neosoft.com  Mon Jul 19 22:59:37 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Jul 1999 02:59:37 GMT
Subject: Python and the Web
References: <3793dd72.0@hoh.olympus.net>
Message-ID: <AFCEC557B580B99F.7F0096BF9FDA5C47.0D15AF43E5EFE822@lp.airnews.net>

In article <3793dd72.0 at hoh.olympus.net>,
Michael Novello <nospam at this.address> wrote:
>Can anyone tell me where to find information on using python to grab form
>data and create dynamic web pages?
			.
			.
			.
<URL:http://starbase.neosoft.com/~claird/comp.lang.python/web_python.html>
operates at a higher level, but each of its references details a specific
approach to your question.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From maedel at nospam_erols.com  Thu Jul  1 18:43:17 1999
From: maedel at nospam_erols.com (Giambattista Vico)
Date: Thu, 01 Jul 1999 18:43:17 -0400
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <377a6bb0.10820509@news.oh.verio.com> <377BDEF7.C34A6088@compaq.com>
Message-ID: <maedel-0107991843320001@209-122-226-23.s277.tnt1.nyw.ny.dialup.rcn.com>

I'd like to hear what people have to say in this thread if we consider
freezing or otherwise creating a standalone app, even if only for Unix and
Win. Has anyone found a Python tool or combination of tools that solves
this problem?

--tom



From Greg.Ewing at p98.f112.n480.z2.fidonet.org  Fri Jul  2 17:05:46 1999
From: Greg.Ewing at p98.f112.n480.z2.fidonet.org (Greg Ewing)
Date: Fri, 02 Jul 1999 22:05:46 +0100
Subject: total idiot question: +=, .=, etc...
Message-ID: <0000163c@bossar.com.pl>

From: Greg Ewing <greg.ewing at compaq.com>

Guido van Rossum wrote:
> 
> Oh, but that check *is* being made!

No, it's not. The check I'm talking about would
cause the call to A.foo in class C below to
fail. (And it doesn't -- I've just tried it.)

class A:
    def foo(self):
        print "foo of A"
        
class B(A):
    def foo(self):
        print "foo of B"
        A.foo(self)
        
class C(B):
    """Fred Coder originally wrote this
    so it was C(A), then changed it and
    forgot to update the method."""
    def foo(self):
        print "foo of C"
        A.foo(self)

c = C()
c.foo()

Do you see what I mean now?
Greg



From fdrake at cnri.reston.va.us  Fri Jul 16 09:17:03 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 16 Jul 1999 13:17:03 GMT
Subject: [XML-SIG] sgmllib has problems with dots in tag names
In-Reply-To: <19990716125230.A14994@sz-sb.de>
References: <19990716125230.A14994@sz-sb.de>
Message-ID: <14223.12495.910694.301072@weyr.cnri.reston.va.us>

Andreas Jung writes:
 > The SGML parsers from the standard sgmllib and the XML sgmllib war both
 > unable to parse SGML tags with dots in the tag name like <TI.DOC>. The 
 > parsers callback functions only get the first part of the tag name (before
 > the dot) as argument (in this case 'TI'). Because the tags are valid SGML
 > tags this is a bit annoying. Ok, one could get a workaround by replacing
 > all dots in tags with an underscore however that's not a clean solution :-)

Andreas,
  Ok, I've poked at the standard sgmllib a bit to see what the problem
is.  The parser is recognizing the start and end tags.  Once
recognized, it is looking for the handler methods start_*() / end_*()
or do_*().  Since there's a dot in the name, these methods are not
defined, and the unknown_*tag() methods are called instead of the
handle_*tag() methods.
  It should be easy to override the unknown_*tag() methods to use a
table-based dispatcher or performs some form or name mangling, then
passes known tags through to the handle_*tag() methods or whatever.
This seems to be the easiest way to deal with the situation in the
short term.
  If you have any suggestions for a better approach to take, I'd love
to hear it.  It may not be unreasonable to use a mechanism similar to
that used by xmllib (a table of registered handler methods).


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From nick.bower at ssec.wisc.edu  Wed Jul 21 10:56:05 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Wed, 21 Jul 1999 09:56:05 -0500
Subject: Scientific Plotting?
References: <19990721010203.96101.qmail@hotmail.com>
Message-ID: <3795DF85.320EADBE@ssec.wisc.edu>


Michael Huster wrote:
> 
> I am also trying to find an adequate *object* based plotting package.
> I am writing some signal processing software and want to code plotting
> into the classes I am defining.
> 
> <snip>
>
> I am kind of bummed. If we wrap intelligent python wrappers around
> something, we will post it.
> 

I agree.  I think it it's great for people loosely affiliated with the
development community to be writing these packages (and a couple of
early but promising packages have been pointed out in this thread), but
what about the people most intimately involved with the design and
distribution of Python?  Is it a concern?

I see Python as a great possibility for a teaching language, especially
for young science/engineering students to get in and get the job done. 
And it's free, which goes a long way in my struggling physics department
back home.  But how can someone promote the language when people will
most understandably spend hundreds/thousands of dollars to fall back to
the commerical maths packages because the plotting is superb and really
flexible?  Maybe I'm the only one that sees python could reach a wider
audience without too much more effort - I wouldn't suggest I'm right
though! ;)

nick

-- 

Nick Bower
Space Science and Engineering Center
University of Wisconsin, Madison, USA
Phone: (608) 265 8007
Email: nick.bower at ssec.wisc.edu
Web: http://arm1.ssec.wisc.edu/~nickb



From markusk at bidra241.bbn.hp.com  Tue Jul 27 11:23:58 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 27 Jul 1999 17:23:58 +0200
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <p51zdvn4c7.fsf@bidra241.bbn.hp.com> <14236.32616.980578.536614@dolphin.mojam.com> <al8aesin1jr.fsf@sirppi.helsinki.fi>
Message-ID: <p5hfmqm7zl.fsf@bidra241.bbn.hp.com>

Markus Stenberg <mstenber at cc.Helsinki.FI> writes:

> Skip Montanaro <skip at mojam.com> writes:
> > I got a bit confused attributing this message.  I think we have too many
> > Markus's running around loose... ;-)  I think I have the quoted text
> > attributed correctly.  Apologies if not.
> 
> You had it reversed ;)
> 
> > Better yet, since the Python source code is available, you could prove it by
> > speeding up the interpreter.  Many people have optimized different parts of
> > Python over the lifetime of the software.  It's significantly faster than it
> > was a couple years ago.
> 
> Being able to note a problem differs greatly from being able to _fix_ a
> problem ;-) I've glanced thru it, and can't offhand find anything _very_
> badly implemented that I could fix.

I agree. The code looks really good. That's why I think that  there are some design
issues to solve if one want to have a faster Python. 

[deletia]


Markus (Kohler)
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From fredrik at pythonware.com  Mon Jul 26 12:50:02 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 16:50:02 GMT
Subject: Tkinter bug ?
References: <379C8F64.E48BF26@lri.fr>
Message-ID: <001301bed786$e88a7490$f29b12c2@secret.pythonware.com>

Stephane Conversy <conversy at lri.fr> wrote:
> when executing File2.py, there is no error, but no image appears in the
> button.
>
> it work fine. It seems to be a bug, isn't it ?

it's a feature.  see:
http://www.python.org/doc/FAQ.html#4.69

</F>




From olevi at my-deja.com  Fri Jul  2 04:03:27 1999
From: olevi at my-deja.com (olevi at my-deja.com)
Date: Fri, 02 Jul 1999 08:03:27 GMT
Subject: TIFF-images
Message-ID: <7lhroa$62p$1@nnrp1.deja.com>

Is there any way to present TIFF-images through tkinter? After what I've
read, only GIF, PPM and PGM are supported, but maybe there's a way
around?

I'm working with many large images (>8M) both as input and output, so
conversion should be really be avoided. But if anyone knows a good
(loss-less, fast) converter TIFF<->GIF I am also interested in that.

Ole-Kr. Villaboe, ESRF


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From claird at Starbase.NeoSoft.COM  Tue Jul  6 01:40:18 1999
From: claird at Starbase.NeoSoft.COM (Cameron Laird)
Date: 6 Jul 1999 00:40:18 -0500
Subject: Python under Java?
References: <MrUJKLEB0KGWFViX4p@holmes.parc.xerox.com> <p0dg3.895$6M6.293143@news.shore.net>
Message-ID: <7ls4s2$2pg$1@Starbase.NeoSoft.COM>

In article <p0dg3.895$6M6.293143 at news.shore.net>,
Michael P. Reilly <arcege at shore.net> wrote:
>Bill Janssen <janssen at parc.xerox.com> wrote:
>: Has anyone tried embedding Python under Java, using JNI?  When I try it
>: (with JDK 1.2 on Solaris 2.6 and Python 1.5), I get SEGV inside
>: Py_Initialize().
>
>I don't think anyone has trying to _embed_ Python in Java.  But Python
			.
			.
			.
To the contrary:  see <URL:http://
pip.dknet.dk/~pip1848/jpython/Readme_jnios.html>.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From Robin.Becker at p98.f112.n480.z2.fidonet.org  Sun Jul  4 11:34:25 1999
From: Robin.Becker at p98.f112.n480.z2.fidonet.org (Robin Becker)
Date: Sun, 04 Jul 1999 16:34:25 +0100
Subject: ordered threading
Message-ID: <000016c3@bossar.com.pl>

From: Robin Becker <robin at jessikat.demon.co.uk>

In message <1281033720-53905059 at hypernet.com>, Gordon McMillan
<gmcm at hypernet.com> writes
>Robin Becker writes:
>
>> In article <000a01bec583$30332b40$769e2299 at tim>, Tim Peters
>> <tim_one at email.msn.com> writes
>
>> >[Robin Becker]
>> >> I have an unbounded number of worker threads which occasionally require
>> >> access to a unique server thread. I can easily mutex access to the
>> >> server thread, but is there any obvious/easy way to ensure that the
>> >> blocked workers wake up in the right order?
>> >
>> >No threading system on earth allows specifying the order in which blocked
>> >threads will wake up, although some systems (not Python's) allow you to
>> >*influence* the order via assorted "priority" gimmicks.
>
>> the workers are HTTP requests in separate threads from Zope I guess.
>> I want to serialise access to the time/space hog. I guess it could
>> run uniquely in the worker threads ie the locked resource would be
>> access to the hog code. I'm not sure that would let me guarantee
>> unlocking as there are ways for the worker threads to suicide.
>
>If that's what you want, you don't really need to order the 
>unblocking. Just give each thread an input queue protected by a 
>Condition. Worker threads put requests on the Hog thread's input 
>queue (including a ref to the return queue). If you find that some 
>worker threads seem to get better access to the Hog's queue than 
>others, a time.sleep(0.001) will probably fix it - since it will 
>pretty much force a context switch.
>
>
>- Gordon
thanks; I knew already that this could work. I didn't know how to force
equalisation.
-- 
Robin Becker




From nick.bower at ssec.wisc.edu  Sat Jul 10 19:12:42 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sat, 10 Jul 1999 18:12:42 -0500
Subject: Simple Qu: Changing directory
Message-ID: <3787D36A.815B7C29@ssec.wisc.edu>

I feel really dumb asking this question, but how do I change the current
working directory?  For browsing the file system I mean.

I've found exec commands, but these would surely be device dependent and
there doesn't seem to be anything obvious in the os or os.path
documentation.  i looked through the tutorials too.

thanks in advance, nick





From msdemlei at tucana.harvard.edu  Mon Jul 26 15:53:45 1999
From: msdemlei at tucana.harvard.edu (Markus Demleitner)
Date: 26 Jul 1999 14:53:45 -0500
Subject: Extension module mystery
Message-ID: <379caeb9.0@cfanews.harvard.edu>

I know this will embarrass me, but after fiddling around with it
for an afternoon I'm almost desperate.  I and my python interpreter
are having a bad crisis.

I'm trying to write a C extension module for python, basically a
wrapper around a function taking two strings and returning an
integer.  This works like a charm, but was too slow when a problem 
required that this function was called for quite a number of times.

So I thought I'd cut down on python's function call overhead by
letting the C module perform the operation on an entire list of
strings.  So I came up with a function like
static PyObject *list_ldws(char *str,PyObject *strlist)
{	int listlen=PyList_Size(strlist);
	PyObject *intlist=PyList_New(listlen);
	int i,dist;
	char *str2;

	/*Py_INCREF(intlist); -- PyList_New does this for me, right?*/
	for (i=0;i<listlen;i++)
	{	str2 = PyString_AsString(PyList_GetItem(strlist,i));
		dist = WLD(str,str2,40);
		PyList_SetItem(intlist,i,PyInt_FromLong(dist));
	}
	return intlist;
}
that is called from the glue function.  The intlist is returned
like it comes out of this function.

Now, when I try the following:
lis = ldw.ldw("Hol",["Kohl","Pfohl"]) 
a = max(lis)
(ldw.ldw being the glue function), I get
Traceback (innermost last):
  File "zw.py", line 3, in ?
    a = max(lis)
TypeError: argument 2: expected string, list found
(that's python 1.5.2, FWIW).  Odd, isn't it?  Why
should max insist on a string?

But it gets better than that:
When I do 
lis = ldw.ldw("Hol",["Kohl","Pfohl"]) 
print type(lis)
a = max(lis)
suddenly everything is fine, and a has
the correct value, etc.

Now, I strongly suspect that in the C code I did something I
should not do.  The only problem is that I can't see what
that might be.  Of course, I'd also be interested in knowing
why a print statement appearently fixes whatever sacrileg I
commited...

        Markus




From claird at Starbase.NeoSoft.COM  Thu Jul 15 11:16:59 1999
From: claird at Starbase.NeoSoft.COM (Cameron Laird)
Date: 15 Jul 1999 10:16:59 -0500
Subject: Harlequin Is Back!
References: <%vti3.121$95.4053@newse2.tampabay.rr.com> <joswig-1307992158330001@194.163.195.67> <86btdgz2rg.fsf@g.pe <wkn1wzl9kt.fsf@ifi.uio.no>
Message-ID: <71E04BBBB50245BC.87397A36A7EDBBA0.44F0816EBAA77500@lp.airnews.net>

In article <wkn1wzl9kt.fsf at ifi.uio.no>,
Lars Marius Garshol  <larsga at ifi.uio.no> wrote:
>
>* Rainer Joswig
>|
>| How about interfacing to ugly things like MFC?
>
>* Chris Double
>| 
>| The windows version of Python manages this if I remember correctly.
>
>You do remember correctly. This is one of the things I like the most
>about Python, compared to languages like Java and Common Lisp: it
>interfaces to everything, including, but not limited to, CORBA, COM,
>Gtk, ActiveX, MFC, XML-RPC, XML, most IETF network protocols, SGML,
>Java, C/C++, Qt, ASP, Apache, tcl and Perl (through Minotaur) and even
>things like IRIX-specific audio interfaces.
>
>Most languages can do this, but the ease with which you can do it in
>Python impresses me and makes it a very attractive glue language.
>
>| My hat is off too them for managing it!
>
>Mine as well. :)
>
>--Lars M.

Lars is right (as usual).  Actually, compared to *any-
thing*, Python seems to arrive first on the interfacing
scene.  This is interesting; Tcl and Lua emphasize ease
of embedding and extensibility in their design, and Perl
has hordes of interfacing worker bees, but a consistent
pattern emerges that Python does at least as well as any
other language in achieving connection to the latest
toolkits-protocols-...  It's an interesting story, even
though I don't yet understand it.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From tom at krause.dk  Wed Jul  7 04:01:12 1999
From: tom at krause.dk (mrk)
Date: Wed, 07 Jul 1999 10:01:12 +0200
Subject: Python under Java?
References: <MrUJKLEB0KGWFViX4p@holmes.parc.xerox.com> <p0dg3.895$6M6.293143@news.shore.net> <7ls4s2$2pg$1@Starbase.NeoSoft.COM>
Message-ID: <37830948.1CE318DF@pipmail.dknet.dk>

Cameron Laird wrote:

> In article <p0dg3.895$6M6.293143 at news.shore.net>,
> Michael P. Reilly <arcege at shore.net> wrote:
> >Bill Janssen <janssen at parc.xerox.com> wrote:
> >: Has anyone tried embedding Python under Java, using JNI?  When I try it
> >: (with JDK 1.2 on Solaris 2.6 and Python 1.5), I get SEGV inside
> >: Py_Initialize().
> >
> >I don't think anyone has trying to _embed_ Python in Java.  But Python
>                         .
> To the contrary:  see <URL:http://
> pip.dknet.dk/~pip1848/jpython/Readme_jnios.html>.
> --

Not quite. jnios is a reimplementation of the posix module in JNI & java. It
does not attempt to run the python VM under java. I believe that is what Bill
Janssen have in mind.

Regards,
Finn Bock
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bckfnn.vcf
Type: text/x-vcard
Size: 317 bytes
Desc: Card for Finn Bock
URL: <http://mail.python.org/pipermail/python-list/attachments/19990707/931c3679/attachment.vcf>

From flowit at hotmail.com  Thu Jul  1 17:08:21 1999
From: flowit at hotmail.com (Florian W.)
Date: Thu, 1 Jul 1999 23:08:21 +0200
Subject: Newbie has trouble with in/output
Message-ID: <7lglb4$ksu$1@news.vossnet.de>

Hello Python-people

I have two little problems with Python (using WinNt )

1.) Being in a pipe:

I won't to filter the output of the compiler (e.g. only errors and warning
are reported, no logo etc.)
using the pipe-mechanismen :

cl.exe blabla | python filter_compiler_output.py

How do I catch the input and how do print the output ?
The processing requires line-oriented handling, so that I can't catch all
input first, process it as a block and finally display it all at once.
Is the way of calling correct at all ?

2.) Catch the output of an application:

This time I call the compiler myself (e.g. os.system('c:\cl.exe blabla'))
and
want to process the output of the compiler. But how do I catch it ?
os.system just returns the error code. And hacks like c:\cl.exe blabla >
output.txt
don't look very good. (Especially if there are some perl-hackers)

Mfg Florian








From Johan.Wouters at p98.f112.n480.z2.fidonet.org  Fri Jul  2 15:45:13 1999
From: Johan.Wouters at p98.f112.n480.z2.fidonet.org (Johan Wouters)
Date: Fri, 02 Jul 1999 20:45:13 +0100
Subject: Why does shelve make such large files?
Message-ID: <00001677@bossar.com.pl>

From: Johan Wouters <johanw at easics.be>

<snip>
> Ahem, with all due respect to everyone... this is humbug.
> 

Why so?

> Pickle serializes your data in one sweep to/from disk.  It's compact.

So? I don't remember me having any trouble with that statement. As a matter
of fact this is just what I stated.


> But the last claim is horrendously misleading, I'm afraid: data storage
> is not a space/time tradeoff.  It's about throughput (I/O bottlenecks)
> and overhead in managing the supported data and indexing schemes.  There
> are order-of-magnitude performance differences in how several solutions
> work, because of this.

The general idea was some typical engineering tradeoff. Larger database
environments *might* give you better performance for some criteria like
searching or combining data. Say you have an amount of data you want to 
search. You could keep it as a serial stream (minimal space) or you could
use some clever hash method and fixed size records to gain speed. This
way you can easily bypass the overhead of accessing the whole database
(and the slow disks and, ...) by simply calculating where the data should
be. On the other hand, this structuring of the data will incur some extra
space. It looks as if this is just the space/time tradeoff I was talking about
..


> 
> Shudder.  The notion that a large database package, or a large datafile,
> is faster, is so far from reality that it has to be corrected, even in
> this Python-oriented newsgroup.  My apologies for the S/N ratio drop.

I was talking about the datafile, not the package! Also I never stated
that bigger files WILL give you better performance.

Maybe the whole story was a little simplistic, but with all respect: how
does your contribution add to the S?

So Gerrit, I hope you still got some insight from all this!

Kind regards,
Johan Wouters



From ron at library.ucsf.edu  Thu Jul 29 12:36:20 1999
From: ron at library.ucsf.edu (Ron Klatchko)
Date: Thu, 29 Jul 1999 09:36:20 -0700
Subject: Use of variables form outside the class.
References: <Pine.LNX.4.04.9907291736120.31968-100000@emerald.netskate.ru>
Message-ID: <37A08304.A2498313@library.ucsf.edu>

> > How do I reach the "x" variable that is outside any class from inside the
> > class functions?
> 
> x = 1
> 
> class X
>    def set_x(self):
>       self.x = 2
> 
>    def set_global_x(self):
>       global x
>       x = 2

   def get_x(self):
       return self.x

   def get_global_x(self):
       return x

Also, global is only necessary if you want to modify the variable.  If
you only need to access it, the Python namespace rules will find it
without needing the global statement.

moo
----------------------------------------------------------------------
          Ron Klatchko - Manager, Advanced Technology Group           
           UCSF Library and Center for Knowledge Management           
                        ron at library.ucsf.edu



From gmcm at hypernet.com  Mon Jul 19 13:36:17 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 19 Jul 1999 12:36:17 -0500
Subject: debugging MSVC DLLs
In-Reply-To: <37933F54.4AE54D39@cornell.edu>
Message-ID: <1279720534-5719823@hypernet.com>

Tom Adelman asks an MSVC question:

> I'm wondering if someone can help me with this (or should I direct
> this question to a Microsoft group?):
> 
> What I'm doing now:
>     I write classes in MSVC++, compile them as a DLL, and SWIG them.
>      I
> then run python.exe from within the Microsoft Developer Studio, and
> then within Python load this DLL.  This way, I can use the MS
> debugger to look at what's going on within my classes while I
> manipulate them in Python.
>     The problem with this is that I can only debug the DLL whose
>     project
> I used to start Python; I don't know how to, for example, write a
> new DLL and switch to debugging the new one.

Project Setup -> Debug -> Additional DLLs.

When you step into one, you _may_ have to browse for the source (if 
the DLL was copied from where it was built) when MSVC pops up the 
browse dialog.
 

- Gordon



From fdrake at cnri.reston.va.us  Tue Jul  6 10:55:56 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Tue, 6 Jul 1999 14:55:56 GMT
Subject: Updated Python documentation available!
Message-ID: <14210.6396.839797.113760@weyr.cnri.reston.va.us>

  Updated documentation for Python 1.5.2 is now available.  The online
version is available at:

	http://www.python.org/doc/

  Archives of the documentation in HTML, LaTeX, PDF, and PostScript
formats are available at the same location, and may also be downloaded
via FTP from:

	ftp://ftp.python.org/pub/python/doc/

  If you have any questions or comments on the documentation, please
send email to:

	python-docs at python.org

  Many additional modules have been documented (22 more!).  Many small
improvements have been made to other parts of the documentation as
well.  The following modules are now documented:

  asyncore	    dl                sched           
  chunk		    fpformat          SimpleHTTPServer
  cmp		    htmlentitydefs    statcache       
  cmpcache	    mutex             statvfs         
  codeop	    new               sunau           
  curses	    nis               SUNAUDIODEV     
  dircache	    pty               tty
  dl		    rlcompleter   

  Windows users:  If you installed Python using the installer rather
than by building from source, you can get the updated documentation by 
downloading the HTML archive and unpacking it into the "Doc" directory 
of your installation.

  Most of the new sections were contributions from the Python
community; your help makes the Python documentation some of the best
available for any free programming language!

  Enjoy!


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From nobody at newsfeeds.com  Thu Jul 15 00:59:40 1999
From: nobody at newsfeeds.com (Anonymous)
Date: Wednesday, 14 Jul 1999 22:59:40 -0600
Subject: 18 and Over Please 99894
Message-ID: <14079922.5940@Yfgsefs.com>

ADULTS ONLY!

Click the link below:

http://www.filthywhore.com/warning.html

* 18+ Only  Please!


rS at cRCZlON

  --------== Posted Anonymously via Newsfeeds.Com ==-------
     Featuring the worlds only Anonymous Usenet Server
    -----------== http://www.newsfeeds.com ==----------



From hanson at europa.com  Sat Jul 10 01:13:58 1999
From: hanson at europa.com (hanson at europa.com)
Date: Sat, 10 Jul 1999 05:13:58 GMT
Subject: Python Code Repository?
Message-ID: <7m9tne$6qg$56@rks1.urz.tu-dresden.de>

I looked at Troves web site and reviewing the mailing list archives.. I
think that trove is not off the ground yet. 

Using CPAN is not fun. I find it's organization is not intuitive (about intuitive as perl itself). In the python world I find that not having a organized central repository not as bad as it seems. I tend to find new modules by skimming the mailing list and reading www.python.org & freshmeat. Lately, though there is a lot of redundancy on the mailing list where interested parties are inquiring about previously discussed modules. 

One thing that I have found useful using python & the contributed modules
is that they reside where the author has control, so that he/she may provide hints & documentation and divulge the uses & enchancements to his/her module. Also, most of the modules have an interdependency between python and products. 

To the point, I don't think that a CPAN like structure is as useful for
python as it sounds, but I feel that there is something else we can do to
help the python community be more efficient in proliferating modules &
extensions to the language. Where a repository for modules would be
useful is when the author has not provided packages such as RPMs or full
installation & usage instructions.

A perl script that I use (very) frequently called cvsweb.
cvsweb allows you to view your cvs logs & differences. 
We could rewrite and enhance this tool allow for an intuitive query architecture for viewing source & changelogs of python
and it's extensions. This could easily written as a Zope product.

The next idea would be something in python that would make it easy to
manage packages & dependencies. RPM is a example of a good architecture that allows the user to easily install & upgrade his/her tools without losing all of the flexibility of self-compiled packages. If I had a python interface to the redhat rpm's it would do a world of good for trying out new things. 

Notice that there are probably only two  or three installers for windows & theyall know how to deinstall what they install.
If we wrote something up for windows that could recognize these formats,
it would be almost the same as a package manager. (of course you would
need some way of wading through all that microsoft piddle)

I know that linuxconf (redhat) was written mostly in python. (am I wrong?)
what is the status of pythoneers supporting or extending this?
 
On a sidenote one of the best things about python is that most every
module & extension comes with a demo and test facility. This is
very helpful when trying out new things. Thanks :>

,Andrew Hanson


PS. my shift key was in use by my cat.









From skip at mojam.com  Tue Jul 27 13:42:32 1999
From: skip at mojam.com (Skip Montanaro)
Date: Tue, 27 Jul 1999 12:42:32 -0500 (CDT)
Subject: Python classes for Excel Reports?
In-Reply-To: <379DAD65.5FDB5B59@usrconsult.be>
References: <379DAD65.5FDB5B59@usrconsult.be>
Message-ID: <14237.61268.652857.738458@dolphin.mojam.com>

    Christian> I have to set up a system to generate some Reports in Excel.
    Christian> I would like to make it in python, but before writing myself
    Christian> a lot of classes, I wonder if there is already something
    Christian> available in the field...

I'm not aware of any.  If you have to generate spreadsheet files is there
something more universally understood, SYLK format, perhaps?

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From Paul.Prescod at p98.f112.n480.z2.fidonet.org  Thu Jul  1 02:12:18 1999
From: Paul.Prescod at p98.f112.n480.z2.fidonet.org (Paul Prescod)
Date: Thu, 01 Jul 1999 07:12:18 +0100
Subject: Mailbox, tell and read
Message-ID: <00001620@bossar.com.pl>

From: Paul Prescod <paul at prescod.net>

Gordon McMillan wrote:
> 
> How 'bout opening in binary, and munging line endings yourself just
> before returning the message? 

First, are there platforms other than DOS/Windows/Win32 where there is a
difference between text and binary mode? I didn't want to munge things
myself because I'm not sure if what is the "right" munge on the Mac, for
instance. Can I just do a search and replace for CR/LF -> CR on all
platforms, all of the time?

> If you can't use tell(), you can't trust seek(), either.

Good point.

-- 
 Paul Prescod  - ISOGEN Consulting Engineer speaking for only himself
 http://itrc.uwaterloo.ca/~papresco

"I don't want you to describe to me -- not ever -- what you were doing
to that poor boy to make him sound like that; but if you ever do it
again, please cover his mouth with your hand," Grandmother said.
	-- John Irving, "A Prayer for Owen Meany"




From gubitz at netcologne.de  Mon Jul 12 07:45:48 1999
From: gubitz at netcologne.de (Hans Gubitz)
Date: 12 Jul 1999 13:45:48 +0200
Subject: Tk-Problem?
References: <lz1zeej0lk.fsf@netcologne.de> <m3wvw6u56s.fsf@greebo.nodomain.de>
Message-ID: <lz4sjaxf9v.fsf@netcologne.de>

/ Bernhard Herzog <herzog at online.de> wrote:
| > >>> root.mainloop()
| > >>> float("1")
| > Traceback (innermost last):
| >   File "<stdin>", line 1, in ?
| > ValueError: float() literal too large: 1
| 
| I had the same problem on the same system, although this happened with
| pygtk. I finally found that the culprit was the environment variable
| LC_CTYPE which has the value de_DE by default on SuSE 6.1, at least on
| the German version. Setting it to C should help.
| 
| It seems to me that this is really a bug in the C-library.

Thanks. It helps on many problems with Tkinter and Python (on SuSE 6.1).

--
Hans Gubitz
email: gubitz at netcologne.de



From tre17 at pc142.cosc.canterbury.ac.nz  Thu Jul 22 17:50:14 1999
From: tre17 at pc142.cosc.canterbury.ac.nz (Timothy R Evans)
Date: 23 Jul 1999 09:50:14 +1200
Subject: tkinter call backs
References: <01bed202$58879280$e855c3d1@Cyberus.cyberus.ca> <c754siy6dja.fsf@pc142.cosc.canterbury.ac.nz> <3796E463.CFE50301@its.uu.se>
Message-ID: <c75yag8l5gp.fsf@pc142.cosc.canterbury.ac.nz>

Dan Pettersson <dan.pettersson at its.uu.se> writes:

> Timothy R Evans wrote:
> > class Command:
> >     def __init__(self, func, *args, **kw):
> >     ...
> >         apply(self.func, args, kw)
> > 
> > Used as following:
> >
> > Button(text='foo', command=Command(myfunction, myarg))
> > 
> > The args (including keywords) that are passed to the Command class are
> > passed on to myfunction when the callback is activated.
> > 
> > I'm wondering if this should be part of Tkinter in some way?
> 
> You can get the same behaviour by using a lambda and defaults
> 
> Button(text='foo', command=lambda e,x=arg1, y=arg2 : function(e,x,y))
> 
> Apologies if all arguments aren't in the write place my TkInter is a bit
> rusty.
> --
> Dan Pettersson				IT Support, Uppsala University, SWEDEN

What bothers me about this method is readability.  Part of python
appeal is clarity and simplicity, using lambdas like this goes quite a 
way to breaking both these principles.  Sorry if this seems a bit
strong, I just have something personal against using lambda.

--
Tim Evans




From tgfjlc at tiac.net  Thu Jul 29 17:47:26 1999
From: tgfjlc at tiac.net (tgfjlc at tiac.net)
Date: 29 Jul 1999 16:47:26 -0500
Subject: Can you believe he got away with it?  1363
Message-ID: <37a0bdde$1@news.mdc.net>

This is not and ad. It is just me trying to help other people, so I will keep this VERY short and to the point. Anyone who owns a small business like myself should go to this URL http://www.mdc.net/~rick/index2.html

also anyone who is a webmaster looking for extra cash to keep there site going will love it as well
zfpqgvxkeykutydosmomoovfdflhkvcqpqxbhjtwyzxjwmztyozcdcvttivvov




From MHammond at skippinet.com.au  Fri Jul  9 08:36:37 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 9 Jul 1999 22:36:37 +1000
Subject: Python and Paradox
References: <7m2mrj$hes$1@nnrp1.deja.com> <7m2o2b$aq0@dfw-ixnews12.ix.netcom.com> <37852804.54D5D2AD@quasar.ipa.nw.ru>
Message-ID: <7m4q97$dpq$1@m2.c2.telstra-mm.net.au>

Alexander V. Voinov wrote in message <37852804.54D5D2AD at quasar.ipa.nw.ru>...
>Aahz Maruch wrote:


>> >I am looking for a way to access Paradox tables from Python without
>> >using ODBC. Does anyone have any ideas?
>>
>> Nope, from an ex-Paradox tech support engineer.

Yep, from someone who has :-)  Well, Im probably not technically correct,
but if you:
* Install MSAccess
* Install Paradox drivers from the "Value Pack" (on the CD) - I think this
does install an ODBC driver even tho we dont use it directly (hence my "not
technically correct")
* Attach the Paradox tables to an Access database.

You can use the COM extensions to talk directly to the Jet API, which in
turn will manage the paradox tables.  I find the COM Jet interface far
better than ODBC.

Mark.






From tim_one at email.msn.com  Mon Jul 26 02:47:08 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 26 Jul 1999 02:47:08 -0400
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have code.)
In-Reply-To: <p5n1woncm0.fsf@bidra241.bbn.hp.com>
Message-ID: <000d01bed732$ae6761a0$492d2399@tim>

Following up on just a couple points:

[Markus Kohler]
> ...
> IMHO it's [Python] very close to Smalltalk, also Guido probably wouldn't
> agree with me ...
> And It has all essential OO features.

Python is something of an ink blot test that way:  people coming from Scheme
think it's very close to that (and just needs macros and continuations and
lexical closures <wink>); people coming from pure functional languages have
a similar tale (and it just needs non-strict functions and a pile of
combinator notation); people coming from Perl -- well, let's stop before it
gets silly <wink>.

Python takes a lot of ideas from a lot of languages, but it has its own
worldview, and its own corresponding implementation.  It really isn't "like"
anything else at heart, in philosophy or implementation.

I'm looking forward to the day when I drop in on another comp.lang.* group
and read "you know, this language is very close to Python, and all it really
needs is namespaces modeled as explicit dicts" <wink>.

>>> Lately we have had already some examples in this group where loop
>>> overhead was a major factor.

[Michael Hudson]
>> Really? When? Sorry, I don't want to sound disbelieving, but I don't
>> remember that.

[back to Markus]
> It was about reading a file and put the words into some hash table.
> As far as I can remember to loop overhead was a dominating factor.
> And Python was much slower than perl for example.

Ah, that one.  That's not loops.  It's input.  Line-at-a-time input on my
platform is about 3 times faster in Perl than Python; indeed, it's faster in
Perl than in C!  It's not loop mechanisms or even refcounting "to blame"
here (btw, Perl does refcounting too ...).  It's that Perl breaks into the
stdio abstraction, going under the covers in platform + compiler specific
ways, peeking and poking C stdio FILE structs directly.

Very clever, but a lot of work.  Most fgets implementations suck, and Perl
deserves all the credit for not settling for that.  I doubt Python will do
it too, not as a matter of principle, but because nobody capable of it will
make time to do it.

would-rather-have-faster-functions-myself-but-if-wishes-were-bananas-
    we'd-be-swimming-in-mushy-fruit-ly y'rs  - tim





From Sven.Drescher at dlr.de  Wed Jul  7 04:32:36 1999
From: Sven.Drescher at dlr.de (Sven Drescher)
Date: Wed, 07 Jul 1999 10:32:36 +0200
Subject: Look for popen3 example!
Message-ID: <378310A4.5816685B@dlr.de>

Hallo!

I looked for an example to use os.popen3() in Deja-News and on
Python-Homepage, but I didn't find anything.
Can anyone give me an example how I can use popen3?

- I want to have the output from an other process started by my little
supervisor. I'm only interessted in the output and not the process by
himself.

- I also have to kill the child with all the childs of my child. But
that's not possible with execv, I think. There's only the PID from the
child, but the child of the child won't be killed!

Thanx for hints and examples or Web-pages!

Sven






From tim_one at email.msn.com  Tue Jul 13 00:03:41 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 13 Jul 1999 00:03:41 -0400
Subject: idle
In-Reply-To: <378a1221@news.xtml.co.uk>
Message-ID: <000c01becce4$b1799e60$31a02299@tim>

[Nick Trout]
> I can't get breakpoints to work in the version I have, 0.4. I'm running it
> under Windows.
>
> Is this just me?!

I doubt it, Nick.  I don't use debuggers much (& haven't used IDLE's at
all), but you're not the first to mention this.  I installed a recent CVS
version of IDLE for my coworkers a couple weeks ago, and they complain
breakpoints (under Windows) can't be set under that either.  So my guess is
that it's been broken since at least IDLE 0.4.

Luckily, you have the source code <wink>.

not-something-i'm-likely-to-fix-anyway-ly y'rs  - tim





From donn at u.washington.edu  Wed Jul 28 12:38:38 1999
From: donn at u.washington.edu (Donn Cave)
Date: 28 Jul 1999 16:38:38 GMT
Subject: Problem with signals & exceptions...
References: <2BB05D8530326D0C.420888E8906528AF.C6BF325D12E9ED90@lp.airnews.net>
Message-ID: <7nnbme$ns2$1@nntp6.u.washington.edu>

Eric Lee Green <e_l_green at hotmail.com> writes:
| I'm having trouble catching my thrown exceptions under FreeBSD 3.2. I am
| using the Python 1.5.2 that came with FreeBSD 3.2 (well, on disk 1 of
| the <n>-disk set). The following program works properly under Red Hat
| Linux version 6.0 running a self-compiled Python 1.5.2 (in place of the
| 1.5.1 which comes with Red Hat 6.0, which does not properly handle
| signals during socket i/o calls). How it works: it listens to port 6664.
| When it gets a connection, it then sets an alarm handler and alarm and
| then tries to read from the socket. If it does not get data within 10
| seconds, it then times out, at which point the alarm handler throws an
| exception and aborts the socket read.  

I tried this exception-from-handler notion on Digital UNIX and BeOS,
and it does seem to work the way I guess you expect, but I do find it
weird. A signal handler executes outside the normal flow of control,
so it's not obvious to me where its exceptions ought to surface.
Probably it should be expected to work the same way on various platforms
for any given version of Python, but would future versions of Python be
guaranteed to work this way also?

Now I can't really make out what you are seeing on FreeBSD.  In the
example output - is that from FreeBSD? - I think you are seeing a
problem with a change to exceptions in recent versions of Python.
Your "why" object prints "READ_TIMEOUT", but it's not that string,
rather an Exception instance with that string as args[0].  So you
need to test if why.args[0] == READ_TIMEOUT.  I get the same output
from your program, on Digital UNIX.

What if your signal handler simply returns, without raising an exception?
C socket I/O functions should return error status with errno EINTR, which
translates to a Python exception.  So you shouldn't need to raise your
own exception, you'd expect one anyway, because the signal interrupts the
function.  Use global state to disambiguate various possible interrupts,
if necessary.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu
-----------------------------
| The test:
|   In one shell window, type:
|      python testalrm.py
|   in the other shell window type:
|      telnet localhost 6664
| type NOTHING just let it time out...
|
| The output in the first window:
|
| bash-2.03$ python testalarm.py
| Got alarm!
| READ_TIMEOUT
| Traceback (innermost last):
|   File "testalarm.py", line 30, in ?
|     raise IOError, why
| IOError: READ_TIMEOUT
| bash-2.03$ 
|
| The program:
| ----------------------------testalarm.py------------------
|
| # A routine to test the 'alarm' signal handling.
|
| import socket
| import signal
|
| READ_TIMEOUT="READ_TIMEOUT"
|
| def alarmhandler(signum,stackframe):
|     print "Got alarm!"
|     raise IOError,READ_TIMEOUT
|
|
| s=socket.socket(socket.AF_INET,socket.SOCK_STREAM)
| addr=("",6664)
| s.bind(addr)
| s.listen(5)
|
| connection,address=s.accept()
|
| oldhandler=signal.signal(signal.SIGALRM,alarmhandler)
|
| signal.alarm(10)  # wait 10 seconds for an alarm!
| try:
|         str=connection.recv(4096) # get a string
| except IOError, why:
|         print why
|         if why == READ_TIMEOUT :
|                 print "why = read timeout"
|         else:
|                 raise IOError, why
| signal.alarm(0)
| print str
| ----------------end program-------------------------------------



From thamza.nospam at geocities.com  Sat Jul 10 02:10:05 1999
From: thamza.nospam at geocities.com (Tim Hamza)
Date: Sat, 10 Jul 1999 06:10:05 GMT
Subject: idle
Message-ID: <7m9tng$6qg$58@rks1.urz.tu-dresden.de>

Have there been any updates to idle.  I thought I saw something about a 0.5
version, but I haven't been able to find anything at python.org.





From jasbahr at REMOVETHIS.origin.ea.com  Thu Jul 29 17:15:39 1999
From: jasbahr at REMOVETHIS.origin.ea.com (Jason Asbahr)
Date: 29 Jul 1999 16:15:39 -0500
Subject: Python "Interfaces"
Message-ID: <m3n1wf88ec.fsf@death.crash.org>

Jim and folks,

What is the current status of the (Scarecrow) Interfaces proposal?  

Thanks,

Jason




From janssen at parc.xerox.com  Fri Jul 16 21:01:33 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Sat, 17 Jul 1999 01:01:33 GMT
Subject: Calling C code from Python
In-Reply-To: <378F93E3.C4F37BED@online.de>
References: <378F93E3.C4F37BED@online.de>
Message-ID: <YrXxLhAB0KGWNkQ1tT@holmes.parc.xerox.com>

Excerpts from ext.python: 16-Jul-99 Calling C code from Python
meikel at online.de (856*)

> That would probably
> mean that I would have to embed Python into the C-application, make
> the function available to Python (in a magic way?) and then execute a
> script (which has to be a text file on disk) which can call the
> mentioned C-function.

Sounds like the standard Python embedding/extending stuff.  This works
quite well.  You embed Python in your application, then you can use
Python as a scripting language for the application.  Where you need to
call back into C to get various application values, you provide Python
accessors to those values through the standard Python extending
mechanism.  You can also execute Python code from C space quite easily. 
See the document on extending/embedding at http://www.python.org/doc/.

Bill



From tismer at appliedbiometrics.com  Thu Jul  8 11:35:05 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Thu, 8 Jul 1999 15:35:05 GMT
Subject: New Python language abuse: ++x with bytecodehacks
References: <slrn7o9625.tk.jepler@craie.inetnebr.com>
Message-ID: <3784C529.78FD22D2@appliedbiometrics.com>


Jeff Epler wrote:
> 
> I know just everyone has always neede to write
> def f(y):
>         x=0
>         while ++x < y:
>                 print x,
>         print
> 
> in Python.  Well, you always could write it, but it printed the rather
> boring
> 0 0 0 0 0 0 0 ...........................
> which is probably not what you wanted.  Well, presenting some code using
> Michael Hudson's bytecodehacks.  You can still write the above code, but
> now it will print
> 1 2 3 4 5 6 7 8 9
> instead.

Thanks for the big laugh!
This should go into Python's humor pages. :)

but-not-into-prodction-code-please - chris

p.s.: I wondered why Guido didn't optimize a double
unary positive. But he can't, since he has no idea what
this operator does.

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From gherron at aw.sgi.com  Fri Jul 23 17:27:09 1999
From: gherron at aw.sgi.com (Gary Herron)
Date: Fri, 23 Jul 1999 21:27:09 GMT
Subject: Xemacs Python Mode.
References: <199907231949.PAA07568@python.org> <3798CF69.1FC04C23@jps.net>
Message-ID: <3798DE2D.C008286D@aw.sgi.com>

Chris Barker wrote:
> 
> Hi,
> 
> Does anyone know if I can set emacs mode to indent with ONLY tabs? (and
> if so, how?) While I'm at it I'd like the tabs to show up as four spaces
> while in Python mode also.
> 

I have my emacs setup to use one tab per indent level, and to display
tabs as four spaces.  I believe that this is all controlled by the
following three settings found in the customization section of my .emacs
file.


(custom-set-variables
 '(py-indent-offset 4)
 '(tab-width 4)
 '(py-smart-indentation nil)
 ...
)

-- 
Dr. Gary Herron <gherron at aw.sgi.com>
206-287-5616
Alias | Wavefront
1218 3rd Ave, Suite 800, Seattle WA 98101



From eric.kidd at pobox.com  Sat Jul  3 14:02:17 1999
From: eric.kidd at pobox.com (Eric Kidd)
Date: Sat, 3 Jul 1999 14:02:17 -0400
Subject: Curses, Ncurses, Slang, etc?
Message-ID: <19990703140217.D14374@moebius.dartmouth.edu>

I'd like to write a Python application which uses either curses, ncurses,
slang or a similar tty interface library. Unfortunately, the built-in
curses module is hopeless broken on my RedHat 6.0 system (it doesn't export
'addch', for example).

Neither the ncurses module or the pwin module work out of the box on my
Linux system. Are there any well-supported Python modules for
curses/ncurses/slang which do work out of the box?

Thank you for any help you can provide!

Cheers,
Eric



From gmcm at hypernet.com  Wed Jul 21 10:58:01 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 21 Jul 1999 09:58:01 -0500
Subject: [Q] In-Browser technology
In-Reply-To: <FE280F7BBBA37CB8.59F593825701D528.AC56447361430CED@lp.airnews.net>
Message-ID: <1279557217-15543582@hypernet.com>

Cameron Laird wrote:
> In article <932522335.684.43 at news.remarQ.com>,
> Vladik <reason at shadow.net> wrote:
...
> >I do not know what it is supposed to be
> >a plugin or not --- but a technology that
> >allows to develop not-trivial GUI applications
> >that run withing a browser.

To amplify a bit on what Cameron wrote:

 Despite all the hype, the only things you can do reasonably portably 
in browsers (NS & IE) is vanilla JScript / JavaScript and vanilla 
Java applets. Even there you will almost certainly have to detect 
which browser you're in, and code accordingly. You may even have to 
detect which version of which browser. And things get even worse as 
soon as you need to deal with the browser's security. And it's very 
difficult to do anything useful without running up against a security 
rule.

 If you have long term relationships with your clients (so you can 
ask them to install something) and if you only have to deal with a 
limited (the smaller the better) number of browser / versions, you 
can do some neat things. (But, as Cameron points out, why bother with 
a 15 meg browser when they could just install an app - HTML forms are 
easy to write, but don't make particularly good GUIs.)

 This whole noise about browsers-as-thin-clients <choke> was started 
by the Java folks. Well, at this point, according to the same folks, 
applets are "out" and server-side is "in". And they're right - Java 
makes much more sense on the server than on the client. You can 
blame it on the browsers, but the JVM spec pushes all the hard 
work into the JVM implementation, so it should be no surprise that 
the promise was hollow. 

On the server, Zope makes even more sense.

 I found Juice very intriguing when I looked at around 5 yrs ago. 
It's not encouraging that they're still using the same wire-frame 
demo they used back then.

> What's the benefit of the browser?  It sounds
> as though you want something that works over
> a network, and operates something like Java
> and JavaScript, but programmed in a language
> "better" than Java or JavaScript, and with no
> particular server-side component.  Why are we
> talking about Web applications at all?  Is
> there some reason you're not choosing your own
> language and designing your own protocol?  Many
> of us do that sort of work daily.

- Gordon



From donn at u.washington.edu  Tue Jul 20 11:59:17 1999
From: donn at u.washington.edu (Donn Cave)
Date: 20 Jul 1999 15:59:17 GMT
Subject: popen3 not works with xterm???
References: <7n1g3h$pou@news.go.dlr.de>
Message-ID: <7n26cl$136c$1@nntp6.u.washington.edu>

"Sven Drescher" <Sven.Drescher at dlr.de> writes:
| I have a question to the following situation.
| I wrote a little C-programm to test the popen3() class.
| That works fine. I write into the pipe and I can read from the pipe, what my
| programm wants to print to stdout.
| But now I thought it's also possible with the xterm. But nothing happens.
| There was only a comment: "pipe was broken - error", but with my C-programm
| it works errorless.
| What did I wrong???

It's xterm.  If you want to display the output of your program on an
xterm window, xterm wants to run the program.  Xterm sets up its own
simulated terminal device (pseudo-tty, or pty), and runs some program,
usually a shell, with input and output connected to the tty side of
that pty.  And of course it opens a connection to the X11 server to
draw its window.  But it just ignores the input and output units it
was started with.  That's why you get the broken pipe, you can't talk
to it that way.

If you're bound and determined, you might be able to start an xterm
running "cat", reading from a named pipe that you create in a convenient
location (see mkfifo.)

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From jblake at eskimo.com  Wed Jul  7 19:57:17 1999
From: jblake at eskimo.com (Jonathon)
Date: 7 Jul 1999 23:57:17 GMT
Subject: Error Trapping In Python
Message-ID: <7m0pgt$cij$1@eskinews.eskimo.com>

	The script I am writing expects the user to input an
	integer.   Being on the safe side, it reads the raw
	data as a string.   However, to convert that string
	to an integer, is perplexing me.

	string.atoi.(read_data) will give a value error if
	read_data contains something other than a number.

	int ( read_data ) gives a Value Error, if non numerical
	data is included.  <<  Or even if 10.0 is written instead
	of 10  >>

	Somewhere, in that vast documentation of Python, I thought
	I read a solution. Can somebody point me to where I meant
	have read it.  << Searching python.org, for "error trapping"
	and "integer" resulted in a nice essay on mathematical languages.
	"Error trapping" alone served out over 600 examples, none of 
	which, from the synopsis, seemed relevent.

	xan

	jonathon

	

-- 
	I'm still looking for a good book on
		3:	The Recent Unpleasantness
		1:	The War Of Northern Aggression.
		2:	The War of Southern Rebellion.



From wlfraed at ix.netcom.com  Fri Jul 16 00:33:09 1999
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Fri, 16 Jul 1999 04:33:09 GMT
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net> <19990715175320.B20155@toast.internal> <378EABDC.EE9A3A46@Lugoj.Com>
Message-ID: <378eb42c.1475054@nntp.ix.netcom.com>

On Thu, 15 Jul 1999 20:49:48 -0700, James Logajan <JamesL at Lugoj.Com>
declaimed the following in comp.lang.python:


> 
> P.S. If Mary Jeannette can give a few more specific details, post them or
> even e-mail them direct to me and I or others will try to help.

	I'm making a wild guess that the Python interpreter itself was
never invoked... Perhaps a UNIX style file with a bang path for Python
fed to an MS-DOS shell command line -- or worse, the source code is
being typed at an MS-DOS shell command line itself.

--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <



From arcege at shore.net  Wed Jul 14 14:33:20 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Wed, 14 Jul 1999 18:33:20 GMT
Subject: Redirect stdout to file
References: <816010E2456BD111A48700805FBBE2EEFDF47B@ex-quebec-u1.baan.com>
Message-ID: <QJ4j3.1847$6M6.614993@news.shore.net>

Gaetan Corneau <Gaetan_Corneau at baan.com> wrote:
: Hello,

: I'm in Barneveld Holland right now (hey, maybe I'll go visit the famous
: Stichting Mathematisch Centrum ;) and I would like to redirect output from
: the print statement to a text file. I think a saw a message about that some
: time ago, but I can't find it right now.

: Anyone could help?

You can change sys.stdout and/or sys.stderr.

import sys  # do not use from sys import stdout
sys.old_stdout = sys.stdout
sys.stdout = open('foo.txt', 'w')

  -Arcege




From fredrik at pythonware.com  Mon Jul 19 08:29:26 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Jul 1999 12:29:26 GMT
Subject: [Q] Type checking...
References: <7mv000$sun$1@feed.teaser.fr>
Message-ID: <004901bed1e2$57a97510$f29b12c2@secret.pythonware.com>

Olivier Deckmyn <olivier.deckmyn at mail.dotcom.fr> wrote:
> Abstract : [Q]How to check the type of an instance before(after?) sending it
> to a method...

also see:

    http://www.python.org/doc/FAQ.html#4.47
    "4.47 How do I check if an object is an instance
    of a given class or of a subclass of it?"

which also points out that explicit class testing
might indicate bad design...  it's usually more
"pythonish" to just expect a valid object, and
trust the "user" to figure out that what the
AttributeError exceptions mean...

</F>




From fredrik at pythonware.com  Sun Jul 25 17:09:13 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 25 Jul 1999 21:09:13 GMT
Subject: simplebeginnerquestion(**args)
References: <379afff7.131877652@news.mindspring.com>
Message-ID: <001c01bed6e1$f2d93760$f29b12c2@secret.pythonware.com>

using a bogus mail address, "optional (Brian)" wrote:
> I'm trying to figure out how to use the (**args) thing. I can manage
> to pass an arbitrary # of arguments to a function with (*stuff) as a
> tuple, but if I use (**stuff)  to do it like this:
> 
> myfunc(**stuff):
>  for i in stuff.keys():
>   print i 
> 
> myfunc(1,2,3,4)
> 
> the interpreter raises an error saying it expected..0 arguments and
> got..4 if I use (*stuff) (without .keys()...) it works. 
> Shouldn't (**stuff) cause a dictionary filled with myfunc(stuff here)
> to happen and the 'print i' print up the keys?

interesting. what did you expect the "keys" to
be in this case?

> What I don't get is why does it expect..0 arguments? Why doesn't it
> realize that there is a ** there?

reading the manual might help:

    If the form ``**identifier'' is present, it is initialized
    to a new dictionary receiving any excess keyword
    arguments.

if you don't pass the function any keyword arguments
at all, there are obviously no excess keyword arguments
to put in that dictionary...

</F>




From phd at emerald.netskate.ru  Wed Jul 21 08:31:55 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Wed, 21 Jul 1999 16:31:55 +0400 (MSD)
Subject: gadfly RDBMS ???
In-Reply-To: <3795B86E.86229F6@arcus.tel.hr>
Message-ID: <Pine.LNX.4.04.9907211629470.12211-100000@emerald.netskate.ru>

On Wed, 21 Jul 1999, Darko Cvagic wrote:
> What does it means that GADFLY is Relational ?

   Because Gadfly is almost Yet Another SQL Engine (not that I am against
Gadfly or SQL).

> How can I make relations between tables in database?

SELECT t1.bor, t1.belt, t2.der
   FROM table1 t1, table2 t2
WHERE t1.plux = t2.flax

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From wolford at enews.nrl.navy.mil  Wed Jul  7 17:26:36 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Wed, 07 Jul 1999 17:26:36 -0400
Subject: win32 python as a dll?
References: <3783C1A0.9F80D390@enews.nrl.navy.mil>
Message-ID: <3783C60C.E4311395@enews.nrl.navy.mil>

Now I'm a little confused. Python does seem to be a dll (?); I found
python15.dll in winnt/system32 directory. But I can't find python15_d.dll
anywhere. Was that supposed to be installed? I think python15.dll bombs
because it was compiled as optimized using the M$ compiler and that
optimizer has problems (at least w/ O2; I've been pretty successfull w/ O1).
So I want to use python15_d but I can't find the dll.

Any help would be appreciated,
Scott




From pmoran at nas.nasa.gov  Mon Jul 12 16:54:25 1999
From: pmoran at nas.nasa.gov (Patrick J. Moran)
Date: 12 Jul 1999 20:54:25 GMT
Subject: bug with Irix configure
Message-ID: <7mdkm1$4c3$1@sun500.nas.nasa.gov>

Hi,

I've encounted a bug with the configure on Irix.  I'm not sure
where to report it, so to the newsgroup it goes.

The problem is under Irix 6.4 with the MIPSpro 7.2 compiler.
The problem is that configure concludes that size_t, mode_t
and pid_t are not defined in <sys/types.h>, when in reality
they are.  I suspect configure comes to this conclusion because
there are #if's around the definitions, and _MIPS_SZLONG has to
be defined as 32 (or 64) in order for the typedef's to be visible.
I presume _MIPS_SZLONG isn't defined when configure does its
tests.  So configure generates #define's in config.h for the
names it thinks are missing, and the defines cause problems
later when the compiler encounters the typedefs.  In particular,
the define for size_t causes errors later.

My current work-around is to just go in to config.h after configure
has been run and comment out to offending defines.

Pat

------------------------------------------------------------
Patrick J. Moran, Ph.D.      http://www.nas.nasa.gov/~pmoran
MRJ Technology Solutions           NASA Ames Research Center




From jlj at cfdrc.com  Fri Jul  2 13:57:22 1999
From: jlj at cfdrc.com (Lyle Johnson)
Date: Fri, 02 Jul 1999 12:57:22 -0500
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net><3773c3a0.102825445@news.oh.verio.com><xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net><xJRc3.292$Of3.67094@news.shore.net><7laovo$c4s$1@Starbase.NeoSoft.COM><377A24B7.AD6FAA1A@cfdrc.com><3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net><377B711A.E8B65E0B@cfdrc.com><377BC620.F5553FF5@easystreet.com> <377BF583.6846FBF9@quasar.ipa.nw.ru> <NF6f3.6498$_A4.93414@news2.rdc1.on.home.com>
Message-ID: <377CFD82.8EA084A2@cfdrc.com>

FOX's grid widget (called FXTable) is under development and will probably be
usable soon (although I can't comment on how its features will compare with the
same widget in wxWindows). As Mike pointed out, one of the test programs for the
C++ library demonstrates what's there so far. I didn't really intend to work up
an FXTable demo for FXPy until the FXTable was really usable.

I think any kind of rich text (i.e. RTF-compliant) or HTML widget is much
farther off for FOX.

"Mike C. Fletcher" wrote:

> None of the demonstrations I've seen are using a "grid" method.  There is
> one which looks similar, the "header" demo, but that is a set of "list"
> widgets standing next to each other and a sitting below a header widget.  To
> create a real grid, you would need to do some (likely C++) programming.  You
> would probably also want to make such thing allow for editing in-field, and
> have the capability of updating directly from an application table to the
> on-screen representation.  It should be possible to create such a thing in
> Python by dropping data into a matrix of entry fields in response to
> scrolling/moving etc., but a grid object generally wants a high-performance
> solution with all sorts of funky options.
>
> A "rich" text editor is evident in the Fox IDE, FXEditor, a separate project
> from the Core Fox libraries, though I have not yet downloaded that project
> to test it.  Something similar should likely become part of a base Fox
> distribution if it is to become a widely deployed Python GUI.
>
> Fox has a basic GUI editor under development (it is a separate project
> again), though my tests with it so far have been unsuccessful, as the
> application exits silently about every-other time I click on a widget.
>
> Neither tkinter nor Fox has any perceptible "slowness" on my system, so I
> cannot comment in that direction.
>
> I'm continuing my experiments, will continue to report as time permits.
>
> Enjoy,
> Mike
>
> Alexander V. Voinov <avv at quasar.ipa.nw.ru> wrote in message
> news:377BF583.6846FBF9 at quasar.ipa.nw.ru...
> ...
> > Did you look at the demos for the underlying toolkit? Some of them
> presents a
> > 'table' which is probably less rich than that of wxWindows, but it does
> exist.
> >
> > What it really lacks is a 'richtext'/html widget, sort of which is present
> and
> > actively developed in wxWindows. As I understand it's just about to become
> > available in wxPython.
> ...




From avv at quasar.ipa.nw.ru  Thu Jul  8 18:38:37 1999
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Thu, 08 Jul 1999 15:38:37 -0700
Subject: Python for the PalmPilot and Other Devices
References: <37852547.E5AC00F4@ssec.wisc.edu>
Message-ID: <3785286D.E7640CC0@quasar.ipa.nw.ru>

Nick Bower wrote:

> Is there any information, or even interest, in making a scaled-down
> python interpretter for 3Com's PalmPilot series?  A new compact (<100k)
> Micro-Java VM has just been released for it btw.
>
> And more generally speaking, are there any plans for a micro-java-like
> python version?  What is the feasibility of this?

Why not to try just to recompile JPython on that Micro-Java, giving up
everything which is not supported by that Micro.

Alexander





From nascheme at ucalgary.ca  Thu Jul 15 05:29:00 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 15 Jul 1999 09:29:00 GMT
Subject: Getters and Setters
References: <000701bece99$6dbeede0$51a22299@tim>
Message-ID: <7mk9ks$9jm@ds2.acs.ucalgary.ca>

Tim Peters <tim_one at email.msn.com> wrote:
>[Neil Schemenauer]
>> Recently I found myself wishing that Python had automatic getters
>> and setters like CLOS.
>
>Bleech.

Bleech to CLOS or to Lisp or to automatic getters and setters?

>Python goes one better: you can write an explicit method if
>needed, but it doesn't encourage bad practice by automatically
>supplying methods that expose every data attribute, crying "use
>me! abuse me!" <wink>.

Point taken.  However, currently I am writing some simulation
software that has objects with lots of publicly accessible,
simple, attributes (ie. reals).  Only a few need special
treatment.  I'm getting bored writing all these little functions.

>And some later corrected code boosted the factor to 10.

Actually the code had to be corrected a third time.  There is
nothing like showing off your incompetence in front of hundreds
(thousands?) of people. :(

>I'll attach code that gets to the same end, but where the
>automatic methods actually run faster than explicit ones. I
>won't say why, though <wink -- but you'll figure it out>.

Your code is fast and elegant but it causes reference loops.  If
only Python had real GC (and closures and macros and generators
and coroutines and ...).


    Neil (running very quickly and hiding)



From heather at thalamus.wustl.edu  Wed Jul 21 12:31:27 1999
From: heather at thalamus.wustl.edu (Heather A. Drury)
Date: 21 Jul 1999 11:31:27 -0500
Subject: how do I find memory leaks?
Message-ID: <7n4skv$8do@thalamus.wustl.edu>

Hi,

I'm relatively new to python and am using it (and Tkinter, 
VTK, NumPy, numpyio, and some other modules that I've probably
forgotten) to write a volume/slice visualization package.

The program has gotten to a fairly advanced stage but I've run
into what appears to be a major memory leak. I come
from the C programming world and have previously used
Purify to debug memory leaks. Alas, I thought using python/vtk
would take care of all this memory garbage for me (although
I'm probably not doing something I should be). 

Here's what I see on a SGI Indigo2 Max Impact with 448 Meg of memory:

	before I start application	350 meg free
	start application		342 
	read volume 1			284
	re-read volume 1		229	
	re-read volume 1		192	
	re-read volume 1		173	
	re-read volume 1		155	

Each time the user reads a new volume it uses numpyio to read
the volume in and then recreates the slice views. 

Over time as the user continues to use the application, the
computer starts using up swap space and eventually the program 
crashes.  I haven't the faintest idea how to begin to debug this
problem. I'm sure I need to provide some more detail but the
code is about 2000 lines now and I'm not sure what detail
would be useful.

Any hints/suggestions/pointers would be greatly appreciated.

TIA,

heather



From tob at world.std.com  Thu Jul  1 19:55:25 1999
From: tob at world.std.com (Tom Breton)
Date: Thu, 1 Jul 1999 23:55:25 GMT
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserv
Message-ID: <m36744kjrm.fsf@world.std.com>

Paul Duffin <pduffin at mailserver.hursley.ibm.com> writes:

> Please enlighten me as to what my misconception is. I am obviously not
> a Lisp expert and I realise that there are plenty of different flavours
> around but the above code works inside Emacs which says that it is
> based on Common Lisp.
[...]
> > 
> > >         (defun inner (b)
> > >           (+ a b))
> > >
> > >         (defun outer (a)
> > >           (inner 3))
> > >
> > >         (outer 2)

In elisp, all variables are "special", which is to say, dynamically
scoped.  Thus inner can find "a" even tho "a" is declared in outer.
This is not so in Common Lisp.

> > > A simple macro version of setq would convert from
> > >         (setq symbol '(list))
> > > to
> > >         (set 'symbol '(list))
[...]
> > > A macro setq is not interpreted the same way as set is because if it
> > > was an error would occur when the Lisp interpreter tried to get the
> > > value of the variable symbol before symbol was created.

What seems to be the problem?  An error in the timing of symbol
dereferencing isn't it.  Perhaps you are thinking of some other macro
system, not Lisp's.

BTW, In elisp which you refer to, there is no problem with setting a
new symbol, (set 'a-new-symbol some-value), because intern both looks
up a symbol and creates it if it doesn't find it.  

-- 
Tom Breton, http://world.std.com/~tob
Ugh-free Spelling (no "gh") http://world.std.com/~tob/ugh-free.html



From drdmgee at geocities.com  Mon Jul  5 03:57:25 1999
From: drdmgee at geocities.com (David Gee)
Date: Mon, 5 Jul 1999 08:57:25 +0100
Subject: newbie question about tkFileDialog.py
References: <3BE7847CF88DD211893100A0C9D43E1A0259E1ED@hqmail2.clarify.com> <19990702094928.A129582@epa.gov>
Message-ID: <7lpof9$vig@dax.unn.ac.uk>

Then the man pages are incomplete - because it *does* exist, and is covered
in the help file supplied with the Win NT version of Tcl/Tk.

--
David Gee
drdmgee at geocities.com

Randall Hopper wrote in message <19990702094928.A129582 at epa.gov>...
>Dinesh Shanbhag (c):
> |Did not find anything like tk_getOpenFile in the Tcl/Tk man pages for
> |selecting Directories...
> |
> |Am i missing something obvious ?
>
>I don't think so.  I looked for the same and didn't find it either.
>
>Randall
>
>





From flight at mathi.uni-heidelberg.de  Fri Jul  9 05:27:14 1999
From: flight at mathi.uni-heidelberg.de (Gregor Hoffleit)
Date: Fri, 9 Jul 1999 11:27:14 +0200
Subject: PyApache and Apache 1.3.6 anyone ???
In-Reply-To: <3785836D.3BE3EBC7@cs.ucsd.edu>; from Pavel Velikho on Thu, Jul 08, 1999 at 10:06:53PM -0700
References: <19990708221333.B16851@hoffleit.org> <3785836D.3BE3EBC7@cs.ucsd.edu>
Message-ID: <19990709112714.A2615@hoffleit.org>

On Thu, Jul 08, 1999 at 10:06:53PM -0700, Pavel Velikho wrote:
> Did you check the error log? If it is a configuration error it will
> usually keep quiet about
> it and just write it in the log file.
> Hope this helps

Nope, there's no entry at all in the error log for all the silently
dying starting attempts. I don't know how to debug apache with gdb,
therefore I'm comletely lost where to look for hints about the
failure.

	Gregor



From tim_one at email.msn.com  Thu Jul 15 12:00:12 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 15 Jul 1999 12:00:12 -0400
Subject: Getters and Setters
In-Reply-To: <7mk9ks$9jm@ds2.acs.ucalgary.ca>
Message-ID: <000101becedb$1f4c3400$91a02299@tim>

[Neil Schemenauer]
> ...
> Bleech to CLOS or to Lisp or to automatic getters and setters?

Just to automagic attr manglers for class instances.  Those grew out of
earlier gimmicks for modeling C-like structs, where they were more
appropriate.

> ...
> Your code is fast and elegant but it causes reference loops.

?  Not under my Python <wink> -- stick a __del__ method in the test class
that prints something and you'll see that objects go away when you expect.
The function object assigned to self.set/get_XXX contains one ref to
self.__dict__ and another to the attr's string name, and that's it; it is
*not* a bound method object (which would also hold a reference to self),
just a function object.  So there's no cycle unless you force one, like

    instance.setWhatever(instance)

But then you would have had a cycle without this code too.

> only Python had real GC (and closures and macros and generators
> and coroutines and ...).
>
>     Neil (running very quickly and hiding)

It's OK to want everything!  It's just ill-advised to pine for it <wink>.

common-lisp-was-god's-experiment-to-see-whether-"everything"-
    was-what-people-really-wanted-ly y'rs  - tim





From pduffin at mailserver.hursley.ibm.com  Thu Jul  1 06:39:08 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Thu, 01 Jul 1999 11:39:08 +0100
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <lwoghyexy5.fsf@copernico.parades.rm.cnr.it>
Message-ID: <377B454C.6F59@mailserver.hursley.ibm.com>

Marco Antoniotti wrote:
> 
> Paul Duffin <pduffin at mailserver.hursley.ibm.com> writes:
> 
> > Marco Antoniotti wrote:
> > >
> > > Nope. *best* is the Common Lisp Macro system, with Scheme/Dylan syntax
> > > stuff a second best.  The notion of 'upvar' in Tcl makes my head
> >
> > I am surprised at this because Tcl's [upvar] is simply an explicit form
> > of dynamic binding which I seem to remember is what Lisp uses. (That

Mistake, I think I meant dynamic scoping.

> > is how the 'let' function can be (is) implemented as a lambda
> > function).
> 
> Ahem!  LET is essentially a macro built on top of LAMBDA application.

I know that.

> Dynamic binding is another beast altogether, which Common Lisp and
> Scheme (statically scoped languages) allow in through a back door.

If they are statically scoped how does the following work.

Look at the following code.

	(let ((a 2))
	  (let ((b 3))
	    (+ a b)))

When you expand the macro let you get something like.

	((lambda (a)
	   ((lambda (b) 
	      (+ a b)) 3)) 2)

Which is equivalent (apart from side effects) of.

	(defun inner (b)
	  (+ a b))

	(defun outer (a)
	  (inner 3))

	(outer 2)

So what is the process which associates the use of "a" inside inner
with the formal argument of outer. I 

The Tcl equivalent is

	proc inner {b} {
	    upvar 1 a a;
	    expr {$a + $b}
	}

	proc outer {a} {
	    inner 3
	}

	outer 2

> The notion of [upvar] and [uplevel] (with the optional numeric argument which allows
> you to inspect N levels of stack is truly hackish.  The utility of

As I have said before it is simply an explicit form of the Lisp mechanism which allows
the let macro to be implemented using lambda. It is very rare that N is more than 1.

> this construct is not in question.  But its use as a
> 'macro-approximating' device is questionable.
> 
> BTW. Is TCL statically or dynamically scoped?
> 

Tcl is statically scoped but you can use upvar to get access to variables
in enclosing stack frames.

> > > spin. :) Finally, AFAIU, a Tcl "macro" must run as an interpreter of
> > > the spec. A Common Lisp (Scheme) macro is compiled into regular code
> > > by read-time expansion.
> > >
> >
> > I find that Lisp macros (while a very powerful and necessary mechanism)
> > are sooo confusing. They are in essence another language inside Lisp,
> > and as such introduce inconsistencies.
> 
> You are very mistaken.  Common Lisp macros are definitively not
> another language.  They manipulate S-expressions, which are what Lisp

Lisp macros manipulate S-expressions but the macro definitions themselves
are interpreted differently to Lisp expressions.

A simple macro version of setq would convert from
	(setq symbol '(list))
to
	(set 'symbol '(list))

A macro setq is not interpreted the same way as set is because if it
was an error would occur when the Lisp interpreter tried to get the
value of the variable symbol before symbol was created.

This is where the inconsistencies come from.

> is made of.  They generate inconsistencies insofar as you are careless
> when dealing with possible name capture problems. GENSYM and GENTEMP
> are there to help you.
> 
> > Tcl on the other hand doesn't
> > need a macro language and as such is much more consistent than Lisp.
> 
> You severely underestimate the power of Common Lisp (Scheme and Dylan)
> macro systems and ascribe to them inconsistencies which are not

I don't underestimate the power of Lisp macros I am just saying that I
find the way they warp the otherwise simple Lisp syntax / semantics.

Lisp does not really need macros, they are really just a way of
creating short cuts.

> there.  It is true that Tcl does not need a macro system. But mostly
> because, since you have Common Lisp, you do not need Tcl altogether :)

Not a particularly effective argument.

> Apart from some good ole flaming, why doesn't Tcl need a Lisp style
> macro system? Just because you have [eval]?  Not a very good
> argument. Especially when brought up in front of a Lisp audience.
> 

Tcl does not need a macro system because it has a simple consistent
syntax / semantics and all commands are treated equally and it is
possible to change the behaviour of existing commands.

> >
> > If I needed a macro language in Tcl I can just write one.
> 
> I just recieved an email from Cameron Laird, citing a 'procm' form in
> Tcl, which is supposed to do 'scan time' macros - probably something
> in line with the real thing. However, the manual pages for 8.1 at Scriptics
> does not mention it.  Yet, I suppose that it is still an experimental
> feature that maybe will appear in Tcl in a later edition - a few
> lustres after Lisp had them :)
> 

I would say that that is probably something Cameron has created. Despite
what you think Tcl is not crying out for a Lisp like macro system. What
it is crying out is for more data types and it will soon be getting them.
Including a 'proper' [lambda] implementation.

I would say that Tcl (without macro system) can do anything that Lisp
(with macro system) can do.

> > Lisp and Tcl have a lot more in common than Lispers seem to want to
> > acknowledge.
> 
> Ahem!  Do you remember the "revised" Tcl paper by Ousterhout (sorry
> for the spelling mistakes :) ), where the essential
> addition/correction was the mention of the L-word? :)
> 

No I don't.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From phil at ricochet.net  Fri Jul  9 03:18:19 1999
From: phil at ricochet.net (phil)
Date: Fri, 09 Jul 1999 07:18:19 GMT
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com> <37850F5B.F80CB0A7@enews.nrl.navy.mil> <7m3ajc$s3q$1@m2.c2.telstra-mm.net.au>
Message-ID: <7m9tls$6qg$10@rks1.urz.tu-dresden.de>

On Fri, 9 Jul 1999 09:03:03 +1000, "Mark Hammond"
<MHammond at skippinet.com.au> wrote:

>Scott Wolford wrote in message <37850F5B.F80CB0A7 at enews.nrl.navy.mil>...
>
>>msvcrt.lib and the application can't then bring in msvcrtd.lib. I do think
>>this is a problem because python15_d.lib is provided but no python15_d.dll
>>is. I would think you'd provide (at the very least) python.lib linked with
>>msvcrt.lib, and python_d.lib linked with msvcrtd.lib (and the dlls), or
>>(to be really nice :) all four combinations.
>
>The main reason we dont is size - we would need to ship the _d, and for
>people to really use it, also the .pdb.  These add too many megs.

This bit me as well. The confusion i had was that the _d.lib was
available but not the _d.dl. Couldn't there be two downloads, one with
the optimised/non-debug and one without? I rather like the
msdog/widow's way where no compiling is involved and suspect that it
is the future. Easy for me to say, i know..

No problemo either way'ly,

phil.




From robin at jessikat.demon.co.uk  Fri Jul 23 19:24:22 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sat, 24 Jul 1999 00:24:22 +0100
Subject: VC++ 6.0 exe file sizes
References: <o$h7sBAVZij3Ew7i@jessikat.demon.co.uk>
 <37981400.1980005@news.cswnet.com> <7nacj7$ge1$1@newsin-1.starnet.net>
Message-ID: <XiYk7DAmmPm3Ewkl@jessikat.demon.co.uk>

In article <7nacj7$ge1$1 at newsin-1.starnet.net>, Martin Bertolino
<martinb at talx.com.no.spam> writes
....
what size for the distibution pythom.exe  what size for your version?
-- 
Robin Becker



From Aahz.Maruch at p98.f112.n480.z2.fidonet.org  Thu Jul  1 11:12:30 1999
From: Aahz.Maruch at p98.f112.n480.z2.fidonet.org (Aahz Maruch)
Date: Thu, 01 Jul 1999 16:12:30 +0100
Subject: Why must path be in args in os.execv?
Message-ID: <00001617@bossar.com.pl>

From: aahz at netcom.com (Aahz Maruch)

In article <377A2590.D95D72FE at phoenix-edv.netzservice.de>,
Holger Jannsen  <holger at phoenix-edv.netzservice.de> wrote:
>
>there's again something I don't understand. Hope you do.;-)
>
>I'd like to start a windows-executable with some arguments.
>Ok. 'os.execv' or anything like that is the thing.
>
>But if I don't repeat the path with the arguments, the execution fails.
>Better: Something happens, but not the correct execution.

If you look in a standard C-library docs, you'll see the same thing.
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From puetsch at gmx.de  Thu Jul 15 13:35:09 1999
From: puetsch at gmx.de (Felix Puetsch)
Date: 15 Jul 1999 17:35:09 GMT
Subject: Printing the output screen
Message-ID: <7ml64d$tbg$5@chief.death.de>

Using Python 1.5.1 on Windows, I'm having trouble using the results in the 
output window. Copy and Paste doesn't work as there appear strange line 
endings when pasting to e.g. notepad.

Printing this window isn't supported by Python either.

Any idea?

Thanks

-- 
Felix P?tsch



From nick.bower at ssec.wisc.edu  Sat Jul 10 22:02:58 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sat, 10 Jul 1999 21:02:58 -0500
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787D8DA.3774AA7D@ssec.wisc.edu> <7m8p7t$o7s@dfw-ixnews10.ix.netcom.com>
Message-ID: <3787FB51.12FE0048@ssec.wisc.edu>

yeah, that's what i originally thought.  sure enough there were some updates in the
os documentation, but not os.chdir.  it's something that would be really helpfull
if anyone linked to the documentation sees this.

nick

Aahz Maruch wrote:

> In article <3787D8DA.3774AA7D at ssec.wisc.edu>,
> Nick Bower  <nick.bower at ssec.wisc.edu> wrote:
> >
> >Incidentally, I went back to the library reference and still couldn't find
> >os.chdir
>
> You probably have an old version of the docs; I suggest you go to
> www.python.org and download a new set.
> --
>                       --- Aahz (@netcom.com)
>
> Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)

--

Nick Bower
Space Science and Engineering Center
University of Wisconsin, Madison, USA
Phone: (608) 265 8007
Email: nick.bower at ssec.wisc.edu
Web: http://arm1.ssec.wisc.edu/~nickb





From ivanlan at callware.com  Fri Jul 23 11:36:49 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 23 Jul 1999 15:36:49 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
References: <001101bed3f2$5e8874a0$642d2399@tim>
		<7n675p$qsk@dfw-ixnews7.ix.netcom.com>
		<14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>
		<379798D9.FF3F811B@compaq.com> <14232.32203.792051.822513@weyr.cnri.reston.va.us>
Message-ID: <37988C11.DB7D67C7@callware.com>

Hi All--

"Fred L. Drake" wrote:
> 

[bobbitt]

>   Yes.
>   Something that I'd like to see, that I don't think has been
> mentioned for a long time, is some sort of "import ... as" statement.
> I like qualifying the names from a module, but if I'm using them a lot
> I want to give the module a short name.  For instance,
> 
>         import Tkinter as Tk
> 
> would be more comfortable, and is much more clear than:
> 
>         import Tkinter
>         _tk = Tkinter
>         del Tkinter
> 
>   Variants could include:
> 
>         import package.module as pkgmod # pkgmod is the actual
>                                         # module, not the package
> 
>         from package import module as pkgmod
>                                         # the same, possibly cleaner
> 
>         from module import function as func
> 
>   As I said, I haven't noticed this come up for a while, so maybe I'm
> the only one, but I think this would encourage clarity and
> "localizing" the names of imported modules (adding a leading _ so when
> others "import *" from our modules their namespace isn't polluted too
> much).
> 

Now, this I like.  I haven't seen this proposal before (or maybe I have
and senility is setting in), but you can now count me in as a proud
supporter of this idea.

Any reason it couldn't be done?  What bad things would happen if this
were implemented and "from spam import *" went away?

<except-for-import-spam-as-foo-will-silently-dump-core,-nothing>-ly
y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From wyc at mothra.lbl.gov  Wed Jul 28 18:04:48 1999
From: wyc at mothra.lbl.gov (Bill Crutchfield)
Date: Wed, 28 Jul 1999 15:04:48 -0700
Subject: INT16 on a cray t3e
Message-ID: <379F7E80.ABD@mothra.lbl.gov>

The Python Imaging Library (PIL) seems to require 16 bit integers. 
Unfortunately, a cray t3e doesn't have anything shorter than a 32-bit
integer.  

Has anyone solved this problem?
-- 
W. Y. Crutchfield                                   
WYcrutchfield at lbl.gov
Center for Computational Sciences and Engineering    PHONE 510-486-5263
Lawrence Berkeley National Laboratory                FAX   510-486-6900



From spud200 at earthlink.net  Mon Jul 26 19:40:46 1999
From: spud200 at earthlink.net (Spud)
Date: Mon, 26 Jul 1999 16:40:46 -0700
Subject: Question: Python and API?
Message-ID: <7nirq3$ipu$1@oak.prod.itd.earthlink.net>

I was wondering if Python (in Windows 9x) supports API functions. Would
anyone here happen to know?

Thanks a lot!






From r.skowronek at fz-juelich.de  Fri Jul 23 08:47:20 1999
From: r.skowronek at fz-juelich.de (r.skowronek at fz-juelich.de)
Date: 23 Jul 1999 12:47:20 GMT
Subject: how to access ports ?
Message-ID: <7n9o8o$41e@zam201.zam.kfa-juelich.de>

Well, how do I access arbitrary ports?
I managed to read and write to serial ports easy, but I would like to access
an mca card at addresses 0x220 - 0x31f.
I checked the docs but did not see any reference to hardware access - or did
I not look in the right places?

Thanks for any hints.


Rolf




From gmcm at hypernet.com  Wed Jul 21 16:33:36 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 21 Jul 1999 15:33:36 -0500
Subject: pow()
In-Reply-To: <37961668.FDF78B51@callware.com>
Message-ID: <1279537084-16754559@hypernet.com>

Ivan Van Laningham wrote:

> I couldn't find an explanation for the three-argument version of
> pow() on DejaNews, the Python 1.5.2 docs or the Python FAQ, so I'll
> ask now.
> 
> Can someone explain (in simple terms, for a bear of small brain) to
> me what the modulo argument to __pow__ is for?
> 
>  __pow__(self,other[,modulo])

It's for the number of points on the stars that revolve around your 
head. bam() needs no such arg, because birdies all have exactly 2 
wings. socko() is no longer supported; which is probably why 
Universal Studios dropped out of the Python Consortium.

leaving-a-bunch-of-looney-tunes-behind-ly y'rs

- Gordon



From bwinton at tor.dhs.org  Thu Jul 22 23:51:28 1999
From: bwinton at tor.dhs.org (Blake Winton)
Date: Fri, 23 Jul 1999 03:51:28 GMT
Subject: [Tutor] I want to learn to hack
References: <19990721154610.B3663@optiplex.palga.uucp> <Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org> <14231.11726.916541.492153@nobot.cnri.reston.va.us>
Message-ID: <slrn7pfq6v.f9h.bwinton@tor.dhs.org>

On Thu, 22 Jul 1999, little Roger <rmasse at cnri.reston.va.us> wrote:
>> (In other words, Tim was kidding, Roger was kidding, I'm not
>> kidding!).
>Errr, it seems I've inadvertantly 'cracked' opened a can of worms.  I
>really just wanted to know how the Python tutors deal with such
>requests, since they are arriving with alarming frequency to
>webmaster at python.org.

This was the first one I've seen in all my many years as a reader of the
tutor list...  Are you planning on forwarding them to the tutor list,
and wanted to test the waters, or did you just want some ideas on how to
handle them when they came your way?

>I'm not quite sure how or why this became a thread on python-list,

Just lucky, I suppose...

>but it sure makes me feel tingley all over when Tim calls me 'little
>Roger'. <blush, giggle>

More information than I needed, thanks.  ;)

>hacking-is-what-I-do-for-work-cracking-cyphertext-is-just-plain-fun-ly
>y'rs

See-now-if-you-had-included-a-tag-line-like-this,-we-would-have-known-
right-away-ly-y'rs,
Blake.

-- 
One Will.  One Dream.  One Truth.  One Destiny.  One Love.



From m.faassen at vet.uu.nl  Wed Jul 21 05:10:48 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 21 Jul 1999 09:10:48 GMT
Subject: [Q] In-Browser technology
References: <932337314.611.103@news.remarQ.com> <BE2245BA1DCDAE31.8804258C1F60F9A7.2EB4BCD57DBA05D1@lp.airnews.net> <932522335.684.43@news.remarQ.com>
Message-ID: <7n42qo$6ug$2@news.accu.uu.nl>

[this reply only posted to comp.lang.python -- hope Vladik sees it :)]

In comp.lang.python Vladik <reason at shadow.net> wrote:
> OK, I will try to clarify my question.
> With the warning that I may not be speaking
> authoritatively about the web technologies yet:
> I am just starting this research.

Having followed this thread a little, I still am not sure how much 
in browser technology the system you're describing actually _needs_. You
can do a lot with dynamically generated HTML from the server, and some
JavaScript.

Client side action in browsers is hard to do with Python (and Perl and
Eiffel), though I'd wish. Although I must mention jpython here, a version
of Python that runs on top of a Java virtual machine, and can therefore
be used in 'Java' applets (www.jpython.org).

For the server side, where you'd have your business rules and database
functions presumably, I recommend you take a look at Zope. Of course I'm 
biased to Python, and Zope uses Python, but it's really very
very nice. Check it out at: http://www.zope.org

Browser technology is hard to do with Python (and Perl and Eiffel) -- 
browsers don't tend to support these languages. 

Regards,

Martijn




From alan at littleford.net  Sat Jul 10 23:12:54 1999
From: alan at littleford.net (alan at littleford.net)
Date: Sun, 11 Jul 1999 03:12:54 GMT
Subject: Another reason assignment expressions would be nice (?)
References: <000301becb50$8c2045a0$479e2299@tim>
Message-ID: <7m9tof$6qg$90@rks1.urz.tu-dresden.de>

On Sat, 10 Jul 1999 23:50:42 -0400, "Tim Peters"
<tim_one at email.msn.com> wrote:

........
>How about
>
>    a = foo()
>    a.setAttr('i', 'self.SomeFunction()')
>    a.setTemplate("'A test line %d' % self.i")
>    a.setTemplate("'Same value as above %d' % self.i")
>
>?  Then (a) the code is easier to write; and, (b) when it all falls apart
>you can at least find your mutations in an editor via searching for
>"setAttr", or even turn on a debug flag causing setAttr to log what it's
>doing.
>
>different-operations-different-methods-ly y'rs  - tim
>

I'd thought along this line -- I'd even though about implementing a
stack within foo and then you could push, pop+push and pop args to and
fro -- this would do away with having to declare stuff (which the
setAttr() is really still doing) but would lead to the usual stack
management problems (as well as being hellish to understand, as you
intimated). Still and all, having i = 3 return 3 would be the nicest
for this app ..

Tnx
Alanl



From Sue at p98.f112.n480.z2.fidonet.org  Sat Jul  3 18:08:24 1999
From: Sue at p98.f112.n480.z2.fidonet.org (Sue)
Date: Sat, 03 Jul 1999 23:08:24 +0100
Subject: How do you make clickable Python/Win9x scripts?
Message-ID: <000016b7@bossar.com.pl>

From: Sue <smalleys at gte.net>

I have an NT hack - not sure about Win95.  Note that this
does NOT require registry/admin access, as far as I know.

1.  Make sure the .pyc version of your file exists using CLI
"python.exe compileall.py $cwd" or whatever.

2. Create a new filetype (Python Executable), pointing to
your python binary, with suffix .pyc.

3. Add .pyc to CMD_SUFFIXES (not 100% sure)

4. Drop a link to the pyc file on your desktop.

5.  Pass script parameters using standard Win hack.

6. I tried a PIF file but gave up.  Looked like overkill.

Sue.



From tim_one at email.msn.com  Sun Jul 18 04:46:23 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 18 Jul 1999 04:46:23 -0400
Subject: Python and Boehm-Demers GC, I have code.
In-Reply-To: <7mr0n1$dmk@ds2.acs.ucalgary.ca>
Message-ID: <000e01bed0fa$04316520$36a02299@tim>

[Tim, speculates on why pystone/pybench aren't much affected by BDW w/
 refcounting left on]

[Neil Schemenauer]
> What could I use as a better benchmark?  Real applications would
> be best I suppose.

This is a real bind!  People advocating changes in core areas are always
challenged to prove they wouldn't slow things down, but there's no clear way
to do that.  Python is used by so many people for so many things on so many
platforms now, a credible study would be a huge undertaking.

So screw it <wink>.  Find someone with an important program that's leaking
memory and see whether this fixes *that*.  This (not speed) is BDW's real
strength anyway.  Chalk up a few victories there, and selling will be much
easier.

A cautionary tale:  Over the past month I've been poking away at tracking
leaks in IDLE.  After building a capable-enough cycle-finding tool (a
feature-bloated but much faster offshoot of Lars's Plumbo.py), I found
cycles all over the place.  As things turned out, though, hardly any of them
could be considered trash!  The way functions point to their module globals,
which in turn point to imported modules, some of which eventually point into
sys, from whose sys.modules everthing can be reached ... an isolated cycle
was very rare indeed.

It came as a real surprise to me that "real GC" wouldn't have helped.  This
may be a particularly nasty problem for GUI programs, so I'm not too quick
to generalize.  OTOH, don't be too quick to discount it either <wink>.

>> Python doesn't ask anyone to change anything about how they like to
>> handle their memory today, and Guido has pronounced that a Major Feature
>> on multiple occasions.  Vladimir Marangozov's PyMalloc (see his Starship
>> page) caters to that, letting Python use its own malloc without anyone
>> else getting involved.

> Perhaps I misunderstand, but I thought that the PyMem_* functions
> were created for this purpose.  Extension modules can use any
> malloc they like but they can't mix malloc/free/realloc with
> PyMem_*.

No, I misunderstood you.  But now that I don't, it's not worth going back to
explain how I did <wink>.

> ...
> Yes, one platform, no threads.  Thanks a lot Tim.  Here I was
> feeling pretty good about myself.  Now I'm depressed.

Excellent -- if you can persist beyond that, this may yet turn into
something useful instead of Yet Another 48-Hour Abandoned Grand Scheme <0.9
wink>.  Reading the BDW docs and comments closely should give you another
sharp jolt of helpful depression:  this is > a megabyte of code spread over
> 100 files, fighting a gazillion platform peculiarities:  there's nothing
easy about it.

OTOH, they've made oodles of progress over the years. and looks like it's
usable on most major platforms now.  Keep pitching it as a compile-time
config option and I think it's got a chance.

[about Tkinter callbacks]
> <homer>Mmmm, documentation</homer>.  I think I have the problem
> licked now.  I use GC_malloc_uncollectable for these things [ClientData
> structs].

It doesn't matter whether that appeared to work, it only matters whether I
*think* that should work <wink>.  Yup!  That should work.  Nice job, Neil --
that's got to be the minimal hack around this one.

>> btw-keep-in-mind-that-software-doesn't-work<wink>-ly y'rs  - tim

> No kidding.  I tried to compile this stuff under Windows.  What a
> nightmare.  I'm going to have to take some time off and lick my
> wounds.  :)

If it's any consolation, last time I looked hard at BDW was about 6 years
ago, when I gave up after about 40 hours of trying to port it to the KSR
architecture.  This was after the guy in the next office gave up after about
20 hours of trying to stop it dumping core on a SPARC/SunOS box.  It's come
a long way since then!

or-we-could-view-it-as-5x-as-many-lines-to-go-wrong<wink>-ly y'rs  - tim





From billtut at microsoft.com  Fri Jul 16 03:22:01 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Fri, 16 Jul 1999 07:22:01 GMT
Subject: win32api.ShellExecute howto?
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F5F@RED-MSG-50>

You can use 0 as a HWND if you don't have one.

Bill

> From: Jeffrey Kunce [mailto:kuncej at mail.conservation.state.mo.us]
> 
> I'd like to use win32api.ShellExecute to open a 
> data file with its associated application. The help
> file says:
> 
> int = ShellExecute( hwnd, op , file , params , dir , bShow )
> 
> I think I understand all the parameters except:
> 
>    hwnd : int
>    The handle of the parent window.
> 
> If I am not running this from a pythonwin application, what
> do I use for hwnd and where do I get it?
> 
> Thanks.
> 
>   --Jeff Kunce
> 
> 
> 
> 



From phd at emerald.netskate.ru  Mon Jul 26 05:25:37 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Mon, 26 Jul 1999 13:25:37 +0400 (MSD)
Subject: How to confirm if mod_pyapache is working.
In-Reply-To: <379B7F5A.B0843658@freenet.co.uk>
Message-ID: <Pine.LNX.4.04.9907261325000.720-100000@emerald.netskate.ru>

Hi!

   Look for related thread in PyApache mailing list archive.

On Sun, 25 Jul 1999 sreekantk at freenet.co.uk wrote:

> Hi all
> 
> I for the first time compiled, installed etc of Apache 1.3.6 and
> PyApache module. Then configured the AddHandler etc. It is running now.
> However when I move then Python executable in to a hiding place all the
> cgi files fail . I thought the python has been embedded in to
> mod_pyapache and it only needs library files but not python it self.
> Well I may be wrong.
> 
> Anyway is there a way I can confirm that the .py cgi files are going
> through the pyapache module rather than
> regular python interpreter.
> 
> Any advice greatly appreciated.
> 
> Thanks
> sreekant

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From robin at jessikat.demon.co.uk  Wed Jul  7 08:45:03 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Wed, 7 Jul 1999 13:45:03 +0100
Subject: interaction between modules... globals?
References: <19990707100532.A951@optiplex.palga.uucp>
Message-ID: <eJEWxAAPv0g3EwH6@jessikat.demon.co.uk>

In article <19990707100532.A951 at optiplex.palga.uucp>, Gerrit Holl
<gerrit.holl at pobox.com> writes
>Hello,
>
>I have a program which consists of several little modules, but I encounter a
>problem with that. I have one module which returns a class instance, with
>messages as members, either Dutch or English. Every module needs that instance,
>but I don't want to pass it as an argument to every little function, the
>modules should find out theirselves. But I _also_ don't want to make that
>instance 5 times.
>
>So I thougt: I need a sort of global between the modules. Is this possible?
>
...
>regards,
>Gerrit.
>
I guess you need a singleton class. There was a c.l.p thread about this
recently.

I suggest something of the form

_LangGlobal=None

def getMessages():
        global _LangGlobal
        if _LangGlobal is None:
                # calculate which language
                #...
                #...
                _LangGlobal=result
        return _LangGlobal

-- 
Robin Becker



From kuncej at mail.conservation.state.mo.us  Thu Jul 15 11:11:27 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Thu, 15 Jul 1999 10:11:27 -0500
Subject: win32api.ShellExecute howto?
Message-ID: <s78db3e3.018@mail.conservation.state.mo.us>

I'd like to use win32api.ShellExecute to open a 
data file with its associated application. The help
file says:

int = ShellExecute( hwnd, op , file , params , dir , bShow )

I think I understand all the parameters except:

   hwnd : int
   The handle of the parent window.

If I am not running this from a pythonwin application, what
do I use for hwnd and where do I get it?

Thanks.

  --Jeff Kunce




From wtanksle at dolphin.openprojects.net  Sun Jul 25 21:48:38 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Mon, 26 Jul 1999 01:48:38 GMT
Subject: simplebeginnerquestion(**args)
References: <379afff7.131877652@news.mindspring.com>
Message-ID: <slrn7pnfjm.oso.wtanksle@dolphin.openprojects.net>

On Sun, 25 Jul 1999 12:47:46 GMT, Brian wrote:

>I'm trying to figure out how to use the (**args) thing. I can manage
>to pass an arbitrary # of arguments to a function with (*stuff) as a
>tuple, but if I use (**stuff)  to do it like this:

>myfunc(**stuff):
> for i in stuff.keys():
>  print i 

This is 100% right (whoops, I just noticed the missing 'def').

>myfunc(1,2,3,4)

Whoops!  This is wrong.  Why?  Because you're calling a **argument as
though it were a *argument.  Watch this:

def myfunc(*stuff, **more):
  print "stuff ", stuff
  print "more ", more

You see?  The language has to fill some arguments into the tuple stuff,
and other arguments into the dictionary more.  Any arguments given by
themselves will be placed in order into stuff; arguments assigned to names
will be placed in the more dictionary.

>>> myfunc(1,2,3,4)
stuff (1,2,3,4)
more {}

>>> myfunc(x=3, duh="no")
stuff ()
more {x:3, duh:'no'}

>the interpreter raises an error saying it expected..0 arguments and
>got..4	if I use (*stuff) (without .keys()...) it works. 
>Shouldn't (**stuff) cause a dictionary filled with myfunc(stuff here)
>to happen and the 'print i' print up the keys?
>What I don't get is why does it expect..0 arguments?Why doesn't it
>realize that there is a ** there?

-- 
-William "Billy" Tanksley



From victor at idaccr.org  Fri Jul  9 16:12:31 1999
From: victor at idaccr.org (Victor S. Miller)
Date: 09 Jul 1999 16:12:31 -0400
Subject: _tkinter problem
Message-ID: <7m9tn0$6qg$41@rks1.urz.tu-dresden.de>

I had successfully built Python 1.5.2 with _tkinter using the system
installed versions of tk and tcl (4.1 and 7.5 respectively).  However,
I decided to download the new versions of tcl and tk (8.1 for both),
and build them.  They both built without problems, and ran through the
test sucessfully.  I then rebuilt Python with them (I'm running on
Solaris 2.6), which appeared to build successfully, however, when I
try Tkinter._test(), I get:

Python 1.5.2 (#5, Jul  9 1999, 15:36:18)  [GCC 2.7.2.2] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import _tkinter
>>> import Tkinter
>>> Tkinter._test()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/local/coop/lib/python1.5/lib-tk/Tkinter.py", line 1947, in _test
    root = Tk()
  File "/usr/local/coop/lib/python1.5/lib-tk/Tkinter.py", line 886, in __init__
    self.tk = _tkinter.create(screenName, baseName, className)
TclError: Can't find a usable init.tcl in the following directories: 
    



This probably means that Tcl wasn't installed properly.

>>> 


Now there's a perfectly good init.tcl in the directory
/usr/local/coop/lib/tcl8.1, which is where make install placed it.
Who is actually complaining about this, and how can I fix it?
-- 
Victor S. Miller     | " ... Meanwhile, those of us who can compute can hardly
victor at idaccr.org    | be expected to keep writing papers saying 'I can do the
CCR, Princeton, NJ   | following useless calculation in 2 seconds', and indeed
    08540 USA        | what editor would publish them?"  -- Oliver Atkin



From chris at frostnet.advicom.net  Fri Jul 23 18:19:21 1999
From: chris at frostnet.advicom.net (Chris Frost)
Date: Fri, 23 Jul 1999 22:19:21 GMT
Subject: Converting itegers to strings
Message-ID: <19990723171921.A5205@pooh.frostnet.net>

How can you convert numbers (an interger in this case) to strings? (i.e. I
have some_number = 3, and would like to use it as a string)

thanks,
-- 
Chris Frost  |  <http://www.frostnet.advicom.net/chris/>
-------------+------------------------------------------
Public PGP Key:
   Email chris at frostnet.advicom.net with the subject "retrieve pgp key"
   or visit <http://www.frostnet.advicom.net/chris/about/pgp_key.phtml>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 256 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19990723/143795dd/attachment.sig>

From Stefan.Schwarzer at tu-clausthal.de  Wed Jul 14 05:17:29 1999
From: Stefan.Schwarzer at tu-clausthal.de (Stefan Schwarzer)
Date: Wed, 14 Jul 1999 10:17:29 +0100
Subject: OS/2 Python: editing in interactive use
Message-ID: <378C55A9.411BE9AB@tu-clausthal.de>

Hi there :)

It seems to me that the OS/2 port of python lacks the
"sophisticated" features for input editing (history,
C-A, C-E, etc.)

If I have overlooked a possibility to activate it, please
give me a hint. If not, how difficult would it be to extend
Python/2 appropriately? I would appreciate this very much. :)

Regards
 Stefan



From rboylan at my-deja.com  Mon Jul 19 15:13:07 1999
From: rboylan at my-deja.com (Ross Boylan)
Date: Mon, 19 Jul 1999 19:13:07 GMT
Subject: Installation Aborted! on NT
Message-ID: <7mvtbm$55r$1@nnrp1.deja.com>

I've run into a strange problem I thought I'd report, in case it might
affect other users.  When I try to install py152.exe or win32all-
125.exe on my system from my account, I get the blue installation
background and then a box with the title "Installation Aborted!" and
the message "Internal Error"

The same install works fine from the Administrator account.

My account is a member of the Administrators group.

The system in NT 4.0 WS, SP5.  Software is minimal, including MSIE 5.0,
Exceed 6.1 (X window system), 3Com Nic software, Matrox drivers, and
Lexmark print utility.  There was a password authentication system on
the machine (Ascend AppServNT v2.0Ci1), but we removed it.  40 bit
security.

Since I've been able to do the install, I don't need a solution.  I'm
curious what might be the problem, though.

P.S. The Tcl installer puts its program group items on the user's menu,
not everyone's.  It would be nice if it put itself under "All Users" as
Python does.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Brian at digicool.com  Thu Jul  8 13:34:23 1999
From: Brian at digicool.com (Brian Lloyd)
Date: Thu, 8 Jul 1999 17:34:23 GMT
Subject: PyImport_ImportModule, PyModule_GetDict
Message-ID: <613145F79272D211914B0020AFF6401914DBBE@gandalf.digicool.com>

> If I understand this correctly, the equivalent of doing this:
> 
> >>> import mymodule
> >>> dir(mymodule)
> 
> is sort of this, in the C API:
> 
> PyObject *mod = PyImport_ImportModule("mymodule");
>  PyObject *myDict = PyModule_GetDict(mod);
> 
> Q:  what is the C API equivalent of this:
> 
> >>> from mymodule import *
> >>> dir()
> 
> 
> --Randy

Ok, I'll bite - though I'll note that the 'from module import *'
form is often considered evil :^) Here's an offhand and untested
way you might try to do it:


int fromModuleImportStar(PyObject *modsrc, PyObject *modto) {
  PyObject *srcdict=PyModule_GetDict(modsrc);
  PyObject *tgtdict=PyModule_GetDict(modtgt);
  PyObject *keys=PyDict_Keys(modsrc);
  PyObject *key=NULL;
  PyObject *val=NULL;
  int size=PyList_Size(keys);
  int n=0;

  while (n < size) {

    /* I _think_ that all the references are borrowed here,
       so we don't need to INCREF/DECREF...
    */
    key=PyList_GetItem(keys, n);

    /* should check this for errors */
    PyDict_SetItem(tgtdict, key, PyDict_GetItem(srcdict, key));

    n++;
  }
  PY_DECREF(srcdict);
  PY_DECREF(tgtdict);
  PY_DECREF(keys);
  return 0;
}

Hope this helps!

Brian Lloyd        brian at digicool.com
Software Engineer  540.371.6909              
Digital Creations  http://www.digicool.com 





From Andrew.Cooke at p98.f112.n480.z2.fidonet.org  Sun Jul  4 06:23:44 1999
From: Andrew.Cooke at p98.f112.n480.z2.fidonet.org (Andrew Cooke)
Date: Sun, 04 Jul 1999 11:23:44 +0100
Subject: Dumb beginner question; what is IDLE?
Message-ID: <000016bd@bossar.com.pl>

From: Andrew Cooke <andrew at andrewcooke.free-online.co.uk>

In article <377D8BFB.3700379 at afn.org>,
  Andy L Krietemeyer <afn08995 at afn.org> wrote:
> I've seen several references to IDLE.  Is this an IDE for python? Or
> what is it?  Thanks,  Andy.
>

Yup.  It has simple editing and debug support but no GUI builer or
"wizards".  A kind of poor-mans emacs :-) although I must admit that
I use it rather than emacs for Python code.  I'm not sure why -
I tried it and never changed back.  The only thing I miss is not
having several windows on one piece of code (maybe this is possible,
I've never been able to do it).    Also, the key bindings are a
compromise between emacs/Unix and Dev Studio/MS, so if you use all
three you can type the wrong thing at times....

It appears to be written in Tkinter, but has a very nice file browser,
which I keep intending to track down, but haven't, yet - but which
you can presumably use in other Tkinter apps.

Andrew


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Farrell.Kenimer at p98.f112.n480.z2.fidonet.org  Fri Jul  2 12:43:36 1999
From: Farrell.Kenimer at p98.f112.n480.z2.fidonet.org (Farrell Kenimer)
Date: Fri, 02 Jul 1999 17:43:36 +0100
Subject: Tkinter just WON'T work right!
Message-ID: <0000167b@bossar.com.pl>

From: Farrell Kenimer <phxken at netzone.com>


--------------EFFFA832FCF37C205A7994BB
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

joakal at my-deja.com wrote:

> Sorry to be so inspecific about my subject =) Also, I'm sure this
> question is asked frequently, but I couldn't find what I was looking for
> in any of the 1400 or so messages here (too many of them start with
> ``Help!'' or ``Python Newbie''). Anyway, here it is:
>
> I tried to use the Grail browser under Slackware 4.0 and Python 1.5.1. I
> got the error saying it's not configured for Tkinter, so I went into my
> /usr/lib/python1.5/config/Setup file and uncommented the appropriate
> stuff.
>
> I tried again, same problem. I read a bit more, and it said I have to
> rebuild Python to add Tkinter (don't remember having to do that before).
> So, argh, I tried to do that. Now I get an error when I use 'make'
> saying there's an error when it gets to sedscript.
>
> Can anyone help?
> Thanks in advance!
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.

Have just gotten grail.py to work on my COL2.2 Linux installation.
Spent hours trying to remake python for tk/tcl and then found out
that for COL2.2, it was already done.  tk8.0 and tc.8.0 are required.
grail package has two errors or had at the time I got it 2 weeks back.
Start python and see if you can import Tkinter.  We can talk in
e-mail as I am not a programmer and am only interested in grail
as a browser for another OS.

--
Farrell Kenimer of Phoenix, Arizona  USA



--------------EFFFA832FCF37C205A7994BB
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
joakal at my-deja.com wrote:
<blockquote TYPE=CITE>Sorry to be so inspecific about my subject =) Also,
I'm sure this
<br>question is asked frequently, but I couldn't find what I was looking
for
<br>in any of the 1400 or so messages here (too many of them start with
<br>``Help!'' or ``Python Newbie''). Anyway, here it is:
<p>I tried to use the Grail browser under Slackware 4.0 and Python 1.5.1.
I
<br>got the error saying it's not configured for Tkinter, so I went into
my
<br>/usr/lib/python1.5/config/Setup file and uncommented the appropriate
<br>stuff.
<p>I tried again, same problem. I read a bit more, and it said I have to
<br>rebuild Python to add Tkinter (don't remember having to do that before).
<br>So, argh, I tried to do that. Now I get an error when I use 'make'
<br>saying there's an error when it gets to sedscript.
<p>Can anyone help?
<br>Thanks in advance!
<p>Sent via Deja.com <a href="http://www.deja.com/">http://www.deja.com/</a>
<br>Share what you know. Learn what you don't.</blockquote>
Have just gotten grail.py to work on my COL2.2 Linux installation.
<br>Spent hours trying to remake python for tk/tcl and then found out
<br>that for COL2.2, it was already done.&nbsp; tk8.0 and tc.8.0 are required.
<br>grail package has two errors or had at the time I got it 2 weeks back.
<br>Start python and see if you can import Tkinter.&nbsp; We can talk in
<br>e-mail as I&nbsp;am not a programmer and am only interested in grail
<br>as a browser for another OS.
<pre>--&nbsp;
Farrell Kenimer of Phoenix, Arizona&nbsp; USA</pre>
&nbsp;</html>

--------------EFFFA832FCF37C205A7994BB--



From da at ski.org  Mon Jul 26 18:17:16 1999
From: da at ski.org (David Ascher)
Date: Mon, 26 Jul 1999 15:17:16 -0700 (Pacific Daylight Time)
Subject: MGL graphics wrappers -- anyone interested?
In-Reply-To: <002a01be8508$49213c20$22990281@wam.umd.edu.bailey>
Message-ID: <Pine.WNT.4.04.9907211345000.285-100000@rigoletto.ski.org>

I've restarted work on my Python wrappers around the MGL (megagraphics
game library, from scitech software -- www.scitechsoft.com), which
provides a fast graphics library with all kinds of nice features.  I've
wrapped version 4.05 and only tested it on Windows NT, but version 4.5
(reportedly out soon) will work on a variety of platforms.

I use the library for psychophysics (if you don't know what that is, you
don't care =), but I expect that it would be of interest as well to folks
who want to write games in Python -- MGL's fullscreen support is very
complete, it interfaces to OpenGL as well as has a 2-d graphics library,
sprite and event manager, etc.

My current needs are met by the stuff I've got working, but I have no
doubt that the wrapping is incomplete and imperfect.  If you'd be
interested in helping flesh it out (and test it with 4.5 when that comes
out, and/or on e.g. Linux, etc.) and know or are willing to learn SWIG,
let me know and I'll setup a CVS site so that concurrent development can
proceed.

--david ascher









From fredrik at pythonware.com  Thu Jul 22 06:41:32 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Jul 1999 10:41:32 GMT
Subject: Fw: ANNOUNCE: Mailing list for Australian Python Users
Message-ID: <007201bed42e$c4316c30$f29b12c2@secret.pythonware.com>

> I have (well, Chris did - thanks!) created a mailing list for Australian
> users of the Python language.  This mailing list should initially be a forum
> for anything uniquely of interest to Aussie Python users.  First order of
> business should be a contest for which city can get the first informal
> Python users meeting together - I think that Melbourne should be able to get
> something happening pretty soon :-)

python-se, n?gon?  Link?ping?  :-)

</F>




From avv at quasar.ipa.nw.ru  Thu Jul  8 18:28:23 1999
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Thu, 08 Jul 1999 15:28:23 -0700
Subject: Python and Fortran code
References: <7m23i3$ktr$1@bignews.shef.ac.uk> <3784D688.A6541F39@enews.nrl.navy.mil> <3785120E.A476F208@starship.python.net>
Message-ID: <37852607.BD16D44F@quasar.ipa.nw.ru>

Berthold H?llmann wrote:

> > Look at a combination of the SWIG tool (I can't remember where it's at but it
> > stands for Simplified Wrapper Interface Generator so you can search for it) and
> > a C wrapper for you fortran function. Basically, make it so you can call the
> > fortran from C. Then swig the C function and link the cstuff.o and fstuff.o with
> > cstuff_wrap.o (from SWIG) into a shared library then you can follow choice 2
> > from the other response to this post.
>
> For the Fortran wrapping you might take a look at cfortran.h avaible
> from http://www-zeus.desy.de/~burow/ which is helpfull if you want
> making the wrapping platform independent.

But if you are interested mainly in g77, you may take a look at a SWIG 'library'
fortran.i, found at:
 ftp://starship.python.net/pub/crew/avv/fortran_i-0.0.1.tgz
It's use is somewhat sophisticated, but for Fortran it does more than it would for C.
You may take a look at the example, supplied with the package to get an idea how to
use it and what would be the benefit.

Alexander





From mwh21 at cam.ac.uk  Thu Jul  1 03:14:13 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Jul 1999 08:14:13 +0100
Subject: strange behaviour at termination time
References: <377765F0.38230773@steria.fr> <37783685.C7BB5B2E@compaq.com> <377ABCF3.3C222F37@cup.hp.com>
Message-ID: <m3iu84depm.fsf@atrus.jesus.cam.ac.uk>

Ovidiu Predescu <ovidiu at cup.hp.com> writes:

> Greg Ewing wrote:
> 
> > To avoid problems like this, you have to design
> > your __del__ methods so that they can do their
> > work without having to refer to any module-level
> > names at the time they are called. One way is
> > to use the default-argument trick:
> > 
> > class B (A) :
> >     def __del__ (self, A = A) :
> >         A.__del__ (self)
> 
> Can you explain why this trick works? When are the default arguments
> stored in the method instance? I guess this happens when the method
> instance for __del__ is created by the interpreter, right?

Yup, that's it. If you want to find out about these things, read the
language reference or import rlcompleter and bash the completion key a
lot. I this case you'll find

B.__del__.im_func.func_defaults

which will contain something like

<class A at xxxx in xxx>

HTH
Michael
 
> Thanks,
> -- 
> Ovidiu Predescu <ovidiu at cup.hp.com>
> http://www.geocities.com/SiliconValley/Monitor/7464/



From mumac at wiu.edu  Mon Jul 19 10:02:33 1999
From: mumac at wiu.edu (Mark Clark)
Date: Mon, 19 Jul 1999 09:02:33 -0500
Subject: How to send text to a Windows program
Message-ID: <37932FF9.47E8FA2D@wiu.edu>

	I am starting a Windows program with ShellExecute.  I would like to
send some text to this program, similar to the way you would if you used
"SendKeys" in other programming languages.  Can someone give me an
example of how this would be done?
	Thanks for your time.



From gherman at my-deja.com  Thu Jul  1 07:27:45 1999
From: gherman at my-deja.com (Dinu C. Gherman)
Date: Thu, 01 Jul 1999 11:27:45 GMT
Subject: How to obtain an instance's name at runtime?
References: <7kqpu8$7u1$1@nnrp1.deja.com> <3771127C.E40F9BF3@home.com> <7kvd8b$ueb$1@nnrp1.deja.com> <37769B92.47EBE99C@compaq.com>
Message-ID: <7lfjb9$aun$1@nnrp1.deja.com>

In article <37769B92.47EBE99C at compaq.com>,
  Greg Ewing <greg.ewing at compaq.com> wrote:
>
> The best solution by far would be to give the
> object a name explicitly when you create it.
> Is this possible, or do you want to retrofit
> it onto existing code?

The latter is what I'm planning to do.

> It looks like you want it to be something like
> "the name of the first variable to be assigned
> a reference to the object". Does this apply to
> local variables as well as globals? Instance
> variables of another object? What if the object
> is never assigned to a variable? What if it
> is, but somewhere far removed from where it
> is created?

You're guessing very well... ;-) Actually, I will
postpone some issues until some advance is made at
other places of the code.

Just that much: an object, that has not been assigned
a name to, cannot exist (having a reference count of
zero), can it (at least for user-defined objects)?

In any case thanks so far! I think the problem is not
a really fundamental one for the type of application
I'm having in mind, after all.

Bye,

Dinu





Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From jepler at inetnebr.com  Thu Jul  1 11:26:52 1999
From: jepler at inetnebr.com (Jeff Epler)
Date: Thu, 01 Jul 1999 15:26:52 GMT
Subject: Why must path be in args in os.execv?
References: <377A2590.D95D72FE@phoenix-edv.netzservice.de>
Message-ID: <slrn7nn2bb.81d.jepler@craie.inetnebr.com>

Holger,

On Unix, read the manpage for execve and friends.  You will see that the
first argument is the path to the executable, and the subsequent arguments
describe the arguments (including argv[0]) and optionally the environment.
By convention, argv[0] is the basename of the executable when the
executable is on the PATH, and the full path otherwise.  (Another unix
convention is that when the first letter of argv[0] is "-", the shell is a
login shell)

DOS doesn't really have execve.  It's possible that the imitation is less
perfect than it could be.  Instead, it may do the equivalent of entering
the string
	string.join(argv, " ")
in the shell, i.e., ignoring the path argument.

This looks like a shortcoming in the DOS (windows 9x, NT) implementation of
exec*, in a way that makes it work differently from the Unix it attempts to
imitate.  This wouldn't surprise me, since I recently discovered that NT's
spawnl takes a unix-like array of arguments (const char *argv[]) but when
any element contains a space, it is split into two arguments in the child
process.  So, I had to break some code that had worked fine on Unix to
generate argv, by quoting each argument (something that would *not* work on
Unix) ..

Bad enough that Windows wants to copy Unix, they have to consistently get
it subtly wrong too.

Jeff



From mgushee at havenrock.com  Wed Jul 21 18:28:16 1999
From: mgushee at havenrock.com (Matt Gushee)
Date: 22 Jul 1999 07:28:16 +0900
Subject: TEST -- please ignore
Message-ID: <m3wvvtskn3.fsf@havenrock.com>

Having trouble posting -- will this make it?

Matt



From tismer at appliedbiometrics.com  Mon Jul 19 14:13:56 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Mon, 19 Jul 1999 20:13:56 +0200
Subject: How fast can we multiply?
References: <000201bed181$a5bf0be0$4ba02299@tim>
Message-ID: <37936AE4.810F2152@appliedbiometrics.com>


Tim Peters wrote:
> 
> [Christian Tismer, hiding from continuations <wink>]

Oh, just a break.

> > Version 0.2 is now 4 times faster at 100000 bits.
> 
> Factor of 5.7 on my machine.

And the cutoff should better be set to 5000, again better.

> > It is a well known fact that by some factorization, integer
> > multiplication's complexity can be brought from O(n**2) down to
> > O(n**1.5).
> 
> The exponent is actually log2(3) ~= 1.58.

I dumbhead thought of 4 mults -> 3 and just said 1.5 - but yes,
sure, sorry.

[Karatsuba, GMP,...]

> Right, GMP does do this.  There's a much faster method (O(n)) based on
> Fourier transforms, but it's much harder to code and I've only seen it used
> in David Bailey's MPFUN package (a highly optimized Fortran pkg for
> arbitrary-but-fixed precision float arithmetic + the std elementary
> functions).

Can you give me a pointer? What I know is O(n log n loglog n),
based on convolutions, but I may be 20 years behind state of the art.

[we agree that coding in Python leads to etter solutions]

> > ...
> > A function to obtain the bit length of a long would
> > be most helpful.
> 
> The lack of this has gotten unbearable, eh?

This is my old fault. If there is a good algorithm with a better Oh,
that's one thing. But then it hurts so much to see an unnecessary
constant factor left over. The counter (number of words) is even
part of the long number and cries to be used.

> > Some kind of split function as well.
> 
> That one is much weaker; unlike bit length, the obvious ways to split and
> join already have the right O() behavior.

Sure. I should stop thinking about it. Just so tempting
to have a functon to break a long into two peaces, at
a word boundary, no shifts... sigh. No, I resist [now].

> 
> > ...
> > (1) Aho/Hopcroft/Ullmann:
> >     "The Design and Analysis of Computer Algorithms"
> 
> Knuth Vol 2 is a better reference for fast integer mult, although it goes
> off the deep end on theoretical refinements of no practical value ("what if
> the number of bits is bigger than the number of electrons in the
> universe?").

I know, also the subject "how..." is from there. But they re-arranged
my office. I found the Knuth corner again five minutes ago.
Yes, there is also the O(n) thing, and I'm not so behind.

> >         if not val1 and val2:

> The test only succeeds when val1 == 0 and val2 != 0 -- not what you
> intended.  OTOH, it's debatable whether programs multiply by 0 often enough
> to pay back the cost of testing for it!

How could I forget the braces. Well, the "why" was in fact to save
time when my "max" is a bad guess.

...
> Need a fancier strategy when nbits(val1) and nbits(val2) are far apart.
> This strategy takes strong advantage of the special-case-for-0 test above
> <wink>, but can easily run 100s of times slower than the builtin mult; e.g.,
> change your test case to lambda x:x*5.

I was undecided what's best. Obviously you are right. The asymmetric
case has a small and a large side, it already doesn't suffer from
the quadratic "flesh". I see it.

Also I was tempted to use "&" on the numbers, in order to save
one nbits call. It was just unclear what happens with negative
numbers and dropped it. Was happy that negatives went correct without
special treatment.

> This is all the more reason to write in Python, though!  Much clumsier to
> handle imbalance in C.  In the meantime, using "min" instead of "max" to
> compute bitcount greatly speeds the worst cases, and doesn't hurt the best
> ones.

> > def nbits(x):
> >     """not clean, rather fast, inaccurate"""
> >     s=marshal.dumps(x)
> >     return abs(struct.unpack("i", s[1:5])[0])*15 -7

Yes it is ugly. I just needed to lower the break-even point,
since I was impatient for the result. :-)

[nbits with hex...]

brrr. I can't help it. This should really be a quick log2()
call, this seems quite natural to me.

> >       time=time.time
> 
> Not time.clock?

One of my very first scripts. Don't ask me, perhaps an old
Python/Win32 problem.

> > # the end ----------------------------------------------------
> 
> looks-more-like-the-beginning-to-me<wink>-ly y'rs  - tim

It was impressive for me to see that with an hour of
coding, I can outperform C in Python. (Actually is was some
more hours, since I didn't remember how, and undusted some books).
My intent was to encourage people to play with stuff like
this, and to see how it can continue. The last word btw. keeps
me busy enough :)

At the same time, the reacton of the list was a little
disappointing to me. There seems to be near to zero interest.
Is this the hot summer? Holidays? Bad title of my post?

Then I also had a look into NumPy and its matrix multiplication.
I stunned. Again no attempt to optimize big O. It could be done
with Python, the algorithms are known, the fruit is waiting
to be picked.

Challenge: I claim Python could be a competitor to Mathematica,
if only some people were interested.

*** Hey Python, where are your ambitioned Math students? ***

(did they wake up, finally?)

waiting-for-the-beginning-to-show-up-ly y'rs  - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From robin at jessikat.demon.co.uk  Thu Jul  8 17:05:28 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 8 Jul 1999 22:05:28 +0100
Subject: scoping assertion
References: <37839491.3F7C1F6@mikemccandless.com>
 <3783C878.B8C3790B@bioreason.com> <19990708211651.A17658@optiplex.palga.uucp>
Message-ID: <7cQP0BAYKRh3Ew6K@jessikat.demon.co.uk>

In article <19990708211651.A17658 at optiplex.palga.uucp>, Gerrit Holl
<gerrit.holl at pobox.com> writes
>On Wed, Jul 07, 1999 at 03:36:56PM -0600, Andrew Dalke wrote:
>> From: Andrew Dalke <dalke at bioreason.com>
>> Subject: Re: scoping assertion
>> Date:        Wed, 07 Jul 1999 15:36:56 -0600
>> To: python-list at python.org
>> 
>> 
>>   try:
>>     1/0
>>   except ZeroDivisionError:
>>
>
>isn't the following much easier, faster and readabler?
>try:
>  raise err
>except err:
>
>regards,
>Gerrit.
>
I use
try:
        raise None
except:
        ....
-- 
Robin Becker



From ivanlan at callware.com  Mon Jul 26 15:02:32 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 26 Jul 1999 19:02:32 GMT
Subject: Editors for Macintosh
Message-ID: <379CB0C8.FA84F91C@callware.com>

Hi All--
Are there any other good editors (i.e., Python-aware) besides Alpha out
there for the Macintosh?  I don't want to _use_ one, I just want to know
if there are any that people recommend.

<i-used-a-lisa-once>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From nbecker at fred.net  Tue Jul 20 08:10:13 1999
From: nbecker at fred.net (nbecker at fred.net)
Date: 20 Jul 1999 08:10:13 -0400
Subject: python vs. lisp (whereis quote?)
Message-ID: <x88r9m3cyje.fsf@rpppc2.hns.com>

One feature of lisp (emacs lisp actually) that I think is pretty cool
is the "hooks".  A hook is basically unevaluated code that will be run 
by eval later.

I suspect python lacks this capability.  As I understand it, the only
similar capability is "eval".  But I think eval expects a string,
which it will then parse.  While the capability is similar, having to
parse a string could be inefficient.

Am I correct that what is missing is the ability to postpone
evaluation using "quote"?



From ivnowa at hvision.nl  Thu Jul  8 01:17:26 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Thu, 8 Jul 1999 05:17:26 GMT
Subject: Two Pythons talking to each other?
In-Reply-To: <199907080019.TAA22244@buffalo.fnal.gov>
References: <199907080019.TAA22244@buffalo.fnal.gov>
Message-ID: <199907080516.HAA19264@axil.hvision.nl>


On 7 Jul 99, Charles G Waldman wrote:

> There's no reason, just because the two programs are running on the same
> machine, that they can't talk to each other using network protocols
> (TCP/IP, sockets, etc).  You can use the "loopback" address of 127.0.0.1
> to make all connections.  You really don't want to develop and test using
> a completely different communications scheme (e.g. pipes, files) then
> re-implement everything for the networked version, and debug it all over
> again.  Client and server can both run on the same box.

Hmm, I didn't know that. My programs usually start out as 
client/server (one listening, one connecting), but once the 
connection is established they'll talk to each other on a peer-to-
peer basis. Will this be a problem with this address 127.0.0.1?

Thanks,

--Hans Nowak (ivnowa at hvision.nl)
Homepage: http://fly.to/zephyrfalcon



From kas at maps.magnetic-ink.dk  Sat Jul 17 10:29:38 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Sat, 17 Jul 1999 16:29:38 +0200
Subject: Python 1.5.2 + Linux 2.0.36 == Bad Combo?
References: <slrn7ou0ru.3sp.kas@titan.magnetic-ink.dk> <7mq2q1$7q8$1@brick.cswv.com>
Message-ID: <slrn7p14qi.1l3.kas@titan.magnetic-ink.dk>

Neel Krishnaswami <neelk at brick.cswv.com> wrote:

> >  *  DejaGrabber
> >     -- The program connects without problem but seems to only
> >     retrieve the headers, not the full articles.
>
> I wrote DejaGrabber, and when I looked at the documentation, I think
> you might have been bitten by a bug in DejaGrabber's documentation
> rather than a bug in your system.

Ah, this explains the behaviour!  I'll try your suggestion and also get a
copy of your newssearch thingey.  Thanks a lot. :-)


  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From fdrake at cnri.reston.va.us  Fri Jul 23 16:48:29 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 23 Jul 1999 20:48:29 GMT
Subject: Python documentation available as ZIP files
In-Reply-To: <m2ogh39l9w.fsf@netmeg.net>
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
	<14232.48512.648609.496768@lrz.de>
	<14232.48816.993092.9754@weyr.cnri.reston.va.us>
	<m2ogh39l9w.fsf@netmeg.net>
Message-ID: <14232.54557.297113.405722@weyr.cnri.reston.va.us>

Les Schaffer writes:
 > 	raise " WARNING: it must be friday afternoon!!!  "


  Oh, most definately!  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From gmcm at hypernet.com  Tue Jul 13 18:32:51 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 13 Jul 1999 17:32:51 -0500
Subject: problem with SocketServer -- address already in use
In-Reply-To: <931899928snz@vision25.demon.co.uk>
Message-ID: <1280221141-7269722@hypernet.com>

Phil Hunt wrote:

> I'm trying to use SocketServer.py to write a TCP server. I ran
> my program with ``python serv.py'', which went into an infinite loop
> (which is what it should do). I exited my program with ^Z, and tried
> to start it up again. I got an error message:
> 
> socket.error: (98, 'Address already in use')
> 
> which I presume is because something thinks the old invokation of my
> program is still using port 7070. 
> 
> How do I clear the port, before I start up my server?

Three non-exclusive optons:

1) Wait. It will eventually go away.

2) Use a try:... finally: ... and close the bound socket in the 
finally clause.

3) Before binding the socket, use setsockopt(SOL_SOCKET, 
SO_REUSEADDR, 1)

Once it has happened, (1) is your only viable alternative.

- Gordon



From nbecker at fred.net  Thu Jul 22 15:17:23 1999
From: nbecker at fred.net (nbecker at fred.net)
Date: 22 Jul 1999 15:17:23 -0400
Subject: Repost: Re: Scientific Plotting?
References: <379392E7.EB0F29E7@ssec.wisc.edu> <7n5fhn$9e5$1@nnrp1.deja.com> <7n76dn$rbc$1@nnrp1.deja.com>
Message-ID: <x889088o5oc.fsf@rpppc2.hns.com>

A *really* fine plotting package:

http://plasma-gate.weizmann.ac.il/Grace

Just use os.popen



From shillcoc at mry.rbd.com  Fri Jul  9 12:43:59 1999
From: shillcoc at mry.rbd.com (Scott Shillcock)
Date: Fri, 09 Jul 1999 09:43:59 -0700
Subject: console show in Python?
Message-ID: <378626CF.9495FAB0@mry.rbd.com>

Hello,

I am moving to Python from Tcl/Tk.
Is there something similar to the Tcl/Tk 'console show' command in
Python.

I am using Tkinter for my gui and once I call mainloop() I no longer
have access to the Python command line. I want to be able to interact
with my gui from the command line to test new code.

>From Tcl/Tk I could turn on and off the interactive command line with
'console show' and 'console hide'. That is all I want to do with Python.
Can anyone help me with this or point me in the right direction.

Thanks,
Scott Shillcock



From gmcm at hypernet.com  Sat Jul  3 19:09:27 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 3 Jul 1999 18:09:27 -0500
Subject: calldll & PyArg_ParseTuple strings
In-Reply-To: <7llqr8$crd$1@nnrp1.deja.com>
Message-ID: <1281082558-50965892@hypernet.com>

Scott McClellan writes:
> 
> I'm attempting to use Sam Rushing's calldll module to call a C
> function in a dll, with the following prototype:
> 
> long __declspec(dllexport)func(long *x, long *y, long *z)
> 
> I expect the function to give certain values back in x, y, and z,
> but so far have been unable to make that happen. Let's say that I
> have three long integers in Python that I initialize as follows:
> 
> x=y=z=0L

A Python int is a C long. You won't find a Python long in C.

> My question is, in the following call from Python,
> 
> call_foreign_function(pFunc, 'lll', 'l', (<?>,<?>,<?>))
> 
> what should the tuple (<?>,<?>,<?>) contain? 

params = calldll.membuf(12)
params.write('\000'*12)
px = params.address()
call_foreign_function(pFunc, 'lll', 'l', (px, px+4, px+8))
rslt = params.read(0,12)
x,y,z = struct.unpack('lll', rslt)

Actually, Sam's got a slicker way of doing it, but I don't understand 
how it works, so I'll stick with the dumb-but-effective method.

> If I pass (x,y,z), the
> Python interpreter (Idle in my case) crashes, and this is
> understandable, since it results in the contents of x, y, and z
> being used as addresses to write to.  I've tried
> (id(x),id(y),id(z)), since id() is supposed to return the address of
> a Python object - Python didn't crash as before, but nothing was
> copied into x, y, and z, either.

id returns the address as a string. You wouldn't want this anyway. 
The Python int 0 is not a slot - it IS the one and only int object 
with value 0. If the foreign function were to return a 3 into the 
Python int (which it wouldn't - it would clobber something else in 
the struct) you would then have

>> 3 + 0
6

(The most commonly used ints are static singletons - try 
 sys.getrefcount(0))

- Gordon



From jhspies at futurenet.co.za  Tue Jul  6 12:27:34 1999
From: jhspies at futurenet.co.za (Johann Spies)
Date: Tue, 6 Jul 1999 16:27:34 GMT
Subject: AttributeError - a dumb question answered
Message-ID: <Pine.LNX.3.96.990706182551.806C-100000@Johann>

Sorry that I have bothered you.  I just now saw my typo "Krylys" and
"KryLys" which caused the problem.

Johann

 --------------------------------------------------------------------------
| Johann Spies                                 Windsorlaan 19              |
| jhspies at futurenet.co.za                3201 Pietermaritzburg             |
| Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
 --------------------------------------------------------------------------

     "Ask, and it shall be given you; seek, and ye shall
      find; knock, and it shall be opened unto you."        
                                  Matthew 7:7 






From wtanksle at dolphin.openprojects.net  Wed Jul 28 15:25:50 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Wed, 28 Jul 1999 19:25:50 GMT
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <slrn7psisn.da.wtanksle@dolphin.openprojects.net> <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com>
Message-ID: <slrn7pum9t.2ro.wtanksle@dolphin.openprojects.net>

On Wed, 28 Jul 1999 00:50:14 GMT, Fredrik Lundh wrote:
>William Tanksley <wtanksle at dolphin.openprojects.net> wrote:
>> The very worst part of the current else: behavior is that it changes the
>> meaning of else.  In other constructs, else is an alternate path to take
>> if the data being tested fails a single expression test.  In this
>> contruct, else is a path taken if the code block belonging to the previous
>> test executes a certain instruction.

>nope.  you've got it all backwards.  consider this:

>    for an "if" statement, "else" is a path taken if the
>    expression evaluates to false.

Close enough -- I would say it's the path taken if the 'if' block isn't
executed.

>    for a "while" statement, "else" is a path taken if the
>    expression evaluates to false.  or in other words,
>    when the loop terminates by natural causes.

In other words, it executes when you'd expect an 'else' to not execute --
'else' doesn't mean "natural causes".

>    for a "for" statement, "else" is a path taken when
>    there are no more elements to loop over.  or
>    in other words, when the loop terminates by
>    natural causes.

In addition to the above carping, I have to add that the code following
the for (or while) loop is what I'd expect to execute after the loop
terminates.

>not that complicated, was it?

I didn't think it was either, but it seems that in spite of my initial
understanding, I still managed to reverse it accidentally.

>now, what you seem to have trouble with isn't
>the "else"-clause -- it's the behaviour of "break":

>"else" always works the same way.  and so does
>"break".  what else did you expect from Guido?

I expect consistency, and I expect that the major features (else with if
and else with catch) will define consistency, not minor features like else
with looping.

try:
   ...
except something:
   ...
else:
   ...

The else block here executes iff the except blocks don't execute.

></F>

-- 
-William "Billy" Tanksley



From heg at softbrew.com  Sun Jul 18 17:55:58 1999
From: heg at softbrew.com (Harold Gottschalk)
Date: Sun, 18 Jul 1999 21:55:58 GMT
Subject: Strange side effect from reading a list of numbers from argv
Message-ID: <O3sk3.6047$Sg6.52046@news1.rdc1.sdca.home.com>

I just wrote this program to become familiar with python and I do not
understand why it works with a list I create with in the program and not
when I use sys.argv[1:].

I am sure there is some nuance I am missing, but an explanation to its
behavior would be helpful.

If you switch the comment between the two for statements you will see the
different results.  Code follows below.

When using the command line to get your list:
findCurrent1.py 3 4 5 90 33 59604 1
largeNum =  90  smallNum =  999999999

When using the internal list the results are:
largeNum =  59604  smallNum =  1

Thanks for the help,

Harold


_____Code______

import sys

#Intialize Varables
largeNum = -999999999
smallNum = 999999999
list = [3,4,5, 90, 33, 59604, 1]

for value in sys.argv[1:]:
#for value in list:
    if value >= largeNum:
        largeNum = value
    if value <= smallNum:
     smallNum = value

print "largeNum = ",largeNum," smallNum = ",smallNum

_______END CODE_______





From jlj at cfdrc.com  Mon Jul 19 11:01:05 1999
From: jlj at cfdrc.com (Lyle Johnson)
Date: Mon, 19 Jul 1999 10:01:05 -0500
Subject: Are postings to comp.lang.python.announce mirrored here?
Message-ID: <37933DB1.47814E84@cfdrc.com>

Hi All,

 1. For some reason I'd convinced myself that posts
    to comp.lang.python.announce were automatically
    copied here; not true?

 2. Is it considered bad form to cross-post new S/W
    announcements to both lists? Or is that the Right
    Thing to do?

Lyle



From mwh21 at cam.ac.uk  Wed Jul  7 14:32:00 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 07 Jul 1999 19:32:00 +0100
Subject: scoping assertion
References: <37839491.3F7C1F6@mikemccandless.com>
Message-ID: <m31zekpazj.fsf@atrus.jesus.cam.ac.uk>

Michael McCandless <mail at mikemccandless.com> writes:

> I wonder if Python could be extended to allow a prefix in front of
> variable names (lvalues?) to assert that the name did not previously
> exist in this scope.  EG, call it "nonexist":
> 
>   nonexist x = 0
>   while x < 10:
>     ...
> 
> Or
> 
>   for nonexist x in xrange(1, 1000):
>     ...
> 
> If x had previously existed in this scope at the time of assignment, an
> appropriate exception would be raised.
> 
> Jim Hugunin pointed out that this could be achieved as follows:
> 
>   assert not locals().has_key('x')
> 
> But because there are no macros in Python, and because you can't make
> this a function (?), I think you'd need to type this out every time.
> 
> The reason I ask is because I've had some sneaky bugs whereby I assigned
> to what I thought was a new variable name, but inadvertantly wiped out a
> value that I later used.
> 
> Statically typed languages basically give you this "for free" ;),
> meaning if you try to "declare" the variable again, the
> compiler/interpreter will typically generate an error saying the
> variable is already defined in this scope.
> 
> Mike

Well, it's a bit baroque, but it works (if you have my bytecodehacks
package: http://starship.python.net/crew/mwh/):

from bytecodehacks.macro import add_macro, expand
from bytecodehacks.code_editor import EditableCode,Function
EditableCode.AUTO_RATIONALIZE = 0

def notexist((x)):
    try:
        x
    except:
        return
    del x
    x

ff = Function(notexist)
ff.func_code.rationalize()

add_macro(ff.make_function())

if __name__ == '__main__':
    def test():
        notexist(y) # should succeed
        y = 1
        notexist(y) # should fail
    test=expand(test)
    test()

and to prove it:

[mwh21 at atrus python]$ python notexist.py
Traceback (innermost last):
  File "notexist.py", line 24, in ?
    test()
  File "notexist.py", line 22, in test
    notexist(y) # should fail
NameError: y

unhelpful-ly y'rs
Michael



From bernhard at alpha1.csd.uwm.edu  Tue Jul  6 17:52:57 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 6 Jul 1999 21:52:57 GMT
Subject: IGNORE: Test of posting software
References: <37822EB4.7D0E602E@ipro.com>
Message-ID: <slrn7o4ulo.eul.bernhard@alpha1.csd.uwm.edu>

On Tue, 06 Jul 1999 09:28:36 -0700, I/PRO Survey Research <survey at ipro.com> wrote:
>This is a test.
Please don't do that again!!!
Use a *.test newsgroup to test your software.
That's what they are for.
	Bernhard


-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.    (www.uwm.edu/~bernhard)
Association for a Free Informational Infrastructure              (ffii.org)



From gmcm at hypernet.com  Wed Jul 21 15:49:58 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 21 Jul 1999 14:49:58 -0500
Subject: gadfly RDBMS ???
In-Reply-To: <3795EA7F.4A452490@arcus.tel.hr>
Message-ID: <1279539701-16597123@hypernet.com>

Darko Cvagic writes:

> Thanks, I think that I'm exchange terms like RELATION and LINK
> So, Gadfly IS RELATIONAL, but I can't make something like:
>  CREATE LINK order_customer
>  FOREIGN KEY order(cno)
>  REFERENCES customer
>  ON DELETE RESTRICT
> 
> I can't do it in gadfly, am I wright?  
> What is name for construction like this? 
> I was thought that it is RELATIONAL , I was wrong!

CREATE LINK is not in the ANSI standard at all. In the standard, 
you'd do this in the CREATE TABLE statment.

The 86 level of the standard doesn't even have FOREIGN KEY or
REFERENCES. 

The term "relational" can mean all kinds of things. I've seen it used 
(by big-wigs in the field) to discard DBs that were based on 
inverted-list technology as "non-relational" despite their having 
full ANSI SQL capabilities.

- Gordon



From fsnm at nmia.o_r_g  Thu Jul 15 17:51:20 1999
From: fsnm at nmia.o_r_g (David Hanley)
Date: Thu, 15 Jul 1999 15:51:20 -0600
Subject: Harlequin Is Back!
References: <%vti3.121$95.4053@newse2.tampabay.rr.com> <joswig-1307992158330001@194.163.195.67> <86btdgz2rg.fsf@g.pe <wkn1wzl9kt.fsf@ifi.uio.no> <71E04BBBB50245BC.87397A36A7EDBBA0.44F0816EBAA77500@lp.airnews.net>
Message-ID: <378E57D8.A77A9F9C@nmia.o_r_g>


Cameron Laird wrote:

Lars is right (as usual).  Actually, compared to *any-

> thing*, Python seems to arrive first on the interfacing
> scene.  This is interesting; Tcl and Lua emphasize ease
> of embedding and extensibility in their design, and Perl
> has hordes of interfacing worker bees, but a consistent
> pattern emerges that Python does at least as well as any
> other language in achieving connection to the latest
> toolkits-protocols-...  It's an interesting story, even
> though I don't yet understand it.

Hmmm, maybe the right approach is to get lisp to talk
to python, then, eh?

dave




From graham at sloth.math.uga.edu  Mon Jul 19 12:09:46 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 19 Jul 1999 16:09:46 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com> <170719990020380651%pereira@research.att.com> <7mqaht$r27$2@cronkite.cc.uga.edu> <170719991835331004%pereira@research.att.com> <7mssne$m8r$1@cronkite.cc.uga.edu> <180719992337546352%pereira@research.att.com>
Message-ID: <7mvika$cpj$1@cronkite.cc.uga.edu>

<graham at sloth.math.uga.edu> wrote:
> expression parsers as primitive parsers". There is no need for the compiler
> to "solve the halting problem". The programmer is free to use regular
> expression/automaton techniques where he/she wants. Such parsers take
> strings and return ASTs. They can then be combined with other parsers
> as if they were primitive parsers.
Fernando Pereira (pereira at research.att.com) wrote:
: One last try. I understood your point from the beginning. However, you
: did not understood *mine*. If you push the complex
: pattern-matching/parsing machinery into black boxes, there's not much
: for the monadic glue to do, is there? 

What your combinators do depends on your application. If all you are
doing is parsing using primitive automaton based parsers then your
combinators won't have much to do. But if you are doing more than
that then they will have more to do, and hence be more complex. For
example you can use the same combinators you define for parsing to do
type checking, model state, etc -- things I would not want to do with
an automaton. That is why I said that combinators are really sequencing
constructs.

Fernando Pereira (pereira at research.att.com) wrote:
: And even then, some
: optimizations, those involving the various black boxes, will be
: blocked. Here's a simple example. Assume that there are two black boxes
: A and B, and that you use monadic glue to say "A or B". A and B may be
: deterministic, but the result will not be, and exhaustive search will
: be required in the glue. Sure, I can recognize such cases by hand and
: create a new black box "A or B" which is optimized, but that is an
: error-prone hand optimization. And one may not notice subtler
: inefficiencies.

Your point appears to be that if you were programming only using automaton
for parsing you wouldn't miss the above optimisation, but that somehow
using combinators means you would miss the above optimisation. You
and I must program differently then. If I were using automaton, even
in the context of also using combinators, I would be aware of the
potential for such optimisations and look for them.

graham
-- 
            The girls is all salty, the boys is all sweet, 
                      the food ain't too shabby 
                    an' they piss in the streets
    In France, way down in France, way on down, way on down in France



From bruce_dodson at bigfoot.com  Thu Jul 29 13:27:51 1999
From: bruce_dodson at bigfoot.com (Bruce Dodson)
Date: Thu, 29 Jul 1999 14:27:51 -0300
Subject: global myfunc; def myfunc():
Message-ID: <002e01bed9e7$afe9f210$8a8fa7d1@dawson.esricanada.com>

Thanks.  Actually, the embedding interface also has a Python.ExecGlobal
method
which binds everything to __main__, so the recommended way of binding a
function to the global namespace is Python.ExecGlobal("import modulename").
I think I will say in the documentation that you can use the python global
statement
to bind specific global variables, or you can use ExecGlobal to make
everything
global.  Then I will say that the global statement is the recommended way
for
variables, whereas ExecGlobal is recommended for defining functions.  Users
can draw the conclusion that recommendations are not requirements, but I
won't spell it out for them. :-)


-----Original Message-----
From: Gordon McMillan <gmcm at hypernet.com>
To: Bruce Dodson <bruce_dodson at bigfoot.com>; python-list at python.org
<python-list at python.org>
Date: Thursday, July 29, 1999 8:23 AM
Subject: Re: global myfunc; def myfunc():


Bruce Dodson wrote:

> I have noticed that, when using python as an embedded language in
> another program, using a temporary dictionary for a private
> namespace, and using __main__'s dictionary for my global namespace,
> I can say:
>
> global myfunc
> def myfunc():
>   ...
>
> to bind myfunc into the global namespace, which allows it to persist
> across calls.  But I want to make sure that is "intended behavior"
> and not just some implementation quirk, so I can decide whether to
> document that as a feature.

It may be an unusual use, but there's nothing quirky about the
mechanism being used. You're safe.

- Gordon





From skip at mojam.com  Tue Jul 27 13:29:33 1999
From: skip at mojam.com (Skip Montanaro)
Date: Tue, 27 Jul 1999 12:29:33 -0500 (CDT)
Subject: Python Performance
In-Reply-To: <379D89BC.DA91BA8C@hp.com>
References: <3794C642.AB6A3E5D@freenet.co.uk>
	<7n5f0c$snu$1@ffx2nh5.news.uu.net>
	<al81zdxkktf.fsf@sirppi.helsinki.fi>
	<p51zdvn4c7.fsf@bidra241.bbn.hp.com>
	<14236.32616.980578.536614@dolphin.mojam.com>
	<379D89BC.DA91BA8C@hp.com>
Message-ID: <14237.60407.130426.461846@dolphin.mojam.com>

    >> I think one of the impediments to improving the speed of the language has
    >> been the dramatic performance improvement in hardware.  There's not a lot of
    >> incentive to pour lots of time into speeding up method calls when for a few
    >> hundred bucks you can simply drop in a faster processor, disk or networking
    >> card.

    Markus> Also it's true that hardware still improves, it's also true that
    Markus> if I would try sell an application to a customer with thousands
    Markus> of users and I would tell him to buy a new machine because I use
    Markus> a slow interpreter, they would probably kill me.  There are also
    Markus> applications where every second you save is worth a lot of many.
    Markus> Production testing of IC's comes to my mind ...

True.  I was thinking about the situation I'm in (small developer, a few
machines to maintain/upgrade/replace).

Skip




From fdrake at cnri.reston.va.us  Mon Jul 12 12:47:25 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Mon, 12 Jul 1999 16:47:25 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
In-Reply-To: <199907121445.HAA09860@netcom9.netcom.com>
References: <14217.63931.75257.566266@weyr.cnri.reston.va.us>
	<000101becc7c$fe2c9c20$8e9e2299@tim>
	<3787D36A.815B7C29@ssec.wisc.edu>
	<3787FB51.12FE0048@ssec.wisc.edu>
	<7m923i$dem@dfw-ixnews5.ix.netcom.com>
	<199907121445.HAA09860@netcom9.netcom.com>
Message-ID: <14218.7197.219561.903989@weyr.cnri.reston.va.us>

Aahz Maruch writes:
 > Given that one of Python's goals is to be accessible to non-CS geeks,
 > os.chdir() and os.getcwd() *must* be documented in the section dealing
 > with files and directories.  As I said originally, I think a reasonable
 > compromise is to leave the process parameter docs alone and add a
 > pointer to that area in the files & directories docs (possibly with an
 > explanation of why they are documented somewhere else).

Tim Peters writes:
 > They don't have to be moved, provided they're also mentioned where people
 > first look <wink>.  The notion that the current directory is "a process
 > table thing" is a bit Unix-centric, and e.g. a Windows person is more likely
 > to think of it as "a directory" first.

  Well, after long consideration and a quick check the Guido, it turns 
out that Guido thinks they should be moved to the "Files and
Directories" section.  Since I disagree, I've contacted several local
bus companies to help in steering the course of future Python
development.  Fortunately for Guido, bus drivers don't like those late 
night jobs, so....
  I'm going to move the descriptions to the "Files and Directories"
and add a link from "Process Parameters" to the actual descriptions.
  Now, I'm going to start saving up for a bus of my own.  Harumph.  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From tim_one at email.msn.com  Sat Jul 10 23:50:39 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 10 Jul 1999 23:50:39 -0400
Subject: Simple Qu: Changing directory
In-Reply-To: <3787FB51.12FE0048@ssec.wisc.edu>
Message-ID: <000201becb50$8afb4da0$479e2299@tim>

[Nick Bower]
> yeah, that's what i originally thought.  sure enough there were
> some updates in the os documentation, but not os.chdir.  it's something
> that would be really helpfull if anyone linked to the documentation
> sees this.

Well, I'm copying Python's massive Documentation Department on this <wink>.

os.chdir is directly linked to from the index, under "directory, changing".
That's not a terribly unlikely place to look for it, either <wink>.

However, it's documented in section 6.1.1 (Process Parameters) instead of in
section 6.1.4 (Files and Directories).  That does seem off; I'd move chdir
and getcwd from the former to the latter; yes, "the current directory" is a
process concept, but it's more directly "a directory".

otoh-"current"-suggests-it-be-documented-under-"time"<wink>-ly y'rs  - tim





From gregm at iname.com  Tue Jul 20 20:06:17 1999
From: gregm at iname.com (Greg McFarlane)
Date: Wed, 21 Jul 1999 10:06:17 +1000
Subject: where is gfplus?
Message-ID: <19990721100617.07994@nms.otc.telstra.com.au>

I have been trying to get through to the gfplus site
<http://shamrock.med.nyu.edu/~jeff/gfplus/> but it seems to be down. 
Does anyone know of anywhere else I can get gfplus (an interactive
gadfly shell)?  Or can anyone send it too me (it can't be that big:).

Also, is there a gui interface to gadfly?  Something that lists all
the tables, all the columns in a table, all the data in a table,
indexes, etc.

Oh, and a cure for cancer would be nice too.

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From tim_one at email.msn.com  Mon Jul 26 21:22:50 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 26 Jul 1999 21:22:50 -0400
Subject: Incompatibility.py
In-Reply-To: <7niuqf$rpm$1@m2.c2.telstra-mm.net.au>
Message-ID: <000a01bed7ce$8b320ae0$eea22299@tim>

[Colin J. Williams]
> IDLE returns:                        The name is: Incompatibility

[Mark Hammond]
> Not for IDLE 0.5 it doesnt.  Are you sure you are not using
> "Import Script".
> ...
> The behaviour you describe sounds like you are simply confusing
> "Run Script" with "Import Script".

Except that "Run Script" doesn't exist in the released IDLE -- you have to
get the CVS version for that.

Since Guido doesn't have anything else to do <wink>, I vote he fixes it so
breakpoints can be set under Windows again, then put out a new IDLE release
+ updated Lib/pyclbr.py.  The CVS is version is gads better in several key
ways.

release-early-&-often-ly y'rs  - tim





From johanw at easics.be  Fri Jul  2 09:41:16 1999
From: johanw at easics.be (Johan Wouters)
Date: Fri, 02 Jul 1999 15:41:16 +0200
Subject: Why does shelve make such large files?
References: <19990701211450.A6827@optiplex.palga.uucp> <377BF01D.16F8AA33@cup.hp.com> <19990702123152.B9975@optiplex.palga.uucp>
Message-ID: <377CC17C.4487A36A@easics.be>

Gerrit Holl wrote:

> Ah, I understand.
> So pickle is useful for very small datases, but when they're really huge, one
> should use shelve. Isn't it?

Pickle will serialise your data into a certain format. This way you can
"store" objects like lists as a whole. The serialising has the benefit that
you can eg. put your object (be it a list or something else) through a pipe
or socket and depickle it at the other side. This way you can have python 
processes that exchange data in a sort of native python format.

Shelve gives you alot more than just storing the data as a long sequence.
Imagine the extra space being the administration that is needed for all
the extra functionality (like tagging what data items are valid etc)

The overhead induced by a database management system can be enormous
compared to the bare data you want to access, but this is mostly a
space/time tradeoff: the more space you use the faster you can do
things like searching and sorting etc etc.

kind regards,
===================================================================
Johan Wouters             ===              Easics               ===
ASIC Designer             ===  System-on-Chip design services   ===
Tel: +32-16-395 616          ===================================
Fax: +32-16-395 619      Interleuvenlaan 86, B-3001 Leuven, BELGIUM
mailto:johanw at easics.be           http://www.easics.com



From tim_one at email.msn.com  Tue Jul 20 01:06:36 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 20 Jul 1999 01:06:36 -0400
Subject: Killing a thread
In-Reply-To: <7mvml3$u5k$1@nntp6.u.washington.edu>
Message-ID: <001001bed26d$a4c08060$139e2299@tim>

[Donn Cave]
> Python's thread support is evolving.  As a previous followup points out,
> you shouldn't kill threads, and I guess that's fair enough - in terms
> of normal program design, one should look for other ways to manage
> threads.
>
> But in default of any alternative, I suspect sometimes we have to violate
> this commandment, and I think one of these days Python will provide the
> tools to do what you have to do, not just what you want to do.

It's unlikely Python will ever allow thread A to kill thread B, or even
suspend it.  Hygiene aside, not all native thread packages support this, and
Python's portable semantics can't be lower-level than the most restrictive
thread platform's.

Note too that Java-- which got to define every bit of its thread
semantics --eventually removed inter-thread carnage from the language.  For
them, allowing it simply proved to be such a bad idea in practice that it
was retroactively declared a design error.

not-every-tool-is-applicable-as-is-to-every-problem-ly y'rs  - tim





From nick at NOSPAMvideosystem.co.uk  Thu Jul  1 15:21:09 1999
From: nick at NOSPAMvideosystem.co.uk (Nick)
Date: Thu, 1 Jul 1999 20:21:09 +0100
Subject: Tuples & lists
Message-ID: <377bbde1@news.xtml.co.uk>

I'm a new user and Python looks great.

Why both having tuples and lists? As far as I can see there're the same, but
tuples are read only..? Have I missed something in the tutorial?

Also, has anyone written a tcl/tk dialog/resource editor type thing so you
could save a form out as code? Could save a lot of time...

Thanks,
Nick.







From g_will at cyberus.ca  Fri Jul 16 12:44:17 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Fri, 16 Jul 1999 16:44:17 GMT
Subject: Radio Buttons and Variables
Message-ID: <01becfaa$4fc73d80$be54c3d1@Cyberus.cyberus.ca>

I'm having trouble getting radio buttons and variables working properly. 
The code works properly the first time it is run from Pythonwin but the
second time the variable does not update properly.  This means that
foovar.get()=0 regardless of the radio button pressed.

Each time the script is run, print foovar indicates a new variable (eg
PY_VAR0 then PY_VAR1 etc).  Could this be the problem?  

The script below prints the value of the radio button variable when the
button at the bottom is pushed.

I'm sure this is a common problem but could not find any reference to it.

Any ideas to solve this?

Thanks,
Gordon Williams

"""
 
Contents:
---------

HorizRadioGroup:
  horizontal and vertical radiobutton groups,

HorizButtonSet:
  horizontal and vertical sets of buttons

MenuBar:
  quick and easy menubar management using too complex data structures,
  but it works.

"""

from Tkinter import *
from types import IntType
import sys

class MenuBar(Frame):
    def __init__(self, menus, master=None):
	Frame.__init__(self, master, relief=RAISED, bd=2)
	for menu in menus:
	    title = menu['title']
	    items = menu['items']
	    del menu['items']
	    del menu['title']
	    button = Menubutton(self, text = title)
	    button.pack( side = LEFT, padx = '1m')

	    themenu = Menu(button, menu, tearoff=0)  # should make tearoff be
configureable

	    for item in items:
		try:
		    mtype = item['type']
		    del item['type']
		except: mtype = 'command'
		try:
		    mlabel = item['label']
		    del item['label']
		except: mlabel = None
		try: 
		    mcommand = item['command']
		    del item['command']
		except: mcommand = None
		cnf = item
		if mlabel: cnf['label'] = mlabel
		if mcommand: cnf['command'] = mcommand
		
		themenu.add(mtype, cnf)

	    button['menu'] = themenu
	

class RadioGroup(Frame):
    def __init__(self, elements, default, theanchor = W, master = None):
	Frame.__init__(self, master)
	self.elements = []
	for i in range(len(elements)):
	    e = elements[i]
	    thetext = e['text']
	    thevar = e['variable']
	    print thetext, thevar, i, thevar.get()   #diagonstic only
	    b = Radiobutton(self, text=thetext, variable=thevar,
			    value=`i`, anchor=theanchor)
	    self.elements.append(b)
	    if i == default:
		self.elements[i].select()

	    if theanchor == W:
		theside = LEFT
	    else: 
		theside = TOP
		
	    self.elements[i].pack(side=theside, expand=1, fill=BOTH)


class HorizRadioGroup(RadioGroup):
    def __init__(self, elements, thedefault, master = None):
	RadioGroup.__init__(self, elements, thedefault, W, master)


class ButtonGroup(Frame):
    def __init__(self, elements, theanchor = W, master = None):
	Frame.__init__(self, master)
	self.elements = []
	for i in range(len(elements)):
	    e = elements[i]
	    thetext = e['text']
	    thecommand = e['command']
	    
	    b = Button(self, text=thetext, command=thecommand)
	    self.elements.append(b)
	    if theanchor == W:
		theside = LEFT
	    else: 
		theside = TOP
		
	    self.elements[i].pack(side=theside, expand=1,
				  anchor = S, fill=BOTH)


class HorizButtonGroup(ButtonGroup):
    def __init__(self, elements, master = None):
	ButtonGroup.__init__(self, elements, W, master)




if __name__ == '__main__':
    root=Tk()
    def newfile():
	print 'New File...'
    def openfile():
	print 'Open File...'
    def spam1():
	print 'Spam! 1'
	print 'the variable is: Choice', foovar.get()+1
    def spam2():
	print 'Spam! 2'
	print 'the variable is: Choice', foovar.get()+1
    def spam3():
	print 'Spam! 3'
	print 'the variable is: Choice', foovar.get()+1

    mdef = [{'title':'File',
	     'items':[{'command':newfile, 'label':'New'},
		      {'command':openfile, 'label':'Open'},
		      {'type':'separator'},
		      {'command':sys.exit, 'label':'Quit'},
		      ],
	     },
	    {'title':'Edit',
	     'items':[{'label':'Cut'},
		      {'label':'Copy'},
		      {'label':'Paste', 'state':'disabled'}
		      ]
	     },
	    ]
    m = MenuBar(mdef,master=root)
    m.pack(side=TOP, fill = BOTH)

    foovar = IntVar()
    
    elements = [
	{'text' : 'Choice 1',
	 'variable' : foovar},
	{'text' : 'Choice 2',
	 'variable' : foovar},
	{'text' : 'Choice 3',
	 'variable' : foovar},
	]

    elements2 = [
	{'text' : 'Choice 1',
	 'command' : spam1},
	{'text' : 'Choice 2',
	 'command' : spam2},
	{'text' : 'Choice 3',
	 'command' : spam3},
	]

    a = HorizRadioGroup(elements, 0,master=root)
    a.pack(side=TOP, fill=BOTH)

    c = HorizButtonGroup(elements2,master=root)
    c.pack(side=BOTTOM, fill=X)
    
    root.mainloop()





From Gerrit.Holl at p98.f112.n480.z2.fidonet.org  Thu Jul  1 19:35:28 1999
From: Gerrit.Holl at p98.f112.n480.z2.fidonet.org (Gerrit Holl)
Date: Fri, 02 Jul 1999 00:35:28 +0100
Subject: print color strings?
Message-ID: <00001633@bossar.com.pl>

From: Gerrit Holl <gerrit.holl at pobox.com>

Hello,

is there an easy way to print color strings in python?

A function like printcolor('brightyellow', 'red', 'spam'), that finds
out on what kind of terminal you are and prints it to stdout?

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!




From manini at ies.it  Fri Jul 16 02:20:04 1999
From: manini at ies.it (Luca Manini)
Date: Fri, 16 Jul 1999 06:20:04 GMT
Subject: Generating python code in glade
Message-ID: <3790ce56.1276725@news.eclipse-net.it>

Hi, 

is there anybody out there using glade ? 

Which "software" do I need apart from glade and a standard python
distribution ?

Glade has an option for generating code in various languages, and
somewhere int he GUI there is a combo box that (in my installation)
lists C, C++ and Ada. What should I do to get python appear there ?


	Thanks a lot for any hint, Luca. 

-----------------------------
Luca Manini                    
GIS and Database Developer     
ArcView and Delphi Programmer  
mail: manini at ies.it            
-----------------------------
Informatica e Servizi
via IV Novembre, 100
38014 Gardolo (Trento)
      Italy
tel:  +39 (461) 992 040
fax:  +39 (461) 991 666
www:  www.ies.it
-----------------------------
                        



From Aahz.Maruch at p98.f112.n480.z2.fidonet.org  Thu Jul  1 12:19:58 1999
From: Aahz.Maruch at p98.f112.n480.z2.fidonet.org (Aahz Maruch)
Date: Thu, 01 Jul 1999 17:19:58 +0100
Subject: Mailbox, tell and read
Message-ID: <00001621@bossar.com.pl>

From: aahz at netcom.com (Aahz Maruch)

In article <377B7742.F39CB0A4 at prescod.net>,
Paul Prescod  <paul at prescod.net> wrote:
>
>First, are there platforms other than DOS/Windows/Win32 where there is a
>difference between text and binary mode? I didn't want to munge things
>myself because I'm not sure if what is the "right" munge on the Mac, for
>instance. Can I just do a search and replace for CR/LF -> CR on all
>platforms, all of the time?

I ran into this last year (while I was still a Perl heretic).
Unfortunately, you *have* to use binary mode in opening a mail file
because one of the legal MIME attachment types is non-encoded binary --
and that means you can't just blindly do a search/replace, either.
Never did come to a completely satisfactory answer.  :-(

I think CR will break on Unix, but I'm not sure.
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From Dinu.C..Gherman at p98.f112.n480.z2.fidonet.org  Thu Jul  1 07:33:52 1999
From: Dinu.C..Gherman at p98.f112.n480.z2.fidonet.org (Dinu C. Gherman)
Date: Thu, 01 Jul 1999 12:33:52 +0100
Subject: TopSort in Python?
Message-ID: <00001608@bossar.com.pl>

From: Dinu C. Gherman <gherman at my-deja.com>

Does anybody have a simple-minded-but-working full-Python
implementation of topsort, the topological sorting algorithm?
Or maybe *any* topsort? I remember only one such algorithm...
python.org/search doesn't reveal any...

Thanks,

Dinu


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From donn at u.washington.edu  Thu Jul  1 17:55:55 1999
From: donn at u.washington.edu (Donn Cave)
Date: 1 Jul 1999 21:55:55 GMT
Subject: print color strings?
References: <19990701213528.A8877@optiplex.palga.uucp> <14203.53386.710688.87678@weyr.cnri.reston.va.us>
Message-ID: <7lgo5b$15is$1@nntp6.u.washington.edu>

"Fred L. Drake" <fdrake at cnri.reston.va.us> writes:
|Gerrit Holl writes:
|> is there an easy way to print color strings in python?
|> 
|> A function like printcolor('brightyellow', 'red', 'spam'), that finds
|> out on what kind of terminal you are and prints it to stdout?
|
| Gerrit,
|   If you're doing this on Unix, take a look at the ncurses library;
| there's a Python interface to that.  (I think Oliver Andrich did it;
| searching DejaNews should find it fairly quickly since this keeps
| coming up.)
|   The standard curses module will be documented in the next release of 
| the Library Reference.

And if that doesn't work, probably your termcap or terminfo definitions
lack the necessary information about color escape sequences.  That's a
widespread problem.

The color escape sequences that have worked for me on a couple of terminal
types have been of the form <ESC>[32m, where 32 is a number from 30 to 37.
I use only 31 to 36.  30 and 37 are more or less black and white, but the
details vary, so I switch from color to non-color with <ESC>[m instead of
trying to sort out the details of black and white vs. background and
foreground.

Of course these are terminal specific escape sequences.  They're evil and
you must not use them, especially if you have a choice.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu






From greg.ewing at compaq.com  Thu Jul 22 18:19:05 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 23 Jul 1999 10:19:05 +1200
Subject: Maintainability (was Re: Stackless & String-processing)
References: <001101bed3f2$5e8874a0$642d2399@tim>
		<7n675p$qsk@dfw-ixnews7.ix.netcom.com> <14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>
Message-ID: <379798D9.FF3F811B@compaq.com>

Skip Montanaro wrote:
> 
> Some modules
> in the standard distribution were designed specifically to support this
> (types and Tkinter come to mind).

In the case of Types, it could easily have been designed
differently. Writing Types.List would not be much harder than
typing Listtype and would be easier to read as well.

As for Tkinter, I've gotten into the habit of always
writing Tkinter.Button, etc. and I think the qualification
is a good idea. It means, for example, I can have a
module called Fancy which implements a custom button
subclass and call it Fancy.Button. I like that better
than having to call the class Fancybutton or something.

So I wouldn't mind much if the * form of import
went away, either. You can easily write a function
which does the same thing if you really, really want it.

Greg



From Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org  Fri Jul  2 01:51:43 1999
From: Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 06:51:43 +0100
Subject: Python GUIs: Portability or not?
Message-ID: <00001654@bossar.com.pl>

From: mikael at toolwood.com (Mikael Lyngvig)

On Thu, 1 Jul 1999 23:16:00 -0500, Travis Oliphant <olipt at mayo.edu>
wrote:

>It struck me that all this talk of a cross-platform GUI might be
>preventing a lot of very good GUI's from being used.

Well, it probably is.  But, I, for one, have this fobia about
portability - it's so easy to give up and so hard to get back :)  My
experience is that the moment you give up portability, you give up
lots of chances to verify the architecture and implementation of your
application.  No chance to compile with, say, five different compilers
on different platforms and benefit from the fact that the XXX compiler
is so great about detecting unreferenced variables in switch
statements, while the YYY compiler catches memory allocation errors in
a brilliant way.  No chance to try it out on your Y box, because
there's such a nifty profiler.  No chance to check if your binary file
formats are sound by exchanging data files between little-endian and
big-endian machines.  All in all, I think portability is a much bigger
thing that the ability to compile and run your app on some obscure
machine nobody you know has ever seen in real life. 

(and portability also makes it possible to distinguish the pro from
the amateur, in these times of GUI IDEs with wizards and all :-)

>So, I looked in to PyGtk and PyGnome (with its Gnome Canvas) and I'm
>impressed.  Right now it fits what I want to do very well and it works on
>systems with an X server (aren't X servers available for MAC's and
>Windows?)  

I don't know about the Mac, but I'm sure there are X servers available
for NT.  However, not like they are available for Linux and other Unix
boxes - you don't just download the X11 source, compile it, and that's
it (unless a lot has changed the last few years).

You'd probably have to buy a commercial X server for NT or, even
worse, ask your users to do so.

>So, I'm happy with it. If you're in the same situation you
>might like it too.  

Well, I'm happy that you're happy!  

Even though I probably won't even get to need the Unix portability, I
still sleep much better if I know that I don't have to discard by GUI,
the day I'm asked to do a Linux port.  And, then again, if the GUI is
isolated cleanly enough from the application, unlike Tk's
implementation (...), it wouldn't be such a big task to switch GUI. 

However, for many projects I agree with you - portability is not a big
issue.  For instance, I've been considering to do a decent Microsoft
Visual SourceSafe client (a  Win32-specific revision control system)
in Python.  In such a case, there would be no need whatsoever for
portability outside Win32 (i.e., between Windoze 9x and Windows NT).

Gotta think this over for a while, because you're probably more right
than I'd like to think.  Anyway, it appears that the best candidates,
for my needs, wxWindows and FOX (which I haven't evaluated yet), are
both portable to Win32 and a lot of other platforms. 

Can PyGTK be used on Win32 without an X server?

>I tend to agree that there is plenty of room for more than one GUI.

And, luckily, there are more than one!


-- Mikael



From aaron_watters at my-deja.com  Tue Jul 27 11:37:56 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Tue, 27 Jul 1999 15:37:56 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <7nhseu$gf0$1@nnrp1.deja.com> <al8d7xen311.fsf@sirppi.helsinki.fi>
Message-ID: <7nkjoj$a86$1@nnrp1.deja.com>

In article <al8d7xen311.fsf at sirppi.helsinki.fi>,
  Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:

> > Note please that gadfly queries have a certain amount of fixed
overhead.
> > You *might* find that queries over a table with one entry are not
> > much faster than queries over tables with 1000 entries, or maybe
even
> > tables with 10000 entries that have appropriate indices.
...
> > On second thought 21 queries per second ain't too bad...
> > What are you comparing it to?  Updates are naturally slower
> > since they involve writes to a log file for recovery purposes.
>
> Single select/update's on Solid or MySQL with Python interface do
easily
> 1000/second on trivial tables. (on most boxen)

Hmmm.  Just out of curiousity does gadfly's relative performance
improve if the queries/tables are not trivial?  I tried to make it such
that the fixed overhead for each query would amortize well over
larger datasets and to some extent more complex queries (at least
of some varieties).  I'm not at all surprised that the msql
and Solid fixed overheads are much lower for trivial queries & tables.
(and I wouldn't be surprised if they were generally faster too
esp. since, eg mysql doesn't support transactions and recovery :) ).
   -- Aaron Watters

===

It was a typical military/political decision.
The navy wanted 6 ships and the minister offered 4,
so they compromised on 8.  -- Churchill


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From MHammond at skippinet.com.au  Tue Jul 13 10:37:28 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Wed, 14 Jul 1999 00:37:28 +1000
Subject: making a Windows NT service out of a python program
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F20@RED-MSG-50> <378B2AA8.58477120@appliedbiometrics.com>
Message-ID: <7mfirr$rn$1@m2.c2.telstra-mm.net.au>

Christian Tismer wrote in message
<378B2AA8.58477120 at appliedbiometrics.com>...

>This is all great stuff if you have understood it already.
>For me, it is still unclear what makes up a service, and
>what I need at the minimum to turn a Python program into
>an NT service.
>
>I need nothing more than to run Medusa as a service. Did
>anybody do this before? Do I need special calls at all?

Here are services explained in as few lines as possible:

NT designed services to be capable of responding asynchronously to requests.
While a service is working, NT can send a stop/pause etc request.  The
service can take pretty much as long as it likes to complete the request,
but it must be very quick about telling NT it is responding, and the state
change is pending (and continue to tell it until the transition is
complete).

At the raw API level, a service starts up by passing a callback function
back to NT.  Whenever NT needs to control the service, it calls this
function.

Thus, services are almost always threaded - one thread doing the work of the
service, and another responding to service control messages.  Could be
single threaded, but the single thread must be fully asynchronous for this
to work.

So, the short answer is that to _really_ work as a service properly, a
program must be written as a service (then the next thing people want are
event log messages and performance monitor data!).  Short of that, it would
actually be trivial to write a service that could run any arbitary Python
script.  You would not have the same level of control (eg, how would you
shut down the script - you cant target exceptions at threads in Python).  It
would actually be even simpler to write a "srvany" in Python - it could use
the win32process functions to execute any arbitary program, and take
whatever action was necessary if it terminated prematurely.

Services get a decent going over in our book :-)

Mark.





From cjw at connection.com  Tue Jul 27 09:01:23 1999
From: cjw at connection.com (Colin J. Williams)
Date: Tue, 27 Jul 1999 09:01:23 -0400
Subject: Incompatibility.py
References: <379C8BB3.12D50F27@connection.com> <7niuqf$rpm$1@m2.c2.telstra-mm.net.au>
Message-ID: <379DADA3.33D7CD82@connection.com>

Mark,

I enter IDLE, from the W95 Desktop, with the Target set to IDLE.py
and the Start In set to the working directory.  Similarly, for PythonWin,
the Target is PythonWin.exe.  The results are as reported below.

At one time one could set breakpoints in IDLE, this capability was lost
in some later download.

I applaud the Tim Peters suggestion in this thread.  It would also be nice
if the IDLE line number were displayed, as is done in PythonWin.

Thanks,

Colin W.

Mark Hammond wrote:

> Colin J. Williams wrote in message <379C8BB3.12D50F27 at connection.com>...
>
> >IDLE returns:                        The name is: Incompatibility
>
> Not for IDLE 0.5 it doesnt.  Are you sure you are not using "Import Script".
> IDLE definately does this, as Python uses the "if __name__=='__main__':"
> idiom extensively - and IDLE would be useless for running scripts if it did
> as you describe.
>
> The behaviour you describe sounds like you are simply confusing "Run Script"
> with "Import Script".
>
> Mark.





From Mark.Hammond at p98.f112.n480.z2.fidonet.org  Sat Jul  3 19:34:38 1999
From: Mark.Hammond at p98.f112.n480.z2.fidonet.org (Mark Hammond)
Date: Sun, 04 Jul 1999 00:34:38 +0100
Subject: Waiting for an external process
Message-ID: <0000169e@bossar.com.pl>

From: "Mark Hammond" <MHammond at skippinet.com.au>

Juan M. Gonzalez wrote in message <377D9E05.E4F91D2F at globalserve.net>...
>Is there a better way to determine when an external process has finished
>than using
>win32process.CreateProcess / GetExitCodeProcess ?

Depends what you mean by "better".  Using win32process, it is possible to
have an event set when the process terminates, which can avoid polling.  Im
not sure if this meets your definition of "better" tho - certainly not
"simpler".

Mark.





From arcege at shore.net  Mon Jul  5 09:27:34 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Mon, 05 Jul 1999 13:27:34 GMT
Subject: Is it just me, or..
References: <000017cd@bossar.com.pl> <al84sjj8ukt.fsf@myntti.helsinki.fi>
Message-ID: <ap2g3.849$6M6.278663@news.shore.net>

Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
: "Michael P. Reilly" <Michael.P..Reilly at p98.f112.n480.z2.fidonet.org> writes:
:> From: "Michael P. Reilly" <arcege at shore.net>
:> Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
:> : Is there a bug in the handling of pipes with the Python? Esp. in
:> : non-blocking way.
:> : Example:
:> :  (r,w)=os.pipe()
:> :  r = os.fdopen(r, "r", 0)
:> :  w = os.fdopen(w, "w", 16384)
:> :  w.write('foobarbaz')
:> :  select.select([r],[],[]) # returns r, as it should
:> :  r.read(1) # returns 'f'
:> :  select.select([r],[],[]) # blocks!
:> 
:> : Now select _blocks_, despite there being obviously some data that at least
:> : should not have been buffered elsewhere. Also, fdopened stuff's reading
:> : features are bit annoying, as doing read(bigNumber) results in blocking
:> : call, as opposed to recv(bigNumber)'s semantics. Thus instead of pipes
:> : (=fast) I am using TCP/UNIX sockets (=slow) but they seem to work better
:> : for what I need. Am I missing something really obvious?
:> 
:> : This is on Linux, if it matters.
:> 
:> At first glance, I'd say that it is the buffering in the fdopen(w, "w",
:> 16384) call.  Put the bufsize to 0 or call the flush method.

: Oh, yes, I was flushing it (that example was done by hand out of existing
: code that is actually in use). Yet, still no dice. Also, buffer size for
: "r"does not matter (0,1,N); behavior is same always. Frustrating. (Oh
: well. UNIX sockets work, I suppose)

No, the buffer size of the writer ("w") should be 0.

Just tried the following on Linux (and on Solaris 2.6):

Python 1.5.2 (#1, Jun 15 1999, 01:24:40)  [GCC 2.7.2.3] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import sys, select
>>> import os
>>> os.uname()
('Linux', 'starship.python.net', '2.0.36', '#1 Tue Oct 13 22:17:11 EDT 1998', 'i586')
>>> (r, w) = os.pipe()
>>> rf = fdopen(r, 'r', 0)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
NameError: fdopen
>>> rf = os.fdopen(r, 'r', 0)
>>> wf = os.fdopen(w, 'w', 0)
>>> wf.write('foobarbaz')
>>> select.select([rf], [], [])
([<open file '(fdopen)', mode 'r' at 80e9cc8>], [], [])
>>> rf.read(1)
'f'
>>> rf.read(1)      -- success, why? no buffering in the writer
'o'
>>> rf.read()
[hang - what happened to 'obarbaz'?]

The data must be in the pipe or the read will hang.  A read(-1) will
always block (since there is never enough data).

Personally, I never trusted the C standard I/O library for anything
other than plain files.  But that's just me. ;)

I would suggest using the POSIX calls (os.read, os.write):

>>> (r, w) = os.pipe()
>>> os.write(w, 'foobarbaz')
9
>>> select.select([r], [], [])
([3], [], [])
>>> os.read(r, 1)
'f'
>>> os.read(r, 1)
'o'
>>>

A rule of thumb is: use Python file objects with objects created as
file object (popen, socket), but when using POSIX files, use the POSIX
functions, too many buffering and timing consideration need to be taken
care of directly.

You might also want to think about named pipes (os.mkfifo).

Tips&Tricks with UNIX pipes (named or otherwise):
* when the read side is closed, processes opened for writing will
  receive a SIGPIPE signal; this is important when using Python
  file objects which close when destroyed.
* when the write side is closed, processes opened for reading will
  receive an EOF.
* there is a OS imposed limit to how much data can sit in a pipe,
  be concious of this when you write your protocol.
* POSIX write is atomic, STDIO write is not necessarily atomic.
* reads must contain a length; the read will wait until all data
  is in the pipe - think about non-blocking reads:
    import fcntl, FCNTL; fcntl.fcntl(r, FCNTL.F_SETFL, FCNTL.O_NDELAY)

  -Arcege




From fredrik at pythonware.com  Fri Jul 30 04:38:01 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 30 Jul 1999 08:38:01 GMT
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <slrn7psisn.da.wtanksle@dolphin.openprojects.net> <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com> <slrn7pum9t.2ro.wtanksle@dolphin.openprojects.net> <001501bed935$561b8210$f29b12c2@secret.pythonware.com> <slrn7q1c10.5jp.wtanksle@dolphin.openprojects.net>
Message-ID: <00ad01beda66$d6bdb9b0$f29b12c2@secret.pythonware.com>

William Tanksley <wtanksle at dolphin.openprojects.net> wrote:
> >for ALL these statements, the else clause is exe-
> >cuted if and only if ALL controlling expressions
> >evaluate to FALSE.
> 
> But that's not even superficially true for the current behavior of a loop.
> A loop's else clause is executed even when there were TRUE results of the
> controlling expressions.

huh?  there's only ONE *controlling* expression in
the loop: the internal test that checks if the
sequence has any more items. the else clause
is executed if and only if THAT expression eval-
Traceback (innermost last):
  File "effbot.py", line 12032, in effbot
  File "effbot.py", line 3428, in generate_reply
    raise ImpedanceMismatch # bailing out
effbot.ImpedanceMismatch




From esr at thyrsus.com  Fri Jul 30 12:04:31 1999
From: esr at thyrsus.com (Eric S. Raymond)
Date: Fri, 30 Jul 1999 12:04:31 -0400
Subject: Need help with TCP/IP client access from Windows
In-Reply-To: <1278772913-26085625@hypernet.com>; from Gordon McMillan on Fri, Jul 30, 1999 at 11:49:38AM -0500
References: <1278785133-25350622@hypernet.com>; <19990730113931.D17721@thyrsus.com> <1278772913-26085625@hypernet.com>
Message-ID: <19990730120431.A17841@thyrsus.com>

Gordon McMillan <gmcm at hypernet.com>:
> > OK.  Will giving the filename ".\\intriguer" work for this?  And,
> > given that the filename is going through the C library, will leaving
> > it as "./intriguer" work?
> 
> Yes and yes. File Manager doesn't know what to make of an 
> extensionless file name, but that's OK. For years I created comma 
> delimited files as ".cdf". Then one day I installed IE 4, and they 
> magically became "Channel Definition Files".
> 
> This is good. It means that while Ionesco is dead, his spirit 
> lives on.

:-)

"Six Filenames In Search Of An Editor".  (Yes, I know, that's Pirandello.)

I've been considering using .ini format anyway; it's a good match to the
problem domain and the module to parse it is just lying there in 1.5.2.
So I guess I'll go with "./intriguer.ini".

Thanks!
-- 
		<a href="http://www.tuxedo.org/~esr">Eric S. Raymond</a>

"The state calls its own violence `law', but that of the individual `crime'"
	-- Max Stirner



From gmcm at hypernet.com  Mon Jul 19 12:52:27 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 19 Jul 1999 11:52:27 -0500
Subject: [Q] Accessors...
In-Reply-To: <7mvbkj$pl$1@feed.teaser.fr>
Message-ID: <1279723165-5561579@hypernet.com>

Olivier Deckmyn writes:
> 
> Is there a way to do the following :
> 
> class MyClass :
>     def __init__(self):
>         self._myAttr = None
> 
>     def get_myAttr(self):
>         return self._myAttr
> 
>     def set_myAttr(self, value)
>         self._myAttr = value # might do more interresting checkings
>         for
> examples
> 
> 
> and then use this like that:
> 
> m=MyClass()
> m.myAttr="foo"
> print m.myAttr

You need to get used to Python's dynamism.

class A:
  pass

a = A()
a.foo = 'bar'

If I've misunderstood your question, there was a recent thread
"Getters and Setters" about tricky ways to automatically create get
and set methods for attributes, so that given an instance of class A
with attribute foo, the user could write:

a = A()
a.getfoo()

without the programmer creating the getfoo method.

But if that's not what you want, don't go there!


- Gordon



From phd at emerald.netskate.ru  Fri Jul 16 08:10:38 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Fri, 16 Jul 1999 16:10:38 +0400 (MSD)
Subject: Python 1.5.2 + Linux 2.0.36 == Bad Combo?
In-Reply-To: <slrn7ou0ru.3sp.kas@titan.magnetic-ink.dk>
Message-ID: <Pine.LNX.4.04.9907161605560.16889-100000@emerald.netskate.ru>

Hello!

   I run python 1.5.1 and (mostly) 1.5.2 on many computers with different
hardware and/or software configurations, among them there are Linux 2.0.36
(RedHat 5.1 and (mostly) Debian 2.1). I write hundreds of programs and CGIs
and haven't found any minor problem with Python. (But I've never touched
the programs you mentioned.)
   Run stress-tests on you hardware, esp. memory...

On Fri, 16 Jul 1999, Klaus Alexander Seistrup wrote:
> I'm running Python 1.5.2 on a Linux 2.0.36 kernel (RedHat 5.1 distro partly
> upgraded to v5.2) and have recently experienced malfunctioning and/or non-
> working python applications -- specifically:
> 
>   *  PageCast
>      -- This one causes a Kernel Oops, though not a serious one.
>      The author, Preston Landers, have managed to invoke a similar
>      Kernel Oops using the same python/kernel combination.
>   
>   *  PyGCS
>      -- The program exits silently as soon as the second is loggin in.
>      The author, Jeff Blaine, hasn't been able to reproduce the behaviour.
>      I have asked Preston Landers to try out the program on his set-up
>      but have yet to hear the results
>   
>   *  DejaGrabber
>      -- The program connects without problem but seems to only retrieve
>      headers, not the full articles.
>      Haven't contacted the author yet.
> 
> I would like to hear if anybody else have experienced similar problems using
> the same Python / Linux kernel combination and, if so, whether is has helped
> shifting to another kernel version or downgrading python.
> 
> 
> Cheers,
> 
>   //Klaus
> 
> -- 
> ???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
> ???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From aahz at netcom.com  Thu Jul  8 13:08:12 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 8 Jul 1999 17:08:12 GMT
Subject: Efficient string manipulation , Help please
References: <37856BF1.BD046003@freenet.co.uk>
Message-ID: <7m2lts$oen@dfw-ixnews16.ix.netcom.com>

In article <37856BF1.BD046003 at freenet.co.uk>,
Sreekant Kodela  <sreekantk at freenet.co.uk> wrote:
>
>I seem to be in trouble here. I am trying to make some cgi scripts for
>my web site. I am using the postgresql as the backend db and a form
>based address book and news forum etc. The problem is everything is
>working ok as long as noone uses ' and " in the text. If they do, bingo
>, postgres refuses to accept. I tried using string library [mainly
>string.join(string.split('awkward'"'st'ring')). It works sometimes and
>doesn't some times.

It has problems with *both* ' and "?  IME with MS SQL Server, you should
only have problems with one of them.  In our code, we replace each
instance of "'" with "''".
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From jasbahr at mephisto.origin.ea.com  Tue Jul 27 12:03:06 1999
From: jasbahr at mephisto.origin.ea.com (jasbahr at mephisto.origin.ea.com)
Date: 27 Jul 1999 18:03:06 +0200
Subject: Hiring Python Talent
Message-ID: <7yso6agjwl.fsf@Mephisto.i-did-not-set--mail-host-address--so-shoot-me>

Electronic Arts (and its subsidiary Origin Systems) is looking to
hire talented Python programmers who are interested in working on 
massively multiplayer online games.  Top salaries, benefits, and 
relocation (to Austin, Texas) are included.  Please contact
Jeff Anderson of Origin Systems, Inc. at 512-434-6726.  Resumes
by email are encouraged, please send them to janderson at origin.ea.com


-- 
Jason L. Asbahr
Origin Systems, Inc.
jasbahr(@)origin.ea.com



From arcege at shore.net  Fri Jul  9 09:08:02 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Fri, 09 Jul 1999 13:08:02 GMT
Subject: range(Instance) vs. foo[Instance]
References: <slrn7o43v5.qkm.jepler@craie.inetnebr.com>
Message-ID: <Sumh3.1344$6M6.423882@news.shore.net>

Jeff Epler <jepler at inetnebr.com> wrote:
: Suppose I have a class with __int__ defined:

: class c:
:   def __init__(self, val):
:     self.val = int(val)

:   def __int__(self): return self.val

:>>> x=c(1)
:>>> range(x)
: [0]
:>>> (0,1,2)[x]
: Traceback (innermost last):
:   File "<stdin>", line 1, in ?
: TypeError: sequence index must be integer

: It seems desirable that the subscript operator call x.__int__ just like
: range() must effectively do.  Is there some reason why not?

: (Not the subscript operator in general, but the subscript operator for
: the builtin types of lists and tuples)

: I ask because I recently wrote a 5-minute hack on top of bytecodehacks that
: makes "integers" function as rational numbers in the interactive Python
: console---functions like "range" work, but subscripting of tuples and lists
: is broken.

: Jeff

Yup, this sounds like a bug, since [].insert( c(0), "spam") would work
as you expect.  Some reasons to not have this would be:
* How do you handle complex numbers as subscripts,
* Strings containing numbers as subscripts.

But this looks like a holdover from older code.  Myself, I'm not sure
if strings (containing a number) should be accepted.

I've submitted a patch to Guido and am including it here.  This will
make all built-in sequences (strings, tuples and lists) perform an
implicit seq[int(subscr)] if subscr is not an int.  The subscripting
will still fail if int() fails.

  -Arcege

Index: Objects/abstract.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Objects/abstract.c,v
retrieving revision 2.27
diff -c -r2.27 abstract.c
*** abstract.c	1999/01/10 16:56:58	2.27
--- abstract.c	1999/07/09 12:30:03
***************
*** 228,233 ****
--- 228,234 ----
  	PyObject *key;
  {
  	PyMappingMethods *m;
+ 	PyObject *tmp, *v;
  
  	if (o == NULL || key == NULL)
  		return null_error();
***************
*** 239,244 ****
--- 240,250 ----
  	if (o->ob_type->tp_as_sequence) {
  		if (PyInt_Check(key))
  			return PySequence_GetItem(o, PyInt_AsLong(key));
+ 		else if ((tmp = PyNumber_Int(key)) != NULL) {
+ 			v = PySequence_GetItem(o, PyInt_AsLong(tmp));
+ 			Py_DECREF(tmp);
+ 			return v;
+ 		}
  		return type_error("sequence index must be integer");
  	}
  
***************
*** 252,257 ****
--- 258,265 ----
  	PyObject *value;
  {
  	PyMappingMethods *m;
+ 	PyObject *tmp;
+ 	int v;
  
  	if (o == NULL || key == NULL || value == NULL) {
  		null_error();
***************
*** 264,269 ****
--- 272,282 ----
  	if (o->ob_type->tp_as_sequence) {
  		if (PyInt_Check(key))
  			return PySequence_SetItem(o, PyInt_AsLong(key), value);
+ 		else if ((tmp = PyNumber_Int(key)) != NULL) {
+ 			v = PySequence_SetItem(o, PyInt_AsLong(tmp), value);
+ 			Py_DECREF(tmp);
+ 			return v;
+ 		}
  		type_error("sequence index must be integer");
  		return -1;
  	}
***************
*** 278,283 ****
--- 291,298 ----
  	PyObject *key;
  {
  	PyMappingMethods *m;
+ 	PyObject *tmp;
+ 	int v;
  
  	if (o == NULL || key == NULL) {
  		null_error();
***************
*** 290,295 ****
--- 305,315 ----
  	if (o->ob_type->tp_as_sequence) {
  		if (PyInt_Check(key))
  			return PySequence_DelItem(o, PyInt_AsLong(key));
+ 		else if ((tmp = PyNumber_Int(key)) != NULL) {
+ 			v = PySequence_DelItem(o, PyInt_AsLong(tmp));
+ 			Py_DECREF(tmp);
+ 			return v;
+ 		}
  		type_error("sequence index must be integer");
  		return -1;
  	}



From phd at emerald.netskate.ru  Wed Jul 14 10:03:28 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Wed, 14 Jul 1999 18:03:28 +0400 (MSD)
Subject: WWW/urllib.urlretrieve problems
Message-ID: <Pine.LNX.4.04.9907141751130.9626-100000@emerald.netskate.ru>

Hello!

   I ran an URL checker, based on urllib, and it reported some errors. I
investigated what was going on, and I need some advices.


   First problem URL is http://www.expert.ru/ - it just timeouts.
When I pointed Netscape to the address - the page appeared well. Lynx and
telnet timed out too!
   Don't understand it. What in Netscape Communicator is so good?


   Second problem is with http://w3.one.net/~alward/. Netscape and lynx
showed the page, urllib.urlretrieve() and telnet returned error403 -
forbidden:

---------- Session ----------
phd at emerald 204 >>> t w3.one.net 80
Trying 206.112.192.125...
Connected to w3.one.net.
Escape character is '^]'.
GEt /~alward/ HTTP/1.0
Host: w3.one.net

HTTP/1.1 403 Forbidden
Date: Wed, 14 Jul 1999 13:48:03 GMT
Server: Apache/OneNet-W3 (LoadBal/2.1-s1) mod_perl/1.18 PHP/3.0.6
Connection: close
Content-Type: text/html

<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<HTML><HEAD>
<TITLE>403 Forbidden</TITLE>
</HEAD><BODY>
<H1>Forbidden</H1>
You don't have permission to access /~alward/
on this server.<P>
<P>Additionally, a 403 Forbidden
error was encountered while trying to use an ErrorDocument to handle the
request.
</BODY></HTML>
Connection closed by foreign host.
---------- /Session ----------

   It seems telnet and urllib need some additional HTTP/1.1 headers, but what
headers?


   Third problem is on http://www.tucows.com/. Netscape and lynx showed the
page, urllib returned exception:

---------- Session ----------
('http error', -1, '<html>\012', None)
Traceback (innermost last):
  File "./test.py", line 12, in ?
    fname, headers = urllib.urlretrieve(url)
  File "/usr/local/lib/python1.5/urllib.py", line 66, in urlretrieve
    return _urlopener.retrieve(url, filename, reporthook)
  File "/usr/local/lib/python1.5/urllib.py", line 184, in retrieve
    fp = self.open(url)
  File "/usr/local/lib/python1.5/urllib.py", line 157, in open
    return getattr(self, name)(url)
  File "/usr/local/lib/python1.5/urllib.py", line 272, in open_http
    return self.http_error(url, fp, errcode, errmsg, headers)
  File "/usr/local/lib/python1.5/urllib.py", line 289, in http_error
    return self.http_error_default(url, fp, errcode, errmsg, headers)
  File "/usr/local/lib/python1.5/urllib.py", line 295, in
http_error_default
    raise IOError, ('http error', errcode, errmsg, headers)
IOError: ('http error', -1, '<html>\012', None)
---------- /Session ----------

   Telnet session showed the server didn't return any HTTP header - just
sent HTML. Should urllib test for this and how it should behave? I am not
sure...

   Thanks in advance to anyone who is willing to discuss.

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From nekotron at pobox.com  Wed Jul  7 11:37:24 1999
From: nekotron at pobox.com (nekotron at pobox.com)
Date: Wed, 07 Jul 1999 15:37:24 GMT
Subject: win32api RegisterWindowMessage, GetMessage?
Message-ID: <7lvs7i$fri$1@nnrp1.deja.com>

Hi, I'm trying to port a Python script that I wrote on Mac that
automates the execution of another program.  On Mac, I'm using apple
events to drive the app and determine when the job is done.  On windows,
this program can use PostMessage to indicate when it's job is done.

The problem I'm having is that I see no way with Python to be able to
get this message.  I can use RegisterWindowMessage to create the message
and pass this to the program so it knows where to Post to, but I can't
figure out how to get this message.

is this possible?

thanks
-chris


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gmcm at hypernet.com  Sat Jul 10 14:34:12 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 10 Jul 1999 13:34:12 -0500
Subject: Win32 installer / "standalone" packager
In-Reply-To: <378781fe.0@news.dbtech.net>
Message-ID: <1280494591-86350953@hypernet.com>

Karl Putland writes:

> Installer install breaks ;-)
> 
> The traceback goes away too fast, but I think it says...
> 
> Traceback... yada-yada-yada
> File "launch.py", line 12, in ?
>     support = os.path.join(installdir, "support")
> NameError: installdir

If you have this problem, change the "idif" at line 10 of install.py 
to "installdir", then run "install.py". It should create a good 
installation.

Meanwhile, I will beg python.org's ftpmad^Hster to expedite getting 
the fix in (as long as the bribe is free, I'll pay it).

- Gordon



From markusk at bidra241.bbn.hp.com  Wed Jul 21 06:25:12 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 21 Jul 1999 12:25:12 +0200
Subject: Python and Boehm-Demers GC, I have code.
References: <000601bed335$893916e0$a72d2399@tim>
Message-ID: <p5hfmymh9z.fsf@bidra241.bbn.hp.com>

"Tim Peters" <tim_one at email.msn.com> writes:

> [Tim]
> >     for i in xrange(10000000):
> >         pass
> > [and RC vs GC implications]
> 
> [Markus Kohler]
> > True.
> > IMHO this is one of Pythons major design flaws.
> 
> What, specifically?

That flaw is that loops are not implemented using closures. 
Instead as far as I understand   loops in Python are implemented such that
the loop counter is created and destroyed for each step of the loop. 
Even if you optimize this heavily you will probably slower than with a design
that "recycles" the loop variable. 

You could compare this to tail recursive versus non-tail recursive functions. 

Also closures make the implementation of efficient loops more difficult they
have the advantage that loops and other control structures can be better integrated
into the OO concept. Smalltalk for example doesn't need Iterators for that reason. 

> 
> > There is no need to create a new object for each step of the loop.
> > Smalltalk has loops at least as flexible as Python and it doesn't
> > have this problem.
> 
> Sorry, I've missed what "the problem" is thought to be here.  Loop overhead
> is typically lareger in Python programs than in other languages, but is also
> typically trivial all the same.
> 
> > Loops in squeak (www.squeak.org) are already 2 to 3 times faster
> > than in Python.
> 
> All possible loops?  All loops you've personally timed?  One loop you timed
> in a moment of boredom <wink>?  An empty loop that counts to 10000000?  I
> don't doubt that Squeak is faster in many areas and for many reasons, but
> I've been running Python for most of the decade and reducing loop overhead
> is about the last thing I'd bother to aim at.

Lately ee have had already some examples in this group  where loop overhead was 
a major factor. 
I measured different kind of loops and even the more general while loop was faster
in squeak than anything else in python. 
Functions calls seems also to be much slower in Python ...


> 
> > ...
> > Is there anything that would prevent someone to reimplement loops
> > such that only one variable is allocated ?
> 
> In the loop above, there are actually two int objects created per iteration:
> the one explicitly created by xrange and bound to "i", and a hidden "loop
> counter" int object used to drive the for/__getitem__ protocol.  The latter
> can be (& has been, experimentally) reimplemented via an internal "mutable
> int" type created for that purpose, but it didn't buy enough to be worth
> adopting.  Most loop bodies contain more than "pass" <wink>, and the more
> work they do the less the fixed overhead matters.

Another problem with loops could be reference counting. 
I'm wondering how often a reference count get's changed when running a loop ... 

Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From nick.bower at ssec.wisc.edu  Sat Jul 17 17:50:14 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sat, 17 Jul 1999 16:50:14 -0500
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net> <378FE5F3.A8394D50@bellsouth.net>
Message-ID: <3790FA96.9403AA17@ssec.wisc.edu>

> Since you indicate you are a beginner, I would like to recommend to
> you O'Reilly's Learning Python by Mark Lutz and David Ascher for
> $23.96 plus shipping from Amazon.com.

When I started out on Python maybe 1.5 years ago, there were only 2
books available in Australia.  So thanks for mentioning this new
"Learning Python" book - it looks more like what I would have wanted at
the time.  But one thing - why the hell is there a rat on the cover??? 
Anyone??? ;)

> For a more advanced treatment O'Reilly's Programming Python is good.

Have to strongly disagree there :)  This book is soooo long-winded.  I
think the original poster said she had C++ experience, so would have
been confronted with the same reality as me - I need to skim all
this???  I had similar programming experience so found that this book
just didn't cater to these needs.  I actually got up and going on the
smaller web programming book before I bothered to buy Programming
Python.

> of these books are very good at presenting Python is an understandable
> way.

I would lean towards saying this is because Python is understandable,
rather than Programming Python being overly creditable <:)  ie  The
books don't need to be good when Python is as simple as it is imho.

nick



From john.michelsen at gte.net  Thu Jul 22 10:55:04 1999
From: john.michelsen at gte.net (John Michelsen)
Date: Thu, 22 Jul 1999 14:55:04 GMT
Subject: Is anybody aware of a tree GUI widget for tkinter or Pmw?
References: <3796E213.9F326412@eng.cam.ac.uk>
Message-ID: <chGl3.118$XG6.6791@dfiatx1-snr1.gtei.net>

>Could anybody please tell me if there exits any TREE control widget for
>tkinter or Pmw? I know that Pmw provides a rather rich set of
>'megawidgets' but the tree control has not been implemented yet as far
>as I know.


There is one at:

http://www2.zyvex.com/OpenChem/Widgets0.1.1.zip

Which also has a MDI, dockable toolbars, and a progress bar.
You can see how they look below:
http://www.worldforge.org/website/tools/mapEditor.html

John






From chad at vision.arc.nasa.gov  Tue Jul 27 18:24:41 1999
From: chad at vision.arc.nasa.gov (Chad Netzer)
Date: Tue, 27 Jul 1999 15:24:41 -0700
Subject: 1.5.2 for: else:
References: <Pine.LNX.4.04.9907271612140.9849-100000@emerald.netskate.ru> <19990727142943.V4068@xs4all.nl> <slrn7ps8ej.vm0.wtanksle@dolphin.openprojects.net>
Message-ID: <379E31A9.8566746@vision.arc.nasa.gov>

William Tanksley wrote:

> Well, I'm suprised.  I had also expected the else clause to run on empty
> loop.  It seems so natural, in line with the meaning of a loop (loop on
> this; otherwise do that).

    Hmmm, it does seem that "finally" might have been a more descriptive
choice (but historians may point out that "else" appeared before "finally"
and was thus used instead).  Of course, if one breaks out of the loop, they
might reasonably expect to avoid the "finally" statement (otherwise, it
becomes redundant), and this is somewhat counter to it's "try" semantics...
Hmmm.





From emile at fenx.com  Sat Jul 31 09:11:14 1999
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 31 Jul 1999 06:11:14 -0700
Subject: os.name
References: <1278785134-25350540@hypernet.com> <V5ho3.4340$VL2.463886@news.direcpc.com> <001e01beda8e$2111d2e0$f29b12c2@secret.pythonware.com> <Ggto3.4361$VL2.467372@news.direcpc.com> <7nuac5$3rs$1@chief.death.de>
Message-ID: <RECo3.4365$VL2.468864@news.direcpc.com>

It sure would.  Any idea how you would get the result back into python
without having the dos command windows flash up on the screen?

but-the-you-wouldn't-have-the-fun-of-parsing-the-file-ly yr's

--

Emile van Sebille
emile at fenx.com
-------------------


Felix Puetsch <puetsch at gmx.de> wrote in message
news:7nuac5$3rs$1 at chief.death.de...
> emile at fenx.com (Emile van Sebille) wrote in
> <Ggto3.4361$VL2.467372 at news.direcpc.com>:
>
> >By the way, in windows at a command prompt:
> >
> >95,98:    net config
> >nt:       net config [server|workstation]
> >
> >gives a line that starts : Software version
>
> Wouldn't "ver" be easier?
>
> --
> Felix P?tsch





From Vladimir.Marangozov at inrialpes.fr  Tue Jul 20 14:58:15 1999
From: Vladimir.Marangozov at inrialpes.fr (Vladimir Marangozov)
Date: Tue, 20 Jul 1999 19:58:15 +0100
Subject: IDLE's PyShell
Message-ID: <3794C6C7.9A212A97@inrialpes.fr>

Just noticed that IDLE's interactive shell eats an unusually big
amount of memory. Just pressing the RETURN key a couple of times,
or even better ;-) -- closing and reopening the PyShell window
does not seem to reuse the prevously allocated memory...

I'm very far from this code, so don't blame me for not proposing a
patch, but here are two things I experimented with that seem to help
partially:

1. Clear the linecache when closing the PyShell window
2. Finalize properly the undo machinery in EditorWindow
   (yes -- the XXX sign is there to remind that the _close
    method needs more work regarding undo)
   For instance, I tried to finalize undo before self.io.close
   which seemed to work.

This helps a bit, but still a repeted open/close of PyShell eats
(less) memory. Sorry for not being able to help you more at this stage,
IDLE hackers! (ah, forgot to mention -- I refer to the latest version,
the one in the CVS tree).

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov at inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252



From emile at fenx.com  Tue Jul 20 09:08:29 1999
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 20 Jul 1999 06:08:29 -0700
Subject: [Q] __methods__ and __members__ special functions...
References: <7n1n7c$7ao$1@feed.teaser.fr>
Message-ID: <Cz_k3.3779$VL2.411701@news.direcpc.com>

Hi Olivier,

It looks like dir(a.__class__) returns the methods available to your
instance, although you may want to iterate through a.__class__.__bases__
if available as well.

--

Emile van Sebille
emile at fenx.com
-------------------


Olivier Deckmyn <olivier.deckmyn at mail.dotcom.fr> wrote in message
news:7n1n7c$7ao$1 at feed.teaser.fr...
> Hi all !
>
> I want to know all the methods available to a instance.
>
> Here is a copy of the official FAQ :
> """
> 4.60. Why doesn't dir() work on builtin types like files and lists?
> It should have -- and it does starting with Python 1.5 (currently in
> development -- see Questions 1.13 and 2.10).
> Using 1.4, you can find out which methods a given object supports by
looking
> at its __methods__ attribute:
>
>
>     >>> List = []
>     >>> List.__methods__
>     ['append', 'count', 'index', 'insert', 'remove', 'reverse',
'sort']
> """"
>
> if I make the test with [].__methods__ it works as expected.
> But when I use my own object :
>
> class ClassA:
>     def __init__(self):
>         pass
>
>     def myMethod(self, a)
>         return a*2
>
> if I try __methods__ on a fresh instance :
>
> a=ClassA()
> a.__methods__
>
> this raises a NameError exception :(
>
> Why ? Where is __methods__ declared ?
> Why does this work for any instance in the brower.py provided with
pythonwin
> ?
>
> Please help !
>
> Olivier.
>
>
>
>
>





From lvirden at cas.org  Fri Jul  2 08:35:02 1999
From: lvirden at cas.org (lvirden at cas.org)
Date: 2 Jul 1999 12:35:02 GMT
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com>
Message-ID: <7liblm$k0f$1@srv38.cas.org>

According to Paul Duffin  <pduffin at mailserver.hursley.ibm.com>:
:in Python and YOU find it easier than doing it in Tcl. I do not
:believe that you can do everything in Python that you can do
:in Tcl (at least as regards extending the language itself).

Paul, what about the opposite - can one do everything in Tcl that can
be done in Python - or are there any fundamental features of Python
which could not be added into Tcl via an extension?

-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.



From johnston.p at worldnet.att.net  Sun Jul 18 15:00:59 1999
From: johnston.p at worldnet.att.net (Paul Johnston)
Date: Sun, 18 Jul 1999 12:00:59 -0700
Subject: what is python?
Message-ID: <7mt81n$s6e$1@bgtnsc03.worldnet.att.net>

Hello people.

I'm a relatively new programmer with a pretty strong java background.
Although I've never used python, I keep hearing more and more about it, and
I can't help but wonder what I'm missing.  I want to learn it, but I think
the fastest way is to put it to work right away.  Therefore, I guess I'm
trying to figure out what python excels in and then I'll allocate a
problem-domain (in my own work) to apply it to -- hopefully this will be the
most pragmatic approach.

So what are people using python for?  What types of problems is python
better suited relative to, say, java?  I do alot of server-side programming
& xml stuff.

Thanks for your input,

Paul






From da at ski.org  Wed Jul 14 12:12:44 1999
From: da at ski.org (David Ascher)
Date: Wed, 14 Jul 1999 09:12:44 -0700 (Pacific Daylight Time)
Subject: togl widget?
In-Reply-To: <FeWi3.824$Dk6.3405@news.flash.net>
Message-ID: <Pine.WNT.4.04.9907140912170.212-100000@rigoletto.ski.org>

On Wed, 14 Jul 1999, Tim Hamza wrote:

> I am trying to track down the togl widget.  The link from the Starship
> OpenGL page is dead.

It's moved to: http://www.cs.umd.edu/~bederson/Togl.html

--david ascher

(yes, i'll update the starship page)




From nascheme at ucalgary.ca  Tue Jul 20 00:56:38 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 20 Jul 1999 04:56:38 GMT
Subject: Cyclops 0.9.4
References: <000c01bed257$84f52a80$139e2299@tim>
Message-ID: <7n0vi6$75c@ds2.acs.ucalgary.ca>

Tim Peters <tim_one at email.msn.com> wrote:
>Cyclops.py implements a CycleTracker class that provides major help in
>finding and analyzing runtime cycles in Python programs.

Cool.  How about a feature that displays only cycles created
since a marking function was called.  I think this would be most
useful in programs that leak memory.  You don't care if cycles
exist, you just don't want them to be continually created.  The
way I see this working is you start up something like IDLE, do
some stuff, call the marking function, do more stuff, call cycle
showing function.  Anyhow, thanks for this tool.


    Neil



From jhendry at shrike.depaul.edu  Thu Jul  1 14:53:32 1999
From: jhendry at shrike.depaul.edu (Jonathan W Hendry)
Date: 1 Jul 1999 12:53:32 -0600
Subject: Installing Python/XML on Windows?
Message-ID: <377bab1c@news.depaul.edu>

Does any documentation exist on this? How do I set up the
paths so that Python can use all the various components of
the package?

Thanks,

- Jon



From Sven.Drescher at dlr.de  Mon Jul 26 10:43:12 1999
From: Sven.Drescher at dlr.de (Sven Drescher)
Date: Mon, 26 Jul 1999 16:43:12 +0200
Subject: problems with signal-handler...
Message-ID: <7nhrda$doi@news.go.dlr.de>

Hallo!

I have a little problem and hope, there is an easy answer to solve it.

I wrote a class 'process'. With this class I can start other programms and
commands. I installed a signal-handler to catch the signals send by any
processes.
If a process will be killed or stops, it sends the signal 'SIGCHLD' to
parent. To chatch that I have the handler.
But my problem is, I cannot distinguish which process sends the signal. Only
the signal-handler of the last started process will be active and if any
process sends a signal, the handler only acts for the last process and not
for the right one.

Is there a possibility to catch the pid from the process, which sends the
signal?
Is there an other way to get what I want?

Thanks for help and comments!

Sven

______________________________________
German Aerospace Research Establishment
Institute of Flight Guidance
Air Traffic Management System
e-mail: Sven.Drescher at dlr.de






From chad at vision.arc.nasa.gov  Sat Jul 24 02:07:30 1999
From: chad at vision.arc.nasa.gov (Chad Netzer)
Date: Fri, 23 Jul 1999 23:07:30 -0700
Subject: int(long(-sys.maxint-1)) fails on Linux
References: <000f01bed58c$4abb9e20$832d2399@tim>
Message-ID: <37995821.8D8A3363@vision.arc.nasa.gov>

Tim Peters wrote:

> Anyway, the lea is simply using a scaled-index addressing mode, with a 0
> base, a 0 offset, and with %edx treated as a "scale by 2" (shift left by 1)
> scaled index.
>
> Check out this link:
>
> http://www.comp.it.bton.ac.uk/burks/burks/language/asm/artofasm/artof001.htm
>

Thanks for the reference, Tim.  I read it for ~20 minutes, and was enlightened. :-)

Cheers,
Chad





From ze5yr at my-deja.com  Thu Jul  1 12:32:47 1999
From: ze5yr at my-deja.com (ze5yr)
Date: Thu, 01 Jul 1999 16:32:47 GMT
Subject: theads & global namespaces
References: <377A5C88.32A1FAB6@notes.cba.ufl.edu> <7ldqi2$nc1$1@nnrp1.deja.com>
Message-ID: <7lg57f$i7p$1@nnrp1.deja.com>

In article <7ldqi2$nc1$1 at nnrp1.deja.com>,
  rdev at my-deja.com wrote:

> - If you're running on Solaris, the thread won't get to run
> unless you give up some time.  So try putting a time.sleep(.001)
> into your while loop.  On NT, this is not necessary (don't know
> why.)

It works on NT, but the looping thread will eat up horrendous amounts
of CPU time (at least it does if under pthreads, don't know if Python
threads would do the same).  So it's a good idea to use time.sleep
under NT as well.

--
Cliff Crawford                           -><-
http://www.people.cornell.edu/pages/cjc26
                            empty-handed I go,
)O(       and behold the spade is in my hands


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From greg.ewing at compaq.com  Wed Jul  7 23:29:04 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 08 Jul 1999 15:29:04 +1200
Subject: Slicing beyond list --  bug or featurette?
References: <377D234A.4C9BDAC0@cisco.com> <m3673wub7a.fsf@chinon.cnrs-orleans.fr>
Message-ID: <37841B00.D37920A2@compaq.com>

Konrad Hinsen wrote:
> 
>   Sequences also support slicing: a[i:j]selects all items with index k
>   such that i <=k < j.

That definition could apply to dictionaries,
too! Pity it doesn't...

Greg



From rmuschall.fih at t-online.de  Wed Jul 14 19:29:07 1999
From: rmuschall.fih at t-online.de (Ralf Muschall)
Date: Thu, 15 Jul 1999 01:29:07 +0200
Subject: console show in Python?
References: <7m9tmi$6qg$30@rks1.urz.tu-dresden.de>
Message-ID: <378D1D43.7FD2578F@t-online.de>

Scott Shillcock schrieb:

> I am moving to Python from Tcl/Tk.
> Is there something similar to the Tcl/Tk 'console show' command in
> Python.

A simple method is the following:
Make a Toplevel with an Entry and a Text, and bind <Key-Return>
to a function which takes the string from the Entry, looks
whether it starts with "p ", and print(eval(...)) if yes,
exec(...) otherwise.

This is far from perfect, but helped me.

Ralf



From arcege at shore.net  Wed Jul 14 21:24:03 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 15 Jul 1999 01:24:03 GMT
Subject: Expand tuple into argument list?
References: <87aesy4w3f.fsf@solano.in.renlabs.com>
Message-ID: <TKaj3.23$ui4.9684@news.shore.net>

Steven Work <steve at renlabs.com> wrote:
: I'm new to Python.  I want a dispatch function -- takes a function
: name and other arguments, calls the named function with the remaining
: arguments.  This works for limited length argument lists:

: def dispatch(fnname, *rest):
:     fn = dispatch.func_globals[fnname]
:     if len(rest)==0:
:         return fn()
:     elif len(rest)==1:
:         return fn(rest[0])
:     elif len(rest)==2:
:         return fn(rest[0], rest[1])
:     elif len(rest)==3:
:         return fn(rest[0], rest[1], rest[2])
:     elif len(rest)==4:
:         return fn(rest[0], rest[1], rest[2], rest[3])

: but is there a general way to expand the tuple into an argument list?
: (I know about keyword arguments already.)

: Lest you ask why: I'm embedding a Python interpreter in a C program; I 
: need to call Python functions directly from C, and I'd like a single
: point of contact to minimize the required housekeeping.

Inside Python you can replace the if-elif construct with apply:
  def dispatch(fnname, *rest):
    fn = dispatch.func_globals[fnname]
    return apply(fn, rest)

But from C, you will want to use PyObject_CallObject:
  result = PyObject_CallObject(fn, args);
(http://www.python.org/doc/current/api/object.html)

  -Arcege




From d at d.dd  Mon Jul 26 05:08:08 1999
From: d at d.dd (nom)
Date: Mon, 26 Jul 1999 11:08:08 +0200
Subject: accent (special char) with python
Message-ID: <932980082.887262@sofia.magic.fr>

hi

i'm from france then i use accent : i try to convert a string :
'Soci\351t\351' to 'Soci?t?'
how can do it?
thanks





From tismer at appliedbiometrics.com  Fri Jul 16 05:28:14 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Fri, 16 Jul 1999 09:28:14 GMT
Subject: Meta-language for mxTextTools
References: <011e01becf6c$07c2d820$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <378EFB2E.DB50EC6C@appliedbiometrics.com>


"Tony J. Ibbs (Tibs)" wrote:
> 
> Christian Tismer wrote:
> >I remember very well! BCPL is mentioned very rarely on this list!
> 
> It is/was a Neat Language - my project at university was written in BCPL,
> and we still used it at Laser-Scan for various utility programs for several
> years after I first joined the company (back in, ye gods, 1980-ish). On the
> other hand, getting lectured by Martin Richards was *not* terribly
> interesting.

At that time, I was involved in a SNOBOL4-ish patternmatcher for
an editing system written in - well, bicipiell.

> Ah, the backwards steps that C took:
> 
> "OK - first off, let's reintroduce the dangling else problem. Oh, and whilst
> we're at it, let's use block delimiters that are hard to distinguish from
> parentheses on just about all available printers and terminals [of the time,
> and often still!]."

Whow, yes. That answers the open question which language
should be used for Python 2000. Not C, not C++, but BCPL.
Spitting out C-- assembly code, of course.
(he last one is no joke but the latest work of Peyton Jones)

> Humph. Now, anyone for Algol68?

Sorry to dissappoint you: me again. My very first language
was A68. My second one was Fortran IV.

> Tibs
> 
> (time for a plug for the Retrocomputing Museum? I have a link to a mirror
> at:
>     http://www.pdc.kth.se/~jas/retro/
> They have a BCPL compiler, although I haven't tried it...)

Thanks for the good link, I *will* try - ciao - granny chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From cgw at fnal.gov  Fri Jul  2 14:08:28 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Fri, 2 Jul 1999 18:08:28 GMT
Subject: C shared object files
In-Reply-To: <199907021717.KAA03837@ogl.llnl.gov>
References: <199907021717.KAA03837@ogl.llnl.gov>
Message-ID: <14205.28.172064.730589@buffalo.fnal.gov>

I'm sure there are plenty of people on this list, including myself,
who'd be glad to help you.  As far as building an extension module
from your .c file, it would help to know what kind of platform you're
on.  If you're on a Unix platform with gcc, the command would look
something like:

gcc -I/path/to/python/includes -c myfile.c
gcc -shared -o mymodule.so myfile.o

As far as checking for presence of a module at runtime, the simplest
thing would be:

try:
	import module
except ImportError:
       #it doesn't exist...

Aren't there any local Python gurus at llnl.gov?






From mgushee at havenrock.com  Thu Jul 22 21:01:15 1999
From: mgushee at havenrock.com (Matt Gushee)
Date: 23 Jul 1999 10:01:15 +0900
Subject: Printable Python reference
References: <3797782A.FD39EF04@newbridge.com>
Message-ID: <m31ze0kwmc.fsf@havenrock.com>

Nick Taylor <ntaylor at newbridge.com> writes:

> I decided to take a look at Python today for the first time, and it
> sounds very interesting.  

It is!

> Does anyone know of a printable (i.e. PDF,
> PostScript, or plain-text) reference out there?

Try ftp://ftp.python.org/pub/python/doc . You'll find PS, PDF, and
LaTeX documents. They're packaged as .tar.gz archives, though, which
might be a problem if you run Windoze. Maybe there are ZIP versions
somewhere.

By the way, once you get through or around the tutorial, the Library
reference is probably the one you'll use most.

Best of luck w/ Python.

Matt Gushee
Portland, Maine, USA



From fredrik at pythonware.com  Thu Jul  8 04:02:18 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 8 Jul 1999 08:02:18 GMT
Subject: Bizarre Tkinter radio button behaviour
References: <378415EC.9F7A87A5@compaq.com>
Message-ID: <007401bec918$34125f90$f29b12c2@secret.pythonware.com>

Greg Ewing <greg.ewing at compaq.com> wrote:
> I have some sets of radio buttons linked to
> Tkinter variables. I set the initial values of
> the variables, expecting the corresponding 
> buttons to come up selected when the window
> first appears.
> 
> But it doesn't happen.
> 
> I mess around trying different things. I
> add a command to one of the radio buttons to
> print out the values of the variables when
> they change. All of a sudden the buttons
> start appearing with default selections.
> 
> Great, I think, and take out my debugging code...
> and they stop working again!
> 
> After some more trial and error I find that
> if *one* of the radio buttons has a command
> attached -- even just an empty procedure --
> then *all* of the default buttons come up
> properly selected. Otherwise, none of them
> do.

sure sounds like a reference counting problem.
but if you want to know exactly what happens,
I think you have to post some code...

</F>




From kistler at fnmail.com  Thu Jul  8 04:48:41 1999
From: kistler at fnmail.com (Per Kistler)
Date: Thu, 08 Jul 1999 10:48:41 +0200
Subject: Two Pythons talking to each other?
References: <199907072203.AAA09601@axil.hvision.nl>
Message-ID: <378465E9.7762BE9A@fnmail.com>

Hi Hans

One could use sockets with AF_INET, if it runs on another
machine, and sockets with AF_UNIX, if it's on the the same
host.

You seems to separate strings by time. But it's better to have
a defined delemiter, like the typical \r\n (\012\015). If you
read from the socket, while for instance, 1.5 lines have arrived,
you just process the first line, and leave the rest in a buffer,
until the lineending of the second line has arrived.

Regards, Per.

Hans Nowak wrote:
> 
> Hello,
> 
> I am currently experimenting with networking and want to write some
> games (where, obviously, two players compete over a network). Now I
> want to test my programs offline. One way would be to start two
> separate Python sessions (on the same machine), both running the same
> game program, and let them somehow talk to each other like they would
> over a network. But, what would be a good way to do this? I was
> thinking about passing strings (commands etc) to a file, but two
> programs cannot read from and write to the same file at the same
> time. Does someone have any ideas about how to do this?
> 
> TIA,
> 
> --Hans Nowak (ivnowa at hvision.nl)
> Homepage: http://fly.to/zephyrfalcon

-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net
------------------------------------------------------------



From posselt at daimi.au.dk  Tue Jul  6 08:44:35 1999
From: posselt at daimi.au.dk (Peter Posselt Vestergaard)
Date: Tue, 06 Jul 1999 14:44:35 +0200
Subject: changing some lines in a file
Message-ID: <3781FA33.92876E30@daimi.au.dk>

Hi, I'm quite a newbie in Python-programming, so I hope you'll forgive
me asking about some pretty things. :-)
My problem is that I have a file in which i want to find some lines and
replace them with some other data. I've tried a lot of things but
nothing seems to work.
Below is my best try, but it seems that the script stops when it is
about to execute the line:

line=('fi='+Form["fi"].value)

I'll be thankfull for any kind of help.
Best regards

Peter Vestergaard

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


db = open("db.dat","r+")						 
data = db.readlines()
db.close

found=0
for line in data:
    print found
    if found==1:
        if line[0:2]=='fi':
            line=('fi='+Form["fi"].value)
        elif line[0:2]=='la':
            line='la='+Form["la"].value
        elif line[0:2]=='em':
            line='em='+Form["em"].value
        elif line[0:2]=='ho':
            line='ho='+Form["ho"].value
        elif line[0:2]=='st':
            line='st='+Form["st"].value
            found=2
    if line[0:9]==('id='+Form["id"].value):
        found=1
if found==0:
    print '<H1>Fejl i ?rskortnummeret!</H1>'
    stop 

db = open("db.dat","w")
db.writelines(data)
db.close



From bkc at Murkworks.com  Mon Jul 19 11:59:34 1999
From: bkc at Murkworks.com (Brad Clements)
Date: Mon, 19 Jul 1999 11:59:34 -0400
Subject: Swig typemaps and me == bananas - can't override a previous %apply
Message-ID: <7mvi1f$nn4$1@news.clarkson.edu>

I'm encapsulating Novell's NDS API using swig 1.1-685. So far it's going
well and I have about 177 functions wrapped.

One annoying problem I have is figuring out which typemap has been applied
for a particular function. There doesn't seem to be any debug output that
will say something like:

int NWDSLogin(
    NWDSContextHandle context,  /* %apply NDSContext ndscore.i line 77 */
   pnstr8 userid,    /* %apply char *OUTPUT ndsmap.i line 128 */
)

For example, I have this in a .i file

----
%apply STRINGOUT {pnstr8 canonName};
%apply char *INPUT {pnstr8 objectName};

NWDSCCODE
NWDSCanonicalizeName
(
   NWDSContextHandle context,
   pnstr8            objectName,
   pnstr8            canonName
);
------

The above doesn't work. objectName is treated as in output parameter
(STRINGOUT is applied to it).

To make this work, I had to do this:
---
%apply STRINGOUT {pnstr8 canonName};

NWDSCCODE
NWDSCanonicalizeName
(
   NWDSContextHandle context,
   pnstr8            INPUT, /*  objectName, */
   pnstr8            canonName
);
---

Anyone have any ideas on why I can't override the typemap? I've tried adding
this before the function:

%clear STRINGOUT;
%clear STRINGOUT;
%clear pnstr8;
%clear pnstr8;


But that doesn't help. (I saw in the source some mention that typemaps are
stacks that are pushed on %apply and pop'd on %clear)

Does anyone have patches to swig that will produce better diagnostic output
to see how typemaps are applied?

Thanks,

--
Brad Clements,
bkc at murkworks.com





From aahz at netcom.com  Wed Jul  7 15:54:54 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 7 Jul 1999 19:54:54 GMT
Subject: What's ZOPE anyway?
References: <3scg3.1267$B95.91660@typ42b.nn.bcandid.com> <7m064a$921@journal.concentric.net> <slrn7o7b7u.6la.wtanksle@dolphin.openprojects.net>
Message-ID: <7m0bae$11l@dfw-ixnews4.ix.netcom.com>

In article <slrn7o7b7u.6la.wtanksle at dolphin.openprojects.net>,
William Tanksley <wtanksle at dolphin.openprojects.net> wrote:
>
>...and what's more is that it's EASY to use.  When I started I didn't know
>much HTML (only the basics).  Thanks to Zope, I can learn bit by bit.

It's a little more difficult if you're trying to do Real Work while
you're learning.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From akuchlin at mems-exchange.org  Mon Jul 12 10:15:42 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: Mon, 12 Jul 1999 10:15:42 -0400 (EDT)
Subject: Python re module
In-Reply-To: <7m5qev$lue$1@nnrp1.deja.com>
References: <199907120337.XAA06072@eric.cnri.reston.va.us>
Message-ID: <14217.63630.458437.222314@amarok.cnri.reston.va.us>

catlee at my-deja.com
>Try the following:
>import re
>re.search("((.)\\1+)","a")
>Now, I know this isn't proper syntax (you shouldn't reference a group
>inside itself), but on my machine python hangs.

	What's happening is a bit complicated.  When you refer to a
group from inside itself, it's treated as if the group has matched the
null string.  You thus wind up asking "how many times does the null
string match at this point?"  The answer is infinity, of course.  

       However, GvR reports that it isn't actually hanging, just
taking a long time to finish.  I think it escapes the infinite loop
because \1+ is actually treated the same as \1{0,INT_MAX}, so once
it's counted up to INT_MAX (2**31-1) = 2.1 billion, it then returns.
     
     The latest version of PCRE, v2.06, does not have a problem with
this (more checks for repeated null-string matches have been added),
so I think the problem will be fixed when the Python module is
resynchronized with the latest version.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
There is a difference between art and life and that difference is readability.
    -- Marian Engel, in the _Toronto Globe and Mail_, Dec. 28, 1974



From claird at starbase.neosoft.com  Wed Jul 21 08:01:22 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Jul 1999 12:01:22 GMT
Subject: Tcl/Python: Who was first?
References: <932552685.620225@news.adtranzsig.de> <3795ACD5.FCFF67A1@equi4.com>
Message-ID: <16E7930929E7C7E5.430278EC189FFB38.BB6793F4819F7C95@lp.airnews.net>

In article <3795ACD5.FCFF67A1 at equi4.com>,
Jean-Claude Wippler  <jcw at equi4.com> wrote:
			.
			.
			.
>Tcl-URL! is also tracked and stored on the Scriptics site and on Linux
>Weekly News, and a few others I believe.  If you scroll to the bottom of
>the first URL page just mentioned, there is a little blurb about the
>background of this.
>
>Cameron and I, both multi-lingual and scripting enthousiasts by choice,
>considered this simple idea to be easily graftable onto other scripting
>communities.  So we started a similar setup for Python, but somehow it
>never reached a stable weekly state.
>
>Which is a big, big, pity if you ask me.  All it would take is for one
>of the leading Pythoneers to take the Python-URL! stick for a month or
>so, and I'm sure we'd be having Python-URL! back on track in no time. 
>Soooo... GvR, TP, DA, MH, CT, GM, anyone want to bring this thing back
>on track?  I'm sure many others would be more than happy to follow your
Let's see; Ivan just finished his book, Barry
has Mailman and JPython and starship under con-
trol, ...
>lead and take things from there later on.
>
>If you're interested, please contact Cameron or me.  We can take care of
>all the details and the mechanics, and I'd be happy to commit to do just
>that for a long time, but the real issue here is that volunteers need to
>be willing to come up with a list of 5 to 10 one-liners and URLs they
>wish to highlight each Monday.
>
>Tcl-URL! works well IMO, because it gives an interesting peek into what
>key players choose and find important, and because it is gradually
>building up a stable set of links to main topics, announcements, and
>discussions.  It has become part of the Tcl folklore by now, I guess,
			.
			.
			.
Python-URL! gossip:  I've been close to resurrecting
it all month.  That is, there are so many good posts
to c.l.p, that I've been bursting with desire to
resume writing up the weekly digests.  As a personal
discipline, I've deferred allowing myself this plea-
sure until I wrap up a few other long-standing
obligations.

At least one well-known magazine wants to feature
Python-URL! in its pages.  There's a LOT more we can
do with the concept, once it regains its vigor and
autonomy.  Jean-Claude has slick ideas about making
tiny little searchable packages and cross-references
and ...  It was always quite popular when fresh.

My summary:  if you like Python-URL!, let us know,
and we'll show you how you can contribute to it at
perhaps negative net cost (weekly compilers regular-
ly report they get more out of the process than the
minimal time they invest in it).  We'll take care of
the infrastructure.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From janssen at parc.xerox.com  Mon Jul  5 21:02:15 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Tue, 6 Jul 1999 01:02:15 GMT
Subject: Python under Java?
Message-ID: <MrUJKLEB0KGWFViX4p@holmes.parc.xerox.com>

Has anyone tried embedding Python under Java, using JNI?  When I try it
(with JDK 1.2 on Solaris 2.6 and Python 1.5), I get SEGV inside
Py_Initialize().

Bill



From beall at psych.ucsb.edu  Sat Jul 17 16:27:39 1999
From: beall at psych.ucsb.edu (Andy Beall)
Date: Sat, 17 Jul 1999 13:27:39 -0700
Subject: importing a Win32 DLL module w/resources
Message-ID: <3790E73B.C969A223@psych.ucsb.edu>

Hi, I'm trying to extend Pythonwin from MSVC with a DLL module that I
can import, and everything is working except with I try and access some
resources, such as a dialog box that I made.  It seems that when the DLL
is built in MSVC, it doesn't bundle the .rc or .res file into the DLL. 
Anybody know a way to force the linker to bind this or how to let
Pythonwin know where these resources are?

Thanks,
Andy Beall



From phil at ricochet.net  Fri Jul  9 03:18:19 1999
From: phil at ricochet.net (phil)
Date: Fri, 09 Jul 1999 07:18:19 GMT
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com> <37850F5B.F80CB0A7@enews.nrl.navy.mil> <7m3ajc$s3q$1@m2.c2.telstra-mm.net.au>
Message-ID: <37859ef4.11675718@news.ricochet.net>

On Fri, 9 Jul 1999 09:03:03 +1000, "Mark Hammond"
<MHammond at skippinet.com.au> wrote:

>Scott Wolford wrote in message <37850F5B.F80CB0A7 at enews.nrl.navy.mil>...
>
>>msvcrt.lib and the application can't then bring in msvcrtd.lib. I do think
>>this is a problem because python15_d.lib is provided but no python15_d.dll
>>is. I would think you'd provide (at the very least) python.lib linked with
>>msvcrt.lib, and python_d.lib linked with msvcrtd.lib (and the dlls), or
>>(to be really nice :) all four combinations.
>
>The main reason we dont is size - we would need to ship the _d, and for
>people to really use it, also the .pdb.  These add too many megs.

This bit me as well. The confusion i had was that the _d.lib was
available but not the _d.dl. Couldn't there be two downloads, one with
the optimised/non-debug and one without? I rather like the
msdog/widow's way where no compiling is involved and suspect that it
is the future. Easy for me to say, i know..

No problemo either way'ly,

phil.




From nathan at islanddata.com  Tue Jul 13 14:43:40 1999
From: nathan at islanddata.com (Nathan Clegg)
Date: Tue, 13 Jul 1999 11:43:40 -0700 (PDT)
Subject: MySQL Module
In-Reply-To: <19990713122901.A2538@pooh.frostnet.net>
Message-ID: <XFMail.990713114340.nathan@islanddata.com>

Check out http://www.mysql.com/Contrib/MySQLmodule-1.4.tar.gz
I haven't tried that one myself.  I use MySQLdb which you can get at
http://starship.python.net/crew/adustman/MySQLdb.html.  It's very recent
and seems to be kept up to date.

On 13-Jul-99 Chris Frost wrote:
> Are than any modules for connecting to a MySQL sever? I didn't see
> anything in the contributed section of python.org. (If there aren't,
> what
> other databases do people think are better, why?)


----------------------------------
Nathan Clegg
 nathan at islanddata.com





From rodisi01 at my-deja.com  Tue Jul 27 18:54:02 1999
From: rodisi01 at my-deja.com (Isidor)
Date: Tue, 27 Jul 1999 16:54:02 -0600
Subject: String2List: odd function
In-Reply-To: <001701bed395$16de7c60$f29b12c2@secret.pythonware.com>
Message-ID: <199907272152.QAA01065@cro.ots.ac.cr>

Hello all -

In response to my nice offer of a strange function that in ~80 lines 
of code slices and dices your list-like strings into *real* lists, I 
received three nice replies pointing out that the same thing could be 
done in ~1 line. Wow. Talk about optimization!:

From:           	Gaetan Corneau <Gaetan_Corneau at baan.com>

> >>> s = '[1, 2, 3, 4, 5, ["a", "b", "c"], 6, 7, 8]'
> >>> s = "l = " + s
> >>> exec(s)
> >>> l
> [1, 2, 3, 4, 5, ['a', 'b', 'c'], 6, 7, 8]

From:           	"Fredrik Lundh" <fredrik at pythonware.com>

> v = eval(s, {"__builtins__": {}})


From:           	"Emile van Sebille" <emile at fenx.com>
 
> >>> a = eval ("[['itema1', 'itema2'], 'itemb1', ['cheese1',
> ['cheese1a','cheese1b'], '2'], 'blah', 'rest of itemb2', 'zib']")
> >>> a
> [['itema1', 'itema2'], 'itemb1', ['cheese1', ['cheese1a', 
'cheese1b'], 
> '2'], 'blah', 'rest of itemb2', 'zib']
> 

When I read these messages, the first thing I thought was: "wow! 
*that's* what exec and eval do?!" I had read about them in the 
documentation, and I suspected that they were important, but I 
couldn't figure out what the heck they really did. A few unsuccessful 
experiments didn't help much either. Well, now I know a little more. 
However, I suspect that this use is not among the common uses for 
which exec and eval were designed. Would you be willing to reply with 
some examples of how you use these functions normally? What are they 
*really* used for? (As I collect more of these examples from others 
and from my own banging around, they will go on a web page w. 
credits, of course, to the example providers...and then someday...I 
will join the PSA...and then...and then...)

Thanks for setting me straight, and thanks in advance for any 
examples!

Isidor



From ikezsbom at eecs.tufts.edu  Fri Jul 23 10:36:45 1999
From: ikezsbom at eecs.tufts.edu (Ian Kezsbom)
Date: Fri, 23 Jul 1999 10:36:45 -0400
Subject: Jpython Question
In-Reply-To: <wkr9m3846k.fsf@ifi.uio.no>
References: <Pine.GSO.4.05.9907191525580.20433-100000@allegro> <wkr9m3846k.fsf@ifi.uio.no>
Message-ID: <Pine.GSO.4.05.9907231035200.14333-100000@allegro>

Just out of curiosity...lets say I've written a Java program, with many
classes and I want to run that program through JPython and have access to
all its classes...how would I go about doing this...

Ian

On 20 Jul 1999, Lars Marius Garshol wrote:

> 
> * Ian Kezsbom
> | 
> | I was just wondering if jpython can instantiate java classes and
> | invokes methods on them...any help would be welcome...
> 
> It can, and it's very easy too. Here's a snippet from an email I sent
> earlier today:
> 
> C:\Mine dokumenter>jpython
> JPython 1.1beta1 on java1.1.7B
> Copyright (C) 1997-1999 Corporation for National Research Initiatives
> >>> from java.util import Hashtable
> >>> h=Hashtable()
> >>> h.put("larsga at ifi.uio.no","Lars Marius Garshol")
> >>> h.get("larsga at ifi.uio.no")
> 'Lars Marius Garshol'
> >>>
> 
> It's actually as easy as this.
> 
> --Lars M.
> 
> 




From godzilla at netmeg.net  Fri Jul 23 19:38:22 1999
From: godzilla at netmeg.net (Les Schaffer)
Date: Fri, 23 Jul 1999 23:38:22 GMT
Subject: Python documentation available as ZIP files
In-Reply-To: <14232.57133.949353.741301@buffalo.fnal.gov>
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
	<14232.48512.648609.496768@lrz.de>
	<14232.48816.993092.9754@weyr.cnri.reston.va.us>
	<m2ogh39l9w.fsf@netmeg.net>
	<14232.57133.949353.741301@buffalo.fnal.gov>
Message-ID: <14232.64750.246232.405333@gargle.gargle.HOWL>

> I agree with you.  But shouldn't that be "print self.smile" ?  I got
> a NameError when I tried to run your code ;-)

damn, i cant even tell a joke without screwing up the punchline!


> But seriously, this is a mistake I've seen a lot of people make with
> class variables, just want to make sure we don't spread around any
> misconceptions.

unfortunately, i didnt -- ahem -- test the class before posting. shame 
on me.

actually, i thought the interpreter would crash on the '$' in $0.02. i
figured the dollar sign wouldnt have been allowed. shows ya what i
know.

happy weekend everyone!

les

-- 
____        Les Schaffer              ___| --->> Engineering R&D <<---
Theoretical & Applied Mechanics          |  Designspring, Inc. 
Center for Radiophysics & Space Research |  http://www.designspring.com/
Cornell Univ.  schaffer at tam.cornell.edu  |  les at designspring.com



From k_kal at my-deja.com  Tue Jul 13 11:48:33 1999
From: k_kal at my-deja.com (K Kal)
Date: Tue, 13 Jul 1999 15:48:33 GMT
Subject: OK python gurus!
Message-ID: <7mfn4c$m80$1@nnrp1.deja.com>

Hey Guys,
     I'm new to python.  Where should I get started?  Any good
references on-line.  Tutorials/sample code would be appreciated.

K.Kal

--
/**************
Keep On Smiling
**************/


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From cbarker at jps.net  Fri Jul 23 16:24:09 1999
From: cbarker at jps.net (Chris Barker)
Date: Fri, 23 Jul 1999 13:24:09 -0700
Subject: Xemacs Python Mode.
References: <199907231949.PAA07568@python.org>
Message-ID: <3798CF69.1FC04C23@jps.net>

Hi,

Does anyone know if I can set emacs mode to indent with ONLY tabs? (and
if so, how?) While I'm at it I'd like the tabs to show up as four spaces
while in Python mode also.

Personally, I prefer the only spaces method, but I'm transfering code
between Linux and the Mac a lot, and all the Mac editors seem to use
tabs. combining the two makes for a big mess if you use different
editors, so I guess I need to just go with all tabs.

-Thanks,

-Chris




-- 
Christopher Barker,
Ph.D.                                                           
cbarker at jps.net                      ---           ---           ---
http://www.jps.net/cbarker          -----@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Water Resources Engineering       ------   @    ------   @   ------   @
Coastal and Fluvial Hydrodynamics -------      ---------     --------    
------------------------------------------------------------------------
------------------------------------------------------------------------



From fredrik at pythonware.com  Tue Jul  6 12:36:49 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 6 Jul 1999 16:36:49 GMT
Subject: problem with lists
References: <378220B1.9E0C272@daimi.au.dk>
Message-ID: <004a01bec7cd$bfdbd1b0$f29b12c2@secret.pythonware.com>

Peter Posselt Vestergaard wrote:
> It seems like I've completely misunderstood something or at least done
> something wrong. There's noway I can see why these two scripts are
> outputting nothing or more correctly only the 4 first lines which cannot
> be seen in a browser. Anybody knows why?

> ----script 1------------------------------------------------
>
> #!/usr/local/bin/python
> import cgi
>
> print "Content-type: text/html"
> print
>
> print "<HTML><HEAD>"
> print "<TITLE> ?ndring af data </TITLE>"
> print '</HEAD><BODY BGCOLOR="#FFFFFF"><FONT FACE="Helvetica">'
>
> nydata.insert(0,'hej')

nydata must exist before you can insert things into it.
if it doesn't, you get a NameError exception (which
probably ends up somewhere in your server logs...)

by the way, you would have noticed this yourself if you
had executed this script from the command line...

anyway, to fix this, change this line to:

nydata = ['hej']

(this creates a single-item list)

> print nydata[0]
> print 'strange'

well, not really...

>
> print '</BODY></HTML>'

</F>




From MHammond at skippinet.com.au  Wed Jul 21 18:35:05 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Thu, 22 Jul 1999 08:35:05 +1000
Subject: Repost: Re: Scientific Plotting?
References: <379392E7.EB0F29E7@ssec.wisc.edu> <7n5fhn$9e5$1@nnrp1.deja.com>
Message-ID: <7n5hrq$4vv$1@m2.c2.telstra-mm.net.au>

mhuster at hotmail.com wrote in message <7n5fhn$9e5$1 at nnrp1.deja.com>...

Have you checked out www.python.org/topics/scicomp ? It has references to a
plot-sig - a SIG dedicated to this topic.  It is almost certain that the
people who know about this stuff will hang out there...  If you dont get an
answer there, it is almost certain that what you are looking for does not
exist.

Mark.

>Repost:
>My apologies for mangling my return address on the post. I wanted to
>avoid more spam by the mangling. My e-mail address was included at the
>bottom. Hmmm. If only I could count the number of spam e-mails I will
>get from this repost,...






From Donn.Cave at p98.f112.n480.z2.fidonet.org  Fri Jul  2 13:37:04 1999
From: Donn.Cave at p98.f112.n480.z2.fidonet.org (Donn Cave)
Date: Fri, 02 Jul 1999 18:37:04 +0100
Subject: Socket programming.
Message-ID: <00001684@bossar.com.pl>

From: donn at u.washington.edu (Donn Cave)

"Kjetil ?degaard" <kjetilod at stud.ntnu.no> writes:
| I'm trying to implement a simple irc module as the base for a simple
| irc client.  However, I'm having some design problems, having never
| done any socket programming before.
|
| My current implementation is threaded (using fork()); one thread for
| user interaction and one for server output.  It uses blocking
| sockets.  It works ok, but I've read some docs here and there (amongst
| them the Socket Programming HOWTO on www.python.org) and I get the
| feeling I should be using select() and non-blocking sockets.  Also,
| I'm not sure I want to use fork() -- it seems very low-level and
| depends on a Unix-like system.

These are questions everyone wrestles with, and the only thing we can
say for sure is ``it depends.''  If your current design works, it sounds
fine to me.

As you know, select comes in when you want to dispatch from two or more
I/O event sources, in a single execution thread.  (With select, you really
don't need non-blocking sockets, because select's job is to predict whether
a read would block.)  If the event handling can be simplified to the point
where with a modest amount of global state information, the program can
handle any event in no time and get back to the dispatch loop more or less
right away, then it's worth thinking about a single threaded design.

But it can be more complicated than you expect - for example, when you
write to a socket device, write() or send() may return after writing less
than the full amount, and you have to keep sending until everything goes.
So your global state may have to queue outbound data.  (Use select to find
out whether write will block, too.)  This is the kind of thing you'd need
to consider in a single threaded service, not necessarily in a client.

This kind of worry is avoided with a multi-threaded design, and then
the question is whether to use processes or threads.  Processes are
more expensive and indeed a bit less portable.  My favorite OS, BeOS,
supports fork - but a child fork doesn't inherit an open socket, so
fork is a very difficult way to make a multi-threaded socket application!
Moreover, select is available for sockets, but the last I checked it
doesn't work on all types of file descriptors.  (I present this as
an example of a non-UNIX socket system that I think even its designers
would admit is slightly lame.  I expect something better is on the way.)

BeOS enthusiasts aren't troubled by this, they feel that threads are
available to solve all these dispatching problems.  But threads carry
their own problems, as there's a lot more potential for collision when
two threads run in the same address space, so you need to sweat interlocks
that aren't an issue in separate forks.  (And Python itself has to deal
with this, as a threaded C program, which complicates the issue.)
Meanwhile the larger program sort of loses control over its threads,
compared to the single threaded select dispatching described above,
because threads blocking on I/O can't process other events, they can't
tell you whether they're actually reading or just blocking, etc.

So the old ways are best, is what I say.  Unless they aren't, and at
times they aren't.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From stidolph at origin.ea.com  Mon Jul 19 10:31:51 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Mon, 19 Jul 1999 14:31:51 GMT
Subject: [Q] Type checking...go further...
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AABFC@molach.origin.ea.com>

You cannot have typed method parameters (Python is typeless), but can only
check the types of parameters within the method.

from types import * #Needed for different types like IntType

def foo(a, b, c):
  if type(a) is IntType:
    print a,'is a integer'
  elif type(a) is LongType:
    print a,'is a long integer'
  elif type(a) is StringType:
    print a,'is a integer'
  elif type(a) is FloatType:
    print a,'is a floating point number'
  elif type(a) is ListType:
    print a,'is a list'
  elif type(a) is TupleType:
    print a,'is a tuple'

or

def foo(a,b,c):
  if not (type(a) is IntType):
    raise "Param a must be integer"
  if not (type(b) is StringType):
    raise "Param b must be string"
  if not (type(c) is ListType):
    raise "Param c must be list"
  ...

Hope this helps,

David Stidolph
Origin

-----Original Message-----
From: Olivier Deckmyn [mailto:olivier.deckmyn at mail.dotcom.fr]
Sent: Monday, July 19, 1999 8:36 AM
To: python-list at cwi.nl
Subject: [Q] Type checking...go further...


Thank you all for your answers !
I finally use isinstance in the methods of my classes...

But to go further, is there a way in python to have typed methods parameters
?
aka :

class MyClass:
    def myMethod(string myString):
        print myString

?

Thanx again all for your help !

Olivier,
happy python newbie.





From nick.bower at ssec.wisc.edu  Sat Jul 10 19:35:54 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sat, 10 Jul 1999 18:35:54 -0500
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu>
Message-ID: <7m9to4$6qg$79@rks1.urz.tu-dresden.de>

thanks for the help all.  exactly what i thought would have existed.

Incidentally, I went back to the library reference and still couldn't find
os.chdir

cheers, nick.


Nick Bower wrote:

> I feel really dumb asking this question, but how do I change the current
> working directory?  For browsing the file system I mean.
>
> I've found exec commands, but these would surely be device dependent and
> there doesn't seem to be anything obvious in the os or os.path
> documentation.  i looked through the tutorials too.
>
> thanks in advance, nick

--

Nick Bower
Space Science and Engineering Center
University of Wisconsin, Madison, USA
Phone: (608) 265 8007
Email: nick.bower at ssec.wisc.edu
Web: http://arm1.ssec.wisc.edu/~nickb





From andrew at andrewcooke.free-online.co.uk  Sun Jul  4 06:23:44 1999
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Sun, 04 Jul 1999 10:23:44 GMT
Subject: Dumb beginner question; what is IDLE?
References: <377D8BFB.3700379@afn.org>
Message-ID: <7lncne$ps5$1@nnrp1.deja.com>

In article <377D8BFB.3700379 at afn.org>,
  Andy L Krietemeyer <afn08995 at afn.org> wrote:
> I've seen several references to IDLE.  Is this an IDE for python? Or
> what is it?  Thanks,  Andy.
>

Yup.  It has simple editing and debug support but no GUI builer or
"wizards".  A kind of poor-mans emacs :-) although I must admit that
I use it rather than emacs for Python code.  I'm not sure why -
I tried it and never changed back.  The only thing I miss is not
having several windows on one piece of code (maybe this is possible,
I've never been able to do it).    Also, the key bindings are a
compromise between emacs/Unix and Dev Studio/MS, so if you use all
three you can type the wrong thing at times....

It appears to be written in Tkinter, but has a very nice file browser,
which I keep intending to track down, but haven't, yet - but which
you can presumably use in other Tkinter apps.

Andrew


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Nick at p98.f112.n480.z2.fidonet.org  Thu Jul  1 17:21:09 1999
From: Nick at p98.f112.n480.z2.fidonet.org (Nick)
Date: Thu, 01 Jul 1999 22:21:09 +0100
Subject: Tuples & lists
Message-ID: <0000162f@bossar.com.pl>

From: "Nick" <nick at NOSPAMvideosystem.co.uk>

I'm a new user and Python looks great.

Why both having tuples and lists? As far as I can see there're the same, but
tuples are read only..? Have I missed something in the tutorial?

Also, has anyone written a tcl/tk dialog/resource editor type thing so you
could save a form out as code? Could save a lot of time...

Thanks,
Nick.






From ivnowa at hvision.nl  Wed Jul  7 18:04:23 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Wed, 7 Jul 1999 22:04:23 GMT
Subject: Two Pythons talking to each other?
Message-ID: <199907072203.AAA09601@axil.hvision.nl>

Hello,

I am currently experimenting with networking and want to write some 
games (where, obviously, two players compete over a network). Now I 
want to test my programs offline. One way would be to start two 
separate Python sessions (on the same machine), both running the same 
game program, and let them somehow talk to each other like they would 
over a network. But, what would be a good way to do this? I was 
thinking about passing strings (commands etc) to a file, but two 
programs cannot read from and write to the same file at the same 
time. Does someone have any ideas about how to do this?

TIA,

--Hans Nowak (ivnowa at hvision.nl)
Homepage: http://fly.to/zephyrfalcon



From hnowak at cuci.nl  Sat Jul 10 09:10:27 1999
From: hnowak at cuci.nl (Hans Nowak)
Date: Sat, 10 Jul 1999 13:10:27 GMT
Subject: Python Code Repository?
In-Reply-To: <378645AA.5430A026@msfc.nasa.gov>
References: <378645AA.5430A026@msfc.nasa.gov>
Message-ID: <199907101311.PAA13089@dionysus.fw.cuci.nl>

On 9 Jul 99, Drew McDowell wrote:

> After seeing the "Brensenham line algorithm" post above, I'm 
curious
> if there is a public "python code repository" (something akin to 
CPAN) out
> there?  Or does Starship assume that role?

I do have a snippet repository, but this is by no means exhaustive 
(yet), and does not aim to be (although it's growing). It's not 
really comparable to CPAN because it's not a repository for modules, 
but rather for simple (and not-so-simple) pieces of code.

I believe the URL is in my signature. :o)

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
Python Snippets: http://www.hvision.nl/~ivnowa/snippets/
The Purple Kookaburra Forum: http://www.delphi.com/kookaburra/



From tgm at math.uni-bonn.de  Fri Jul  9 08:48:00 1999
From: tgm at math.uni-bonn.de (Thomas Ackermann)
Date: Fri, 9 Jul 1999 12:48:00 GMT
Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
In-Reply-To: <19990708223856.A29441@optiplex.palga.uucp>; from Gerrit Holl on Thu, Jul 08, 1999 at 10:38:57PM +0200
References: <tot2m7.su.ln@final.math.uni-bonn.de> <19990708223856.A29441@optiplex.palga.uucp>
Message-ID: <19990709144800.46076@devel.math.uni-bonn.de>

On Thu, Jul 08, 1999 at 10:38:57PM +0200, Gerrit Holl wrote:
> On Thu, Jul 08, 1999 at 09:22:05PM +0200, Thomas Ackermann wrote:
> > I have not much time to experiment with those modules and would like
> > to hear opinions about to use what module     ;-)
> 
> Well,
> I used HTMLgen and I don't really understand why to use it. If you don't
> know HTML, you can't use it: you need to understand tags like <B>, <I>, <U>
> and such. I strongly recommend to have a look at HTML: if you look at it only
> one hour, you'll already understand the basics. Another hour and you understand
> frames and tables, and another hour to understand how it interacts with CGI.
> So you'll only need 3 hours, if not less, to learn web programming. When you're
> able to do that, you'll understand much more about HTMLgen (but won't need it,
> I think...)

Thanx - But i know HTML (and Java) itself already	;-)
What i do not know are the Python modules for it and wanted to get some
information from people who used several of the alternatives already.
But thanx anyway ...


> > So, what to use to allow for an Internet-Shop with pages created from
> > databases and allowing customers to purchase over net?
> > 
> 
> If you're doing such serious things, you should 
> 
>                                      **really**
> 
> learn web programming!!

Again, i want to use Python and a decent HTML/CGI module, i never said to
be completely unexperienced with WWW stuff ...

	Byebye,
-- 
Thomas Ackermann | Tel. +49-(0)228/631369|73-7773 | <tgm at math.uni-bonn.de>
             finger tgm at rhein.math.uni-bonn.de for public key
               GNU LINUX Python gtk pygtk MySQL FUDGE GURPS



From MHammond at skippinet.com.au  Mon Jul 26 19:41:40 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Tue, 27 Jul 1999 09:41:40 +1000
Subject: COM bites - help please
References: <eZUl3.1080$wm4.135734@news.uswest.net>
Message-ID: <7nirjd$n52$1@m2.c2.telstra-mm.net.au>

Karl Putland wrote in message ...
> >>> c1.fields        # This is a util.Collection(list of GMField
instances)
> Traceback (innermost last):
>   File "<pyshell#56>", line 1, in ?
>     c1.fields

Have you tried using the debugging flags?  This can help alot.  If you
register your object with "--debug", you can use either Pythonwin or
win32traceutil.py to see exceptions from the server side of Python COM.

This is likely to show a traceback saying something like "object does not
support the policy" - ie, as Bill said, you need a list of wrapped objects,
not Python classes.

[However, win32com.server.util really should grow helpers for this - a
reasonable and common requirement - feel free to propose a change!]

> 3. PyUnicode objects bit me in the ass when strings are passed into
methods and manipulated.
> class Foo:

After reading Bills reply and your followup, Im unsure what your remaining
problem is with this.  You have discovered that "str()" is the simplest way
to work with them, and we know that Python 1.5 and Unicode integration
suck - but is there some other Unicode problem you are having beyond this?

> >>>foo = win32com.client.Dispatch("Python.Foo")
> >>>foo.bar("Hello World")

> Some traceback about TypeError PyUnicode.

Again, use the "--debug" option, and you will also see a traceback on the
server side.  This will point you at the problem.

Mark.





From thomar at ifi.ntnu.no  Wed Jul 14 08:36:14 1999
From: thomar at ifi.ntnu.no (Thomas Røberg)
Date: Wed, 14 Jul 1999 14:36:14 +0200
Subject: Embedding Python in C
Message-ID: <7mi084$r80$1@kopp.stud.ntnu.no>

I am writing a program in C/C++ and want to embed Python. The one problem I
encounter is that when I try to import a module using
Pyrun_SimpleString("import mymodule\n"), this module is not found. I have
specified the correct paths at the correct places in my Makefile (at least I
think I have). Anyone had problem with this??

Thomas R?berg
thomar at ifi.ntnu.no





From n89553 at stud.uni-paderborn.de  Fri Jul 30 07:44:00 1999
From: n89553 at stud.uni-paderborn.de (Mirko Liss)
Date: Fri, 30 Jul 1999 13:44:00 +0200
Subject: Problem compiling Python on Suse 6.1
References: <37A01B2B.2CF78A89@gmx.net>
Message-ID: <37A19000.2C94CCC5@stud.uni-paderborn.de>

Tomaz Ficko wrote:
> 
> I have a problem compiling Python 1.5.2 with Tk and readline support.
> I'm using Suse 6.1 (evaluation iso):
> gcc version egcs-2.91.66 19990314 (egcs-1.1.2 release)
> Tcl/Tk 8.0.5
> bash-2.02
Download the source package of python151 from the suse dist and
look at the changes they made. Http://www.suse.de/ftp.html
lists several ftp-servers.  
For suse60, it boils down to:
 
- grep the python source for <tcl.h> and <tk.h> 
  and replace with <tcl80.h> resp. <tk80.h>
- edit Module/Setup. Uncomment several line with respect to the 
  software you installed (Tix, BLT, gdbm etc). This includes:
     readline readline.c -lreadline # -ltermcap
     curses cursesmodule.c -lncurses # -ltermcap
     zlib zlibmodule.c -$(prefix)/include -L$(exec_prefix)/lib -lgz #
-lz

- consider installing the ncurses-module by Oliver Andrich
- suse comes with a static lib readline.a .
  I built readline.so from source long before
  I installed Python 152. This might not be
  necessary.

However, your mileage may vary.

regards, 
Mirko

-- 
M Li?, <n89553 at hrz.uni-paderborn.de>



From thomas at xs4all.nl  Wed Jul 21 09:35:39 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Wed, 21 Jul 1999 15:35:39 +0200
Subject: Merging two lists as a dictionary
In-Reply-To: <19990720211946.A12202@pooh.frostnet.net>; from Chris Frost on Wed, Jul 21, 1999 at 02:19:46AM +0000
References: <19990720211946.A12202@pooh.frostnet.net>
Message-ID: <19990721153539.S6170@xs4all.nl>

On Wed, Jul 21, 1999 at 02:19:46AM +0000, Chris Frost wrote:

> I have two lists and would like to merge them into a dictionary, with one
> list forming the keywords and the other the values in the dictionary. How could
> this be done? "map(None, keyword_list, value_list)" looked great, but it
> returns a list of pairs (I'd like to be able to specify values according to
> keywords...).

>>> keywords = ['s','p','a','m']           
>>> values = ['eggs', 'ham', 'toast', 'omelet']
>>> result = {}
>>> for i in range(min(len(keywords), len(values))):
...     result[keywords[i]] = values[i]
... 
>>> result
{'m': 'omelet', 's': 'eggs', 'p': 'ham', 'a': 'toast'}

the 'min()' is unecessary if you are 100% certain the lists are same length,
or if you know which list will be the shorter one. (iter over the shortest
one, in that case)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From rasumner at my-deja.com  Sun Jul 18 09:34:49 1999
From: rasumner at my-deja.com (rasumner at my-deja.com)
Date: Sun, 18 Jul 1999 13:34:49 GMT
Subject: memory leaking?!
References: <slrn7op5jj.cva.rasumner@math5-pc.wisdom.weizmann.ac.il>
Message-ID: <7msl5q$2pt$1@nnrp1.deja.com>

In article <slrn7op5jj.cva.rasumner at math5-pc.wisdom.weizmann.ac.il>,
  rasumner at math5-pc.wisdom.weizmann.ac.il (Reuben Sumner) wrote:
> I have some code creating a somewhat complicated data structure.
> When I am using Python 1.5.2 compiled with Py_REF_DEBUG I can see
> that reference counts are dissapearing (although I can't figure out
> what is getting lost, all the objects that I can think of I have
> confirmed are being released).  When I compile with Py_TRACE_REFS
> then all of a sudden the reference counts don't increase between
> subsequent calls.  What on earth is going on here?  Running "top"
> and generating a lot of very large structures then memory does seem
> to leak no matter which way I compile python (although I think it is
> slower with Py_TRACE_REFS).

Here is a really small test case demonstrating the problem.

import sys

class C:
	def __init__(self): pass

print sys.getrefcount()
for i in range(100): c=C()
print sys.getrefcount()
for i in range(100): c=C()
print sys.getrefcount()
for i in range(100): c=C()
print sys.getrefcount()

Realy simple.  When I run this in an interpreter compiled
with Py_REF_DEBUG but _not_ Py_TRACE_REFS there is a leakage
of exactly 100 references between the 2nd and 3rd getrefcounts().
This use of sys.getrefcount() requires a small patch I posted
a while ago.  Alternatively run this interactively and look
a the printed values for the total number of references.  My
Python is 1.5.2 on Linux (RedHat 6.0).

Can other people reproduce this?  Can anybody _not_ reproduce
this?  I get the same effect on IRIX6.

Reuben



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From heiland at ncsa.uiuc.edu  Tue Jul  6 23:52:23 1999
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Wed, 07 Jul 1999 03:52:23 +0000
Subject: Python array --> C
Message-ID: <3782CEF6.97D7C5FB@ncsa.uiuc.edu>

In my continuing quest to better understand embedding/extending, I have
the following question...

in Demo/embed, I have modified demo.c so that the static module
xyzzy_foo(self, args)
now does the following (instead of returning 42L):
{
        unsigned char retval[5] = {0,2,4,8,16};
        return Py_BuildValue("s#", retval,5);
}
i.e., it returns an array of unsigned chars.

>From main(), I would like to invoke xyzzy.foo() and get this array of
unsigned chars into a C array.
How do I do this?  I assume I'll need to use PyRun_String(...) somehow -
true?

thanks,
--Randy




From tismer at appliedbiometrics.com  Tue Jul 13 04:48:05 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 13 Jul 1999 08:48:05 GMT
Subject: ANN: Stackless Python 0.3
References: <378A40AD.B4FDB79F@appliedbiometrics.com> <m37lo5kpd0.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <378AFD45.8793D507@appliedbiometrics.com>


Michael Hudson wrote:

> Very nice, but

> >>> apply(l.sort,())
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> SystemError: error return without exception set
> >>> l
> [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

Here is the patch.
It was one of those "=" instead of "==" bugs.

Line 131 of bltinmodule.c should read now

	if (retval == Py_UnwindToken) {
	-----------!!------------------

I'm updating the files right now, without changing the version number.

Thanks, Michael!

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From Mark.Hammond at p98.f112.n480.z2.fidonet.org  Fri Jul  2 14:56:49 1999
From: Mark.Hammond at p98.f112.n480.z2.fidonet.org (Mark Hammond)
Date: Fri, 02 Jul 1999 19:56:49 +0100
Subject: Newbie has trouble with in/output
Message-ID: <00001643@bossar.com.pl>

From: "Mark Hammond" <MHammond at skippinet.com.au>

Florian W. wrote in message <7lglb4$ksu$1 at news.vossnet.de>...
>1.) Being in a pipe:
>
>I won't to filter the output of the compiler (e.g. only errors and warning
>are reported, no logo etc.)
>using the pipe-mechanismen :
>
>cl.exe blabla | python filter_compiler_output.py
>
>How do I catch the input and how do print the output ?

You simply read() to get the input, and print to write output!!

>2.) Catch the output of an application:

Check out the win32pipe module in the win32all package.  Documentation
(sort-of :) is in the help file, or searching dejanews for that module name
will return plenty of hits...

Mark.





From mnot at pobox.com  Fri Jul  9 22:52:32 1999
From: mnot at pobox.com (Mark Nottingham)
Date: Sat, 10 Jul 1999 02:52:32 GMT
Subject: Python webcrawler?
References: <99Jul9.164357pdt."3569"@watson.parc.xerox.com>
Message-ID: <7m9tnc$6qg$55@rks1.urz.tu-dresden.de>

I started writing a crawler, but never finished. However, the engine was
used in the Cacheability Engine -- http://www.mnot.net/cacheability/

It might be useful to rip the core out as a start for one if you don't have
anything else.

Cheers,




From holger at phoenix-edv.netzservice.de  Tue Jul  6 09:45:59 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Tue, 6 Jul 1999 13:45:59 GMT
Subject: Win32: silent script
References: <37809062.6D463928@phoenix-edv.netzservice.de> <7ls61c$3nu$1@Starbase.NeoSoft.COM>
Message-ID: <37820897.93F4ED5D@phoenix-edv.netzservice.de>

Ok, that's a possibility.
Problem is that Windows link files with extensions like .py and .py
to python.exe (take a look at registry!). This causes in installation
of python under Windows. Don't like to change registry!

To start pythonw.exe in that links in windows-menu I've to specify
the whole path to that file... Ouch, I really didn't want that!;-(

What's about the other way:
How could I set the 'minimized window' option in that link by python-code?

That the source how I create that links:

def createLink(filePath, linkPath, args, descr, wrkngDir):
        "creates a link to a file"
        import pythoncom
        from win32com.shell import shell
        #get the shell interface
        sh = pythoncom.CoCreateInstance(shell.CLSID_ShellLink, None,
pythoncom.CLSCTX_INPROC_SERVER, shell.IID_IShellLink)
        #get an IPersist interface
        persist = sh.QueryInterface(pythoncom.IID_IPersistFile)
        #set the data
        sh.SetPath( filePath )
        sh.SetDescription(descr)
        sh.SetWorkingDirectory(wrkngDir)
        sh.SetArguments(args)
        #save the link itself.
        persist.Save(linkPath, 1)  

I think there must be a 'sh.SetWindowsMode' or something like that?!?

>...
> Use pythonw.exe.  Its entire purpose in life is to make the console
> window unnecessary.
>...

Ciao,
Holger



From ed at sis.yorku.ca  Mon Jul 26 11:59:45 1999
From: ed at sis.yorku.ca (Ed F. McCurdy)
Date: Mon, 26 Jul 1999 11:59:45 -0400
Subject: CGI ImportError
Message-ID: <199907261559.LAA09924@sprawl.sis.yorku.ca>

>Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]

You could modify the environment of the web server to
include and pass on (mod_env?) PYTHONHOME=/Python-1.5.2.
You should verify that the server (with uid "nobody") 
has permissions to read $PYTHONHOME.

What may be simpler is to re-build python and
leave the default paths unchanged. 

Hope this helps,
Ed
-- 

Edward F. McCurdy,               | ed at sis.yorku.ca
York University, Toronto, Canada | (416) 736-2100x22685






From mhagger at blizzard.harvard.edu  Wed Jul 21 19:14:00 1999
From: mhagger at blizzard.harvard.edu (Michael Haggerty)
Date: 21 Jul 1999 19:14:00 -0400
Subject: Scientific Plotting?
References: <19990721010203.96101.qmail@hotmail.com> <37952384.14766416@news.erols.com> <3795c243.764244896@news.ufl.edu> <37963f6d.2239819@news.erols.com>
Message-ID: <xj0btd57g07.fsf@cyclone.harvard.edu>

On Wed, 21 Jul 1999 13:03:52 GMT, lazrnerd at ufl.edu (Craig Schardt)
wrote:

>I have modified Michael Haggerty's Gnuplot.py to work under Windows.
>The Windows version of Gnuplot doesn't support pipes but there is a
>helper application, pgnuplot, which works around this problem and
>provides a pipe interface. The combination works well on NT and 98
>with PythonWin.

kernr at mail.ncifcrf.gov (Robert Kern) writes:

> I've done the same thing, but I don't think I required PythonWin (as
> in, Mark Hammond's win32 extensions).

(I wonder how many other people have reinvented this wheel.)

The latest version of Gnuplot.py has been downloaded about 500 times,
but almost the only feedback I've gotten has been from a few people
asking for help to get it working under windows.  I don't use windows
so I couldn't much help them.  If somebody (anybody!) would be willing
to send me their patches and agree to put them under GPL I would be
happy to include them in the next release...

By the way, I just added support for binary-format data to a
yet-unreleased version of Gnuplot.py.  It only works for splot grid
data because AFAIK that's the only place gnuplot allows binary data.
But that is the most data-intensive type of plot and it substantially
helps the speed issue for that case.

I'm working on allowing data for all kinds of plots to be passed
through pipes, too (currently commands go through pipes but data go
through temporary files).  I fear it won't make a big speed difference
though because I suppose more time is spent converting data to ascii
and back than on the disk I/O.  If only there were a decent gnuplot
library that could be easily linked to.

Michael

-- 
Michael Haggerty
mhagger at blizzard.harvard.edu



From s341625 at student.uq.edu.au  Mon Jul 19 22:56:28 1999
From: s341625 at student.uq.edu.au (Anthony Pfrunder)
Date: Tue, 20 Jul 1999 12:56:28 +1000
Subject: Python and the Web
In-Reply-To: <3793dd72.0@hoh.olympus.net>
References: <3793dd72.0@hoh.olympus.net>
Message-ID: <Pine.OSF.4.10.9907201255050.4911-100000@student.uq.edu.au>

On Mon, 19 Jul 1999, Michael Novello wrote:

> Can anyone tell me where to find information on using python to grab form
> data and create dynamic web pages?

Try Zope: http://www.zope.org/ Windows binaries are located at:
http://student.uq.edu.au/~s341625/  Note, you will need the python source.

Zope can do everything you could possibly want ;)

> 
> Thanks,
> 
> Michael

Cheers,

Anthony Pfrunder
Computer Systems Engineer
University of Queensland




From dolana at logica.com  Thu Jul 29 08:50:30 1999
From: dolana at logica.com (Andrew Dolan)
Date: Thu, 29 Jul 1999 13:50:30 +0100
Subject: Is "shelve" thread-safe on Windows?
Message-ID: <7npimc$l9b@romeo.logica.co.uk>

Hi All,

I hope to use the shelve module to store data which may be accessed from
several Python CGI scripts at once.  I have studied the source-code and it
seems that the anydbm code (that shelve relies on) falls-back to using the
dumbdbm module on Windows which is not safe for concurrent access (according
to comments in its source code).

Can anyone suggest a simple way to gain the functionality of "shelve" on a
Windows platform, but which is safe for handling concurrent accesses (like
from CGI scripts)?

Cheers,

Ian.

--
Any views expressed in this email are my own and not those of Logica UK.





From C.Kotsokalis at ece.ntua.gr  Sun Jul  4 20:06:24 1999
From: C.Kotsokalis at ece.ntua.gr (Constantinos A. Kotsokalis)
Date: 5 Jul 1999 00:06:24 GMT
Subject: Why have 300 articles just appeared in c.l.py?
References: <m33dz4oxxx.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <slrn7nvtfr.dm.C.Kotsokalis@ppp031.dialup.ntua.gr>

On 04 Jul 1999 23:24:42 +0100, Michael Hudson wrote:
>I just had the slightly alarming experience of about 300 articles, all
>of which I had read, appearing unthreaded in this newsgroup.
>
>Did this happen to everyone, or is it some local snafu?

Same here...

  --C

-- 
Constantinos A. Kotsokalis || C.Kotsokalis at ece.ntua.gr
National Technical University of Athens - GREECE
Electrical and Computer Engineering Department
"Bus error -- driver executed."



From neelk at brick.cswv.com  Thu Jul  1 22:17:48 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 1 Jul 1999 21:17:48 -0500
Subject: [Q] Static type checker for Python
References: <7lee1g$sod$1@news.etri.re.kr> <377ADAD7.EA105C50@mojam.com>
Message-ID: <7lh7gc$2du$1@brick.cswv.com>

In article <377ADAD7.EA105C50 at mojam.com>,
Skip Montanaro  <skip at mojam.com> wrote:
>> Can I make static type checker for Python?
>> or Is there any previous work about this?
>
>In theory, you could do some type inference and generate some warning
>messages, but since you don't declare variables to be of a particular
>type in Python, it's unlikely it would be very useful.  The same
>variable can point to many different types at runtime.

I'm about to go off on a tangent, so first I want to warn the original
poster that there aren't too many actual problems with dynamic typing
-- people have written huge and complex apps with no problems using
Smalltalk and Lisp, for example. 

Now, it is possible to teach Python to check the argument types of
functions at runtime, and I've written a little module called
TypeChecker to do just that. It's a wrapper class with callable
instances, and the instance checks the declaration you make in a
docstring against the actual types and classes of the arguments. 
For example:

def fact(n):
    """foo(n :: LongType) <-- here's the type declaration; it can be
			      either a type from the types module or a
			      class name.
    A factorial that only takes longs as arguments."""
    f = 1L
    for i in range(1L, n+1):
        f = f * i
    return f

f = TypeChecker(fact) 

>>> fact(3)
6L

>>> f(3L)
6L

>>>f(3)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/tmp/python-6-237", line 25, in __call__
  File "/usr/tmp/python-6-237", line 118, in check_args
TypeError: Argument 1 of fact wrong type, <type 'int'>, declared <type 'long int'>

What it's doing is parsing the type declaration in the doc string to
generate a list of types and classes, and then the __call__ method of
the instance is comparing the declared types against the runtime
arguments it receives before running the function. I've always wanted
to abuse docstrings, and now I've gotten my chance. :) Though this is
a fairly mild abuse, considering that it doesn't block you from
putting other things into the docstring.

Obviously, this is not terribly useful. (It doesn't handle subclasses
correctly, f'ex, and can't yet be used on class methods.) But it is
interesting to see what Python static types might look like. (The
answer to my eyes is "fairly ugly;" real lexical scoping and
type/class unification seem to me to be prerequisites before static
typing will behave gracefully in Python.)

The actual code for TypeChecker is a little too long to put in the
post, so here's a URL:

    http://www.sff.net/people/neelk/free-software/TypeChecker.py

Neel



From MHammond at skippinet.com.au  Fri Jul  9 08:36:37 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 9 Jul 1999 22:36:37 +1000
Subject: Python and Paradox
References: <7m2mrj$hes$1@nnrp1.deja.com> <7m2o2b$aq0@dfw-ixnews12.ix.netcom.com> <37852804.54D5D2AD@quasar.ipa.nw.ru>
Message-ID: <7m9tm3$6qg$17@rks1.urz.tu-dresden.de>

Alexander V. Voinov wrote in message <37852804.54D5D2AD at quasar.ipa.nw.ru>...
>Aahz Maruch wrote:


>> >I am looking for a way to access Paradox tables from Python without
>> >using ODBC. Does anyone have any ideas?
>>
>> Nope, from an ex-Paradox tech support engineer.

Yep, from someone who has :-)  Well, Im probably not technically correct,
but if you:
* Install MSAccess
* Install Paradox drivers from the "Value Pack" (on the CD) - I think this
does install an ODBC driver even tho we dont use it directly (hence my "not
technically correct")
* Attach the Paradox tables to an Access database.

You can use the COM extensions to talk directly to the Jet API, which in
turn will manage the paradox tables.  I find the COM Jet interface far
better than ODBC.

Mark.






From aahz at netcom.com  Wed Jul  7 20:38:33 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 8 Jul 1999 00:38:33 GMT
Subject: Error Trapping In Python
References: <7m0pgt$cij$1@eskinews.eskimo.com>
Message-ID: <7m0ru9$gvt@dfw-ixnews3.ix.netcom.com>

In article <7m0pgt$cij$1 at eskinews.eskimo.com>,
Jonathon <jblake at eskimo.com> wrote:
>
>	The script I am writing expects the user to input an
>	integer.   Being on the safe side, it reads the raw
>	data as a string.   However, to convert that string
>	to an integer, is perplexing me.
>
>	string.atoi.(read_data) will give a value error if
>	read_data contains something other than a number.

try:
	string.atoi.(read_data)
except:     # or except ValueError:
	do_something_or_other()
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From billtut at microsoft.com  Thu Jul 15 10:58:15 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Thu, 15 Jul 1999 14:58:15 GMT
Subject: making a Windows NT service out of a python program
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F4B@RED-MSG-50>

> From: Christian Tismer [mailto:tismer at appliedbiometrics.com]
> 
> 
> This is all great stuff if you have understood it already.
> For me, it is still unclear what makes up a service, and
> what I need at the minimum to turn a Python program into
> an NT service.
> 

Not knowing ANYTHING about Medusa... I'll provide another example service
right here, and try and explain things as I go... 

The Win32 support for services is encompassed by 3 major pieces.
* win32service.pyd
This provides the standard "Glue Win32 API functions to Python" code.
* PythonService.exe
NT services are executables that must follow special rules. 
PythonService.exe follows these special rules. 
It also exports Python bindings to the "servicemanager" module.
This module exports the following methods:
	* CoInitializeEx (COM initialization)
	* CoUnitialize (COM unitilialization)
	* RegisterServiceCtrlHandler 
	  (This is the lynchpin function of getting 
         a Python class to run as a service.)
	* LogMsg, LogInfoMsg, LogErrorMsg, LogWarningMsg
	   (These all log messages to the eventlog, and mark them as
		coming from the "Python Service Manager") To make eventlog
messages
        appear to come from your own application you'll need to use Python's
        other eventlog support.
	* PumpWaitingMessages
	   (Dispatches any incoming window messages, 
	    and returns 1 if WM_QUIT was recieve, otherwise returns 0.)
	   Most services shouldn't need this, I take it Mark stuck it here
for
       some wierd service he wrote that actually needs access to the
desktop.	* Debugging
	   (This returns 1 if the service is running in debug mode.
          More on this later....)
* win32serviceutil.py
	This sucker provides the following sub-groups of useful code:
	* ServiceFramework
  	    This is just a useful base class to build services from.
	* Provides a handly command line handling function that handles:
	    Installing, Removing, Updating, Starting, Stoping, Restarting,
	    pausing, setting the user the service logs on as, specifying
        auto, manual, or disabled startup policy, and something about 
	    perfmon stuff. 
	* Handy wrapper functions to the ugly API interfaces in
win32service.pyd.
	    (See QueryServiceStatus for a good example.)
	* Handy utility functions for storing configuration data in the 
	  registry.
	    (SetServiceCustomOption, and GetServiceCustomOption)

Ok, so those are the pieces, but Bill, how does all of this nonsense fit
together? Good question. :)

NT Services are managed by the SCM (Service Control Manager).
The SCM is what manages almost everything about services...
Installing, removing, starting, stopping, etc...
The SCM is also in charge of dealing with the service interdependancy chain.
Classic example:
  IIS is composed of multiple services: iisadmin, w3svc, ftpsvc, etc....
  In order for w3svc, and ftpsvc to run correctly iisadmin must already
  have been started. The SCM handles all of this detail for us.

An NT service must have special entry points inorder to get along with
the SCM nicely. All of these entry points for most Python services
should just be in PythonService.exe to make your life simple.

* main: Most NT services are console applications, so this is where things
start.
	PythonService.exe also has some code to handle command line
arguments.  i.e. register, and debug. register handles keeping track
of some python/eventlog bookkeeping it needs internally. After
handling any command line arguments, main calls the NT API
StartServiceCtrlDispatcher function, which connects the main thread of
PythonService.exe to the SCM. The main thread is the thread on which
the service management code gets called on.

* ServiceMain: This what gets passed into StartServiceCtrlDispatcher
	As soon as this gets called, we call RegisterServiceCtrlHandler()
to let NT know what C function of ours is going to handle dealing with
notifications from the SCM. We then tell NT that we're in the process
of starting up, chill out and wait a bit while we initliaze ourselves,
and suggest to NT that 5 seconds might be a nice length to wait for us
to finish starting up before complaining too loudly.
	ServiceMain then calls the Python class that you've designated's
"SvcRun" method. "SvcRun" must not return until the service is ready
to terminate. The default "SvcRun" in win32serviceutil.py tells NT
that the service is up and running, calls your definition of
"SvcDoRun".

* ServiceCtrlHandler:
	This handles communications between the SCM and the rest of your
code. This function gets called by the SCM when it wants you do
something, or wants to inform you of something. Examples: start, stop,
pause/continue, machine shutdown, update your status, power event
(Win2k only), etc....

NOTE: ServiceCtrlHandler gets called on the orignal main thread of your
process. SvcRun is actually running on a seperate thread, so you must use
some kind of synchronization object between your registered
ServiceCtrlHandler function, and your "SvcDoRun" function. An example of
this is in the attached TCPServerService class.

Hopefully this helps explain an overview of how NT services need to work.
Now, lets get into the nitty gritty of this example service I cooked up.

The necessary files in this little example are:
* edna.py: Greg Stein's Edna
    A simple streaming MP3 server, and a subclass of BaseHTTPServer) (only
a simple change to handle how configuration is handled was all that
was necessary....)  See http://www.lyra.org/greg/edna for more
details.

* TCPServerService.py: This contains two useful classes you need to inherit
from.
	* TCPServerService:
      This provides all the hairy glue of meshing the Python service
framework, and SocketServer.TCPServer subclasses together.

	* SvcTrackingThreadMixin:
      This is almost identical to the normal threading mixin, except
that it keeps track of thread handles, so shutdown can be postponed
until all of our pending requests are done.

* ednaNTSvc.py:
    This is all that is necessary to fill in the rest of the details
to enable an NT service. All this module does is implement policy for
when the service should shut down with pending requests, and handle
installation/configuration issues.

* win32eventmodule_win32.cpp:
    It turned out that the Win32 extensions were missing one API
function. WSAEventSelect(), so its been added to the win32event
module. WSAEventSelect allows you to do a WaitForMultipleObjects on
sockets which is nice sine you can't use select() on NT events. :)

How to setup this cool example so that you can try it out.
1. Alter edna.conf to taste.
2. Do the following command: python ednaNTSvc.py -c ...\edna.conf install
    This saves the location of the configuration file in the registry, and
installs the service, and marks it as not starting automatically.  To
make it start automatically when your machine starts add "--startup
auto" before "install" in the command line.

3. Run it.
You can run it more than one way.
	* ednaNTSvc.py start
		This kicks off the service, and then ednaNTSvc.py exits.

	* ednaNTSvc.py debug
		This kicks off PythonService.exe passing in -debug on the
command line, this has the nice effect of being able to see the output
of stdout/stderr in your console window, and being interuptible via
Ctrl-C.

	* net start edna (or any other way to start an NT service)
		This just let NT do its thing.

Since there are so many files that are involved, you can pick up the
whole lot of them, including an updated win32event.pyd, at:
http://lima.mudlib.org/~rassilon/ednaSvc.zip

Mark:
The .pyd hasn't changed, but the rest of the files have been
excessively commented. Feel free to drop this note & the
TCPServerService

The rest of the code is heavily commented, so it should be fairly
simple to see whats going on, if not scream and yell with specific
questions/comments and I will endeavor to answer... :)

I hope this helps,
Bill



From jepler at inetnebr.com  Thu Jul  8 08:19:46 1999
From: jepler at inetnebr.com (Jeff Epler)
Date: Thu, 08 Jul 1999 12:19:46 GMT
Subject: New Python language abuse: ++x with bytecodehacks
Message-ID: <slrn7o9625.tk.jepler@craie.inetnebr.com>

I know just everyone has always neede to write
def f(y):
	x=0
	while ++x < y:
		print x,
	print

in Python.  Well, you always could write it, but it printed the rather
boring
0 0 0 0 0 0 0 ...........................
which is probably not what you wanted.  Well, presenting some code using
Michael Hudson's bytecodehacks.  You can still write the above code, but
now it will print
1 2 3 4 5 6 7 8 9
instead.

The trick (and it will only let you implement preincrement and
predecriment) is that the source
	++x
will produce distinctive code, along the lines of
	LOAD_FAST x
	UNARY_POSITIVE
	UNARY_POSITIVE
which can be turned by bytecodehacks into
	LOAD_FAST x
	LOAD_CONST 1
	BINARY_ADD
	STORE_FAST x
	LOAD_FAST x

Hey, it's abusive, and hey, I sure won't use it in my daily life, and hey,
it'll go wrong in at least one circumstance ("+(+x)").  But that doesn't
matter, bytecodehacks made it possible.

Jeff
PS Who wants to suggest inventive uses for +-x and -+x?  Heck, why not make
+++---+++x call x.sos() ?  I can see it now, calling python methods through
morse code!

from bytecodehacks.code_editor import Function,EditableCode
from bytecodehacks.ops import *

def make_preinc(function):
	f = Function(function)
	code = f.func_code
	cs = code.co_code
	i = 0
	while i < len(cs)-2:
		op = cs[i]
		if op.__class__ in (LOAD_FAST, LOAD_GLOBAL, LOAD_ATTR)  \
				and cs[i+1].__class__ is UNARY_POSITIVE\
				and cs[i+2].__class__ is UNARY_POSITIVE:
			if 1 in code.co_consts:
				a = code.co_consts.index(1)
			else:
				a = len(code.co_consts)
				code.co_consts.append(1)
			if op.__class__ is LOAD_FAST:
				new_op = [LOAD_FAST(op.arg), LOAD_CONST(a), BINARY_ADD(), STORE_FAST(op.arg), LOAD_FAST(op.arg)]
			elif op.__class__ is LOAD_GLOBAL:
				new_op = [LOAD_GLOBAL(op.arg), LOAD_CONST(a), BINARY_ADD(), STORE_GLOBAL(op.arg), LOAD_GLOBAL(op.arg)]
			elif op.__class__ is LOAD_ATTR:
				new_op = [DUP_TOP(), DUP_TOP(), LOAD_ATTR(op.arg), LOAD_CONST(a), BINARY_ADD(),
					STORE_ATTR(op.arg), LOAD_ATTR(op.arg)]
			cs[i:i+3] = new_op
			i=i+len(new_op)
		i=i+1
	return f.make_function()

# Test
import dis

x=lambda x: (x, ++x)
dis.dis(x)
print

y=make_preinc(x)
dis.dis(y)

print x(1), y(1)


def f(x=10):
	y=0
	while ++y < x:
		print y,
	print
dis.dis(Function(f).make_function()); print

g=make_preinc(f)
dis.dis(g)
g()

# vim:ts=4:sw=4
	




From joyce at netcard.com  Mon Jul 19 20:42:42 1999
From: joyce at netcard.com (Joyce Stack)
Date: Tue, 20 Jul 1999 00:42:42 GMT
Subject: python + bash + new user
References: <3793BE52.43E023BB@netcard.com> <7n0f3d$1d7@dfw-ixnews10.ix.netcom.com>
Message-ID: <3793C729.55DE42D0@netcard.com>

Sorry about being vague...i meant the Bash prompt...i have gotten other simple
programs to work...changed mode and executed the programs ...but i dont know how to
run the program and pass a variable at the same time...

Thanking you for you help..

Joyce

Aahz Maruch wrote:

> In article <3793BE52.43E023BB at netcard.com>,
> Joyce Stack  <joyce at netcard.com> wrote:
> >
> >I am not new to programming but new to python...i want to know how do i
> >call functions from the prompt....
>
> From which prompt, python or bash?  The python prompt should look
> something like this:
>
> % python
> Python 1.5.2 (#6, Jun 29 1999, 15:08:27)  [GCC 2.7.2] on sunos5
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>>
>
> You can then type any valid python command or expression at the ">>>"
> python prompt.  For example:
>
> >>> int("3")
> 3
> >>> x=int("3")
> >>> x*x
> 9
>
> Note that commands do not print output, but expressions do.  And, yes,
> startling as it is for a C programmer, "=" is *not* an expression but a
> statement/command.
> --
>                       --- Aahz (@netcom.com)
>
> Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)




From kernr at mail.ncifcrf.gov  Fri Jul  2 18:41:41 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Fri, 02 Jul 1999 22:41:41 GMT
Subject: GUI's etc: PyGtk on Windows
References: <Pine.LNX.4.10.9907021307420.21537-100000@us2.mayo.edu>
Message-ID: <377d3f2c.6001412@news.erols.com>

On Fri, 2 Jul 1999 13:14:42 -0500, Travis Oliphant <olipt at mayo.edu>
wrote:

>
>Previously I mentioned that Gtk works on any system with an X server,
>which is true.
>
>I also want to point out that Gtk is a cross-platform solution
>because there is a port of the GDK (the underlying windowing calls that
>Gtk uses) for Win32.  You can find a link at http://www.gtk.org
>
>From what I've read it looks like Gtk was intentionally written on top of
>GDK so that a port of GDK would make Gtk available.  The person who wrote
>the GDK port to Win32 did it to get the GIMP to run on Windows.
>There are screenshots of the result available on the site.
>
>So if someone develops mainly for X and just wants to make sure that it is
>not impossible to run on Windows, you can use PyGTK.  In fact a great
>program using GTK can be a strong motivator to get the ports of GDK going.

Last time I tried to compile PyGTK on windows (about a week ago with
the June 21 snapshot), there were some symbols missing from the
libraries.

Soon, perhaps.

>Just some thoughts,
>
>Travis

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From junkster at nospam.rochester.rr.com  Sun Jul 25 19:31:57 1999
From: junkster at nospam.rochester.rr.com (Benjamin Schollnick)
Date: Sun, 25 Jul 1999 23:31:57 GMT
Subject: Anti-Port scan.... (Port Monitor / Logger)
References: <rNHvjEdhm5Pp-pn2-vyeUbHbYW9pR@d185d18f8.rochester.rr.com> <14235.39649.786121.5135@dolphin.mojam.com>
Message-ID: <rNHvjEdhm5Pp-pn2-VyPdfEUnvmf7@d185d18f8.rochester.rr.com>

On Sun, 25 Jul 1999 23:18:59, Skip Montanaro <skip at mojam.com> wrote:

>     Benjamin> I'm having some activity on my modem that I can't explain, but
>     Benjamin> I run OS/2....  So I can't use NUKENABBER, etc, for Windows to
>     Benjamin> monitor the ports on my "server" box.

> Can you run Linux?  If so, you can use tcpdump to monitor your ppp
> connection.  It's the best way to look at what's really happening on your
> network. (No Python content, but then there are some things with good
> non-Python solutions already.)

I could use IPTRACE which I believe is the OS/2 equivalent of 
tcpdump...
And I've already isolated it to being non-tcpip traffic.....

But I really would like to try to make a python program to monitor the
ports anyway.....In a nice clean clear format, unlike IPTRACE.... Just
isolate
the "dangerous" bits, and leave me in piece...

Plus IPTRACE dumps way toooooo much data to the drive/log.  I would
easily start to eat up tooo much disk space...

			- Benjamin



From skip at mojam.com  Fri Jul 23 03:22:33 1999
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 23 Jul 1999 02:22:33 -0500 (CDT)
Subject: python + bash + new user
Message-ID: <14232.6016.91164.241221@64.chicago-11-12rs.il.dial-access.att.net>

Spam detection software, running on the system "albatross.python.org", has
identified this incoming email as possible spam.  The original message
has been attached to this so you can view it (if it isn't spam) or label
similar future email.  If you have any questions, see
the administrator of that system for details.

Content preview:  Blake> So my question is this. Can I get python working in
   such a way Blake> that if it's being run from /opt, it will check /opt before
   /usr? Blake> I'm not averse to changing c files, and recompiling. [...] 

Content analysis details:   (5.0 points, 5.0 required)

 pts rule name              description
---- ---------------------- --------------------------------------------------
 0.5 RCVD_IN_PBL            RBL: Received via a relay in Spamhaus PBL
                            [12.75.152.64 listed in zen.spamhaus.org]
 1.6 RCVD_IN_SORBS_DUL      RBL: SORBS: sent directly from dynamic IP address
                            [12.75.152.64 listed in dnsbl.sorbs.net]
 2.0 FH_DATE_IS_19XX        The date is not 19xx.
 0.9 AWL                    AWL: From: address is in the auto white-list


-------------- next part --------------
An embedded message was scrubbed...
From: Skip Montanaro <skip at mojam.com>
Subject: Re: python + bash + new user
Date: Fri, 23 Jul 1999 02:22:33 -0500 (CDT)
Size: 2239
URL: <http://mail.python.org/pipermail/python-list/attachments/19990723/640d5b76/attachment.mht>

From fredrik at pythonware.com  Fri Jul 30 17:18:02 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 30 Jul 1999 21:18:02 GMT
Subject: Include statements with python scripting
References: <7nsu3t$kui$1@nw001t.infi.net>
Message-ID: <001b01bedad1$0230b880$f29b12c2@secret.pythonware.com>

Sean Robertson <seanr at infi.net> wrote:
> How do you do an include statement in python?  I know that in Cold Fusion it
> would be "<cfinclude template="blue_header_noad.html">" and in an shtml file
> it would be "#include http://data.pilotonline.com/connect2/blue_header.html"
> but I need to know how this is done in Python scripting.  This script is
> being embedded in a template for a classifieds page.  I have something like
> this so far:
> 
> <html><head><title><!--#var name--> Search Results</title></head>
> 
> <!--#if mainclass == Employment:-->
>     <!--#include http://data.pilotonline.com/connect2/blue_header.html-->
> <!--#else:-->
>     <!--#include http://data.pilotonline.com/connect2/default_header.html-->
> 
> <font size="+1">Main classification: <!--#var mainclass--> .....

if I'm not mistaken, that's zope DHTML code.  it's
definitely not an ordinary Python script...

it probably works better if you send zope questions 
to one of the zope mailing lists:

    http://www.zope.org/Community/

</F>




From billtut at microsoft.com  Thu Jul 15 10:05:52 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Thu, 15 Jul 1999 14:05:52 GMT
Subject: making a Windows NT service out of a python program
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F4A@RED-MSG-50>


> From: Haimo G. Zobernig [mailto:Haimo.Zobernig at cern.ch]
> 
> Am I doing something wrong, or does that really not work?
> I'd like to be sure that the examples work before I adapt 
> them to my own
> programs.
> 

There are at least one thing wrong with the code, and another thing wrong
with Python 1.5.2 (race condition in deleting thread states).
If you ran the service in debug mode, you might see a message that says
something about trying to delete an invalid threadstate, and python
execution.
You can run the service in debug mode via:
	python pipeTestService.py debug

The sample code for testing large messages is broken.
The service only sees the first 256 bytes of the message.
(Not sure why yet...)
The other problem, is that the call to send the large message is broken as
well.

Here's the bogus line in question:
	data = CallPipe(CallNamedPipe, ("\\\\%s\\pipe\\PyPipeTest" % server,
msg, 512, NMPWAIT_WAIT_FOREVER))

This should be something like:
	data = CallPipe(CallNamedPipe, ("\\\\%s\\pipe\\PyPipeTest" % server,
msg, len(msg)+256, NMPWAIT_WAIT_FOREVER))

If the service had actually read all 4k of the incoming data, CallNamedPipe
would have thrown an exception.

So there's nothing wrong with the service code, just with the sample pipe
handling code. :)

Bill



From jbauer at rubic.com  Thu Jul 22 22:08:26 1999
From: jbauer at rubic.com (Jeff Bauer)
Date: Fri, 23 Jul 1999 02:08:26 GMT
Subject: Non-ODBC MS Sql Connect
References: <3797B848.74D04740@rl.gov>
Message-ID: <3797CE9A.23A6CCAC@rubic.com>

"Roberts, Robert J" wrote:
> How to do I make a non-odbc connection to a MS SQL Server?
> Am running Python Windows NT/IIS.

Argh!  Your post said MS SQL, but my brain saw MySQL.
Sorry for the previous response.

-Jeff



From DOUGS at oceanic.com  Thu Jul  8 14:40:42 1999
From: DOUGS at oceanic.com (Doug Stanfield)
Date: Thu, 8 Jul 1999 18:40:42 GMT
Subject: Two Pythons talking to each other?
Message-ID: <5650A1190E4FD111BC7E0000F8034D26A0EEA3@HUINA>


> From: Roy Smith
> Much as I think evil about NT and all things Microsoft, I would be
> surprised (astonished, actually) if NT didn't handle loopback.

No need for surprise, ping localhost works as expected.

-Doug-



From gerrit.holl at pobox.com  Fri Jul 23 06:47:41 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 23 Jul 1999 10:47:41 GMT
Subject: Fw: ANNOUNCE: Mailing list for Australian Python Users
In-Reply-To: <379798BC.33FC0CA9@equi4.com>; from Jean-Claude Wippler on Fri, Jul 23, 1999 at 12:18:36AM +0200
References: <007201bed42e$c4316c30$f29b12c2@secret.pythonware.com> <wkogh56j2o.fsf@ifi.uio.no> <7n78nm$lic$1@news.accu.uu.nl> <379798BC.33FC0CA9@equi4.com>
Message-ID: <19990723124741.A29309@optiplex.palga.uucp>

On Fri, Jul 23, 1999 at 12:18:36AM +0200, Jean-Claude Wippler wrote:
> Martijn Faassen wrote:
> > 
> > Lars Marius Garshol <larsga at ifi.uio.no> wrote:
> > 
> > > * Fredrik Lundh
> > > |
> > > | python-se, n?gon?  Link?ping?  :-)
> > 
> > > Tjah. Si Oslo, s? kan vi bli en del. :-)
> > 
> > Utrecht is veel beter denk ik, Python-NL en zo..
> 
> Doen.  Ik kom op de fiets.
> 
> > An-large-percentage-of-pythoneers-speaks-Dutch-anyway-ly yours,
> 
> De meerderheid, toch?
> 

Yep.

Toch zie ik in nl.comp.programmeren bijna niemand die python gebruikt...

Yes.

But I don't see anyone using python in nl.comp.programmeren...

groeten,
regards,

Gerrit Holl.



From olevi at my-deja.com  Fri Jul  2 04:03:27 1999
From: olevi at my-deja.com (olevi at my-deja.com)
Date: Fri, 02 Jul 1999 09:03:27 +0100
Subject: TIFF-images
Message-ID: <00001659@bossar.com.pl>

From: olevi at my-deja.com

Is there any way to present TIFF-images through tkinter? After what I've
read, only GIF, PPM and PGM are supported, but maybe there's a way
around?

I'm working with many large images (>8M) both as input and output, so
conversion should be really be avoided. But if anyone knows a good
(loss-less, fast) converter TIFF<->GIF I am also interested in that.

Ole-Kr. Villaboe, ESRF


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From tismer at appliedbiometrics.com  Fri Jul 16 04:48:54 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Fri, 16 Jul 1999 08:48:54 GMT
Subject: Meta-language for mxTextTools
References: <010501becf66$6a316b30$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <378EF1F6.34561319@appliedbiometrics.com>


"Tony J. Ibbs (Tibs)" wrote:
> 
> Christian Tismer wrote:
> >Can you express the parser for the little language with the tagging
> >engine? That would be a funny bootstrap :-)
> 
> My intention is to write the parser to the little language *in* the little
> language, and then to hand-translate it to tuples. It may be that I don't
> have to implement all of the little language before I can use it to
> bootstrap the rest (I seem to remember languages like BCPL worked in this
> manner).

I remember very well! BCPL is mentioned very rarely on this list!
Great, exactly what I meant.

...

> (it's actually just an excuse to understand how mxTextTools
> works, believe it or not!).

I believe it, have been though that tuple assembly, too. :-)
good luck - chris

-- 
     (just increasing the counter. BCPL BCPL BCPL BCPL BCPL BCPL).



From fredrik at pythonware.com  Thu Jul 29 08:26:25 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 29 Jul 1999 12:26:25 GMT
Subject: Why not standard ? (Re: Building a shared libpython)
References: <Pine.GSO.4.03.9907281701150.15012-100000@spica> <001001bed938$b6b34330$f29b12c2@secret.pythonware.com> <37A02B7E.4083@mailserver.hursley.ibm.com>
Message-ID: <007c01bed9bd$df81baa0$f29b12c2@secret.pythonware.com>

Paul Duffin <pduffin at mailserver.hursley.ibm.com> wrote:
> > www.python.org => search => newsgroup search for
> > "shared libpython" brings up lots of hits.  start with
> > these:
> > 
> >     http://www.deja.com/=dnc/getdoc.xp?AN=488431463
> >     http://www.deja.com/=dnc/getdoc.xp?AN=398994871
> > 
> 
> Why is this not available with the standard distribution ?

I thought
http://www.deja.com/=dnc/getdoc.xp?AN=398994871
explained why...

</F>




From gregm at iname.com  Sat Jul 17 10:34:25 1999
From: gregm at iname.com (Greg McFarlane)
Date: Sun, 18 Jul 1999 00:34:25 +1000
Subject: Pmw activate method
In-Reply-To: <7m0q5b$6ba@dfw-ixnews11.ix.netcom.com>; from Reggie Dugard on 07 Jul 1999 at 05:08:10PM
References: <7m0q5b$6ba@dfw-ixnews11.ix.netcom.com>
Message-ID: <19990718003425.53476@nms.otc.telstra.com.au>

Please forgive the slow response...

You have found a bug in Pmw.  I have fixed it and it will be released
in the next version of Pmw.  Based on your solution, I modified
Pmw.initialise() so that it initialises Pmw module data whenever it is
called to initialise a new Tkinter interpreter.  This means that Pmw
will now support repeated calls to (Tkinter.Tk(); Pmw.initialise();
Tkinter.mainloop()) as you are doing, but it will not (yet) support
multiple simultaneous Tk interpreters.

A couple of other points about your code:  Note that __del__ does not
destroy Tkinter widgets (and hence Pmw megawidgets).  This means that
your dialogs are not being deleted at the end of the dialog() method. 
You could either call dialog.destroy() before returning from the
method, or do something like this, which saves re-creating the dialog
each time:

    def dialog(self):
        if not hasattr(self, 'messagedialog'):
            self.messagedialog = Pmw.MessageDialog(
                    title = 'Warning', message_text = 'Testing')
        self.messagedialog.activate()

On 7 Jul, Reggie Dugard wrote:
> Please forgive me if this is a stupid question, because I'm pretty new to
> Python and really new to Pmw.
> 
> I'm writing an GUI application using Pmw on NT 4.0 SP5 and, during
> debugging, I call it repeatedly during a given python interactive session.
> As part of the application, I use some dialogs which I display using the
> activate method.  My problem is, once I've activated a dialog during a
> "run", any subsequent "runs" in the same python session will generate the
> following background error if I attempt to activate a dialog:
> 
> Error: 1
> TclError Exception in Tk callback
>   Function: <method test.dialog of test instance at 898fe0> (type: <type
> 'instance method'>)
>   Args: ()
> Traceback (innermost last):
>   File "...\Python\Lib\Pmw\Pmw_0_8\lib\PmwBase.py", line 1349, in __call__
>     return apply(self.func, args)
>   File "test2.py", line 18, in dialog
>     dialog.activate()
>   File "...\Python\Lib\Pmw\Pmw_0_8\lib\PmwBase.py", line 846, in activate
>     showbusycursor()
>   File "...\Python\Lib\Pmw\Pmw_0_8\lib\PmwBase.py", line 1033, in
> showbusycursor
>     if window.state() != 'withdrawn':
>   File "...\Python\Lib\lib-tk\Tkinter.py", line 861, in wm_state
>     return self.tk.call('wm', 'state', self._w)
> TclError: can't invoke "wm" command:  application has been destroyed
> 
> I have found that the root window does not seem to get cleared from the
> Pmw._toplevelBusyInfo dictionary and if I manually delete it, the error goes
> away, but I'm sure there is a better way of fixing this problem.
> 
> A typical run looks like:
> 
> H:\python>python
> Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import test2
> >>> a=test2.test()  <-- Deletion necessary if dialog was activated
> Deleting root
> >>> a=test2.test()  <--- No deletion if I don't activate the dialog during
> the run
> >>>
> 
> 
> Here is my test2 code to give you a better idea of what I'm doing.  Any
> advice would be appreciated.
> import Tkinter, Pmw
> 
> class test:
>     def __init__(self):
>         root = Tkinter.Tk()
>         Pmw.initialise(root)
>         button = Tkinter.Button(text = 'Dialog', command = self.dialog)
>         button.pack(pady = 20)
>         exit = Tkinter.Button(text = 'Exit', command = root.destroy)
>         exit.pack(pady = 20)
>         root.mainloop()
>         if hasattr(Pmw,'_toplevelBusyInfo') and
> Pmw._toplevelBusyInfo.has_key(root):
>             print 'Deleting root'
>             del Pmw._toplevelBusyInfo[root]
> 
>     def dialog(self):
>         dialog = Pmw.MessageDialog(title = 'Warning', message_text =
> 'Testing')
>         dialog.activate()
> 
> --
> Reggie Dugard
> 
> Merfin, LLC                             Voice:  925-937-4560
> 1460 Maria Lane, Suite 420              FAX:    925-937-4530
> Walnut Creek, CA 94596                  Mailto:reggie at merfinllc.com
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From Alex.Maranda at p98.f112.n480.z2.fidonet.org  Fri Jul  2 08:28:38 1999
From: Alex.Maranda at p98.f112.n480.z2.fidonet.org (Alex Maranda)
Date: Fri, 02 Jul 1999 13:28:38 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001661@bossar.com.pl>

From: Alex Maranda <amaranda at nospam.com>

Giambattista Vico wrote:
> 
> I'd like to hear what people have to say in this thread if we consider
> freezing or otherwise creating a standalone app, even if only for Unix and
> Win. Has anyone found a Python tool or combination of tools that solves
> this problem?
> 
> --tom
Read my previous post on this thread.

Cheers,
-- 
Alex Maranda         mailto: amaranda at spider dot com
Spider Software Ltd. Tel: +44 (0)131 4757036
Edinburgh, UK        http://members.xoom.com/Alex_Maranda
STREAMS based communications protocols for embedded systems



From herzog at online.de  Thu Jul 15 14:00:18 1999
From: herzog at online.de (Bernhard Herzog)
Date: 15 Jul 1999 20:00:18 +0200
Subject: Getters and Setters
References: <000101becedb$1f4c3400$91a02299@tim>
Message-ID: <m3emi9x07h.fsf@greebo.nodomain.de>

"Tim Peters" <tim_one at email.msn.com> writes:

> [Neil Schemenauer]
..
> > ...
> > Your code is fast and elegant but it causes reference loops.
> 
> ?  Not under my Python <wink> -- stick a __del__ method in the test class
> that prints something and you'll see that objects go away when you expect.
> The function object assigned to self.set/get_XXX contains one ref to
> self.__dict__ and another to the attr's string name, and that's it; it is
> *not* a bound method object (which would also hold a reference to self),
> just a function object.  So there's no cycle unless you force one, like
> 
>     instance.setWhatever(instance)
> 
> But then you would have had a cycle without this code too.

The cycle doesn't involve self, but self's dict will hold an indirect
reference to itself. The instance object will be deleted, so it's
__del__ method is called, but its __dict__ won't be deleted:

Python 1.5.2 (#3, Apr 21 1999, 16:51:54)  [GCC 2.7.2.3] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> class C:
..     def __del__(self):
..             print '*del*'
.. 
>>> c = C()
>>> c.x = C()
>>> del c
*del*
*del*
>>> c = C()
>>> c.x = C()
>>> c.y = c.__dict__
>>> del c
*del*
>>> 



-- 
Bernhard Herzog	  | Sketch, a python based drawing program
herzog at online.de  | http://www.online.de/home/sketch/



From arcege at shore.net  Mon Jul 19 20:21:21 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Tue, 20 Jul 1999 00:21:21 GMT
Subject: Embedding: Defining packages?
References: <7mq9r5$irj$1@news.rz.uni-karlsruhe.de>
Message-ID: <5iPk3.527$ui4.177861@news.shore.net>

Bjoern Giesler <un4e at rz.uni-karlsruhe.de> wrote:
: Hi,

: I'm embedding Python in a rather complex C program. To represent the
: complexity (and to keep the thing modular), I want to define several 
: Python modules and group them in a package hierarchy:

:          MyProgram
:          +- Server
:          |       +- Communications
:          |       +- HardwareAccess
:          +- Client
:                  +- Communications
:                  +- AbstractObjects

: ...etc. How do I do that? Is there a way to pass hierarchical module names
: to Py_InitModule, such as

: theModule = Py_InitModule(methodDefs, "MyProgram.Server.Communications");

: My attempts so far haven't worked.

: Thanks in advance,
:          --Bjoern

: PS I'm posting this via the newsgroup, as sending mail to python-list at cwi.nl
:    doesn't seem to work. Has anything changed there that I'm not aware of?

You probably need to reverse the order of the arguments
  (http://www.python.org/doc/current/ext/methodTable.html)

theModule = Py_InitModule("MyProgram.Server.Communications", methodDefs);

(I've tried this format successfully.)

  -Arcege




From nick.bower at ssec.wisc.edu  Sat Jul 10 22:02:58 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sat, 10 Jul 1999 21:02:58 -0500
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787D8DA.3774AA7D@ssec.wisc.edu> <7m8p7t$o7s@dfw-ixnews10.ix.netcom.com>
Message-ID: <7m9to7$6qg$83@rks1.urz.tu-dresden.de>

yeah, that's what i originally thought.  sure enough there were some updates in the
os documentation, but not os.chdir.  it's something that would be really helpfull
if anyone linked to the documentation sees this.

nick

Aahz Maruch wrote:

> In article <3787D8DA.3774AA7D at ssec.wisc.edu>,
> Nick Bower  <nick.bower at ssec.wisc.edu> wrote:
> >
> >Incidentally, I went back to the library reference and still couldn't find
> >os.chdir
>
> You probably have an old version of the docs; I suggest you go to
> www.python.org and download a new set.
> --
>                       --- Aahz (@netcom.com)
>
> Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)

--

Nick Bower
Space Science and Engineering Center
University of Wisconsin, Madison, USA
Phone: (608) 265 8007
Email: nick.bower at ssec.wisc.edu
Web: http://arm1.ssec.wisc.edu/~nickb





From da at ski.org  Wed Jul 14 14:04:19 1999
From: da at ski.org (David Ascher)
Date: Wed, 14 Jul 1999 11:04:19 -0700 (Pacific Daylight Time)
Subject: newbie questions about sys.path, wxpython, IDLE and pythonwin
In-Reply-To: <378cc9b6.22412677@aplnews>
Message-ID: <Pine.WNT.4.04.9907141102050.212-100000@rigoletto.ski.org>

On Wed, 14 Jul 1999, Matt Feinstein wrote:
> >Python uses a combination of the values you found in the registry, and some
> >default values - things like ".\lib" are added always, regardless of the
> >registry.  Remember, you have the same source to see how it works that I do.
> >
> 
>  Well, with all due respect, this can't be the whole story--  after
> installing the numeric extension, I found the numeric directory in
> sys.path, but not in the registry list.  So, what -is- going on here?

NumPy uses a .pth file which is used by Python at startup to modify the
path as well.  .pth files are discussed at:
 
     http://www.python.org/doc/essays/packages.html

(With all due respect, Mark's comment necessarily is the whole story,
since he pointed you at the code =)

--david ascher




From mstenber at cc.Helsinki.FI  Tue Jul  6 02:58:55 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 06 Jul 1999 09:58:55 +0300
Subject: COM with Python?
Message-ID: <al8wvwe6z80.fsf@myntti.helsinki.fi>

Is there some form of _somewhat_ more in-depth, but less abstract source of
information for COM+Python than the 'quick client/server'/the slides
respectively. 

Spesifically, I'm looking for information on activation policies; how to
implement idispatch objects that are singletons for a single machine.

-Markus

-- 
"The truth is out there."					-X-files
"The truth may be out there, but lies are inside your head."	-T.Pratchett



From tgabriel at bellsouth.net  Fri Jul 16 22:09:55 1999
From: tgabriel at bellsouth.net (Xenophanes)
Date: Fri, 16 Jul 1999 22:09:55 -0400
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net>
Message-ID: <378FE5F3.A8394D50@bellsouth.net>

Hi,

First try to get your money back for the Unix books you bought. Python
is not an o/s, and unless you are on a system for with the Unix
operating system, you do not need to know it. If you are on a windows
system, Python runs fine, you just have to have the correct version.
Also, there is no relationship at the beginning level with the operating
system as long as the version of Python you have is ported to whatever
o/s you are using on your own pc. At some advanced level, you will be
interfacing with the o/s but for now it does not matter too much.

There are several good tutorials on the Python web site (www.python.org)
you can download. Just click on documentation and you will see it listed
there. The program software for the Python language you downloaded has a
complete suite of documents with it you can read and print out. This
also includes Guido van Rossum's own tutorial which is very good. One
caution about it is that you will need a little background in
programming in order to get the most benefit out of it. There is one
tutorial in particular that is good for a quick start for a non
programmer called A Non Programmer's Tutorial for Python. It is at
www.python.org/doc/intros.html, if you are interested.

Since you indicate you are a beginner, I would like to recommend to you
O'Reilly's Learning Python by Mark Lutz and David Ascher for $23.96 plus
shipping from Amazon.com. For a more advanced treatment O'Reilly's
Programming Python is good. Search Amazon for it too. Both of these
books are very good at presenting Python is an understandable way.

bye,

X
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19990716/adf315da/attachment.html>

From Fredrik.Lundh at p98.f112.n480.z2.fidonet.org  Fri Jul  2 04:02:06 1999
From: Fredrik.Lundh at p98.f112.n480.z2.fidonet.org (Fredrik Lundh)
Date: Fri, 02 Jul 1999 09:02:06 +0100
Subject: embedding; return array
Message-ID: <00001657@bossar.com.pl>

From: "Fredrik Lundh" <fredrik at pythonware.com>

Randy Heiland wrote:
> I'd welcome any pointers into being able to return an (unsigned char)
> array from an embedded module.

how about:

    return Py_BuildValue("s#", buffer, buffer_size);

(gives you a string; you can use the array module to
process it further).

if you really want to return it as a python list, you
have to build the list yourself.  something like:

    PyObject* list;

    list = PyList_New(buffer_size);
    if (list)
        for (i = 0; i < buffer_size; i++) {
            PyObject* item;
            item = PyInt_FromLong(buffer[i]);
            if (item == NULL) {
                Py_DECREF(list);
                list = NULL;
                break;
            }
            PyList_SetItem(list, i, item);
        }
    }

    return list;

should do the trick.

</F>



From dan.pettersson at its.uu.se  Thu Jul 22 05:29:07 1999
From: dan.pettersson at its.uu.se (Dan Pettersson)
Date: Thu, 22 Jul 1999 11:29:07 +0200
Subject: tkinter call backs
References: <01bed202$58879280$e855c3d1@Cyberus.cyberus.ca> <c754siy6dja.fsf@pc142.cosc.canterbury.ac.nz>
Message-ID: <3796E463.CFE50301@its.uu.se>

Timothy R Evans wrote:
> class Command:
>     def __init__(self, func, *args, **kw):
>     ...
>         apply(self.func, args, kw)
> 
> Used as following:
>
> Button(text='foo', command=Command(myfunction, myarg))
> 
> The args (including keywords) that are passed to the Command class are
> passed on to myfunction when the callback is activated.
> 
> I'm wondering if this should be part of Tkinter in some way?

You can get the same behaviour by using a lambda and defaults

Button(text='foo', command=lambda e,x=arg1, y=arg2 : function(e,x,y))

Apologies if all arguments aren't in the write place my TkInter is a bit
rusty.
--
Dan Pettersson				IT Support, Uppsala University, SWEDEN



From phd at emerald.netskate.ru  Thu Jul  8 11:44:02 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Thu, 8 Jul 1999 19:44:02 +0400 (MSD)
Subject: Help with strings please
In-Reply-To: <378555CC.8B65F809@freenet.co.uk>
Message-ID: <Pine.LNX.4.04.9907081942540.13521-100000@emerald.netskate.ru>

On Fri, 9 Jul 1999, Sreekant Kodela wrote:
> I seem to be in trouble here. I am trying to make some cgi scripts for
> my web site. I am using the postgresql as the backend db and a form
> based address book and news forum etc. The problem is everything is
> working ok as long as noone uses ' and " in the text. If they do, bingo
> , postgres refuses to accept. I tried using string library [mainly
> string.join(string.split('awkward'"'st'ring')). It works sometimes and
> doesn't some times.
> 
> Any better and efficient way to deal with this situation !
> Any suggestions and ideas will be greatly appreciated.

s = string.replace(s, "'", "''")
s = string.replace(s, '"', '\\"')

> Thanks
> sreekant

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From nick.bower at ssec.wisc.edu  Wed Jul 21 15:16:51 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Wed, 21 Jul 1999 14:16:51 -0500
Subject: Scientific Plotting?
References: <19990721010203.96101.qmail@hotmail.com> <3795DF85.320EADBE@ssec.wisc.edu> <Pine.LNX.4.10.9907211108100.11968-100000@us2.mayo.edu>
Message-ID: <37961CA3.186124E7@ssec.wisc.edu>

> I totally agree with you.  I'm encouraged by the efforts made and
> the progress of PIDDLE and Graphite.  You should look in to them
> if you want to help out to make python more plotting-friendly.

Sorry to ask a silly question, but where can I find this Graphite?  I've
been searching the python site with no success.

cheers, nick

-- 

Nick Bower
Space Science and Engineering Center
University of Wisconsin, Madison, USA
Phone: (608) 265 8007
Email: nick.bower at ssec.wisc.edu
Web: http://arm1.ssec.wisc.edu/~nickb



From mumac at wiu.edu  Thu Jul 15 23:08:02 1999
From: mumac at wiu.edu (Mark Clark)
Date: Thu, 15 Jul 1999 22:08:02 -0500
Subject: Newbie ? - Starting Windows program/Sending text to window
Message-ID: <7mm7ne$97v$1@news1.ecnet.net>

    I have *just* started trying out Python.  I would like to start a
Windows program (any program - not necessarily one that can accept DDE) and
then just send some text to the window (like SendKeys does).   I am also
using the Win32 extensions from Mark Hammond.
    To test, I have gotten notepad to start with ShellExecute (great!).  Now
I cannot figure out how to send text to the notepad window.  I thought some
sort of "print" or "write" would work, but it doesn't seem to.   Can you
help me?  Thank you!





From parkw at better.net  Fri Jul  9 13:02:28 1999
From: parkw at better.net (William Park)
Date: Fri, 9 Jul 1999 13:02:28 -0400
Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
In-Reply-To: <19990709144800.46076@devel.math.uni-bonn.de>; from Thomas Ackermann on Fri, Jul 09, 1999 at 12:48:00PM +0000
References: <tot2m7.su.ln@final.math.uni-bonn.de> <19990708223856.A29441@optiplex.palga.uucp> <19990709144800.46076@devel.math.uni-bonn.de>
Message-ID: <19990709130228.A1036@better.net>

On Fri, Jul 09, 1999 at 12:48:00PM +0000, Thomas Ackermann wrote:
    <more>...</more>
> Thanx - But i know HTML (and Java) itself already	;-)
> What i do not know are the Python modules for it and wanted to get
> some information from people who used several of the alternatives
> already.  But thanx anyway ...

I downloaded HTMLgen and Zope (and few other packages from
www.python.org), read all their documentations, and installed them.
But, I just couldn't figure out how to use them for my need.  And, no,
I'm not a newbie.

Some time ago, I wanted to generate HTML output with LaTeX-like
structures (ie. numbered sections, math expressions as transparent GIF
images, references, footnotes).  So, I wrote 'HTMLtag' and uploaded to
'ftp.python.org'.  It's not converter like 'html2latex'.  It only spit
out HTML <tag>...</tag> format, as opposed to manually editing .html
file or using lots of 'print' in .py file.

> 
> 
> > > So, what to use to allow for an Internet-Shop with pages created from
> > > databases and allowing customers to purchase over net?
> > > 
> > 
> > If you're doing such serious things, you should 
> > 
> >                                      **really**
> > 
> > learn web programming!!
> 
> Again, i want to use Python and a decent HTML/CGI module, i never said to
> be completely unexperienced with WWW stuff ...

As long as your ISP has Python and mSQL, mySQL, ..., it should be
straight-forward.  There are modules in the python distribution for
parsing HTML and CGI input.  For producing HTML output, it's sort of
up to you, as long as you know HTML.

William Park



From conrad at vanir.srl.caltech.edu  Sat Jul 10 14:24:33 1999
From: conrad at vanir.srl.caltech.edu (Conrad Steenberg)
Date: 10 Jul 1999 18:24:33 GMT
Subject: Printing returned object in embedded app
References: <378694FC.D857F68A@srl.caltech.edu> <6wxh3.1472$6M6.450919@news.shore.net>
Message-ID: <7m9tnq$6qg$68@rks1.urz.tu-dresden.de>

On Sat, 10 Jul 1999 01:40:18 GMT, Michael P. Reilly <arcege at shore.net> wrote:
>
>Try:
>  PyObject_Print(object, stdout, Py_PRINT_RAW);

Thanks! 

Conrad
-- 



From bwinton at tor.dhs.org  Sat Jul 24 16:57:07 1999
From: bwinton at tor.dhs.org (Blake Winton)
Date: Sat, 24 Jul 1999 20:57:07 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
References: <001101bed3f2$5e8874a0$642d2399@tim> 	<7n675p$qsk@dfw-ixnews7.ix.netcom.com> 	<14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net> 	<379798D9.FF3F811B@compaq.com> <14232.32203.792051.822513@weyr.cnri.reston.va.us> <37988C11.DB7D67C7@callware.com>
Message-ID: <slrn7pkanm.gn7.bwinton@tor.dhs.org>

On Fri, 23 Jul 1999, Ivan Van Laningham <ivanlan at callware.com> wrote:
>"Fred L. Drake" wrote:
>>         import Tkinter as Tk
>>         from module import function as func
>Now, this I like.  I haven't seen this proposal before (or maybe I have
>and senility is setting in), but you can now count me in as a proud
>supporter of this idea.

I seem to remember this idea from a few months ago (I checked deja, but
couldn't find it).  I also seem to remember Tim or Guido convincing me
that it wasn't really necessary, and things could be done almost as well
another way.  Or perhaps I'm just getting old and set in my ways.  ;)

>Any reason it couldn't be done?  What bad things would happen if this
>were implemented and "from spam import *" went away?

Well, billions of Tkinter programs would break...

(Speaking of which, other weirdness just happened...
>>> import urllib
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/opt/lib/python1.5/urllib.py", line 25, in ?
    import socket
ImportError: No module named socket
>>> print sys.path
Traceback (innermost last):
  File "<stdin>", line 1, in ?
NameError: sys
>>> import sys
>>> print sys.path
['', '/opt/lib/python1.5/', '/opt/lib/python1.5/test',
'/opt/lib/python1.5/plat-linux-i386', '/opt/lib/python1.5/lib-tk',
'/opt/lib/python1.5/lib-dynload']
>>> import urllib
>>>

Funky, eh?)

Later,
Blake.

-- 
One Will.  One Dream.  One Truth.  One Destiny.  One Love.



From donn at u.washington.edu  Wed Jul  7 11:53:15 1999
From: donn at u.washington.edu (Donn Cave)
Date: 7 Jul 1999 15:53:15 GMT
Subject: Py_BuildValue, PyArg_ParseTuple
References: <37835CB3.2B876BE3@ncsa.uiuc.edu>
Message-ID: <7lvt5b$141q$1@nntp6.u.washington.edu>

Randy Heiland <heiland at ncsa.uiuc.edu> writes:
| Here's a very simple program where I want to get back out
| (PyArg_ParseTuple) what I put in (Py_BuildValue) - however, I can't seem
| to do that.  Can someone PLEASE enlighten me?
...
|  pobj = Py_BuildValue("s#", ba,3);
...
|  PyArg_ParseTuple(pobj,"s#", &cstr,&len);

I'm just guessing, from the name, that PyArg_ParseTuple() expects a tuple.
If it's important to get a tuple here, you could try
Py_BuildValue("(s#)", ba, 3)

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From tgabriel at bellsouth.net  Mon Jul 19 07:00:56 1999
From: tgabriel at bellsouth.net (Xenophanes)
Date: Mon, 19 Jul 1999 07:00:56 -0400
Subject: what is python?
References: <7mt81n$s6e$1@bgtnsc03.worldnet.att.net>
Message-ID: <37930568.78059AC0@bellsouth.net>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19990719/87f43145/attachment.html>

From medley8086 at aol.com  Wed Jul 28 05:27:04 1999
From: medley8086 at aol.com (Medley8086)
Date: 28 Jul 1999 09:27:04 GMT
Subject: a little motivation, anyone?
Message-ID: <19990728052704.10608.00002260@ng-fq1.aol.com>

Ok, I'm running a win98 partition and a linux partition.  I have all sorts of
programming tools, and have studied how to program in assembly and c++.  In
addition, I plan to know c (should I bother after c++?  give me your opinion),
pascal, euphoria, and some scripting languages by the end of the summer. 
Through all this planning, however, I keep running into the same problem with
learning how to program.  I'm teaching myself, so the motivation isn't there. 
I've tried thinking of things that i think would be challenging for me to
create, rather than necessary, so now the most off-putting thing is not knowing
whether my progress is good.  What I'm asking is, could i please have some
posts of what some of you other programmers made when you were starting out, in
what languages, and how long it took.  I'm looking for stories to convince me
while trying to work with stubborn code that I'm not algorithmically
handicapped or just not cut out to be a coder at all.  Thanks in advance for
anyone who decides to help.

E. Leron Culbreath




From jim at interet.com  Wed Jul  7 09:41:17 1999
From: jim at interet.com (James C. Ahlstrom)
Date: Wed, 07 Jul 1999 09:41:17 -0400
Subject: mxCGIPython question (was: Build the Python library into the binary)
References: <376F9518.D3252CEE@interet.com> <7kolnv$gp6$1@nnrp1.deja.com> <015e01bebe1f$202e9560$4500a8c0@thomasnotebook> <37821C75.405CD9C5@lemburg.com>
Message-ID: <378358FD.DB5D2186@interet.com>

Thomas Heller wrote:
> > Is it possible to build a python15.dll containing the
> > standard library or is it only possible to freeze executables?

Yes, it is possible and easy to freeze the library into python15.dll
and this is preferred, since its version must match the binary.  Just
link the frozen code into python15.dll, and assign the pointer
PyImport_FrozenModules to your array of struct _frozen.  You don't
even have to export this symbol, because it is only used by functions
which are already in python15.dll.

The above guarantees you have the correct libraries for the Python
interpreter.  The problem is that you can not use freeze for anything
else, since there is only one frozen array currently allowed.  If this
is a problem, let me know because I have code to fix it.

Jim Ahlstrom



From fredrik at pythonware.com  Mon Jul 26 09:37:23 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 13:37:23 GMT
Subject: accent (special char) with python
References: <932980082.887262@sofia.magic.fr>
Message-ID: <00a401bed76b$ff8e2210$f29b12c2@secret.pythonware.com>

"nom <d at d.dd>" wrote:
> i'm from france

hmm. last time I checked, france was ".fr",
not ".dd"...

> then i use accent : i try to convert a string :
> 'Soci\351t\351' to 'Soci?t?'
> how can do it?

print it?

>>> print 'Soci\351t\351'
Soci?t?

...

Python is mostly agnostic when it comes to
character sets -- strings just contain 8-bit
characters, and it's up to the programmer
to make sure they're interpreted in the right
way on input or output.   in your case,
'Soci\351t\351' is an ISO Latin 1 string (also
known as ISO 8859-1). so if your environment
uses ISO Latin 1, it just works. on the other
hand, if your environment were to use, say,
IBM's old PC encoding (like in the MS-DOS
window under Windows), it would come out
as:

>>> print 'Soci\351t\351'
Soci?t?

and in an UTF-8 environment, it's an
illegal string:

>>> print unicode('Soci\351t\351')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
ValueError: invalid UTF-8 code

...

so I guess the answer to your question is
"depends on what you're doing..."

but before you try to explain that, please
take a look at Jukka Korpela's character
code tutorial, available from:

http://www.hut.fi/~jkorpela/chars.html

    "This document in itself does not
    contain solutions to practical problems
    with character codes; rather, it gives
    background information needed for
    understanding what solutions there
    might be, what the different solutions
    do - and what's really the problem in
    the first place."

</F>




From S.I.Reynolds at cs.bham.ac.uk  Tue Jul 13 07:48:57 1999
From: S.I.Reynolds at cs.bham.ac.uk (Stuart Reynolds)
Date: Tue, 13 Jul 1999 12:48:57 +0100
Subject: Pickling C objects
Message-ID: <378B27A9.4201@cs.bham.ac.uk>

I'm writing some new Python classes with a C implementation. How do you
make a C object pickleable?

Is it possible to use the __repr__ method? It would be easy for me to
output some Python code as a string which generated the object:

 'MyClass( MyComponent1(1,2,3), MyOtherComponent2('a','b','c') )'


Stuart



From arcege at shore.net  Thu Jul 22 21:07:44 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Fri, 23 Jul 1999 01:07:44 GMT
Subject: Embedding: Defining packages?
References: <7mq9r5$irj$1@news.rz.uni-karlsruhe.de> <5iPk3.527$ui4.177861@news.shore.net> <7n1a37$s1$1@news.rz.uni-karlsruhe.de>
Message-ID: <AfPl3.958$ui4.298570@news.shore.net>

Bjoern Giesler <un4e at rz.uni-karlsruhe.de> wrote:
: Hi,

: Michael P. Reilly <arcege at shore.net> wrote:
: : Bjoern Giesler <un4e at rz.uni-karlsruhe.de> wrote:
: : : theModule = Py_InitModule(methodDefs, "MyProgram.Server.Communications");

: : theModule = Py_InitModule("MyProgram.Server.Communications", methodDefs);

: ...was of course what I meant. Mine was just a typo. Still, this gives me a
: module named "MyProgram.Server.Communications", but _no_ module "MyProgram".
: What I want is to be able to say
: 	from MyProgram import *
: and then address functions via
: 	Server.Communications.foo()

: This doesn't work; it seems that Python can't define _packages_ from C, but
: only _modules_. Is that correct?

Yes, I see what you mean now; sorry.  Correct, you cannot create
"package modules" that can be imported through "from", but you can
create the naming conventions.  Most of the reason is how the importer
works, it is looking for a file with the submodule name, not a module
with that dotted name.  You would have to install a replacement for the
__builtin__.__import__ function.

  -Arcege




From gmcm at hypernet.com  Tue Jul  6 23:41:16 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 6 Jul 1999 22:41:16 -0500
Subject: COM with Python?
In-Reply-To: <7lsrsp$fh2$1@m2.c2.telstra-mm.net.au>
Message-ID: <1280807192-67537824@hypernet.com>

[Markus Stenberg] 
> >Spesifically, I'm looking for information on activation policies; how to
> >implement idispatch objects that are singletons for a single machine.
[Mark Hammond]
> Your best bet is DejaNews - the singleton subject was done fairly
> thourghly a while ago, and no significan changes have been made to
> the COM package to make it any eaiser (alas)

Hmm. I've read those posts, and never did find a resolution.

There's a thread "connect to running python COM server", which ends 
with code from Andy that doesn't look to me like it would really 
create a singleton (it looks like it would only allow one instance 
within a process, but not one instance on a machine). Does it work?

There's also a thread "I'm BaaaAAAAAAaaaaaaak" where you find some 
code (horribly trashed by being posted with leading tabs and getting 
htmlized) that appears to try to use Policy objects to create a 
singleton. Again, to me it looks like it would work in-process, but 
not machine wide.

I have had success with using the ROT, which is very easy to do. 
Unfortunately, if I leave the (singleton) server running for an 
extended period of time (say 6 to 24 hrs), at some point I find my 
machine locked up. This is NT 4 SP4 on a _very_ stable machine - the 
only BSODs I've ever seen were when I had a parallel port Zip disk 
attached. (I moved the Zip disk to my Linux box, where it's performed 
flawlessly ever since :-)).

- Gordon



From nascheme at ucalgary.ca  Thu Jul 15 02:22:38 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 15 Jul 1999 06:22:38 GMT
Subject: Getters and Setters
References: <7mjrqg$e4o@ds2.acs.ucalgary.ca>
Message-ID: <7mjune$e2g@ds2.acs.ucalgary.ca>

It seems my implementation had a bug.  It did not work when the
function was stored and called multiple times.  Premature
optimization really is the root of all evil.


    Neil

======================================================================
#
# A mixin class to automagicly create getter/setter methods for
# class attributes. By default the method names are get* and
# set*. Modify as desired. Using the automatic methods seems to
# be about 10 times slower than explicit methods.
#
# Neil Schemenauer, July 1999
#

class _Getter:
    def __init__(self, value=None):
        self.value = value

    def __call__(self):
        return self.value

class _Setter:
    def __init__(self, dict=None, name=None):
        self.dict = dict
        self.name = name

    def __call__(self, value):
        self.dict.__dict__[self.name] = value
        

class GetterSetter:
    def __getattr__(self, name):
        try:
            if name[:3] == 'get':
                return _Getter(self.__dict__[name[3:]])
            elif name[:3] == 'set':
                return _Setter(self, name[3:])
        except KeyError:
            pass
        raise AttributeError


def test():
    import time
    class A(GetterSetter):
        def __init__(self):
            self.X = self.Y = 1
        def getY(self):
            return self.Y
        def setY(self, v):
            self.Y = v
    a = A()
    t = time.time()
    n = 1000
    print 'Comparing', n, 'iterations'
    for x in range(n):
        a.getX()
        a.setX(x)
    print 'GetterSetter', time.time()-t
    t = time.time()
    for x in range(n):
        a.getY()
        a.setY(x)
    print 'explicit function', time.time()-t


if __name__ == '__main__': test()



From thomas at xs4all.nl  Wed Jul 21 10:17:05 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Wed, 21 Jul 1999 16:17:05 +0200
Subject: uu.py - get result of uuencode() in a string
In-Reply-To: <3795CDC1.40F5E36@freemail.nl>; from Richard van de Stadt on Wed, Jul 21, 1999 at 03:40:17PM +0200
References: <3795CDC1.40F5E36@freemail.nl>
Message-ID: <19990721161705.U6170@xs4all.nl>

On Wed, Jul 21, 1999 at 03:40:17PM +0200, Richard van de Stadt wrote:

> Is it possible to get the output of the uuencode() function
> into a string, instead of a file or stdout?

You can give uu.encode() File objects, so you can also give them StringIO
objects.

>>> import StringIO
>>> import uu
>>> import operator
>>> infile = open("/etc/hosts.allow",'r') 
>>> outfile = StringIO.StringIO()
>>> uu.encode(infile, outfile)
>>> outfile.buflist  
['begin 666 -\012',
'M(PHC(&AO<W1S+F%L;&]W"51H:7, at 9FEL92!D97-C<FEB97,@=&AE(&YA;65S\012',
'M(&]F(\'1H92!H;W-T<R!W:&EC:"!A<F4*(PD)86QL;W=E9"!T;R!U<V4@=&AE\012',
'M(&QO8V%L($E.150@<V5R=FEC97,L(&%S(&1E8VED960*(PD)8GD@=&AE("<O\012',
':=7-R+W-B:6XO=&-P9"<@<V5R=F5R+ at HC"@H \012', ' \012end\012']

>>> print reduce(operator.add, outfile.buflist)
begin 666 -
M(PHC(&AO<W1S+F%L;&]W"51H:7, at 9FEL92!D97-C<FEB97,@=&AE(&YA;65S
M(&]F('1H92!H;W-T<R!W:&EC:"!A<F4*(PD)86QL;W=E9"!T;R!U<V4@=&AE
M(&QO8V%L($E.150@<V5R=FEC97,L(&%S(&1E8VED960*(PD)8GD@=&AE("<O
:=7-R+W-B:6XO=&-P9"<@<V5R=F5R+ at HC"@H 
 
end

>>> 

StringIO objects are simple string objects that behave like filehandles
(they support the same functions) but store their data in strings.
My apologies for the ugly print reduce(operator.add) etc, but i haven't had
much more than 5 hours sleep in the last three days. work work work ;-P

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From lvirden at cas.org  Fri Jul  2 08:26:13 1999
From: lvirden at cas.org (lvirden at cas.org)
Date: Fri, 02 Jul 1999 13:26:13 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <0000166b@bossar.com.pl>

From: lvirden at cas.org


According to Fredrik Lundh <fredrik at pythonware.com>:
:What native API?  As of 8.0, Tk still emulates all widgets
:except the scrollbar, and use a somewhat ineffective
:X emulation layer for graphics.  AFAIK, they haven't
:changed that in 8.1...

I suspect that Tk features such as these are the native 'api's to which
the original poster was referring.  The following are a few of
the entries from Tk 8.x changes file.


8/22/96 (new feature) There are four new commands for opening common
dialog boxes: tk_chooseColor, tk_getOpenFile, tk_getSaveFile and
tk_messageBox. Native dialog boxes are used wherever available. (IL)

11/7/96 (feature change) Under Windows, buttons and scrollbars now
have native look and feel.  This affects the default class bindings
and the way the some configuration options are interpreted.  Refer to
the widget manual pages for more details. (SS)
*** POTENTIAL INCOMPATIBILITY ***

12/17/96 (feature change) Native Windows labels do not get a focus-ring
border.  (CS)

7/30/97 (feature change) As a result of native menus, you can no
longer drag through a frame of menubuttons on Macintosh and Windows
and have the menus pop down. You can still click on individual
menubuttons and their menus will pop down.  Applications needing to
present a menubar should consider using the new "-menu" configuration
of the toplevel widget to set up menubar which behaves correctly on
Macintosh, Windows and X Windows. (SRP)

10/21/97 (bug fix) Tk did not reset the result after native dialog
modal loops on Window so background events could perturb the dialog
result. (SS)

-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.



From mal at lemburg.com  Thu Jul  8 08:31:59 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 08 Jul 1999 14:31:59 +0200
Subject: .pyo's without .pyc's?
References: <19990526140137.A66609@vislab.epa.gov> <374C6C6E.1D07326B@lemburg.com> <19990708074113.A61389@epa.gov>
Message-ID: <37849A3F.28C0EF21@lemburg.com>

[CCed to Guido, because I think what you've found is true bug in 1.5.2
 and earlier versions]

Randall Hopper wrote:
> 
> M.-A. Lemburg:
>  |Looks like you have setup PYTHONPATH to point to some old
>  |extensions.py file (the compiled versions seemed to be ok).
> 
> It doesn't appear this is the case (please see below).
> 
> Unsuccessful exceptions module search -- *.py and *.pyc pruned:
> 
> > find . -name 'exceptions*'
> ./lib/python1.5/exceptions.pyo
> 
> > bin/python -O -vv | & grep except
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.so
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptionsmodule.so
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.py
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.pyc
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/plat-irix646-n32/exceptions.so
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/plat-irix646-n32/exceptionsmodule.so
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/plat-irix646-n32/exceptions.py
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/plat-irix646-n32/exceptions.pyc
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-tk/exceptions.so
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-tk/exceptionsmodule.so
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-tk/exceptions.py
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-tk/exceptions.pyc
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-dynload/exceptions.so
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-dynload/exceptionsmodule.so
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-dynload/exceptions.py
> # trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-dynload/exceptions.pyc
> 'import exceptions' failed; traceback:
> ImportError: No module named exceptions
> Warning!  Falling back to string-based exceptions

I've checked this and indeed you're right. exceptions.py is imported
prior to all other modules at a time where the switch from .pyc to
.pyo hasn't been done yet (this is done in _PyImport_Init(), yet
exceptions.py is imported by _PyBuiltin_Init_2()):

>From pythonrun.c:

	bimod = _PyBuiltin_Init_1();
	if (bimod == NULL)
		Py_FatalError("Py_Initialize: can't initialize __builtin__");
	interp->builtins = PyModule_GetDict(bimod);
	Py_INCREF(interp->builtins);

	sysmod = _PySys_Init();
	if (sysmod == NULL)
		Py_FatalError("Py_Initialize: can't initialize sys");
	interp->sysdict = PyModule_GetDict(sysmod);
	Py_INCREF(interp->sysdict);
	_PyImport_FixupExtension("sys", "sys");
	PySys_SetPath(Py_GetPath());
	PyDict_SetItemString(interp->sysdict, "modules",
			     interp->modules);

	/* phase 2 of builtins */
	_PyBuiltin_Init_2(interp->builtins);
	_PyImport_FixupExtension("__builtin__", "__builtin__");

	_PyImport_Init();

To work around this, simply add the excpetions.py file (and only
this one) to the python1.5/lib/ directory.

Still, I think you're right in that the .pyc -> .pyo switch should
be done prior to importing *any* Python module, meaning that
_PyImport_Init() should be called prior to importing anything.

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   176 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From tony at lsl.co.uk  Fri Jul 16 04:24:39 1999
From: tony at lsl.co.uk (Tony J. Ibbs (Tibs))
Date: Fri, 16 Jul 1999 08:24:39 GMT
Subject: Meta-language for mxTextTools
Message-ID: <010001becf64$a615fa50$f0c809c0@lslp7o.lsl.co.uk>

M.-A. Lemburg wrote:
> This looks very promising :-)


(*blush*)

> One itch that I have with the syntax: since you are
> already using Pythonish syntax, why not simply write
>
> def keyword:
>     ...
>
> instead of
>
> keyword is:
>     ...


"is" was what I originally thought of, because I was trying to think of
"reading" the tag table definition back to myself, and that sounded better.
When I realised that I was wanting to be "Python-esque" I realised that
using "def" would perhaps be more consistent, but by then I'd got used to
"is", and also I couldn't help feeling that "def" was less clear *to me* - I
wanted it to be obvious to me that I wasn't defining a function (perhaps I'm
still (vaingloriously) hoping to allow Python code in tag-table files...).
However, it's probably not a religious issue, and may perhaps change before
things get released. How strong an itch is it?

> An alias "back" for -1 would help reading the tables:
>
> Skip back
> vs.
> Skip -1


I would be happy with that - would one bother with "forward" as well, even
though I doubt that is as often used (I tend to assume not)? Or should one
reuse "previous":

    Skip previous

which is explicit in meaning -1 (whereas to me "back" just indicates the
direction), but seems to be missing the word "to".

Hmm. I'll think on it - both have disadvantages for me, but I definitely
like "hiding" that -1 (although since this is moving the read-head it is
less important to me than losing what I think of as "line numbers" in the
tag table).

> Hmm, I guess Tim will like this ;-)


That would be nice - I bet he won't comment until some code is produced,
though!

Tibs
--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.demon.co.uk/
Feet first with 5 wheels...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)





From olipt at mayo.edu  Thu Jul 15 15:32:51 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Thu, 15 Jul 1999 14:32:51 -0500
Subject: Translating Matlab into Python
In-Reply-To: <ul7lo196wn.fsf@harder.princeton.idaccr.org>
References: <ul7lo196wn.fsf@harder.princeton.idaccr.org>
Message-ID: <Pine.LNX.4.10.9907151431260.22646-100000@us2.mayo.edu>

> Does anyone know if there's a program out there to translate (or help
> in translating) MATLAB code into Python (probably with NumPy extensions)?

I don't know of any.  The Numeric extensions package has a file called
MLab.py that implements some maltab calls but it is just the beginnings.

I would be very interested in something like this too, if there are any
language experts who wan't to take a stab at a very useful tool.  

Travis





From gmcm at hypernet.com  Fri Jul 16 10:21:03 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 16 Jul 1999 09:21:03 -0500
Subject: I envy newbies
In-Reply-To: <378F2304.ACC1AF7E@callware.com>
Message-ID: <1279991447-21085630@hypernet.com>

Ivan Van Laningham writes:
> 
> Tim "The Insufferably Right" Peters wrote:

Pah. He's not always right. He _is_ always... nevermind...

> Besides, I don't think IDLE will work if you don't have Tcl in your
> path, and if you have to add that, why not add Python?  

I think the Tcl install does modify your path. But Python will go 
looking for it in the default places. Which is, of course, C:\Program 
Files\... [I think it's hilarious that MS was so enamored of their 
"advance" in being able to put spaces in paths that they made the 
"proper" place to install everything a demo of this "feature"... 
Which demos the fact that spaces in path names are a pain in the 
ass]. And if tcl80.dll and tk80.dll are on your path (which they 
should be, since they get installed to your system directory), then 
what needs to be found in the lib\*.tcl stuff. You can point to those 
with TCL_LIBRARY and TK_LIBRARY and not mess with PATH at all.

> Well,
> obviously I should go forth and re-format my hard drive and
> re-install W98 and then re-install Python. ...  But I really don't
> feel like a little root canal today, 

I think you can get ahold of the install script.

> > can-just-feel-the-linux-folks-melting-with-jealousy<wink>-ly y'rs  - tim
> 
> "Yearning for NT when you're running Linux is like lusting for a Big
> Wheel when you're driving a BMW." --Harrison Picot.

More apt would be an cut-rate underpowered "luxury" car (you 
know, where you have to turn off the A/C to go up a hill) and a John 
Deere. Now Win9x, that's a tinker toy.

- Gordon



From kistler at fnmail.com  Sun Jul  4 06:34:21 1999
From: kistler at fnmail.com (Per Kistler)
Date: Sun, 04 Jul 1999 12:34:21 +0200
Subject: Tkinter._test() problem
Message-ID: <377F38AD.C6CDBA6D@fnmail.com>

Hi All

I installed py1.5.2 with tcl/tk.8.1 on SuSE6.1 Linux.

After:

import Tkinter
Tkinter._test()

I get the following error, which I do not unerstand:

-----cut
>>python
Python 1.5.2 (#4, Jul  4 1999, 12:25:56)  [GCC egcs-2.91.66 19990314
(egcs-1.1.2  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import Tkinter
>>> Tkinter._test()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/per/lib/python1.5/lib-tk/Tkinter.py", line 1947, in _test
    root = Tk()
  File "/per/lib/python1.5/lib-tk/Tkinter.py", line 886, in __init__
    self.tk = _tkinter.create(screenName, baseName, className)
TclError: Can't find a usable init.tcl in the following directories:

This probably means that Tcl wasn't installed properly.
----cut

It seems, that python looks in some directories, which it does not
have. What do I have to do?

Thanks, Per.
-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net
------------------------------------------------------------



From never at matavnet.hu  Sun Jul 11 14:26:05 1999
From: never at matavnet.hu (Zsolt Hever)
Date: Sun, 11 Jul 1999 20:26:05 +0200
Subject: database
Message-ID: <199907111829.OAA04318@python.org>

> From: dzsi <granadam at mail.datanet.hu>
> To: python-list at python.org
> Subject: database
> Date: 1999. j?lius 11. 0:11
> 
> hi,
> 
> i am new to python.......
> 
> i have a database (comma separated text), and i would like to get some
data
> out of it based on some query. let's say there are seven fields, f1...f7,
> and i am searching for entries in f1, and want to display f1..f7  for the
> field that matches the query. how do i go about doing that?

By the help with the following you can convert your text database
into an SQL table.
Use  Gadfly: SQL Relational Database written in Python, in this case there
is no need to configure a server.
Search for Database / Gadfly on the page <URL:
http://www.python.org/download/Contributed.html >


For converting your text database into SQL table:

You can work on it by module string / function split() /. 
More about it in the standard documentation.

But there are CVS modules. 
The addresses were mentioned on this mailing lists:
<URL: http://eh.org/~laurie/comp/python/csv/ >
<URL: http://yi.com/home/TrattLaurence/comp/python/csv/ >
ftp://ftp.visi.com/users/mdc/csvsplit.py


> also, can i access db databases, or any other common pc type databases
with

dbf
<URL: http://www.stud.ifi.uio.no/~larsga/download/python/ >
or check for it at the < http://www.python.org/ > website.

> perl scripts only on a unix system? without any server configuration,
> assuming python is installed. so far, the comma separated text file is
the
> only way i could figure out.

I hope this is what you are searching for.

Zsolt

> thanks,
> 
> adam.
> -- 
> http://www.python.org/mailman/listinfo/python-list







From pierrette at p98.f112.n480.z2.fidonet.org  Sat Jul  3 13:04:00 1999
From: pierrette at p98.f112.n480.z2.fidonet.org (pierrette)
Date: Sat, 03 Jul 1999 18:04:00 +0100
Subject: Upgrade or not ??
Message-ID: <000016a7@bossar.com.pl>

From: pierrette <pserrand at worldnet.fr>

First a general question :
Actually we're working with Python 1.5.1 and we're satisfied with it.
Should we upgrade to 1.5.2 and if yes WHY?

No a Tkinter guestion :
We developped a prototype of an application with python an tkinter, now
we have to
decide if we're gonna write the real app in python or something else !
We now allready that in principle we could do it with python, the thing
we're missing
in tk is a 'on get focus' ,a 'on lost focus' event and a 'set focus to'
command
Pe: on lost focus of inputfield 1:
            if inputfield1 = A:
                set focus to field2
            elif inputfield1 = B:
                set focus to field 3

If those things are not available in tkinter what alternatives do we
have ?
- Changing the GUI (would be a shame but feasable), knowing that is has
to be multi-platform WIN-LINUX? If so to what ?
- Not write in python ? In this case we lose the work done for the
prototype

Thanks for comments and remarks

Pierrette Serrand
pserrand at worldnet.fr





From jiwils at acxiom.com  Fri Jul 23 03:06:40 1999
From: jiwils at acxiom.com (James "Jimmy" Wilson)
Date: Fri, 23 Jul 1999 07:06:40 GMT
Subject: VC++ 6.0 exe file sizes
References: <o$h7sBAVZij3Ew7i@jessikat.demon.co.uk>
Message-ID: <37981400.1980005@news.cswnet.com>

On Thu, 15 Jul 1999 19:20:05 +0100, Robin Becker
<robin at jessikat.demon.co.uk> wrote:

>Any VC++ gurus out there help with a problem I'm getting building the
>python.exe file with VC++ 6.0 & win95 osr2. My hd has 4k clusters.
>
>No matter what optimisations etc I do the python.exe seems to come out
>as 16384. I reckon it should be around 5k like the 1.5.2 distributed
>version (which I think is built using VC++ 5.x).
>
>I see this particular problem, but what about all the other bits of exe
>code etc?
>-- 
>Robin Becker

You actually got it to compile?  Each time I tried, I got an error
message close to "MASSIVE COMPILER ERROR".  I tried updating to VS SP3
without any luck.  Did you happen to encounter this problem?
-- 
James "Jimmy" Wilson
Software Engineer
Acxiom Corporation
ADN Development
E-Mail: jiwils at acxiom.com



From m.faassen at vet.uu.nl  Wed Jul 28 08:02:33 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 28 Jul 1999 12:02:33 GMT
Subject: IDLE mini-hack & ? (adding menu items, understanding *args, self)
References: <199907272152.QAA01078@cro.ots.ac.cr>
Message-ID: <7nmrgp$bcu$1@newshost.accu.uu.nl>

Isidor <rodisi01 at my-deja.com> wrote:

[monstrous snip]
> One thing I don't really understand here is why it is necessary to 
> raise an exception when an argument is passed to a function that 
> isn't prepared to handle it. If the argument is needed *inside* the 
> function, I could understand an exception being raised then (e.g., 
> "hey, where's my argument"), but in my case i'm doing nothing (or at 
> least I *think* I'm doing nothing) with those arguments. Why do I 
> need to make a "landing strip" for them?

def myfunction(foo, bar):
     print foo
     print bar

myfunction(1, 2, 3)

You send '3' to myfunction. If myfunction now just ignored '3' silently,
you'd be endlessly confused why your program doesn't work, especially if
you didn't write myfunction yourself. But luckily, you get an exception
stating that you do something wrong. Python just tells you when you do
something that *can't* be right.

Unrelated to this, a common confusion with Python occurs with methods
(as opposed to functions):

class Myclass:
    def mymethod(self, foo, bar):
        print foo
        print bar

myobj = Myclass()
myobj.mymethod(1, 2, 3)

This is going to give an exception that too many arguments were passed.
This is because according to Python, *4* arguments were passed instead of
3.

If you call a method on an object, that object is always passed as the
first argument to that object, implicitly (usually it's called 'self'). 
Inside the method you can then use self to do various things to the
object (for instance, call other methods on it, or set attributes).

[snip]
> So, maybe by allowing the unused arguments 
> to pass through my function, everything flows nicely and nothing gets 
> stopped up. Hmm, strange. Am I on basically the right track here with 
> this idea? 
[big snip]

Actually, you can make functions that allow extra arguments -- but only when
you explicitly tell Python:

(file: test.py)
def foo(a, b, *rest):
    print a
    print b
    print rest

>>> from test import foo
>>> foo(1, 2)
1
2
()
>>> foo(1, 2, 3)
1
2
(3,)
>>> foo(1, 2, 3, 4, 5)
1
2
(3, 4, 5)
 
You can also do this (this only works with keyword arguments):

(file: test2.py)
def foo(a, b, **rest):
    print a
    print b
    print rest

>>> from test2 import foo
>>> foo(1, 2)
1
2
{}
>>> foo(1, 2, 3)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: too many arguments; expected 2, got 3
>>> foo(1, 2, something=3)
1
2
{'something': 3}
>>> foo(1, 2, something=3, other=4)
1
2
{'other': 4, 'something': 3}

So, as long as you tell Python what you want, Python allows lots of
flexibility. But if you don't tell Python that you want it, Python assumes
you don't want it -- it gives an error, which is good, as it probably
*was* an error.

Some languages take this further and also require you to add the *types* of
the arguments you espect in your function (they are strongly typed). So, if
you have a function argument 'foo', you need to specify in your function
definition if 'foo' is a string, or a number, or an object made from some
class, etc. This can also be used to prevent errors (and compile into
faster code); the language compiler can detect if you do something that's
not intended (pass a number where you specified only a string could go).

But dynamic typing like Python has other advantages -- you
have more flexibility and you don't get caught up messing with argument
types all the time (it can cost lots of time early in development, and it's
harder to change something).

The last discussion is just for random information and does not have much
to do with Python. :)

I hope my explanation was useful.

Regards,

Martijn




From holger at phoenix-edv.netzservice.de  Mon Jul  5 06:42:46 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Mon, 5 Jul 1999 10:42:46 GMT
Subject: file.tell problems (1.5.2/Win32)
References: <7krd1r$2oa$1@holly.csv.warwick.ac.uk> <7l3aao$65c$1@nnrp1.deja.com> <3776ED63.35D87F89@compaq.com> <37782C9E.3061AF25@geekspace.com> <37793097.17D1BDA6@compaq.com>
Message-ID: <37808C26.1693E239@phoenix-edv.netzservice.de>

Hi Greg,


Greg Ewing schrieb:
> 
> Joshua Rosen wrote:
> >
> > This leads me to believe that `append mode' is intended to `append' at the
> > *current* point in the file, which should, for consistency's sake, start at 0.

I think there is the answer: "Some Unix-Systems" properly would append the write-string
directly at the current file position. 

> I just tried an experiment (on Windows) using "a+"
> mode. It seems that f.tell() is telling the truth,
> but the truth is strange.
> 
> What "a"/"a+" modes seem to mean is "open the
> file positioned at 0, but if the first operation
> is a write, go to the end of file before doing

I think even this is `bizarre`!
Because you could 1st work with "read(1)" or readline()
on the file, getting correct fileposition (tell), but then a
write even wouldn't work like expected!;-( It doesn't work!!
Take the test:
Create a small file with e.g. 3 lines a 10 characters.
>>> f=open('test.txt','a+')
>>> f.tell()
0
>>> f.read(12)
'111111111111'
>>> f.write("sssss")
>>> f.tell()
17
>>> f.close()
You won't even see the obviosly written string inside the file!!
I think, here's something wrong!?

> it". Which seems quite bizarre to me - wouldn't
> it be simpler to just go to the end of the
> file on opening?

You've seen: I could even read on a 'a+'-mode-opened file
getting new file-position.
And there are 'some Unix-OSs' that do the IMHO right thing to 
append directly at the current file-position without jumping
to the end of file.

Ciao,
Holger
[Python1.52, WinNT4.0]



From cgw at fnal.gov  Fri Jul 23 17:31:25 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Fri, 23 Jul 1999 21:31:25 GMT
Subject: Python documentation available as ZIP files
In-Reply-To: <m2ogh39l9w.fsf@netmeg.net>
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
	<14232.48512.648609.496768@lrz.de>
	<14232.48816.993092.9754@weyr.cnri.reston.va.us>
	<m2ogh39l9w.fsf@netmeg.net>
Message-ID: <14232.57133.949353.741301@buffalo.fnal.gov>

Les Schaffer writes:

 > class IMHO  ( $0.02 ):
 >   
 >     """of course, one can ENCOURAGE "common enough" by supplying bzip2
 >     for something popular like documentation and watch your community
 >     GROW acceptance of the new compression standard"""
 > 
 >     smile = "  :-)  "
 > 
 >     def __init__(self, agree_wth_me):
 > 
 >        if agree_with_me == YEP:
 >           os.system('tar cvf dem_docs dem_docs.tar.bz2 --use_compress-program=bzip2')
 > 
 > 	print self.__doc__
 > 
 > 	print smile
 > 
 >         print   "Les Schaffer godzilla at netmeg.net"
 > 
 > 	raise " WARNING: it must be friday afternoon!!!  "
 > 

I agree with you.  But shouldn't that be "print self.smile" ?  I got a 
NameError when I tried to run your code  ;-)

But seriously, this is a mistake I've seen a lot of people make with
class variables,  just want to make sure we don't spread around any
misconceptions.




From greg.ewing at compaq.com  Wed Jul 28 18:33:08 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 29 Jul 1999 10:33:08 +1200
Subject: IDLE mini-hack & ? (adding menu items, understanding *args, self)
References: <199907272152.QAA01078@cro.ots.ac.cr>
Message-ID: <379F8524.40B89D81@compaq.com>

Isidor wrote:
> 
> One thing I don't really understand here is why it is necessary to
> raise an exception when an argument is passed to a function that
> isn't prepared to handle it.

It isn't strictly necessary, but it helps to catch
errors earlier. Python is so dynamic that there aren't
many opportunities to catch errors until the last
moment, but when there is one, Python takes it.
It's a language design decision, and in my opinion,
it's a good one. The vast majority of the time, passing
the wrong number of arguments to a function is an error.
For the rare cases where it isn't, Python lets you
explicitly say that it's okay.

> My guess is that insert is a method of text, and "insert" (first
> argument) is the location of the insertion point

That's right. To add text to the end of a text widget,
you can use "end" instead of "insert", e.g.

  self.text.insert("end", "something to append")

> a function knows nothing about its context that it isn't told

There are only three sets of names that can be
directly referred to in a function:

(1) Its parameters and local variables
(2) Names defined in the top level of the module where the
    function was defined
(3) Names of built-in Python functions

Note that this does *not* include the parameters and
locals of any *other* function, even if you nest one
function definition inside another. (This surprises people
who are used to languages like Pascal or Scheme.)

In the case of a method, it also doesn't include
any attributes of any class or instance - hence the
need for the "self" argument.

> 2       def funcx(self, *args, **namedargs):
> 3               zip = "googoo"
> 
> 4       def funcy(self, *args, **namedargs):
> 5               print zip
> 
> Something causes me to suspect that this won't work.

You're right, it won't work - whatever you mean
by "work" (it's not clear what you want it to do).

> Can I attach zip to "self" somehow in funcx, e.g., line 3:
> 'self.zip = "googoo"'?

Yes, you can certainly do that. It's the usual way
of assigning to an instance variable in Python.

> Will it only work if self.zip is defined in the
> __init__ func?

No, there's no need to do that. You can add a new
attribute to any instance of any class at any time
just by assigning to it. There is no notion of
"declaring" instance variables.

> "read chap x, sect y of tutorial or language reference").

You might like to read through the chapters of the
language reference dealing with classes and instances.
The principles are really very simple.
Once you have the right mental model of these, you
should be able to answer all those "what will happen
if I do this" questions for yourself.

Hope that helps,
Greg



From nick.bower at ssec.wisc.edu  Sat Jul 10 23:26:13 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sat, 10 Jul 1999 22:26:13 -0500
Subject: Simple Tk8 Problems
Message-ID: <37880ED5.2BD7C347@ssec.wisc.edu>

Hi,

I'm currently having 2 simple, but I think fundamental, problems with my
understanding of Tk 8.

1.  I can't control the size of the Tk root window.  I've tried putting
".width=xxx" statements everywhere in the following snippet, but nothing
seems to work.  (The FileList is a ListBox next to a scrollbar).  Is
there something I'm missing?.

class FileList(Frame):
  def __init__(self,parent):
    Frame.__init__(self,parent)
    self.filelist=Listbox(self)
    self.scroll=Scrollbar(self)
    self.filelist.config(yscrollcommand=self.scroll.set,relief=SUNKEN)
    self.filelist.pack(side=LEFT,expand=YES,fill=BOTH)
    self.scroll.config(command=self.filelist.yview,relief=SUNKEN)
    self.scroll.pack(side=RIGHT,expand=NO,fill=Y)

<etc>

if __name__ == '__main__':
  toplevel = Tk()
  afilelist=FileList(toplevel)
  afilelist.pack(expand=YES,fill=BOTH)
  afilelist.mainloop()


2. Also, the Listbox in the above code is scrolling all over the place
tracking the mouse movement as it runs across the top of it.  I don't
even need to touch the scrollbar and it seems to track mouse movement so
somehow it's been bound to the mouse as well as the scroll bar.  Weird.


Thanks in advance for help (again)  :)

nick


--

Nick Bower
Space Science and Engineering Center
University of Wisconsin, Madison, USA
Phone: (608) 265 8007
Email: nick.bower at ssec.wisc.edu
Web: http://arm1.ssec.wisc.edu/~nickb






From aahz at netcom.com  Sat Jul  3 13:38:24 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 3 Jul 1999 17:38:24 GMT
Subject: Upgrade or not ??
References: <377E0A3B.CA3B258C@worldnet.fr>
Message-ID: <7llhqg$b60@dfw-ixnews9.ix.netcom.com>

In article <377E0A3B.CA3B258C at worldnet.fr>,
pierrette  <pserrand at worldnet.fr> wrote:
>
>First a general question :
>Actually we're working with Python 1.5.1 and we're satisfied with it.
>Should we upgrade to 1.5.2 and if yes WHY?

I think so.  We're sure planning on upgrading to 1.5.2 as soon as we can
spare the cycles.  There are a bunch of bug fixes across everything
(e.g. one recent example we uncovered was whitespace handling in
sgmllib) plus tons of new features (e.g. time.strptime()).
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From tim_one at email.msn.com  Mon Jul 19 00:12:34 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 19 Jul 1999 00:12:34 -0400
Subject: Stackless & String-processing
In-Reply-To: <180719992337546352%pereira@research.att.com>
Message-ID: <000a01bed19c$edcea9c0$f0a02299@tim>

[Graham Matthews]
> You have totally misunderstood what I said.
> ...

[Fernando Pereira]
> One last try. I understood your point from the beginning. However, you
> did not understood *mine*.
> ...

Trust me on this one, fellows:  you each understand the other, and have for
several (interesting) msgs already.  Graham has an unusual way of expressing
his appreciation for a good discussion <wink>.

    http://www.deja.com/getdoc.xp?AN=316262255

leaping-thru-hoops-ly y'rs  - tim





From tismer at appliedbiometrics.com  Sun Jul 18 14:48:21 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Sun, 18 Jul 1999 18:48:21 GMT
Subject: How fast can we multiply?
References: <37920F41.185E012B@appliedbiometrics.com>
Message-ID: <37922175.FEEBE203@appliedbiometrics.com>

Sorry,
I detected an oversight, which made me loose a lot of speed.
Version 0.2 is now 4 times faster at 100000 bits.

"""
fastlong.py

Fast long integers. 

Christian Tismer
990718

Version 0.2: 
Adding some forgotten fastlong calls made it much faster.

>From time to time I have argued on the main list that Python's
long integers should have a faster multiplication. It is a well
known fact that by some factorization, integer multiplication's
complexity can be brought from O(n**2) down to O(n**1.5) .

The method is sometimes called Karatsuba algorithm. But studying
chapter 12.7 of (1), it turns out to be nothing else but a special
case of preconditioning polynomials.

I think to remember that the GMP library uses this algorithm,
but I didn't look it up and tried my own straightforward way.

The reason why I don't aim to modify the C implementation any
longer is: It will not be reasonably faster in C. The real
working horse is still the basic internal long multiplication
(sometimes called "highschool algorithm") which is fast for small
long integers. The only effect of a C port would be to move the
break_even point to something smaller.

My proposal to get something like this into the C code:
For very long numbers to multiply, the C version could
be enabled to call back into the Python implementation.

A function to obtain the bit length of a long would
be most helpful. Some kind of split function as well.
Currently, with a "clean" nbits implementation, the
break even point appears to be 50000 bits.
With a marshal-based inaccurate hack, break even is 
around 10000 bits.

For a bit length of 100000, my algorithm outperforms
the built in long numbers by a factor of 4.


(1) Aho/Hopcroft/Ullmann: 
    "The Design and Analysis of Computer Algorithms"

"""

class fastlong:
    """a class around long numbers which can multiply faster than
long"""
    break_even = 50000 # clean nbits
    break_even = 10000 # hacked nbits
    def __init__(self, val):
        if isinstance(val, fastlong):
            self.val = val.val
        else:
            self.val = long(val)
        
    def __repr__(self):
        return repr(self.val)
        
    def __add__(self, other):
        return fastlong(self.val + fastlong(other).val)
        
    def __sub__(self, other):
        return fastlong(self.val - fastlong(other).val)

    def __div__(self, other):
        # not optimized yet
        return fastlong(self.val / fastlong(other).val)
        
    def __pos__(self):
        return self
        
    def __neg__(self):
        return fastlong(-self.val)
        
    def __lshift__(self, bits):
        return fastlong(self.val << bits)
 
    def __rshift__(self, bits):
        return fastlong(self.val >> bits)
 
    def __mul__(self, other):
        """Karatsuba alogrithm
        (a+bX)(c+dX)=(X^2+X)(bd)+(X)(a-b)(d-c)+(X+1)(ac)
        """
        val1 = self.val
        val2 = fastlong(other).val
        if not val1 and val2:
            return 0
        bitcount = max(nbits(val1), nbits(val2))
        if bitcount < self.break_even:
            return val1 * val2
        shift = bitcount / 2
        (hi1, lo1) = split_num(val1, shift)
        (hi2, lo2) = split_num(val2, shift)
        sum1 = lo1-hi1
        sum2 = hi2-lo2
        mul1 = hi1 * hi2
        mul2 = lo1 * lo2
        mul3 = sum1 * sum2
        res = mul2 + ((mul1 + mul2 + mul3) << shift) + (mul1 <<
shift+shift)
        return res
        
def split_num(longval, shift):
    scale = 1L << (shift)
    mask = scale - 1
    lo = longval & mask
    hi = longval >> shift
    return fastlong(hi), fastlong(lo)
        
def nbits(x):
    """clean, accurate, not so fast"""
    if x <= 2:
        if x < 0:
            return nbits(-x)+1
        return x
    sum = 1
    while x:
        shift = 1
        x = x2 = x >> 1
        while x2:
            x = x2
            sum = sum + shift
            shift = shift + shift
            x2 = x >> shift
    return sum

import marshal, struct

def nbits(x):
    """not clean, rather fast, inaccurate"""
    s=marshal.dumps(x)
    return abs(struct.unpack("i", s[1:5])[0])*15 -7

#---------------------------------------------

def test():
        import sys
        be = fastlong.break_even
        for i in range(be/2, be*16, be/4):
            arg = (1l << i+1) -1
            fastarg = fastlong(arg)
            tim1, res1 = timing(lambda x:x*x, arg)
            tim2, res2 = timing(lambda x:x*x, fastarg)
            if res1 <> res2:
                raise ValueError, "different results fo i=" % i
            print i, tim1, tim2
            sys.stdout.flush()

def timing(func, args, n=1, **keywords) :
	import time
	time=time.time
	appl=apply
	if type(args) != type(()) : args=(args,)
	rep=range(n)
	before=time()
	for i in rep: res=appl(func, args, keywords)
	return round(time()-before,4), res
	
# the end ----------------------------------------------------

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From skip at mojam.com  Wed Jul 28 17:49:06 1999
From: skip at mojam.com (Skip Montanaro)
Date: Wed, 28 Jul 1999 16:49:06 -0500 (CDT)
Subject: Problem with signals & exceptions...
In-Reply-To: <DA4B3986E94F79C9.E3C10FB44F334D06.D3443001EBA1A2FA@lp.airnews.net>
References: <2BB05D8530326D0C.420888E8906528AF.C6BF325D12E9ED90@lp.airnews.net>
	<7nnbme$ns2$1@nntp6.u.washington.edu>
	<DA4B3986E94F79C9.E3C10FB44F334D06.D3443001EBA1A2FA@lp.airnews.net>
Message-ID: <14239.30920.777566.596051@dolphin.mojam.com>

    Eric> I changed my code to set a global variable rather than throw an
    Eric> exception and that works the way you'd expect. A socket.error
    Eric> exception is thrown, and I catch it, check the global variable, if
    Eric> the global variable is set I print "got a timeout!" else re-throw
    Eric> the exception.

I think this is fairly standard practice in Python.  You can avoid the
global variable using a class like:

    class Server(SocketServerTCPServer):
	def __init__(self):
	    self.serving = 1

	def exit(self):
	    do_cleanup_stuff()
	    self.serving = 0

	def server_forever(self):
	    while self.serving:
		self.handle_request()

    server = Server(server_address, request_handler_class)
    signal.signal(signal.SIGHUP, server.exit)
    signal.signal(signal.SIGQUIT, server.exit)
    signal.signal(signal.SIGTERM, server.exit)
    server.serve_forever()

Presumably the Server class is actually a bit more substantial than this,
but this serves to demonstrate hiding the global variable in a bit of state
inside a probably already existing server object.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758






From markusk at bidra241.bbn.hp.com  Thu Jul 22 07:32:55 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 22 Jul 1999 13:32:55 +0200
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have code.)
References: <000601bed335$893916e0$a72d2399@tim> <p5hfmymh9z.fsf@bidra241.bbn.hp.com> <Pine.LNX.4.10.9907211223270.11985-100000@atrus.jesus.cam.ac.uk>
Message-ID: <p5n1woncm0.fsf@bidra241.bbn.hp.com>

Michael Hudson <mwh21 at cam.ac.uk> writes:

> On 21 Jul 1999, Markus Kohler wrote:
> > "Tim Peters" <tim_one at email.msn.com> writes:
> > 
> > > [Tim]
> > > >     for i in xrange(10000000):
> > > >         pass
> > > > [and RC vs GC implications]
> > > 
> > > [Markus Kohler]
> > > > True.
> > > > IMHO this is one of Pythons major design flaws.
> > > 
> > > What, specifically?
> > 
> > That flaw is that loops are not implemented using closures. 
> 
> Sorry, how would that help?

It wouldn't help regarding performance, but increase the
expressiveness of the language. 

> 
> > Instead as far as I understand   loops in Python are implemented such that
> > the loop counter is created and destroyed for each step of the loop. 
> > 
> > You could compare this to tail recursive versus non-tail recursive functions. 
> > 
> > Also closures make the implementation of efficient loops more
> > difficult they have the advantage that loops and other control
> > structures can be better integrated into the OO concept. Smalltalk for
> > example doesn't need Iterators for that reason.
> 
> Python is not a functional programming language. 

It could be extended by some functional features.  
All I want is the integration of loops into the OO concept of the language. 

Loops are cannot be used as any other polymorph
function if you don't implement them with closures (please tell me
if there's another way). 

Smalltalk Example 

iterate over a container and do something for each element looks like this:

aContainer do:[:each | each doSomething]

because do: is just a method like any other method you can redefine it for each
class. 
This is very usefull because it allows to for example to loop over a Container
that contains two other Containers ("a" and "b" of an unknown type very easily:

do: aBlock
 a do: aBlock. 
 b do: aBlick. 


> As I understand it, you
> can't really have convenient closures because it is not lexically scoped
> (I can give you inconvenient closures...) and there were lengthy
> discussions about that some months back. I didn't follow it too closely
> because I didn't know what lexically scoped meant then, but the upshot was
> (I think) that lexical scoping isn't really the "Python way".
> 
> Python isn't really an OO language either, in the sense of something like
> Smalltalk or Eiffel.

IMHO it's very close to Smalltalk, also Guido probably wouldn't agree with me ...
And It has all essential OO features.  


The features that make it very similiar to Smalltalk are :
byte code interpreter, 
all data are objects (is that really true ?), 
crossplatform, 
reflection api, 
dynamically type checked, 

In fact I cannot remember any successfull language that is closer to 
Smalltalk than Python is. 


> 
> I have heard it said that Python is a language that "gets it's compromises
> exactly right" and that about sums it up for me.

I also like some things about Python. For example the module system, and how
easily external C functions can be integrated. 

> 
> > Lately ee have had already some examples in this group where loop
> > overhead was a major factor. 
> 
> Really? When? Sorry, I don't want to sound disbelieving, but I don't
> remember that.

It was about reading a file and put the words into some hash table. 
As far as I can remember to loop overhead was a dominating factor. 
And Python was much slower than perl for example. 

> 
> > I measured different kind of loops and
> > even the more general while loop was faster in squeak than anything
> > else in python.
> 
> Python, as I said, is neither functional nor purely OO; 

Ok it's not purely OO because you can define global functions. 
But it's anything else is pretty much OO. 

> it's not
> particularly fast either. Fast enough, generally.
> 
> > Functions calls seems also to be much slower in Python ...
> 
> *That* I believe. Not clear what to do about it though.

It's IMHO the overly complex function call mechanism, with it's 
default values on all that other stuff. I really can imagine it's difficult to 
make this fast. IMHO it's not worth the effort and maybe one should
invent Python lite that uses a simpler call mechanism :-O

> 
> > Another problem with loops could be reference counting.  I'm wondering
> > how often a reference count get's changed when running a loop ...
> 
> Not vastly more often than in regular code, I'd guess. It's true that
> executing
> 
> 1+1
> 
> spends longer twiddling refcounts than doing the work, but that's Python.
> 
> I can't decide which Peters-style ending I want here, so you're all
> getting two:
> 
> not-OO-functional-or-fast-but-still-great-ly y'rs
> it's-not-OO-or-functional-it's-Python-ly y'rs
> 


Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From mithy at my-deja.com  Wed Jul  7 02:17:54 1999
From: mithy at my-deja.com (Preston Landers)
Date: Wed, 07 Jul 1999 06:17:54 GMT
Subject: total idiot question: +=, .=, etc...
References: <000015b4@bossar.com.pl> <76u2rha5qr.fsf@four.npl.uiuc.edu>
Message-ID: <7lurec$41m$1@nnrp1.deja.com>

In article <76u2rha5qr.fsf at four.npl.uiuc.edu>,
  miller5 at uiuc.edu wrote:

> Does anyone have a tweak for python-mode.el that will highlight
> occurrences of self and self.?

Hey Mike,

Yeah, I posted my 3 line patch to python-mode.el in a separate thread
just a minute ago.

If you can't find it or whatever, check here:

http://askpreston.com/projects/python/

I'd be cool to implement this as a mode hook but I haven't found a
clean, simple way to do this yet.  So, patch it is.

--
||  Preston Landers <mithy at my-deja.com>  ||
||        http://askpreston.com          ||


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From greg.ewing at compaq.com  Thu Jul 22 16:53:37 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 23 Jul 1999 08:53:37 +1200
Subject: Spreading the word about Python; ?s and ideas
References: <199907211307.IAA14980@cro.ots.ac.cr>
Message-ID: <379784D1.13A69983@compaq.com>

Isidor wrote:
> 
>   The whole time I was asking myself...you
>   mean I have to *load* something just to change the directory?!

Maybe it would help if certain well-known built-in
modules (sys, os, string...) were pre-imported
into the builtin namespace. You'd still have to
write os.chdir, etc, but at least you wouldn't
have to write import statements for them all the
time.

> Download 15megs to run a 2k script?! 

You have to keep in mind that the 2k script is
so small because it's making use of a huge library
of modules written in Python and C.

Any given script is only going to use a small part
of the library, so you might be able to get the user
just to download the Python core and include the
required standard modules with the script. But
then if they download several scripts there is
likely to be some duplication.

It should be possible to produce an ultra-minimal
Python runtime that's quite small. The executable
and its DLLs only come to about 170K on Windows.
If you leave out the documentation, the rest is
mostly library .py files, most of which are relatively
rarely used.

A minimal runtime could perhaps include only the
most commonly used library modules, the rest being
stubs which display a message telling you what you
need to download.

Greg



From gerrit.holl at pobox.com  Thu Jul  8 16:38:57 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 8 Jul 1999 20:38:57 GMT
Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
In-Reply-To: <tot2m7.su.ln@final.math.uni-bonn.de>; from Thomas Ackermann on Thu, Jul 08, 1999 at 09:22:05PM +0200
References: <tot2m7.su.ln@final.math.uni-bonn.de>
Message-ID: <19990708223856.A29441@optiplex.palga.uucp>

On Thu, Jul 08, 1999 at 09:22:05PM +0200, Thomas Ackermann wrote:
> From: Thomas Ackermann <tgm at math.uni-bonn.de>
> Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
> Date: 	Thu, 8 Jul 1999 21:22:05 +0200
> To: python-list at python.org
> 
> Hi!
> 
> I have not much time to experiment with those modules and would like
> to hear opinions about to use what module     ;-)
> 

Well,
I used HTMLgen and I don't really understand why to use it. If you don't
know HTML, you can't use it: you need to understand tags like <B>, <I>, <U>
and such. I strongly recommend to have a look at HTML: if you look at it only
one hour, you'll already understand the basics. Another hour and you understand
frames and tables, and another hour to understand how it interacts with CGI.
So you'll only need 3 hours, if not less, to learn web programming. When you're
able to do that, you'll understand much more about HTMLgen (but won't need it,
I think...)

> Also, i am not exactly sure what Zope is - it seems to be something to
> create WWW-Pages and allow access to databases and it has it's own
> WebServer (a variant of Medusa) ...
> 

Me neither.

> So, what to use to allow for an Internet-Shop with pages created from
> databases and allowing customers to purchase over net?
> 

If you're doing such serious things, you should 

                                     **really**

learn web programming!!

> Can do Zope such things? Is it overkill at first? What can it do
> beyond? (And yes, i read some documentation, but i do not even
> understand what an "Application Server" is    ;-)
> 

Zope seems to be quite powerful.

>                GNU LINUX Python gtk pygtk MySQL FUDGE GURPS

GNU LINUX Python Civ III discoverb X much more.

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From nascheme at ucalgary.ca  Thu Jul 15 01:33:04 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 15 Jul 1999 05:33:04 GMT
Subject: Getters and Setters
Message-ID: <7mjrqg$e4o@ds2.acs.ucalgary.ca>

Recently I found myself wishing that Python had automatic getters
and setters like CLOS. For those who are not familiar, CLOS (and
Dylan) create a "getter" and "setter" method for each attribute
of a class. This saves the programmer from writing "getX" and
"setX" for every attribute when no special action is needed.
Later, an explict method can be created if needed.

An implementation of this idea is attached.  With this
implementation, these automatic methods seem to be about 8 times
slower than explicit methods.  Maybe Tim and friends can do
something about that.  I would appreciate any comments.

On a related note, has anyone considered adding hygenic macros to
Python?  Due to the high cost of function calls in Python they
would probably be a useful (and for other things too).  I have
the feeling that hygenic macros are much easier to do in a
language like Lisp.  Maybe Dylan can give us some ideas.


    Neil


======================================================================
#
# A mixin class to automagicly create getter/setter methods for
# class attributes. By default the method names are get* and
# set*. Modify as desired. Using the automatic methods seems to
# be about 8 times slower than explicit methods.
#
# Neil Schemenauer, July 1999
#

class _Getter:
    def __init__(self, value=None):
        self.value = value

    def __call__(self):
        return self.value

class _Setter:
    def __init__(self, klass=None, name=None):
        self.klass = klass
        self.name = name

    def __call__(self, value):
        self.klass.__dict__[self.name] = value
        del self.klass

class GetterSetter:
    def __getattr__(self, name, getter=_Getter(), setter=_Setter()):
        try:
            if name[:3] == 'get':
                getter.value = self.__dict__[name[3:]]
                return getter
            elif name[:3] == 'set':
                setter.klass = self
                setter.name = name[3:]
                return setter
        except KeyError:
            pass
        raise AttributeError


def test():
    import time
    class A(GetterSetter):
        def __init__(self):
            self.X = self.Y = 1
        def getY(self):
            return self.Y
        def setY(self, v):
            self.Y = v
    a = A()
    t = time.time()
    n = 10000
    print 'Comparing', n, 'iterations'
    for x in range(n):
        a.getX()
        a.setX(1)
    print 'GetterSetter', time.time()-t
    t = time.time()
    for x in range(n):
        a.getY()
        a.setY(1)
    print 'explicit function', time.time()-t


if __name__ == '__main__': test()



From jblake at eskimo.com  Thu Jul  8 17:54:22 1999
From: jblake at eskimo.com (Jonathon)
Date: 8 Jul 1999 21:54:22 GMT
Subject: Error Trapping In Python
References: <7m0pgt$cij$1@eskinews.eskimo.com> <199907080535.HAA19715@axil.hvision.nl>
Message-ID: <7m36me$bpa$1@eskinews.eskimo.com>

Hans Nowak (ivnowa at hvision.nl) wrote:

: On 7 Jul 99, Jonathon wrote:
: >  int ( read_data ) gives a Value Error, if non numerical
: You can trap such errors with exceptions... A simple example:
	
	Now I feel like an idiot.   Especially since I had
	been looking at that page in the pocket reference. 

: There are others ways, of course. For instance, you could scan that 
: string and filter out any unwanted characters (other than 0-9, 
	
	I'll do that for the field that has financial data in it.

	xan

	jonathon
	
-- 
	I'm still looking for a good book on
		3:	The Recent Unpleasantness
		1:	The War Of Northern Aggression.
		2:	The War of Southern Rebellion.



From sreekantk at freenet.co.uk  Fri Jul  9 06:35:41 1999
From: sreekantk at freenet.co.uk (Sreekant Kodela)
Date: Fri, 09 Jul 1999 11:35:41 +0100
Subject: Help with strings please
References: <Pine.LNX.4.04.9907081942540.13521-100000@emerald.netskate.ru>
Message-ID: <3785D07D.D33700D8@freenet.co.uk>

Many thanks Dear Oleg

Regards
sreekant

Oleg Broytmann wrote:

> On Fri, 9 Jul 1999, Sreekant Kodela wrote:
> > I seem to be in trouble here. I am trying to make some cgi scripts for
> > my web site. I am using the postgresql as the backend db and a form
> > based address book and news forum etc. The problem is everything is
> > working ok as long as noone uses ' and " in the text. If they do, bingo
> > , postgres refuses to accept. I tried using string library [mainly
> > string.join(string.split('awkward'"'st'ring')). It works sometimes and
> > doesn't some times.
> >
> > Any better and efficient way to deal with this situation !
> > Any suggestions and ideas will be greatly appreciated.
>
> s = string.replace(s, "'", "''")
> s = string.replace(s, '"', '\\"')
>
> > Thanks
> > sreekant
>
> Oleg.
> ----
>     Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
>            Programmers don't die, they just GOSUB without RETURN.




From fredrik at pythonware.com  Wed Jul 28 18:17:58 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Jul 1999 22:17:58 GMT
Subject: String2List: odd function
References: <199907272152.QAA01065@cro.ots.ac.cr> <379F731B.C2C39989@compaq.com>
Message-ID: <001701bed947$0d4fedc0$f29b12c2@secret.pythonware.com>

Greg Ewing <greg.ewing at compaq.com> wrote:
> > However, I suspect that this use is not among the common uses for
> > which exec and eval were designed.
> 
> Using eval for something like this is probably overkill.
> Eval and exec will accept *any* valid Python expression
> or statement, respectively - which can do some nasty
> things to the internals of your program if you're not
> careful what you feed into them!

well, last time I checked, eval didn't accept statements.
and you can pass in your own dictionary, to prevent the
expression from being evaluated in the current name-
space.

but that's not enough; consider this:

    >>> d = {}
    >>> eval("__import__('os').remove('precious-file')", d)
   OSError: No such file or directory: 'precious-file'
    >>> print d
    {'__builtins__': {...lots of stuff...}}

here, the __import__ builtin is another way of spelling
"import" (that function is used by the import state-
ment to do the actual work...).  and if you can access
arbitrary modules via eval, there's no limit to what you
can do...

but there's a way around this:

if you look in the dictionary, you'll find that python
automatically inserts an entry called "__builtins__"
which contains the current set of builtin functions
etc.

now, what happens if you add such an entry your-
self, before calling eval.  let's try this:

    >>> d = {"__builtins__": {}}
    >>> eval("__import__('os').remove('precious-file')", d)
    NameError: __import__
    >>> print d
    {"__builtins__": {}}

much better!

to sum this up: if you need to parse arbitrary python
expressions, and you're not 100% sure you can trust
the source, you can use:

    v = eval(s, {"__builtins__": {}})

(which, of course, was what I suggested in my original
reply to isidor...)

</F>




From MHammond at skippinet.com.au  Tue Jul 13 19:11:27 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Wed, 14 Jul 1999 09:11:27 +1000
Subject: newbie questions about sys.path, wxpython, IDLE and pythonwin
References: <378B7A69.12874AAC@gnet.tn>
Message-ID: <7mggvo$aug$1@m2.c2.telstra-mm.net.au>

MB wrote in message <378B7A69.12874AAC at gnet.tn>...
>I need to know exactly how the sys.path is built. In my system I have no
>pythonpath variable but the sys.path returns a list. I suppose it was
>built by default. This list doesn't contain all my python
>subdirectories. So I want to know when and how this list was built ?
>(yes I examined the registry and found references to python directories
>but they don't match my sys.path)

Python uses a combination of the values you found in the registry, and some
default values - things like ".\lib" are added always, regardless of the
registry.  Remember, you have the same source to see how it works that I do.

>Another question in relation with the first. I downloaded wxPython an
>tried to run the demo.py from the pythonwin GUI (opening
>c:\wxpython\demo\demo.py and running it). It runs perfectly.

>From what I hear, you dont want to try running it a second time or it
crashes Pythonwin :-)

>Other thing. I examined demo.py through the debugger under pythonwin I
>found this statement:
>'import miscc'
>The interpreter import this module but is not able to locate it. When I
>type >>> miscc it says that is a built-in module. Are there modules that
>are not python files ? What is a built-in module ? I didn't find its
>definition in the docs.

I have no idea - may be wxPython thing.  Certainly not built-in to Python
1.5.2.

>Last question. About the  syntaxic ':' at the end of a line when it
>begins with def, class,if ,when,etc... Is it absolutely necessary for
>the language ?
Yes.

Mark.






From bhoel at server.python.net  Sun Jul 11 11:53:45 1999
From: bhoel at server.python.net (Berthold =?iso-8859-1?Q?H=F6llmann?=)
Date: Sun, 11 Jul 1999 17:53:45 +0200
Subject: Extending ConfigParser
Message-ID: <3788BE09.26379078@starship.python.net>

Hello,

For my application I need a config file parser. ConfigParser comes close
to my needs, so I made my own Parser class derived from the
ConfigParser:

class MyParser(ConfigParser.ConfigParser):

First I wanted to change the parsing routine to allow white space in
section and option names, so I added the lines:
    
    __SECTCRE = re.compile(r'\[(?P<header>[-\w ]+)\]')
    __OPTCRE = re.compile(r'(?P<option>[-.\w ]+)(?P<value>.*)$')

but section names containing blanks still lead to an error. I found the
problem when I tried to add my own "read" method. Looking to the
original class I found, that the main work is done in a method "__read",
so I added 

    def __read(self, fp):
        print "TEST"
        return

to my class but still got:

>python cfg_tst.py
Traceback (innermost last):
  File "cfg_tst.py", line 5, in ?
    C.read("dat/Lines_my.cfg")
  File "/usr/local/lib/python1.5/ConfigParser.py", line 175, in read
    self.__read(fp)
  File "/usr/local/lib/python1.5/ConfigParser.py", line 326, in __read
    raise e
ConfigParser.ParsingError: File contains parsing errors:
dat/Lines_my.cfg
        [line 47]: '[station -3]\012'

So it seems, that class members whos names begin with "__" can't be
overloaded? Is this right, or have I overseen something?

Greetings

Berthold
-- 
bhoel at starship.python.net / http://starship.python.net/crew/bhoel/
        It is unlawful to use this email address for unsolicited ads
        (USC Title 47 Sec.227). I will assess a US$500 charge for
        reviewing and deleting each unsolicited ad.



From arcege at shore.net  Wed Jul  7 16:24:26 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Wed, 07 Jul 1999 20:24:26 GMT
Subject: Getting contents of Text / Entry widgets
References: <7ltmfu$nbt$1@nnrp1.deja.com> <c75pv252spo.fsf@pc142.cosc.canterbury.ac.nz> <7lvt4k$g8a$1@nnrp1.deja.com>
Message-ID: <_HOg3.1147$6M6.363209@news.shore.net>

catlee at my-deja.com wrote:
: In article <c75pv252spo.fsf at pc142.cosc.canterbury.ac.nz>,
:   Timothy R Evans <tre17 at pc142.cosc.canterbury.ac.nz> wrote:
:> The class below does this, see the example at the bottom of the script
:> for how to use it.  This example just prints the contents of the entry
:> widget.
:>
:> ==============================================================
:> #!/usr/bin/python
:>
:> from Tkinter import *
:>
:> class CallbackEntry(Entry):
:>     '''Replacement of Entry widget.  Set the "callback" attribute to
:>     your function, which will be called whenever the entry widget
:>     changes and will be passed the new contents of the widget'''
:>
:>     def __init__(self, *args, **kw):
:>         apply(Entry.__init__, (self,)+args, kw)
:>         self.stringvar = StringVar(self)
:>         self.stringvar.trace_variable('w', self.changed)
:>         self.configure(textvariable=self.stringvar)
:>         self.callback = None
:>
:>     def changed(self, *args):
:>         'Note that the args given to this method are useless'
:>         if self.callback != None:
:>             self.callback(self.stringvar.get())
:>

: Great!  This works perfectly!  Thanks for your help!  I've come up with
: a replacement for the Text widget.  The same method could be used with
: almost any widget, I imagine.  This doesn't pass the text back to the
: callback function since I only want to know when the text changes, not
: necessarily what it is.

: from Tkinter import *

: class CallbackText(Text):
:    '''Replacement of Text widget.  Set the "callback" attribute to
:    your function, which will be called whenever the text widget
:    changes.'''
:    def __init__(self,*args,**kw):
:       apply(Tkinter.Text.__init__,(self,)+args,kw)
:       self.bind("<Key>",self.active)
:       self.callback=None

:    def active(self,event):
:       self.after_idle(self.idle)

:    def idle(self):
:       if self.callback != None:
:          self.callback()


If you really wanted to "replace" your use of the Text widget, then
you could do the following:

  ### CallbackText.py

  import Tkinter

  class Text(Tkinter.Text):
    # the rest of the class from above

Then in your code, you only have to change:
  from Tkinter import *
  from CallbackText import Text # 'Text' binds to newer widget

None of the code that uses a Text widget would have to change.

  -Arcege




From rodisi01 at my-deja.com  Wed Jul 21 10:03:59 1999
From: rodisi01 at my-deja.com (Isidor)
Date: Wed, 21 Jul 1999 08:03:59 -0600
Subject: Spreading the word about Python; ?s and ideas
Message-ID: <199907211307.IAA14980@cro.ots.ac.cr>

Hello python-list -

I am a newbie, and I am learning python (1.5.2) with what is
available over the web ("look list, no books!")...and I am having
a good time doing it. Doing this is possible because (a) python
is a nice language, (b) there is a lot of good documentation on
the web, and (c) my background. I don't have a formal programming
background. In fact, in any serious meaning of the word, I have
*no* programming background. The background I *do* have is from
working with the various MS Office scripting environments:
WordBasic, VB-Applications for Excel and Access (all old
versions...I'm running office95). Why do I mention this? Well,
because I think it may call attention to a potentially large
group of people who would be excited by and could easily learn
python.  MS-office and its components are on a lot of people's
machines (yes/no?), and therefore, I suspect that the office
scripting languages are probably the most commonly used
"programming" tools of "non-programmers". If there really does
exist such a "class" of newbies, it may be worthwhile to design
training materials for them, worthwhile in the sense that you
could produce a lot of programmers for relatively little effort.
I am thinking here of the goals stated in GvR's essay/proposal:
"Computer Programming for Everybody" (
http://www.python.org/doc/essays/everybody.html ).
How many people out there (non-cs-ers) have gotten into
programming because they discovered they could write macros in
Word or apps like Word? MS makes it fairly easy (or at least I
thought so) by providing the "record macro" facility. How many
office users have found themselves doing the same find and
replace over and over? How many of them say, "hmm, maybe I'll
record a macro for doing this"? Of course, as you would guess,
any "recorded" macro is a piece of smelly cow dung. This is a
good (and unavoidable?) thing, I think: because recorded macros
are so bad, users are forced to edit them. It was that first
macro editing experience that turned on the "empowerment" light
for me. "You mean....I can do all this stuff! You mean...that's
all that word is doing...is running these little commands that I
can run too?!" (Laugh if you must ;-) All the bad, obfuscatory
magic evaporated and was replaced by the magic of engineering and
craftwork: "how do I find the command to do x? What is *really*
going on when y happens?" That's the point at which I went off to
the installed help files. (I'm always amazed how many people
don't install those help files.) It's not the easiest thing to
do, but these languages can be learned from the help files.
However, not much time passes before any macro writer realizes,
"Hey, Word is cheating! I can't do all of the things it can! I
can't gain access to all of the things it can do and that I want
to do!" This is an incredibly frustrating realization. I hate it
when the help file leads me to "for more information, buy the
word-basic programmers toolkit available from your ms authorized
blah blah"...and then I realize that I have to buy hundreds of
dollars of stuff to do what should be fairly easy to do.
Workarounds suffice for a while, but they run out soon too...And
that's where Python came in for me. And maybe that's where it can
come in for a *lot* of other people. Maybe?

To reiterate (more concisely):

There may be a lot of office scripters out there, and they are
probably more than ready to hear the word of python. Targeting
them through marketing and educational materials might be an easy
way to get a lot of people quickly ramped up to more serious
programming.

(On the other hand: most of the office users I know still have
half their screens filled with inane toolbars. I don't know how
many times I've been asked "Is there any way to stop Word from
capitalizing x?" I'm stunned that they didn't hunt down and
destroy that default setting in the first five minutes of using
Word. So maybe there aren't a lot of macro writers out there.
While they're all getting their jobs done, I'm tweaking and
configuring to the nth degree. Don't know. However, the huge
selection of "power user" and "tweaker" tools on the shareware
sites indicates that there are a lot of people champing at the
bit with their machines. An aside: Is this a Windoze only
phenomenon? Are macs so well-designed that they don't cause this
sort of angst? I am guessing that all folks using unix already
know that they can make things be the way they want them...a unix
user *is* a power user, by definition(?).)

There are two sub-issues I would like to bring up under the
heading of "python for office scripters":

1.     Things that office scripters will probably need help with
2.     Distributing python scripts


1. Things that office scripters will probably need help with:

a)     object orientation

  There is a lot of object oriented-looking stuff in office, but
  compared to python, it looks like tiddly-winks. (I think...I'm
  just starting the oo thing myself.) I.e., in office,
  documents, cell ranges, etc. are objects and can be
  manipulated as such. This is nice. Can you create your own
  objects? Yeah, I think so. I never got that far. Does it make
  sense to do so in office? I don't know.
  
  So, I often find myself asking: "what is a class *really*?" I
  am able to answer the question sometimes, but I find the
  answer lacking somehow. An in depth discussion (with pithy
  examples) of what classes are and why we should care would be
  really helpful. (There are probably lots of pages out on the
  web that do this already. Putting together a list of pages and
  examples would be useful.)

b)     how do I do x in Python? In WordBasic/Excel/Access, it
  was blah blah...

  It strikes me that there is a key set of tasks that office
  scripters quickly learn how to do. Many of these key tasks are
  not available in python's built-in functions; they need to be
  imported from separate modules. This seemed like a big deal to
  me at first. The importance (and normal-ness ;)  ) of
  importing needs to be explained early on. An "office-
  scripter's index" to the module documentation would be
  invaluable. "Change directory? See os.chdir()") (See the
  recent debate in this list about where to put os.chdir() in
  the documentation. The whole time I was asking myself...you
  mean I have to *load* something just to change the directory?!
  whoa. ;) *Whatever* the correct place for os.chdir() is, it is
  going to be buried too deep...I suspect that "correctness" is
  audience specific.) Some nice "office-scripter-targetted"
  faq's could help a lot along these lines. (I'm already
  starting my own collection, of course. When they're worth
  anything, I'll be happy to share them.)

c)     lots of other really obvious things that I can't think of
  right now 'cause I'm too tired.

2.     Distributing Python scripts

A recent question in the tutor-list was "can I make .exe files
from python?" The answer surprised me (it shouldn't have): "you
can, but it's a lot of work."

(Disclaimer: I am really ignorant on this topic, and I will try
to take that into account in my comments. Please feel free to
suggest appropriate instructional reading.)

As a Word-scripter, it is very easy for me to make my macros
available to other word users. They already have the runtime (is
that the right word?) installed.

If I want to distribute my python script, I have to tell people
that they need to download python first. Maybe this is no big
deal for most people. I am working on a 486 w/ 500meg hard drive.
I was maxed out before python and I needed to remove a bunch of
stuff to fit it in. Not including the documentation, the win32
binary was ~5mb for python and ~5mb for tcl. Add in the 5mb of
docs (wow, that's a lot of docs...what an incredible gift!)...and
you've got ~15megs. That seems like a lot (to me). (Yeah, I know,
I need to get a several gigabyte harddrive and then I'll shut
up...) Before someone gets the wrong idea here, I need to stress:
I am *not* complaining about the size of the python distribution.
I *am* suggesting that its size may reduce the likelihood of
someone downloading it *solely* for the purpose of running a
little script that parses their web-page bibliographic search
results into something that endnote can read. Download 15megs to
run a 2k script?! So, my question to you wise ones: is there some
way to make a nice tiny package to distribute with a script?
<blinking marquee ignorance tag> What I'm imagining is that when
someone stumbles across your handy script on cnet.com downloads,
there will be a little link at the end of the script description
saying: "this script requires the python runtimes ver.x (3 megs);
download them here." It would probably be mostly impossible for
new users to write functioning scripts of their own just with
these runtimes, but they might be able to wangle off some. More
importantly, new users will probably look at the source and
realize "wow, it's really easy to write a program to do x!" and
kapow, you have a new python convert. Every script could carry a
few lines saying something like: "If you want to write your own
scripts, just download www.python.etc... and get to it." Or if
you try to do something impossible with the runtime, a little
message box will pop up saying..."to do that, you should download
the full python at www.etc...." <more reeking ignorance> I have
no idea whether any of this makes sense / is possible. I have
never compiled anything in my life. I have never even seen a
compiler...so this could all just be a bunch of wishful thinking.

Ok, that's the end of my rant. I hope this is a useful catalyst
for discussion.

Take care

Isidor

(note: I am not going to be so impudent as to *close* my
<ignorance> tags! ;)




From arcege at shore.net  Wed Jul  7 12:07:04 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Wed, 07 Jul 1999 16:07:04 GMT
Subject: returning Python Lists from C++
References: <tKGg3.149$FH1.16131@newsr2.u-net.net>
Message-ID: <IWKg3.1099$6M6.349834@news.shore.net>

Mark Butterworth <mark.butterworth at cis.co.uk> wrote:
: Can anyone give me a quick example of how to return a python list object
: from a C++ function?

There are two ways,  Py_BuildValue and building your own PyListObject
yourself.

Using Py_BuildValue:
  return Py_BuildValue("[si]", "Spam, eggs and spam", 42);

The square brankets tell the function to build a python list and return
it with the contents; in this case a two element list with a Python string
and a Python integer.

Read about Py_BuildValue in section 1.9 of the "Extending and Embedding
the Python Interpreter" manual (http://www.python.org/doc/current/ext/
buildValue.html).

Making PyListObject objects:
  PyObject *list;

  list = PyList_New(2);
  PyList_SetItem(list, 0, PyString_FromString("Spam, eggs and spam"));
  PyList_SetItem(list, 1, PyInt_FromLong(42));
  return list;

Notice that I did not change the reference counts.  We are returning a
new list object, so we do not increment the reference count.  Also,
the SetItem function does not increment the reference count of the
elements, which are new objects anyway.

You can read about these in section 7.2.3 "List Objects" of the
"Python/C API Reference Manual" (http://www.python.org/doc/current/api/
listObjects.html).

  -Arcege




From claird at starbase.neosoft.com  Sun Jul 18 13:17:11 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 18 Jul 1999 17:17:11 GMT
Subject: Harlequin Is Back!
References: <%vti3.121$95.4053@newse2.tampabay.rr.com> <378E57D8.A77A9F9C@nmia.o_r_g> <302747D6133FC474.B42081F80C6F00FB.672284C91DB0497B@lp.airnews.net> <3791b247.269188472@judy.x-ray.local>
Message-ID: <883C1D0C4B5A6107.A5862CDFC4AF10C9.11150CC24C1CF324@lp.airnews.net>

In article <3791b247.269188472 at judy.x-ray.local>,
Reini Urban <rurban at sbox.tu-graz.ac.at> wrote:
>claird at Starbase.NeoSoft.COM (Cameron Laird) wrote:
>>Coincidentally, I report today on several
>>related projects of cross-language co-operation in <URL:
>>http://www.sunworld.com/swol-07-1999/swol-07-regex.html#2>.
>
>this article unfortunately has several broken links:
>
>http://starbase.neosoft.com/~claird/comp.object.corba/COM.html
>-> http://starbase.neosoft.com/~claird/comp.object.corba/CORBA.html
>
>http://starbase.neosoft.com/~claird/comp.protocols.misc/COM.html 
>-> ??
>--                                         
>Reini

My apologies for these errors, and thanks to you for spotting them.
The COM.html URL was indeed empty; I've begun to populate it.  The
CORBA.html page has been available all along, but at <URL:http://
starbase.neosoft.com/~claird/comp.object.corba/CORBA.html>, as you
clevely discovered.  I expect to correct the reference in the
*SunWorld* article tomorrow.
  
Incidentally, I have ancestors surnamed Urban who apparently were
in Alsace in the early 19th century.  That seems quite far from
Graz.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From Lyle.Johnson at p98.f112.n480.z2.fidonet.org  Fri Jul  2 03:57:22 1999
From: Lyle.Johnson at p98.f112.n480.z2.fidonet.org (Lyle Johnson)
Date: Fri, 02 Jul 1999 08:57:22 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001687@bossar.com.pl>

From: Lyle Johnson <jlj at cfdrc.com>

FOX's grid widget (called FXTable) is under development and will probably be
usable soon (although I can't comment on how its features will compare with the
same widget in wxWindows). As Mike pointed out, one of the test programs for
the
C++ library demonstrates what's there so far. I didn't really intend to work up
an FXTable demo for FXPy until the FXTable was really usable.

I think any kind of rich text (i.e. RTF-compliant) or HTML widget is much
farther off for FOX.

"Mike C. Fletcher" wrote:

> None of the demonstrations I've seen are using a "grid" method.  There is
> one which looks similar, the "header" demo, but that is a set of "list"
> widgets standing next to each other and a sitting below a header widget.  To
> create a real grid, you would need to do some (likely C++) programming.  You
> would probably also want to make such thing allow for editing in-field, and
> have the capability of updating directly from an application table to the
> on-screen representation.  It should be possible to create such a thing in
> Python by dropping data into a matrix of entry fields in response to
> scrolling/moving etc., but a grid object generally wants a high-performance
> solution with all sorts of funky options.
>
> A "rich" text editor is evident in the Fox IDE, FXEditor, a separate project
> from the Core Fox libraries, though I have not yet downloaded that project
> to test it.  Something similar should likely become part of a base Fox
> distribution if it is to become a widely deployed Python GUI.
>
> Fox has a basic GUI editor under development (it is a separate project
> again), though my tests with it so far have been unsuccessful, as the
> application exits silently about every-other time I click on a widget.
>
> Neither tkinter nor Fox has any perceptible "slowness" on my system, so I
> cannot comment in that direction.
>
> I'm continuing my experiments, will continue to report as time permits.
>
> Enjoy,
> Mike
>
> Alexander V. Voinov <avv at quasar.ipa.nw.ru> wrote in message
> news:377BF583.6846FBF9 at quasar.ipa.nw.ru...
> ...
> > Did you look at the demos for the underlying toolkit? Some of them
> presents a
> > 'table' which is probably less rich than that of wxWindows, but it does
> exist.
> >
> > What it really lacks is a 'richtext'/html widget, sort of which is present
> and
> > actively developed in wxWindows. As I understand it's just about to become
> > available in wxPython.
> ...



From mstenber at cc.Helsinki.FI  Mon Jul 26 07:42:31 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 26 Jul 1999 14:42:31 +0300
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com>
Message-ID: <al86737ocwo.fsf@sirppi.helsinki.fi>

"M.-A. Lemburg" <mal at lemburg.com> writes:
> Markus Stenberg wrote:
> > ...
> > Thus I've had great deal of interest in getting the speed up; however,
> > speeding up dynamic interpreted languages is nontrivial, especially
> > languages as dynamic as Python.
> There are various ways to enhance Python's performance. Here
> are some links:
> 
> 	http://www.automatrix.com/~skip/python/
> 	http://lima.mudlib.org/~rassilon/p2c/
> 
> The most trivial trick is to eliminate global lookups by "localizing"
> them. Others include turning to C functions (see mxTools on my
> Python Pages) or using specialized Python extensions (calls
> to builtin functions are noticeably faster than Python function
> calls).

Skip's page contains some valid hints, although haven't tried the optimizer
et al. Python2C page doesn't seem to work, and it does _not_ speed up
object-oriented programs very much (non-OO, yes; OO, negligibly) yet
consumes space like hell.

None of them touch my primary speed problem; the function/method
invocation 'speed'; C++ function/method invocation seems 200+ times as fast
as Python's, and it really limits the proper OO nature of things,
especially while using apply (another 4x slowdown). Deep class hierarchies
seem no-go, as do several other things.

Although even with 1000x slowdown, my Python runs almost as fast as
assembly on A500.. hmm. ;-)

To my regret I've had to muck too much with C++ recently :P

-Markus

-- 
"One World, One Web, One Program" - Microsoft Promotional Ad
"Ein Volk, Ein Reich, Ein Fuhrer" - Adolf Hitler

We don't see a similarity. We really don't.



From fredrik at pythonware.com  Thu Jul  8 04:09:18 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 8 Jul 1999 08:09:18 GMT
Subject: PIL anim gif
References: <378364C4.E18D7C1E@fnmail.com>
Message-ID: <008e01bec919$2e919210$f29b12c2@secret.pythonware.com>

Per Kistler <kistler at fnmail.com> wrote:
> How to make an animated gif from several Image
> objects in PIL? According to the docu, it seems,
> that it can only anim gifs which are such already.

see the Scripts/gifmaker.py sample script.

</F>




From olipt at mayo.edu  Wed Jul 21 12:09:43 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Wed, 21 Jul 1999 11:09:43 -0500
Subject: Scientific Plotting?
In-Reply-To: <3795DF85.320EADBE@ssec.wisc.edu>
References: <19990721010203.96101.qmail@hotmail.com> <3795DF85.320EADBE@ssec.wisc.edu>
Message-ID: <Pine.LNX.4.10.9907211108100.11968-100000@us2.mayo.edu>

> 
> I agree.  I think it it's great for people loosely affiliated with the
> development community to be writing these packages (and a couple of
> early but promising packages have been pointed out in this thread), but
> what about the people most intimately involved with the design and
> distribution of Python?  Is it a concern?

I don't think Guido want's to micromanage every extension to his
"extension language"

> 
> I see Python as a great possibility for a teaching language, especially
> for young science/engineering students to get in and get the job done. 
> And it's free, which goes a long way in my struggling physics department
> back home.  But how can someone promote the language when people will
> most understandably spend hundreds/thousands of dollars to fall back to
> the commerical maths packages because the plotting is superb and really
> flexible?  Maybe I'm the only one that sees python could reach a wider
> audience without too much more effort - I wouldn't suggest I'm right
> though! ;)

I totally agree with you.  I'm encouraged by the efforts made and the
progress of PIDDLE and Graphite.  You should look in to them if you want
to help out to make python more plotting-friendly.


Travis





From gmcm at hypernet.com  Thu Jul 15 12:21:54 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 15 Jul 1999 11:21:54 -0500
Subject: Getters and Setters
In-Reply-To: <7mk9ks$9jm@ds2.acs.ucalgary.ca>
Message-ID: <1280070597-16324768@hypernet.com>

Neil Schemenauer writes:
> >[Neil Schemenauer]
> >> Recently I found myself wishing that Python had automatic getters
> >> and setters like CLOS.
[Tim]
> >Bleech.
[Neil]
> Bleech to CLOS or to Lisp or to automatic getters and setters?

I vote for the last. Note that Python is in the unusual position that 
it's easier to fool a class user into thinking he's doing a straight 
attribute access (while in fact being shunted through methods) than 
the other way 'round.

[code contest]
> Your code is fast and elegant but it causes reference loops.  If
> only Python had real GC (and closures and macros and generators and
> coroutines and ...).

Without bothering with petty details like timing, I claim moral 
victory <wink> based on the following:
----------------------------------------------------------
from types import FunctionType

class GetSetMeta:
    def __init__(self, name, bases, nmspace):
        self._name, self._bases, self._nmspace = name, bases, nmspace
    def __call__(self, *args):
        return GetSetHelper(self, args)

class GetSetHelper:
    def __init__(self, klass, args):
        self.__class__.__name__ = klass._name
        for (k,v) in klass._nmspace.items():
            self.__class__.__dict__[k] = v
            if not type(v) is FunctionType:
                def getter(self, nm=k):
                    return getattr(self, nm)
                def setter(self, value, nm=k):
                    setattr(self, nm, value)
                self.__class__.__dict__['get%s'%k] = getter
                self.__class__.__dict__['set%s'%k] = setter
        init = self.__class__.__dict__.get('__init__', None)
        if init:
            apply(init, (self,) + args)

MC = GetSetMeta('GetSetMeta', (), {})

class A(MC):
    foo = 1
    def __init__(self):
        bar = 2

a = A()

print `a.getfoo()`
a.setfoo(3)
print `a.getfoo()`
try:
    print `a.getbar()`
except AttributeError:
    print 'Only attributes with class level defaults acquire get/set
    methods'

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

Timings, complaints and further one-upmanships > dev/null


- Gordon



From kperacles at geocities.com  Mon Jul 12 22:26:22 1999
From: kperacles at geocities.com (Karl Putland)
Date: Mon, 12 Jul 1999 21:26:22 -0500
Subject: Python and COM?
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F0F@RED-MSG-50>
Message-ID: <378aa53f.0@news.dbtech.net>

Thanks for the info everyone.

I think that my first generation I might just release with some docs and examples.  Then work on the IDL.  I signed up on the pycom-dev list... I'd love to see this happen.  It's so much nicer in Python that C++ or VB for creating objects.  If anyone knows how far along Christian Tismer is with what he was working on, or what's left to be done, I'd like to help if I could.  I just don't know if I'm ready to play with the big dogs yet ;-)

Karl Putland
kperacles at geocities.com

P.S.
Start a big drive tomorrow. I'm moving back to Denver, CO, from AL, so I will be gone for a couple of weeks (switching providers, looking for broadband, etc...).  E-mail to my geocities address will still reach me when we get there.

Bye.


Bill Tutt <billtut at microsoft.com> wrote in message news:4D0A23B3F74DD111ACCD00805F31D8100DB90F0F at RED-MSG-50...
> 
> 
> > From: Mark Hammond [mailto:MHammond at skippinet.com.au]
> > 
> > 
> > Karl Putland wrote in message 
> > <7m9to7$6qg$82 at rks1.urz.tu-dresden.de>...
> > > For a COM server implemented in python, can a type library 
> > be generated?
> > 
> > No, but nearly :-)  Python has the framework and even some
> > "proof-of-concept" code, but isnt done.
> 
> Err.. Mark means from within Python. That's not to say you can't build a
> type library using an IDL file and MIDL.
> 
> Here's an example for the Python.Interperter COM server that exists in
> win32com\servers\interp.py.
> 
> Unfortuantely, Python.Interpreter needs a slight change inorder to write an
> IDL file.
> Replace the  _public_methods_ line with:
> _dispid_to_func = { 1 : Eval, 2 : Exec }
> A matching IDL that should work is attached below:
> (As an added bonus, this should actually work)
> The only GUID from interp.py is the coclass one. Every other GUID was just
> generated.
> Hope the below helps.... The MIDL docs in MSDN are certainly less than
> useful. :(
> 
> // System idl include nonsense
> import "oaidl.idl";
> import "ocidl.idl";
> 
> [
> uuid(EB048AA4-C2D1-11d2-855D-00C04F797DBA),
> helpstring("Python.Interepreter dispinterface")
> ]
> dispinterface IPythonInterpreter
> { properties: 
> methods:
> [id(1)] HRESULT Eval([in] BSTR bstrExpression, [out, retval] VARIANT
> pvarReturnValue);
> [id(2)] HRESULT Exec([in] BSTR bstrExpression);
> };
> 
> [
> uuid(60240FE4-C2D2-11d2-855D-00C04F797DBA),
> version(1.0),
> helpstring("Python.Interpreter Type Library")
> ]
> library PythonInterpreterLib
> {
> importlib("stdole32.tlb");
> importlib("stdole2.tlb");
> 
> [
> uuid(30BD3490-2632-11cf-AD5B-524153480001),
> helpstring("Python.Interpreter COClass")
> ]
> coclass PythonInterpreter
> {
> [default] dispinterface IPythonInterpreter;
> };
> 
> }
> 
> Bill




From pfortin at cisco.com  Fri Jul  2 16:38:34 1999
From: pfortin at cisco.com (Pierre Fortin)
Date: Fri, 02 Jul 1999 20:38:34 +0000
Subject: Slicing beyond list --  bug or featurette?
Message-ID: <377D234A.4C9BDAC0@cisco.com>

Just ran into a situation which can be summarized as follows:

list = [ 0,1,2,3,4 ]

list[<0..4>] returns the appropriate item

list[m] where m is >= len(list) expectedly fails

list[2:50] returns [ 2,3,4 ]
list[20:] or list[20:30] returns []
 
Can I count on this, or should I code try/except for the day it's
fixed?  This 'feature' reduces the error situations; but is it
correct/intended?

Curious,
Pierre
--
Linux -- The C(r)ash-Free Operating System



From bwarsaw at cnri.reston.va.us  Wed Jul  7 11:12:27 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Wed, 7 Jul 1999 11:12:27 -0400 (EDT)
Subject: total idiot question: +=, .=, etc...
References: <000015b4@bossar.com.pl>
	<76u2rha5qr.fsf@four.npl.uiuc.edu>
Message-ID: <14211.28251.23566.522880@anthem.cnri.reston.va.us>

>>>>> "MAM" == M A Miller <miller at nialas.npl.uiuc.edu> writes:

    MAM> Does anyone have a tweak for python-mode.el that will
    MAM> highlight occurrences of self and self.?

You could probably add "self" to python-font-lock-keywords.  I'm not
entirely sure this is a good idea since self is just a convention, but 
if a lot of people would like to see this, I might consider adding it.

-Barry



From andrew at starmedia.net  Thu Jul  1 14:31:27 1999
From: andrew at starmedia.net (Andrew Csillag)
Date: Thu, 1 Jul 1999 18:31:27 GMT
Subject: Name of function within the function?
References: <377B8C2F.80F44F9@americasm01.nt.com>
Message-ID: <377BB3FF.53A1CB62@starmedia.net>

"Dula, Debbie (EXCHANGE:RICH2:2C25)" wrote:
> 
> I've looked and looked, but can't find the answer to this ... I'm
> sure it's right in front of me ...
> 
> I want the current function's name to be output in my error
> messages, but can't figure out how a function can get its own name
> (without already knowing it).  For example,
> 
>         def f1():
>                 print "Error in function", f1.__name__
> 
> is not really useful for what I need as I might as well print "Error
> in function f1".  Is there any generic way I can get the current
> function's name?
> 
> Thanks,
> Debbie

How's about this:
import sys
def y():
    print callerFunc()
def y2():
    print callerFunc()

def callerFunc():
    try:
        raise "arg"
    except "arg":
        func=sys.exc_traceback.tb_frame.f_back.f_code.co_name
    return func

y()
y2()

Should print:
y
y2

Cheers,
Drew Csillag
-- 
  Writing perl programs is like pounding stakes into your feet:
  You could if you had to, but would be better off if you didn't.
Corallary:
  Using Microsoft products is like pounding stakes into your feet:
  You could if you had to, but would be better off if you didn't.



From ivanlan at callware.com  Sat Jul 10 12:45:27 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sat, 10 Jul 1999 16:45:27 GMT
Subject: Longs
Message-ID: <378778A7.C6BAF5C3@callware.com>

Hi All--
When were longs added to Python?  Whose idea was it?

Thanks,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From puetsch at gmx.de  Fri Jul  2 10:29:44 1999
From: puetsch at gmx.de (Felix Puetsch)
Date: 2 Jul 1999 14:29:44 GMT
Subject: test - ignore
References: <377C7879.428A3E19@worldnet.fr> <377C92FC.9F3E6A08@worldnet.fr>
Message-ID: <8DF7A7E56424242@chief>

[posted and mailed]

pserrand bucodi at worldnet.fr wrote in <377C92FC.9F3E6A08 at worldnet.fr>:

>pserrand wrote:
>
>> test with python
>
>reply test
>
>

Could you pleaser:
a) either use alt.test or similar or
b) cancel your messages afterwards?!

-- 
Felix Puetsch



From lvirden at cas.org  Fri Jul  2 08:04:54 1999
From: lvirden at cas.org (lvirden at cas.org)
Date: 2 Jul 1999 12:04:54 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773CA67.7057F797@quasar.ipa.nw.ru> <3773D55B.8C966174@spacenet.tn.cornell.edu> <slrn7naiqb.8ll.zeitlin@seth.lpthe.jussieu.fr>
Message-ID: <7li9t6$j2k$1@srv38.cas.org>

According to Vadim Zeitlin <zeitlin at dptmaths.ens-cachan.fr>:
:
: The sad truth is that macs are not very common nowadays. I would like to help
:develop wxMac, but I don't have one (nor at home nor anywhere near me in the

I wonder why various industry rags have proclaimed during the past
10 months about Apple's G3 machines outselling other vendor
machines.  ...


-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.



From thomas at xs4all.nl  Mon Jul 19 08:29:10 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Mon, 19 Jul 1999 14:29:10 +0200
Subject: [Q] Type checking...
In-Reply-To: <19990719142024.C6170@xs4all.nl>; from Thomas Wouters on Mon, Jul 19, 1999 at 02:20:24PM +0200
References: <7mv000$sun$1@feed.teaser.fr> <slrn7p637a.4v2.kas@titan.magnetic-ink.dk> <19990719142024.C6170@xs4all.nl>
Message-ID: <19990719142910.D6170@xs4all.nl>

On Mon, Jul 19, 1999 at 02:20:24PM +0200, Thomas Wouters wrote:
> |>>> b=ClassB()
> |>>> a=ClassA

> |>>> b.mymethod(a)
> |Yes :)

Erp, this is a cut-and-paste error (my xterms obviously doesn't talk python
yet.) While 'a=ClassA' is valid, the 'b.mymethod(a)' would have yielded 'No
:(', not 'Yes :)', because ClassA is not an instance of ClassA. It IS
ClassA.

Anyway, i meant to say

> |>>> a=ClassA()

The rest of the example was copied right.

(c)-ly yours,

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From tim_one at email.msn.com  Sat Jul 10 23:50:43 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 10 Jul 1999 23:50:43 -0400
Subject: Dreams.
In-Reply-To: <slrn7ofkrt.55r.sdossey@dolphin.openprojects.net>
Message-ID: <000401becb50$8d3d9c80$479e2299@tim>

[Scott Dossey]
> ...
> What are dreams, you say?  Look at the following python code
> fragment for an idea, first of all, of how they are put into code:
>
> #START PSEUDOCODE FRAGMENT
> import sys
> import dreams
>
> aDream=dreams.dream()
> try:
> 	#Insert code here.
> except:
> 	aDream.wakeup()
> 	sys.exit(1)
>
> aDream.realize()
> #END PSEUDOCODE FRAGMENT
>
> The code, at the location denoted by "#Insert code here." would
> run, but if an exception was thrown, it will have been as if the
> code had "happened in a dream".  No variables will have been changed/
> deleted.  Nothing will have changed from before the "dream" in terms
> of Python's variable state.  Ideally, (but this is much harder), all
> file changes, all things that are "possible" to return to their
> original state, would be returned as well.
> (You can't take back network traffic).
>
> Dream.wakeup() would restore to the state before the dream, and continue
> execution from that line forward.  Dream.realize() would commit to the
> dream, stop saving change-state, and delete the change state it had
> already recorded.
> [and sketches an implementation based on changing code all over the
>  interpreter]

I expect you could get most of this with little effort and little cost under
Unix via fork and a little IPC:  fork the process and try the speculative
stuff in the child, while the parent waits for the child to tell it how
things turned out.  If the child is happy, it tells the parent to die; if
the child has a nightmare, it kills itself after telling the parent "it was
all a bad dream", and the parent goes on as if nothing had happened.  Modulo
problems with external resources (like files indeed), the semantics of fork
guarantee that neither process can affect the other's view of reality.

Another, more portable, but more difficult approach, is adding a
checkpoint/restart facility to Python.  Checkpoint the interpreter state at
the start of a dream, and either delete it or restart from it (as
appropriate) when the dream ends.  This may make every dream expensive to
try (depending on how much state there is to be saved away), but confines
the expense to the start & the end.  While the dream is in progress, or if
nobody is dreaming at all, everything runs as it does now.

Or ...

I'm encouraging you to think about other approaches only partially because
the one you have in mind sucks <wink>:  the chances of getting massive code
changes into the distributed Python these days are vanishingly small.  OTOH,
if you just want to play with it to see how it works out, Python's code base
is a pleasant one to play with!

who-would-take-your-dreams-away-
    takes-your-soul-another-day-ly y'rs  - tim





From jhspies at futurenet.co.za  Wed Jul 14 08:09:24 1999
From: jhspies at futurenet.co.za (Johann Spies)
Date: Wed, 14 Jul 1999 12:09:24 GMT
Subject: Printing from a wxpython application under Windows98
Message-ID: <Pine.LNX.3.96.990714140708.3078A-100000@Johann>

I am a Linux user and have no experience in programming printer functions
in a windows-environment.

I am writing an application that would have to run on Windows 98.  I see
in the documentation that the printer setup dialog is no longer necessary
for windows 98.  How I handle printing in Windows?  How do you set up the
printer?

I have tried the PrintFramework demo on Linux, and it ended in a
segmentation fault.  I do not know why and that prevents me from
experimenting which is the way I learn most of my programming.

To be specific: If I want to print from a database on a preprinted
certificate, which would be the best way to handle it?

1. Using wxPrinter
2. Writing my own classes in python printing using the os module (of which
I am not certain in a Windows environment)
3. Create a LaTeX file and print using LaTeX and dvips.
4. Any other way?

Johann

 --------------------------------------------------------------------------
| Johann Spies                                 Windsorlaan 19              |
| jhspies at futurenet.co.za                3201 Pietermaritzburg             |
| Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
 --------------------------------------------------------------------------

     "God is faithful, by whom ye were called unto the 
      fellowship of his Son Jesus Christ our Lord."         
                                   I Corinthians 1:9 





From arcege at shore.net  Tue Jul  6 08:50:29 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Tue, 06 Jul 1999 12:50:29 GMT
Subject: Python under Java?
References: <MrUJKLEB0KGWFViX4p@holmes.parc.xerox.com> <p0dg3.895$6M6.293143@news.shore.net> <7ls4s2$2pg$1@Starbase.NeoSoft.COM>
Message-ID: <pYmg3.921$6M6.305678@news.shore.net>

Cameron Laird <claird at Starbase.NeoSoft.COM> wrote:
: In article <p0dg3.895$6M6.293143 at news.shore.net>,
: Michael P. Reilly <arcege at shore.net> wrote:
:>Bill Janssen <janssen at parc.xerox.com> wrote:
:>: Has anyone tried embedding Python under Java, using JNI?  When I try it
:>: (with JDK 1.2 on Solaris 2.6 and Python 1.5), I get SEGV inside
:>: Py_Initialize().
:>
:>I don't think anyone has trying to _embed_ Python in Java.  But Python
: .
: .
: .
: To the contrary:  see <URL:http://
: pip.dknet.dk/~pip1848/jpython/Readme_jnios.html>.

Guess I thought wrong. :)

  -Arcege




From ovidiu at cup.hp.com  Thu Jul  1 18:47:57 1999
From: ovidiu at cup.hp.com (Ovidiu Predescu)
Date: Thu, 01 Jul 1999 22:47:57 +0000
Subject: Why does shelve make such large files?
References: <19990701211450.A6827@optiplex.palga.uucp>
Message-ID: <377BF01D.16F8AA33@cup.hp.com>

Gerrit Holl wrote:

> Is it really necesarry for shelve to make such large files?
> Have a look at this:
> /tmp> python
> Python 1.5.2 (#1, Apr 18 1999, 00:16:12)  [GCC 2.7.2.3] on linux2
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import shelve
> >>> d = shelve.open('database')
> >>> d['key'] = 'value'
> >>>
> /tmp> ls -l database
> -rw-rw-r--   1 gerrit   gerrit      16384 Jul  1 21:13 database
>                                     ^^^^^
> 
> 16 KB for only one key!!?
> pickle seems to make _much_ smaller files!
> 
> Why is this?

The shelve module uses DBM which is like a small database that allows
you to store objects and search it for a given key. DBM allows you to
store millions of objects and search for them later without requiring
you to load all the file in memory first.

The pickle module on the other hand is serializing the objects with the
purpose of deserializing them _all_ from the file later. Pickle does not
offer you any way to search for data based on a key, you have to do this
yourself after the objects have been created from the file. This is
opposed to the way shelve handles this, all the key accesses and
insertions in a shelve object are actually reads or writes to or from
the DBM file.

And to answer your question, DBM is creating these big files because of
the way it manages the database. The data in the database file could
have gaps as a result of multiple insertions and deletions. Pickle's
data in files is a simple representation of the objects that were
written and there is no way to update the file other than rewriting it
entirely.

-- 
Ovidiu Predescu <ovidiu at cup.hp.com>
http://www.geocities.com/SiliconValley/Monitor/7464/



From chrisr at synthesys.com  Tue Jul 27 16:35:20 1999
From: chrisr at synthesys.com (Chris Reed)
Date: Tue, 27 Jul 1999 15:35:20 -0500
Subject: Editors for Macintosh
References: <379CB0C8.FA84F91C@callware.com> <kantel-2707991052470001@demokritos.mpiwg-berlin.mpg.de>
Message-ID: <7nl52g$r92$1@hiram.io.com>

>In article <379CB0C8.FA84F91C at callware.com>, Ivan Van Laningham
><ivanlan at callware.com> wrote:
>
>> Hi All--
>> Are there any other good editors (i.e., Python-aware) besides Alpha out
>> there for the Macintosh?  I don't want to _use_ one, I just want to know
>> if there are any that people recommend.
>
>I use BBEdit on the Mac and do to the fact that there is no syntax
>colouring mode for python I've mapped the ending "py" to Tcl syntax
>colouring.
>
>BBEdit does it's job very well because I can save the Python-Files as
>"Python Files" so they run by double click ;-) (Non Mac Users won't
>understand that I suppose :o)
>
>I don't use the IDE very often because I mostly deal with Tkinter on the
>Mac too, but if you don't need that the IDE will be a very good choice
>also.

I remember hearing rumours that the next major release of BBEdit will
include much more powerful syntax colouring capabilities, possibly with user
editing. The reason for such limited support in the current versions is that
each language is hard-coded into the (Object Pascal!) source code and thus
very hard extend.

Just rumours, of course, but I learned much of this from talking with the
author a couple of years ago while trying to convince him that he needed to
provide hooks so we could support our company's web scripting language.

-chris





From tim_one at email.msn.com  Wed Jul 21 19:06:47 1999
From: tim_one at email.msn.com (tim_one at email.msn.com)
Date: Wed, 21 Jul 1999 23:06:47 GMT
Subject: How fast can we multiply?
References: <001801bed279$92c348a0$139e2299@tim> <Pine.GSO.3.96.990720231414.4445A-100000@saga1.Stanford.EDU>
Message-ID: <7n5jq0$b16$1@nnrp1.deja.com>

[Jeffrey Chang <jefftc at leland.Stanford.EDU>]
> Oh, shoot.  I'm 2 days into a week long calculation of normalizing
> a dense ~6000x6000 matrix.  Do you mean I'm the only person doing
> this kind of thing?  ;)

Yes <wink>.

> I've been toying around with the idea of working on optimized
> matrix functions for NumPy (while waiting for the computations to
> finish!).   Since I often work with large matrices, there could
> be big wins for me.  However, it's a little disheartening to hear
> that few other people would benefit from it.
>
> I know the O(n**log2(7)) algorithm for matrix multiply and am
> aware that there are faster ones.  Does anyone know what they are?
> Can anyone recommend a good general book for optimized matrix
> algorithms?  Is Knuth 2 the best source?

It's probably the best accessible source for a quick overview of the
theory.

In practice, though, you're pretty much hosed:  I don't know what your
6000x6000 matrices *contain*, but if it's doubles or even floats then
each consumes a significant fraction of a gigabyte.  So unless you're
running on a real-memory machine, cache effects are likely as least as
important as operation count; on many platforms, much more important.
When high performance requires coding to the most obscure details of a
platform's memory hierarchy, portability goes out the window.

There's an interesting paper about how all this relates to Strassen-
Winograd matmult (the simplest of the "advanced" methods) at:

     http://www.cs.duke.edu/~alvy/papers/sc98/index.htm

Here's part of the abstract:

    Strassen's algorithm for matrix multiplication gains its lower
    arithmetic complexity at the expense of reduced locality of
    reference, which makes it challenging to implement the algorithm
    efficiently on a modern machine with a hierarchical memory
    system. We report on an implementation of this algorithm that
    uses several unconventional techniques to make the algorithm
    memory-friendly.
    ...
    Performance comparisons of our implementation with that of
    competing implementations show that our implementation often
    outperforms the alternative techniques (up to 25%). However, we
    also observe wide variability across platforms and across matrix
    sizes, indicating that at this time, no single implementation is
    a clear choice for all platforms or matrix sizes.
    ...

Par for the course.

buy-an-old-cray-while-they're-still-cheap<wink>-ly y'rs  - tim


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From tony at lsl.co.uk  Fri Jul 16 05:17:28 1999
From: tony at lsl.co.uk (Tony J. Ibbs (Tibs))
Date: Fri, 16 Jul 1999 09:17:28 GMT
Subject: Meta-language for mxTextTools
Message-ID: <011e01becf6c$07c2d820$f0c809c0@lslp7o.lsl.co.uk>

Christian Tismer wrote:
>I remember very well! BCPL is mentioned very rarely on this list!


It is/was a Neat Language - my project at university was written in BCPL,
and we still used it at Laser-Scan for various utility programs for several
years after I first joined the company (back in, ye gods, 1980-ish). On the
other hand, getting lectured by Martin Richards was *not* terribly
interesting.

Ah, the backwards steps that C took:

"OK - first off, let's reintroduce the dangling else problem. Oh, and whilst
we're at it, let's use block delimiters that are hard to distinguish from
parentheses on just about all available printers and terminals [of the time,
and often still!]."

Humph. Now, anyone for Algol68?

Tibs

(time for a plug for the Retrocomputing Museum? I have a link to a mirror
at:
    http://www.pdc.kth.se/~jas/retro/
They have a BCPL compiler, although I haven't tried it...)

--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.demon.co.uk/
Feet first with 5 wheels...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)





From thornley at visi.com  Sun Jul  4 14:50:49 1999
From: thornley at visi.com (David Thornley)
Date: Sun, 04 Jul 1999 18:50:49 GMT
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.i <m3btdvhrlt.fsf@localhost.localdomain>
Message-ID: <d2Of3.1233$U5.237262@ptah.visi.com>

In article <m3btdvhrlt.fsf at localhost.localdomain>,
Lieven Marchand  <mal at bewoner.dma.be> wrote:
>cbbrowne at news.brownes.org (Christopher B. Browne) writes:
>
>The difference between Moore and "just another of the opinionated
>fools" is indeed that he gets results. I'm more into Common Lisp these
>days but "Thinking Forth" is a book that has radically changed many of
>my ideas about programming. 
>
I tend to learn new ways of thinking with new languages.  I came up with
a really nice way of doing one particular thing in a C++ environment,
and I'm positive it was inspired by "Lispy" thinking.  (Now if I can
get my boss to buy in on it.....)

"Thinking Forth" is an excellent book to read.  So is Stroustrup's
"Design and Evolution of C++".  I haven't seen a similar book about
the Lisp way of thinking, not counting books specifically on learning
Lisp and doing it well.  Then again, I find Common Lisp revolutionary
enough sometimes.

>gets the work done. Perhaps Forth and Lisp are really duals in the
>language space with Common Lisp and Moore's minimalist Forth on the
>extreme ends. Forth and Lisp (with Basic) are also the languages with

Lisp and Forth are surprisingly similar in some ways, and way different
in others.

--
David H. Thornley                        | If you want my opinion, ask.
david at thornley.net                       | If you don't, flee.
http://www.thornley.net/~thornley/david/ | O-



From nascheme at ucalgary.ca  Sat Jul 17 10:08:21 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 17 Jul 1999 14:08:21 GMT
Subject: Python and Boehm-Demers GC, I have code.
References: <000401bed023$d08d3360$06a22299@tim>
Message-ID: <7mq2ol$hk8@ds2.acs.ucalgary.ca>

[Tim]
>Just as it appeared <wink>.  So it's really no surprise that the overall
>timing didn't change much in pybench, right?

I'm surprised.  The collection of reference cycles is coming for
free.  I expected there to be some speed tradeoff.  So far the
only problem I see is sorting out the mallocs (ie. extension
modules may use their own malloc).  This should be done anyhow in
order for alternate Python malloc implemetations to be used.  The
other problem is objects that expect not to be collected even
though Python has no references to them (Tkinter callbacks).  I
don't think this problem is unsolvable.  Things are much better
than I expected.

With some tuning maybe the gc version of Python will perform
better than the regular version.

[Tim on Tkinter callbacks being collected]
>It's convoluted for sure, but I do expect you'll find "the answer" at the
>bottom of it.
>
>rc-is-so-simple<wink>-ly y'rs  - tim

This is really nasty.  I am trying to allocate this data with
regular malloc.  If I use malloc and free then this data should
not get collected and the GC should realize that the Python
functions still have references to them.  So far I am not having
much luck with this approach.  It would help if I totally
understood what was happening here.  I have more work to do.


    Neil



From janssen at parc.xerox.com  Fri Jul  9 19:43:52 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Fri, 9 Jul 1999 16:43:52 PDT
Subject: Python webcrawler?
Message-ID: <99Jul9.164357pdt."3569"@watson.parc.xerox.com>

Can anyone point me to a good Python webcrawler system?  I'm looking for
something like the WebSphinx software from CMU.  Thanks.

Bill



From jhspies at futurenet.co.za  Tue Jul  6 11:22:00 1999
From: jhspies at futurenet.co.za (Johann Spies)
Date: Tue, 6 Jul 1999 15:22:00 GMT
Subject: AttributeError - a dumb question
Message-ID: <Pine.LNX.3.96.990706171320.374A-100000@Johann>

Why does the following code produce an attribute error?
class F:
    def __init__(self):

          x = self.Krylys()

    def KryLys(self):
        l = [1,2,3,4,5]
        return l
    

l = F
print dir(l)
print hasattr(l, 'x')
print hasattr(l,'Krypos')
print l.x

What I do not understand is that the output from this code looks like
this:

['KryLys', '__doc__', '__init__', '__module__']
0
0
Traceback (innermost last):
  File "<stdin>", line 14, in ?
AttributeError: x

'Krylys' according to dir(l) is an attribute.  Why does
hasattr(l,'Krypos') report that it is not?

Johann


 --------------------------------------------------------------------------
| Johann Spies                                 Windsorlaan 19              |
| jhspies at futurenet.co.za                3201 Pietermaritzburg             |
| Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
 --------------------------------------------------------------------------

     "Ask, and it shall be given you; seek, and ye shall
      find; knock, and it shall be opened unto you."        
                                  Matthew 7:7 




From cgw at fnal.gov  Thu Jul  8 12:00:09 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Thu, 8 Jul 1999 16:00:09 GMT
Subject: Two Pythons talking to each other?
In-Reply-To: <378465E9.7762BE9A@fnmail.com>
References: <199907072203.AAA09601@axil.hvision.nl>
	<378465E9.7762BE9A@fnmail.com>
Message-ID: <14212.51977.601241.772478@buffalo.fnal.gov>

Per Kistler writes:
 > Hi Hans
 > 
 > One could use sockets with AF_INET, if it runs on another
 > machine, and sockets with AF_UNIX, if it's on the the same
 > host.

There's no particular reason to use AF_UNIX, even if the two processes
are running on the same node.  AF_INET with a loopback IP address
works fine, and is portable (not all platforms support unix-domain
sockets).



From mstenber at cc.Helsinki.FI  Tue Jul 27 00:45:28 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 27 Jul 1999 07:45:28 +0300
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <p51zdvn4c7.fsf@bidra241.bbn.hp.com> <14236.32616.980578.536614@dolphin.mojam.com>
Message-ID: <al8aesin1jr.fsf@sirppi.helsinki.fi>

Skip Montanaro <skip at mojam.com> writes:
> I got a bit confused attributing this message.  I think we have too many
> Markus's running around loose... ;-)  I think I have the quoted text
> attributed correctly.  Apologies if not.

You had it reversed ;)

> Better yet, since the Python source code is available, you could prove it by
> speeding up the interpreter.  Many people have optimized different parts of
> Python over the lifetime of the software.  It's significantly faster than it
> was a couple years ago.

Being able to note a problem differs greatly from being able to _fix_ a
problem ;-) I've glanced thru it, and can't offhand find anything _very_
badly implemented that I could fix.

> I think one of the impediments to improving the speed of the language has
> been the dramatic performance improvement in hardware.  There's not a lot of
> incentive to pour lots of time into speeding up method calls when for a few
> hundred bucks you can simply drop in a faster processor, disk or networking
> card.

True.

>     Kohler> P.S. This might surface as one example for 'why scripting
>     Kohler> languages don't always cut it' in my master's thesis; if I have
>     Kohler> made some grievous error(s?), corrections are welcome.
> I don't recall anyone ever saying scripting languages would always cut it.

Scripting languages are offered as be-all end-all by many people, for some
reason. 

>     Kohler> P.P.S. Too bad Python doesn't have anything that creates _nice_
>     Kohler> code, like Stalin for Lisp ;-)
> There is a substantially shorter history to Python than to Lisp.  Give it
> time.  In addition, Lisp has been knocking around the academic world for
> 40-odd years.  With that many graduate students looking for a thesis topic,
> it was likely that some of those theses would focus on performance
> enhancement.

True; maybe in 30 years, eh?

> give-typewriters-to-enough-chimpanzees-and-you'll-get-Shakespeare-ly y'rs,

-Markus (Stenberg)



From gmcm at hypernet.com  Thu Jul 29 18:21:12 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 29 Jul 1999 17:21:12 -0500
Subject: Need help with TCP/IP client access from Windows
In-Reply-To: <slrn7q1ej3.18d.esr@snark.thyrsus.com>
Message-ID: <1278839420-22085362@hypernet.com>

Eric Raymond writes:

> I need a Windows equivalent of this class for a client I'm writing.
[snip]
> The issues here are:
> 
> (1) How do I get (at least) line-oriented access to an Internet
> server
>     from a Python program running under Windows?

No problem there. (The only difference - when using Python - is that 
you can only select on sockets on Windows).
 
> (2) How does one read and write the Windows equivalent of a dotfile?
>     Must this involve the (shudder) registry?

You could, by using Mark's win32 extensions. MS would prefer you
kept this kind of info in the registry, but that doesn't stop
anybody, least of all you <g>.

The issue is really what is meant by a HOME directory. On NT, this 
has a certain amount of meaning - there's probably a 
%WINDIR%/profiles/%USERNAME% directory tree for that user (where 
those funky %...% things are environment vars), but probably it's 
safer to get that path out of the registry. On Win9x, it's pretty 
meaningless, since a logon and username are pretty much an empty 
formality. There's almost no way to keep multiple users on a Win9x 
box from stepping on each other. (Of course, I mean multiple serial 
users; the only possible kind in Windows).

Probably better just to have a default file, and let the user 
override that if need be. Which may well never happen.

> (3) Is there a reliable way for Python programs under Windows to
> point a local
>     browser at a specified URL?

The best way is to use WinExec from Mark's win32 extensions. You just 
give it the URL and it loads the user's default browser. Using 
os.system is harder, because most of the time a browser is not on the 
user's path.

However, this worked fine for me:

browse='"C:\\Program Files\\Netscape\\Communicator\\Program\\netscape"
%s'

Note the double quotes around the nowhere-but-Windows <snicker> path 
with spaces in it, and that argv[1] is expected to be a URL (same for 
IE). Also note that the path doesn't go thru the c runtime lib, so 
those wonderful backslashes can't be avoided.

- Gordon



From chega_ at yahoo.com  Wed Jul 28 20:08:07 1999
From: chega_ at yahoo.com (Vadim Chugunov)
Date: Wed, 28 Jul 1999 17:08:07 -0700
Subject: 1.5.2 for: else:
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC5B@molach.origin.ea.com>
Message-ID: <7no61e$lrk@news.dns.microsoft.com>

How about this:

8<---------------------------------
class Item:
    def __init__(self,key):
        self.key = key

seq= [Item(1),Item(2),Item(3)]

for item in seq:
    if item.key==42:
        break
else:
    item = Item(42) # oh well, append() does not return a value
    seq.append(item)

print item.key
8<---------------------------------


Stidolph, David <stidolph at origin.ea.com> wrote in message
news:11A17AA2B9EAD111BCEA00A0C9B41793034AAC5B at molach.origin.ea.com...
>
> Please show your example using the else .. I'd like to see it.
>
> -----Original Message-----
> From: Vadim Chugunov [mailto:chega_ at yahoo.com]
> Sent: Wednesday, July 28, 1999 4:58 PM
> To: python-list at cwi.nl
> Subject: Re: 1.5.2 for: else:
>
>
> I would say that else: clause after a loop is really related to the if:
> controlling the break,
> rather than to the loop itself.
>
> When I first saw Python syntax for loops I said: "A-ha!  So in Python I will
> not
> need a goto
> in situations like this:
> -----
> for(Item* pitem=pseq->First(); pitem; pitem=pitem->Next())
>     if (pitem->key==42)
>         goto Found;
>
> pitem = pseq->Insert(new Item());
> Found:
>     // use item pointed to by pitem
> -----
> In fact, I do not see any other good use for else: clause in a loop.
> Maybe the docs should just explicitly say what this feature is good for ?
>
> Vadim






From akuchlin at mems-exchange.org  Thu Jul  8 09:24:33 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: Thu, 8 Jul 1999 13:24:33 GMT
Subject: python core dump (SIGBUS) on Solaris
In-Reply-To: <37849166.DD658F5@uk.sun.com>
References: <3F8BA119B48AD11197470008C7A493CB98B443@HQOBGEX1>
	<37830BB2.94C4613D@uk.sun.com>
	<37837154.7A6889A@ix.netcom.com>
	<slrn7o7nb9.eo.amk@207-172-111-58.s58.tnt1.ann.va.dialup.rcn.com>
	<37849166.DD658F5@uk.sun.com>
Message-ID: <14212.42641.217500.947671@amarok.cnri.reston.va.us>

Gary Pennington - UK Performance Centre writes:
>It would be more efficient to assume that you are processing a string and
>only try to unpickle if the string is detected as a pickle. The tricky bit
>is detecting that it is a pickled object in the string. Does a pickled
>string start with some kind of magic sequence?

	Unfortunately, there's no magic sequence in pickles.  The
relevant function in Cookie.py is: def _debabelize(val, loads=loads):
    str = _unquote(val)
    try:
        return loads(str)
    except:
        return str

Which is fine if loads() just chokes on bad input and raises an
exception, but not if it dumps core.  I don't see how to change this
behaviour in Cookie.py without breaking backward compatibility by
turning unpickling off by default, so maybe someone should look at
fixing cPickle.  The bug seems to be a memory overrun, because it
doesn't fail on the first attempt:

amarok Python-1.5>./python
Python 1.5.2+ (#82, Jun 17 1999, 09:53:36)  [GCC 2.8.1] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import cPickle
>>> cPickle.loads('garyp')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
cPickle.BadPickleGet: path
>>> cPickle.loads('garyp')
Bus Error (core dumped)

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
The biggest difference between Heaven's Gate and mainstream Christianity is
that the Heaven's Gate people never tried to brainwash my kids.
    -- Rev. Ivan Stang




From moshez at server.python.net  Fri Jul 30 13:15:25 1999
From: moshez at server.python.net (Moshe Zadka)
Date: Fri, 30 Jul 1999 13:15:25 -0400 (EDT)
Subject: syntax
In-Reply-To: <002801beda96$89097300$f29b12c2@secret.pythonware.com>
Message-ID: <Pine.LNX.4.10.9907301309190.10808-100000@starship.python.net>


On Fri, 30 Jul 1999, Fredrik Lundh wrote:

[Xavier Bec asks why colons are necessary after if, for or while]

[Fredrik Lundh answers]
> readability.  think about it for a while, and then read:
> http://www.deja.com/=dnc/getdoc.xp?AN=171314015

In which the Guido writes
...
> Python inherits this feature from ABC.  The designers
> of ABC spent quite some time experimenting with language variants and
> showed them to beginning users (with no programming experience).  They
> initially had indentation for block structure, but not the colon.
> Their experience was that first-time users had some trouble
> understanding that the indented block was "subordinate" to the if- or
> for-statement preceding it.
...

How about the same solution as in "pindent"? I.e., have a preprocessor
which automatically adds the colons. While it's at it, it can
automatically correct '=' to '==' and vice versa, so we can forget all
the crutches and just write code <0.9 wink>

heck-why-not-let-it-write-the-damn-code-ly y'rs, Z.




From rceci at my-deja.com  Thu Jul  1 11:04:17 1999
From: rceci at my-deja.com (rceci at my-deja.com)
Date: Thu, 01 Jul 1999 15:04:17 GMT
Subject: 1.5.1 on IRIX
Message-ID: <7lg019$fro$1@nnrp1.deja.com>

Hello,

I built the 1.5.1 distro on one of our irix6.4 systems ( n64-mips4
arch), using the -32 (-o32)  compiler.  This was installed in the
"normal" place - /usr/local/....  and ran fine on that machine.

Now I have moved this installation to a NFS mount, and tried running the
interpreter on another irix6.4 system with similar characterstics.  I am
setting PYTHONPATH to point to the location of new python runtime
libraries:

The libs originally where built/installed  -> /usr/local/lib/python1.5
re-installed as ${somedir}/lib/irix32, and set PYTHONPATH to
${somedir}/lib/irix32.

the interpreter itself is re-located in ${somedir}/bin/irix32/

I can run python in its new location on the old system, just fine:

    setenv PYTHONPATH ${somedir}/lib/irix32
    ${somedir}/bin/irix32/python -v

and get the normal debugging info when the modules are loaded, which
confirms they are loading out of my specified PYTHONPATH.

However, when I try running this installation on another one of our irix
systems, I get:

Could not find platform independent libraries <prefix>
Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
Python 1.5.1 (#3, May 21 1999, 16:24:24) [C] on irix646-32-mips2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
Contacting Registry ...
environment exported ...
-- Using registry version :  10.0

>>>

The messages about Registry, are coming from SWIGged module linked
against the interpreter.  I looks like that is working fine in this
installation on both systems.  It would appear that the warnings on the
second system are non-fatal.

Any ideas how to "fix" or suppress them?

Rob


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Brian at digicool.com  Fri Jul  9 09:58:17 1999
From: Brian at digicool.com (Brian Lloyd)
Date: Fri, 9 Jul 1999 13:58:17 GMT
Subject: Python for the PalmPilot and Other Devices
Message-ID: <7m9tmb$6qg$23@rks1.urz.tu-dresden.de>

> You'd have to ask Brian Lloyd the original reasons for porting
> Python to his pathetic little 2MB CE device. ;-)  Now that CE's
> got 16-32MB direct from the factory, well, grins all around.

Well, to see if I could do it, of course :^) And because I happened
to have a CE device, and not a Pilot.

> The CE devices have some advantages.  There is a wide diversity 
> in models (palm sized, handhelds, even laptop sized units) and
> vendors (at least a dozen or so).  

They also have some disadvantages - for example, there is a wide
diversity in models (palm sized, handhelds, even laptop sized units)
and vendors (at least a dozen or so)...  :^)  Configuration mgmt
has been a big pain in the a** for the CE port, and it's only made
worse by the fact that everyone working on it has a pesky "real job" 
and tend to be very busy.


> The Win32 development heritage
> is reasonably well-understood, if often cursed.  The CE platform 
> includes a lot of built-in features (support for TCP/IP, sockets, 
> IrDa).  Microsoft is using its clout to move CE into automobiles 
> and cable set-tops.  True, CE represents a capitulation to a well
> known software vendor, but ...

He he - when I first embarked on that, I thought that the built in
features would be great help -- until I realized that there were
some missing features like, oh, most of the std C library that
Python takes for granted... :^)

Seriously though, the successive generations of CE devices have
been steadily piling more and more desktop win32 functionality
in - you could say that this is good or bad, but in the long run
it is probably good. Once CE finishes turning into NT in a handheld
(never mind that it will need to come with 256M of ram and need to
be powered by a carry-along car battery), we'll be able to use most
of the existing win32 port code pretty much unchanged :)

There are actually some potentially interesting opportunities for Python

on CE. For example, I believe that the upcoming sega Dreamcast video 
game unit will run CE as it's os, so it's not out of the realm of 
possibility that one could use Python for video game prototyping
or something.

I personally am waiting for a decent AutoPC so I can write a Python
script to drive me to work so I can get some shuteye on the way 
in... :^)


Brian Lloyd        brian at digicool.com
Software Engineer  540.371.6909              
Digital Creations  http://www.digicool.com 






From gerrit.holl at pobox.com  Thu Jul  8 15:16:51 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 8 Jul 1999 21:16:51 +0200
Subject: scoping assertion
In-Reply-To: <3783C878.B8C3790B@bioreason.com>; from Andrew Dalke on Wed, Jul 07, 1999 at 03:36:56PM -0600
References: <37839491.3F7C1F6@mikemccandless.com> <3783C878.B8C3790B@bioreason.com>
Message-ID: <19990708211651.A17658@optiplex.palga.uucp>

On Wed, Jul 07, 1999 at 03:36:56PM -0600, Andrew Dalke wrote:
> From: Andrew Dalke <dalke at bioreason.com>
> Subject: Re: scoping assertion
> Date: 	Wed, 07 Jul 1999 15:36:56 -0600
> To: python-list at python.org
> 
> 
>   try:
>     1/0
>   except ZeroDivisionError:
>

isn't the following much easier, faster and readabler?
try:
  raise err
except err:

regards,
Gerrit.

-- 
The Dutch Linuxgames homepage:	http://linuxgames.nl.linux.org
Personal homepage:		http://www.nl.linux.org/~gerrit/

Discoverb is a python program (in several languages) which tests the words you
learned by asking it. Homepage: http://www.nl.linux.org/~gerrit/discoverb/
Oh my god! They killed init! You bastards!



From cbbrowne at news.hex.net  Sat Jul  3 00:28:53 1999
From: cbbrowne at news.hex.net (Christopher Browne)
Date: Sat, 03 Jul 1999 04:28:53 GMT
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it>
  <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> <wkzp1g2t2e.fsf@ifi.uio.no> <slrn7nnfs3.19k.wtanksle@dolphin.openprojects.net>
Message-ID: <9kgf3.301$F7.23516@news3.giganews.com>

On Thu, 01 Jul 1999 19:20:36 GMT, William Tanksley
<wtanksle at dolphin.openprojects.net> wrote:
>On 01 Jul 1999 19:11:53 +0200, Lars Marius Garshol wrote:
>>* William Tanksley
>
>[I made the joking claim that Lisp was an improvement on Forth, then to
>defuse criticism I claimed:]
>>| (Yes, I know -- the ancestry goes the other way around: Lisp ->
>>| Scheme -> Forth.)
>
>>Huh? Forth dates back to the 60s, whereas Scheme is from 1975, and I'm
>>quite unsure of whether Chuck Moore knew Lisp at all.
>
>I don't think Forth is that old, certainly not in any usable form.  Here,
>according to the history on www.forth.com, "the first program to be called
>Forth was written in about 1970."  That substantiates what you're saying
>about Scheme -- I'm a little suprised.  Okay, toss Scheme out of the
>ancestry ;-).

Yes, please.  If there's relationship between Forth and Lisp, it does
not lie in the Scheme direction.

>Chuck Moore, the inventor of Forth, got a BA in physics from MIT and went
>into grad school at Stanford.  He claims to have taken classes from and
>learned a lot from Lisp.

I don't remember seeing any indication of Moore having encountered
Lisp; if so, that is very interesting.

>Forth is very much like Lisp, with the subtraction of memory management
>and the addition of implicit parameter passing.  Forth is Lisp for people
>who prefer "as simple as possible" to "but no simpler."  (And it IS a very
>good language; I use it more than Scheme.)
>
>A paragraph like that is neccesarily an oversimplification.  Perhaps I can
>get away with saying that Forth learned more from Lisp than it did from
>any other language existing at the time?  Yes, I think that's tenable.

Moore was (and seems still to be) *very* independently-minded; anyone
that remembers "BLOCK wars" or "FP stack wars" or "Is ANSI Forth
uselessly incompatible?" is probably also aware that Moore has been
sufficiently "extreme" as to be above and beyond all such "petty"
disagreements.

The fact that he actually designed some microprocessors with OK makes
the unbelievably sparse architecture of OK seem faintly believable.

(Or, putting it another way, if he *hadn't* done such massive things
as designing several Forth-related languages, founding companies, and
designing, building, and even *selling* families of microprocessors,
the rather radical things he says would probably be regarded as the
ravings of a Net.Kook.)

Forth has enough common features (e.g. - having a stack, which allows
recursion even in the simplest of words) with Lisp as to make an
association not unbelievable.

But the lack of GC, dynamic allocation, and the likes keep them
distant enough that you have to look closely to see the resemblance.
-- 
DOS: n., A small annoying boot virus that causes random spontaneous
system crashes, usually just before saving a massive project.  Easily
cured by UNIX.  See also MS-DOS, IBM-DOS, DR-DOS.  (from David Vicker's
.plan)
cbbrowne at hex.net- <http://www.hex.net/~cbbrowne/langobscure.html>



From jeremy at cnri.reston.va.us  Mon Jul 12 14:31:05 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Mon, 12 Jul 1999 14:31:05 -0400 (EDT)
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
In-Reply-To: <14218.7724.879301.831590@weyr.cnri.reston.va.us>
References: <3787D36A.815B7C29@ssec.wisc.edu>
	<3787D8DA.3774AA7D@ssec.wisc.edu>
	<7m8p7t$o7s@dfw-ixnews10.ix.netcom.com>
	<3787FB51.12FE0048@ssec.wisc.edu>
	<7m923i$dem@dfw-ixnews5.ix.netcom.com>
	<14217.63931.75257.566266@weyr.cnri.reston.va.us>
	<7md2ip$19q2$1@nntp6.u.washington.edu>
	<14218.7724.879301.831590@weyr.cnri.reston.va.us>
Message-ID: <14218.13417.796678.728859@bitdiddle.cnri.reston.va.us>

>>>>> "FLD" == Fred L Drake <fdrake at cnri.reston.va.us> writes:

  FLD> Donn Cave writes:
  >> For what it's worth, me too: current working directory is a
  >> process setting.  It's relevant to files and directories, to be
  >> sure, but whoever put chdir() and getcwd() in process parameters
  >> knew what he was doing.

  FLD>   I am vindicated!

By vindicated, I assume you mean that you and Donn are both twisted
old Unix geeks <wink>.

Jeremy



From catlee at my-deja.com  Mon Jul 19 17:13:39 1999
From: catlee at my-deja.com (catlee at my-deja.com)
Date: Mon, 19 Jul 1999 21:13:39 GMT
Subject: Killing a thread
References: <7mu6a5$h61$1@nnrp1.deja.com> <7mvml3$u5k$1@nntp6.u.washington.edu> <ePJk3.4$Z11.6082@news-dal.corridex.com>
Message-ID: <7n04dm$8ca$1@nnrp1.deja.com>

In article <ePJk3.4$Z11.6082 at news-dal.corridex.com>,
  "Evan Simpson" <evan at tokenexchange.com> wrote:
> This makes me wonder; Zope has a similar problem, which it mostly
handles
> with severe restrictions on what's available to the user (no 'while',
only
> 'for', and no '[x]range' to generate arbitrarily large tuples).
> Nevertheless, a creative malcontent could consume a lot of resources
with
> nested loops.
>
> Bytecodehacks to the rescue?  <waves hands furiously> Wrap the user
code in
> a function (if it isn't already), find all jump-back codes, and
insert a bit
> of test code in front of them which bails out if ?a global variable
is set?
> (not sure what's best, here).  You'll take a slight performance hit,
but
> that probably won't matter in this context.  Perhaps you could inline
a call
> to time.clock and use it to limit the running time of the function.

It looks like it is bytecodebacks to the rescue!  I've never used the
bytecodehacks module before, but WOW!  This has so many uses!  Here's
some code I whipped up in 15 minutes that handles the "while 1: 1+1"
problem:
from bytecodehacks.ops import *
from bytecodehacks.code_editor import *
from threading import Thread
from rexec import RExec
import time,thread

class thing(Thread):
	def __init__(self):
		Thread.__init__(self)
		self.done_flag=0

	def run(self):
		self.renv=RExec()
		self.source="while 1: 1+1"
		self.code=compile(self.source,'<string>','exec')
		self.fix_code()
		self.renv.r_exec(self.code)

	def fix_code(self):
		co=EditableCode(self.code)
		pos=0
		while pos < len(co.co_code):
			if co.co_code[pos].is_jump():
				op=LOAD_NAME("quit_if_done")
				co.co_code.insert(pos,op)
				op=CALL_FUNCTION(0)
				pos=pos+1
				co.co_code.insert(pos,op)
				pos=pos+1
			pos=pos+1
		self.code=co.make_code()
		self.renv.modules['__main__'].__dict__['quit_if_done']
=self.quit_if_done

	def stop(self):
		self.done_flag=1

	def quit_if_done(self):
		if self.done_flag==1:
			thread.exit()

t=thing()
t.start()
time.sleep(1)
t.stop()
t.join()

Now, some things I'm not sure about:
- Everything SHOULD be cleaned up correctly since the thread exited
itself, it wasn't killed forcibly
- Is it conceivable that a user could make write some resource
intensive code without using any kind of jump operations?  I don't
think so, unless s/he writes out a 200 megabytes file, but that's easy
enough to limit.
- Can the user somehow overwrite the quit_if_done function?

Cheers,
Chris


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gmcm at hypernet.com  Thu Jul 22 19:07:34 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 22 Jul 1999 18:07:34 -0500
Subject: string.replace
In-Reply-To: <Pine.BSF.4.10.9907221446520.20655-100000@soda.CSUA.Berkeley.EDU>
Message-ID: <1279441446-22507099@hypernet.com>

marlon asks:

> can anyone tell me why this breaks?
> 
> Python 1.4 (Nov  4 1997)  [GCC 2.7.2.3]
         ^^^

> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import sys, string
> >>> line = 'cat "in" the hat'
> >>> line = string.replace( line, '"', '')
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> AttributeError: replace
> >>> 

I believe string.replace came with 1.5.

- Gordon



From arcege at shore.net  Fri Jul  9 09:08:02 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Fri, 09 Jul 1999 13:08:02 GMT
Subject: range(Instance) vs. foo[Instance]
References: <slrn7o43v5.qkm.jepler@craie.inetnebr.com>
Message-ID: <7m9tm5$6qg$19@rks1.urz.tu-dresden.de>

Jeff Epler <jepler at inetnebr.com> wrote:
: Suppose I have a class with __int__ defined:

: class c:
:   def __init__(self, val):
:     self.val = int(val)

:   def __int__(self): return self.val

:>>> x=c(1)
:>>> range(x)
: [0]
:>>> (0,1,2)[x]
: Traceback (innermost last):
:   File "<stdin>", line 1, in ?
: TypeError: sequence index must be integer

: It seems desirable that the subscript operator call x.__int__ just like
: range() must effectively do.  Is there some reason why not?

: (Not the subscript operator in general, but the subscript operator for
: the builtin types of lists and tuples)

: I ask because I recently wrote a 5-minute hack on top of bytecodehacks that
: makes "integers" function as rational numbers in the interactive Python
: console---functions like "range" work, but subscripting of tuples and lists
: is broken.

: Jeff

Yup, this sounds like a bug, since [].insert( c(0), "spam") would work
as you expect.  Some reasons to not have this would be:
* How do you handle complex numbers as subscripts,
* Strings containing numbers as subscripts.

But this looks like a holdover from older code.  Myself, I'm not sure
if strings (containing a number) should be accepted.

I've submitted a patch to Guido and am including it here.  This will
make all built-in sequences (strings, tuples and lists) perform an
implicit seq[int(subscr)] if subscr is not an int.  The subscripting
will still fail if int() fails.

  -Arcege

Index: Objects/abstract.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Objects/abstract.c,v
retrieving revision 2.27
diff -c -r2.27 abstract.c
*** abstract.c	1999/01/10 16:56:58	2.27
--- abstract.c	1999/07/09 12:30:03
***************
*** 228,233 ****
--- 228,234 ----
  	PyObject *key;
  {
  	PyMappingMethods *m;
+ 	PyObject *tmp, *v;
  
  	if (o == NULL || key == NULL)
  		return null_error();
***************
*** 239,244 ****
--- 240,250 ----
  	if (o->ob_type->tp_as_sequence) {
  		if (PyInt_Check(key))
  			return PySequence_GetItem(o, PyInt_AsLong(key));
+ 		else if ((tmp = PyNumber_Int(key)) != NULL) {
+ 			v = PySequence_GetItem(o, PyInt_AsLong(tmp));
+ 			Py_DECREF(tmp);
+ 			return v;
+ 		}
  		return type_error("sequence index must be integer");
  	}
  
***************
*** 252,257 ****
--- 258,265 ----
  	PyObject *value;
  {
  	PyMappingMethods *m;
+ 	PyObject *tmp;
+ 	int v;
  
  	if (o == NULL || key == NULL || value == NULL) {
  		null_error();
***************
*** 264,269 ****
--- 272,282 ----
  	if (o->ob_type->tp_as_sequence) {
  		if (PyInt_Check(key))
  			return PySequence_SetItem(o, PyInt_AsLong(key), value);
+ 		else if ((tmp = PyNumber_Int(key)) != NULL) {
+ 			v = PySequence_SetItem(o, PyInt_AsLong(tmp), value);
+ 			Py_DECREF(tmp);
+ 			return v;
+ 		}
  		type_error("sequence index must be integer");
  		return -1;
  	}
***************
*** 278,283 ****
--- 291,298 ----
  	PyObject *key;
  {
  	PyMappingMethods *m;
+ 	PyObject *tmp;
+ 	int v;
  
  	if (o == NULL || key == NULL) {
  		null_error();
***************
*** 290,295 ****
--- 305,315 ----
  	if (o->ob_type->tp_as_sequence) {
  		if (PyInt_Check(key))
  			return PySequence_DelItem(o, PyInt_AsLong(key));
+ 		else if ((tmp = PyNumber_Int(key)) != NULL) {
+ 			v = PySequence_DelItem(o, PyInt_AsLong(tmp));
+ 			Py_DECREF(tmp);
+ 			return v;
+ 		}
  		type_error("sequence index must be integer");
  		return -1;
  	}



From aahz at netcom.com  Thu Jul 22 11:01:28 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 22 Jul 1999 15:01:28 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
References: <001101bed3f2$5e8874a0$642d2399@tim> <7n675p$qsk@dfw-ixnews7.ix.netcom.com> <14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>
Message-ID: <7n7bo8$204@dfw-ixnews19.ix.netcom.com>

In article <14230.42741.34295.650213 at 153.chicago-05-10rs.il.dial-access.att.net>,
Skip Montanaro  <skip at mojam.com> wrote:
>>>>>> "Aahz" == Aahz Maruch <aahz at netcom.com> writes:
>
> Aahz> Yup.  Today I had to lay down the law to our programming team: use of
> Aahz> the construct "from foo import *" is now expressly, explicitly, and
> Aahz> completely forbidden.  "from foo import bar" is deprecated; "from foo
> Aahz> import foo" is acceptable; and "import foo" is strongly preferred.
>
> Aahz> I'd be in favor of getting "from foo import *" entirely out of the
> Aahz> language, but I suspect it's just too convenient for short hacks.
>
>What Tim said about programmers not inventing control structures is somewhat
>different than what Aahz put his foot down about.  There's nothing wrong
>with "from foo import *" when it's used in the right context.  Some modules
>in the standard distribution were designed specifically to support this
>(types and Tkinter come to mind).  Used willy-nilly, you can get into
>trouble with it, one-man show or not.

I'm afraid I can't agree that there's *ever* a good time to use "from
foo import *" in a multi-programmer environment.  Too often, you have a
programmer new to a particular chunk of code trying to debug it; not
knowing exactly where a method/function call is coming from vastly
increases debugging time (particularly when there are hundreds of
modules and scripts).  If you want to use an unadorned class/function,
you can just take the trouble to explicitly list it as "from foo import
bar, baz".

My point was related to Tim's in that we're both talking about how
Python's strait-jacket approach is a Good Thing; I'm simply using an
example where I think Python could actually be tightened.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From michels at linax1.mpae.gwdg.de  Thu Jul 22 03:40:26 1999
From: michels at linax1.mpae.gwdg.de (Helmut Michels)
Date: Thu, 22 Jul 1999 09:40:26 +0200
Subject: Repost: Re: Scientific Plotting?
References: <379392E7.EB0F29E7@ssec.wisc.edu> <7n5fhn$9e5$1@nnrp1.deja.com>
Message-ID: <3796CAEA.41C6@linax1.mpae.gwdg.de>

mhuster at hotmail.com wrote:
> 
> ...
> Original reply:
> 
> I am also trying to find an adequate *object* based plotting package. I
> am writing some signal processing software and want to code plotting
> into the classes I am defining.
> 
> Some requirements:
> 1) Postscript output (for inclusion in high quality Word & Powerpoint
> documents) GIF, JPG, etc just do not have the resolution. They default
> to 72 dpi rendering. You can overcome this on Linux using Image Magick's
> convert -dens 300. The resulting files are large.
> 
> 2) Object oriented
> 3) Images
> ... 

What about the plotting library DISLIN? It supports high quality
PostScript files and some elementary image routines. PostScript fonts
can be used for PostScript files. Manuals and software are available
from the DISLIN Homepage

          http://www.linmpi.mpg.de/dislin

   -------------------
  Helmut Michels
  Max-Planck-Institut fuer Aeronomie  Phone: +49 5556 979-334
  Max-Planck-Str. 2                   Fax  : +49 5556 979-240
  D-37191 Katlenburg-Lindau           Mail : michels at linmpi.mpg.de



From matomira at iname.com  Thu Jul  1 08:23:52 1999
From: matomira at iname.com (Fernando Mato Mira)
Date: Thu, 01 Jul 1999 14:23:52 +0200
Subject: Reflection in Python?
References: <7ldglr$ita$1@nnrp1.deja.com> <nxlnd0y6iv.fsf@raillkatt.itek.norut.no>
Message-ID: <377B5DD8.3666A5CA@iname.com>

If you are into reflection, you should really take a look at CLOS,
before setting things in stone in Python.

You can find my stuff in
http://ligwww.epfl.ch/matomira/en/publications.html (the last book was
finally released in '96 as "Advances in Metalevel Architectures and
Reflection"). For DB interface stuff, there was
paper in LUV'95 I think.






From tismer at appliedbiometrics.com  Thu Jul 15 12:28:05 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Thu, 15 Jul 1999 16:28:05 GMT
Subject: Meta-language for mxTextTools
References: <002001beced7$692bc800$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <378E0C15.F3CC8E8C@appliedbiometrics.com>


"Tony J. Ibbs (Tibs)" wrote:
> 
> Christian Tismer <tismer at appliedbiometrics.com> wrote:
> > I got the impression: Why another language?
> 
> Yes, that always makes me uncomfortable too. Although, really, the original
> tuple form is *already* another language (just a very low level one).
> 
> > Can't we express the tagging tuples by Python programs?
> > Marc's tuples are nothing more than predefined actions,
> > conditions, and jumps to labels.
> >
> > Given a restricted subset of Python, and a couple of known
> > functions which are to be used, the tags could be expressed
> > as a small number of Python functions, and
> > it would be a real application for the ByteCodeHacks
> > to turn these Python functions back into tagging tables.
> > We would go without any parsing since Python does it for us.
> 
> Strangely enough, I have had exactly this thought myself.

That's good to hear.
I have often the wish to write something in a python-ish style,
without actually generating Python code but something else.

> I guess the reply is in two parts:
> 
> 1. I wanted to keep this to something simple that I could have a hope of
> implementing myself. In the very, very restricted spare time I have (we have
> two small children!) I *might* just be able to use this little language as a
> way into learning mxTextTools and produce something I can write quicker than
> the tuples themselves. Simple text transforms to produce a "mini-language"
> is something I've done before and can give back seemingly amazing progress
> for little effort.

Now, that's a good thing. Can you express the parser for the little
language with the tagging engine? That would be a funny bootstrap :-)

> 2. I can't help feeling that "jump-to-label" (aka
> skip-forwards/backwards-N-statements) is intrinsically necessary in such an
> engine, and I can't see any way of doing that sensible whilst keeping to
> Python syntax.

if target_of_jump_is("thisone"):
   # do this
if target_of_jump_is("thatone"):
   # do that

... and so on.
This is the syntax which I use for my computed goto example with
continuations. Man, this is all so similar. Where does the yourney go.

> 3. I rather wanted not to hack Python itself, and to keep independent of
> whether one is using CPython, JPython (which OK, can't use the current
> mxTextTools anyway) or some other language accessing the DLL. Perhaps a bit
> spurious, this one.

Yes, and it's number 3 already.

> (BTW, there is no comfy chair)
> 
> Oh, I just thought of:

Is it allowed to go until 4 in this joke? I don't remember :-)

> 4. Also, for me, I think that the way to go from the original tuple-form to
> (perhaps) a final form as Python programs would be via something I can think
> in more easily than the original tuples - so perhaps in the end this "little
> language" will be no more than a half-way house.
> 
> Having said all of the above, I am definitely not averse to someone else
> coming up with something better, or explaining to me why what I am doing is
> the hard way to do it and, look, here is a lazier way...

No, feel encouraged to continue! I was just blathering about what
came to mind.

> > Whoops - I could make the tagging engine into a pluggable
> > interpreter for stackless Python, this is just another engine.
> 
> Please feel free - you do that and I will happily use it...

Isn't a tagging engine a number of comparisons, head
movements, and gotos? Continuations could do that.

> > yes-I-have-to-duck-and-cover-ly y'rs - chris :-)
> 
> Oh, I don't know, it seemed sensible to me (hint: not a recommendation!)

kids-don't-try-that-at-home-ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From tim_one at email.msn.com  Thu Jul 22 23:08:13 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 22 Jul 1999 23:08:13 -0400
Subject: Maintainability (was Re: Stackless & String-processing)
In-Reply-To: <7n7bo8$204@dfw-ixnews19.ix.netcom.com>
Message-ID: <000501bed4b8$9a42e9a0$2c2d2399@tim>

[Aahz Maruch, hijacks a thread & holds it hostage to "import *"]

[Skip Montanaro, regurgitates the company line -- as if that ever stopped a
terrorist]
> ...
> There's nothing wrong with "from foo import *" when it's used in
> the right context.  Some modules in the standard distribution were
> designed specifically to support this (types and Tkinter come to mind).

[Aahz counterstrikes]
> I'm afraid I can't agree that there's *ever* a good time to use "from
> foo import *" in a multi-programmer environment. [followed by the usual
> good reasons]

Political power flows from the barrel of a gun; a language flowers on the
rotting corpse of its creator.  Come the revolution, use of "from foo import
*" will be punishable by exile to Perland!  For that matter, so will use of
the wretched "foo" <wink>.

It's still a valuable feature of the language, though, and the Official
History here is a tad revisionist.  This business about modules being
designed specifically for use with * is a latter-day invention that popped
up around the time Tkinter.py appeared; in the beginning, "import *" served
its One True Purpose:  to save a pile of tedious typing in interactive mode!
Try crunching some numbers interactively with and without "from math import
*", and the intended use of this construct shines.

The other <ahem> One True Purpose was to save a pile of tedious typing when
whipping up a throwaway prototype as quickly as possible.

Their may have been another, but all Sole True "*" Purposes are
characterized by ephemerality and solitude.  Anyone using "import *" in
public, or over extended periods of time, may as well just admit that they
didn't even bother to ask the name of the last cheap hooker they paid for a
5-minute fling.

explicit-import-is-the-condom-of-responsible-programming-ly y'rs  - tim





From billtut at microsoft.com  Tue Jul 13 07:11:18 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Tue, 13 Jul 1999 11:11:18 GMT
Subject: making a Windows NT service out of a python program
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F20@RED-MSG-50>

Take a look at the stuff in win32\demos\service\ of your friendly latest
Win32 extension install tree.

It shows you how to wrap Python code into an NT service quite easily.

The code in win32\lib\win32serivceutil.py appears to do almost everything
you would ever want to do with a service. Very cool stuff...

No, I didn't write it, it's all Mark's fault... :)

Bill



From jbauer at rubic.com  Thu Jul  8 21:02:16 1999
From: jbauer at rubic.com (Jeff Bauer)
Date: Fri, 9 Jul 1999 01:02:16 GMT
Subject: os.kill on NT
References: <3784B57E.C0A57A@cern.ch>
		<3784C95D.45E5E617@rubic.com> <7m39jh$qmu$1@m2.c2.telstra-mm.net.au>
Message-ID: <37854A18.908569C0@rubic.com>

Okay, modified to Mark Hammond's stylistic and pedantically 
correct <0.5 wink> specifications. 

-Jeff

Mark Hammond wrote:
>> http://www.python.org/cgi-bin/todo.py?req=show&file=faq13.005.htp
> 
> I hadnt seen this before.  One style nit:  Im trying to get to the point
> where "pywintypes" really is never imported directly.  My "recommended"
> style is to catch "win32api.error" in that case, rather than
> pywintypes.error.  They are indentical - ie, pywintypes.error is
> win32api.error, and neither will go away, so really a tiny, tiny point.



From tim_one at email.msn.com  Thu Jul  1 11:53:56 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 1 Jul 1999 11:53:56 -0400
Subject: How to obtain an instance's name at runtime?
In-Reply-To: <7lfjb9$aun$1@nnrp1.deja.com>
Message-ID: <000801bec3d9$ecdd53c0$5d9e2299@tim>

[Dinu C. Gherman]
> ...
> Just that much: an object, that has not been assigned
> a name to, cannot exist (having a reference count of
> zero), can it (at least for user-defined objects)?

There must be a way to "get to" it, but access paths needn't be names; e.g.,

week = [Day("Sunday"), Day("Monday"), ...]

The object constructed by Day("Monday") was never given a name, and is
accessible only via the access path week[1] (or equivalent from the end of
the list).

anonymously y'rs  - tim






From fredrik at pythonware.com  Fri Jul  2 03:57:43 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 2 Jul 1999 07:57:43 GMT
Subject: wddx wrapper?
References: <%0Qe3.4034$lr3.164499@news2.jacksonville.net>
Message-ID: <00a201bec460$91ab74f0$f29b12c2@secret.pythonware.com>

Hank Fay <hank at prosysplus.com>:
> Does anyone know a reference to a Python wrapper for wddx (www.wddx.org)?
> It looks like a neat way to pass data around (being able to transport
> recordsets between differnet languages, in particular), and I'd like to be
> able to have Python doing some of the heavy lifting on the server side.

no, but if the protocol doesn't matter, as long as it's
light-weight, cross-platform, XML, and supported by
lots of languages, you might wish to check out:

    http://www.xmlrpc.com

implementations available for python, perl, java, tcl, etc.
and zope supports it too, btw.

</F>




From tim_one at email.msn.com  Sat Jul 31 01:16:00 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 31 Jul 1999 01:16:00 -0400
Subject: NatLink v001:  Talk to your Python!
Message-ID: <000601bedb13$c72058c0$a32d2399@tim>

Apologies if you've seen this before:  a coworker posted it to the python
and announce groups, but I never saw it come across on their mailing-list
versions.  So I'm mailing it to both.

Original msg, w/ HTML citation appended:


     NatLink v001 by Joel Gould

I would like to announce the availability of a package which allows Python
programs to use large vocabulary speech recognition by interfacing to Dragon
NaturallySpeaking.

NatLink is a Python extension module which allows direct access and control
of Dragon NaturallySpeaking.  Dragon NaturallySpeaking is the best selling
large vocabulary (i.e. dictation) speech recognition program commercially
available.

NatLink requires that you already have a copy of Dragon NaturallySpeaking
(sorry).  It also only runs under Windows 95/98/NT (sorry again).

But with NatLink, you can:

(1) add speech recognition to your Python programs
(2) control Dragon NaturallySpeaking from Python
(3) write voice macros in Python for Dragon NaturallySpeaking

For more details, visit:
http://www.synapseadaptive.com/joel/natlink.htm

Note: NatLink is not a product but a freely downloadable package.

Joel Gould
joelg at alum.mit.edu

<P><A HREF="http://www.synapseadaptive.com/joel/natlink.htm">NatLink
v001</A> - speech recognition via Dragon NaturallySpeaking.  (31-Jul-99)





From mail at mikemccandless.com  Wed Jul  7 13:55:29 1999
From: mail at mikemccandless.com (Michael McCandless)
Date: Wed, 07 Jul 1999 13:55:29 -0400
Subject: scoping assertion
Message-ID: <37839491.3F7C1F6@mikemccandless.com>

I wonder if Python could be extended to allow a prefix in front of
variable names (lvalues?) to assert that the name did not previously
exist in this scope.  EG, call it "nonexist":

  nonexist x = 0
  while x < 10:
    ...

Or

  for nonexist x in xrange(1, 1000):
    ...

If x had previously existed in this scope at the time of assignment, an
appropriate exception would be raised.

Jim Hugunin pointed out that this could be achieved as follows:

  assert not locals().has_key('x')

But because there are no macros in Python, and because you can't make
this a function (?), I think you'd need to type this out every time.

The reason I ask is because I've had some sneaky bugs whereby I assigned
to what I thought was a new variable name, but inadvertantly wiped out a
value that I later used.

Statically typed languages basically give you this "for free" ;),
meaning if you try to "declare" the variable again, the
compiler/interpreter will typically generate an error saying the
variable is already defined in this scope.

Mike



From aahz at netcom.com  Sat Jul 10 20:41:33 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Jul 1999 00:41:33 GMT
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787D8DA.3774AA7D@ssec.wisc.edu>
Message-ID: <7m8p7t$o7s@dfw-ixnews10.ix.netcom.com>

In article <3787D8DA.3774AA7D at ssec.wisc.edu>,
Nick Bower  <nick.bower at ssec.wisc.edu> wrote:
>
>Incidentally, I went back to the library reference and still couldn't find
>os.chdir

You probably have an old version of the docs; I suggest you go to
www.python.org and download a new set.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From esr at snark.thyrsus.com  Thu Jul 29 16:32:34 1999
From: esr at snark.thyrsus.com (Eric S. Raymond)
Date: 29 Jul 1999 20:32:34 GMT
Subject: Need help with TCP/IP client access from Windows
Message-ID: <slrn7q1ej3.18d.esr@snark.thyrsus.com>

I need a Windows equivalent of this class for a client I'm writing.

class SessionConnection:
    # Encapsulate the line-by-line level of a session with a server.
    # This class requires Unix services.
    param_names = ('host', 'port', 'username', 'password', 'browse')

    def __init__(self, file=".intriguer"):
        # Initialize a SessionConnection object from user's stored paramters
        self.dotfile = os.path.join(os.environ["HOME"], file)
        self.host = "snark.thyrsus.com"
        self.port = 99999
        self.username = "esr at thyrsus.com"
        self.password = "fooup"
        self.browse = "netscape -remote 'openURL(%s)'"
        try:
            dfp = open(self.dotfile);
            lexer = shlex.shlex(dfp)
            while 1:
                variable = lexer.get_token()
                if not variable:
                    break
                if not variable in  Session.param_names:
                    raise SyntaxError
                if lexer.get_token() != '=':
                    raise SyntaxError
                value = lexer.get_token()
                if not value:
                    raise SyntaxError
                setattr(self, variable, value);
            dfp.close()
        except IOError:
            pass	# if file is absent, fall back to defaults.
        self.socket = None
        self.sockfp = None

    def __repr__(self):
        # Dump current session parameters
        res = "# intriguer configuration\n"
        for p in Session.param_names:
            if getattr(self, p):
                res = res + "%s=%s\n" % (p, `getattr(self, p)`)
        return res

    def wrapup(self):
        # Update the dotfile to reflect parameter edits
        dfp = open(self.dotfile, "w")
        dfp.write(`self`);
        dfp.close()

    # Establish and break server connections

    def connect(self):
        # Initialize a server connection.
        # Not part of the class initialization sequence only because
        # we might want to create an instance just to examine or
        # edit the parameters.
        self.socket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
        self.socket.connect(self.host, self.port)
        self.sockfp = self.socket.makefile("r+")
        
    def disconnect(self):
        # Terminate a server connection
        if self.socket:
            self.socket.close()
            self.sockfp.close()

    # Communication with server

    def readline(self):
        # Read a raw line from the server
        return self.sockfp.readline()

    def writeline(self, line):
        # Write a raw line to the server
        return self.sockfp.write(line + "\n")

    # Local service invocations

    def browse_url(self, url):
        # Point a local browser at a specified URL 
        os.system(self.browse % url)

The issues here are:

(1) How do I get (at least) line-oriented access to an Internet server
    from a Python program running under Windows?

(2) How does one read and write the Windows equivalent of a dotfile?
    Must this involve the (shudder) registry?

(3) Is there a reliable way for Python programs under Windows to point a local
    browser at a specified URL?
-- 
		<a href="http://www.tuxedo.org/~esr">Eric S. Raymond</a>



From gherman at my-deja.com  Thu Jul  1 07:33:52 1999
From: gherman at my-deja.com (Dinu C. Gherman)
Date: Thu, 01 Jul 1999 11:33:52 GMT
Subject: TopSort in Python?
Message-ID: <7lfjn0$b1v$1@nnrp1.deja.com>

Does anybody have a simple-minded-but-working full-Python
implementation of topsort, the topological sorting algorithm?
Or maybe *any* topsort? I remember only one such algorithm...
python.org/search doesn't reveal any...

Thanks,

Dinu


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gregm at iname.com  Sun Jul  4 08:49:12 1999
From: gregm at iname.com (Greg McFarlane)
Date: Sun, 4 Jul 1999 12:49:12 GMT
Subject: Upgrade or not ??
In-Reply-To: <377E0A3B.CA3B258C@worldnet.fr>; from pierrette on 03 Jul 1999 at 03:04:00PM
References: <377E0A3B.CA3B258C@worldnet.fr>
Message-ID: <19990704224912.16776@nms.otc.com.au>

Depending on exactly what you want to do, it may be possible in
Tkinter.  You can use something like this to find out when a widget
gets and loses focus:

    import Tkinter

    def gotFocus(event):
	print 'Focus in'
    def lostFocus(event):
	print 'Focus out'

    root = Tkinter.Tk()

    e = Tkinter.Entry()
    e.bind('<FocusIn>', gotFocus)
    e.bind('<FocusOut>', lostFocus)
    e.pack()

    root.mainloop()

Also, you can use this to set the focus to a widget.

    e.focus_set()

On 3 Jul, pierrette wrote:
> First a general question :
> Actually we're working with Python 1.5.1 and we're satisfied with it.
> Should we upgrade to 1.5.2 and if yes WHY?
> 
> No a Tkinter guestion :
> We developped a prototype of an application with python an tkinter, now
> we have to
> decide if we're gonna write the real app in python or something else !
> We now allready that in principle we could do it with python, the thing
> we're missing
> in tk is a 'on get focus' ,a 'on lost focus' event and a 'set focus to'
> command
> Pe: on lost focus of inputfield 1:
>             if inputfield1 = A:
>                 set focus to field2
>             elif inputfield1 = B:
>                 set focus to field 3
> 
> If those things are not available in tkinter what alternatives do we
> have ?
> - Changing the GUI (would be a shame but feasable), knowing that is has
> to be multi-platform WIN-LINUX? If so to what ?
> - Not write in python ? In this case we lose the work done for the
> prototype
> 
> Thanks for comments and remarks
> 
> Pierrette Serrand
> pserrand at worldnet.fr
> 
> 
> 
> 
> 

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at nms.otc.telstra.com.au)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From tgabriel at bellsouth.net  Mon Jul 19 06:57:52 1999
From: tgabriel at bellsouth.net (Xenophanes)
Date: Mon, 19 Jul 1999 06:57:52 -0400
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net> <378FE5F3.A8394D50@bellsouth.net> <3790FA96.9403AA17@ssec.wisc.edu>
Message-ID: <379304B0.13D316F4@bellsouth.net>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19990719/91526899/attachment.html>

From eugene.leitl at lrz.uni-muenchen.de  Thu Jul  1 20:08:11 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Thu,  1 Jul 1999 17:08:11 -0700 (PDT)
Subject: Why is tcl broken?
In-Reply-To: <86pv2cwco4.fsf@g.pet.cam.ac.uk>
References: <1283099377-77267484@hypernet.com>
	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com>
	<375FDCA7.62FF@mailserver.hursley.ibm.com>
	<7josm1$bgc$1@Starbase.NeoSoft.COM>
	<37628EBF.207F7A8F@iname.com>
	<37682DD3.43923243@compaq.com>
	<7ka2be$qja$1@nntp6.u.washington.edu>
	<37776FFA.2D85@mailserver.hursley.ibm.com>
	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it>
	<876748mcxg.fsf@home.ivm.de>
	<37788BB9.ABD@mailserver.hursley.ibm.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.ibm.com>
	<lwoghyexy5.fsf@copernico.parades.rm.cnr.it>
	<377B454C.6F59@mailserver.hursley.ibm.com>
	<377B593B.F7F7504B@iname.com>
	<377B8CFE.6113@mailserver.hursley.ibm.com>
	<86pv2cwco4.fsf@g.pet.cam.ac.uk>
Message-ID: <14204.686.875737.910962@lrz.de>

Gareth McCaughan writes:

 > Emacs Lisp, however, is not a recent Lisp.

There are plans to convert XEmacs to use a Scheme engine instead of elisp.



From pserrand at p98.f112.n480.z2.fidonet.org  Fri Jul  2 10:21:02 1999
From: pserrand at p98.f112.n480.z2.fidonet.org (pserrand)
Date: Fri, 02 Jul 1999 15:21:02 +0100
Subject: newbie
Message-ID: <0000165f@bossar.com.pl>

From: pserrand <bucodi at worldnet.fr>

cyberescue wrote:

> I am new to programming. Where can I get a copy of Python?
>
> mailto: ashfordv at usa.net

www.python.org



From robin at jessikat.demon.co.uk  Thu Jul 15 14:20:05 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 15 Jul 1999 19:20:05 +0100
Subject: VC++ 6.0 exe file sizes
Message-ID: <o$h7sBAVZij3Ew7i@jessikat.demon.co.uk>

Any VC++ gurus out there help with a problem I'm getting building the
python.exe file with VC++ 6.0 & win95 osr2. My hd has 4k clusters.

No matter what optimisations etc I do the python.exe seems to come out
as 16384. I reckon it should be around 5k like the 1.5.2 distributed
version (which I think is built using VC++ 5.x).

I see this particular problem, but what about all the other bits of exe
code etc?
-- 
Robin Becker



From jonathan at onegoodidea.com  Fri Jul 30 10:31:00 1999
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Fri, 30 Jul 99 14:31:00 GMT
Subject: RTSP streaming media proxy
Message-ID: <mt2.0-24150-933345060@news.informatik.uni-bonn.de>

hot on the heels of my last post, here is my next contribution. again,
it can be found on the projects page of our company web-site:

   <http://www.onegoodidea.com/projectscore.html>

RTSP_Proxy is a proxy server for the RTSP/RTP streaming media protocols.
this is a quick-and-dirty hack i threw together a couple of weekends ago
in order to be able to watch the MacWorld keynote address which was
netcast using QuickTime 4 video/audio streaming.

for information on QuickTime 4, take a look at:

   <http://www.apple.com/quicktime/>

unfortunately, whilst apple chose a non-proprietry standard protocol for
implementing their streaming architecture, they also chose one which is
not particularly established and is not IP masquerading friendly. apple
provide a proxy server for running on a firewall/NAT machine, but it's
pretty flaky so i had a stab at writing my own.

i've only run it on an i386 linux box, but i've been told it works on
freebsd too. it requires threads and sockets, but should work on anything
that provides them.

it's not a pretty program, but it worked enough for me to watch the new
iBook being launched :-)  i'm in the process of writing a new version, but
thought this might be of interest/amusement.

again, all comments welcome,

:-j

== 
jonathan hogg, one good idea limited, 25 clouston street, glasgow g20 8qr, uk
jonathan at onegoodidea.com www.onegoodidea.com t:(0)976-614338 f:(0)7970-537451

<P><A HREF="http://www.onegoodidea.com/projectscore.html">RTSP proxy</A> -
quick-and-dirty proxy server for the RTSP/RTP streaming media protocols,
written in Python.  (29-Jul-99)

-- 
----------- comp.lang.python.announce (moderated) ----------
Article Submission Address:  python-announce at python.org
Python Language Home Page:   http://www.python.org/
Python Quick Help Index:     http://www.python.org/Help.html
------------------------------------------------------------



From nick at #spam_me_not#videosystem.co.uk  Wed Jul 14 09:13:34 1999
From: nick at #spam_me_not#videosystem.co.uk (Nick Trout)
Date: Wed, 14 Jul 1999 14:13:34 +0100
Subject: wxPython Form editor
References: <378c488e@news.xtml.co.uk>
Message-ID: <378c8b3c@news.xtml.co.uk>

Is there a wxPython mail group/news group?

N

Nick Trout <nick at NOSPAMvideosystem.co.uk> wrote in message
news:378c488e at news.xtml.co.uk...
>Has anybody written a dialog/form editor for wxPython?
>
>Thanks,
>Nick.
>
>
>





From fredrik at pythonware.com  Wed Jul 21 12:21:28 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Jul 1999 16:21:28 GMT
Subject: String2List: odd function
References: <199907211304.IAA14937@cro.ots.ac.cr>
Message-ID: <001701bed395$16de7c60$f29b12c2@secret.pythonware.com>

Isidor <rodisi01 at my-deja.com> wrote:
> Takes a list-like string, e.g.,:
> 
> "[['itema1', 'itema2'], 'itemb1', ['cheese1', ['cheese1a',
> 'cheese1b'], '2'], ''blah' rest of itemb2', 'zib']"
> 
> and returns *real* lists:
> 
> [['itema1', 'itema2'], 'itemb1', ['cheese1', ['cheese1a',
> 'cheese1b'], '2'], "'blah' rest of itemb2", 'zib']
> 
> Which, in text-world, looks like nothing more than the removal of
> some quotation marks, but in Python-world...well, I don't have to
> tell you what the difference is.

v = eval(s, {"__builtins__": {}})

</F>




From arcege at shore.net  Sun Jul 18 08:45:52 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Sun, 18 Jul 1999 12:45:52 GMT
Subject: upload file with cgi
References: <7mro7g$kr5@newsops.execpc.com>
Message-ID: <40kk3.351$ui4.121238@news.shore.net>

Sim & Golda Zacks <simngolda at voyager.net> wrote:
: I am uploading a file using the cgi module and I can not figure out how
: to get the name of the file.
: The code I am using is
: form=cgi.FieldStorage()
: pic=form["picture"]
: Now when I print pic it gives me the name of the File  and the actual
: file.
: When I get pic.value it is the actual picture.
: pic.key doesn't give me anything and I have absolutely no clue how to
: get the filename.

: Please help.

There should be a "filename" attribute along with the "file" object.
  print 'You gave me the file:', pic.filename

  -Arcege




From jepler at inetnebr.com  Thu Jul  8 22:02:43 1999
From: jepler at inetnebr.com (Jeff Epler)
Date: Fri, 09 Jul 1999 02:02:43 GMT
Subject: Uses for unary + (was Re: New Python language abuse: ++x with bytecodehacks)
References: <slrn7o9625.tk.jepler@craie.inetnebr.com> <3784C529.78FD22D2@appliedbiometrics.com>
Message-ID: <slrn7oam9a.1it.jepler@craie.inetnebr.com>

On Thu, 8 Jul 1999 15:35:05 GMT, Christian Tismer
 <tismer at appliedbiometrics.com> wrote:
>p.s.: I wondered why Guido didn't optimize a double
>unary positive. But he can't, since he has no idea what
>this operator does.

Because nobody has a reason to type it?

I also overloaded the meaning of unary + when I was working with classes
that acted like numeric types.  ref[:] is a nice way to do a shallow copy
of sequence types, and "+ref" is a nice way to denote a shallow copy of a
numeric type.  (You just have to write the method __pos__ to do so)

Jeff
PS Too bad that dictionaries don't support dict[:] to make a copy too!
-- 
\/ http://incolor.inetnebr.com/jepler/           Jeff Epler jepler at inetnebr.com
"But I don't like Spam!!!!"



From Greg.Ewing at p98.f112.n480.z2.fidonet.org  Fri Jul  2 17:14:27 1999
From: Greg.Ewing at p98.f112.n480.z2.fidonet.org (Greg Ewing)
Date: Fri, 02 Jul 1999 22:14:27 +0100
Subject: total idiot question: +=, .=, etc...
Message-ID: <0000163e@bossar.com.pl>

From: Greg Ewing <greg.ewing at compaq.com>

Bernhard Herzog wrote:
> 
>         .last = .lines[-1]
>         .lines = .lines[:-1]

Hey, I've just had an idea: allow an *empty*
name as an identifier. Then we could write:

  class C:
    def foo(,x,y):
      .z = .z + x * y

and kill two birds with one syntax change!

Greg



From garry at research.att.com  Wed Jul 14 13:42:10 1999
From: garry at research.att.com (Garrett G. Hodgson)
Date: Wed, 14 Jul 1999 17:42:10 GMT
Subject: Running 2 gui libs simultaneously
References: <7mf2rn$dj9$1@news1.saix.net>
Message-ID: <378CCBF2.B199281C@research.att.com>

Riaan Booysen wrote:

> How difficult would it be to run 2 window libraries simultaneously (Tcl/Tk /
> wxPython) in one Python program.
> The problem is that both need an event processing loop ( lib.mainloop() ).
> Are threads my only solution?

why would you want to do that?
perhaps an explanation of the problem you're
trying to solve would invoke some better solutions.


-- 
Garry Hodgson			comes a time
garry at sage.att.com		when the blind man
Software Innovation Services	takes your hand, says,
AT&T Labs			"don't you see?"



From mwh21 at cam.ac.uk  Tue Jul 20 07:56:21 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: Tue, 20 Jul 1999 12:56:21 +0100
Subject: Python and Boehm-Demers GC, I have code.
In-Reply-To: <p5k8rvmymj.fsf@bidra241.bbn.hp.com>
References: <000901becffa$b0f25e00$91a22299@tim> <p5k8rvmymj.fsf@bidra241.bbn.hp.com>
Message-ID: <Pine.LNX.4.10.9907201250340.6803-100000@atrus.jesus.cam.ac.uk>

On 20 Jul 1999, Markus Kohler wrote:
[snippety snip]
> 
> Is there anything that would prevent someone to reimplement loops such
> that only one variable is allocated ?
> 

Yes. The fact that ints are immutable. Consider:

for i in xrange(10):
    if i == 2:
        j = i

if the loop only allocated one variable, then at the end of the loop
'print j' would give the answer '9'!

I guess this could be worked around by inspecting refcounts and such...
but not easily.

there's-the-small-integer-cache-to-worry-about-too-ly y'rs
Michael





From gmcm at hypernet.com  Thu Jul  8 23:15:40 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 8 Jul 1999 22:15:40 -0500
Subject: Python/PythonWin and DDE?
In-Reply-To: <3785515c.0@news.dbtech.net>
Message-ID: <1280636026-77839847@hypernet.com>

Karl Putland writes:
> 
> Mark Hammond <MHammond at skippinet.com.au> wrote in message
> news:7m39vq$r5o$1 at m2.c2.telstra-mm.net.au...

> > Yes.  Pythonwin acts as both a client as a server as it communicates between
> > processes.
> 
> Where to start?  Does PythonWin's DDE usage follow that found in M$
> VC++?  If so those docs are weak.  I only have examples in VB and
> VBA.  Where would be the best place to start with PythonWin and DDE?

Here's some code I got working somehow. I think from Pegusas Mail's 
DDE docs, the VB examples:

    server = dde.CreateServer()
    server.Create("ME")  
    conversation = dde.CreateConversation(server)
    try:
        conversation.ConnectTo("WinPMail", "TCP")
    except:
        win32api.WinExec("C:\\PMAIL\\winpm-32.exe -Z 8192")
        conversation.ConnectTo("WinPMail", "TCP")
    try:
        conversation.Poke("TCP", "Get")
    except:
        print curtime(), "Poke failed. Maybe a timeout "  
        time.sleep(600)

- Gordon



From Andy.L.Krietemeyer at p98.f112.n480.z2.fidonet.org  Fri Jul  2 16:05:15 1999
From: Andy.L.Krietemeyer at p98.f112.n480.z2.fidonet.org (Andy L Krietemeyer)
Date: Fri, 02 Jul 1999 21:05:15 +0100
Subject: Dumb beginner question; what is IDLE?
Message-ID: <000016a8@bossar.com.pl>

From: Andy L Krietemeyer <afn08995 at afn.org>

I've seen several references to IDLE.  Is this an IDE for python? Or
what is it?  Thanks,  Andy.



From martin.lemburg at bigfoot.de  Sun Jul 11 07:03:33 1999
From: martin.lemburg at bigfoot.de (martin.lemburg at bigfoot.de)
Date: Sun, 11 Jul 1999 13:03:33 +0200
Subject: python, tkInter -> tcl/tk problems
Message-ID: <7m9tid$2sl$1@rzsun02.rrz.uni-hamburg.de>

Hi,

I would be glad to get some help belonging to the use of tkInter with python 1.5.2!

The situation:

1. I already have tcl/tk 8.05 installed
2. tcl/tk is inside the path of WindowsNT
3. starting idle returns following error message: 
>Traceback (innermost last):
>  File "E:\programming\Python\Tools\idle\idle.py", line 3, in ?
>    PyShell.main()
>  File "E:\programming\Python\Tools\idle\PyShell.py", line 611, in main
>    root = Tk()
>  File "E:\programming\Python\Lib\lib-tk\Tkinter.py", line 893, in __init__
>    self.tk = _tkinter.create(screenName, baseName, className)
>TclError: Can't find a usable init.tcl in the following directories:
>    {} ./lib/tcl8.0 E:/programming/Python/tcl8.0/library E:/programming/Python/Tools/library

My tries to solve the problem:

1. I copied the lib/tcl8.0 directory from my tcl/tk installation into the python directory:
    <tcl/tk>/lib/tcl8.0    =>    <python>/tcl8.0/library
2. I copied the lib/tk8.0 directory from my tcl/tk installation into the python directory:
    <tcl/tk>/lib/tk8.0    =>    <python>/tk8.0/library

New situation:

1. starting idle returns following error message:
>Traceback (innermost last):
>  File "E:\programming\Python\Tools\idle\idle.py", line 3, in ?
>    PyShell.main()
>  File "E:\programming\Python\Tools\idle\PyShell.py", line 611, in main
>    root = Tk()
>  File "E:\programming\Python\Lib\lib-tk\Tkinter.py", line 893, in __init__
>    self.tk = _tkinter.create(screenName, baseName, className)
>TclError: Can't find a usable tk.tcl in the following directories:
>    . E:/programming/Python/tcl8.0/tk8.0 ./tk8.0 ./lib/tk8.0 ./library ./tk8.0/library
> ./library E:/programming/Python/tk8.0/library
>
>E:/programming/Python/tk8.0/library/tk.tcl: bad event type or keysym "MouseWheel"
>bad event type or keysym "MouseWheel"
>    while executing
>"bind Listbox <MouseWheel> {
>    %W yview scroll [expr - (%D / 120) * 4] units
>}"
>    (file "E:/programming/Python/tk8.0/library/listbox.tcl" line 179)
>    invoked from within
>"source [file join $tk_library listbox.tcl]"
>    (file "E:/programming/Python/tk8.0/library/tk.tcl" line 151)
>    invoked from within
>"source E:/programming/Python/tk8.0/library/tk.tcl"
>    ("uplevel" body line 1)
>    invoked from within
>"uplevel #0 [list source $tkfile]"

What know?

The tcl/tk-version delivered by python 1.5.2 is 8.0.5 too, so what goes wrong?
Why does python/tkInter looks for tcl/tk in non tcl/tk-standard directories?
What could I do to activate tkInter to use it with python 1.5.2?

I would be glad reading something from you - but it would be nice to email me your answers too!

Bye

Martin Lemburg

____________________________________________

Martin Lemburg
Bielfeldtstra?e 12 - D-22763 Hamburg
Tel./Ph.   +49 (40)  8803486
Fax.        +49 (441) 800 990 6462
eMail.    martin.lemburg at bigfoot.de





From rodisi01 at my-deja.com  Wed Jul 21 09:53:53 1999
From: rodisi01 at my-deja.com (Isidor)
Date: Wed, 21 Jul 1999 07:53:53 -0600
Subject: IDLE mini-hack & ?
Message-ID: <199907211257.HAA14822@cro.ots.ac.cr>

Hello Pythonates -

I like IDLE. It's very nice. Thank you GvR.

I am working on learning Tkinter, so I thought, hey, why not look
at this cool app sitting in front of me called IDLE? (Can anyone
say "in over your head"?). (All of the following material refers
to the IDLE that came with 1.5.2 win32 binary.)

Here's a little something I did that someone else might like to
do as well: add a "start up date/time stamp" after the copyright
notice.

1. Start-up date/time stamp

In PyShell.py (your python-directory|Tools|idle), line 371 is
where the "begin" function of class PyShell(OutputWindow) is
defined. A few lines down, we find:

self.write("Python %s on %s\n%s\n" %
                   (sys.version, sys.platform, sys.copyright))

which does something pretty obvious. So, to add a startup
date/time stamp, I added the following line after the "Python
ver, platform, copyright" line:

self.write("Start time: " + time.strftime("%Y%m%d.%H%M",
time.localtime(time.time())))

which prints the date and time like so: "19990719.1422". You may
not like that format (probably don't), but you can read the docs
on the time module to find out about all the other formatting
options.

Of course, for this to work, you need to import the time module
somewhere in PyShell, either at the top of the module, at the
start of the class, or the start of the function. I suspect that
the first option is the best, but I'm not sure.

2.     Adding a command to a menu

Well, I figured, if I can add a date/time stamp there, why not
make it a command that can be executed anytime (in comments or
where ever.) Yeah, right. This is where I got really lost. I came
close...I got stuff to appear on the menu drop down list, but
when I clicked on it, all hell would break loose (actually, an
exception was raised):

>>> Exception in Tkinter callback
Traceback (innermost last):
  File "C:\Program Files\Python152\Lib\lib-tk\Tkinter.py", line
764, in __call__
    return apply(self.func, args)
TypeError: no arguments expected

Ok, what did I do to make this mess (in a second copy of the idle
directory):

A. To Bindings.py

in the menudefs dictionary, at the end of the list of the edit
tuple, I added:

('_DateTime', "<<date-time-stamp>>")

B. To EditorWindow.py

i. at the beginning, where all the "#$ [menu definitions]"
reside, I added

#$ event <<date-time-stamp>>
#$ win <Control-Shift-T>
#$ unix <Control-Shift-T>

to the end of the edit menu.

ii. in the beginning of the __init__ function of the class
EditorWindow, I added the line that follows the remark

text.bind("<<remove-selection>>", self.remove_selection)
##Next 1 line added by IFR 19990716:0212
text.bind("<<date-time-stamp>>", self.date_time_stamp)
text.bind("<3>", self.right_menu_event)


iii. later on, near the bottom of class EditorWindow, I added the
following function. The commented-out parts are different things
I've tried unsuccessfully.

##def date_time_stamp(self):
def date_time_stamp():
        ##from OutputWindow import OutputWindow
        ##import time
        ##OutputWindow.write("##DateTime: " +
time.strftime("%Y%m%d:%H%M", time.localtime(time.time())))
        ##print "cheese"
     ##self.write("cheese")
     tkMessageBox.showerror(
                "blah",
                "blah2",
                master=self.text)
     self.text.write("cheese")

Basically, I'm really clueless and I'm banging around inside IDLE
like a ...like a...madman at a cocktail party (?). I'm sure that
if I got a decent night's sleep and followed all the function
calls from one module to another and one class to another, I
could figure it out. Until that happens, though, I'd love to
receive some small clues or hints.   ;^)

Thanks in advance,

Isidor

"hacking IDLE, idly hacking"





From tismer at appliedbiometrics.com  Thu Jul 15 09:19:10 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Thu, 15 Jul 1999 13:19:10 GMT
Subject: ANN: Stackless Python 0.3
References: <378A40AD.B4FDB79F@appliedbiometrics.com> <m37lo5kpd0.fsf@atrus.jesus.cam.ac.uk> <378AFD45.8793D507@appliedbiometrics.com> <Pine.LNX.4.10.9907151242590.16482-100000@atrus.jesus.cam.ac.uk>
Message-ID: <378DDFCE.A145AC43@appliedbiometrics.com>


Michael Hudson wrote:

...

> echo "fudge" | stackless-python
> 
> does *not* print an error!
> 
> echo "fudge" | stackless-python -i
> 
> does!

Yes, thanks. This has been found yesterday, and it was an oversight
in PyRun_SimpleFile. There I call the _nr version, which actually
doesn't execute its code but returns a Py_UnwindToken.
I always felt that I've overdone here, but it was simple to solve.
I just had to repeat the errorprint code, and it works.
In other words, the program just misses to print the error,
everything else is fine. I didn't update the source since
I'm again making a huge change, Version 0.4 this week. This
one solves all continuation problems which can occour, and
the final goal (continuationmodule.c) is going to become public.

Here the patch:

pythonrun.c must be changed this way, from line 548 on:
--------------------------------------------------------
int
PyRun_SimpleFile(fp, filename)
	FILE *fp;
	char *filename;
{
	int retval;
	PyObject * result = PyRun_SimpleFile_nr(fp, filename);
	if (result == Py_UnwindToken) {
		result = PyEval_Frame_Dispatch();
		/* proper handling added 990714 */
		if (result == NULL)
			PyErr_Print();
		if (Py_FlushLine())
			PyErr_Clear();
	}
	if (result==NULL)
		retval = -1;
	else
		retval = 0;
	Py_XDECREF(result);
	return retval;
}
--------------------------------------------------------
> 
> As you can see I have built stackless python into an executable called
> stackless-python. This required extensive hackery of the build process, so
> that could be at fault. This makes even less sense to me.

Really so hard? I thought you can rename it when it's ready,
just in the install case?

> A consequence of all this is that stackless-python cannot install itself
> without error on my system (compileall.py "fails").

> PS: Did you see stackless python got a mention on Linux Weekly News Today?
> Cool.

Oh, they did it again?
What will they say when I claim full blown first
class continuations for python, next week? :-))

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From skip at mojam.com  Mon Jul 26 11:46:45 1999
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 26 Jul 1999 10:46:45 -0500 (CDT)
Subject: Python Performance
In-Reply-To: <p51zdvn4c7.fsf@bidra241.bbn.hp.com>
References: <3794C642.AB6A3E5D@freenet.co.uk>
	<7n5f0c$snu$1@ffx2nh5.news.uu.net>
	<al81zdxkktf.fsf@sirppi.helsinki.fi>
	<p51zdvn4c7.fsf@bidra241.bbn.hp.com>
Message-ID: <14236.32616.980578.536614@dolphin.mojam.com>

I got a bit confused attributing this message.  I think we have too many
Markus's running around loose... ;-)  I think I have the quoted text
attributed correctly.  Apologies if not.

    Stenberg> Yes that is much too slow. Good OO style requires a lot of
    Stenberg> function calls.  I'm sure Visualworks a commercial Smalltalk
    Stenberg> implementation can do millions of calls on this machine. Ok
    Stenberg> it's a JIT and it's commercial but's still Smalltalk and
    Stenberg> therefore at least as dynamic as python. I guess even Squeak
    Stenberg> (www.squeak.org) would be faster. If you would sent us code, I
    Stenberg> could prove it.

Better yet, since the Python source code is available, you could prove it by
speeding up the interpreter.  Many people have optimized different parts of
Python over the lifetime of the software.  It's significantly faster than it
was a couple years ago.

I think one of the impediments to improving the speed of the language has
been the dramatic performance improvement in hardware.  There's not a lot of
incentive to pour lots of time into speeding up method calls when for a few
hundred bucks you can simply drop in a faster processor, disk or networking
card.

    Kohler> P.S. This might surface as one example for 'why scripting
    Kohler> languages don't always cut it' in my master's thesis; if I have
    Kohler> made some grievous error(s?), corrections are welcome.

I don't recall anyone ever saying scripting languages would always cut it.

    Kohler> P.P.S. Too bad Python doesn't have anything that creates _nice_
    Kohler> code, like Stalin for Lisp ;-)

There is a substantially shorter history to Python than to Lisp.  Give it
time.  In addition, Lisp has been knocking around the academic world for
40-odd years.  With that many graduate students looking for a thesis topic,
it was likely that some of those theses would focus on performance
enhancement.

give-typewriters-to-enough-chimpanzees-and-you'll-get-Shakespeare-ly y'rs,

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From aahz at netcom.com  Thu Jul  8 00:18:55 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 8 Jul 1999 04:18:55 GMT
Subject: Slicing beyond list --  bug or featurette?
References: <377D234A.4C9BDAC0@cisco.com> <m3673wub7a.fsf@chinon.cnrs-orleans.fr> <37841B00.D37920A2@compaq.com>
Message-ID: <7m18rf$f36@dfw-ixnews19.ix.netcom.com>

In article <37841B00.D37920A2 at compaq.com>,
Greg Ewing  <greg.ewing at compaq.com> wrote:
>Konrad Hinsen wrote:
>> 
>>   Sequences also support slicing: a[i:j]selects all items with index k
>>   such that i <=k < j.
>
>That definition could apply to dictionaries, too! Pity it doesn't...

You could easily subclass UserDict to create this effect, I think.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From gmcm at hypernet.com  Thu Jul  8 14:57:13 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 8 Jul 1999 13:57:13 -0500
Subject: win32 python as a dll?
In-Reply-To: <3784D0CC.C8130207@enews.nrl.navy.mil>
Message-ID: <1280665923-76040383@hypernet.com>

Scott Wolford writes:

> I'm beginning to question my sanity. Try this:
> 
> #include "Python.h"
> #include <stdio.h>
> int main(int argc, char **argv)
> {
> //  return Py_Main(argc, argv);
>   FILE *file = fopen(argv[1], "rt");
>    Py_Initialize();
>   PyRun_AnyFile(file, argv[1]);
> }
> 

Getting rid of the 't' in "rt", adding a "return 0;" to get rid of 
the warning and (in the absence of error checking) making sure 
argv[1] is a valid file, it works.

- Gordon



From ovidiu at cup.hp.com  Wed Jul 28 21:35:37 1999
From: ovidiu at cup.hp.com (Ovidiu Predescu)
Date: Wed, 28 Jul 1999 18:35:37 -0700
Subject: any way to know when the program is exiting?
Message-ID: <199907290135.SAA22743@hpcll563.cup.hp.com>

Hi,

I was wondering if there is any way in a __del__ method to know when it's
invoked on the way out, as a result of the program going away?

I have a __del__ method of a class that sends a message to a global variable.
It happens that on the way out, that value in that variable is destroyed before
the __del__ methods of my instances are invoked. The result is that I have a
bogus value that doesn't respond to the usual methods anymore. How can I
prevent sending messages to that object?

A possible solution would be to use atexit() but this doesn't work if the user
program that's using my library replaces it with its own function.

Any ideas?


Thanks,
-- 
Ovidiu Predescu <ovidiu at cup.hp.com>
http://andromeda.cup.hp.com/  (inside HP's firewall only)
http://www.geocities.com/SiliconValley/Monitor/7464/





From dl231 at eng.cam.ac.uk  Thu Jul 22 05:19:15 1999
From: dl231 at eng.cam.ac.uk (D. Lopez-De-Ipina)
Date: Thu, 22 Jul 1999 10:19:15 +0100
Subject: Is anybody aware of a tree GUI widget for tkinter or Pmw?
Message-ID: <3796E213.9F326412@eng.cam.ac.uk>

Hi,

Could anybody please tell me if there exits any TREE control widget for
tkinter or Pmw? I know that Pmw provides a rather rich set of
'megawidgets' but the tree control has not been implemented yet as far
as I know.

Thanks a lot,

Diego




From cgw at fnal.gov  Wed Jul 28 00:30:42 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Wed, 28 Jul 1999 04:30:42 GMT
Subject: threading question
Message-ID: <199907280430.XAA17615@buffalo.fnal.gov>

When I run the trivial example attached below, which simply
starts a daemon thread that does nothing, and then type "ps",
I see *three* python processes.  I can understand two, but
why three?

Platform is Linux 2.2.10 with the most recent glibc (2.1.1),
if that matters.

Thanks in advance for any pointers.  Or, since this is Python,
better make that "references".

from threading import *        
import time

class Test:

    def __init__(self):
        self.test = Thread(group=None,
                               target=self.thread_fn,
                               name="thread1",
                               )
        self.test.setDaemon(1)
        self.test.start()

    def thread_fn(self):
        while 1:
            time.sleep(1)


if __name__ == "__main__":

    x = Test()
    while 1:
        time.sleep(1)
        




From MHammond at skippinet.com.au  Mon Jul 26 20:16:46 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Tue, 27 Jul 1999 10:16:46 +1000
Subject: Memory leaks whilst embedding python
References: <1279335510-28879097@hypernet.com> <7nclk9$sg1$1@news.clarkson.edu>
Message-ID: <7nitl6$q2r$1@m2.c2.telstra-mm.net.au>

Brad Clements wrote in message <7nclk9$sg1$1 at news.clarkson.edu>...

>This, however, is going to be a giant problem for me.

Its a little problem for a few of us too.

>Because Python doesn't finalize modules, I can't clean up this memory...

This has been discussed before, and there seems to be general agreement
Python _should_ take some steps to finalize modules.

Further, although PyInit/PyFinalize leaks "static" memory, I believe
(although have not confirmed) that repeatedly calling this pair _will_
leak - although it is uncommon for apps to need to do this, it can happen.

>I'm going to have to write my own memory allocator and fake out Python and
>all add-on modules to use it, what a major pain.

Well, I would like to suggest that you help with getting Python to clear
this stuff up.  I dont think there is anything seriously hard about the
cleanup, other than a general feeling that this memory is "not a problem".
Still quite a pain, but then your efforts help mankind :-)

I dont think anyone would complain about a Python that freed up every single
memory allocation it made.  Someone just has to do it.

Re the module finalization:  Maybe a proposal could be put forward for a
module to optionally register a "module term" function (by passing a pointer
to a function at module init time - Python could stash this away somewhere,
calling it as the module is unloaded).  This would be backward compatible,
and allow modules to cleanup all their allocations - something that is
non-trivial and hacky for an extension module to do today.

You would probably find a number of people would be interested in this, and
be willing to help out with the tricky stuff once you (or someone!) takes
the lead.  I know Pythonwin and PythonCOM could benefit, and believe from
recent discussions wxPython could too.  Im sure the list is quite large.
Unfortunately for you, it seems you are the only one for which it is a
"giant problem"

Mark.






From algaba at my-deja.com  Sat Jul 10 06:04:24 1999
From: algaba at my-deja.com (algaba at my-deja.com)
Date: Sat, 10 Jul 1999 10:04:24 GMT
Subject: Again, small utility docSor
Message-ID: <7m9tnj$6qg$61@rks1.urz.tu-dresden.de>

Sorry if you get this twice !!

I've done this:

"""
1998 Manuel Gutierrez Algaba
You are free to use , modify, distribute and copy this piece of
python code, if you keep this free copyright notice in it.

arbor.py Version 1.0

It makes trees of anything you want to. Class depencencies, calls...

For more related doc py programs look at:
http:\\www.ctv.es\USERS\irmina\texpython.htm

or email me:
irmina at ctv.es

Something like this :
|||->a
|||  |||->1
|||  |||  |->e
|||  ||-->2
|||  |--->4
||-->b
|--->c

just describing it as:
    a.add(['a','b','c'],'letras', 'letras del abecedario')
    a.add([1,2,4], 'a' )
    a.add(['e'],1)

    a.dibuja('letras')

It's so extremely easy and small ( the python code )
that it doesn't deserve a translation...

TODO:
- Read from a properties file.

"""

import sys

class atree:
    def __init__(self, cabecera, hojas, diccionario, inf):
        self.cabecera = cabecera
        self.inf = inf
        self.hojas = hojas
        self.diccionario = diccionario
        self.anyadido = ""
        self.anyadido_hijo = ""

    def dibuja(self, anyadido=""):
        self.anyadido = anyadido
        if not self.inf is None:
            self.imprimeln( self.cabecera + "("+ self.inf +")")
        else:
            self.imprimeln( self.cabecera)
        k = 0
        cuantas = len(self.hojas)
        for i in self.hojas:
            self.dibuja_hoja(i, k, cuantas)
            k = k + 1

    def dibuja_hoja(self , hoja, pos, cuantas):
        self.anyadido_hijo = ""
        self.imprime( self.anyadido )
        composicion_vert = self.compone_vert(cuantas, pos)
        self.imprime( composicion_vert )
        composicion_vert = self.escribe_flecha(composicion_vert)
        if hoja in self.diccionario.keys():
            self.diccionario[hoja].dibuja(self.anyadido +
self.anyadido_hijo)
        else:
            self.imprimeln( hoja)

    def compone_vert(self, cuantas, pos):
        a = ""
        for i in xrange(0, cuantas):
            if (i < cuantas - pos) :
                b = "|"
                c = "|"
            else:
                b = " "
                c = "-"
            a = a + c
            self.anyadido_hijo = self.anyadido_hijo + b

        return a

    def escribe_flecha(self, anyadido):
        self.imprime( "->")
        self.anyadido_hijo = self.anyadido_hijo + "  "
        return anyadido + "  "


    def imprime(self, s):
        sys.stdout.write(str(s))

    def imprimeln(self, s):
        sys.stdout.write(str(s)+"\n")

class rep:
    def __init__(self):
        self.diccionario  = {}
        #self.anyade()

    def add(self, lista_hojas, cabecera, comentario = None ):
        self.diccionario[cabecera] = atree(cabecera,
                                           lista_hojas,
                                           self.diccionario,
                                           comentario)

    def dibuja(self, l):
        self.diccionario[l].dibuja()

if __name__=='__main__':
    a = rep()
    a.add(['a','b','c'],'letras', 'letras del abecedario')
    a.add([1,2,4], 'a' )
    a.add(['e'],1)

    a.dibuja('letras')

Sorry again, if you get this twice.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From NO_SPAMnpirzkal at hq.eso.org  Tue Jul  6 15:56:38 1999
From: NO_SPAMnpirzkal at hq.eso.org (NO_SPAMnpirzkal at hq.eso.org)
Date: 6 Jul 1999 19:56:38 GMT
Subject: Inserting empty space into an existing file/pointer
Message-ID: <7ltn1m$nen$1@snatch.hq.eso.org>

Hi,

	I have the need to insert some empty space inside an already existing binary 
file. Is there a convenient, quick way of doin gthis with Python? I would 
like for example to add 100K of space starting at offset 2047K of a 8 meg 
file.
	The only way I could think of doing this in Python i susing some temporary 
files but ths is a bit slow and not necessarily very portable.
	If anybody knows a cool way to do this it would be nice... :-)

		Cheers,
		Nor




From Robert.Kern at p98.f112.n480.z2.fidonet.org  Fri Jul  2 18:30:22 1999
From: Robert.Kern at p98.f112.n480.z2.fidonet.org (Robert Kern)
Date: Fri, 02 Jul 1999 23:30:22 +0100
Subject: Python GUIs: Portability or not?
Message-ID: <00001695@bossar.com.pl>

From: kernr at mail.ncifcrf.gov (Robert Kern)

On Fri, 02 Jul 1999 09:32:48 GMT, Georg Mischler
<schorsch at schorsch.com> wrote:

>Mikael Lyngvig wrote:

[snip]

>There are other X servers around, but those who blend into the
>host GUI (instead of creating their own X desktop within a window)
>all seem to be commercial.
>
>Now *this* would be a task for an open software project... ;)

http://siddiqi.webjump.com/
It doesn't work yet, but it *is* an open source project.  Not sure if
they are aiming for the "blending," though.

>Have fun!
>
>-schorsch
>
>--
>Georg Mischler  --  simulation developper  --  schorsch at schorsch.com
>+schorsch.com+  --  lighting design tools  --  http://www.schorsch.com/

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From jsight at mindspring.com  Sat Jul 10 00:15:11 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Sat, 10 Jul 1999 00:15:11 -0400
Subject: Python re module
References: <7m5qev$lue$1@nnrp1.deja.com>
Message-ID: <3786C8CF.14DBF51B@mindspring.com>

catlee at my-deja.com wrote:
> 
> Try the following:
> 
> import re
> re.search("((.)\\1+)","a")
> 
> Now, I know this isn't proper syntax (you shouldn't reference a group
> inside itself), but on my machine python hangs.
> 
> Is this a bug, or a feature? =)

Yes.

-- 
---------------
Jesse D. Sightler
http://www3.pair.com/jsight/

"Do not use a hatchet to remove a fly from your friend's forehead." 
      - Chinese Proverb



From holger at phoenix-edv.netzservice.de  Wed Jul 14 05:59:29 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Wed, 14 Jul 1999 09:59:29 GMT
Subject: getting values of global variables?
Message-ID: <378C5F81.E6F8B014@phoenix-edv.netzservice.de>

Hi,

ok, sounds easy, but I get a little confused...;-(

I'd like to parse over a file reading line per line.
I want to write everything through a special filter
which exchanges known tokens. This may be special
dictionaries or global variables, too.
The tokens have to be surrounded by '%'-charcter.

Getting a problem with the global varaibles. I couldn't
get their values as string and havn't found the help
in books yet. I think it's quite simple and I'm just a 
"four blind mice".;-)

To check the filter I wrote:

-------------
import string, os, re
HEIN="Bl?d"

def egalizeTokens(aString):
	"changes known tokens in a string"
	try:
		splitter=re.split('(%.*?%)',aString)
	except:
		return aString
	for s in splitter:
		print s
		if len(splitter)==1:
			aTok = checkForToken(s)
		else:
			ox = string.split(s, '%')
			if len(ox)==1:
				aTok = ox[0]
			else:
				aTok = checkForToken(ox[1])
		if aTok==None:
			splitter[splitter.index(s)] = s
		else:			
			splitter[splitter.index(s)] = aTok	
	try:
		aTok = string.joinfields(splitter,'')
	except:
		aTok = splitter[0] #only one element!
	return aTok


def checkForToken(aString):
	"checks a string for known tokens"
	try: 
		aTok = eval(aString)
		if aTok==float(aString): #don't exchange numbers!
			return None
		return aTok
	except:
		aTok = os.environ.get(aString)
		if aTok==None:			
			aTok = directories.get(aString)
			if aTok==None:
				aTok = files.get(aString)
		return aTok

if __name__=='__main__':
	print egalizeTokens("This is a test! %WINDIR% %1 %2 %HEIN%. That's it")
	raw_input ("Press RETURN!")
-------------------------------

Perhaps there's a much easier way to get what I want????
(Still I am a newbie..;-)))



many thanx,
Holger



From tim_one at email.msn.com  Mon Jul 26 00:58:33 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 26 Jul 1999 00:58:33 -0400
Subject: Dreaded Newbie Question
In-Reply-To: <S3i8NRAzoFm3EwEN@epstein.co.uk>
Message-ID: <000701bed723$82f0d100$492d2399@tim>

[Dani Epstein]
> ...
> What I have found in the docs at hand (oh dear) is that there is a
> definite unix bias to them, which I suppose is understandable, but makes
> life terribly complicated for us Windoze fellows (and fellowettes of
> course).
>
> I presume this has changed in the latest docs.

It's a little less Unix-centric now, but that does go deep.  Python was
first written on a Mac!  Which is why Unix shows through everywhere <wink>.

It's not really Unix at work here so much as C:  The std C library + POSIX
define the only vaguely portable "system level" API there is.  Only MS
implements the Win32 API, and only Apple the Mac API.  So you go with
C/POSIX and look like Unix, mimic someone else's and get sued, or invent
your own API and scare *everyone* away <0.7 wink>.

> Thanks to all for your replies. This adventure looks like its going to
> mean good company as well as fun!

I found both to be true.  Let us know when you get stuck.

c.l.py-is-open-24-hours-a-day-ly y'rs  - tim





From ivanlan at callware.com  Tue Jul 20 23:35:55 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 21 Jul 1999 03:35:55 GMT
Subject: Merging two lists as a dictionary
References: <19990720211946.A12202@pooh.frostnet.net>
Message-ID: <3795401B.CA793358@callware.com>

Hi All--

Chris Frost wrote:
> 
> I have two lists and would like to merge them into a dictionary, with one
> list forming the keywords and the other the values in the dictionary. How could
> this be done? "map(None, keyword_list, value_list)" looked great, but it
> returns a list of pairs (I'd like to be able to specify values according to
> keywords...).
> 
> thanks for any pointers!

Here's a simplistic implementation:
----------------------------------------------
#!/usr/local/bin/python

k = ["one","two","three"]
v = [1,2,3]
d = {}

n = 0
for i in k:
    d[i] = v[n]
    n = n + 1

print "Keys:", k
print "Values:", v
print "Dictionary:", d
----------------------------------------------

Alternatively, you could take the list of pairs (tuples) you get back
from map() and iterate over that:

for i in pairs:
	dict[i[0]] = i[1]

<you're-in-trouble-if-the-lists-aren't-the-same-length>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From conversy at lri.fr  Mon Jul 26 12:40:04 1999
From: conversy at lri.fr (Stephane Conversy)
Date: Mon, 26 Jul 1999 18:40:04 +0200
Subject: Tkinter bug ?
Message-ID: <379C8F64.E48BF26@lri.fr>

File1.py:

from Tkinter import *

photo=None

def create_photo():
    global photo
    photo=PhotoImage(image="img.gif")



File2.py:

from Tkinter import *
from File1 import *

root = Tk()
create_photo()
b= Button(root, image=photo)
b.pack()
root.mainloop()


when executing File2.py, there is no error, but no image appears in the
button.

now if you chanage File2.py:

from Tkinter import *
import File1

root = Tk()
File1.create_photo()
b= Button(root, image=File1.photo)
b.pack()
root.mainloop()

it work fine. It seems to be a bug, isn't it ?

--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr






From miller at nialas.npl.uiuc.edu  Wed Jul  7 12:38:45 1999
From: miller at nialas.npl.uiuc.edu (M.A.Miller)
Date: 07 Jul 1999 11:38:45 -0500
Subject: total idiot question: +=, .=, etc...
References: <000015b4@bossar.com.pl> <76u2rha5qr.fsf@four.npl.uiuc.edu> <14211.28251.23566.522880@anthem.cnri.reston.va.us>
Message-ID: <76k8sc9zze.fsf@four.npl.uiuc.edu>

>>>>> "BAW" == Barry A Warsaw <bwarsaw at cnri.reston.va.us> writes:

    > I'm not entirely sure this is a good idea since self is
    > just a convention, but if a lot of people would like to see
    > this, I might consider adding it.

You could add a defcustom variable for people who use something
other than 'self'?

Mike



From johnh at phm.gov.au  Thu Jul  1 16:19:00 1999
From: johnh at phm.gov.au (John Hirsch)
Date: Thu, 01 Jul 99 15:19:00 EST
Subject: Python in the public eye !
Message-ID: <377B86E9@pcmgw.phm.gov.au>

Thanks to Drew and the Crew at the ANU :) who created a fantastic exhibit   
which has been very popular with all the visitors.
This is not the first exhibit at the Powerhouse Museum that has been   
developed using Python. Mainly due to the programming prowess of one Tim   
Docker (Ex engineer at the Museum) who is now on a mission to convert   
fellow programmers at his new job.

>Anyone who visits the Powerhouse Museum (Sydney, Australia) might be
>interested to know that the "Wedge" virtual environment exhibit in the
>new "Universal Machine" permanent exhibition which opened last night
>is Python powered.  The whole program for the exhibit was put together
>over the last fortnight on top of our in house "pSpace" VE system
>which use python as it's embedded language. I got to watch primary
>school kids using it for the first time today and it was a real
>buzzzz. Thanks Guido and Co !!!


 ----------------------------------------
Jonathan Hirsch
Powerhouse Museum
Sydney, Australia
E-mail : johnh at phm.gov.au
Phone: (02) 92170233
 ----------------------------------------  



From Felix.Puetsch at p98.f112.n480.z2.fidonet.org  Fri Jul  2 10:29:44 1999
From: Felix.Puetsch at p98.f112.n480.z2.fidonet.org (Felix Puetsch)
Date: Fri, 02 Jul 1999 15:29:44 +0100
Subject: test - ignore
Message-ID: <00001670@bossar.com.pl>

From: puetsch at gmx.de (Felix Puetsch)

[posted and mailed]

pserrand bucodi at worldnet.fr wrote in <377C92FC.9F3E6A08 at worldnet.fr>:

>pserrand wrote:
>
>> test with python
>
>reply test
>
>

Could you pleaser:
a) either use alt.test or similar or
b) cancel your messages afterwards?!

-- 
Felix Puetsch



From tismer at appliedbiometrics.com  Wed Jul 21 12:27:16 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Wed, 21 Jul 1999 16:27:16 GMT
Subject: Stackless & String-processing
References: <000701bed335$8a812740$a72d2399@tim> <7n4nrl$sfe$1@cronkite.cc.uga.edu>
Message-ID: <3795F4E4.8330755F@appliedbiometrics.com>

Hi friends,

My understanding of this thread is limited and reduces
to the following sketch:

Graham Matthews wrote:
> [Fernando Pereira] [arguing]
> [Tim Peters] [digging in]
[Graham] [ elaborating ]
> [Tim Peters] [syntax mixed with sematix]
[Graham] [ elaborating ]
> Tim Peters [insists]
[Graham] [ combinatores are robust, is Tim theorising? ]
> Tim Peters [ papers, not tried ]
[Graham] [ made a point, Tim is theorising ]
> Tim Peters [ combinators as assembly language ]
[Graham, now at Hutton/Meijer and the old CLean book]

can't see an end of this and would like to understand more.

I believe this thread would not have evolved if I had not
begun the Stackless Python / continuation/coroutine/generator
talk. Now I'm asking to excuse my limited understanding.

Besides my person, I think there are more interested readers
who would be happy to know what you are talking about at all.
I may assume that you would use private email if this
is not your intent. Also I'm sure that we'll see its
relevance for Python, soon.

Can you please be all a little more concrete?
How would your patterns look in Python?
Please show some small example code pieces.

This would also be valuable input for my implementation,
and would give me more reason to finish it at all.

gimme-a-reason-to-read-then-I'll-give-you-a-reason-to-write - ly y'rs

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From Eric.Kidd at p98.f112.n480.z2.fidonet.org  Sat Jul  3 07:15:50 1999
From: Eric.Kidd at p98.f112.n480.z2.fidonet.org (Eric Kidd)
Date: Sat, 03 Jul 1999 12:15:50 +0100
Subject: Curses, Ncurses, Slang, etc?
Message-ID: <000016b2@bossar.com.pl>

From: Eric Kidd <eric.kidd at pobox.com>

On Sat, Jul 03, 1999 at 06:14:02PM +0000, olli at rhein-zeitung.de wrote:
> I can only tell you that the ncurses module I wrote works for me on my
> RedHat 6 and Mandrake 6 systems. What problems do you have?

Nothing dramatic. I just need to set VERSION=1.5 and do something about
Setup.in:

  cp ./Setup.in Setup
  cp: ./Setup.in: No such file or directory
  make[1]: *** [Setup] Error 1
  make[1]: Leaving directory `...'
  make: *** [boot] Error 2

But I'm writing an end-user application and I didn't want to require anyone
to edit Makefiles.

> If support and so on is an issue, then I would use the slang bindings at
> the moment.

Slang bindings would be wonderful. Is there such a thing for Python, and if
so, where does it live?

Many thanks for your help!

Cheers,
Eric




From Sven.Drescher at dlr.de  Fri Jul 30 08:58:09 1999
From: Sven.Drescher at dlr.de (Sven Drescher)
Date: Fri, 30 Jul 1999 14:58:09 +0200
Subject: Libraries needed, please help!!!
Message-ID: <7ns6o5$jo2@news.go.dlr.de>

Hi!

I wanted to embed Python into C. I looked for the libraries to link them to
my program. But I haven't the libraries. I looked for it in the Internet and
everywhere, but never found.

Please give me a hint, where I can download the files:
libModules.a
libPython.a
libObject.a
libParser.a

Thank you very much!!!
Sven

--
______________________________________
German Aerospace Research Establishment






From MHammond at skippinet.com.au  Mon Jul 26 20:36:37 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Tue, 27 Jul 1999 10:36:37 +1000
Subject: Incompatibility.py
References: <379C8BB3.12D50F27@connection.com>
Message-ID: <7niuqf$rpm$1@m2.c2.telstra-mm.net.au>

Colin J. Williams wrote in message <379C8BB3.12D50F27 at connection.com>...

>IDLE returns:                        The name is: Incompatibility

Not for IDLE 0.5 it doesnt.  Are you sure you are not using "Import Script".
IDLE definately does this, as Python uses the "if __name__=='__main__':"
idiom extensively - and IDLE would be useless for running scripts if it did
as you describe.

The behaviour you describe sounds like you are simply confusing "Run Script"
with "Import Script".

Mark.






From ivanlan at callware.com  Fri Jul 23 19:23:07 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 23 Jul 1999 23:23:07 GMT
Subject: Converting itegers to strings
References: <19990723171921.A5205@pooh.frostnet.net>
Message-ID: <3798F95B.69310C38@callware.com>

Hi All--

Chris Frost wrote:
> 
> How can you convert numbers (an interger in this case) to strings? (i.e. I
> have some_number = 3, and would like to use it as a string)
> 

some_number = `3`

;-)

or

some_number = str(3)

You might want to consider joining the python-tutor mailing list.  You
can find out how at

	http://www.python.org

<python-slices-and-dices-too>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From ashish_swarup at my-deja.com  Thu Jul 29 15:12:46 1999
From: ashish_swarup at my-deja.com (ashish_swarup at my-deja.com)
Date: Thu, 29 Jul 1999 12:12:46 -0700
Subject: Still Not Getting Comp.Object.Moderated?
References: <36B6392C.1D23C4C0@oma.com>
Message-ID: <37A0A7AE.738C42@my-deja.com>

The address is 
ftp://ftp.isc.org/pub/usenet/control/comp/comp.object.moderated.Z

Regards,
Ash

Tim Ottinger wrote:
> 
> Several people have written me (in my capacity as
> c.o.m software janitor) about problems in accessing
> comp.object.moderated.
> 
> We already have multiple reliable points of insertion to
> the 'net. We're getting propagation internationally, and
> sometimes with amazing speed. There's really no problem
> at our end :-)
> 
> So the problem is at your end, or somewhere in the middle.
> News propagation can be a tricky thing. :-(
> 
> Right now, the very best thing you can do is to gently,
> kindly, politely, and patiently ask your ISP to carry
> the moderated group. You may point them to :
> <URL:ftp://ftp.isc.org/pub/control/comp/comp.object.moderated>
> to see the valid control messages and for comfirmation that this is a
> valid newsgroup.
> 
> Quite often, you can reach them at news at YOUR_ISP or less
> directly at support at YOUR_ISP. The ISP may not answer back
> the same day, so I reiterate, be patient. Some ISPs took
> several weeks to add the group. Also remember the problem
> may not even be their fault, but someone upstream from
> them.
> 
> If your ISP cannot get a reliable feed, they may contact the
> comp.object.maintainer at oma.com. Maybe we can work something
> out.
> 
> In the meantime, there's DejaNews and NewsGuy, and Doc's
> list of open news servers (but don't use servers without
> asking -- there can be trouble there)
> 
> --
> Your Janitor
> Tim
> 
> --------- There is no craftite conspiracy -------------
> Tim Ottinger         Object Mentor      OO Training and
> ottinger at oma.com      www.oma.com          Mentoring
> -------------------------------------------------------
> We can interpret a bad temper as a sign of inferiority.
>                                         -- Alfred Adler



From billtut at microsoft.com  Mon Jul 12 02:10:08 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Mon, 12 Jul 1999 06:10:08 GMT
Subject: Python and COM?
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F0F@RED-MSG-50>


> From: Mark Hammond [mailto:MHammond at skippinet.com.au]
> 
> 
> Karl Putland wrote in message 
> <7m9to7$6qg$82 at rks1.urz.tu-dresden.de>...
> > For a COM server implemented in python, can a type library 
> be generated?
> 
> No, but nearly :-)  Python has the framework and even some
> "proof-of-concept" code, but isnt done.

Err.. Mark means from within Python. That's not to say you can't build a
type library using an IDL file and MIDL.

Here's an example for the Python.Interperter COM server that exists in
win32com\servers\interp.py.

Unfortuantely, Python.Interpreter needs a slight change inorder to write an
IDL file.
Replace the  _public_methods_ line with:
	_dispid_to_func = { 1 : Eval, 2 : Exec }
A matching IDL that should work is attached below:
(As an added bonus, this should actually work)
The only GUID from interp.py is the coclass one. Every other GUID was just
generated.
Hope the below helps.... The MIDL docs in MSDN are certainly less than
useful. :(

// System idl include nonsense
import "oaidl.idl";
import "ocidl.idl";

[
	uuid(EB048AA4-C2D1-11d2-855D-00C04F797DBA),
	helpstring("Python.Interepreter dispinterface")
]
dispinterface IPythonInterpreter
{ properties: 
methods:
	[id(1)] HRESULT Eval([in] BSTR bstrExpression, [out, retval] VARIANT
pvarReturnValue);
	[id(2)] HRESULT Exec([in] BSTR bstrExpression);
};

[
	uuid(60240FE4-C2D2-11d2-855D-00C04F797DBA),
	version(1.0),
	helpstring("Python.Interpreter Type Library")
]
library PythonInterpreterLib
{
	importlib("stdole32.tlb");
	importlib("stdole2.tlb");

	[
		uuid(30BD3490-2632-11cf-AD5B-524153480001),
		helpstring("Python.Interpreter COClass")
	]
	coclass PythonInterpreter
	{
		[default] dispinterface IPythonInterpreter;
	};

}

Bill



From hnowak at cuci.nl  Sat Jul 10 09:10:27 1999
From: hnowak at cuci.nl (Hans Nowak)
Date: Sat, 10 Jul 1999 13:10:27 GMT
Subject: Python Code Repository?
In-Reply-To: <378645AA.5430A026@msfc.nasa.gov>
References: <378645AA.5430A026@msfc.nasa.gov>
Message-ID: <7m9tnk$6qg$63@rks1.urz.tu-dresden.de>

On 9 Jul 99, Drew McDowell wrote:

> After seeing the "Brensenham line algorithm" post above, I'm 
curious
> if there is a public "python code repository" (something akin to 
CPAN) out
> there?  Or does Starship assume that role?

I do have a snippet repository, but this is by no means exhaustive 
(yet), and does not aim to be (although it's growing). It's not 
really comparable to CPAN because it's not a repository for modules, 
but rather for simple (and not-so-simple) pieces of code.

I believe the URL is in my signature. :o)

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
Python Snippets: http://www.hvision.nl/~ivnowa/snippets/
The Purple Kookaburra Forum: http://www.delphi.com/kookaburra/



From fuuvid at NO.MAIL  Fri Jul 23 18:31:30 1999
From: fuuvid at NO.MAIL (fuuvid at NO.MAIL)
Date: 23 Jul 1999 22:31:30 GMT
Subject: EVERYTHING YOU WANT !!!!  1418
Message-ID: <7naqg2$7o7$1731@news1.xs4all.nl>

LOVE FROM SANDY
http://wetdreams.free-teen.com

18+ only !!!!!
bnrugosddprjkxqyksecfduxrswhedqguwngwpvj




From jepler at inetnebr.com  Mon Jul 19 20:30:40 1999
From: jepler at inetnebr.com (Jeff Epler)
Date: Tue, 20 Jul 1999 00:30:40 GMT
Subject: Python Bug:os.spawnvc
References: <7mto88$r8e$1@bunyip.cc.uq.edu.au>
Message-ID: <slrn7p7h37.fnd.jepler@craie.inetnebr.com>

On 18 Jul 1999 23:33:28 GMT, Anthony Pfrunder
 <s341625 at student.uq.edu.au> wrote:
>Hi,
>
>I think I have found a bug in python but it could be a Windows bug.  With
>os.spawnvc in this form:

It's Windows brain damage.

%0 should be what is the first argument in "args", but Windows overwrites
it with the actual path to the executable (.bat file in this case)

Another spawn* gotcha on Windows when I was writing C for it would be that
	os.spawnvc(P_WAIT, "test.bat", ["test", "a b", "c"], [])
would actually have 3 arguments, "a", "b", and "c".  I don't know if this
dane brammage is worked around in the Python source, but we worked around
it in our program.  The equivalent of
	os.spawnvc(P_WAIT, "test.bat", ["test", "\"a b\"", "c"], [])
works, for some value of work.  (but who knows what other gotchas lie in
wait)

Jeff
-- 
\/ http://www.infidels.org/                      Jeff Epler jepler at inetnebr.com
Fights between cats and dogs are prohibited by statute in Barber, North
Carolina.



From kyliu at idiomtech.com  Fri Jul  2 22:25:37 1999
From: kyliu at idiomtech.com (Ken Liu)
Date: Fri, 2 Jul 1999 22:25:37 -0400
Subject: Problem with informixdb and python1.5.2
Message-ID: <7ljsbc$srn$1@news.harvard.net>

I've been using the informixdb module (version 1.2a1) with python1.5.1 for a
while now with no problems.  Just a few days ago we installed python1.5.2,
and now informixdb won't work any more.  when I attempt to construct a
Connection object, I get an error about "unable to get name of current
working directory."  Here's a log of my interaction with the interpreter

Python 1.5.2 (#8, Jul  1 1999, 20:22:25)  [GCC 2.7.2.3] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import informixdb
>>> con = informixdb.informixdb('acme_online_demo')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "informixdb.py", line 47, in __init__
    self.conn = _informixdb.informixdb(logon)
InformixdbError: Error -922 performing LOGON: Cannot get name of current
working directory.
>>>


Informix's finderr informs me that "-922" means:

"A problem exists with the current working directory on this system or
on another computer system that your application is using. Possibly
your account does not have read access to the directory, or possibly
the file system that contains it is improperly mounted. Exit your
application, and retry. If the problem persists, see your system
administrator for assistance."

which does not make any sense.  Our Informix setup has not changed for the
last few months, and
the informixdb modules continues to work great under python1.5.1.

Does anyone have any suggestions on what might be wrong?





From rodh at hou.valmet.com  Mon Jul 19 09:52:54 1999
From: rodh at hou.valmet.com (Rod Haper)
Date: Mon, 19 Jul 1999 13:52:54 +0000
Subject: python?
References: <378DEB78.34747906@hou.valmet.com> <3792612A.57801B43@freenet.co.uk>
Message-ID: <37932DB6.469CC311@hou.valmet.com>

Sreekant,

Yes it's an adult site.  But you've apparantly missed the whole point of
the subject line.  That is reason for the "?" in the subject line. 
Someone posted three anonymous messages to the comp.lang.python list
last week advertising the URL and I was just wondering what this junk
has to do with Python and why it was posted on the comp.lang.python
list.  I don't want to see this junk on this news list.

Rod Haper


sreekant kodela wrote:
> 
> This is an adult site. Subject header is misleading, saying python.
> 
> Regards
> sreekant
> 
> Rod Haper wrote:
> 
> > http://207.240.225.250



From winkjj at my-deja.com  Tue Jul 13 11:32:15 1999
From: winkjj at my-deja.com (winkjj at my-deja.com)
Date: Tue, 13 Jul 1999 15:32:15 GMT
Subject: DLL load failed with SWIG DLL
References: <7mfd42$hrq$1@nnrp1.deja.com>
Message-ID: <7mfm5v$lql$1@nnrp1.deja.com>

In article <7mfd42$hrq$1 at nnrp1.deja.com>,
  winkjj at my-deja.com wrote:
> I am attempting to use SWIG to create an
> extension DLL for Python on a WinNT machine.
>
> I have successfully compiled and linked my DLL,
> which uses the -shadow option in SWIG.  My DLL is
> called Examplec, and the SWIG generated .py
> module is Example.
>
> When I attempt to import either the .py module or
> the DLL itself, I get the following error:
>
> ImportError: DLL load failed: The specified
> module could not be found.
>
> Now, I have searched all over this newsgroup and
> have found similiar problems, but no one has
> complained about this exact message.
>
> Other problems have been things like not being
> able to find DLLs that this DLL is dependent on,
> or the DLL not having a correct init<module>
> signature.
>
> I have verified that all this looks correct.  The
> DLL and .py module are on the PYTHONPATH.  And
> the init entry point is correct.  All the
> dependacy DLLs can be found on the system path as
> well.
>
> I also compiled the DLL with the __WIN32__ symbol.
>
> The only thing I question is whether or not I
> built SWIG correctly.  After you build SWIG you
> run 'nmake install'.  This installs SWIG into the
> target directory.
>
> I built SWIG on machine A, and then copied the
> install directory to the exact location on
> machine B.
>
> I don't see how this could make any difference,
> however, because the DLL compiles and links fine.
>
> Has anyone seen this before, or could someone
> give some suggestions?  I am at a loss at to what
> to try next.
>
> Thanks in advance for your help.
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.
>

I guess I answered my own question.  I was linked a against the
python15_d.dll, instead of python15.dll.  Once I changed this, it
worked fine.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gmcm at hypernet.com  Tue Jul  6 00:52:54 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 5 Jul 1999 23:52:54 -0500
Subject: Win32: silent script
In-Reply-To: <37809062.6D463928@phoenix-edv.netzservice.de>
Message-ID: <1280889250-62599454@hypernet.com>

Holger Jannsen writes:
> 
> I'd like to use a little python-script to start win-programs
> with special parameters. So I've made some links up to that 
> python-script in windows-menu. 
> But I'd like to let the programs start without the 
> rising dos-window, that occurs for a few centiseconds while
> python-script is working. Or even let it start with a minimized
> window outside the real screen or something like that....
> 
> How could I realize that?

pythonw.exe


- Gordon



From Gary.Pennington at uk.sun.com  Tue Jul  6 12:33:37 1999
From: Gary.Pennington at uk.sun.com (Gary Pennington - UK Performance Centre)
Date: Tue, 06 Jul 1999 17:33:37 +0100
Subject: python core dump (SIGBUS) on Solaris
Message-ID: <37822FE1.62CB4C8@uk.sun.com>

Hi,

I've encountered a fairly unusual problem using python on Solaris.
Before I investigate further, has anyone else already resolved this one?

Env:
Solaris 2.6 (all latest relevant patches)
compiler egcs-2.91.66

Test case to reproduce core is pretty straightforward :-

python
>>> import Cookie
>>> C=Cookie.Cookie("webJudgeUser=garyp;")
>>> C

At this point when using python interactively, a sigbus is generated.

Here's some trace from gdb

foucalt:/home/garyp 82 $ gdb python
GNU gdb 4.18
Copyright 1998 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you
are
welcome to change it and/or distribute copies of it under certain
conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for
details.
This GDB was configured as "sparc-sun-solaris2.6"...
(gdb) run
Starting program: /usr/local/bin/python
[New LWP    2        ]
[New LWP    3        ]
Python 1.5.2 (#1, Jul  6 1999, 16:50:14)  [GCC egcs-2.91.66 19990314
(egcs-1.1.2  on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
[New LWP    4        ]
>>> import Cookie
>>> C=Cookie.Cookie("webJudgeUser=garyp;")
>>> C

Program received signal SIGBUS, Bus error.
0xef2c5ab8 in _smalloc () from /usr/lib/libc.so.1
(gdb) info threads
  7 Thread 3          0xef647c14 in _swtch () from
/usr/lib/libthread.so.1
  6 Thread 2 (LWP 2)  0xef2b98d0 in _signotifywait () from
/usr/lib/libc.so.1
* 5 Thread 1 (LWP 1)  0xef2c5ab8 in _smalloc () from /usr/lib/libc.so.1
  4 LWP    4          0xef2b973c in ___lwp_cond_wait () from
/usr/lib/libc.so.1
  3 LWP    3          0xef2b699c in door_restart () from
/usr/lib/libc.so.1
  2 LWP    2          0xef2b98d0 in _signotifywait () from
/usr/lib/libc.so.1
  1 LWP    1          0xef2c5ab8 in _smalloc () from /usr/lib/libc.so.1
(gdb) where
#0  0xef2c5ab8 in _smalloc () from /usr/lib/libc.so.1
#1  0xef2c5b00 in malloc () from /usr/lib/libc.so.1
#2  0x4c094 in PyString_FromString (str=0x9d698 "<stdin>")
    at stringobject.c:145
#3  0x4d8d8 in PyString_InternFromString (cp=0x9d698 "<stdin>")
    at stringobject.c:1098
#4  0x2a434 in jcompile (n=0x0, filename=0xe0e90 "", base=0xe0ed8)
    at compile.c:3423
#5  0x2a328 in PyNode_Compile (n=0x11d938, filename=0x9d698 "<stdin>")
    at compile.c:3382
#6  0x32f94 in run_node (n=0x11d938, filename=0x9d698 "<stdin>",
    globals=0xe5b30, locals=0xe5b30) at pythonrun.c:883
#7  0x322f8 in PyRun_InteractiveOne (fp=0xffffffff, filename=0x9d698
"<stdin>")
    at pythonrun.c:528
#8  0x32120 in PyRun_InteractiveLoop (fp=0xd7430, filename=0x9d698
"<stdin>")
    at pythonrun.c:472
#9  0x32038 in PyRun_AnyFile (fp=0xd7430, filename=0x9d698 "<stdin>")
    at pythonrun.c:449
#10 0x1dfc4 in Py_Main (argc=1, argv=0xefffe844) at main.c:287
#11 0x1d9e4 in main (argc=1, argv=0xefffe844) at python.c:12
(gdb)


You can see that the problem is being manifested in _smalloc.

I started thinking about what the problem could be and I think it may be
related to the mymalloc routines which are used in the pypcre routines.
I seem to recall encountering problems before when there are multiple
malloc functions being used.

Does an instant remedy suggest itself to anyone?

Investigations to date:-

Rebuilt python without using threads - MADE NO DIFFERENCE
Tried different combinations of regex and it seems to fail only when 31
bytes are required.


Yours,

Gary Pennington
Sun UK




From aahz at netcom.com  Sat Jul 10 20:41:33 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Jul 1999 00:41:33 GMT
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787D8DA.3774AA7D@ssec.wisc.edu>
Message-ID: <7m9to5$6qg$80@rks1.urz.tu-dresden.de>

In article <3787D8DA.3774AA7D at ssec.wisc.edu>,
Nick Bower  <nick.bower at ssec.wisc.edu> wrote:
>
>Incidentally, I went back to the library reference and still couldn't find
>os.chdir

You probably have an old version of the docs; I suggest you go to
www.python.org and download a new set.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From ed at sis.yorku.ca  Tue Jul  6 13:05:19 1999
From: ed at sis.yorku.ca (Ed F. McCurdy)
Date: Tue, 06 Jul 1999 13:05:19 -0400
Subject: loading C++ into Python; static instance initialization?
Message-ID: <199907061705.NAA17054@sprawl.sis.yorku.ca>

The	quick answer probably is
link using the C++ compiler rather than ld or the C compiler.

The longer answer is that the static C++ objects with ctors
are named in a special way, and that during linking
code to call the ctors (and dtors) is generated. gcc calls this
process "collect", others have called it "munch".

Hope this helps,
Ed
-- 

Edward F. McCurdy,               | ed at sis.yorku.ca
York University, Toronto, Canada | (416) 736-2100x22685




From steve at renlabs.com  Wed Jul 14 20:04:20 1999
From: steve at renlabs.com (Steven Work)
Date: 14 Jul 1999 17:04:20 -0700
Subject: Expand tuple into argument list?
Message-ID: <87aesy4w3f.fsf@solano.in.renlabs.com>

I'm new to Python.  I want a dispatch function -- takes a function
name and other arguments, calls the named function with the remaining
arguments.  This works for limited length argument lists:

def dispatch(fnname, *rest):
    fn = dispatch.func_globals[fnname]
    if len(rest)==0:
	return fn()
    elif len(rest)==1:
	return fn(rest[0])
    elif len(rest)==2:
	return fn(rest[0], rest[1])
    elif len(rest)==3:
	return fn(rest[0], rest[1], rest[2])
    elif len(rest)==4:
	return fn(rest[0], rest[1], rest[2], rest[3])

but is there a general way to expand the tuple into an argument list?
(I know about keyword arguments already.)

Lest you ask why: I'm embedding a Python interpreter in a C program; I 
need to call Python functions directly from C, and I'd like a single
point of contact to minimize the required housekeeping.
-- 
Steven Work
Renaissance Labs
steve at renlabs.com
360 647-1833



From fredrik at pythonware.com  Sat Jul 24 13:16:34 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 24 Jul 1999 17:16:34 GMT
Subject: HTTP 1.1 anyone?
References: <37997331.0@pink.one.net.au>
Message-ID: <016601bed5f8$48c135a0$f29b12c2@secret.pythonware.com>

Arnold de Vos <adv at langdale.com.au> wrote:
> Is anyone working on an HTTP 1.1 implementation in Python?   I am
> considering creating variants of BaseHTTPServer and httplib to handle the
> 1.1 protocol.  But I would not like to re-invent the wheel.

httplib: http://www.lyra.org/greg/

</F>




From catlee at my-deja.com  Sun Jul 18 23:33:25 1999
From: catlee at my-deja.com (catlee at my-deja.com)
Date: Mon, 19 Jul 1999 03:33:25 GMT
Subject: Killing a thread
Message-ID: <7mu6a5$h61$1@nnrp1.deja.com>

Hi,

I'm writing a progrma where some user code is executed in a restricted
environment running in a separate thread.  Now, if the user is
malicious, s/he can just put 'while 1: 1+1'.  I'd like to be able to
recover from this.  Is there any way I can kill the thread that is stuck
in the infinite loop?

Thanks,
Chris


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From dubois1 at llnl.gov  Wed Jul  7 10:19:32 1999
From: dubois1 at llnl.gov (Paul F. Dubois)
Date: Wed, 07 Jul 1999 14:19:32 GMT
Subject: returning Python Lists from C++
References: <tKGg3.149$FH1.16131@newsr2.u-net.net>
Message-ID: <UlJg3.5798$w4.79939@news1.frmt1.sfba.home.com>

Using CXX (part of the LLNL distribution at ftp-icf.llnl.gov/pub/python) you
do:

List w;
...
w.append(...);
...
return new_reference_to(w);


Mark Butterworth <mark.butterworth at cis.co.uk> wrote in message
news:tKGg3.149$FH1.16131 at newsr2.u-net.net...
> Can anyone give me a quick example of how to return a python list object
> from a C++ function?
>
>





From ivanlan at callware.com  Thu Jul 22 12:31:20 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Thu, 22 Jul 1999 16:31:20 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
References: <001101bed3f2$5e8874a0$642d2399@tim>
		<7n675p$qsk@dfw-ixnews7.ix.netcom.com>
		<14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net> <7n7bo8$204@dfw-ixnews19.ix.netcom.com>
Message-ID: <37974758.FF2DAE55@callware.com>

Hi All--

Aahz Maruch wrote:
> 
> 
> I'm afraid I can't agree that there's *ever* a good time to use "from
> foo import *" in a multi-programmer environment.  Too often, you have a
> programmer new to a particular chunk of code trying to debug it; not
> knowing exactly where a method/function call is coming from vastly
> increases debugging time (particularly when there are hundreds of
> modules and scripts).  If you want to use an unadorned class/function,
> you can just take the trouble to explicitly list it as "from foo import
> bar, baz".
> 

All true, except for perhaps Tkinter.  And my soon-to-be-released
mayalib-1.2.  Both are specifically designed to use 'from xxx import *'
and both carefully refrain from polluting your namespace.  But possibly
you don't use Tkinter in your projects, Aahz?  (I think I can safely
assume you don't use any Mayan dates. ...;-)

> My point was related to Tim's in that we're both talking about how
> Python's strait-jacket approach is a Good Thing; I'm simply using an
> example where I think Python could actually be tightened.
> 

I think it's exactly as tight as it needs to be.  If it were tighter,
people would spend more time and energy finding ways around the
restrictions, and it wouldn't be so attractive--and fun.

<houdini-died-in-a-straight-jacket>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From andrew at andrewcooke.free-online.co.uk  Thu Jul 22 09:30:31 1999
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Thu, 22 Jul 1999 13:30:31 GMT
Subject: Repost: Re: Scientific Plotting?
References: <379392E7.EB0F29E7@ssec.wisc.edu> <7n5fhn$9e5$1@nnrp1.deja.com>
Message-ID: <7n76dn$rbc$1@nnrp1.deja.com>

In article <7n5fhn$9e5$1 at nnrp1.deja.com>,
  mhuster at hotmail.com wrote:
[...]
> Plplot: uses stroke fonts, even in postscript output. We cannot get a
> uniform look with matlab produced plots. Bummer. Otherwise it looked
> good. I am afraid when blown up the fonts will look even worse. For
> example. To get a bold font you have to play with pen width and font
> size. Isn't "Helevetica-Bold" easier to specify? Also, it is at a
> developmental (& FORTRAN) deadend. No chance to change things.
>
[...]
> Pgplot: looks like plplot.

Just thought I'd add that pgplot is pretty much a standard in some
astronomy circles, and is probably used as much via it's perl package
(from Karl Glazebrook) as from Fortran (which, believe it or not, is not
a dead language in the same circles.... ;-)

So, while I don't think it is either OO, or being further developed, it
is in use (and does have it's own particular style which some find quite
attractive)...

(Just in case you are scared it will disappear in a few months, really).

Andrew


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Travis.Oliphant at p98.f112.n480.z2.fidonet.org  Fri Jul  2 04:14:42 1999
From: Travis.Oliphant at p98.f112.n480.z2.fidonet.org (Travis Oliphant)
Date: Fri, 02 Jul 1999 09:14:42 +0100
Subject: GUI's etc: PyGtk on Windows
Message-ID: <00001689@bossar.com.pl>

From: Travis Oliphant <olipt at mayo.edu>


Previously I mentioned that Gtk works on any system with an X server,
which is true.

I also want to point out that Gtk is a cross-platform solution
because there is a port of the GDK (the underlying windowing calls that
Gtk uses) for Win32.  You can find a link at http://www.gtk.org

>From what I've read it looks like Gtk was intentionally written on top of
GDK so that a port of GDK would make Gtk available.  The person who wrote
the GDK port to Win32 did it to get the GIMP to run on Windows.
There are screenshots of the result available on the site.

So if someone develops mainly for X and just wants to make sure that it is
not impossible to run on Windows, you can use PyGTK.  In fact a great
program using GTK can be a strong motivator to get the ports of GDK going.

Just some thoughts,

Travis



From hinsen at cnrs-orleans.fr  Mon Jul  5 04:52:38 1999
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 05 Jul 1999 10:52:38 +0200
Subject: C shared object files
References: <199907021717.KAA03837@ogl.llnl.gov>
Message-ID: <m3so73jx61.fsf@chinon.cnrs-orleans.fr>

Martin Casado <casado at ogl.llnl.gov> writes:

>    I am trying to get python to dynamically load a C object module, and 
> first off I can't figure out how to compile myCollProggy.c into a 
> shareable object file, and secondly I was curios if there is a way to check
> at runtime if a module is on the system or not.  

The answer to the first question can be found in section 2 (Building C
and C++ Extensions on Unix) of "Extending and Embedding the Python
Interpreter", available at www.python.org and many other places that
have the standard Python documentation. Assuming of course that you
are using Unix, for other systems the procedure is different and
probably compiler dependent.

As for the second question, the simplest solution is a test import:

  try:
     from foo import bar
  except ImportError:
     print "Sorry, module 'foo' is not available."

-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------



From bwarsaw at cnri.reston.va.us  Mon Jul 19 23:20:47 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Mon, 19 Jul 1999 23:20:47 -0400 (EDT)
Subject: Are postings to comp.lang.python.announce mirrored here?
References: <37933DB1.47814E84@cfdrc.com>
	<14227.22618.779710.843912@anthem.cnri.reston.va.us>
	<JNPk3.532$ui4.178537@news.shore.net>
Message-ID: <14227.60175.280720.305790@anthem.cnri.reston.va.us>

>>>>> "MPR" == Michael P Reilly <arcege at shore.net> writes:

    MPR> FYI, I've attempted some cross-posts to both which didn't
    MPR> even make it to the moderator.  Cross-posting is the Right
    MPR> Thing(TM), but not all servers seem to handle it when one is
    MPR> moderated.  Seperate posts is at times the only option (if
    MPR> you want it announced in both newsgroups).

Ah!  You can always just email it to both python-list at python.org and
python-announce at python.org.

-Barry



From claird at starbase.neosoft.com  Wed Jul 21 07:49:29 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Jul 1999 11:49:29 GMT
Subject: [Q] In-Browser technology
References: <932337314.611.103@news.remarQ.com> <BE2245BA1DCDAE31.8804258C1F60F9A7.2EB4BCD57DBA05D1@lp.airnews.net> <932522335.684.43@news.remarQ.com>
Message-ID: <FE280F7BBBA37CB8.59F593825701D528.AC56447361430CED@lp.airnews.net>

In article <932522335.684.43 at news.remarQ.com>,
Vladik <reason at shadow.net> wrote:
>OK, I will try to clarify my question.
>With the warning that I may not be speaking
>authoritatively about the web technologies yet:
>I am just starting this research.
			.
			.
			.
>I do not know what it is supposed to be
>a plugin or not --- but a technology that
>allows to develop not-trivial GUI applications
>that run withing a browser.
What's the benefit of the browser?  It sounds
as though you want something that works over
a network, and operates something like Java
and JavaScript, but programmed in a language
"better" than Java or JavaScript, and with no
particular server-side component.  Why are we
talking about Web applications at all?  Is
there some reason you're not choosing your own
language and designing your own protocol?  Many
of us do that sort of work daily.
			.
			.
			.
>
>> 2.  "... a prototype of the client
>>     in MS ASP (active server pages)
>>     and, I think but not sure, blended
>>     with Visual Basic ..."  I have no
>>     definite idea of what you're say-
>>     ing here.  Is it that your
>>     customer has a picture of the
>>     visual appearance of the desired
>>     deliverable?  What is there about
>>     the ASP prototype that does *not*
>>     meet the requirements?  Is VB
>>     blended on the server or client
>>     side?  Does it matter?  Is it really
>>     VB, or VBScript?
>
> I meant that VB script was used in the prototype, not VB
I'll ask again:  what constrains your team from
rewriting the display-handling in JavaScript,
refining the prototype business rules, and
declaring that you have won the war?  Is there
something in the architecture of the prototype,
apart from VBScript's non-portability, that
doesn't meet your requirements.
>
>> 3.  Are you aware that *lots* of browsers
>>     read pages served as ASP?  You can
>>     use ASP without requiring IE.
>NO I was not aware.  Thanks for pointing out
>I will try to find more about netscape (the only
>two browser we are to support are Netscape and IE).
What does it mean to support the browsers?
It sounds as though you're willing to deploy
plugins on every desktop anyway; what benefit
do the browsers give?
			.
			.
			.
>I guess what I was looking for is something like
>this:
>"Here is a technology like Java (or different
>but accomplises the same thing), and it is based on this
>OO programing language and here is the page where you
>can read more about that also includes technical comparasings
>between this technology and JavaScript (or other)
>"
OK, sure.  Tcl has a plugin.  There's a whole book
on Perl-coded Web clients.  The Grail Web browser
is written in and fully programmable in Python.
You've already found out about Juice.  People have
done wonderful Web and more general client-server
things with Eiffel and Forth and Scheme and ...  I
have unpacked a few of my notes on language compar-
isons at <URL:http://
starbase.neosoft.com/~claird/comp.lang.misc/language_comparisons.html>.

Still, I feel I'm just not getting it.  You want
something just like Java, except faster; you're not
going to find it.  I acknowledge Juice is zippier
and more compact for particular test cases; I have
no way of knowing, though, whether those isolated
measurements have *any*thing to do with your domain.
I rather suspect they don't.
			.
			.
			.
>I have never used JavaScript for development, for comparising Juice
>applet gets to my computer faster and draws the wireframe
>example faster.
>The client app we will have to develop will have to
>a) maintain a limited set of bussiness rules
>b) apply a to relatively large sets of data from the database
Are you quite sure you want all this processing to
be on the client side?  Why is this a Web application
at all?
>
>
>> 6.  How much client-side programming are
>>     you doing?  Why does object-orienta-
>>     tion there matter to you?
>
>Yes it does. Because the app will contain some bussiness
>rules, some data access and data presentation layers.
I'll take this as a religious matter.
			.
			.
			.
>Well, I guess what I was asking the ability to have
>the server run on any platform and this web client
>be executable withing a browser that runs on any platform.
Everything we've discussed, apart from
VBScript, is as portable as you seem to
need.
			.
			.
			.
>They make SRC available, I downloaded it,
>the license does not seem to be restrictive, but I am
>clear on that 100% yet.
I repeat:  talk with them.  Perhaps Juice
is more portable than you realize.
			.
			.
			.
There's something about this picture that
doesn't make sense to me.  You are rela-
tively inexperienced in designing Web
architectures, so you're asking for help.
Good.  It sounds as though you're respon-
sible for quite a serious project.  If
this work is as important as it seems,
and you don't have local experts to offer
seasoned judgment, and you're in the hurry
I gather you are, then I advise you to fly
in someone you trust for a couple days of
consultation.  It will be far less expen-
sive than trying to recreate on short
notice the expertise that comes only with
practice.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org  Fri Jul  2 11:28:25 1999
From: Mikael.Lyngvig at p98.f112.n480.z2.fidonet.org (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 16:28:25 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001675@bossar.com.pl>

From: mikael at pobox.com (Mikael Lyngvig)

On 2 Jul 1999 12:04:54 GMT, lvirden at cas.org wrote:

>I wonder why various industry rags have proclaimed during the past
>10 months about Apple's G3 machines outselling other vendor
>machines.  ...

That makes two of us.


-- Mikael



From rodisi01 at my-deja.com  Tue Jul 27 18:54:00 1999
From: rodisi01 at my-deja.com (Isidor)
Date: Tue, 27 Jul 1999 16:54:00 -0600
Subject: IDLE mini-hack & ? (adding menu items, understanding *args, self)
In-Reply-To: <37969A3D.EEDC201@compaq.com>
Message-ID: <199907272152.QAA01078@cro.ots.ac.cr>

Greg Ewing wrote:

> Isidor wrote:
> > 
> > ##def date_time_stamp(self):
> > def date_time_stamp():
> 
> I think the first version of that line was better.
> If it's a method of a class, it'll need a self
> argument. That would account for the "No arguments
> expected" exception - it's being called with one
> argument (self) but it's not expecting any.

Hello Greg - 

Thanks for responding to my query. I also thought that having "self" 
in there would help. I didn't think this because I knew what I was 
doing, but, because like a good monkey, I see that most funcs in a 
class have a "self" argument, so I figured, yup, better try it out. 
Unfortunately, it didn't work. In fact, it brought up an "odd" error: 


case 1: using: def date_time_stamp(self):  ...generates... 

>>> Exception in Tkinter callback
Traceback (innermost last):   
File "C:\Program Files\Python152\Lib\lib-tk\Tkinter.py", line 764, in 
__call__     
return apply(self.func, args) 
TypeError: too many arguments; expected 1, got 2 


Why do I call it an "odd" error message? Well, because to the 
untrained eye (e.g., mine) the previous error message is completely 
at odds with the following message: 

case 2: using   def date_time_stamp():  ...generates... 

>>> Exception in Tkinter callback
Traceback (innermost last):   
File "C:\Program Files\Python152\Lib\lib-tk\Tkinter.py", line 764, in 
__call__     
return apply(self.func, args) 
TypeError: no arguments expected 


To me it sounds as if in case 1, when I gave an argument, it said I 
gave too many while in case 2, when I didn't give an argument, it 
said it didn't want any. (Kind of like life?..;) 

But of course, what's wrong here is my totally ass-backwards 
interpretation of the error messages. I would never have figured this 
out if it hadn't been for an email sent to me personally by a kind 
gent who shall go unnamed here (he is being bcc'ed (hello and 
thanks!) (am i correct in assuming that a private emailer does not 
want public acknowledgment?)). He wrote: 

:::::::::::::::::::: 
You seem to have done everything right (I don't use TKInter enough to 
help with the real debugging), then ignored the information in the 
exception...  

Traceback (innermost last):   
File "C:\Program Files\Python152\Lib\lib-tk\Tkinter.py", line 764, in 
__call__     
return apply(self.func, args) 
TypeError: no arguments expected 

Which is trying to say: "Tkinter tried to call a function with too 
many arguments"  

So... 

def date_time_stamp( *args, **namedargs)
::::::::::::::::: 

Since at the time I understood nothing about the uses of *args, 
**namedargs , I did not understand his clue. Thanks to some recent 
posts and replies discussing these argument-passing-methods, I have a 
better grip on what's happening here. So, with a little time on my 
hands (yeah, right ;), I did a simple copy and paste from his 
message, and voila, it worked. I was able to pass through the 
function definition and into the interior of the function. He wasn't 
giving me a *clue*, he was giving me the freakin' *answer*! (Again, 
thanks.) 

error messages, a newbie's eye-view:

So now that it works, I am looking back, trying to reinterpret the 
error messages to fit with reality...not my ass-backwards version of 
them. 

case 1: def blah(self): 
error:  too many arguments; expected 1, got 2 

case 2: def blah(): 
error: no arguments expected 

It seems to me that the error message is being "spoken" by the 
"receiving" function (i.e., the "blah" func). In case 1, it is set up 
to handle one argument, but gets more than that and balks. In case 2, 
it is set up to handle no arguments, gets some, and balks. You see, I 
was thinking that the error was being "spoken" by something *after* 
the function had been defined (um, I have mentioned several times 
that I'm new at this, right?;)...so that whatever was being invoked 
after the def statement was balking in a very strange way...telling 
me that no arguments were expected when I didn't *give* it any 
arguments. Maybe this would have been clearer to me if the error 
message in case 2 were (similar to the error message in case 1): "too 
many arguments; expected 0, got 2", then I would have realized that 
both errors were variations on a theme rather than opposites. Of 
course, this new interpretation could be totally wrong too. If that 
is the case, I hope someone will point it out to me. 

One thing I don't really understand here is why it is necessary to 
raise an exception when an argument is passed to a function that 
isn't prepared to handle it. If the argument is needed *inside* the 
function, I could understand an exception being raised then (e.g., 
"hey, where's my argument"), but in my case i'm doing nothing (or at 
least I *think* I'm doing nothing) with those arguments. Why do I 
need to make a "landing strip" for them? To add more metaphors to the 
blender, something about this reminds me of the trick of putting a 
potato in a car exhaust pipe. If the exhaust has nowhere to go, it 
backs up until nothing can move in or out of cylinders and the motor 
dies (or at least it did in Beverly Hills Cop...or whatever that 
Eddie Murphy movie was). So, maybe by allowing the unused arguments 
to pass through my function, everything flows nicely and nothing gets 
stopped up. Hmm, strange. Am I on basically the right track here with 
this idea? 

Understanding the "self" ("the toughest thing of all" - stated in a 
philosophy of science seminar on reflexivity i once attended):

So, anyway, what did I have inside that function? Well, basically 
just some test code to see if I was getting that far:

print "cheese"

It worked, but poorly. It pushed whatever text was already on the 
line (at the command prompt) on to the next line and jumped to the 
front of the line. Hmm. I suspected that using "print" was like using 
a sledgehammer where a scalpel was needed. Scanning through the other 
functions in EditorWindow.py, I saw a lot of things being done to/by 
self.text. In the init constructor, I saw that "self.text = 
Text(...lots of arguments that would make sense for a text widget 
type thing...)", and I realized, aha, "text" must be the arena in 
which we are actually playing when using IDLE. So, then I wondered, 
what "methods" does "text" provide me to insert text? I looked at the 
Tkinter text man page and found a section on insert that looked 
reasonably like what I wanted to do, but it didn't look anything like 
what I was seeing in EditorWindow.py (to me, at least). So then I id 
a "Find in Files..." (oh my, what a wonderful feature!) for 
text.insert in the idle files and found a nice example in line 51 of 
Autoexpand.py. Something like:

self.text.insert("insert", newwords)

My guess is that insert is a method of text, and "insert" (first 
argument) is the location of the insertion point and will be the 
location where the second argument gets injected into "text". So, 
just for kicks, I copied that line into my newly-functioning def, 
changed "newwords" to "timestr" (which contained the date-time string 
I wanted), and tried it out. (I suspected it wasn't going to work, 
but I wanted to try anyway.)

case 3: 

def date_time_stamp( *args, **namedargs):
	timestr = ...stuff...
	self.text.insert("insert", timestr)

>>> Exception in Tkinter callback
Traceback (innermost last):
  File "C:\Program Files\Python152\Lib\lib-tk\Tkinter.py", line 764, 
in __call__
    return apply(self.func, args)
  File "C:\PROGRA~1\PYTHON15\TOOLS\IDLE\EditorWindow.py", line 586, 
in date_time_stamp
    self.text.insert("insert", timestr)
NameError: self

I was curious: I wanted to know whether a function inside a class 
could automatically see the methods and attributes of that class. 
Well, this error message makes me think that the answer is no. Self 
has to be "created" in the arguments of the function. Ok, so now I'm 
starting to understand *why* everyone always says you have to have 
"self" as the first argument. But, I wondered, maybe "self" doesn't 
work here because I haven't created it within the context of this 
func, but "text" would be recognized because it is created in the 
class that houses the func. So I tried the following (hey, give me a 
break, I'm a newbie! ;) :

case 4:
def date_time_stamp( *args, **namedargs):
	timestr = ...stuff...
	text.insert("insert", timestr)


>>> Exception in Tkinter callback
Traceback (innermost last):
  File "C:\Program Files\Python152\Lib\lib-tk\Tkinter.py", line 764, 
in __call__
    return apply(self.func, args)
  File "C:\PROGRA~1\PYTHON15\TOOLS\IDLE\EditorWindow.py", line 587, 
in date_time_stamp
    text.insert("insert", timestr)
NameError: text

Nope, as I suspected, that didn't work either. If I am correct, then, 
these results point to the interesting fact that a function knows 
nothing about its context that it isn't told, and that the first 
argument (conventionally "self") of an argument within a class is 
used by the class to pass itself and its context to the function, and 
is used by the func to "catch" everything the class has to offer. So 
I did what I should have done all along:

case 5:

def date_time_stamp(self, *args, **namedargs): 
	timestr = ...stuff...
	self.text.insert("insert", timestr) 

result upon selecting the menu/or hitting the right keys:

>>> 19990726.1735

Yay. It works. Just to see if I understood what was going on here, I 
tried the following (pushing my luck, I know!):

case 6:
def date_time_stamp(blah, *args, **namedargs): 
	timestr = ...stuff...
	blah.text.insert("insert", timestr) 

Yup, that worked too. So now I understand what they mean when they 
say that calling the first argument "self" is just a convention. 
*Having* the argument there isn't the convention, in fact, in most 
cases it's *necessary*, but it can be *called* almost anything. "Aha, 
the grasshopper is beginning to understand." ;^)

"The self is just an empty vessle into which...."
"The self must be empty before...."

	no no no no... ;) 


Ok, seriously now. Here is something I would like to know. Is it 
possible for function y in a class to have access to a variable 
defined in function x of that same class (without having to 
explicitly pass the variable with a return statement)?:

def class something:

1	...stuff, including __init__...

2	def funcx(self, *args, **namedargs):
3		zip = "googoo"
 
4	def funcy(self, *args, **namedargs):
5		print zip

Something causes me to suspect that this won't work. What would make 
it work? Can I attach zip to "self" somehow in funcx, e.g., line 3: 
'self.zip = "googoo"'?  If I then change line 5 to 'print self.zip', 
will this work? Will it only work if self.zip is defined in the 
__init__ func? Until I trust myself to create a working class, I 
don't think I can reliably test this out myself. In the meantime, I 
will welcome any insights anyone is willing to provide (including 
"read chap x, sect y of tutorial or language reference").

Ok, I hope this rambling troubleshooting log helps some other newbie 
understand what the hell is going on with passing arguments to 
functions, including class "context".

Thank to Greg and my friendly e-mailer for the helpful advice! With 
this post and my original one, it should be possible to make a "date- 
time stamp" (or any other "insert something here") menu item. If you 
can't find the original post and want to do this, email me, and I'll 
put it together more coherently for you. 

Thanks again, and take care. 

Isidor 



From ovidiu at cup.hp.com  Thu Jul 22 16:24:52 1999
From: ovidiu at cup.hp.com (ovidiu at cup.hp.com)
Date: Thu, 22 Jul 1999 13:24:52 -0700
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have 
 code.)
In-Reply-To: Your message of "22 Jul 1999 13:32:55 +0200."
             <p5n1woncm0.fsf@bidra241.bbn.hp.com> 
Message-ID: <199907222024.NAA29172@hercules.cup.hp.com>

On 22 Jul 1999 13:32:55 +0200, Markus Kohler <markusk at bidra241.bbn.hp.com> 
wrote:

> IMHO it's very close to Smalltalk, also Guido probably wouldn't agree with me ...
> And It has all essential OO features.  
> 
> 
> The features that make it very similiar to Smalltalk are :
> byte code interpreter, 
> all data are objects (is that really true ?), 
> crossplatform, 
> reflection api, 
> dynamically type checked, 

One problem though with Python is the fact that it does make a distinction
between types and classes. Another is the garbage collection scheme that's
using.

> In fact I cannot remember any successfull language that is closer to 
> Smalltalk than Python is. 

Objective-C. Is a dynamically typed language and is compiled.


Greetings,
--
Ovidiu Predescu <ovidiu at cup.hp.com>
http://andromeda.cup.hp.com/  (inside HP's firewall only)
http://www.geocities.com/SiliconValley/Monitor/7464/





From fredrik at pythonware.com  Mon Jul 19 11:58:09 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Jul 1999 15:58:09 GMT
Subject: Problems with PyRun_SimpleFile()
References: <379336D5.799D83E0@gmx.net>
Message-ID: <001501bed1ff$80ac1720$f29b12c2@secret.pythonware.com>

From: Markus Ewald <Markus_Ewald at gmx.net> wrote:
> Excuse this dumb question, but how can I use PyRun_SimpleFile() ?
> I'm using VC6 on Win98, where I've declared some simple functions for
> testing.
> 
> Well -- everything works fine when I use PyRun_SimpleString().
> 
> If I put the same lines I gave PyRun_SimpleString() in a File and then
> use PyRun_SimpleFile() to execute it, I get an access violation.
> 
> I did everything exactly as I've seen in Modules\Main.C (except that I'm
> using CPP), what's wrong ?

make sure you use the same runtime library for
all modules (Python is usually built with -MD).

creating a file with one runtime library and using
it with another is a sure way to get into trouble...

(in other words, this is a MSVC question, not
a Python question per se).

</F>




From nick at NOSPAMvideosystem.co.uk  Wed Jul 14 04:29:02 1999
From: nick at NOSPAMvideosystem.co.uk (Nick Trout)
Date: Wed, 14 Jul 1999 09:29:02 +0100
Subject: wxPython Form editor
Message-ID: <378c488e@news.xtml.co.uk>

Has anybody written a dialog/form editor for wxPython?

Thanks,
Nick.






From fdrake at cnri.reston.va.us  Fri Jul  9 17:28:17 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 9 Jul 1999 21:28:17 GMT
Subject: Grail won't work
In-Reply-To: <3786527C.6230C4D2@ssec.wisc.edu>
References: <3786527C.6230C4D2@ssec.wisc.edu>
Message-ID: <7m9tn4$6qg$45@rks1.urz.tu-dresden.de>

Nick Bower writes:
 > I just installed Grail 0.6 to take a look.  I have Py1.5 and Tk8 but it
 > won't run:

  Can either try it with Python 1.5.2.  I thought there was a patch on 
the Web site, but I don't see it now.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From fredrik at pythonware.com  Wed Jul 28 16:11:10 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Jul 1999 20:11:10 GMT
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <slrn7psisn.da.wtanksle@dolphin.openprojects.net> <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com> <slrn7pum9t.2ro.wtanksle@dolphin.openprojects.net>
Message-ID: <001501bed935$561b8210$f29b12c2@secret.pythonware.com>

William Tanksley <wtanksle at dolphin.openprojects.net> wrote:
> >    for an "if" statement, "else" is a path taken if the
> >    expression evaluates to false.
> 
> Close enough -- I would say it's the path taken if the 'if' block isn't
> executed.

you can say what you want, but it isn't defined
that way; the language reference says:

    "If all expressions are false, the suite of
    the else clause, if present, is executed." 

    (where "all expressions" are the expressions
    for all if/elif clauses in the statement)

if you continue reading, you'll find:

    "if the expression is false (which may be
    the first time it is tested) the suite of the
    else clause, if present, is executed"

and

    "when the items are exhausted (which is
    immediately when the sequence is empty),
    the suite in the else clause, if present, is
    executed"

for ALL these statements, the else clause is exe-
cuted if and only if ALL controlling expressions
evaluate to FALSE.

> >    for a "while" statement, "else" is a path taken if the
> >    expression evaluates to false.  or in other words,
> >    when the loop terminates by natural causes.
> 
> In other words, it executes when you'd expect an 'else' to not execute --
> 'else' doesn't mean "natural causes".

huh?  the "else" clause isn't involved in the loop's
termination at all; it's executed when the con-
trolling expression becomes false.  if you raise
an exception, or use the break statement, or
return to the caller, you don't end up in the
"else" clause by a very simple reason:  the con-
trolling expression isn't false.

> >    for a "for" statement, "else" is a path taken when
> >    there are no more elements to loop over.  or
> >    in other words, when the loop terminates by
> >    natural causes.
> 
> In addition to the above carping, I have to add that the code following
> the for (or while) loop is what I'd expect to execute after the loop
> terminates.

huh?  the "else" clause is part of the if/while/for
statement. the code that follows the loop isn't.

I repeat:

for ALL these statements, the else clause is exe-
cuted if and only if ALL controlling expressions
evaluate to FALSE.

whether blocks are executed or not isn't part
of the definition; only the conditions are.

> I expect consistency, and I expect that the major features (else with if
> and else with catch) will define consistency, not minor features like else
> with looping.

you have consistency.  for-else and while-else works
EXACTLY like if-else.  for ALL these statements, the
else clause is executed if and only if ALL controlling
expressions evaluate to FALSE.

you can argue as much as you want, but that's the
way it is.  you gotta live with it (unless you have your
own time machine, of course ;-)

</F>




From gmcm at hypernet.com  Wed Jul  7 11:04:15 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 7 Jul 1999 10:04:15 -0500
Subject: Python array --> C
In-Reply-To: <378354DA.D45DF723@ncsa.uiuc.edu>
Message-ID: <1280766240-70002628@hypernet.com>

Randy Heiland writes:

> Gordon McMillan wrote:
> 
> > Randy Heiland writes:
...
> > > From main(), I would like to invoke xyzzy.foo() and get this array
> > > of unsigned chars into a C array. How do I do this?  I assume I'll
> > > need to use PyRun_String(...) somehow - true?
> >
> > Not at all. You'll want to use PyString methods to get its size and
> > char *, etc. You can use python as a lib without ever involving the
> > interpreter. See the C API reference manual (and be prepared to grep
> > the objects and modules subdirectories).
...
> Can you please be more specific (sorry, I need a lot of
> hand-holding). Exactly what PyString method would I call in main()
> to (1) invoke the xyzzy.foo() method and (2) get the resulting byte
> array into a C array?

In Python, you would first "import xyzzy", right? The C API docs 
mention PyImport_ImportModule(). So we grep and see it should look 
like:

module = PyImport_ImportModule("xyzzy");
....
func = PyObject_GetAttrString(module, "foo");
....
PyObject *res = PyObject_CallFunction(func, ...);

Now you should probably do

if (PyString_Check(res)) {
  int sz = PyString_Size(res);
  char *p = PyString_AsString(res);
  memcpy(myarray, p, sz);
  ...
}

I've left out error handling and the all important refcounting. 
Basically it's a matter of figuring out what would happen in Python, 
looking through the C API docs, and using grep to find examples.

Pretty soon you'll realize it's a lot more fun to write it in pure 
Python <wink>.

- Gordon



From alberto at totallygames.com  Wed Jul 14 20:01:46 1999
From: alberto at totallygames.com (Alberto Fonseca)
Date: Wed, 14 Jul 1999 17:01:46 -0700
Subject: Building Python static lib in WIN32
Message-ID: <931997051.156.53@news.remarQ.com>

Hello, I would like to know if anyone has built Python in Windows 95/98/NT
as a static lib instead of a dll.  If anyone has I would like to know how
you set up your project workspace.  I am using MS Visual C++ 6.  I have
looked at the Python workspace included in the Win32 version of Python but
that is set up to build the .dll.  As far as I know there is no way to
change that.  I'd have to make a new project and add all the necessary files
but since they're all in different directories I'd like to know if anyone
has done this and has a project file I could use to simplify this.  Any help
would be greatly appreciated.

Thank you,

Alberto
alberto at totallygames.com





From arcege at shore.net  Thu Jul 15 17:36:58 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 15 Jul 1999 21:36:58 GMT
Subject: Embedding Python in C
References: <7mi084$r80$1@kopp.stud.ntnu.no>
Message-ID: <_vsj3.138$ui4.39372@news.shore.net>

In comp.lang.python Thomas R*berg <thomar at ifi.ntnu.no> wrote:
: I am writing a program in C/C++ and want to embed Python. The one problem I
: encounter is that when I try to import a module using
: Pyrun_SimpleString("import mymodule\n"), this module is not found. I have
: specified the correct paths at the correct places in my Makefile (at least I
: think I have). Anyone had problem with this??

It might be that the sys.path is not set correctly in the C program
before calling the PyRun_SimpleString() function.  Did you insert your
library directory into the path before attempting to import your module?

  PyRun_SimpleString("from sys import path; path.insert(0, 'Mydir')");

  -Arcege




From kosullivan at snet.wit.ie  Mon Jul 26 06:02:41 1999
From: kosullivan at snet.wit.ie (Kieran O' Sullivan)
Date: Mon, 26 Jul 1999 11:02:41 +0100 (IST)
Subject: CGI ImportError
Message-ID: <Pine.LNX.4.10.9907261044180.72-100000@snet.wit.ie>

I am using the BSCW software on a S.u.S.E 5.3 Linux server and every time
I try to use one of the python scripts I get the following error in the
apache error log.

Could not find platform independent libraries <prefix>
Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
'import exceptions' failed; use -v for traceback
Warning!  Falling back to string-based exceptions
'import site' failed; use -v for traceback
Traceback (innermost last):
  File "wrapped", line 2, in ?
    import os, sys
ImportError: No module named os
[Mon Jul 26 09:23:49 1999] [error] Premature end of script headers:
/BSCW3/BSCW

I thought that there was some thing wrong with os and sys so I wrote and
successfully compiled the following program.

#!/usr/bin/python
import os, sys
print 'test'

I have never used python before so this test program might not be worth
a whole lot but it seems to be importing os and sys without any
difficulty.

My python directory is Python-1.5.2 on the / directory.
There is a soft link in the /usr/bin directory.
Python is in the path.
I downloaded the lot form the www.python.org web site.





From 02rmm at williams.edu  Wed Jul 14 17:28:31 1999
From: 02rmm at williams.edu (Robert McGehee)
Date: Wed, 14 Jul 1999 17:28:31 -0400
Subject: Reading from a COM port?
References: <7me6ci$5dg$1@nnrp1.deja.com>
Message-ID: <7mivdm$75e$1@news.duke.edu>

Yes, I'm curious about this also. How does one access data from a com port
and then manage the data through python?

Robert

<achatham at hotmail.com> wrote in message news:7me6ci$5dg$1 at nnrp1.deja.com...
What would be a good way to access a COM port in Solaris? I'm trying to
read data from a temperature probe and filter it. Can I just access the
device as a file or is there a more elegant solution? Sorry, I'm just a
beginner at Python. Any help would be appreciated.

Thanks,
Andrew





From ergo at merlin.in-berlin.de  Wed Jul 28 14:16:09 1999
From: ergo at merlin.in-berlin.de (Olaf Foellinger)
Date: Wed, 28 Jul 1999 18:16:09 GMT
Subject: Problems building Python1.5.2 for OpenStep
Message-ID: <FFLEqx.5v3@merlin.in-berlin.de>

Hi,

I'm writing this without much knowledge of python but in the hope to get 
things running well.

As a tool for another project (Zope)  I've tried to compile python 1.5.2  for 
OpenStep. I've configured using 

$ ./configure --without-gcc --with-dyld

without any errors. Make brought the first error

cc  -g -O2 -OPT:Olimit=0 -I./../Include -I.. -DHAVE_CONFIG_H -c 
./posixmodule.c
./posixmodule.c: In function `posix_ttyname':
/posixmodule.c:613: warning: assignment makes pointer from integer without a 
cast
./posixmodule.c: In function `posix_fsync':
./posixmodule.c:656: `fsync' undeclared (first use this function)
./posixmodule.c:656: (Each undeclared identifier is reported only once
./posixmodule.c:656: for each function it appears in.)
make[1]: *** [posixmodule.o] Error 1
make[1]: Leaving directory `/private/tmp/O_py152/Python-1.5.2/Modules'
make: *** [Modules] Error 2

which I've solved by inserting the line 

extern int fsync(int);

in  Modules/posixmodule.c. Including <libc.h> didn't help because of 
conflicting types for 'chown'. After that everything compiled fine.

Now the test: 

make test

resulted in

test_cmath
/bin/sh: 7411 Bus error
make: [test] Error 138 (ignored)
PYTHONPATH= ./python ./Lib/test/regrtest.py

gdb showed, that the error occurs in the last return statement of

Py_complex
PyComplex_AsCComplex(op)
	PyObject *op;
{
	Py_complex cv;
	if (PyComplex_Check(op)) {
		return ((PyComplexObject *)op)->cval;
	} else {
		cv.real = PyFloat_AsDouble(op);
		cv.imag = 0.;
		return cv;  /* error occurs here */
	}   
}
 
After removing test_cmath.py python doesn't crash any more but fails with two 
trsts:

test test_strftime failed -- Writing: 'Conflict for %j (julian day 
(001-366)):', expected: ''
test test_time failed -- Writing: 'time.mktime(time.localtime(t)) <> t', 
expected: ''

Any hints solving these problems are welcome.

Greetings Olaf





From Voynow  Thu Jul  1 01:10:43 1999
From: Voynow (Voynow, Sasha)
Date: Thu, 01 Jul 1999 06:10:43 +0100
Subject: wddx wrapper?
Message-ID: <0000163b@bossar.com.pl>

From: Voynow, Sasha <anonymous at web.remarq.com>


If you're in a Windows environment you can use Python's COM
support to to use the COM object they offer.

Sasha Voynow

sasha at ofoto.com



**** Posted from RemarQ - http://www.remarq.com - Discussions Start Here (tm)
****



From fredrik at pythonware.com  Thu Jul 15 20:18:16 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 16 Jul 1999 00:18:16 GMT
Subject: Oneliners?
References: <378D99C4.BDACE15B@serop.abb.se> <c75iu7ljyfu.fsf@pc142.cosc.canterbury.ac.nz>
Message-ID: <001d01becf20$b4c36540$f29b12c2@secret.pythonware.com>

Timothy R Evans <tre17 at pc142.cosc.canterbury.ac.nz> wrote:
> Shouldn't be too hard, if you count:
> 
> import SimpleHTTPServer
> 
> as only one line!

better make that:

import SimpleHTTPServer; SimpleHTTPServer.test()

(don't ask why, but that server produced something
that caused my browser to completely lose it...)

</F>




From jbauer at rubic.com  Thu Jul 22 22:05:46 1999
From: jbauer at rubic.com (Jeff Bauer)
Date: Fri, 23 Jul 1999 02:05:46 GMT
Subject: Non-ODBC MS Sql Connect
References: <3797B848.74D04740@rl.gov>
Message-ID: <3797CDFA.C2B95CC@rubic.com>

RJR,

1. Grab mysql-python151-win32.zip from /pub/crew/jbauer
via anon ftp at starship.python.net.

2. Put libmySQL.dll somewhere in your PATH, and put MySQL.pyd
somewhere Python will find it -- e.g. the same directory
python.exe is possible.

3. It works okay with 1.5.1 or higher.  For more info
about the MySQL library, download the source from the
Contrib area at python.org.

>>> import MySQL
>>> DBH = MySQL.connect('myhost','mydatabase','mypassword')
>>> DBH.listdbs()

[['mydb1'], ['mysql'], ['test']]

Best regards,

Jeff Bauer
Rubicon, Inc.

>> How to do I make a non-odbc connection to a MS SQL Server?
>> Am running Python Windows NT/IIS.



From avv at quasar.ipa.nw.ru  Fri Jul  9 12:27:09 1999
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Fri, 09 Jul 1999 09:27:09 -0700
Subject: Python and Paradox
References: <7m2mrj$hes$1@nnrp1.deja.com> <7m2o2b$aq0@dfw-ixnews12.ix.netcom.com> <37852804.54D5D2AD@quasar.ipa.nw.ru> <7m4q97$dpq$1@m2.c2.telstra-mm.net.au>
Message-ID: <7m9tmh$6qg$29@rks1.urz.tu-dresden.de>

Mark Hammond wrote:

> Alexander V. Voinov wrote in message <37852804.54D5D2AD at quasar.ipa.nw.ru>...
> >Aahz Maruch wrote:
>
> >> >I am looking for a way to access Paradox tables from Python without
> >> >using ODBC. Does anyone have any ideas?
> >>
> >> Nope, from an ex-Paradox tech support engineer.
>
> Yep, from someone who has :-)  Well, Im probably not technically correct,
> but if you:
> * Install MSAccess
> * Install Paradox drivers from the "Value Pack" (on the CD) - I think this
> does install an ODBC driver even tho we dont use it directly (hence my "not
> technically correct")

Another interesting question, for those who can check it: Does this Paradox DDL
also expose those older direct API, for those who still remember them? If this
engine was _not_ rewritten from the scratch, but just licensed, this would be
possible. But anyway, it's too slow even on the direct level. If one likes to
work with a C API database engine, I'd suggest to purchase Raima Data Manager.
Some time ago I moved a large office application from Paradox to db_Vista
relational submodel and gained great performance benefit. Certainly, I made
Python wrappers for a resonably higher level abstraction over its plain C API.
And RDM's network model would give you even more benefit if you realize how to
use it properly. (I speak it for myself, it's just an experience...)

Alexander





From ivanlan at callware.com  Fri Jul 16 08:18:12 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 16 Jul 1999 12:18:12 GMT
Subject: I envy newbies
References: <000201becf46$d79e78e0$3fa02299@tim>
Message-ID: <378F2304.ACC1AF7E@callware.com>

Hi All--

Tim "The Insufferably Right" Peters wrote:
> 
> [Ivan Van Laningham, making an IDLE desktop shortcut under Windows]
> > ...
> > (I quote from _Teach Yourself Python in 24 Hours_ (forthcoming from
> > SAMS)):
> 
> Monday <wink>?
> 

Sure.  I just won't guarantee you what month or year;-)

> > ``The first thing you need to do is to find where IDLE lives, and to do
> > that you need to know where Python lives;
> > [instructions for navigating the directory tree in Explorer until
> >  finding idle.pyw]
> 
> Try this:  Start -> Programs -> Python 1.5 then rightclick+drag the IDLE
> entry to the desktop.  Everything else:
> 

But it _is_ a good idea to know where Python lives!

> > Right click ... and select "Create Shortcut" on the popup menu ...
> 
> is the same, but this avoids any mucking with Explorer.  Could be you need
> the Active Desktop enabled for this to work (but it's your book, so I'm not
> going to take the time to disable the Active Desktop to check that out for
> you <wink>).
> 

I suspect you do need the Active Desktop enabled, because it sure don't
work for me.  The Active Desktop is a metaphor best left unexplored ...
although I admit that every time I see one of these things "working," I
tend to think of Korak the Killer's interminable journey through the
endless quagmires surrounding the land of Pal-ul-Don. ... ;-)

> > The installer does modify your autoexec.bat to include the Python path
> > on W95/98.
> 
> Eh?  It never touched autoexec.bat on my Win95 -- and darned glad it didn't.
> That's a dangerous thing for any installer to presume to do.
> 

Oh, poo, Timmy.  Everything else under the M$ sun mucks with your
autoexec.bat file, why shouldn't Python?  "Backups are for p'takh!"
--Klingon Programmer Aphorism;-)

Besides, I don't think IDLE will work if you don't have Tcl in your
path, and if you have to add that, why not add Python?  Well, obviously
I should go forth and re-format my hard drive and re-install W98 and
then re-install Python. ...  But I really don't feel like a little root
canal today, so I'll take your word for it that my autoexec.bat file is
left virgin and pristine by Python's installer.

> The good news is that new users introduced via IDLE have no immediate need
> to worry about the DOS path -- the IDLE shortcut knows how to invoke Python
> regardless.  In fact, I've been known to use Python from time to time
> <wink>, and
> 

Gee, not even a *qualified* wink!

> C:\temp>python
> Bad command or file name
> 
> C:\temp>
> 
> That is, I still haven't bothered to put python.exe on my PATH!  Although I
> do have an MS-DOS desktop shortcut with D:\PYTHON as its working directory
> and "doskey /INSERT" as its "batch file".
> 

I still maintain that the installer *ought* to modify your path; really,
that should be safe enough, and, if you don't want to put all your
pythons in one basket, necessary.

> can-just-feel-the-linux-folks-melting-with-jealousy<wink>-ly y'rs  - tim

"Yearning for NT when you're running Linux is like lusting for a Big
Wheel when you're driving a BMW." --Harrison Picot.

<peddling-like-mad>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From viofis at my-deja.com  Thu Jul  1 20:33:35 1999
From: viofis at my-deja.com (viofis at my-deja.com)
Date: Fri, 02 Jul 1999 01:33:35 +0100
Subject: please help with simple regexp
Message-ID: <00001646@bossar.com.pl>

From: viofis at my-deja.com

I'm a beginner at Python, so this may be a stupid
question:

I have a string, for example, myString="the
Potato".
1. Convert it to all lowercase ("the potato")
2. Match it against a list of strings -
"potato,tomato,corn,etc...", extract "potato" (if
it matches with something on the list), then save
into myNewString.

Any help is greatly appreciated.

thanks,
Vadim



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From poinot at onera.fr  Wed Jul 21 08:50:39 1999
From: poinot at onera.fr (Marc POINOT)
Date: Wed, 21 Jul 1999 14:50:39 +0200
Subject: gadfly RDBMS ???
References: <Pine.LNX.4.04.9907211629470.12211-100000@emerald.netskate.ru>
Message-ID: <3795C21F.A0F13648@onera.fr>

Oleg Broytmann wrote:

> On Wed, 21 Jul 1999, Darko Cvagic wrote:
> > What does it means that GADFLY is Relational ?
>
>    Because Gadfly is almost Yet Another SQL Engine (not that I am against
> Gadfly or SQL).
>
> > How can I make relations between tables in database?
>
> SELECT t1.bor, t1.belt, t2.der
>    FROM table1 t1, table2 t2
> WHERE t1.plux = t2.flax
>

SQL is a query langage, OSQL is an extension to OODBMS (Object oriented).

The "relationnal" model is an old model of data [1] but it as a theorical
background
with set theory, and it is widely spread in the industry.
Say you have a set of tables, each table can be viewed as
the subset of the cartesian product of to sets.

For more info, I think you can read some good books about
DBMS history, such as [2].

[1] The Entity-Relationship model - toward a unified view of data
P.P.-S. Chen, MIT, Readings in Database systems, M.Stonebraker,
Morgan Kaufmann, 1988
[note: paper is from ACM 1976]

[2] Principle of database systems - J. D. Ullman, Computer
science press1982

Marcvs [alias Hum, I'm quite pleased about this pedantic and obscure answer. ]





From olipt at mayo.edu  Fri Jul  2 00:16:00 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Thu, 1 Jul 1999 23:16:00 -0500
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
In-Reply-To: <14204.4556.177837.851763@lrz.de>
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net> <377B711A.E8B65E0B@cfdrc.com> <377BC620.F5553FF5@easystreet.com> <b%Pe3.4033$lr3.164497@news2.jacksonville.net> <14204.4556.177837.851763@lrz.de>
Message-ID: <Pine.LNX.4.10.9907012309060.6263-100000@us2.mayo.edu>

This has been a very interesting discussion on GUI's.

It struck me that all this talk of a cross-platform GUI might be
preventing a lot of very good GUI's from being used.

So, I looked in to PyGtk and PyGnome (with its Gnome Canvas) and I'm
impressed.  Right now it fits what I want to do very well and it works on
systems with an X server (aren't X servers available for MAC's and
Windows?)  So, I'm happy with it. If you're in the same situation you
might like it too.  I tend to agree that there is plenty of room for more
than one GUI.

Travis Oliphant








From pserrand at worldnet.fr  Fri Jul  2 04:29:45 1999
From: pserrand at worldnet.fr (pserrand)
Date: Fri, 02 Jul 1999 10:29:45 +0200
Subject: test - ignore
Message-ID: <377C7879.428A3E19@worldnet.fr>

test with python




From daniels at mindspring.com  Sat Jul  3 00:51:47 1999
From: daniels at mindspring.com (Alan Daniels)
Date: 3 Jul 1999 04:51:47 GMT
Subject: Python suitable for a game engine?
References: <slrn7nlpuk.eu.daniels@localhost.localdomain> <377B7E17.5BD7@nospam.com> <slrn7noa2j.hg.daniels@localhost.localdomain> <377C9904.41AB@nospam.com>
Message-ID: <slrn7nr5uq.qi.daniels@localhost.localdomain>

On Fri, 02 Jul 1999 11:48:36 +0100, the infinitely wise Alex Maranda
(amaranda at nospam.com) spoke forth to us, saying...

[snip...]
>Alan Daniels wrote:
>> I have the (VERY!) preliminary code at:
>>     http://www.mindspring.com/~daniels/engine/index.html
>> if you're interested, although there isn't much to look at.
>I'm downloading it now. You can use negative version numbers;
>SmallEiffel is now at version -0.78. They'll stop when they'll reach 0.0
>:-)

In that case, I'm going to restart with negative numbers.
I was just going to call it the "Zebra" engine, but that name is
taken. So, I'm going to rename it to the "Zeroth" engine. Seems
appropriate, considering the state its in. :=)

-- 
=======================
Alan Daniels
daniels at mindspring.com
daniels at cc.gatech.edu



From jam at newimage.com  Tue Jul 13 18:18:41 1999
From: jam at newimage.com (jam)
Date: Tue, 13 Jul 1999 18:18:41 -0400
Subject: HTML parser/reformatter/filter wanted
In-Reply-To: <MPG.11f5b895bd447488989685@news.mop.no>; from Alexander Staubo on Tue, Jul 13, 1999 at 09:57:29PM +0200
References: <MPG.11f5b895bd447488989685@news.mop.no>
Message-ID: <19990713181841.E11617@toast.internal>

On Tue, Jul 13, 1999 at 09:57:29PM +0200, Alexander Staubo wrote:
> 
> I'm looking for a piece of Python code that can filter HTML and tie up 
> unclosed tags.
> 
> By way of explanation: I'm building a bulletin-board-ish solution where 
> users should be able to post HTMl-formatted submissions; I only want to 
> enable a subset of HTML (eg., <b>, <i> and so forth), plus I want to make 
> sure text-level tags are ended correctly. I could sit down and write a 
> smaller HTML parser, but I don't want to reinvent wheels if I can just 
> grab this stuff off the shelf.
> 

greetings,

I'm not sure if it will do *exactly* what you want, but a good starting
point is probably 'htmllib.py' in the standard distribution. it will at
least give you something to build on.

regards,
J
-- 
|| visit gfd <http://quark.newimage.com/> 
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From mikael at toolwood.com  Fri Jul  2 01:51:43 1999
From: mikael at toolwood.com (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 05:51:43 GMT
Subject: Python GUIs: Portability or not?
References: <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <3A174E56AD89F7F6.47941772B9D545B9.852EB0CCFCA68614@lp.airnews.net> <377B711A.E8B65E0B@cfdrc.com> <377BC620.F5553FF5@easystreet.com> <b%Pe3.4033$lr3.164497@news2.jacksonville.net> <14204.4556.177837.851763@lrz.de> <Pine.LNX.4.10.9907012309060.6263-100000@us2.mayo.edu>
Message-ID: <5DEB4D7CBF56EC8D.37F29301ECE96FC0.D3C7C2F2D9068D65@lp.airnews.net>

On Thu, 1 Jul 1999 23:16:00 -0500, Travis Oliphant <olipt at mayo.edu>
wrote:

>It struck me that all this talk of a cross-platform GUI might be
>preventing a lot of very good GUI's from being used.

Well, it probably is.  But, I, for one, have this fobia about
portability - it's so easy to give up and so hard to get back :)  My
experience is that the moment you give up portability, you give up
lots of chances to verify the architecture and implementation of your
application.  No chance to compile with, say, five different compilers
on different platforms and benefit from the fact that the XXX compiler
is so great about detecting unreferenced variables in switch
statements, while the YYY compiler catches memory allocation errors in
a brilliant way.  No chance to try it out on your Y box, because
there's such a nifty profiler.  No chance to check if your binary file
formats are sound by exchanging data files between little-endian and
big-endian machines.  All in all, I think portability is a much bigger
thing that the ability to compile and run your app on some obscure
machine nobody you know has ever seen in real life. 

(and portability also makes it possible to distinguish the pro from
the amateur, in these times of GUI IDEs with wizards and all :-)

>So, I looked in to PyGtk and PyGnome (with its Gnome Canvas) and I'm
>impressed.  Right now it fits what I want to do very well and it works on
>systems with an X server (aren't X servers available for MAC's and
>Windows?)  

I don't know about the Mac, but I'm sure there are X servers available
for NT.  However, not like they are available for Linux and other Unix
boxes - you don't just download the X11 source, compile it, and that's
it (unless a lot has changed the last few years).

You'd probably have to buy a commercial X server for NT or, even
worse, ask your users to do so.

>So, I'm happy with it. If you're in the same situation you
>might like it too.  

Well, I'm happy that you're happy!  

Even though I probably won't even get to need the Unix portability, I
still sleep much better if I know that I don't have to discard by GUI,
the day I'm asked to do a Linux port.  And, then again, if the GUI is
isolated cleanly enough from the application, unlike Tk's
implementation (...), it wouldn't be such a big task to switch GUI. 

However, for many projects I agree with you - portability is not a big
issue.  For instance, I've been considering to do a decent Microsoft
Visual SourceSafe client (a  Win32-specific revision control system)
in Python.  In such a case, there would be no need whatsoever for
portability outside Win32 (i.e., between Windoze 9x and Windows NT).

Gotta think this over for a while, because you're probably more right
than I'd like to think.  Anyway, it appears that the best candidates,
for my needs, wxWindows and FOX (which I haven't evaluated yet), are
both portable to Win32 and a lot of other platforms. 

Can PyGTK be used on Win32 without an X server?

>I tend to agree that there is plenty of room for more than one GUI.

And, luckily, there are more than one!


-- Mikael



From nbecker at fred.net  Tue Jul 20 10:39:38 1999
From: nbecker at fred.net (nbecker at fred.net)
Date: 20 Jul 1999 10:39:38 -0400
Subject: python vs. lisp (whereis quote?)
References: <x88r9m3cyje.fsf@rpppc2.hns.com> <lbzk8rva3to.fsf@morgoth.uio.no>
Message-ID: <x88zp0rbd1x.fsf@rpppc2.hns.com>

>>>>> "Rolf" == Rolf Marvin B?e Lindgren <roffe at morgoth.uio.no> writes:

    Rolf> [ nbecker
    Rolf> | One feature of lisp (emacs lisp actually) that I think is pretty cool
    Rolf> | is the "hooks".  A hook is basically unevaluated code that will be run 
    Rolf> | by eval later.
    Rolf> |
    Rolf> | I suspect python lacks this capability.  As I understand it, the only
    Rolf> | similar capability is "eval".  But I think eval expects a string,
    Rolf> | which it will then parse.  While the capability is similar, having to
    Rolf> | parse a string could be inefficient.
    Rolf> |
    Rolf> | Am I correct that what is missing is the ability to postpone
    Rolf> | evaluation using "quote"?

    Rolf> in Lisp, there is no syntactic difference between code and data, and
    Rolf> that is what makes the hook mechanism possible the way it is implemented
    Rolf> in Emacs.  if Python had a way of representing code as data, Python
    Rolf> could have had the same support for hooks. 

    Rolf> if eval expects a string, and if the string can contain code, then you
    Rolf> have it, haven't you?

Sort of - but it has to be parsed - so it may not be very efficient.



From tismer at appliedbiometrics.com  Sat Jul 17 15:20:18 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Sat, 17 Jul 1999 19:20:18 GMT
Subject: Snippet: bitcount()
References: <000001bed082$726b19a0$27a22299@tim>
Message-ID: <3790D772.1A19AD82@appliedbiometrics.com>


Tim Peters wrote:
> 
> This has been beaten beyond death so often I don't have the heart to do it
> all again <wink>.  Some factoids:
> 
> + Christian's post summarizes best efforts pre-1.5.2.  For popcount4 it's
> missing the nbits function.  Here's an elegant one that's reasonably fast:
> 
> def nbits(x):   # x >= 0 assumed
>     if x <= 2:
>         return x
>     sum = 1
>     while x:
>         shift = 1
>         x = x2 = x >> 1
>         while x2:
>             x = x2
>             sum = sum + shift
>             shift = shift + shift
>             x2 = x >> shift
>     return sum

Sorry, I forgot this one. It is exactly what was in my file.

> + About once a year someone gets excited about providing a pile of fast,
> basic longint ops (via extension module or adding methods to longs -- like
> popcount, trailing zero count, parity, nbits, long <-> int vector
> conversions, set/clear/test/invert bit number i, find closest 0/1 bit
> before/after bit number i).  Unfortunately, nobody has persisted long enough
> to write a line of code, perhaps because people take such delight in endless
> argument about what to name things <0.1 wink>.

If someone should consider it at all:
Add a fast multiplication (with O(1.5)) or I won't look at it.

> the-next-person-to-get-excited-should-keep-it-a-secret-until-the-code-
>     is-released-ly y'rs  - tim

But I'm willing to help on the fast stuff via private emails :-)

more-code-less-words-ly y?rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     saving one multiply by spending a few adds is the trick
       no I don't mean Schoenhage-Strassen here.



From olivier.deckmyn at mail.dotcom.fr  Mon Jul 19 10:08:17 1999
From: olivier.deckmyn at mail.dotcom.fr (Olivier Deckmyn)
Date: Mon, 19 Jul 1999 16:08:17 +0200
Subject: [Q] Accessors...
Message-ID: <7mvbkj$pl$1@feed.teaser.fr>

HI again...

Is there a way to do the following :

class MyClass :
    def __init__(self):
        self._myAttr = None

    def get_myAttr(self):
        return self._myAttr

    def set_myAttr(self, value)
        self._myAttr = value # might do more interresting checkings for
examples


and then use this like that:

m=MyClass()
m.myAttr="foo"
print m.myAttr

This way, because it exists two methods get_XXX and set_XXX, a special
proporty XXX would be implicitly created, and would make the user call
"automatically" the get/set methods.

(Please note that myAttr is defined nowhere, but is the "guessed" word found
after methods get_XXX and set_XXX)

I other words, would be AUTOMATICALLY "understood" by the interpreter as :

m=MyClass()
m.set_myAttr("foo")
print m.get_myAttr()

This is what I call "accessors" : a way to make the user use automatically
methods to access variables ?
Such a thing is a "property" in Delphi for example.


Thanx all for your support.





From Lyle.Johnson at p98.f112.n480.z2.fidonet.org  Fri Jul  2 03:51:06 1999
From: Lyle.Johnson at p98.f112.n480.z2.fidonet.org (Lyle Johnson)
Date: Fri, 02 Jul 1999 08:51:06 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001683@bossar.com.pl>

From: Lyle Johnson <jlj at cfdrc.com>

Yeah, I got it wrong the first time. It should read:

    http://www.cfdrc.com/FOX/fox.html

I confused it with the ftp URL, which does need the "pub" in there.

Greg Ewing wrote:

> Kevin Dahlhausen wrote:
>
> > Lyle Johnson <jlj at cfdrc.com> wrote:
>
> > >Just to add fuel to the fire, we now have a Python interface to FOX, a
> > >cross-platform GUI toolkit:
>
> > >    http://www.cfdrc.com/pub/FOX/fox.html
>
> This link doesn't work for me. Has
> it been moved somewhere else?
>
> Greg



From wolford at enews.nrl.navy.mil  Thu Jul  8 12:24:44 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Thu, 08 Jul 1999 12:24:44 -0400
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com> <3784BD4B.7938B7BA@enews.nrl.navy.mil> <3784CE0E.6BD46E97@enews.nrl.navy.mil>
Message-ID: <3784D0CC.C8130207@enews.nrl.navy.mil>

I'm beginning to question my sanity. Try this:

#include "Python.h"
#include <stdio.h>
int main(int argc, char **argv)
{
//  return Py_Main(argc, argv);
  FILE *file = fopen(argv[1], "rt");
   Py_Initialize();
  PyRun_AnyFile(file, argv[1]);
}





From kperacles at geocities.com  Fri Jul 23 03:32:27 1999
From: kperacles at geocities.com (Karl Putland)
Date: Fri, 23 Jul 1999 01:32:27 -0600
Subject: COM bites - help please
Message-ID: <eZUl3.1080$wm4.135734@news.uswest.net>

I've got a couple of pyCOM things that are kicking my @$$! 

1.  Trying to wrap a list of instances into a util.Collection.  All instances have been prepped with _public_methods_ and _public_attrs_ and have been wrapped with a call to win32com.server.util.wrap(field) Trying to access the Collection from a com instance produces the following error.

>>> gm=win32com.client.Dispatch("Python.GMConnector")
>>> c1=gm.GMTable('contact1')
getting table=gmc.GMConnector.GMTable(self, TableName) <gmCOM.GMcomConnector instance at 9373c0> contact1

['lower', 'upper']
Wrapping Table

Looping over fields to set _public_attrs_ for table D:\GOLDMINE\DEMO\CONTACT1.DBF

Wrapping Fields
>>> c1.top()
>>> c1.contact        # This is a GMField instance
<COMObject <unknown>>
>>> c1.contact.read()
'Karn Smith'
>>> c1.fields        # This is a util.Collection(list of GMField instances)
Traceback (innermost last):
  File "<pyshell#56>", line 1, in ?
    c1.fields
  File "D:\Python\win32com\client\dynamic.py", line 394, in __getattr__
    raise pythoncom.com_error, details
com_error: (-2147467259, 'Unspecified error', None, None)


2. A computed attribute in __getattr__ does not want to be accessed through com.  It is in the _public_attrs_  list but produces the following error when attempting to access it.

>>> c1.contact.value
Traceback (innermost last):
  File "<pyshell#55>", line 1, in ?
    c1.contact.value
  File "D:\Python\win32com\client\dynamic.py", line 400, in __getattr__
    raise AttributeError, "%s.%s" % (self._username_, attr)
AttributeError: <unknown>.value

3. PyUnicode objects bit me in the ass when strings are passed into methods and manipulated.
class Foo:
    <COM stuff>
    def bar(self, somestring):
        somestring = string.upper(somestring)
        return somestring

>>>foo = win32com.client.Dispatch("Python.Foo")
>>>foo.bar("Hello World")

Some traceback about TypeError PyUnicode.

Causing a need to change bar to
    def bar(self, somestring):
        somestring=str(somestring)
        somestring = string.upper(somestring)
        return somestring
This change needs to be made in ALL methods that accept and manipulate strings???


Is there a better method than COM for allowing someone who uses VB access to Python objects, short of converting all of the VB users to Python or resorting to reimplementing in VB or C++?  The COM interface I've been able to expose is sufficient for now but does not allow some of the access that I have in Python.  Maybe that's good in the end?

The work I've done in Python is short and sweet.  I hate the thought of having to redo it all Not that the Python stuff would be thrown out,  I just want to release some cool stuff to other GoldMine folks before I am finished with the complete application.

Any help appreciated.

Karl Putland
kperacles at geocities.com





From fdrake at cnri.reston.va.us  Mon Jul 12 14:40:41 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Mon, 12 Jul 1999 14:40:41 -0400 (EDT)
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
In-Reply-To: <14218.13417.796678.728859@bitdiddle.cnri.reston.va.us>
References: <3787D36A.815B7C29@ssec.wisc.edu>
	<3787D8DA.3774AA7D@ssec.wisc.edu>
	<7m8p7t$o7s@dfw-ixnews10.ix.netcom.com>
	<3787FB51.12FE0048@ssec.wisc.edu>
	<7m923i$dem@dfw-ixnews5.ix.netcom.com>
	<14217.63931.75257.566266@weyr.cnri.reston.va.us>
	<7md2ip$19q2$1@nntp6.u.washington.edu>
	<14218.7724.879301.831590@weyr.cnri.reston.va.us>
	<14218.13417.796678.728859@bitdiddle.cnri.reston.va.us>
Message-ID: <14218.13993.593282.109454@weyr.cnri.reston.va.us>

Jeremy Hylton writes:
 > By vindicated, I assume you mean that you and Donn are both twisted
 > old Unix geeks <wink>.

Jeremy,
  Works for me!  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From gregm at iname.com  Wed Jul  7 03:32:07 1999
From: gregm at iname.com (Greg McFarlane)
Date: Wed, 7 Jul 1999 17:32:07 +1000
Subject: Problem with BLT busy cursor on NT
Message-ID: <19990707173207.62192@nms.otc.telstra.com.au>

Has anyone noticed that, under some circumstances, the BLT busy cursor
is not displayed under NT?  The (tcl) code below demonstrates this. 
(The equivalent python has the same problem, so it is not a problem
with python, but someone on that mailing list may know that answer.)

Clicking on the top "Show busy" button changes the text of the button
as well as displaying the busy cursor for two seconds, as expected. 
Under Unix, clicking on the second button does the same thing, but
under NT, it only changes the text of the button for two seconds and
does not display the busy cursor.

It seems that, under NT, to actually get the busy cursor displayed,
the interpreter has to return to the main loop after a call to
blt::busy.  If there is a long lasting calculation or a blocked read,
etc, then the cursor will not be not changed.  I would have thought
that the call to "update idletasks" would have been enough to flush
out the cursor change.

Has anyone seen this before?  Is it a bug in BLT or a limitation
of NT?

==================================================================
package require BLT

proc startBeingBusy {} {
    .b1 configure -text "I am busy..."
    ::blt::busy hold .
    update idletasks
    after 2000 stopBeingBusy
}

proc stopBeingBusy {} {
    ::blt::busy release .
    .b1 configure -text "Show busy"
}

proc beBusy {} {
    .b2 configure -text "I am busy..."
    ::blt::busy hold .
    update idletasks
    after 2000
    ::blt::busy release .
    .b2 configure -text "Show busy"
}

button .b1 -text "Show busy" -command startBeingBusy
pack .b1 -padx 40 -pady 40

button .b2 -text "Show busy" -command beBusy
pack .b2 -padx 40 -pady 40
==================================================================

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From bruce_dodson at bigfoot.com  Thu Jul 29 23:24:09 1999
From: bruce_dodson at bigfoot.com (Bruce Dodson)
Date: Fri, 30 Jul 1999 00:24:09 -0300
Subject: any way to know when the program is exiting?
References: <199907290135.SAA22743@hpcll563.cup.hp.com>
Message-ID: <7Z8o3.42069$jl.29810418@newscontent-01.sprint.ca>

It sounds like your class here relies on the existance of some other object
to help it clean up.  Presumably this same object is required to help in
your class's __init__ method, and presumably it is not OK to just skip the
cleanup on the way out for some reason.  In a case like this, I think it
would be safe to store a reference to that "global" as part of your
classes's __init__, as long as the "global" doesn't retain a reference to
your class instances.  Then you can guarantee that the "global's" reference
count won't hit zero until all the objects that need it have been deleted,
even if it is no longer a "global" by that time, so your class can send
messages to the "global" through the reference that it owns.




From Tim.Peters at p98.f112.n480.z2.fidonet.org  Thu Jul  1 03:53:56 1999
From: Tim.Peters at p98.f112.n480.z2.fidonet.org (Tim Peters)
Date: Thu, 01 Jul 1999 08:53:56 +0100
Subject: How to obtain an instance's name at runtime?
Message-ID: <0000161e@bossar.com.pl>

From: "Tim Peters" <tim_one at email.msn.com>

[Dinu C. Gherman]
> ...
> Just that much: an object, that has not been assigned
> a name to, cannot exist (having a reference count of
> zero), can it (at least for user-defined objects)?

There must be a way to "get to" it, but access paths needn't be names; e.g.,

week = [Day("Sunday"), Day("Monday"), ...]

The object constructed by Day("Monday") was never given a name, and is
accessible only via the access path week[1] (or equivalent from the end of
the list).

anonymously y'rs  - tim







From mal at bewoner.dma.be  Fri Jul  2 13:46:38 1999
From: mal at bewoner.dma.be (Lieven Marchand)
Date: 02 Jul 1999 19:46:38 +0200
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> <wkzp1g2t2e.fsf@ifi.uio.no>  <14204.912.383867.589108@lrz.de> <slrn7noclt.429.cbbrowne@knuth.brownes.org>
Message-ID: <m3btdvhrlt.fsf@localhost.localdomain>

cbbrowne at news.brownes.org (Christopher B. Browne) writes:

> If it weren't that he's actually gotten truly useful results out of
> all of these things, he'd probably be considered "just another of the
> opinionated fools" that clutter up the Internet.
> 
> Note that the final item on that list is really crucial; he minimizes
> things to the max, and the rather bizarre results are indeed direct
> results of that philosophy.

The difference between Moore and "just another of the opinionated
fools" is indeed that he gets results. I'm more into Common Lisp these
days but "Thinking Forth" is a book that has radically changed many of
my ideas about programming. 

It's also interesting to see that he resists most of the additions
that have been made to "Standard Forth" in the last decade and still
gets the work done. Perhaps Forth and Lisp are really duals in the
language space with Common Lisp and Moore's minimalist Forth on the
extreme ends. Forth and Lisp (with Basic) are also the languages with
a lot of variations and differences (perhaps they should be called
language families?).

-- 
Lieven Marchand <mal at bewoner.dma.be>
If there are aliens, they play Go. -- Lasker



From mal at lemburg.com  Wed Jul  7 04:38:41 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Wed, 07 Jul 1999 10:38:41 +0200
Subject: Interfacing from Python to the ODBC world
Message-ID: <37831211.5B92EE38@lemburg.com>

Hi everybody,

as you might know, I am the maintainer of mxODBC, a Python C
extension capable of interfacing from Python to most ODBC drivers
and managers in a Python DB-API standard way:

	http://starship.skyport.net/~lemburg/mxODBC.html

Since mxODBC has been around for quite some time now, I would like
to start a small feedback query to find out about your experiences
with it and the setups you use it in. I intend to use this feedback
to promote the use of Python to do database interfacing and to
direct development into new directions.

Some questions:

1. Since when do you use mxODBC ?

2. In what setups does mxODBC work in your company, e.g. which
   databases do you interface to and on which platforms ?

3. If you have previously used the Win32 odbc module, have you had
   any significant troubles switching from the Win32 module to the
   mxODBC package ?

4. Did you know that mxODBC can help you interface to Oracle,
   MS Access, MS SQL Server, Solid, MySQL, Informix ?

5. Did you know that mxODBC provides access to iODBC, the Unix
   ODBC manager, and unixODBC, a new ODBC manager with GUI ?

6. Did you know that I am offering commercial support for
   mxODBC ?

7. Are you satisfied with the documentation ?

8. How troublesome do you find setting up mxODBC, i.e. adjusting
   paths and linking against the target ODBC libs ?

9. Which features would you like to see in future version ?

10. Would you be willing to pay for early release versions ?

The last question is of special importance to me, because the
mxODBC development and support is becoming a pretty time
consuming task. Keeping the package in sync with all the great
ODBC software out there takes many hours of downloading,
installing and testing the various setups.

That's why I plan to distribute future versions in two phases:
Only registered users get access to the latest version,
while all others can freely download the previous versions.
If this setup proves to be a success, I see a long future for
the mxODBC package.

Thanks for your time.

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   177 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From Stefan.Schwarzer at tu-clausthal.de  Wed Jul 14 19:08:47 1999
From: Stefan.Schwarzer at tu-clausthal.de (Stefan Schwarzer)
Date: Thu, 15 Jul 1999 00:08:47 +0100
Subject: OS/2 Python: editing in interactive use
References: <378C55A9.411BE9AB@tu-clausthal.de> <m3673nkuod.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <378D187F.554BB020@tu-clausthal.de>

Hello Michael

Thank you for your help!

Michael Hudson schrieb:
> Stefan Schwarzer <Stefan.Schwarzer at tu-clausthal.de> writes:
> > It seems to me that the OS/2 port of python lacks the
> > "sophisticated" features for input editing (history,
> > C-A, C-E, etc.)
> 
> Those features are provided by readline. I doubt readline works on
> OS/2, though you could have a look. A search on google turned up a
> reference to a Python readline module for win32, though the site seems
> to be off net, at least temporarily.

At least I found the GNU readline lib for the emx compiler for
OS/2 (on "Hobbes"). However, Python/2 is compiled with the IBM
Visual Age compiler, and I don't know how difficult to integrate
it is. BTW, I have no Visual Age for C. And I'm just new to python.

I have emx here, but I can't estimate how difficult it would be
to compile the Visual Age sources with it. I also have rather little
experience in porting. *sigh*

> > If I have overlooked a possibility to activate it, please
> > give me a hint. If not, how difficult would it be to extend
> > Python/2 appropriately? I would appreciate this very much. :)
> 
> Can you use idle? Most Python IDE's have some commandline editing
> features.

I have looked for IDLE, but it needs Tkinter, which is not 
available for (native) OS/2, as far as I know.

After all, I would like to use the "regular" python interpreter
interactively without the GUI stuff. ;-)

Regards
 Stefan



From kperacles at geocities.com  Mon Jul 26 23:41:12 1999
From: kperacles at geocities.com (Karl Putland)
Date: Mon, 26 Jul 1999 21:41:12 -0600
Subject: COM bites - Followup
References: <eZUl3.1080$wm4.135734@news.uswest.net>
Message-ID: <qY9n3.1074$yZ1.108243@news.uswest.net>

WOW what a pain.  Now I know and will share.

>From original post.

[I am frustrated]
> 1.  Trying to wrap a list of instances into a 
> util.Collection.  All instances have been prepped with 
> _public_methods_ and _public_attrs_ and have been wrapped 
> with a call to win32com.server.util.wrap(field) Trying to 
> access the Collection from a com instance produces the 
> following error.
> 
[...]
> >>> c1.fields        # This is a util.Collection(list of 
> GMField instances)
> Traceback (innermost last):
>   File "<pyshell#56>", line 1, in ?
>     c1.fields
>   File "D:\Python\win32com\client\dynamic.py", line 394, in 
> __getattr__
>     raise pythoncom.com_error, details
> com_error: (-2147467259, 'Unspecified error', None, None)
> 

Here is what I was doing.  The following function was wrapping up an object that contains other objects and a list of instances.

def gm_com_table_prep_and_wrap(table_inst):
    ...
    [other table_inst prep]
    ...
    tmpfields=[]
    for field in table_inst.fields:
        field._public_methods_ = ['read',
                                  'write',
                                  'append']
        field._public_attrs_ = ['name',
                                'size',
                                'type'
                                'value']
        name = string.lower(field.name)
        field = win32com.server.util.wrap(field)
        tmpfields.append(field)
        setattr(table_inst, name, field)

    # turn tmpfields into a collection
    tmpfields=win32com.server.util.Collection(tmpfields,1)
    setattr(table_inst, "fields", tmpfields)

    # Return a wrapped table
    wrappedtable=win32com.server.util.wrap(table_inst)
    return wrappedtable


Where this went wrong is that "win32com.server.util.Collection(tmpfields,1)" returns a Collection instance READY to be wrapped.  

So what finnally made the collection work from COM was the addition of ONE line

    # turn tmpfields into a collection
    tmpfields=win32com.server.util.Collection(tmpfields,1)
    tmpfields=win32com.server.util.wrap(tmpfields)    # This line mad it work!!!
    setattr(table_inst, "fields", tmpfields)

Maybe for collections there should be a Factory that returns a wrapped instance

    WrappedList=win32com.server.util.GetWrappedCollection(List, read_only=0)

I might do that here or now that I know I'll NEVER forget to wrap the Collection.

As for the other problems I had... The PyUnicode thing still suck but I understand it, and I haven't yet been able to hunt down the problem with exposing a computed attribute so there is a method that gets the attribute I want instead.

Thanks Bill, Gordon, and Mark.

Now if I can just understand the whole Type Library thing things will be hunky dorey.\

Finally-took-a-bite-out-of-COM-ly y'rs - Karl





From rodisi01 at my-deja.com  Wed Jul 21 09:41:57 1999
From: rodisi01 at my-deja.com (Isidor)
Date: Wed, 21 Jul 1999 07:41:57 -0600
Subject: Hello, Thnx, Projects & ?s TOC
Message-ID: <199907211245.HAA14587@cro.ots.ac.cr>

Hello Pythoneers-

I downloaded 1.5.2 a couple of weeks ago and have been having a
lot of fun with it ever since, um, that is to say, it has taken
over my brain and completely destroyed my capacity to do my work.
Though I have no formal training in CS, I have been dreaming up
software projects for quite a while, and I think I may actually
be able to tackle them with python. I am so excited about this
possibility that, as I mentioned above, it's hard for me to think
about anything else.

I have been scanning the comp.lang.python archives, and I am
amazed at how nice and helpful you all are. Cheers to you all!
You really are a good group of folks, and I hope that someday I
will be able to make a worthy contribution to the python world.

For now, I have some minor contributions to make, along with some
questions and ideas/comments. It's a fairly large tangle of
stuff, so I am going to break it up into separate messages, the
intent being to make it easier for people to find what is useful
for them and to enable question-answerers to address small chunks
of connected ideas rather than wade through a pile of other
stuff. I hope this is not a serious breach of news-group
etiquette.

Table of Contents (other messages):

1.   A Contribution:

     Tool for Newbz (& beyond?): ModuleWriter

          Extract import statements, function defs, and classes
          from a long and messy IDLE session and put them in a
          separate file sorted according to the order of their
          appearance in the original IDLE session.
     
2.   Another, more minor contribution:

     Win95(9x/NT?) file extension & association tricks

3.   A minor (and odd!) contribution:

     String2List.py

          Take a string that is written to *look* like a list
          "[['a', 'b'], '1', ['g', ['h', 'blah's "blah" blah']],
          '5']" but *isn't* a list, and turn it into a *real*
          list! Weird, huh?

4.   A very-minor contribution, and now start the questions:

     IDLE mini-hack & ?s

     a)   Inserting date/time-stamp at startup
     b)   ?: How to add commands to menus: details of an
          unsuccessful attempt
     c)   ?: How to execute a module (not import it) from within
          a python command shell (see problem workaround in
          "Win95 file extensions and associations tricks).

5.   Spreading the word about Python; ?s and ideas

     a)   MS-Office (i.e., vb-applications/word basic) macro-
          writers as a group of newbies who would eagerly &
          easily convert to python, with a little help.
     b)   Distribution of modules/packages on the big share/free-
          ware sites.


A question for this message:

I have noticed that not all the messages of this newsgroup are
archived in the python-list archives. (
http://www.python.org/pipermail/python-list/ ) Is this a known
discrepancy about which nothing can be done? (E.g., the only
messages archived are those that were sent to the mailinglist
while posts directly to the news group cannot be archived? I
don't think the evidence supports that interpretation, but I
don't recall right now.) Is this a problem with recent threads
and messages, i.e., a problem which gets resolved by the time of
a final monthly archive? (I have been looking almost exclusively
at the July archive, so I don?t know if this problem exists in
previous months' archives.)



Ok, thanks in advance for your help, and I hope my contributions
are of use to someone.

Take care

Isidor





From Tim.Peters at p98.f112.n480.z2.fidonet.org  Thu Jul  1 03:53:50 1999
From: Tim.Peters at p98.f112.n480.z2.fidonet.org (Tim Peters)
Date: Thu, 01 Jul 1999 08:53:50 +0100
Subject: TopSort in Python?
Message-ID: <0000161d@bossar.com.pl>

From: "Tim Peters" <tim_one at email.msn.com>

[Dinu C. Gherman]
> Does anybody have a simple-minded-but-working full-Python
> implementation of topsort, the topological sorting algorithm?
> Or maybe *any* topsort? I remember only one such algorithm...
> python.org/search doesn't reveal any...

Searching for "topological" instead turns up two hits there, a few more on
DejaNews.  Apparently "the std" algorithm I posted years ago predates
DejaNews, though.  So here's a fancier version ...

check-out-aaron's-kjbuckets-too-ly y'rs  - tim

# topsort takes a list of pairs, where each pair (x, y) is taken to
# mean that x <= y wrt some abstract partial ordering.  The return
# value is a list, representing a total ordering that respects all
# the input constraints.
# E.g.,
#    topsort( [(1,2), (3,3)] )
# may return any of (but nothing other than)
#    [3, 1, 2]
#    [1, 3, 2]
#    [1, 2, 3]
# because those are the permutations of the input elements that
# respect the "1 precedes 2" and "3 precedes 3" input constraints.
# Note that a constraint of the form (x, x) is really just a trick
# to make sure x appears *somewhere* in the output list.
#
# If there's a cycle in the constraints, say
#    topsort( [(1,2), (2,1)] )
# then CycleError is raised, and the exception object supports
# many methods to help analyze and break the cycles.  This requires
# a good deal more code than topsort itself!

from exceptions import Exception

class CycleError(Exception):
    def __init__(self, sofar, numpreds, succs):
        Exception.__init__(self, "cycle in constraints",
                           sofar, numpreds, succs)
        self.preds = None

    # return as much of the total ordering as topsort was able to
    # find before it hit a cycle
    def get_partial(self):
        return self[1]

    # return remaining elt -> count of predecessors map
    def get_pred_counts(self):
        return self[2]

    # return remaining elt -> list of successors map
    def get_succs(self):
        return self[3]

    # return remaining elements (== those that don't appear in
    # get_partial())
    def get_elements(self):
        return self.get_pred_counts().keys()

    # Return a list of pairs representing the full state of what's
    # remaining (if you pass this list back to topsort, it will raise
    # CycleError again, and if you invoke get_pairlist on *that*
    # exception object, the result will be isomorphic to *this*
    # invocation of get_pairlist).
    # The idea is that you can use pick_a_cycle to find a cycle,
    # through some means or another pick an (x,y) pair in the cycle
    # you no longer want to respect, then remove that pair from the
    # output of get_pairlist and try topsort again.
    def get_pairlist(self):
        succs = self.get_succs()
        answer = []
        for x in self.get_elements():
            if succs.has_key(x):
                for y in succs[x]:
                    answer.append( (x, y) )
            else:
                # make sure x appears in topsort's output!
                answer.append( (x, x) )
        return answer

    # return remaining elt -> list of predecessors map
    def get_preds(self):
        if self.preds is not None:
            return self.preds
        self.preds = preds = {}
        remaining_elts = self.get_elements()
        for x in remaining_elts:
            preds[x] = []
        succs = self.get_succs()

        for x in remaining_elts:
            if succs.has_key(x):
                for y in succs[x]:
                    preds[y].append(x)

        if __debug__:
            for x in remaining_elts:
                assert len(preds[x]) > 0
        return preds

    # return a cycle [x, ..., x] at random
    def pick_a_cycle(self):
        remaining_elts = self.get_elements()

        # We know that everything in remaining_elts has a predecessor,
        # but don't know that everything in it has a successor.  So
        # crawling forward over succs may hit a dead end.  Instead we
        # crawl backward over the preds until we hit a duplicate, then
        # reverse the path.
        preds = self.get_preds()
        from random import choice
        x = choice(remaining_elts)
        answer = []
        index = {}
        in_answer = index.has_key
        while not in_answer(x):
            index[x] = len(answer) # index of x in answer
            answer.append(x)
            x = choice(preds[x])
        answer.append(x)
        answer = answer[index[x]:]
        answer.reverse()
        return answer

def topsort(pairlist):
    numpreds = {}   # elt -> # of predecessors
    successors = {} # elt -> list of successors
    for first, second in pairlist:
        # make sure every elt is a key in numpreds
        if not numpreds.has_key(first):
            numpreds[first] = 0
        if not numpreds.has_key(second):
            numpreds[second] = 0

        # if they're the same, there's no real dependence
        if first == second:
            continue

        # since first < second, second gains a pred ...
        numpreds[second] = numpreds[second] + 1

        # ... and first gains a succ
        if successors.has_key(first):
            successors[first].append(second)
        else:
            successors[first] = [second]

    # suck up everything without a predecessor
    answer = filter(lambda x, numpreds=numpreds: numpreds[x] == 0,
                    numpreds.keys())

    # for everything in answer, knock down the pred count on
    # its successors; note that answer grows *in* the loop
    for x in answer:
        assert numpreds[x] == 0
        del numpreds[x]
        if successors.has_key(x):
            for y in successors[x]:
                numpreds[y] = numpreds[y] - 1
                if numpreds[y] == 0:
                    answer.append(y)
            # following "del" isn't needed; just makes
            # CycleError details easier to grasp
            del successors[x]

    if numpreds:
        # everything in numpreds has at least one predecessor ->
        # there's a cycle
        if __debug__:
            for x in numpreds.keys():
                assert numpreds[x] > 0
        raise CycleError(answer, numpreds, successors)
    return answer






From Johan.Wouters at p98.f112.n480.z2.fidonet.org  Fri Jul  2 13:41:16 1999
From: Johan.Wouters at p98.f112.n480.z2.fidonet.org (Johan Wouters)
Date: Fri, 02 Jul 1999 18:41:16 +0100
Subject: Why does shelve make such large files?
Message-ID: <0000166e@bossar.com.pl>

From: Johan Wouters <johanw at easics.be>

Gerrit Holl wrote:

> Ah, I understand.
> So pickle is useful for very small datases, but when they're really huge, one
> should use shelve. Isn't it?

Pickle will serialise your data into a certain format. This way you can
"store" objects like lists as a whole. The serialising has the benefit that
you can eg. put your object (be it a list or something else) through a pipe
or socket and depickle it at the other side. This way you can have python 
processes that exchange data in a sort of native python format.

Shelve gives you alot more than just storing the data as a long sequence.
Imagine the extra space being the administration that is needed for all
the extra functionality (like tagging what data items are valid etc)

The overhead induced by a database management system can be enormous
compared to the bare data you want to access, but this is mostly a
space/time tradeoff: the more space you use the faster you can do
things like searching and sorting etc etc.

kind regards,
===================================================================
Johan Wouters             ===              Easics               ===
ASIC Designer             ===  System-on-Chip design services   ===
Tel: +32-16-395 616          ===================================
Fax: +32-16-395 619      Interleuvenlaan 86, B-3001 Leuven, BELGIUM
mailto:johanw at easics.be           http://www.easics.com



From pereira at research.att.com  Tue Jul 20 23:48:21 1999
From: pereira at research.att.com (Fernando Pereira)
Date: Wed, 21 Jul 1999 03:48:21 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com> <170719990020380651%pereira@research.att.com> <7mqaht$r27$2@cronkite.cc.uga.edu> <170719991835331004%pereira@research.att.com> <7mssne$m8r$1@cronkite.cc.uga.edu> <180719992337546352%pereira@research.att.com> <7mvika$cpj$1@cronkite.cc.uga.edu> <200719990035531711%pereira@research.att.com> <7n27r8$mm4$1@cronkite.cc.uga.edu>
Message-ID: <200719992348217245%pereira@research.att.com>

In article <7n27r8$mm4$1 at cronkite.cc.uga.edu>, Graham Matthews
<graham at sloth.math.uga.edu> wrote:
> Could you please point out to me where I claimed that "a program could
> optimise your combinator program". If you can't find such a quote then
> what is your point?
You didn't. The point, which is obvious to anyone who has actually done
serious pattern matching and parsing applications operating on
gigabytes of data, is that pattern matchers and parsers *must* be
optimized automatically (in the same way that programs in high-level
languages *must* be optimized automatically) because humans are just
not good enough at doing hand optimization of complex programs. Insofar
as a style of writing pattern-matchers/parsers is incompatible with
automatic optimization, it is only useful for toy examples, or, at,
best, for proofs of concept (although in my experience a proof of
concept that doesn't deal with efficiency issues is not much of a
proof). 
> 
> Fernando Pereira (pereira at research.att.com) wrote:
> : But from my readings (those mentioned in the thread
> : and others) as well as your comments, the whole point of monadic
> : combinators for parsing is to interweave the grammatical structure with
> : other computations, thus making the underlying grammatical structure
> : less accessible to optimization. As I suggested above, it would be
> : interesting to define classes of monadic programs that still allow (an
> : appropriate extension of) those optimizations. But AFAIK that has not
> : been done.
> 
> It hasn't been done, it won't likely be done, and I never said it
> could be done. So what is your point?
That it might be interesting to understand under what conditions it
could be done, building on the work I mentioned in my earlier posting
as well as on the work on attribute grammars, which is directly
relevant. Of course, if what is at stake is religion rather than
science, such curiousity may be frowned upon. Oh well... Enough said.

-- F



From htrd90 at zepler.org  Tue Jul 13 15:47:31 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Tue, 13 Jul 1999 19:47:31 GMT
Subject: Python and COM (long)
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F1D@RED-MSG-50>
Message-ID: <378d8ee2.10645225@news.freeserve.net>

Bill Tutt <billtut at microsoft.com> wrote:

(posted and cc'ed to bill)

>> From: htrd90 at zepler.org [mailto:htrd90 at zepler.org]
>> 
>> Bill Tutt <billtut at microsoft.com> wrote:
>> 
>> >(As an added bonus, this should actually work)
>> >The only GUID from interp.py is the coclass one. Every other 
>> GUID was just
>> >generated.
>> 
>> I _think_ the python code would also need to be changed to allow
>> QueryInterface for the IID of that dispinterface.
>
>Nope. Dispinterfaces are wierd in that they only tell you whats available
>via the IDispatch interface.

Argh - this com terminology is a persistant pain. Certainly it is an
IDispatch interface. The tricky question is whether it is the
IDispatch interface returned by QueryInterface for IID_IDispatch, or a
_different_ IDispatch returned by QueryInterface for
IID_IPythonInterpreter (in this case)

>You have to do a QI for IDispatch, not the
>dispinterface.

OK, that clarifies your meaning.


Bill's advice is something I've heard before, but it contradicts my
experience. I thought I would try some experiments with this object
and some other Automation controllers. The results were mixed, and a
little disturbing.....

For the first tests I followed Bill's recommendations. I compiled the
.idl, registered the .tlb, and manually patching the TypeLib section
of the "Python.Interpreter" object to point to our new LIBID.

Testing the object in this state, VB6 and the COM Object Viewer (ver
2.10.054), produce negative results. VB produces "Run-time error 430"
"Class does not support Automation or does not support expected
interface". The object viewer creates the object but does not list
IPythonInterpreter.

Instrumenting the com server with the method below makes it clear that
both controllers are querying for IID_IPythonInterpreter, and not
IID_IDispatch.
    def _query_interface_(self, iid):
        print iid
        return 0

However, VC6 (sp1) using #import works well with this object. It is
querying for IID_IDispatch, as predicted by Bill, and never
IID_IPythonInterpreter.


For the second set of tests, I added the following implementation of
QueryInterface, as I originally thought was necessary.
    def _query_interface_(self, iid):
        print iid
        if str(iid)=="{EB048AA4-C2D1-11D2-855D-00C04F797DBA}":
            return win32com.server.util.wrap(self)
        return 0

This fixes VB and the Object Viewer (which now correctly lists
IPythonInterpreter). VC6 still works.


So, my conclusion is that you need to support QueryInterface for
_both_ the IID of your dispinterface and IID_IDispatch, and that they
must both return the same implementation of IDispatch.

Comments _really_ appreciated. Thanks for reading,



Toby Dickenson



From olipt at mayo.edu  Thu Jul  8 09:28:58 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Thu, 8 Jul 1999 08:28:58 -0500
Subject: Python and Fortran code
In-Reply-To: <7m23i3$ktr$1@bignews.shef.ac.uk>
References: <7m23i3$ktr$1@bignews.shef.ac.uk>
Message-ID: <Pine.LNX.4.10.9907080813220.4176-100000@us2.mayo.edu>

> Hi!
>        	I have a quetion on using python to do some simple processing on
> some data I pre-process with some Fortran code. I wanted to be able to call
> this subroutine from within python, as it is already "tried and tested" and
> optimized. Basically, my subroutine looks like this:
> 
> 	subroutine ex1(exdata)
> 	dimension exdata(256,200)
> 	print*,'Doing some processing in FORTRAN'
>  C      We said we were doing some processing, so we might as well do it :)
>  C	Processing of exdata
>  	return
> 	end
> 	
> 	This subroutine would need to get input from the user and would also
> need to print stuff out for the user to see several parameters as the
> software runs. I'd like to pass exdata as a Python object (NumPy array
> object), and then recover it with the results. I wonder if I could somehow
> do that compiling this subroutine as an object file and then calling it from
> python (not that i know how to do it, but with scilab, it works :D).

I don't know of any tool to call this routine "automatically" like in
scilab.  Such a tool would be possible and wouldn't be difficult to get
"most of the way" but would be a bear to make work in all cases.  Awhile
ago, I heard someone was working on something like that.

For now there are two options I know of.  

(1) Mess around with dlmodule.c
which is "an experimental device to call arbitrary C functions in
arbitrary shared libraries."  On linux with g77 your function would be
called ex1_ and you would pass it a Python string (which is a sequence of
bytes).  I don't know much more about this option.

(2) Do what I did to construct multipack and other modules which link to
Fortran code:  Write an extension module --- reference the documentation.
Briefly this involves writing two subroutines and compile them as a Python
module: The first is named initxxxxx  and is what Python runs when
you say import xxxxx.  The second is a function that takes two Python
objects.  This is the function that will be run when you call the function
in Python.  In here you get the NumPy array and pass a pointer to the
data of the array to ex1_ (remember the "_" on GNU/linux), then return.

If you need more help for option (2) just email.  Option (1) I know less
about but would be interested to hear how that works.

Best,

Travis




From jsight at mindspring.com  Mon Jul 19 22:32:57 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Mon, 19 Jul 1999 22:32:57 -0400
Subject: Python and the Web
References: <3793dd72.0@hoh.olympus.net>
Message-ID: <3793DFD9.5907157C@mindspring.com>

The Pyhon website at http://www.python.org/ has many things under the
documentation section which may be useful, but to get you right to the
stuff you want, go to http://www.python.org/topics/web/

This is a page full of essentially nothing but the web-related Python
stuff.  :-)

Michael Novello wrote:
> 
> Can anyone tell me where to find information on using python to grab form
> data and create dynamic web pages?
> 
> Thanks,
> 
> Michael

-- 
---------------
Jesse D. Sightler
http://www3.pair.com/jsight/

"Do not use a hatchet to remove a fly from your friend's forehead." 
      - Chinese Proverb



From winkjj at my-deja.com  Tue Jul 13 08:57:44 1999
From: winkjj at my-deja.com (winkjj at my-deja.com)
Date: Tue, 13 Jul 1999 12:57:44 GMT
Subject: DLL load failed with SWIG DLL
Message-ID: <7mfd42$hrq$1@nnrp1.deja.com>

I am attempting to use SWIG to create an
extension DLL for Python on a WinNT machine.

I have successfully compiled and linked my DLL,
which uses the -shadow option in SWIG.  My DLL is
called Examplec, and the SWIG generated .py
module is Example.

When I attempt to import either the .py module or
the DLL itself, I get the following error:

ImportError: DLL load failed: The specified
module could not be found.

Now, I have searched all over this newsgroup and
have found similiar problems, but no one has
complained about this exact message.

Other problems have been things like not being
able to find DLLs that this DLL is dependent on,
or the DLL not having a correct init<module>
signature.

I have verified that all this looks correct.  The
DLL and .py module are on the PYTHONPATH.  And
the init entry point is correct.  All the
dependacy DLLs can be found on the system path as
well.

I also compiled the DLL with the __WIN32__ symbol.

The only thing I question is whether or not I
built SWIG correctly.  After you build SWIG you
run 'nmake install'.  This installs SWIG into the
target directory.

I built SWIG on machine A, and then copied the
install directory to the exact location on
machine B.

I don't see how this could make any difference,
however, because the DLL compiles and links fine.

Has anyone seen this before, or could someone
give some suggestions?  I am at a loss at to what
to try next.

Thanks in advance for your help.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From neelk at brick.cswv.com  Fri Jul  9 19:16:59 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 9 Jul 1999 18:16:59 -0500
Subject: VIM and [Python] block jumping
References: <3781347D.CB2FBAB3@starvision.com> <37815B80.B4B51F94@starvision.com> <3785D753.DB2FF1ED@kami.com> <3785DA06.974567F3@kami.com>
Message-ID: <7m5vtb$2b7$1@brick.cswv.com>

In article <3785DA06.974567F3 at kami.com>,
C. Laurence Gonsalves <clgonsal at kami.com> wrote:
>Of course, right after I posted that, realized that there's a problem.
>The blockMotion function doesn't account for lines that are
>continuations of other lines. That means that triple-quoted string
>literals, lines with more {[('s that }])'s, and lines joined with \ can
>screw it up. Are there any other situations? I'll try to post a
>corrected version later. Some of those problems are pretty nasty though
>(triple-quoted strings in particular).

Being a faithful Emacs-ite, I've never used vim. But since you can
script vim with Python, it might be possible to use the tokenizer in
the standard library (module tokenize) to handle this cleanly. Likely
you'd need to extend it a bit to handle the syntax errors that show up
in half-written code gracefully, but taking vim that one step past
syntax-awareness and into structure-awareness would be a very slick
hack.


Neel



From fuming at saturn.radsci.uci.edu  Mon Jul  5 20:11:18 1999
From: fuming at saturn.radsci.uci.edu (Fuming Wang)
Date: Mon, 5 Jul 1999 17:11:18 -0700
Subject: VIM and block jumping
In-Reply-To: <3781347D.CB2FBAB3@starvision.com>
References: <3781347D.CB2FBAB3@starvision.com>
Message-ID: <Pine.SOL.4.05.9907051548470.13052-100000@saturn.radsci.uci.edu>

Hi, I have a small patch that can make vim do following indentation:
def func():
    print "hello world!"
    if a == b and \
       b == c or  \
       e == f:
        print "got it!"

I can send you the patch if this is what you like.

Cheers,
Fuming Wang
fuming at venus.radsci.uci.edu



On Mon, 5 Jul 1999, Angus MacKay wrote:

> this is a question for all of the VIMers out there.
> 
> is there any way to get [# ]# style logical python block jumping
> to work for python? (to make up for the lack of {} in python)
> 
> cheers, Angus.
> 
> 




From gmcm at hypernet.com  Fri Jul 30 09:25:58 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 30 Jul 1999 08:25:58 -0500
Subject: Need help with TCP/IP client access from Windows
In-Reply-To: <19990730005211.T7149@thyrsus.com>
References: <1278839420-22085362@hypernet.com>; from Gordon McMillan on Thu, Jul 29, 1999 at 05:21:12PM -0500
Message-ID: <1278785133-25350622@hypernet.com>

Eric S. Raymond writes:

> Gordon McMillan <gmcm at hypernet.com>:

> > Probably better just to have a default file, and let the user 
> > override that if need be. Which may well never happen.
> 
> OK, but where to put it?  I don't know what the conventions are.

Probably the same directory where you put your script. (The Win31 
convention of putting all ini files in the Windows directory under 
<appname>.ini is now, thank goodness, passe).

> > > (3) Is there a reliable way for Python programs under Windows to
> > > point a local
> > >     browser at a specified URL?
> > 
> > The best way is to use WinExec from Mark's win32 extensions. You just 
> > give it the URL and it loads the user's default browser. Using 
> > os.system is harder, because most of the time a browser is not on the 
> > user's path.
> 
> Are these in the stock 1.5.2 Windows distribution?

No, They're in Mark Hammond's extensions (link from python.org). The 
"start" trick (which I just tried) is probably the best way in stock 
Python (though it raises a bogus console window for a few seconds 
before the browser starts).


- Gordon



From arcege at shore.net  Mon Jul 19 20:55:05 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Tue, 20 Jul 1999 00:55:05 GMT
Subject: Are postings to comp.lang.python.announce mirrored here?
References: <37933DB1.47814E84@cfdrc.com> <14227.22618.779710.843912@anthem.cnri.reston.va.us>
Message-ID: <JNPk3.532$ui4.178537@news.shore.net>

Barry A. Warsaw <bwarsaw at cnri.reston.va.us> wrote:

:     |  2. Is it considered bad form to cross-post new S/W
:     |     announcements to both lists? Or is that the Right
:     |     Thing to do?

: It's probably okay (as long as it's a real crosspost and not two
: separate posts).  The thing /not/ to do is post announcements only to
: c.l.py.  The `Recent Announcements' part of the Python home page, and
: the corresponding Netscape channel, is only culled from c.l.py.a.

FYI, I've attempted some cross-posts to both which didn't even make it
to the moderator.  Cross-posting is the Right Thing(TM), but not all
servers seem to handle it when one is moderated.  Seperate posts is
at times the only option (if you want it announced in both newsgroups).

  -Arcege




From fredrik at pythonware.com  Fri Jul 16 13:39:04 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 16 Jul 1999 17:39:04 GMT
Subject: I envy newbies
References: <1279991447-21085630@hypernet.com>
Message-ID: <002601becfb2$19ef6400$f29b12c2@secret.pythonware.com>

Gordon McMillan <gmcm at hypernet.com> wrote:
> [I think it's hilarious that MS was so enamored of their 
> "advance" in being able to put spaces in paths that they made the 
> "proper" place to install everything a demo of this "feature"...

fwiw, it's "c:\Program" on my box.  no spaces there.
maybe you should get a non-US version of Windows? ;-)

</F>




From Dinu.C..Gherman at p98.f112.n480.z2.fidonet.org  Thu Jul  1 07:27:45 1999
From: Dinu.C..Gherman at p98.f112.n480.z2.fidonet.org (Dinu C. Gherman)
Date: Thu, 01 Jul 1999 12:27:45 +0100
Subject: How to obtain an instance's name at runtime?
Message-ID: <00001607@bossar.com.pl>

From: Dinu C. Gherman <gherman at my-deja.com>

In article <37769B92.47EBE99C at compaq.com>,
  Greg Ewing <greg.ewing at compaq.com> wrote:
>
> The best solution by far would be to give the
> object a name explicitly when you create it.
> Is this possible, or do you want to retrofit
> it onto existing code?

The latter is what I'm planning to do.

> It looks like you want it to be something like
> "the name of the first variable to be assigned
> a reference to the object". Does this apply to
> local variables as well as globals? Instance
> variables of another object? What if the object
> is never assigned to a variable? What if it
> is, but somewhere far removed from where it
> is created?

You're guessing very well... ;-) Actually, I will
postpone some issues until some advance is made at
other places of the code.

Just that much: an object, that has not been assigned
a name to, cannot exist (having a reference count of
zero), can it (at least for user-defined objects)?

In any case thanks so far! I think the problem is not
a really fundamental one for the type of application
I'm having in mind, after all.

Bye,

Dinu





Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From andy at robanal.demon.co.uk  Tue Jul 27 16:00:22 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Tue, 27 Jul 1999 20:00:22 GMT
Subject: MS-SQL Server:Connectivity Help...
References: <379DD9A7.589CE943@ecom.net>
Message-ID: <37a80e9f.7428511@news.demon.co.uk>

"Edward B. Wilson II" <zero at ecom.net> wrote:

>I would like to introduce Python with a bang to my new work group.  Does
>anyone have a driver to "insert" a single row into an MS-SQL server7.0
>db.  It appears from the main Python page, one must compile their own
>ODBC drivers with VC++ to get the most robust solution.  Can anybody
>comment on this shortcoming.  Getting Python to work for MS_SQL/NT would
>be a big entrance.
>
I'd thoroughly recommend mxODBC.  One huge advantage is the DateTime
objects, which offer capabilities to make database developers drool.
Furthermore, I'm pretty sure you don't lose any performance using
Microsoft ODBC drivers for Microsoft's own database!

You can do two lines of Python to grab a whole query into Python
variables; manipulate it; then do another line to insert the lot into
another table.    One row at a time?  pah!

You can also tell your colleagues that it is working with prepared
statements for optimum performance.

Regards,

Andy



From mstenber at cc.Helsinki.FI  Sun Jul 18 15:42:34 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 18 Jul 1999 22:42:34 +0300
Subject: Getters and Setters
References: <000101becedb$1f4c3400$91a02299@tim> <m3emi9x07h.fsf@greebo.nodomain.de> <7mlfkf$qdh$1@nnrp1.deja.com>
Message-ID: <al8btd9d9sl.fsf@myntti.helsinki.fi>

tim_one at email.msn.com writes:
> [Bernhard Herzog <herzog at online.de]
> > The cycle doesn't involve self, but self's dict will hold an
> > indirect reference to itself. The instance object will be
> > deleted, so it's __del__ method is called, but its __dict__
> > won't be deleted:
> 
> Ah -- got it.  Thanks for the explanation.  self.__del__ is called,
> though, so the cycles are easy to break:
> 
> def __del__(self):
>     self.__dict__.clear()

That's only painful if other classes define __del__ too while using this as
mixin.. (lazily yours ;>)

> even-tested-it-this-time<wink>-ly y'rs  - tim

Just to contrib something to the discussion.. (note that test period wasn't
extensive, but I doubt it'd vary much)

normal get    :  127539.1651/sec [0.24s] (7.8407us/call) (foo.x)
normal set    :  101784.9252/sec [0.29s] (9.8246us/call) (foo.x=0)
normal getattr:  118375.8787/sec [0.25s] (8.4477us/call) (getattr(foo,'x'))
normal setattr:  110274.2723/sec [0.27s] (9.0683us/call) (setattr(foo,'x',1))

FastGet       :   75519.0025/sec [0.40s] (13.242us/call) (instance method wrapper)
FastSet       :   64276.5807/sec [0.47s] (15.558us/call)

SlowGet       :   43534.9160/sec [0.69s] (22.970us/call) (class method wrapper)
SlowSet       :   38138.5309/sec [0.79s] (26.220us/call)

Obviously, the first call is slow, but beyond that, it seems reasonable
enough - I might even use it, one of these days, in my programs ;-)

-Markus

-- 

	"If you want to travel around the world and be invited to speak at
	a lot of different places, just write a Unix operating system."
		-- Linus Torvalds



From kaine_the_bloody at my-deja.com  Thu Jul  8 13:24:19 1999
From: kaine_the_bloody at my-deja.com (Kaine)
Date: Thu, 08 Jul 1999 17:24:19 GMT
Subject: Python and Paradox
Message-ID: <7m2mrj$hes$1@nnrp1.deja.com>

I am looking for a way to access Paradox tables from Python without
using ODBC. Does anyone have any ideas?


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From news at dorb.com  Wed Jul 28 16:19:21 1999
From: news at dorb.com (Darrell)
Date: Wed, 28 Jul 1999 16:19:21 -0400
Subject: a little motivation, anyone?
References: <19990728052704.10608.00002260@ng-fq1.aol.com>
Message-ID: <MBJn3.859$Yk3.6771@newsr1.twcny.rr.com>

Medley8086 <medley8086 at aol.com> wrote in message
news:19990728052704.10608.00002260 at ng-fq1.aol.com...
> What I'm asking is, could i please have some
> posts of what some of you other programmers made when you were starting
out, in
> what languages, and how long it took.  I'm looking for stories to convince
me
> while trying to work with stubborn code that I'm not algorithmically
> handicapped or just not cut out to be a coder at all.  Thanks in advance
for
> anyone who decides to help.
>
> E. Leron Culbreath
>

Writing code is hard.
It's hard for everyone, I think :)
Writing code is fun.
It's not fun for everyone.
If it's fun for you then no doubt you are cut out to be a coder:
Else maybe coding isn't for you.

Hint #1: Look for and steal patterns of how to do things from
    the code of others. A pattern can be as simple as how
    to handle command line switches.

Hint #2: Write lots of code.

Hint #3: Read lots of code.

The code you write in a year will be vastly better than the code you wrote
today.
Hang in there.
--
--Darrell






From un4e at rz.uni-karlsruhe.de  Tue Jul 20 03:56:23 1999
From: un4e at rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 20 Jul 1999 07:56:23 GMT
Subject: Embedding: Defining packages?
References: <7mq9r5$irj$1@news.rz.uni-karlsruhe.de> <5iPk3.527$ui4.177861@news.shore.net>
Message-ID: <7n1a37$s1$1@news.rz.uni-karlsruhe.de>

Hi,

Michael P. Reilly <arcege at shore.net> wrote:
: Bjoern Giesler <un4e at rz.uni-karlsruhe.de> wrote:
: : theModule = Py_InitModule(methodDefs, "MyProgram.Server.Communications");

: theModule = Py_InitModule("MyProgram.Server.Communications", methodDefs);

...was of course what I meant. Mine was just a typo. Still, this gives me a
module named "MyProgram.Server.Communications", but _no_ module "MyProgram".
What I want is to be able to say
	from MyProgram import *
and then address functions via
	Server.Communications.foo()

This doesn't work; it seems that Python can't define _packages_ from C, but
only _modules_. Is that correct?

		--Bjoern
-- 
--------------------------------/\--One OS to rule them all---Windows NT-----
Bjoern Giesler                 /  \  One OS to find them           
<un4e at rz.uni-karlsruhe.de>    / <> \  One OS to bring them all    
-----------------------------/______\--And in the Darkness bind them---------
    Thank you for your time, worship the Antichrist, and have a nice day.



From mark.butterworth at cis.co.uk  Wed Jul  7 07:18:21 1999
From: mark.butterworth at cis.co.uk (Mark Butterworth)
Date: Wed, 7 Jul 1999 12:18:21 +0100
Subject: returning Python Lists from C++
Message-ID: <tKGg3.149$FH1.16131@newsr2.u-net.net>

Can anyone give me a quick example of how to return a python list object
from a C++ function?





From jam at p98.f112.n480.z2.fidonet.org  Fri Jul  2 03:29:33 1999
From: jam at p98.f112.n480.z2.fidonet.org (jam)
Date: Fri, 02 Jul 1999 08:29:33 +0100
Subject: RFC822 & Text of Mail Message
Message-ID: <00001676@bossar.com.pl>

From: jam <jam at newimage.com>

On Fri, Jul 02, 1999 at 03:07:03PM +0000, gcollins99 at my-deja.com wrote:
> 
> Greetings!
> 
>   Here is what I want to do but cannot find out how. I am using RFC822
> to parse out the headers and store them somewhere which works just fine.
> But I am also in need of getting the text of the incoming message. So
> what module would I use to also capture the message text. I see no way
> of doing this via rfc822 module since it's not designed to do such
> things. WOuld I use mimetools? And if so, how?
> 
> Thanks!
> 
> -Glen Collins
> 

hi.

take a look at 'mailbox.py' in the standard library distribution for an example
of how to do this.  it boils down to calling 'rewindbody' and the 'read'ing
the data like any normal file.

good luck ! ;)

regards,
J
-- 
|| visit gfd <http://quark.newimage.com/> 
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>




From jcw at equi4.com  Fri Jul  2 11:50:47 1999
From: jcw at equi4.com (Jean-Claude Wippler)
Date: Fri, 02 Jul 1999 17:50:47 +0200
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773CA67.7057F797@quasar.ipa.nw.ru> <3773D55B.8C966174@spacenet.tn.cornell.edu> <slrn7naiqb.8ll.zeitlin@seth.lpthe.jussieu.fr> <7li9t6$j2k$1@srv38.cas.org> <30CF6C97EED47EAD.8A8612E26FB53A0A.B0666516E77A0F15@lp.airnews.net>
Message-ID: <377CDFD7.3103B1AE@equi4.com>

Mikael Lyngvig wrote:
> 
> On 2 Jul 1999 12:04:54 GMT, lvirden at cas.org wrote:
> 
> >I wonder why various industry rags have proclaimed during the past
> >10 months about Apple's G3 machines outselling other vendor
> >machines.  ...
> 
> That makes two of us.

Sales != installed base.
Also: Sales of one vendor vs. combined sales of N.

-- Jean-Claude, proving that PB G3's are fine for posting (even off-topic)



From aa8vb at yahoo.com  Thu Jul  8 07:41:13 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Thu, 8 Jul 1999 07:41:13 -0400
Subject: .pyo's without .pyc's?
In-Reply-To: <374C6C6E.1D07326B@lemburg.com>; from M.-A. Lemburg on Wed, May 26, 1999 at 11:49:34PM +0200
References: <19990526140137.A66609@vislab.epa.gov> <374C6C6E.1D07326B@lemburg.com>
Message-ID: <19990708074113.A61389@epa.gov>

M.-A. Lemburg:
 |Looks like you have setup PYTHONPATH to point to some old
 |extensions.py file (the compiled versions seemed to be ok).

It doesn't appear this is the case (please see below).

 |Try "python -vv" to have a look at the path scanning Python
 |does on startup.

I've attached the two different search outputs below.  As you can see, even
with -O, it's searching .pyc files even when .pyo is specified.  As far as
I've heard, this is incorrect behavior.

 |Running Python with .pyo files only works just fine if you
 |remember to always use the -O flag on the command line.

I hope it can be made to work that way, but it doesn't seem to.

 |For minimalist Python installs, you could also check out the
 |mxCGIPython project:
 |
 |	http://starship.skyport.net/~lemburg/mxCGIPython.html
 |
 |It's Python wrapped into a single file (with the standard
 |lib compiled with -O BTW).

Alas, I need Tkinter compiled in and IIRC this doesn't work well with Tk
yet.

Thanks,

-- 
Randall Hopper
aa8vb at yahoo.com


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

~/software/python-1.5.2 >  unsetenv PYTHONPATH
~/software/python-1.5.2 >  setenv PATH "/bin::usr/bin"
~/software/python-1.5.2 >  pwd
/home/rhh/software/python-1.5.2
~/software/python-1.5.2 >  ls
bin      demos    include  lib      man      tools
~/software/python-1.5.2 >  bin/python
Python 1.5.2 (#1, Apr 14 1999, 07:55:56) [C] on irix646-n32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 
~/software/python-1.5.2 >  bin/python -O
Python 1.5.2 (#1, Apr 14 1999, 07:55:56) [C] on irix646-n32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 

~/software/python-1.5.2 >  find . -name '*.py' -print | xargs rm
~/software/python-1.5.2 >  bin/python
Python 1.5.2 (#1, Apr 14 1999, 07:55:56) [C] on irix646-n32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 
~/software/python-1.5.2 >  bin/python -O
Python 1.5.2 (#1, Apr 14 1999, 07:55:56) [C] on irix646-n32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 

~/software/python-1.5.2 >  find . -name '*.py' -print | xargs rm
Incorrect usage
Usage: rm [-fir] file ...
~/software/python-1.5.2 >  find . -name '*.pyc' -print | xargs rm
~/software/python-1.5.2 >  bin/python
Could not find platform independent libraries <prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
'import exceptions' failed; use -v for traceback
Warning!  Falling back to string-based exceptions
'import site' failed; use -v for traceback
Python 1.5.2 (#1, Apr 14 1999, 07:55:56) [C] on irix646-n32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 
~/software/python-1.5.2 >  bin/python -O
'import exceptions' failed; use -v for traceback
Warning!  Falling back to string-based exceptions
Python 1.5.2 (#1, Apr 14 1999, 07:55:56) [C] on irix646-n32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 
~/software/python-1.5.2 >  exit

script done on Thu Jul  8 07:25:00 1999

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

Successful exceptions module search -- unadulterated install tree:

> find . -name 'exceptions*'
./lib/python1.5/exceptions.py
./lib/python1.5/exceptions.pyo
./lib/python1.5/exceptions.pyc

> bin/python -O -vv | & grep except
# trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptionsmodule.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.py
# /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.pyo matches /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.py
import exceptions # precompiled from /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.pyo

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

Unsuccessful exceptions module search -- *.py and *.pyc pruned:

> find . -name 'exceptions*'
./lib/python1.5/exceptions.pyo

> bin/python -O -vv | & grep except
# trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptionsmodule.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.py
# trying /home/rhh/software/python-1.5.2/lib/python1.5/exceptions.pyc
# trying /home/rhh/software/python-1.5.2/lib/python1.5/plat-irix646-n32/exceptions.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/plat-irix646-n32/exceptionsmodule.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/plat-irix646-n32/exceptions.py
# trying /home/rhh/software/python-1.5.2/lib/python1.5/plat-irix646-n32/exceptions.pyc
# trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-tk/exceptions.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-tk/exceptionsmodule.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-tk/exceptions.py
# trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-tk/exceptions.pyc
# trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-dynload/exceptions.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-dynload/exceptionsmodule.so
# trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-dynload/exceptions.py
# trying /home/rhh/software/python-1.5.2/lib/python1.5/lib-dynload/exceptions.pyc
'import exceptions' failed; traceback:
ImportError: No module named exceptions
Warning!  Falling back to string-based exceptions



From graham at sloth.math.uga.edu  Tue Jul 20 12:26:26 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 20 Jul 1999 16:26:26 GMT
Subject: Stackless & String-processing
References: <001101bed26d$a5bc4580$139e2299@tim>
Message-ID: <7n27vi$mm4$2@cronkite.cc.uga.edu>

Tim Peters (tim_one at email.msn.com) wrote:
: Who said anything about agreeing?  You two obviously disagree, but
: disagreement doesn't imply misunderstanding.  Heck, you've both repeated the
: same unbudging positions so often even *I* understand what you're both
: saying <wink>.
:
: + other equally useful contributions.

What is the point of your post Tim? Getting your jollies from adding
useless material to a thread?

graham

-- 
            The girls is all salty, the boys is all sweet, 
                      the food ain't too shabby 
                    an' they piss in the streets
    In France, way down in France, way on down, way on down in France



From Johan.Wouters at p98.f112.n480.z2.fidonet.org  Thu Jul  1 10:09:33 1999
From: Johan.Wouters at p98.f112.n480.z2.fidonet.org (Johan Wouters)
Date: Thu, 01 Jul 1999 15:09:33 +0100
Subject: Does the first argument of a method *has to be* self?
Message-ID: <00001603@bossar.com.pl>

From: Johan Wouters <johanw at easics.be>

Gerrit Holl wrote:
> 
> Hello,
> 
> I never saw code where the first argument of a method wasn't self. Is it
> in theory possible?

As one thinks about it, the first argument being self (the instance)
is a handy feature. Why would you want to put it somewhere else?

As the documentation states "self" is just some conventional name.
Renaming self into something else is probably a bad idea as it
will make your code less readable for others (and YOU as time passes!).

Experience shows that having conventions is a powerful way to
make life and code maintenance easier.

Kind regards,
===================================================================
Johan Wouters             ===              Easics               ===
ASIC Designer             ===  System-on-Chip design services   ===
Tel: +32-16-395 616          ===================================
Fax: +32-16-395 619      Interleuvenlaan 86, B-3001 Leuven, BELGIUM
mailto:johanw at easics.be           http://www.easics.com



From tim_one at email.msn.com  Fri Jul 16 00:51:18 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 16 Jul 1999 00:51:18 -0400
Subject: I envy newbies
In-Reply-To: <378EA68A.F2D8A7E@callware.com>
Message-ID: <000201becf46$d79e78e0$3fa02299@tim>

[Ivan Van Laningham, making an IDLE desktop shortcut under Windows]
> ...
> (I quote from _Teach Yourself Python in 24 Hours_ (forthcoming from
> SAMS)):

Monday <wink>?

> ``The first thing you need to do is to find where IDLE lives, and to do
> that you need to know where Python lives;
> [instructions for navigating the directory tree in Explorer until
>  finding idle.pyw]

Try this:  Start -> Programs -> Python 1.5 then rightclick+drag the IDLE
entry to the desktop.  Everything else:

> Right click ... and select "Create Shortcut" on the popup menu ...

is the same, but this avoids any mucking with Explorer.  Could be you need
the Active Desktop enabled for this to work (but it's your book, so I'm not
going to take the time to disable the Active Desktop to check that out for
you <wink>).

> The installer does modify your autoexec.bat to include the Python path
> on W95/98.

Eh?  It never touched autoexec.bat on my Win95 -- and darned glad it didn't.
That's a dangerous thing for any installer to presume to do.

The good news is that new users introduced via IDLE have no immediate need
to worry about the DOS path -- the IDLE shortcut knows how to invoke Python
regardless.  In fact, I've been known to use Python from time to time
<wink>, and

C:\temp>python
Bad command or file name

C:\temp>

That is, I still haven't bothered to put python.exe on my PATH!  Although I
do have an MS-DOS desktop shortcut with D:\PYTHON as its working directory
and "doskey /INSERT" as its "batch file".

can-just-feel-the-linux-folks-melting-with-jealousy<wink>-ly y'rs  - tim





From bill.barendse at tag.csiro.au  Mon Jul  5 20:36:05 1999
From: bill.barendse at tag.csiro.au (Bill Barendse)
Date: Tue, 06 Jul 1999 10:36:05 +1000
Subject: Tkinter._test() problem Tcl/Tk8.1 incompatibility python-1.5.2
References: <377F38AD.C6CDBA6D@fnmail.com>
Message-ID: <37814F74.DE625933@tag.csiro.au>

Hi Everyone,

I believe that Tcl/Tk 8.1 is incompatible with python-1.5.2 due to the
following experiences.

I use a SPARC ultra running solaris 2.6 with GCC 2.8.1 as a C compiler.
I downloaded python-1.5.2 earlier this month built and tested it and all
implemented tests worked.

I then downloaded Tcl/Tk8.1, built those in the proper order, tested
them and installed them.  Tcl failed the ASCII vs ISO character set test
(not surprising since I only have ASCII) and Tk failed the focus test.

I made clobber and rebuilt python specifying the locations, and
installed after setting the LD_LIBRARY_PATH variable.  I got the same
answer as Per Kistler - can't find a usable init.tcl in the following
directories - no directories specified.

I checked and rebuilt python (YES) several times since a have a usable
Tcl/Tk.

I then downloaded Tcl/Tk 8.0p2, built and tested (it failed more tests
than Tcl/Tk 8.1), rebuilt python and Tkinter generated the 'Hello World'
widget.

There are significant differences in the init.tcl of the two packages (I
can provide a diff of the files for those that are interested) but I do
not know enough about Tkinter to be able to say what it is about the
init.tcl8.1 or the rest of the Tcl/Tk8.1 distribution that makes it
incompatible.  The distribution notes that come with Tcl/Tk8.1 note
several bug fixes that will generate "***POTENTIAL INCOMPATIBILITY***"

There is no note on the python homepage about this incompatibility.

Cheers,

Bill Barendse

Bill.Barendse at tag.csiro.au
Comonwealth Scientific and Industrial Research Organisation, Australia.

Per Kistler wrote:

> Hi All
>
> I installed py1.5.2 with tcl/tk.8.1 on SuSE6.1 Linux.
>
> After:
>
> import Tkinter
> Tkinter._test()
>
> I get the following error, which I do not unerstand:
>
> -----cut
> >>python
> Python 1.5.2 (#4, Jul  4 1999, 12:25:56)  [GCC egcs-2.91.66 19990314
> (egcs-1.1.2  on linux2
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import Tkinter
> >>> Tkinter._test()
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "/per/lib/python1.5/lib-tk/Tkinter.py", line 1947, in _test
>     root = Tk()
>   File "/per/lib/python1.5/lib-tk/Tkinter.py", line 886, in __init__
>     self.tk = _tkinter.create(screenName, baseName, className)
> TclError: Can't find a usable init.tcl in the following directories:
>
> This probably means that Tcl wasn't installed properly.
> ----cut
>
> It seems, that python looks in some directories, which it does not
> have. What do I have to do?
>
> Thanks, Per.
> --
> Per Kistler kistler at fnmail.com / kistler at gmx.net
> ------------------------------------------------------------






From mstenber at cc.Helsinki.FI  Thu Jul 29 03:26:59 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 29 Jul 1999 10:26:59 +0300
Subject: any way to know when the program is exiting?
References: <199907290135.SAA22743@hpcll563.cup.hp.com>
Message-ID: <al8907zuda4.fsf@sirppi.helsinki.fi>

Ovidiu Predescu <ovidiu at cup.hp.com> writes:
> I was wondering if there is any way in a __del__ method to know when it's
> invoked on the way out, as a result of the program going away?

AFAIK, not..

> I have a __del__ method of a class that sends a message to a global
> variable.  It happens that on the way out, that value in that variable is
> destroyed before the __del__ methods of my instances are invoked. The
> result is that I have a bogus value that doesn't respond to the usual
> methods anymore. How can I prevent sending messages to that object?

Keep reference to the global variable in the classes with __del__.

Example follows: (How I kill child processes, and ensure the os module is
garbage collected _after_ the DelWrapper instance is __del__'d)

class DelWrapper:
    """
quick-n-dirty class to wrap __del__
"""
    def __init__(self, f):
        self.__del__ = f

def foo():
	... # proprietary, super-secret code snipped
	global __tmp
	__tmp = DelWrapper(lambda k=os.kill,pid=pid:apply(k, (pid, 15)))

> A possible solution would be to use atexit() but this doesn't work if the
> user program that's using my library replaces it with its own function.


> Any ideas?

Occassionally, some ;)

-Markus

-- 
   "I am Grey. I stand between the candle and the star.
  We are Grey. We stand between the darkness and the light."
  - Delenn / Babylon 5 [ #1 scifi show ATM, IMO ]



From jjens at primenet.com  Tue Jul 20 09:21:30 1999
From: jjens at primenet.com (John Jensen)
Date: 20 Jul 1999 13:21:30 GMT
Subject: python to perl conversion?
References: <slrn7p7q8j.n4.dyoung@baltazar.baltazar.org>
Message-ID: <7n1t4q$ik$1@nnrp03.primenet.com>

David Young <dyoung at pobox.com> wrote:
: Is there a program that converts Python source to Perl source?  There is
: a program that a client wants written in Perl so that their Perl-only
: programmers (can you imagine?) can hack on it later. I haven't the time or
: inclination to learn Perl, and this project fairly cries out for Python,
: but I cannot justify the time expense of translating from Python to Perl
: by hand.

I think I lean towards Python for beauty, but Perl does have its role in
commerce:

	http://www.robojob.com/lang.html

It might be good to know both, to use the one you prefer on projects you
control, but to use the other when the client specifies.

John



From mwh21 at cam.ac.uk  Wed Jul 21 07:43:42 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: Wed, 21 Jul 1999 12:43:42 +0100
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have code.)
In-Reply-To: <p5hfmymh9z.fsf@bidra241.bbn.hp.com>
References: <000601bed335$893916e0$a72d2399@tim> <p5hfmymh9z.fsf@bidra241.bbn.hp.com>
Message-ID: <Pine.LNX.4.10.9907211223270.11985-100000@atrus.jesus.cam.ac.uk>

On 21 Jul 1999, Markus Kohler wrote:
> "Tim Peters" <tim_one at email.msn.com> writes:
> 
> > [Tim]
> > >     for i in xrange(10000000):
> > >         pass
> > > [and RC vs GC implications]
> > 
> > [Markus Kohler]
> > > True.
> > > IMHO this is one of Pythons major design flaws.
> > 
> > What, specifically?
> 
> That flaw is that loops are not implemented using closures. 

Sorry, how would that help?

> Instead as far as I understand   loops in Python are implemented such that
> the loop counter is created and destroyed for each step of the loop. 
> 
> You could compare this to tail recursive versus non-tail recursive functions. 
> 
> Also closures make the implementation of efficient loops more
> difficult they have the advantage that loops and other control
> structures can be better integrated into the OO concept. Smalltalk for
> example doesn't need Iterators for that reason.

Python is not a functional programming language. As I understand it, you
can't really have convenient closures because it is not lexically scoped
(I can give you inconvenient closures...) and there were lengthy
discussions about that some months back. I didn't follow it too closely
because I didn't know what lexically scoped meant then, but the upshot was
(I think) that lexical scoping isn't really the "Python way".

Python isn't really an OO language either, in the sense of something like
Smalltalk or Eiffel.

I have heard it said that Python is a language that "gets it's compromises
exactly right" and that about sums it up for me.

> Lately ee have had already some examples in this group where loop
> overhead was a major factor. 

Really? When? Sorry, I don't want to sound disbelieving, but I don't
remember that.

> I measured different kind of loops and
> even the more general while loop was faster in squeak than anything
> else in python.

Python, as I said, is neither functional nor purely OO; it's not
particularly fast either. Fast enough, generally.

> Functions calls seems also to be much slower in Python ...

*That* I believe. Not clear what to do about it though.

> Another problem with loops could be reference counting.  I'm wondering
> how often a reference count get's changed when running a loop ...

Not vastly more often than in regular code, I'd guess. It's true that
executing

1+1

spends longer twiddling refcounts than doing the work, but that's Python.

I can't decide which Peters-style ending I want here, so you're all
getting two:

not-OO-functional-or-fast-but-still-great-ly y'rs
it's-not-OO-or-functional-it's-Python-ly y'rs

Michael





From tim_one at email.msn.com  Tue Jul 20 02:32:03 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 20 Jul 1999 02:32:03 -0400
Subject: How fast can we multiply?
In-Reply-To: <37937DF5.519CF1AA@callware.com>
Message-ID: <001901bed279$946e87a0$139e2299@tim>

[Ivan Van Laningham]
> Oh, bad title, most certainly!  ;-)  *I'm* interested!  I don't
> understand most of what you guys are talking about, but I can see that
> if I could afford to spend the time studying it I would be rewarded.

Here's the heart of it.  Suppose you want to multiply 10*A+B by 10*C+D,
where A, B, C, D are all in range(10).  The answer is

    100*A*C + 10*(A*D + B*C) + B*D

Multiplying by 100 or 10 is just tacking zeroes on the end, so that's cheap.
Adding is cheap too.  Multiplying is expensive (well, imagine you already
understood what follows so that A, B, C, D each had hundreds of digits
instead of one <wink>).

Now time for a trick:  consider (A+B)*(C+D).  That's A*C+A*D+B*C+B*D.  So
the answer above is the same as

    100*A*C + 10*((A+B)*(C+D)-A*C-B*D) + B*D

and there are only *three* unique products there instead of 4:  A*C,
(A+B)*(C+D), and B*D.  That means we can compute the product of two B-bit
numbers via 3 multiplications of B/2-bit numbers, plus some cheap stuff.
The same trick applies to the B/2-bit numbers, down & down.  In the end that
implies we can do multiplication in time proportional to B**log2(3) instead
of B**2.

The method Chris actually used computes (A-B)*(D-C) instead for obscure
technical reasons (history, plus that subtracting instead of adding reduces
the size of the multiplicands a little on average).

> On topic, though, doesn't Knuth include in his "How Fast can We
> Multiply?" section a discussion of using modular arithmetic to speed up
> multiplies by a pretty large factor?  Now, maybe I've misunderstood the
> discussion so far so badly that I missed your use of it, but I don't
> recall seeing anything about it.

We didn't mention it.  Python doesn't store numbers in modular form, so it's
not that interesting -- it's expensive to convert in to and out of modular
form.  If it kept them in modular form, then multiplication would be
essentially linear-time.  If you want that-- and can live with numbers in
modular form over long stretches of calculations --it's easy enough to write
a class to do that.

> I would be interested--just not competent!  But a Python version of
> Mathematica could indeed be a killer app.  (Or is that a killer rabbit?)

I thought Mayan calendar analysis was Python's killer app!

or-is-that-an-app-killer<wink>?-ly y'rs  - tim





From marjea at uswest.net  Thu Jul 15 15:20:33 1999
From: marjea at uswest.net (mary jeanette)
Date: Thu, 15 Jul 1999 14:20:33 -0500
Subject: I envy newbies
Message-ID: <378E3481.EE31C97@uswest.net>

I decided to learn python because I heard it was a very easy language to
learn. So I downloaded it and bought a couple of very expensive
paperback books from o'reilly only to discover that they are written for
unix. In other words, typing in the examples from the book into the
interpretor only yields various syntax errors. I have no idea what I am
supposed to change in order to make the examples work, and apparently
there is nowhere to turn in order to find out. Q: Am I supposed to know
DOS as a prerequisite? The idea of having to learn DOS is appalling.
Should I try partitioning my hard drive and installing Linux or freeBSD?
Perhaps I should learn Perl instead, I have seen books on Perl written
for Win32 systems.   Thanks for your advice.




From fdrake at cnri.reston.va.us  Tue Jul 13 15:59:56 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Tue, 13 Jul 1999 19:59:56 GMT
Subject: Pickling C objects
In-Reply-To: <378B27A9.4201@cs.bham.ac.uk>
References: <378B27A9.4201@cs.bham.ac.uk>
Message-ID: <14219.39612.343183.547443@weyr.cnri.reston.va.us>

Stuart Reynolds writes:
 > I'm writing some new Python classes with a C implementation. How do you
 > make a C object pickleable?

  Take a look at the source for the parser module in the standard
source distribution; it uses the copy_reg module to register a
mechanism to support pickling.  It worked when I wrote it, but I
haven't tried it lately.  (A good candidate for a regression test!)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From tismer at appliedbiometrics.com  Tue Jul 20 19:04:22 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 20 Jul 1999 23:04:22 GMT
Subject: How fast can we multiply?
References: <001801bed279$92c348a0$139e2299@tim> <m2wvvvqwjb.fsf@netmeg.net>
Message-ID: <37950076.6A0CAC68@appliedbiometrics.com>


Les Schaffer wrote:
> 
> Tim and Christian spoke:
> 
> C> Then I also had a look into NumPy and its matrix multiplication.  I
> C> stunned. Again no attempt to optimize big O. It could be done with
> C> Python, the algorithms are known, the fruit is waiting to be picked.
> 
> T> Large dense matmult isn't very common.  Sing along:
> 
> i dont see the connection between floating point matrix multiplies and
> the integer arithmetic tim sketched out in the previous post.
> 
> Spock, explain!

The floating point property is not the point.
It is just a similar strategy : divide and conquer.
Multplying long ints is breaking them into vectors of smaller
size, and then figure out where a mult can be saved by some
more adds.
The way for matrices is factorizing the multiplicants
into smaller ones and then figuring out which mults
can be saved.

And for both also applies: The low level math for the small
pieces which are below a threshold, are still done by the
fast, "dumb" C routines.
Elegant problem reduction is done with Python which controls
that, and saves calls to the low level.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From garry at research.att.com  Wed Jul 14 13:37:22 1999
From: garry at research.att.com (Garrett G. Hodgson)
Date: Wed, 14 Jul 1999 17:37:22 GMT
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <377B454C.6F59@mailserver.hursley.ibm.com> <7liguf$58o$1@Starbase.NeoSoft.COM> <m0JTEFA0DTf3EwLV@ugglan.demon.co.uk> <slrn7nss0n.5u0.wtanksle@dolphin.openprojects.net> <378077E1.5E65@mailserver.hursley.ibm.com> <slrn7o58l7.4vr.wtanksle@dolphin.openprojects.net> <3783305C.33FA@mailserver.hursley.ibm.com> <slrn7o7d7o.6la.wtanksle@dolphin.openprojects.net> <3785B997.2001@mailserver.hursley.ibm.com>
	  <w6btdmw4mm.fsf@wallace.nextel.no> <3789ABAB.1E71@mailserver.hursley.ibm.com> <slrn7okdnn.ep3.wtanksle@dolphin.openprojects.net> <n7vi3.21348$AU3.605286@news2.giganews.com>
Message-ID: <378CCAD2.97D4FEA7@research.att.com>

Christopher Browne wrote:
> 
> On Mon, 12 Jul 1999 18:41:59 GMT, William Tanksley
> <wtanksle at dolphin.openprojects.net> wrote:
> >(although I am curious as to whether an indent/dedent based syntax might
> >be amusing in Lisp).
> 
> That would provide opportunity for the implementation to be criticized
> inanely by *absolutely everyone.*
> 
> Not only those that see Lisp as "too many parens," but also those that
> knee-jerk into thinking that Python must be worthless due to
> "indentation-sensitivity."

if you could only add in the required use of all combinations of
[$&@_;{}] as well, you might have the perfect language for flamebait.

-- 
Garry Hodgson			comes a time
garry at sage.att.com		when the blind man
Software Innovation Services	takes your hand, says,
AT&T Labs			"don't you see?"



From wtanksle at dolphin.openprojects.net  Wed Jul 28 15:39:47 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Wed, 28 Jul 1999 19:39:47 GMT
Subject: 1.5.2 for: else:
References: <1278996512-12636218@hypernet.com>
Message-ID: <slrn7pun42.2ro.wtanksle@dolphin.openprojects.net>

On Tue, 27 Jul 1999 21:43:00 -0500, Gordon McMillan wrote:
>William Tanksley wrote:
>> On Tue, 27 Jul 1999 18:35:10 -0500, Gordon McMillan wrote:

>> >Maybe it violated your expectations, but it's much more valuable than 
>> >what you expected. After all, testing for an empty sequence is a 
>> >no-brainer. But finding a match in a list, and testing whether you 
>> >fell off the end without finding one, is (without the else clause) a 
>> >much messier proposal.
>[rant]
>> I would also disagree with you that my interpretation would result
>> in disutility: you'd merely have to use exceptions instead of
>> 'break'.  
>[more rant]

>--------billy.py--------------------
>class neverMind: pass

>try:
> for x in [1,2,3]:
>  print x
>  if x == 2: raise neverMind
>except neverMind:
>  print "modern else clause here"

>for x in [1,2,3]:
> print x
> if x == 2:
>  break
>else:
> print "Hi Billy!"
>---------------------------------------------------

>C:\TEMP>python billy.py
>1
>2
>modern else clause here
>1
>2

>C:\TEMP>

>Notice the difference?
>You need to raise the exception if you've gone through the loop and 
>not found 2. The best way to do that <snicker> is with an else 
>clause.

Sure do!  In spite of the fact that I actually did understand the modern
meaning of 'else', in practice I manage to screw it up anyhow.  I can
choose to make this a commentary on my programming skill, OR on the
obtuseness of the current definition of 'else'.  Hmm, which to choose,
which to choose...

:)

However, you're wrong about the best way to detect a 2 in the list using
this style.  To steal the same example:

try:
 for x in [1,2,3]:
  print x
  if x == 2: raise neverMind
 print "modern else clause here"
except neverMind: print "found two"
else:
 print "another place to put the modern else clause"

Wow, TMTOWTDI.  Oops, did I say the wrong thing?

Of course, by demonstrating both ways to do it, I'm making the code more
complex; but you can see that this solution solves the problem quite
elegantly.

The place to use MY definition of 'else', OTOH, is when you're building a
structure during the iteration, but the result of an empty loop is
different than the initial value the variable has to have when going
through the loop.

Here you go.  This example might convert a Python list into a Lisp
tree-list -- but to imitate Lisp (okay, it's a toy problem, I'm cheating)
an empty list produces NIL ~= None.

var = Node(None,None)
for x in pyList:
  var.car = x
  var.cdr = Node()
else:
  var = None

>insufferably-sanctimoniously-y'rs

>- Gordon

unstoppably-confidantly y'rs

-- 
-William "Billy" Tanksley



From nascheme at ucalgary.ca  Thu Jul 15 04:58:59 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 15 Jul 1999 08:58:59 GMT
Subject: Getters and Setters
References: <7mjrqg$e4o@ds2.acs.ucalgary.ca> <7mjune$e2g@ds2.acs.ucalgary.ca> <378D8E21.957C6AEA@bioreason.com>
Message-ID: <7mk7sj$e5u@ds2.acs.ucalgary.ca>

Andrew Dalke <dalke at bioreason.com> wrote:
>Here's one optimization, don't recreate getter/setter objects.

If you don't recreate the objects you end of with reference
loops.  I don't see any way around this.

There is another problem with your optimizations.  What happens
if the value of the attribute changes?  My second version had
this problem too.  Also, what if someone holds a reference to the
method and calls it periodicly?  This is the best I can do:

class _Getter:
    def __init__(self, dict, name):
        self.dict = dict
        self.name = name

    def __call__(self):
        return self.dict[self.name]

class _Setter:
    def __init__(self, dict, name):
        self.dict = dict
        self.name = name

    def __call__(self, value):
        self.dict.__dict__[self.name] = value
        
class GetterSetter:
    def __getattr__(self, name):
        try:
            if name[:3] == 'get':
                return _Getter(self.__dict__, name[3:])
            elif name[:3] == 'set':
                return _Setter(self, name[3:])
        except KeyError:
            pass
        raise AttributeError


Using lambda instead of __call__ gives a small speed increase
but, ask you point out, changes the symantics.


    Neil



From catlee at my-deja.com  Tue Jul  6 13:20:07 1999
From: catlee at my-deja.com (catlee at my-deja.com)
Date: Tue, 06 Jul 1999 17:20:07 GMT
Subject: problem with lists
References: <378220B1.9E0C272@daimi.au.dk>
Message-ID: <7ltdrp$jn1$1@nnrp1.deja.com>

You need to create a nydata variable before you can insert things into
it or index it.

Try putting this line near the top of your code:
nydata=[]

Cheers,
Chris


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From p.t at iname.com  Thu Jul 29 06:42:39 1999
From: p.t at iname.com (Peter Torstenson)
Date: Thu, 29 Jul 1999 07:42:39 -0300
Subject: How to set text color using pyKDE 0.8?
Message-ID: <7npb63$2fdn@enews3.newsguy.com>

I'm using pyKDE 0.8 binding for python.
I made a QLabel on the main widget. Now I'd like to change the color of the
text inside this QLabel.
I managed to change the backgroundcolor and the font. But I can't find any
info on how to change the color of the text, or the foregroundcolor.


/Peter Torstenson
Rio de Janeiro





From nick.bower at ssec.wisc.edu  Fri Jul  9 16:35:26 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Fri, 09 Jul 1999 15:35:26 -0500
Subject: Grail won't work
References: <Pine.LNX.4.04.9907100016400.3944-100000@emerald.netskate.ru>
Message-ID: <37865D0E.67C2854A@ssec.wisc.edu>


> On Fri, 9 Jul 1999, Nick Bower wrote:
> >     if not isabs(path):
> > NameError: isabs
> >
> > What does this mean?  Cheers, Nick.
>
>    This:
>      if not isabs(path):
>    should be:
>      if not os.path.isabs(path):

thanks.  and "return normpath" to "return os.path.normpath" at a guess.

the browser doesn't display images or let you activate links.  it spits the dummy
when i display frame source.  all in the first 5 seconds of testing.  so i try and
quit - big crash.  i thought this thing had been around for a while!!!  (i'm
running python on NT)

nick





From markusk at bidra241.bbn.hp.com  Mon Jul 26 05:32:56 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 26 Jul 1999 11:32:56 +0200
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi>
Message-ID: <p51zdvn4c7.fsf@bidra241.bbn.hp.com>

Markus Stenberg <mstenber at cc.Helsinki.FI> writes:

> gward at cnri.reston.va.us (Greg Ward) writes:
> > sreekant kodela <sreekantk at freenet.co.uk> wrote:
> > If you're concerned about runtime performance -- ie. how long it takes
> > to mangle strings and loop over lists and so forth -- I wouldn't worry
> > about it.  Yes, Python is slow, but the Internet is so much slower that
> > it hardly matters.
> 
> Even in websites' case, I would worry; Python doesn't use multiCPU very
> well, and with single CPU, it is fairly easy to saturate one with
> pure-Python code, especially if it is badly designed to boot.
> 
> For AMD K6-2 300mhz, your mileage may vary:
> 
> Normal null method call :  156332.1874/sec [0.19s] (6.3966us/call)
> (=simple, def foo(): pass call)
> Apply null method call  :   39783.6246/sec [0.75s] (25.136us/call)
> (=simple, foo-call using apply)

Yes that is much too slow. Good OO style requires a lot of function calls. 
I'm sure Visualworks a commercial Smalltalk implementation can do millions
of calls on this machine. Ok it's a JIT and it's commercial but's still Smalltalk
and therefore at least as dynamic as python. I guess even Squeak (www.squeak.org)
would be faster. If you would sent us code, I could prove it.  

[more interesting numbers deleted]
> - Speed-does-matter-ly yours, Markus

Yes Speed does matter even for scripting languages. 
IMHO the whole advantage of scripting languages is that they
are more productive than other compiled languages like C. 

The faster a scripting languages is the less you need to do in lower
level languages. Therefore making the scripting language faster should
increase productivity. 

> 
> P.S. This might surface as one example for 'why scripting languages don't
> always cut it' in my master's thesis; if I have made some grievous
> error(s?), corrections are welcome. 
>


 
> P.P.S. Too bad Python doesn't have anything that creates _nice_ code, like
> Stalin for Lisp ;-)

Yes Stalin is great but it's a static compiler. That limits it's appliacations and
also really works if the generated code does exactly the same as the interpreted
code. A profiling JIT like Sun's Hotspot compiler for Java, might be a better choice. . 

> 
> -- 
> ----- BEGIN GEEK CODE v3.0 BLOCK (*) -----
> GCS d- s+: !a C+(++) UL+++$ P--- L+++ E++ W-- N+ W--- O---- M-- V- PS+++
> Y+ PGP++>+++ t+@ 5++ X+ R* tv-- b+++ DI+++>++++ D---- G++ e++ h* r% z+
> ----- END GEEK CODE v3.0 BLOCK -----
> * For spoilers, check http://www.geekcode.com

-- 
Markus Kohler  mailto:markus_kohler at hp.com



From jeff at todday.aiai.ed.ac.uk  Fri Jul  2 12:59:23 1999
From: jeff at todday.aiai.ed.ac.uk (Jeff Dalton)
Date: 02 Jul 1999 17:59:23 +0100
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com>
	<Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com>
	<375FDCA7.62FF@mailserver.hursley.ibm.com>
	<7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com>
	<37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu>
	<37776FFA.2D85@mailserver.hursley.ibm.com>
	<lw7looxx7m.fsf@copernico.parades.rm.cnr.it>
	<876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com>
	<lw67471fyi.fsf@copernico.parades.rm.cnr.it>
	<3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org>
	<377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov>
	<slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net>
	<wkzp1g2t2e.fsf@ifi.uio.no> <ey3r9mskvja.fsf@lostwithiel.tfeb.org>
Message-ID: <x27lojoums.fsf@todday.aiai.ed.ac.uk>

Tim Bradshaw <tfb at tfeb.org> writes:

> * Lars Marius Garshol wrote:
> 
> > Huh? Forth dates back to the 60s, whereas Scheme is from 1975, and I'm
> > quite unsure of whether Chuck Moore knew Lisp at all.

Well *Lisp* dates back to the late 50s.  Scheme may be from 75,
but Lisp is a different matter.  And lambda calculus is, what,
1949?




From tgabriel at bellsouth.net  Mon Jul 19 06:51:57 1999
From: tgabriel at bellsouth.net (Xenophanes)
Date: Mon, 19 Jul 1999 06:51:57 -0400
Subject: Terminating Python...
References: <BFC0587634C3D111AACB00805FA766FA02C1C652@exchny12.ny.smb.com>
Message-ID: <3793034D.D670FFB0@bellsouth.net>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19990719/b86036ee/attachment.html>

From tim_one at email.msn.com  Fri Jul 16 21:25:26 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 16 Jul 1999 21:25:26 -0400
Subject: Stackless & String-processing
In-Reply-To: <7mnidi$t2b$1@cronkite.cc.uga.edu>
Message-ID: <000801becff3$3fe92880$91a22299@tim>

[Neel Krishnaswami]
> ...
> I found a nice description of this in a paper called "Monadic Parser
> Combinators" by Graham Hutton and Erik Meijer. Its URL is:
>    http://www.cs.nott.ac.uk/Department/Staff/gmh/monparsing.ps

[Graham Matthews]
> Yes this is a nice place to start. Another nice place to look is in
> the Clean documentation. Somewhere in their docs (I think in the old
> book) there is a Clean implementation of combinators (without all the
> "do" notation).

A nice paper on Clean parsing combinators can be downloaded as item 5 from
the "Case Studies" list at

    http://www.cs.kun.nl/~clean/

Interestingly enough, the next paper on the list ("An Interpreter") declines
to use  combinators for its parsing task, instead building a conventional
"by hand" recursive descent parser.

another-two-years-and-they'll-discover-lex/yacc<wink>-ly y'rs  - tim





From graham at sloth.math.uga.edu  Sun Jul 18 11:43:42 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 18 Jul 1999 15:43:42 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com> <170719990020380651%pereira@research.att.com> <7mqaht$r27$2@cronkite.cc.uga.edu> <170719991835331004%pereira@research.att.com>
Message-ID: <7mssne$m8r$1@cronkite.cc.uga.edu>

<graham at sloth.math.uga.edu> wrote:
> I don't really see why using parser combinators means that you cannot
> optimise. You are not precluded from using regular expression parsers
> as primitive parsers, nor are you precluded from using LALR(n) parsers
> as primitive parsers. Hence you can optimise those parts of your
> parsing with automaton if you want. So I don't really understand your
> criticism.
Fernando Pereira (pereira at research.att.com) wrote:
: The monadic approach mixes up the search structure for pattern matching
: or parsing (what is elsewhere represented as an automaton) with other
: data processing. Unless your optimizer can solve the halting problem,
: it will not be able to determinize otherwise determinizable control
: structures because it cannot figure out the data dependencies.

You have totally misunderstood what I said. I said "you can have regular
expression parsers as primitive parsers". There is no need for the compiler
to "solve the halting problem". The programmer is free to use regular
expression/automaton techniques where he/she wants. Such parsers take
strings and return ASTs. They can then be combined with other parsers
as if they were primitive parsers. This is purely a programming issue
and has nothing to do with the halting problem, compiler optimisations,
etc. To make it clearer a programmer can do

	p = some_parser_that_uses_automaton :: String -> Ast
	q = some_parser_that_uses_some_other_technique :: String -> Ast
	pq = p <&> q
		where <&> is the combinator that sequences parsers

Where does the halting problem enter into this? The program has to write
the autoton code him/herself, but there are fp tools that do this, much
like for example Yacc.

<graham at sloth.math.uga.edu> wrote:
: > Also as I said parser combinators are really more of a sequencing 
: > technique -- one that applies nicely to parsing.
Fernando Pereira (pereira at research.att.com) wrote:
: Sure, but so what? Automata are too a sequencing technique, usable for
: many other things besides pattern matching and parsing. The fundamental
: difference is that automata-theoretic approaches separate neatly the
: representation of sequences from what one wants to do with the parts of
: those sequences, allowing optimizations that are difficult to obtain
: within general-purpose programs.

I still don't understand why you think the combinator approach is so
inefficient, given that you can have automaton based parsers embedded
in it at will. 

The point of saying that combinators are more a of sequence tool than
a parsing tool is to point out that they are more general than automaton.
You can program much richer sequencing with combinators than you can
with automaton (not suprisingly). For those parts of your parser or
sequenceing where you only need automaton you use automaton as outlined
above. But where you need more powerful sequencing you have it in the
combinator approach.


<graham at sloth.math.uga.edu> wrote:
: > Can you explain why? I don't see it, since you can have reg expr parsers
: > and other such fast parsers as primitive parsers that you combine ...
Fernando Pereira (pereira at research.att.com) wrote:
: Either you put all the interesting stuff in separate primitive parsers,
: in which case I'm not sure what's the point of the monadic approach,

Err isn't this what I said you do ... I said "you can have regular
expression parsers as primitive parsers". The point of the monadic
approach is that is allows you to cleanly program arbitrary sequencing
when you need it. Moreover it makes for sound software engineering,
etc. That too is what I said in the last mail ....

graham
-- 
            The girls is all salty, the boys is all sweet, 
                      the food ain't too shabby 
                    an' they piss in the streets
    In France, way down in France, way on down, way on down in France



From fdrake at cnri.reston.va.us  Mon Jul 12 10:20:43 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Mon, 12 Jul 1999 14:20:43 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
In-Reply-To: <7m923i$dem@dfw-ixnews5.ix.netcom.com>
References: <3787D36A.815B7C29@ssec.wisc.edu>
	<3787D8DA.3774AA7D@ssec.wisc.edu>
	<7m8p7t$o7s@dfw-ixnews10.ix.netcom.com>
	<3787FB51.12FE0048@ssec.wisc.edu>
	<7m923i$dem@dfw-ixnews5.ix.netcom.com>
Message-ID: <14217.63931.75257.566266@weyr.cnri.reston.va.us>

Aahz Maruch writes:
 > I think I understand how it happened: os.getcwd() is (probably
 > rightfully) documented there, so someone thought os.chdir() belonged
 > there, too.  AFAICT, that's the only mistake in that area of the docs,
 > though I could make a case that os.getcwd() should be moved (or at least
 > have a pointer in "F&D").

Aahz,
  Are you saying that (either of) these are in the wrong place?  The
notion of "current directory" is tied only to the entry in the process 
table; should it be documented otherwise?
  The "Process Parameters" section contains functions with provide any
direct access to the process table entry, both queries and changes, so
I think it really is the right place for these two functions.
  I can easily add a textual reference to the right place within the
"Files and Directories" section, but you'll need to explain very
clearly why the descriptions should be moved.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From conrad at vanir.srl.caltech.edu  Sat Jul 10 16:08:37 1999
From: conrad at vanir.srl.caltech.edu (Conrad Steenberg)
Date: 10 Jul 1999 20:08:37 GMT
Subject: Printing returned object in embedded app
References: <378694FC.D857F68A@srl.caltech.edu> <6wxh3.1472$6M6.450919@news.shore.net> <slrn7oep57.a57.conrad@vanir.srl.caltech.edu>
Message-ID: <slrn7oev8a.a6k.conrad@vanir.srl.caltech.edu>

On 10 Jul 1999 18:24:33 GMT, Conrad Steenberg <conrad at vanir.srl.caltech.edu>
wrote:
>On Sat, 10 Jul 1999 01:40:18 GMT, Michael P. Reilly <arcege at shore.net> wrote:
>>
>>Try:
>>  PyObject_Print(object, stdout, Py_PRINT_RAW);
>
>Thanks! 
>
>Conrad
>-- 
Hmmm... Sorry to reply to my own message, but it seems like PyRun_String 
returns only the Py_None object! 

How can I get the result of a calculation, say
obj=PyRun_String ("1+1", start, globals, locals);
which should be a numeric object? 

The behaviour I'm trying to get is the same as running python on a terminal:
when an instruction have a printable result, I want to print it.

Cheers

Conrad

-- 



From andrew at andrewcooke.free-online.co.uk  Tue Jul 13 12:18:59 1999
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Tue, 13 Jul 1999 16:18:59 GMT
Subject: making a Windows NT service out of a python program
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F20@RED-MSG-50> <378B2AA8.58477120@appliedbiometrics.com> <378B2C50.B7C2134A@cern.ch>
Message-ID: <7mfot5$n30$1@nnrp1.deja.com>

In article <378B2C50.B7C2134A at cern.ch>,
  "Haimo G. Zobernig" <Haimo.Zobernig at cern.ch> wrote:
> I'm in exactly the same situation - no idea what really distinguishes
a
> "service"
> from any old program, or how programs behave differently when run as a
service.
> Anyway, thanks for Bill's pointer, I will be trying out the
pipeTestService.py
> example and see if I can adapt that for my purpose.

Services are described quite well in the help files that come with
VC/Developer Studio, if you're lucky enough to have that ;-)

They are useful because you can specify an account for them to run in
(assuming you have sufficient priveliges), set whether they should be
started automatically, and let them access the user's screen.

At base, they are just ordinary programs, but they have to meet a
particular interface so that they can be controlled (ie provide methods
to stop, start, pause them etc.)  Writing programs to install/start
services is more complicated, but hopefully you won't need to do that
(the Control panel can stop and start them, and I assume the Python
package others mentioned has an installer).  If you poke around in the
registry you'll find entries that associate services with particular
programs etc.

The worst thing about them is that they can be tricky to debug.  If they
are running in the system account then they have a restricted
environment and the standard output channels send useful data to
somewhere I never found.  So the first thing you want to do is open a
file somewhere for debug information...

All this is from C and Java knowledge, I've never written a Python
service, so I can't give more specific help.

Good luck,
Andrew




Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From tdc0331 at hti.com  Tue Jul 27 14:07:27 1999
From: tdc0331 at hti.com (David Stokes)
Date: Tue, 27 Jul 1999 13:07:27 -0500
Subject: Serial port IO question
Message-ID: <379DF55E.EAFCBBF7@hti.com>

I am attempting to send a ^M (Control-M) down a serial port and then
read back from that same port.  The application is an old UPS that will
send back system status after a ^M.  The code below opens the port but
the readline seems to be returning what I wrote.  My guess is that I
haven't really grabbed I/O stream for the read operation.  Any
suggestions?

Dave Stokes
-------

#!/usr/bin/env python
import sys, string

upsdev = "/dev/ttyS1"   # Serial port attached to ups


class   get_ups:
    def __init__(self):
        print "Getting ups status"

    def get_ups_data_p(self):
        upsdata = open( upsdev, 'r+');      # open port to ups
        upsdata.write("\015")                   # send ^M to ask for
status
        upsdata.flush()
        upsbuffer = upsdata.readline()       # get status back
        upsdata.close()

if __name__ == '__main__':
    ups_status = get_ups()
    ups_status.get_ups_data_p()






From john at mediamanager.com.sg  Wed Jul  7 06:05:28 1999
From: john at mediamanager.com.sg (John)
Date: Wed, 7 Jul 1999 18:05:28 +0800
Subject: Error in python script
Message-ID: <3783f651.0@news.smartnet.com.sg>

Hi,

I have written the program correctlly but I get the following message when I
run the program in Windows NT.

But the same program when I run it in Win 95 it is working find

This is the program

#!c:/program files/python

import string
import sys
import os
import cgi
#import urllib

data_file = open('c:/inetpub/scripts/FilmFestReview.txt', 'r')
form = cgi.FieldStorage()
header = \
"""
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
<html>
<head>
<title>Extract & Write</title>
</head>
<body>
"""

footer = \
"""
</body>
</html>
"""

out = sys.stdout

begin_table = 1
out.write(header)
for line in data_file.readlines():
        if begin_table:
                out.write('<table border = "1">')
                begin_table = 0
        if line[:6] == "------":
                out.write('</table>')
                out.write('<br>')
                begin_table = 1
                continue
        field_name, field_value = string.split(line, ':')
        out.write('\t<tr>\n')
        out.write('\t\t<td>\n')
        out.write('\t\t\t%s\n' % field_name)
        out.write('\t\t</td>\n')
        out.write('\t\t<td>\n')
        out.write('\t\t\t%s\n' % field_value)
        out.write('\t\t</td>\n')
        out.write('\t</tr>\n')


out.write(footer)


This is the following error I encounter when I run it using the Browser

Traceback (innermost last):
  File "C:\InetPub\scripts\ffratingextract.py", line 39, in ?
    field_name, field_value = string.split(line, ':')
ValueError: unpack list of wrong size






From tratt at dcs.kcl.ac.uk  Sat Jul 24 08:12:43 1999
From: tratt at dcs.kcl.ac.uk (Laurence Tratt)
Date: Sat, 24 Jul 1999 12:12:43 GMT
Subject: Python documentation available as ZIP files
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us> <7nal1u$6cb$1@tribune.oar.net>
Message-ID: <b3c1a02649.tratt@tratt.freeuk.co.uk>

In message <7nal1u$6cb$1 at tribune.oar.net>
          "David Robinow" <drobinow at dayton.adroit.com> wrote:

[Fred L. Drake]
>> I've received enough hostile mail from Windows users not understanding
>> how to unpack .tgz archives (er, WinZip works for me...) that I will be
>> making most of the documentation packages available as ZIP archives as
>> well as .tgz archives.
> This is really sad.  You're not really doing Windows users any favors by
> doing this. "tar" and "gzip" are "must know" tools.
> Let me quote from the GNU macs on Windows FAQ
>
> "You can find precompiled versions of all of the compression and archive
> utilities in ftp.cs.washington.edu/pub/ntemacs/utilities/. "
>
> It would seem to make more sense to include gzip and tar.on the web site.

Sad or not, the simple truth is that many authors who distribute Python
programs (that tend to run on many platforms) as .tar.gz seem to end up with
full inboxes of "I can't decompress this" messages. And, in my experience,
not all of them are from Windows users. Ultimately, the simplest and most
effective solution seems to be to ship .tar.gz files for those who want
quick downloads, and .zip files for people who don't.

Admittedly, as a non-Windows user, WinZip has always worked when I've tried
it on .tar.gz files, but maybe I'm not using it correctly <wink>.

If you want to preach the benefits of .tar.gz I'd reccomend preaching about
.tar.bz2... And that way you can try educating lots of Unix users as well as
the Windows gang <wink>.


Laurie
-- 
http://eh.org/~laurie/comp/python/



From nascheme at ucalgary.ca  Fri Jul 16 21:25:20 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 17 Jul 1999 01:25:20 GMT
Subject: Python and Boehm-Demers GC, I have code.
References: <7mo2iu$iqg@ds2.acs.ucalgary.ca>
Message-ID: <7mom20$dgo@ds2.acs.ucalgary.ca>

Neil Schemenauer <nascheme at ucalgary.ca> wrote:
>I have to fix up modules which use other mallocs (readline and
>_tkinter for sure).  Guido, will a patch be accepted which
>changes Python code to use Py_Mem_* instead of the libc names
>where approprate?  I will have to go through the extension
>modules and fix them by hand.

It seems that most of the Python code was already using the
PyMem_* functions.  I only had to fix a few things.  The garbage
collecting version is still as fast as the regular version.  I
think readline, zlib, and _tkinter are all working okay now too.
The patch is at:

	http://www.ucalgary.ca/~nascheme/python/gc-python.patch





From fredrik at pythonware.com  Wed Jul 14 11:12:22 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 14 Jul 1999 15:12:22 GMT
Subject: Recursive method
References: <1280168234-10452044@hypernet.com> <378CA422.B2E99CB7@eml.villa-bosch.de>
Message-ID: <004901bece0b$46982210$f29b12c2@secret.pythonware.com>

Ralph Gauges <ralph.gauges at eml.villa-bosch.de> wrote:
> You are absolutely right. So far I had some methods in my
> class and made the whole gui right inside the class
> definition. This must have created some problems with
> namespaces and 'self'. Now I moved the whole gui creation
> inside the __init__ method and now 'self' works.
> Now I have one more question: Why is there no 'self' right
> inside the class definition?

probably because there is no instance when the
class definition is executed...

("class" is a *statement*, not a declaration, and
is executed when the script is run.  see:
http://www.python.org/doc/current/ref/class.html
for details.  "def" is also a statement, btw...)

</F>




From joakal at my-deja.com  Thu Jul  1 23:11:36 1999
From: joakal at my-deja.com (joakal at my-deja.com)
Date: Fri, 02 Jul 1999 03:11:36 GMT
Subject: Tkinter just WON'T work right!
Message-ID: <7lhal1$7o$1@nnrp1.deja.com>

Sorry to be so inspecific about my subject =) Also, I'm sure this
question is asked frequently, but I couldn't find what I was looking for
in any of the 1400 or so messages here (too many of them start with
``Help!'' or ``Python Newbie''). Anyway, here it is:

I tried to use the Grail browser under Slackware 4.0 and Python 1.5.1. I
got the error saying it's not configured for Tkinter, so I went into my
/usr/lib/python1.5/config/Setup file and uncommented the appropriate
stuff.

I tried again, same problem. I read a bit more, and it said I have to
rebuild Python to add Tkinter (don't remember having to do that before).
So, argh, I tried to do that. Now I get an error when I use 'make'
saying there's an error when it gets to sedscript.

Can anyone help?
Thanks in advance!


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From kas at maps.magnetic-ink.dk  Tue Jul 20 11:43:13 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Tue, 20 Jul 1999 17:43:13 +0200
Subject: Finding absolute pathname?
References: <7n23mm$ukr$1@nnrp1.deja.com>
Message-ID: <slrn7p968h.3i8.kas@titan.magnetic-ink.dk>

rdudfield at my-deja.com <rdudfield at my-deja.com> wrote:

> How do I find the absolute pathname of the current directory?

os.path.abspath()


  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From hnowak at cuci.nl  Thu Jul 29 01:43:38 1999
From: hnowak at cuci.nl (Hans Nowak)
Date: Thu, 29 Jul 1999 05:43:38 GMT
Subject: 1.5.2 for: else:
In-Reply-To: <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com>
References: <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com>
Message-ID: <199907290544.HAA08030@dionysus.fw.cuci.nl>

> William Tanksley <wtanksle at dolphin.openprojects.net> wrote:
> > The very worst part of the current else: behavior is that it changes the
> > meaning of else.  In other constructs, else is an alternate path to take
> > if the data being tested fails a single expression test.  In this
> > contruct, else is a path taken if the code block belonging to the
> > previous test executes a certain instruction.
> 
> nope.  you've got it all backwards.  consider this:
> 
>     for an "if" statement, "else" is a path taken if the
>     expression evaluates to false.
> 
>     for a "while" statement, "else" is a path taken if the
>     expression evaluates to false.  or in other words,
>     when the loop terminates by natural causes.
> 
>     for a "for" statement, "else" is a path taken when
>     there are no more elements to loop over.  or
>     in other words, when the loop terminates by
>     natural causes.
> 
> not that complicated, was it?

I think quite a few people would expect only one of the parts to 
execute, like in if..else, and not both of them. So this extension of 
the meaning of 'else' may come as a surprise for those (including me, 
actually).

But, the feature itself is nice, and the keyword is not going to 
change, is I guess it's no use whining about it. ^_^

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
Python Snippets: http://www.hvision.nl/~ivnowa/snippets/
The Purple Kookaburra Forum: http://www.delphi.com/kookaburra/



From kajiyama at grad.sccs.chukyo-u.ac.jp  Tue Jul 20 17:10:13 1999
From: kajiyama at grad.sccs.chukyo-u.ac.jp (Tamito Kajiyama)
Date: 20 Jul 1999 21:10:13 GMT
Subject: binding keys to slider in Tkinter?
In-Reply-To: heather@thalamus.wustl.edu's message of 19 Jul 1999 11:26:00 -0500
References: <7mvjio$dh8@thalamus.wustl.edu>
Message-ID: <KAJIYAMA.99Jul21061013@pseudo.grad.sccs.chukyo-u.ac.jp>

heather at thalamus.wustl.edu (Heather A. Drury) writes:
| 
| I'm trying to bind the arrow keys to allow a user to increment
| (or decrement) a counter using the Scale widget and Tkinter.
| I seem to be unable to bind anything to the Scale widget.
| I'm sure I'm missing something.

For a widget to get key press/release events, the widget first needs to
get the input focus.  You can pass around the input focus by pressing
the tab key.  The Scale widget has the key bindings that you want
already.

If you want to control the Scale widget using arrow keys even when
another widget has the input focus, you have to let the widget to pass
the relevant events to the Scale widget.

Regards,

--
KAJIYAMA, Tamito <kajiyama at grad.sccs.chukyo-u.ac.jp>



From mal at lemburg.com  Tue Jul  6 13:19:09 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 06 Jul 1999 19:19:09 +0200
Subject: Getting list of data sources from mxODBC?
References: <37818772.1C27503E@compaq.com>
Message-ID: <37823A8D.526B6DDC@lemburg.com>

Greg Ewing wrote:
> 
> Is there a way to get a list of available
> data sources from mxODBC? I can't find any
> mention of this in the supplied docs.

No, the API for this are not yet included in mxODBC. They may appear
in a future version though... if I find time to hack them up (or sell
a few licenses ;).

The needed APIs are:

SQLBrowseConnect, SQLDrivers and SQLDataSources

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   178 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From fredrik at pythonware.com  Fri Jul  2 04:17:30 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 2 Jul 1999 08:17:30 GMT
Subject: TIFF-images
References: <7lhroa$62p$1@nnrp1.deja.com>
Message-ID: <012201bec463$55824640$f29b12c2@secret.pythonware.com>

From: <olevi at my-deja.com>
> Is there any way to present TIFF-images through tkinter? After what I've
> read, only GIF, PPM and PGM are supported, but maybe there's a way
> around?

http://www.pythonware.com/products/pil/
(software distribution)

http://www.pythonware.com/library/pil/handbook/index.htm
http://www.pythonware.com/library/pil/handbook/imagetk.htm
(docs)

</F>




From greg.ewing at compaq.com  Thu Jul  1 17:34:48 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 02 Jul 1999 09:34:48 +1200
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <377a6bb0.10820509@news.oh.verio.com>
Message-ID: <377BDEF7.C34A6088@compaq.com>

Kevin Dahlhausen wrote:

> Lyle Johnson <jlj at cfdrc.com> wrote:

> >Just to add fuel to the fire, we now have a Python interface to FOX, a
> >cross-platform GUI toolkit:

> >    http://www.cfdrc.com/pub/FOX/fox.html

This link doesn't work for me. Has
it been moved somewhere else?

Greg



From janssen at parc.xerox.com  Mon Jul 26 20:05:21 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Tue, 27 Jul 1999 00:05:21 GMT
Subject: Python and Red Hat Linux 6
In-Reply-To: <379CD547.22A10A50@postoffice.pacbell.net>
References: <379CD547.22A10A50@postoffice.pacbell.net>
Message-ID: <srbDT1kB0KGW1ZV7w0@holmes.parc.xerox.com>

Yes, Linux supports #!.  Sounds like /usr/bin/python doesn't exist or
perhaps is not executable.

Bill



From kas at maps.magnetic-ink.dk  Fri Jul 23 06:01:37 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Fri, 23 Jul 1999 12:01:37 +0200
Subject: "libmysqlclient.so.6" not found. (what does /sbin/ldconfig /usr/local/lib/mysql do ?)
References: <7n9c7f$ihd$1@nnrp1.deja.com>
Message-ID: <slrn7pgfc1.pj.kas@titan.magnetic-ink.dk>

sweeting at neuronet.com.my <sweeting at neuronet.com.my> wrote:

> When building Python from source, and with the MySQLmodule in the
> Modules directory, I got the following error :
> 
> Shared object "libmysqlclient.so.6" not found
> *** Error code 1
> 
> But libmysqlclient.so.6 does indeed exist in /usr/local/lib/mysql.

Put the directory in your /etc/ld.so.conf, then run ldconfig without any
arguments and you're all set.


Cheers,

  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From vaton at postoffice.pacbell.net  Mon Jul 26 16:53:20 1999
From: vaton at postoffice.pacbell.net (Rob)
Date: Mon, 26 Jul 1999 13:53:20 -0700
Subject: Python and Red Hat Linux 6.0
Message-ID: <379CCAC0.AD5D1C2F@postoffice.pacbell.net>

When programming in Perl or Python, I am using a header like the
following on an executable file to cause the shell (bash or csh) to run
the scripts as Perl/Python files:
#!/usr/bin/perl   for Perl   and #!/usr/bin/python   for Python. I've
checked and the compilers are located in the appropriate directories. I
keep getting this when I run scripts:
bash: <script name>: command not found. Does Linux support the '#!'
notation in the file header, or do I always have to run the scripts by
explicitly typing in 'perl' or 'python' before the script name?






From Hank.Fay at p98.f112.n480.z2.fidonet.org  Thu Jul  1 08:17:28 1999
From: Hank.Fay at p98.f112.n480.z2.fidonet.org (Hank Fay)
Date: Thu, 01 Jul 1999 13:17:28 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001635@bossar.com.pl>

From: "Hank Fay" <hank at prosysplus.com>

Is it possible to use ActiveX controls in any of these GUI's?  Then you've
got a large choice of controls.

Hank

--
http://www.prosysplus.com
Al Christians <achrist at easystreet.com> wrote in message
news:377BC620.F5553FF5 at easystreet.com...
> I looked at Fox and put it aside because I need a grid control.
> wxPython seems to be the winner for Windows in that category, even
> though the grid is pretty minimal.
>
> Any other candidates with a grid?
>
> Al
>
>
>
>
> Lyle Johnson wrote:
> >
> > I've already worked up the binary distribution of FXPy for Windows; I
just
> > forgot to mention it in the previous post. The binary installer can be
found
> > at:
> >
> >     ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42-win32.exe
> >
> > I haven't worked up a binary distribution of the (C++) FOX distribution
> > itself, primarily because it changes most every day :) But that is in
the
> > plans as well.
> >
> > Thanks,
> >
> > Lyle
> >
> > Mikael Lyngvig wrote:
> >
> > > On Wed, 30 Jun 1999 09:07:51 -0500, Lyle Johnson <jlj at cfdrc.com>
> > > wrote:
> > >
> > > >Just to add fuel to the fire, we now have a Python interface to FOX,
a
> > > >cross-platform GUI toolkit:
> > > >
> > > >    http://www.cfdrc.com/pub/FOX/fox.html
> > > >
> > > >The latest version of the Python wrappers can be picked up at:
> > > >
> > > >    ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.42.tar.gz
> > > >
> > > >I don't know how the speed of FXPy would compare with some of the
others
> > > >mentioned (Tk, wxWindows, Qt, GTK, ...) but it would be worthwhile
for
> > > >some independent opinions.
> > >
> > > I'm downloading it now.  The http URL is wrong, it should be
> > > http://www.cfdrc.com/FOX/fox.html.
> > >
> > > The presentation seems promising, but I'd like to suggest a binary
> > > distribution for PC/Intel computers.  I know this is not common for
> > > Unix distributions, but it's the norm for PCs.
> > >
> > > -- Mikael




From wolford at enews.nrl.navy.mil  Thu Jul  8 12:44:49 1999
From: wolford at enews.nrl.navy.mil (Scott Wolford)
Date: Thu, 08 Jul 1999 12:44:49 -0400
Subject: Two Pythons talking to each other?
References: <199907080019.TAA22244@buffalo.fnal.gov> <199907080529.HAA19502@axil.hvision.nl>
Message-ID: <3784D581.D35B453C@enews.nrl.navy.mil>

How about preceding the string with the length of the string, like if you were
reading and writing from a binary file. Or possibly a NUL character (unless you
plan to embed NUL's in your string.

Scott

Hans Nowak wrote:

> I just tested this, and it works. But now I have another problem &
> question (which may be stupid, but I know hardly anything about
> networking and sockets). My programs send commands to each other in
> IRC style (like "/say this") or just plain text (chat). For receiving
> these strings, I use a non-blocking socket connection which polls the
> server every 0.x seconds (the period may vary). However, when the
> strings are sent too fast, which is the case with the 127.0.0.1
> address, and sometimes with regular connections too, the other side
> will pick two strings up as one string.
>
> So my question is, how does one generally deal with this? Should I
> attach a newline (\n) (or maybe another separator character) after
> every string? Or are there other ways to guarantee that two commands
> will arrive at the other side as two strings?
>
> TIA,
>
> --Hans Nowak (ivnowa at hvision.nl)
> Homepage: http://fly.to/zephyrfalcon




From MHammond at skippinet.com.au  Mon Jul 26 20:30:44 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Tue, 27 Jul 1999 10:30:44 +1000
Subject: NT admin modules?
References: <379C0402.9DA8259D@hk.rfa.org> <7nheqm$81n$1@salome.rfa.org>
Message-ID: <7niufc$rb8$1@m2.c2.telstra-mm.net.au>

Bill Eldridge wrote in message <7nheqm$81n$1 at salome.rfa.org>...
>
>I was thumbing through a Win32 Perl programming book

Soon you will be able to thumb through a Win32 Python programming book, and
we cover all of this :-)

> and found a lot of admin modules available
>for the NT world - is there a similar assortment
>for Python, or is someone working on porting these?

There are a huge number of these functions available in the latest (build
125) win32net module - functions for creating users, shares, groups,
changing server info etc.  We infact go better than Perl for these
functions, as Perl limits you to pre-defined "information levels" which do
not cover the complete information about these objects.  Python supports all
information levels supported by the Windows APIs.  However, build 125 is
missing a few functions that Perl does cover - my latest source tree does
now cover all of these (and more) so I could mail you a later version.

Also, depending on the requirements, win32security, win32evtlog, win32pdh
(perf mon) and win32service modules all provide administrative type
functions  (eg, controlling/querying services, reading event log or perf mon
info, changing security on objects, etc)

Mark.






From iceland at green.com  Fri Jul 30 16:36:27 1999
From: iceland at green.com (iceland at green.com)
Date: 30 Jul 1999 13:36:27 PDT
Subject: ADULTS ONLY 59749
Message-ID: <30079913.3100@green.com>

Adults Only:

http://207.240.225.250

-Free XXX


m4Wxn\9HIF



From neelk at brick.cswv.com  Thu Jul 15 20:27:03 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 15 Jul 1999 19:27:03 -0500
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim>
Message-ID: <7mlu8n$6ug$1@brick.cswv.com>

In article <000601bece8f$b549a420$51a22299 at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>[Neel Krishnaswami]
>
>> (I'd love to be able to define coroutines in Python like so:
>>
>> def evens(z):
>>     for elt in z:
>>         if z % 2 == 0:
>>             suspend(elt)
>
>
>> It would probably require some rethinking of Python's iteration
>> protocol though.)
>
>How do you intend to resume it?

I left this unspecified because I don't know the answer. :) I figured
that it could be rolled up into the iteration protocol and I could
use it without having to completely understand how it all works. 

[The iteration protocol]

>That's not a hangup; Guido already knows how to do it cleanly.  

Good -- now I don't have to worry at all. 

>BTW, generators are much easier to implement than coroutines -- the former
>don't require threads or stacklessness or continuations under the covers,
>just some relatively straightforward changes to Python's current
>implementation (Steven Majewski noted this 5 years ago).

What's the difference, exactly? AFAICT you need to save the execution
state when suspending both coroutines and generators, but I might be
missing something obvious....


Neel



From tismer at appliedbiometrics.com  Thu Jul 22 07:56:46 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Thu, 22 Jul 1999 11:56:46 GMT
Subject: int(long(-sys.maxint-1)) fails on Linux
References: <3796DC4F.224795DA@vision.arc.nasa.gov>
Message-ID: <379706FE.6C5351F@appliedbiometrics.com>


Chad Netzer wrote:
> 
> Compiling python 1.5.2 on Linux Mandrake 6.0,  the "make test" step fails
> because it gets an OverflowError during test_types.py on the following line:
> 
> int(long(-sys.maxint-1))

This went wrong with 1.5 and 1.5.1, and it should be gone
for 1.5.2 .

Have a look at the (Tim Peters) implementation.

Especially at the end:
	if ((long)x < 0 && (sign > 0 || (x << 1) != 0))
		goto overflow;
	return (long)x * sign;

where x is unsigned long.
This cannot get wrong, unless you have one of three problems:

1) The mandrake compiler is buggy
2) You don't have 1.5.2 but 1.5.1
3) Oliver Andrich made a mistake with the package

Your comparison to 1.5.1 is invalid since that definately had a bug.

If you can deduce that case 1) must be true, then I'd like
to send you a test program which I just wrote, to simulate
the above without Python. I'd like to see the generated
assembly code, then.

If it is 2), then please stand in the corner for costing my time.

For 3), Oliver is guilty, and he deserves so many hugs about
his RPM stuff that this is just something that happens.

ciao - pirx

________________________________
long
PyLong_AsLong(vv)
	PyObject *vv;
{
	/* This version by Tim Peters */
	register PyLongObject *v;
	unsigned long x, prev;
	int i, sign;

	if (vv == NULL || !PyLong_Check(vv)) {
		PyErr_BadInternalCall();
		return -1;
	}
	v = (PyLongObject *)vv;
	i = v->ob_size;
	sign = 1;
	x = 0;
	if (i < 0) {
		sign = -1;
		i = -(i);
	}
	while (--i >= 0) {
		prev = x;
		x = (x << SHIFT) + v->ob_digit[i];
		if ((x >> SHIFT) != prev)
			goto overflow;
	}
	/* Haven't lost any bits, but if the sign bit is set we're in
	 * trouble *unless* this is the min negative number.  So,
	 * trouble iff sign bit set && (positive || some bit set other
	 * than the sign bit).
	 */
	if ((long)x < 0 && (sign > 0 || (x << 1) != 0))
		goto overflow;
	return (long)x * sign;

 overflow:
	PyErr_SetString(PyExc_OverflowError,
			"long int too long to convert");
	return -1;
}
________________________________

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From markusk at bidra241.bbn.hp.com  Fri Jul 23 04:15:11 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 23 Jul 1999 10:15:11 +0200
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have code.)
References: <000601bed335$893916e0$a72d2399@tim> <p5hfmymh9z.fsf@bidra241.bbn.hp.com> <Pine.LNX.4.10.9907211223270.11985-100000@atrus.jesus.cam.ac.uk> <p5n1woncm0.fsf@bidra241.bbn.hp.com> <7n8fnr$2al$1@vvs.superst.iae.nl>
Message-ID: <p5k8rrn5o0.fsf@bidra241.bbn.hp.com>

Sorry my Smalltalk example is wrong. 
I must have coded for to long in this poor language called C ...
I must look like this :


do: aBlock
 a do:[:each | aBlock value: each]
 b do:[:each | aBlock value: each]

more complicated but still very easy to understand. 

Greetings,
Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From mal at lemburg.com  Mon Jul 26 06:23:00 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Mon, 26 Jul 1999 12:23:00 +0200
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi>
Message-ID: <379C3704.A6F8984D@lemburg.com>

Markus Stenberg wrote:
> ...
> Thus I've had great deal of interest in getting the speed up; however,
> speeding up dynamic interpreted languages is nontrivial, especially
> languages as dynamic as Python.

There are various ways to enhance Python's performance. Here
are some links:

	http://www.automatrix.com/~skip/python/
	http://lima.mudlib.org/~rassilon/p2c/

The most trivial trick is to eliminate global lookups by "localizing"
them. Others include turning to C functions (see mxTools on my
Python Pages) or using specialized Python extensions (calls
to builtin functions are noticeably faster than Python function
calls).

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   158 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From nospam at bitbucket.com  Sat Jul 24 00:01:31 1999
From: nospam at bitbucket.com (Phil Mayes)
Date: Fri, 23 Jul 1999 21:01:31 -0700
Subject: Converting itegers to strings
References: <19990723171921.A5205@pooh.frostnet.net>
Message-ID: <NUam3.9$8K2.1478@newsfeed.avtel.net>

Chris Frost wrote in message <19990723171921.A5205 at pooh.frostnet.net>...

>How can you convert numbers (an interger in this case) to strings? (i.e. I
>have some_number = 3, and would like to use it as a string)


Yeah, there's no itoa() -- use the printf equivalent:
>>> print "%d" % some_number
3

Multiple numbers are supplied as a tuple:
>>> h = 7
>>> m = 45
>>> print "%02d:%02d" % (h,m)
07:45
--
Phil Mayes    pmayes AT olivebr DOT com






From tbryan at arlut.utexas.edu  Tue Jul 27 06:07:28 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Tue, 27 Jul 1999 05:07:28 -0500
Subject: Python and Red Hat Linux 6
References: <379CD547.22A10A50@postoffice.pacbell.net> <379CD5C2.5BA096A2@mail.danbbs.dk> <379d7fcf.52695@news.tiscalinet.it>
Message-ID: <379D84E0.5F872B@arlut.utexas.edu>

Kranio wrote:

> I have the same problem and this solution won't works. If you type
> 
> $ file.py
> bash: ./file.py : Command not found

file.py is not in your PATH and isn't found.

> $ ./file.py
> 
> but the output is
> 
> bash: file.py : Permission denied

Now you've specified the path, and file.py is found, but you 
haven't made the file executable.

chmod u+x file.py
./file.py
 
> Any good solution?

man chmod
:)

-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From kistler at fnmail.com  Thu Jul  8 18:38:31 1999
From: kistler at fnmail.com (Per Kistler)
Date: Fri, 09 Jul 1999 00:38:31 +0200
Subject: Two Pythons talking to each other?
References: <14212.51977.601241.772478@buffalo.fnal.gov> <Pine.SGI.3.95.990708120053.9140Q-100000@hppc.fnal.gov> <roy-0807991340410001@qwerky.med.nyu.edu>
Message-ID: <37852867.151E9F26@fnmail.com>

Hi Roy

I'm not the original questioner, but I would like to transfer
pictures from a background C++ program to a python program,
which would display it in the GUI. The background program could
run locally or remotely, so I thought, it would be nice, to have
a method, which codes nearly identical, but is applicable for
remote and local usage. On the other hand, it would be much
faster, to use something like shared memory for the local case,
but it makes me code much more. But it's still a difference,
whether I get 1.5 more picts through the pipe....

Regards, Per.

Roy Smith wrote:

> But, in both cases, real time was 10x the CPU time, indicating that I/O
> speed totally swamped everything else.

> Roy Smith <roy at popmail.med.nyu.edu>
> New York University School of Medicine

-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net
------------------------------------------------------------



From svoynow at yahoo.com  Wed Jul 21 16:17:33 1999
From: svoynow at yahoo.com (Sasha Voynow)
Date: Wed, 21 Jul 1999 13:17:33 -0700
Subject: printing with Python/Windows
References: <FP6l3.3374$K3.72023@nuq-read.news.verio.net> <37952079.44083903@news.gte.net> <932533928.18948@www.remarq.com> <379620b8.10675324@news.gte.net>
Message-ID: <8Vpl3.3477$K3.77369@nuq-read.news.verio.net>

 Roger:

    Wow. Excellent.
Another example of the excellence of comp.lang.python and it's
contributors.

Sasha


I'll have to take a look at this script and see if I can
> >modify it for
> >"unattended" execution, as that is what I'm really after.
> >
>
> Okay, done.  See my web site:
>
> http://starship.python.net/crew/roger/
>
> for DisplayImage.py.  From the doc:
>
>
> Image Display and Printing demo.
>
> Demonstrates a number of techniques for displaying and printing images
using
> the PythonWin MFC framework.  This method of printing is suitable for
> "automatic" printing, and so, uses the default printer settings.
>
> As this was ripped out of a larger app, I left some other stuff in as
well:
>     *) The pixel value under the cursor is displayed in the title bar.
>     *) You can draw up to four circles on the image using the mouse.
>     *) After four circles are drawn, you can drag them to new locations.
>     *) You can delete a circle with the delete or backspace key.
>     *) Right click on the image to get an options menu (print only for
now).
>
> Usage:
>     from the windows command line:
>         PythonWin.exe /app path/DisplayImage.py [-p -m mag] [imagePath]
>         where:
>             path:      The directory containing this script.
>             -p:        If present, print the image and exit.
>             -m mag:    Only used if -p present, the magnification to print
at,
>                        defaults to 3.
>             imagePath: The image to open and display.
>         You may need to supply the PythonWin.exe path as well.
>         If you use the -p option, you must supply the imagePath as well.
>         If no imagePath is given, an open file dialog is used to get the
image.
>
>     from the Python/PythonWin command prompt (or, within a script):
>         import DisplayImage
>         DisplayImage.displayImage() # Use file dialog to get and display
an
>                                     # image
>         DisplayImage.displayImage(imgPath) # display given image
>         DisplayImage.printImage(imgPath, mag) # print image at mag.
>
> Requires:  PIL (Python Imaging Library)
>            PythonWin (Build 125 or later)
>
>
>
> Cheers,
>
>
> Roger Burnham
> Cambridge Research & Instrumentation
> rburnham at cri-inc.com
> http://www.cri-inc.com/
> http://starship.python.net/crew/roger/
> PGP Key: http://www.nai.com/default_pgp.asp
> PGP Fingerprint: 5372 729A 9557 5F36 177F  084A 6C64 BE27 0BC4 CF2D





From granadam at mail.datanet.hu  Sat Jul 10 18:11:04 1999
From: granadam at mail.datanet.hu (dzsi)
Date: Sun, 11 Jul 1999 00:11:04 +0200
Subject: database
Message-ID: <7m9tnu$6qg$73@rks1.urz.tu-dresden.de>

hi,

i am new to python.......

i have a database (comma separated text), and i would like to get some data
out of it based on some query. let's say there are seven fields, f1...f7,
and i am searching for entries in f1, and want to display f1..f7  for the
field that matches the query. how do i go about doing that?

also, can i access db databases, or any other common pc type databases with
perl scripts only on a unix system? without any server configuration,
assuming python is installed. so far, the comma separated text file is the
only way i could figure out.

thanks,

adam.







From delgado at Mathematik.Uni-Bielefeld.DE  Fri Jul  9 09:31:54 1999
From: delgado at Mathematik.Uni-Bielefeld.DE (Olaf Delgado)
Date: Fri, 9 Jul 1999 15:31:54 +0200
Subject: Binary pickle portability
In-Reply-To: <Pine.WNT.4.04.9907071517320.248-100000@rigoletto.ski.org>
References: <Pine.WNT.4.04.9907071517320.248-100000@rigoletto.ski.org>
Message-ID: <7m9tm6$6qg$20@rks1.urz.tu-dresden.de>

On Wed, 7 Jul 1999, David Ascher wrote:

> There is a bug in pickling of NumPy arrays, which I expect will be fixed
> in the next release.  Binary pickles works across byte-order and byte-size
> platforms (when it's bug-free =).

Thank you for the information! It's a bit painful to have to convert the
arrays back and forth right now, but good to know that it's not my fault.

I am not sure I understand what you mean by byte-order and byte-size
platforms. What's a counterexample?

Olaf




From MHammond at skippinet.com.au  Thu Jul  8 19:03:03 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 9 Jul 1999 09:03:03 +1000
Subject: win32 python as a dll?
References: <1280719925-72790231@hypernet.com> <37850F5B.F80CB0A7@enews.nrl.navy.mil>
Message-ID: <7m3ajc$s3q$1@m2.c2.telstra-mm.net.au>

Scott Wolford wrote in message <37850F5B.F80CB0A7 at enews.nrl.navy.mil>...

>msvcrt.lib and the application can't then bring in msvcrtd.lib. I do think
>this is a problem because python15_d.lib is provided but no python15_d.dll
>is. I would think you'd provide (at the very least) python.lib linked with
>msvcrt.lib, and python_d.lib linked with msvcrtd.lib (and the dlls), or
>(to be really nice :) all four combinations.

The main reason we dont is size - we would need to ship the _d, and for
people to really use it, also the .pdb.  These add too many megs.

Further, almost _everyone_ who needs to use this _d version already has MSVC
and already has the Python sources.  So building there own is trivial.

Further, "all 4 combinations" doesnt work.  As you found out painfully, is
it _critical_ that all Python systems share the same C runtime library.
This is the whole point of the _d fiasco - to prevent the mixing and
matching.

We obviously need to document this process better - Im not sure where tho -
what documentation did you use to get this far?

It is also worth noting that Gordon pointed this out to you before this
particular thread, and you still had the problem.  Im not sure that
documenting it would be effective in all cases.


> Don't get me wrong; I think
>it's a great service to provide the pre-compiled package. I just want to

Well, FWIW, I will be making these _d versions (and all the win32 extensions
_d versions) available in a seperate .zip file as a Registered User benefit
(http://starship.python.net/crew/mhammond/registration).  I havent started
this particular benefit yet tho.

Mark.






From gmcm at hypernet.com  Fri Jul 23 18:13:54 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 23 Jul 1999 17:13:54 -0500
Subject: Python documentation available as ZIP files
In-Reply-To: <7nal1u$6cb$1@tribune.oar.net>
Message-ID: <1279358265-27510415@hypernet.com>

David Robinow wrote: 
> Fred L. Drake wrote in message
> <14232.40613.53508.888049 at weyr.cnri.reston.va.us>...
> >
> >  I've received enough hostile mail from Windows users not
> >understanding how to unpack .tgz archives (er, WinZip works for me...)
> >that I will be making most of the documentation packages available as
> >ZIP archives as well as .tgz archives.
>  This is really sad.  You're not really doing Windows users any
>  favors by
> doing this.

No, he's doing himself a favor. So far, his only mistake is in 
announcing this.

Fred-you-worthless-slut-<wink/counterwink>-ly y'rs

- Gordon



From tjreedy at udel.edu  Sun Jul  4 23:08:26 1999
From: tjreedy at udel.edu (Terry Reedy)
Date: 5 Jul 1999 03:08:26 GMT
Subject: Why have 300 articles just appeared in c.l.py?
References: <m33dz4oxxx.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <7lp7ja$aih$1@news.udel.edu>

In article <m33dz4oxxx.fsf at atrus.jesus.cam.ac.uk>, mwh21 at cam.ac.uk 
says...
>
>I just had the slightly alarming experience of about 300 articles, all
>of which I had read, appearing unthreaded in this newsgroup.
>
>Did this happen to everyone, or is it some local snafu?

Same thing here.  TJR




From jim at interet.com  Mon Jul 19 11:32:55 1999
From: jim at interet.com (James C. Ahlstrom)
Date: Mon, 19 Jul 1999 11:32:55 -0400
Subject: importing a Win32 DLL module w/resources
References: <3790E73B.C969A223@psych.ucsb.edu>
Message-ID: <37934527.8C9273BB@interet.com>

Andy Beall wrote:
> 
> Hi, I'm trying to extend Pythonwin from MSVC with a DLL module that I
> can import, and everything is working except with I try and access some
> resources, such as a dialog box that I made.  It seems that when the DLL
> is built in MSVC, it doesn't bundle the .rc or .res file into the DLL.
> Anybody know a way to force the linker to bind this or how to let
> Pythonwin know where these resources are?

First make sure the resources are in the DLL.  The file myapp.rc
should be in the list of files, and there should be a tab for
"Resource View" in MSVC next to the file view.

I don't think DLL resource are searched by default (???).  To
use DLL resources instead of Main resources, you call
FindResource() followed by AfxSetResourceHandle() in MFC.

Jim Ahlstrom



From jlj at cfdrc.com  Fri Jul 16 09:53:55 1999
From: jlj at cfdrc.com (Lyle Johnson)
Date: Fri, 16 Jul 1999 08:53:55 -0500
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net>
Message-ID: <378F3973.1AA591D5@cfdrc.com>

You can avoid ever seeing the DOS prompt if you also download the PythonWin
stuff (see the links at http://www.python.org/windows). This includes a very
nice little integrated development environment (IDE) with a built-in Python
shell, debugger, etc.

mary jeanette wrote:

> I decided to learn python because I heard it was a very easy language to
> learn. So I downloaded it and bought a couple of very expensive
> paperback books from o'reilly only to discover that they are written for
> unix. In other words, typing in the examples from the book into the
> interpretor only yields various syntax errors. I have no idea what I am
> supposed to change in order to make the examples work, and apparently
> there is nowhere to turn in order to find out. Q: Am I supposed to know
> DOS as a prerequisite? The idea of having to learn DOS is appalling.
> Should I try partitioning my hard drive and installing Linux or freeBSD?
> Perhaps I should learn Perl instead, I have seen books on Perl written
> for Win32 systems.   Thanks for your advice.




From sjoshi at ingr.com  Thu Jul  8 09:22:24 1999
From: sjoshi at ingr.com (sjoshi at ingr.com)
Date: Thu, 08 Jul 1999 13:22:24 GMT
Subject: Getting/Swaping Drive Letters ??
Message-ID: <7m28m4$auk$1@nnrp1.deja.com>

Hello All
   Just wondering if there's a way to get the current drive letters in
a local machine and be able to swap a drive letter, with an available
one (like U do in NT-Disk Admin)

thanks
Sunit


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Christian.Lemer at usrconsult.be  Tue Jul 27 09:00:21 1999
From: Christian.Lemer at usrconsult.be (Christian Lemer)
Date: Tue, 27 Jul 1999 15:00:21 +0200
Subject: Python classes for Excel Reports?
Message-ID: <379DAD65.5FDB5B59@usrconsult.be>

I have to set up a system to generate some Reports in Excel.

I would like to make it in python, but before writing myself a lot of
classes, I wonder if there is already something available in the
field...

-- 
Christian LEMER                                    tel +32.10.65.44.11
UsrConsult S.P.R.L                                 fax +32.10.65.44.10
rue Margot, 37	                              http://www.usrconsult.be
B-1457 Nil St Vincent             mailto:Christian.Lemer at usrconsult.be



From fredrik at pythonware.com  Wed Jul 21 12:37:28 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Jul 1999 16:37:28 GMT
Subject: ImportError: Unresolved symbol
References: <199907210617.BAA26862@pheno.physics.wisc.edu>
Message-ID: <003b01bed397$553917c0$f29b12c2@secret.pythonware.com>

Nikolas Kauer <kauer at pheno.physics.wisc.edu> wrote:
> When I run a Python CGI script (first line #!/patjh/bin/python), I get the
> following error in the Apache error log:
>
> [Tue Jul 20 16:58:55 1999] [error] [client ...] Premature end of script headers:
/usr/share/test/apache_1.3.6/cgi-bin/db_
> driver.cgi
> Traceback (innermost last):
>   File "/usr/share/test/apache_1.3.6/cgi-bin/db_driver.cgi", line 24, in ?
>     from cgi_namespace import *
>   File "/usr/share/test/python/lib/python1.5/lib-dynload/cgi_namespace.py", line 8, in ?
>     from cgi_state import State
>   File "/usr/share/test/python/lib/python1.5/lib-dynload/cgi_state.py", line 1, in ?
>     import types, DateTime, string
>   File "/usr/share/test/python/lib/python1.5/lib-dynload/DateTime.py", line 8, in ?
>     from mxDateTime import *
> ImportError: Unresolved symbol in /usr/share/test/python/lib/python1.5/lib-dynload/mxDateTime.so:
timegm
>
> When I start an interactive Python session, and do "import DateTime", or execute the Python script
on
> the command line (like tcsh> ./db_driver.cgi) no error messages appear, everything seems to work
fine.
> The Web server runs on Digital Unix 4.0E.
>
> The only solution I can currently think of would be to try to statically link mxDateTime.so ...
>
> Any hints, anybody?

looks like mxDateTime requires library function
named "timegm".  afaik, there's no such function
in the Digital Unix standard libraries (there's an
"gmtime" in there, but no "timegm"...)

check the mxDateTime build instructions (there's
gotta be some way to do without that function),
and rebuild.

</F>




From aulenbac at ucar.edu  Fri Jul 23 15:02:13 1999
From: aulenbac at ucar.edu (Steve Aulenbach)
Date: Fri, 23 Jul 1999 13:02:13 -0600
Subject: Python interface to IBM Data Explorer?? (DX/OpenDX)
References: <3798841F.B9ED6E95@brown.edu>
Message-ID: <7nae7m$aru$1@ncar.ucar.edu>

Have you looked at VTK, the Visualization Toolkit? It already has Python
bindings ( as well as TCL, Java, and not to mention a great open source C++
library, several textbooks and an active community). It is at
http://www.kitware.com/vtk.html.

Steve Aulenbach

Daniel Faken <Daniel_Faken at brown.edu> wrote in message
news:3798841F.B9ED6E95 at brown.edu...
> Hello everyone,
>
>   Greetings from Rhode Island.
>
>   Has anyone developed a python interface (via SWIG or something else)
> for the Data Explorer visualization system?
>   IBM recently made it "OpenSource", and it seems very useful to me.
>   There's a web site at http://www.opendx.org
>
> thanks for any help,
> Daniel Faken
>
> --
> Daniel B. Faken
> Daniel_Faken at brown.edu
> http://www.chem.brown.edu/brochure/people/grads/faken/





From tbryan at arlut.utexas.edu  Sat Jul  3 10:56:20 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Sat, 03 Jul 1999 09:56:20 -0500
Subject: Dumb beginner question; what is IDLE?
References: <377D8BFB.3700379@afn.org>
Message-ID: <377E2494.DD6DD018@arlut.utexas.edu>

Andy L Krietemeyer wrote:
> 
> I've seen several references to IDLE.  Is this an IDE for python? Or
> what is it?  Thanks,  Andy.

>From http://www.python.org/1.5/NEWS-152.txt
"""
I've added IDLE: my Integrated DeveLopment Environment for Python.
Requires Tcl/Tk (and Tkinter).  Works on Windows and Unix (and should
work on Macintosh, but I haven't been able to test it there; it does
depend on new features in 1.5.2....This is very much a work in
progress.  I'd like to hear how people like it compared to PTUI (or
any other IDE they are familiar with).
"""

-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From Alex.Maranda at p98.f112.n480.z2.fidonet.org  Fri Jul  2 08:48:36 1999
From: Alex.Maranda at p98.f112.n480.z2.fidonet.org (Alex Maranda)
Date: Fri, 02 Jul 1999 13:48:36 +0100
Subject: Python suitable for a game engine?
Message-ID: <00001664@bossar.com.pl>

From: Alex Maranda <amaranda at nospam.com>

Alan Daniels wrote:
[snip]
> will be involved. The only justification I can see is better CPU
> utilization on an SMP machine, and I don't happen to have one of those
> lying around. Hence the engine will run in a single thread.
This justification doesn't fly on SMP machines precisely because of the
global interpreter lock. This was debated to death in previous threads.
All we can do is wait till the interpreter is fully re-entrant and the
need for the global lock goes away...
BTW (completely off topic), I wonder what's the current thinking on this
- the Linux kernel went for finer-grained locks.

> I have the (VERY!) preliminary code at:
>     http://www.mindspring.com/~daniels/engine/index.html
> if you're interested, although there isn't much to look at.
I'm downloading it now. You can use negative version numbers;
SmallEiffel is now at version -0.78. They'll stop when they'll reach 0.0
:-)

Cheers,
-- 
Alex Maranda         mailto: amaranda at spider dot com
Spider Software Ltd. Tel: +44 (0)131 4757036
Edinburgh, UK        http://members.xoom.com/Alex_Maranda
STREAMS based communications protocols for embedded systems



From aahz at netcom.com  Thu Jul  8 13:44:43 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 8 Jul 1999 17:44:43 GMT
Subject: Python and Paradox
References: <7m2mrj$hes$1@nnrp1.deja.com>
Message-ID: <7m2o2b$aq0@dfw-ixnews12.ix.netcom.com>

In article <7m2mrj$hes$1 at nnrp1.deja.com>,
Kaine  <kaine_the_bloody at my-deja.com> wrote:
>
>I am looking for a way to access Paradox tables from Python without
>using ODBC. Does anyone have any ideas?

Nope, from an ex-Paradox tech support engineer.

Oh, sure, you could play tricks such as scripting Paradox through COM to
export the tables, but if you want direct table access ... what's wrong
with ODBC?
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From anl at completebbs.com  Thu Jul 22 16:54:26 1999
From: anl at completebbs.com (anl at completebbs.com)
Date: 22 Jul 1999 20:54:26 GMT
Subject: tiles-rus  4135
Message-ID: <7n80e2$ee4$11849@news.fsr.net>

tiles-r-us.com #1 source for ceramic tile, Floor & Wall Ceramic tile, mosaic ceramic tile, tools fo the professioanl & the do it yourself, do it yourself guide, for a complete guide how to install ceramic tile. and much more
ycdfxwyetllrriuskwcbfqwsiwjxzotvzkvdekqfoluencq




From bruce_dodson at bigfoot.com  Wed Jul 28 23:49:56 1999
From: bruce_dodson at bigfoot.com (Bruce Dodson)
Date: Thu, 29 Jul 1999 00:49:56 -0300
Subject: global myfunc; def myfunc():
Message-ID: <ffQn3.40339$jl.29112404@newscontent-01.sprint.ca>

I have noticed that, when using python as an embedded language in another
program, using a temporary dictionary for a private namespace, and using
__main__'s dictionary for my global namespace, I can say:

global myfunc
def myfunc():
  ...

to bind myfunc into the global namespace, which allows it to persist across
calls.  But I want to make sure that is "intended behavior" and not just
some implementation quirk, so I can decide whether to document that as a
feature.

Any advice?

Thanx,
Bruce




From sreekantk at freenet.co.uk  Thu Jul  8 23:26:41 1999
From: sreekantk at freenet.co.uk (Sreekant Kodela)
Date: Fri, 09 Jul 1999 04:26:41 +0100
Subject: Efficient string manipulation , Help please
Message-ID: <37856BF1.BD046003@freenet.co.uk>

Dear all

I seem to be in trouble here. I am trying to make some cgi scripts for
my web site. I am using the postgresql as the backend db and a form
based address book and news forum etc. The problem is everything is
working ok as long as noone uses ' and " in the text. If they do, bingo
, postgres refuses to accept. I tried using string library [mainly
string.join(string.split('awkward'"'st'ring')). It works sometimes and
doesn't some times.

Any better and efficient way to deal with this situation !

Any suggestions and ideas will be greatly appreciated.

Thanks
sreekant






From nkauer at students.wisc.edu  Fri Jul  2 11:53:43 1999
From: nkauer at students.wisc.edu (Nikolas Kauer)
Date: Fri, 2 Jul 1999 10:53:43 -0500
Subject: execfile(), import and namespaces
Message-ID: <000201bec4a3$0f909700$c7326880@goldeneye>

We use a modification of "Listing One" of Andrew Kuchling's "A CGI Framework
in Python"
article in Web Techniques (February 1998) for our dynamic Web site.  We just
made our
Web site public and wanted to switch from files with Python source (under
/htdocs) to compiled files with Python object code (i.e. python -c 'import
file.py' -> file.pyc) as suggested in CGI FAQ 1.8, mainly to make it harder
to (download and) re-use our code.  The object files would be loaded as
modules with 'import'.

Problem: In the following code snipet environment variables are passed to
the
executed Python file via the corresponding global namespace argument of the
execfile() function; whereas the 'import' statement does not allow for such
arguments.

namespace = {
             'headers': headers,
             'webvars': webvars,
             'environ': os.environ
            }

[ more code ]

try:
    execfile(filename, namespace)
except:
    # The page's code raised some sort of exception.

Does anyone know of an elegant way to pass namespace objects to modules in a
similar
way execfile() does?

Thanks,
Nikolas

PS I'm not on the mailing list, please reply directly.





From Alex at p98.f112.n480.z2.fidonet.org  Fri Jul  2 14:38:38 1999
From: Alex at p98.f112.n480.z2.fidonet.org (Alex)
Date: Fri, 02 Jul 1999 19:38:38 +0100
Subject: Problem embedding C++ code
Message-ID: <0000169c@bossar.com.pl>

From: Alex <alex at somewhere.round.here>


I had this problem a while back.  It went away when we made everything
that gets included a shared library.  That included the C++ library, I
think...  it was on linux, and I think we had to include libg++.so
instead of libg++.a.  

HTH, and good luck.
Alex.

> CC -c -KPIC link-wrap.c -I/home/mdeshpan/SW/include/python1.5
> "link-wrap.c", line 13: Warning: String literal converted to char* in
> formal argument 1 in call to Py_BuildValue(char*, ...).
> 1 Warning(s) detected.
> CC -G link-wrap.o -lapic -lpthread -o link.so
> daffy:~/Swig$python
> Python 1.5.2 (#3, Jun 29 1999, 11:17:23) [C] on sunos5
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import link
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> ImportError: ld.so.1: python: fatal: relocation error: file
> ./libapic.so: symbol _ex_rethrow_q: referenced symbol not found

This is a gratuitous signature, included so that the news server won't
complain that I've got more included text than new text.  Snappy, huh?



From larsga at ifi.uio.no  Tue Jul 20 07:52:44 1999
From: larsga at ifi.uio.no (Lars Marius Garshol)
Date: 20 Jul 1999 13:52:44 +0200
Subject: python to perl conversion?
References: <slrn7p7q8j.n4.dyoung@baltazar.baltazar.org> 	<3793F935.C4CEFD44@callware.com> <14228.3147.448161.442831@lrz.de>
Message-ID: <wk908ba67n.fsf@ifi.uio.no>

* Eugene Leitl
| 
| What's the consensus for the best (=less painful) way to call a perl
| module from Python? There is this interesting thing called bioperl
| http://bio.perl.org/ , which is both 1) very useful 2) in vigorous
| development, so porting it is out of question.

It's very new, so there's unlikely to be any consensus on it yet, but
Minotaur looks like it might be (or become) what you're looking for.
See <URL: http://mini.net/pub/ts2/minotaur.html >

--Lars M.



From gmcm at hypernet.com  Fri Jul 23 09:26:52 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 23 Jul 1999 08:26:52 -0500
Subject: COM bites - help please
In-Reply-To: <eZUl3.1080$wm4.135734@news.uswest.net>
Message-ID: <1279389887-25608327@hypernet.com>

Karl Putland is frustrated:

> I've got a couple of pyCOM things that are kicking my @$$! 

Bill Tutt got most of 'em...

> 2. A computed attribute in __getattr__ does not want to be accessed
> through com.  It is in the _public_attrs_  list but produces the
> following error when attempting to access it.

Both you and the COM support are playing attribute games, and not 
agreeing on how the game is played. Can you expose Get methods to 
COM,  and play your games there? That way, COM will find something 
real (a method object), and you can still do lazy evalutaion.

BTW, I find it very helpful to use a separate (wrapper) object:

 class RealWork:
   ...

 class COMRealWorkWrap:
   _public_methods_ = [ ... ]
   def __init__(self, real_obj):
     ...

That also means I can put all my Unicode conversion in one place.

- Gordon



From ajung at sz-sb.de  Tue Jul 13 03:24:59 1999
From: ajung at sz-sb.de (Andreas Jung)
Date: Tue, 13 Jul 1999 07:24:59 GMT
Subject: DCOracle not handling Oracle LONG/LONG RAW datatype
In-Reply-To: <378A1678.7A43720E@email.sps.mot.com>; from Larry Gwinn on Mon, Jul 12, 1999 at 09:23:20AM -0700
References: <378A1678.7A43720E@email.sps.mot.com>
Message-ID: <19990713092458.A5137@sz-sb.de>

On Mon, Jul 12, 1999 at 09:23:20AM -0700, Larry Gwinn wrote:
> Hello All,
> 
> I have been evaluating the DCOracle extension from
> http://www.zope.org/Download/DCOracle/  and have run into an issue when
> querying tables with a LONG or LONG RAW data type and I was wondering if
> anybody else has encountered this problem.  If you execute a SQL select
> statement that retrieves from a column that is defined as a LONG or LONG
> RAW datatype, when you perform a fetchone you can only successfully
> retrieve the first row.  The next fetchone will cause a "IndexError: 1"
> to occur.  This exception is being thrown by the Buffer_item function
> located in the Buffer.c file.
> 
I can confirm this behaviour. I always get the following traceback
when I try to access a column of type LONG:


Traceback (innermost last):
  File "./test.py", line 25, in ?
    data = ORA.curs.fetchall()
  File "/ojs/home/ojs/tools/python-1.5.2/lib/python1.5/DCOracle/ociCurs.py", line 351, in fetchall
    v=f(1)
  File "/ojs/home/ojs/tools/python-1.5.2/lib/python1.5/DCOracle/ociCurs.py", line 311, in fetchone
    if ind[ai] < 0: v=MissingValue
IndexError: 1


Testing env: Solaris 2.6, Oracle 7.3.3.6, DCOracle 1.2.1, Python 1.5.2

Cheers,
Andreas 

-- 
                                _\\|//_
                               (' O-O ')
------------------------------ooO-(_)-Ooo--------------------------------------
   Andreas Jung, Saarbr?cker Zeitung Verlag und Druckerei GmbH
   Saarbr?cker Daten-Innovations-Center
   Gutenbergstr. 11-23, D-66103 Saarbr?cken, Germany
   Phone: +49-(0)681-502-1563, Fax: +49-(0)681-502-1509
   Email: ajung at sz-sb.de (PGP key available)
-------------------------------------------------------------------------------



From mgushee at havenrock.com  Sat Jul 24 20:16:55 1999
From: mgushee at havenrock.com (Matt Gushee)
Date: 25 Jul 1999 09:16:55 +0900
Subject: strptime: unpleasant surprise (py 1.5.2 on Linux)
Message-ID: <m3aesltwg8.fsf@havenrock.com>

Hi, Folks--

I'm working on some date-related functions, and wanted to see what
kind of exception 'strptime' would raise if I fed it a European-style
date string with a US format. But instead of an exception, I get:

>>> import time
>>> time.strptime('31/7/99', '%x')  # US locale, wants D/M/Y
Segmentation fault (core import)

>>> dumped DateTime                 # mxDateTime version
>>> DateTime.strptime('31/7/99', '%x')
Segmentation fault (core dumped)

Surely this isn't what we want, is it? Has anyone else noticed this
behavior?

I'm running Python 1.5.2 on RedHat Linux 5.1 (2.0.34 kernel with glibc 
2.0.7).

Matt Gushee
Portland, Maine, USA
mgushee at havenrock.com




From rmasse at cnri.reston.va.us  Thu Jul 22 10:42:22 1999
From: rmasse at cnri.reston.va.us (Roger Masse)
Date: Thu, 22 Jul 1999 10:42:22 -0400 (EDT)
Subject: [Tutor] I want to learn to hack
In-Reply-To: <Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org>
References: <19990721154610.B3663@optiplex.palga.uucp>
	<Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org>
Message-ID: <14231.11726.916541.492153@nobot.cnri.reston.va.us>

David Ascher writes:

 > Gerrit -- just to fill you in -- Roger was presumably either joking or
 > testing a spam filter -- he is a colleague of Guido's, and knows all about
 > hacking.  Whether he knows about cracking or not is not something he's
 > likely to disclose in public. =)
 > 
 > (In other words, Tim was kidding, Roger was kidding, I'm not kidding!).

Errr, it seems I've inadvertantly 'cracked' opened a can of worms.
I really just wanted to know how the Python tutors deal with such
requests, since they are arriving with alarming frequency to
webmaster at python.org.  I'm not quite sure how or why this became a
thread on python-list, but it sure makes me feel tingley all over
when Tim calls me 'little Roger'. <blush, giggle>

hacking-is-what-I-do-for-work-cracking-cyphertext-is-just-plain-fun-ly y'rs

								    -roj



From ovidiu at cup.hp.com  Wed Jul 14 14:14:28 1999
From: ovidiu at cup.hp.com (Ovidiu Predescu)
Date: Wed, 14 Jul 1999 18:14:28 +0000
Subject: Pickling C objects
References: <378B27A9.4201@cs.bham.ac.uk>
Message-ID: <378CD384.236744BE@cup.hp.com>

Stuart Reynolds wrote:

> I'm writing some new Python classes with a C implementation. How do you
> make a C object pickleable?
> 
> Is it possible to use the __repr__ method? It would be easy for me to
> output some Python code as a string which generated the object:
> 
>  'MyClass( MyComponent1(1,2,3), MyOtherComponent2('a','b','c') )'

You need to register your new C type extension with the pickle module.
Take a look in the copy_reg module in the Python distribution to see how
this works.

As an example I have written a new C type entension, WeakReference that
keeps a reference to a Python object without incrementing the reference
count (this is useful for avoiding cycles). Here is the initialization
function of the module:

void initWeakReference ()
{
  char pickling[] = "\
from copy_reg import *\n\
from WeakReference import *\n\
def pickleWeakReference (ref):\n\
    return WeakReference, (ref.__object__,)\n\
w = WeakReference (1)\n\
pickle (type (w), pickleWeakReference, WeakReference)\n";

  Py_InitModule ("WeakReference", WeakReferenceType_methods);
  if (PyErr_Occurred())
    Py_FatalError ("can't initialize module WeakReference");

  if (PyRun_SimpleString (pickling) == -1)
    Py_FatalError ("can't initialize module WeakReference for
pickling");
}

As I mentioned, the only member that I need to pickle is the object,
which is internally kept in a WeakReference as __object__. The
pickleWeakReference function returns a tuple with one element, the
object. If your C type has more members just add them to the tuple.

Hope this helps.

Greetings,
-- 
Ovidiu Predescu <ovidiu at cup.hp.com>
http://andromeda.cup.hp.com/  (inside HP's firewall only)
http://www.geocities.com/SiliconValley/Monitor/7464/



From sjoerd at oratrix.nl  Thu Jul  8 10:41:46 1999
From: sjoerd at oratrix.nl (Sjoerd Mullender)
Date: Thu, 08 Jul 1999 16:41:46 +0200
Subject: Xlib wrapper for Python
In-Reply-To: Your message of Thu, 08 Jul 1999 13:12:24 +0300.
             <Pine.SUN.3.95-heb-2.07.990708130942.17745A-100000@sunset.ma.huji.ac.il> 
References: <Pine.SUN.3.95-heb-2.07.990708130942.17745A-100000@sunset.ma.huji.ac.il> 
Message-ID: <19990708144147.D99EA301CF2@bireme.oratrix.nl>

On Thu, Jul 8 1999 Moshe Zadka wrote:

> Hi. 
> 
> I saw
> http://www.uni-klu.ac.at/unihome/header/tutorial/python--1.4/
> X-extension/Xlib-functions-and-variables.html
> 
> And started wondering where do I get this Xlib wrapper?

http://www.cwi.nl/ftp/sjoerd/

-- Sjoerd Mullender <sjoerd.mullender at oratrix.com>



From danielt3 at gte.net  Thu Jul  8 19:03:01 1999
From: danielt3 at gte.net (Daniel T.)
Date: Thu, 08 Jul 1999 23:03:01 GMT
Subject: My first Tkinter app. Did I do good?
Message-ID: <danielt3-0807991903200001@1cust103.tnt9.clearwater.fl.da.uu.net>

I would love to hear from experienced Python programmers. Does it look
like I know what I'm doing?

# My first Tkinter app

from Tkinter import *

class Window:
    def __init__(self, master):
        self.window = Toplevel(master)
        self.new_button()
        
    def new_button(self):
        button = Button(self.window, text="New Button", command =
self.new_button)
        button.pack(side = TOP)



class MyApp:
    def __init__(self):
        self.root = Tk()
        b = Button(self.root, text = "Quit", command = self.root.quit)
        b.pack()
        b = Button(self.root, text = "New Window", command = self.make_window)
        b.pack(side = BOTTOM)
        self.window_list = []

    def make_window(self):
        n = Window(self.root)
        self.window_list.append(n)
 
    def run(self):
        self.root.mainloop()
       

app = MyApp()
app.run()

-- 
Daniel T.
danielt3 at gte.net
http://home1.gte.net/danielt3/



From matomira at iname.com  Thu Jul  1 10:20:24 1999
From: matomira at iname.com (Fernando Mato Mira)
Date: Thu, 01 Jul 1999 16:20:24 +0200
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net>
			<3773c3a0.102825445@news.oh.verio.com>
			<xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net>
			<xJRc3.292$Of3.67094@news.shore.net>
			<7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com> <377A204C.D843868D@prescod.net>
Message-ID: <377B7928.CA16CD80@iname.com>


Paul Prescod wrote:

> I find it fascinating that there are so many GUI toolkits. The GUI is more
> than twenty years right and we haven't agreed on the basic programming
> model or API yet??

Most [OO] GUI toolkits are yet more of the same primitive stuff. From the
glimpse I got from Qt I think it
goes in the right direction, but I got an idea it's just reinventing stuff in
a way which has already been superseded (something like GARNET's formulas).
The better one is to use multi-way constraints. There're toolkits based on
that (for Eiffel (unreleased) and Smalltalk). What still needs to be thought
is how to merge that with the CLIM presentation model.





From EXCHANGE  Thu Jul  1 01:41:35 1999
From: EXCHANGE (EXCHANGE:RICH2:2C25)
Date: Thu, 01 Jul 1999 06:41:35 +0100
Subject: Name of function within the function?
Message-ID: <00001627@bossar.com.pl>

From: "Dula, Debbie (EXCHANGE:RICH2:2C25)" <violet at americasm01.nt.com>

I've looked and looked, but can't find the answer to this ... I'm
sure it's right in front of me ...


I want the current function's name to be output in my error
messages, but can't figure out how a function can get its own name
(without already knowing it).  For example,

	def f1():
		print "Error in function", f1.__name__

is not really useful for what I need as I might as well print "Error
in function f1".  Is there any generic way I can get the current
function's name?

Thanks,
Debbie



From claird at Starbase.NeoSoft.COM  Fri Jul  2 10:05:03 1999
From: claird at Starbase.NeoSoft.COM (Cameron Laird)
Date: 2 Jul 1999 09:05:03 -0500
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.i <377B454C.6F59@mailserver.hursley.ibm.com>
Message-ID: <7liguf$58o$1@Starbase.NeoSoft.COM>

In article <377B454C.6F59 at mailserver.hursley.ibm.com>,
Paul Duffin  <pduffin at mailserver.hursley.ibm.com> wrote:
>Marco Antoniotti wrote:
			.
		[lots of interesting stuff]
			.
			.
>Tcl does not need a macro system because it has a simple consistent
>syntax / semantics and all commands are treated equally and it is
>possible to change the behaviour of existing commands.
>
>> >
>> > If I needed a macro language in Tcl I can just write one.
>> 
>> I just recieved an email from Cameron Laird, citing a 'procm' form in
>> Tcl, which is supposed to do 'scan time' macros - probably something
>> in line with the real thing. However, the manual pages for 8.1 at Scriptics
>> does not mention it.  Yet, I suppose that it is still an experimental
>> feature that maybe will appear in Tcl in a later edition - a few
>> lustres after Lisp had them :)
>> 
>
>I would say that that is probably something Cameron has created. Despite
>what you think Tcl is not crying out for a Lisp like macro system. What
>it is crying out is for more data types and it will soon be getting them.
>Including a 'proper' [lambda] implementation.
>
>I would say that Tcl (without macro system) can do anything that Lisp
>(with macro system) can do.
			.
			.
			.
I've tweaked follow-ups to move this out of
the way of Pythoneers.

Paul's right.  I apologize for not writing
more clearly before.  I'll summarize:  Marco
wrote, "Here's the kind of thing that Tcl's
[proc] doesn't do as well as LISP's macros
...", I replied, "No problem--we'll just de-
fine [procm] ..." and left the details as an
exercise to the reader.  No, there is NOT a
standard [procm], and I hadn't written one
explicitly at the time I followed up.

That in itself is interesting.  The point I
was trying to make is that Tclers find it
very natural to reason, "Hmm, I don't have
the control structure I want, so I'll quickly
create it."  This is different from, for ex-
ample, Python.  I'll characterize Python's
attitude as more like, "Hmm, Guido doesn't
provide the control structure I want; I must
be using the wrong object pattern for my
algorithm, so it's time to redesign."

I'll summarize:  the real content of my mes-
sage was that Tclers are generally rather
indifferent to reservations about syntax,
because any disagreement quickly is absorbed
into the, "well, if that's what you want,
let's just write a proc that interprets it
that way" state.  LISPish macro processing
isn't present in Tcl, but anyone who wants
it can have it easily enough, so the Tcl
mentality simply doesn't worry much about it.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From tim_one at email.msn.com  Thu Jul 22 03:07:55 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 22 Jul 1999 03:07:55 -0400
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have code.)
In-Reply-To: <Pine.LNX.4.10.9907211223270.11985-100000@atrus.jesus.cam.ac.uk>
Message-ID: <001601bed410$ec2ba0c0$642d2399@tim>

[Michael Hudson]
> ...
> Python is not a functional programming language.

NOW they tell me <wink>.

> As I understand it, you can't really have convenient closures because it
> is not lexically scoped (I can give you inconvenient closures...) and
> there were lengthy discussions about that some months back. I didn't
> follow it too closely because I didn't know what lexically scoped meant
> then, but the upshot was (I think) that lexical scoping isn't really
> the "Python way".

It wasn't, but it may be.  Python's 3-level scoping system (which Guido
routinely calls a 2-level system these days -- I guess because he thinks 2
sounds less contrived than 3 <wink>) was a thoroughly deliberate decision.
Today he'll tell you that he may have overreacted against the rampant abuse
of deep procedure nesting common in large Pascal programs of that long-ago
time.  I tangled with him on this in email at the time, but having also
suffered the Pascal abuses, reluctantly agreed to stop harassing him about
it and just see how the 3-level scheme worked out in practice.

I think it worked out great!  I was surprised.  But there were two miserable
problems:

1) Despite its simplicity, the implementation of name resolution was much
slower than in most languages that support arbitrarily deep lexical nesting.
By far the worst of this got repaired by assigning locals to fixed slots
(and bytecodehacks wouldn't be nearly as much fun without that to abuse
<wink>).

2) Despite that scoping is 3-level, the *syntax* allows nesting functions &
classes to any level.  This is an eternal source of confusion for people
coming from other languages, and can take a long time to stop tripping over.
So why did the syntax allow it?  Because it would have complicated the
grammar to stop it!  Guido may deny that today, but it's the truth <wink>.

Times are different now.  The only people *likely* to abuse lexical nesting
these days are Scheme/Lisp heads, but Python's lack of a macro system will
drive them away anyway <wink>.

As far as the implementation goes, arbitrarily deep lexical closures with
indefinite extent are easy in Python -- it's a small & simple patch
(frames-- activation records --are heap-allocated anyway; all that's really
missing is a "static link" to crawl up the lexical nest).  Guido even
implemented (but never released) it a few years ago.  An unexpected
technical glitch killed it:  it turned out that the natural implementation
created cyclic implementation structures that refcounting isn't strong
enough to clean up.  Presumably Python2 could repair that by hook or by
crook.

So Schemish lexical scoping may yet become The Python Way!  I wouldn't count
on it, though; and maintaining state via class instances will always be more
Pythonic.

[prematurely wise comments elided]

> not-OO-functional-or-fast-but-still-great-ly y'rs
> it's-not-OO-or-functional-it's-Python-ly y'rs

it's-guido's-way-or-the-highway-ly y'rs  - tim





From wtanksle at dolphin.openprojects.net  Sat Jul 31 20:13:47 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Sun, 01 Aug 1999 00:13:47 GMT
Subject: 1.5.2 for: else:
References: <000301bedb13$c38e5860$a32d2399@tim>
Message-ID: <slrn7q749q.bdg.wtanksle@dolphin.openprojects.net>

On Sat, 31 Jul 1999 01:15:54 -0400, Tim Peters wrote:
>[William Tanksley]

>> Oh, and if you REALLY like the current behavior of else-loop, you can
>> still have it if else-loop were gone -- like this:

>> try:
>>  for x in list:
>>    if iffers(x): raise "no!"
>>    yadda(x)
>> except "no!": pass
>> else:
>>  twiddle()

>> That's two extra lines, and one level of nesting -- for the _exact_ same
>> functionality, only much more likely to be comprehended.

>Except that when I see that loop, I'm at a loss to guess whether the
>"except" clause is intended to catch otherwise-uncaught "no!" exceptions
>raised by list.__getitem__, iffers(), and yadda() too.  Presumably the
>intent is that it should not, but there's no easy way to stop it from
>catching unintended stuff too short of defining a unique exception for each
>loop.  for+break+else doesn't have this problem.

Very good point, Tim.

The best solution of all is to simply not use features of the language
which I find objectionable for whatever reason.  I'm just disappointed in
the odd definition of for-else.  I'm glad to have learned it this way
rather than by painful debugging, and I don't think any of us are likely
to forget now :).

Memo to self -- don't use for-else, it doesn't do what it looks like it's
doing.

-- 
-William "Billy" Tanksley



From ralph.gauges at eml.villa-bosch.de  Thu Jul 29 09:27:22 1999
From: ralph.gauges at eml.villa-bosch.de (Ralph Gauges)
Date: Thu, 29 Jul 1999 15:27:22 +0200
Subject: Use of variables form outside the class.
References: <7npi9i$2j12@enews3.newsguy.com>
Message-ID: <37A056BA.7025526@eml.villa-bosch.de>

Peter Torstenson wrote:
> 
> I declare a variable "x" right in the beginning of a module.
> Later I declare a class with function defs.
> I managed to declare varibles inside the class and address them by putting
> "self." in front of the variable.
> How do I reach the "x" variable that is outside any class from inside the
> class functions?
> 
> Peter Thorstenson
> The Newbie

By referencing it via 'classname.x' instead of
`instancename.x' (or 'self.x').
This variable does not belong to an instance of the class
but to the class itself. This also means that you always use
the same variable no matter if you use different instances
of the class. But I guess this is what you want.

Ralph



From fredrik at pythonware.com  Mon Jul 26 12:19:59 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 16:19:59 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com> <al86737ocwo.fsf@sirppi.helsinki.fi> <p5n1wjmmxc.fsf@bidra241.bbn.hp.com>
Message-ID: <01d901bed782$b5c41740$f29b12c2@secret.pythonware.com>

Markus Kohler <markusk at bidra241.bbn.hp.com> wrote:
> My guess is that what calling speed makes slow on Python are all those
> different kinds of ways to call the same function. 
> Examples are default arguments and named arguments. 
> IHMO these features may save some typing but they are not worth
> sacrifying performance nor do they help increasing readability of the
> code. 

really?  back in the 1.2 days, you had to write:

    button = Button(frame, {"text": "QUIT", "fg": "red", "command": frame.quit})
    button.pack({"side": LEFT})

instead of:

    button = Button(frame, text="QUIT", fg="red", command=frame.quit)
    button.pack(side=LEFT)

are you sure you find the first one being easier
to read?

on the other hand, 1.2 was much slower than the
current releases.  go figure...

</F>




From tbryan at arlut.utexas.edu  Wed Jul 28 10:22:39 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Wed, 28 Jul 1999 09:22:39 -0500
Subject: Python and Red Hat Linux 6
References: <379CD547.22A10A50@postoffice.pacbell.net> <379CD5C2.5BA096A2@mail.danbbs.dk> <379d7fcf.52695@news.tiscalinet.it> <379D84E0.5F872B@arlut.utexas.edu> <Pine.LNX.4.10.9907271220420.19094-100000@atrus.jesus.cam.ac.uk>
Message-ID: <379F122F.D89B8FD1@arlut.utexas.edu>

Michael Hudson wrote:
> 
> On Tue, 27 Jul 1999, Thomas Bryan wrote:
> 
> > Kranio wrote:
> >
> > > I have the same problem and this solution won't works. If you type
> > >
> > > $ file.py
> > > bash: ./file.py : Command not found
> >
> > file.py is not in your PATH and isn't found.
> >
> 
> No, no, that would make sense; you get an error like this if file.py is
> found and is executable and has a first line of the form
> 
> #!/path/to/executable
> 
> and /path/to/executable doesn't exist.

I think I may have misread the post and answered somewhat hastily.
The point I was trying to make was:
1. Make sure the file containing the python script is in your PATH
   or supply an absolute path to that file.
2. Make sure the #!/path/to/executable actually points to the correct 
   executable.
3. Make sure the file containing the python script is executable.

On my Red Hat 6.0 system, problems #1 and #2 report an identical 
error message.  I lost track of the thread, but I meant to suggest
that if one has already checked #2, then he should also check #2.

the-gremlins-have-been-modifying-my-PATH-ly yours
Tom

-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From pereira at research.att.com  Tue Jul 20 00:35:53 1999
From: pereira at research.att.com (Fernando Pereira)
Date: Tue, 20 Jul 1999 04:35:53 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com> <170719990020380651%pereira@research.att.com> <7mqaht$r27$2@cronkite.cc.uga.edu> <170719991835331004%pereira@research.att.com> <7mssne$m8r$1@cronkite.cc.uga.edu> <180719992337546352%pereira@research.att.com> <7mvika$cpj$1@cronkite.cc.uga.edu>
Message-ID: <200719990035531711%pereira@research.att.com>

In article <7mvika$cpj$1 at cronkite.cc.uga.edu>, Graham Matthews
<graham at sloth.math.uga.edu> wrote:

> <graham at sloth.math.uga.edu> wrote:
> > expression parsers as primitive parsers". There is no need for the compiler
> > to "solve the halting problem". The programmer is free to use regular
> > expression/automaton techniques where he/she wants. Such parsers take
> > strings and return ASTs. They can then be combined with other parsers
> > as if they were primitive parsers.
> Fernando Pereira (pereira at research.att.com) wrote:
> : One last try. I understood your point from the beginning. However, you
> : did not understood *mine*. If you push the complex
> : pattern-matching/parsing machinery into black boxes, there's not much
> : for the monadic glue to do, is there? 
> 
> What your combinators do depends on your application. If all you are
> doing is parsing using primitive automaton based parsers then your
> combinators won't have much to do. But if you are doing more than
> that then they will have more to do, and hence be more complex. For
> example you can use the same combinators you define for parsing to do
> type checking, model state, etc -- things I would not want to do with
> an automaton. That is why I said that combinators are really sequencing
> constructs.
*What* I've been saying is that there is a generality-efficiency
tradeoff, and that the monadic approach goes for generality at the
expense of efficiency, by precluding optimizations in those very
important cases in which combinators are used for parsing. After all,
the papers cited earlier in this discussion were all about parsing, and
the thread was originally about string processing, coroutines and Icon
generators. Here's an interesting question. A context-free grammar can
be seen as a set of simultaneous recursive definitions in which the
RHSs are polynomials (where + is alternation and * is concatenation;
more generally, this can be expressed in terms of semirings and formal
power series, with some very interesting advantages). A regexp pattern
matcher is just a special case (modulo complications with non-regular
extensions like matching against pattern variables set by earlier \(
... \) ). Now, could the mappings of (special cases of) such grammars
to deterministic automata be extended to interesting cases of monadic
expressions? I have some vague ideas on how this might go based on
recent work on weighted transducer determinization by my colleague
Mehryar Mohri -- basically, one defines an algebraic structure over
"outputs" so that it makes sense to talk about the greatest common
factor of alternative outputs generated from the same input prefix, and
one can give sufficient conditions for determinizability. 
> 
> Fernando Pereira (pereira at research.att.com) wrote:
> : And even then, some
> : optimizations, those involving the various black boxes, will be
> : blocked. Here's a simple example. Assume that there are two black boxes
> : A and B, and that you use monadic glue to say "A or B". A and B may be
> : deterministic, but the result will not be, and exhaustive search will
> : be required in the glue. Sure, I can recognize such cases by hand and
> : create a new black box "A or B" which is optimized, but that is an
> : error-prone hand optimization. And one may not notice subtler
> : inefficiencies.
> 
> Your point appears to be that if you were programming only using automaton
> for parsing you wouldn't miss the above optimisation, but that somehow
> using combinators means you would miss the above optimisation. You
> and I must program differently then. If I were using automaton, even
> in the context of also using combinators, I would be aware of the
> potential for such optimisations and look for them.
You might be, by looking at the code and changing it by hand. But a
program would not, because in general the optimizable skeleton of
sequence, alternation and Kleene closure combinators is interspersed
with arbitrary functional code. The point of automata-based compilation
of regexps and CFGs is that it is *automatic*. Clearly, a monadic
program that is a direct transcription of a regexp or a CFG could be
similarly compiled. But from my readings (those mentioned in the thread
and others) as well as your comments, the whole point of monadic
combinators for parsing is to interweave the grammatical structure with
other computations, thus making the underlying grammatical structure
less accessible to optimization. As I suggested above, it would be
interesting to define classes of monadic programs that still allow (an
appropriate extension of) those optimizations. But AFAIK that has not
been done.

-- F



From claird at Starbase.NeoSoft.COM  Thu Jul  8 09:39:12 1999
From: claird at Starbase.NeoSoft.COM (Cameron Laird)
Date: 8 Jul 1999 08:39:12 -0500
Subject: [PSA MEMBERS] Using (physical) units and quantities
References: <1280719924-72790315@hypernet.com> <3784268C.340CAD46@callware.com>
Message-ID: <7m29m0$f57$1@Starbase.NeoSoft.COM>

In article <3784268C.340CAD46 at callware.com>,
Ivan Van Laningham  <ivanlan at callware.com> wrote:
			.
			.
			.
>> NT has a POSIX (har har) subsystem which will give you case
>> sensitivity. But none of the tools work, so you're dead in the water.
>> 
>
>Oh, I didn't know that.  So, how does anyone know it exists?  Rumour? 
>It's documented but not implemented?  You mean people actually *read*
>MSDN docs????
			.
			.
			.
1.  There are *two* POSIX subsystems.  There's one
    Microsoft did with its first releases of NT.
    It somehow passes requirements, so that MS is
    eligible for certain US government contracts
    (or equivalents), but no one uses it.  NO one.
    Seriously.  Now that I think of it, I'm not
    even sure it is compliant; it could well be
    just that MS *says* it is.  It's a non-issue,
    because you have in effect to disable NT to get
    at the POSIXness.

    I'll decline this morning to give a URL for this
    travesty.
2.  Softway Systems <URL:http://www.interix.com> (also
    is a mostly independent corporation that sells a
    very nice Unix (now called Interix, originally
    OpenNT) operating under NT.  This is good stuff,
    technically.  Microsoft's relation to Interix is
    far too byzantine for me to explain today.
3.  Many people *pay* to read MS documents.  Smart
    people, well familiar with Unix, whom I respect,
    do so, and then enthusiastically describe the
    experience to me.  While the phenomenon continues
    to puzzle me, I can report multiple instances.
    I'm not much opposed to any of this in principle,
    but, when I've tried, the experience hasn't been
    happy.  All the documents I've needed have been
    frustrating.
4.  Ivan, I haven't yet solved the mystery of why
    e-mail from you doesn't reach me.  Let's try some
    experiments.  Please send test messages to
      <claird at neosoft.com>
      <cameron at lairds.com>
      <claird at www.guardtower.net>
      <claird at place.net>
      <test4 at oneplace.com>.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From wtanksle at dolphin.openprojects.net  Thu Jul 29 13:10:19 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Thu, 29 Jul 1999 17:10:19 GMT
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <379F6D43.3531FE0@compaq.com>
Message-ID: <slrn7q12nr.552.wtanksle@dolphin.openprojects.net>

On Thu, 29 Jul 1999 08:51:15 +1200, Greg Ewing wrote:
>Gordon McMillan wrote:

>> But finding a match in a list, and testing whether you
>> fell off the end without finding one, is (without the else clause) a
>> much messier proposal.

>Not if you go about it the right way. I always
>put such loops in a procedure of their own,
>and use return:

>Not only does this not require a weird else
>clause, it doesn't require break either. I
>almost never use break - and when I do I usually
>regret it later. I wouldn't mind if there weren't
>any break - and without break, there would be no
>need for the weird else.

I have to admit that I do the same -- this is why I've never been tripped
up by the else.

>Greg

-- 
-William "Billy" Tanksley



From r.skowronek at fz-juelich.de  Thu Jul  1 02:43:45 1999
From: r.skowronek at fz-juelich.de (r.skowronek at fz-juelich.de)
Date: Thu, 01 Jul 1999 07:43:45 +0100
Subject: Decent Tutorial
Message-ID: <000015fa@bossar.com.pl>

From: r.skowronek at fz-juelich.de

Newbie <8499 at greenhead.ac.uk> wrote:

: Where can I find a decent tutorial which is 
: for beginners to Python, preferably including 
: some documented source code and avaliable for 
: download.

Have a look at http://www.spindoczine.com/folder5

There is an (almost?) complete book which I as a newbee to python found very
helpful. It is in MS-Word 97 format though.

Rolf

-- 
r.skowronek at fz-juelich.de



From catlee at my-deja.com  Wed Jul  7 11:52:57 1999
From: catlee at my-deja.com (catlee at my-deja.com)
Date: Wed, 07 Jul 1999 15:52:57 GMT
Subject: Getting contents of Text / Entry widgets
References: <7ltmfu$nbt$1@nnrp1.deja.com> <c75pv252spo.fsf@pc142.cosc.canterbury.ac.nz>
Message-ID: <7lvt4k$g8a$1@nnrp1.deja.com>

In article <c75pv252spo.fsf at pc142.cosc.canterbury.ac.nz>,
  Timothy R Evans <tre17 at pc142.cosc.canterbury.ac.nz> wrote:
> The class below does this, see the example at the bottom of the script
> for how to use it.  This example just prints the contents of the entry
> widget.
>
> ==============================================================
> #!/usr/bin/python
>
> from Tkinter import *
>
> class CallbackEntry(Entry):
>     '''Replacement of Entry widget.  Set the "callback" attribute to
>     your function, which will be called whenever the entry widget
>     changes and will be passed the new contents of the widget'''
>
>     def __init__(self, *args, **kw):
>         apply(Entry.__init__, (self,)+args, kw)
>         self.stringvar = StringVar(self)
>         self.stringvar.trace_variable('w', self.changed)
>         self.configure(textvariable=self.stringvar)
>         self.callback = None
>
>     def changed(self, *args):
>         'Note that the args given to this method are useless'
>         if self.callback != None:
>             self.callback(self.stringvar.get())
>

Great!  This works perfectly!  Thanks for your help!  I've come up with
a replacement for the Text widget.  The same method could be used with
almost any widget, I imagine.  This doesn't pass the text back to the
callback function since I only want to know when the text changes, not
necessarily what it is.

from Tkinter import *

class CallbackText(Text):
   '''Replacement of Text widget.  Set the "callback" attribute to
   your function, which will be called whenever the text widget
   changes.'''
   def __init__(self,*args,**kw):
      apply(Tkinter.Text.__init__,(self,)+args,kw)
      self.bind("<Key>",self.active)
      self.callback=None

   def active(self,event):
      self.after_idle(self.idle)

   def idle(self):
      if self.callback != None:
         self.callback()


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From rodisi01 at my-deja.com  Wed Jul 21 09:48:42 1999
From: rodisi01 at my-deja.com (Isidor)
Date: Wed, 21 Jul 1999 07:48:42 -0600
Subject: Win95(9x/NT?) file extension / association tricks
Message-ID: <199907211252.HAA14699@cro.ots.ac.cr>

Hello Pythologists-

Here are some things I've done on my win95 boxito. They may or
may not work for win98 and NT, I have no idea.

When I right click on a .py file, I now have the following
options:

1)     run (same as installed "open", but I edited its "name" in
  the registry)
2)     open in IDLE
3)     open in notepad
4)     run w dos
5)     run w.o console

One of these may be of particular interest to you: "Run w DOS". I
made this one because I got really frustrated with buggy modules
running, printing out their error messages in the console, and
then the console disappearing moments later. I tried fiddling
with pifs and such, but had no luck. This file extension seems to
solve that problem. It runs a dos shell first, then python, then
your selected module. The dos shell remains open after the module
finishes running so you can see the error messages.

To make any of these additional actions, open a directory window,
and View --> Options --> File Types. Scroll down the file types
until you get to Python File. Double click on it. In the "Edit
File Type" dialog, click the "new" button.

In the text box titled: "Action", type the name of the action you
want, e.g., "open in IDLE", w/o the quotation marks. Put an
ampersand (&) in front of the letter that you want underlined in
the popup menu. Be sort of careful here because you won't be able
to edit this field in the future without going into the registry.
(Can you say "annoying"?)

Here is a list of action and "application used to perform action"
pairs:

1)     Open w &IDLE

  "C:\your python directory name here\pythonw.exe" "C:\your
  python directory name here\Tools\idle\idle.pyw" "%1" %*

  on my machine:

  "C:\Program Files\Python152\pythonw.exe" "C:\Program
  Files\Python152\Tools\idle\idle.pyw" "%1" %*

  This one opens an IDLE python shell and opens the module you
  selected as a separate window.

  NOTE: if IDLE is already open, this action will start from
  scratch and open yet *another* python shell etc. If anyone
  knows how to avoid this, I would be very happy to learn about
  it! (See below for more details on things I've tried.)

2)     Open w &Notepad

  Notepad.exe

3)     Run w &DOS

  C:\WINDOWS\command.com /k python.exe "%1"

  I like this one a lot for the reasons explained above.

  NOTE: you need to have your python directory on your system's
  "path" in order for this one to work. Putting or not putting
  python on that path has been a big subject of debate on this
  list. I'm not going to get into it here though. If you don't
  want to put python on your path, try this instead:

  C:\WINDOWS\command.com /k "C:\your python directory name
  here\python.exe" "%1"

  It works on my machine as well.

4)     Run w/o &console

  "C:\your python directory name here\pythonw.exe" "%1" %*


NOTE: some of your files may not be stored in the same places as
mine. I have no idea whether command.com still exists in win98 or
where it exists if it does, and where it exists on NT.

Most of this kind of trickery I figured out for myself, but I
found the following site somewhat helpful as well:  (
http://cpcug.org/user/clemenzi/technical/  go to the pages about
win explorer and win registry...good stuff!) It has info for
different win versions, too.


IDLE hack:

Ok, you idle, ahem, I mean, IDLE hackers, how do I do the
following?: a pop-up menu action item that (a) checks whether
IDLE is running. (b1) If it is, open the file in the already
running session (as a separate window, of course). (b2) If it
isn't, do what my other IDLE extension does: open an IDLE python
shell and the selected module.

What I tried:

After bungling my way through the various modules of IDLE,
following calls from one place to another (for more details on
that adventure, see a forthcoming post to this list), I
discovered: FileList.py. The test function for that module looked
like it might do the trick. "What an easy solution!" I thought. I
made the following extension action thing-a-ma-bob:

"C:\Program Files\Python152\pythonw.exe" "C:\Program
Files\Python152\Tools\idle\FileList.py" "%1"

When I ran it, up popped my module in an IDLE window without an
accompanying IDLE python shell. Yay. But....wait...it did this
*whether or not* IDLE was already open. No good. It *looked* like
I solved the problem, but I hadn't. The lone module window was
running in its own session of IDLE. If IDLE is already open, none
of the other existing IDLE windows appear in the "window list" of
the new module window. (Did that sentence make sense?)

(Does IDLE do DDE? Is that a *really* dumb question?)

Why do I want to do this? you ask. It probably seems pretty
stupid to you. Well, I like having a folder for each python
project. I don't want to have to open the registry to add a new
path to the python path every time I add a new project.) Also, I
tend to have a lot of directory windows open. If I am looking at
a directory with a module in it that I want to run, I don't want
to have to go to the IDLE window, go to "open file", and browse
all over kingdom come just to get back to the directory I was
just looking at. ;)

It would also be fun to do drag and drop...oooh...(not very
useful, but intriguing)

Anybody got any ideas about that?

Ok, I hope this is of some use to someone.

Take care-

Isidor






From mikael at toolwood.com  Thu Jul  1 23:47:15 1999
From: mikael at toolwood.com (Mikael Lyngvig)
Date: Fri, 02 Jul 1999 03:47:15 GMT
Subject: win32com.client -- beginner question
References: <3779BC79.720C1281@uni-bonn.de>
Message-ID: <FF880675767F8C91.3D82743A8B0CE885.EF0E36D9A525E1DE@lp.airnews.net>

On Wed, 30 Jun 1999 08:43:06 +0200, Axel Schmitz-Tewes
<uzs59g at uni-bonn.de> wrote:

>Hello,
>
>I ' ll try to use the comclient ability of python. In the Doku I have
>found
>the example:
>
>o = win32com.client.Dispatch("Excel.Application")
>o.Visible = 1
>o.Workbooks.Add() # for office 97 ? 95 a bit different!
>o.Cells(1,1).Value = "Hello"
>
>Because I have installed Excel95 I have given it a try and that's the
>result:
>
>Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
>Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>> import win32com.client
>>>> o= win32com.client.Dispatch("Excel.Application")
>>>> o.Workbooks.Add()
>Traceback (innermost last):
>  File "<stdin>", line 1, in ?
>  File "E:\Programme\Python\win32com\client\dynamic.py", line 400, in
>__getattr_
>_
>    raise AttributeError, "%s.%s" % (self._username_, attr)
>AttributeError: Excel.Application.Workbooks
>
>But Workbooks.Add is a available method as I can test with VBScript.
>Is my installation corrupt or do I do some silly mistake or perhaps some
>more serious?

Did you remember to run the makepy.py script in
Python\win32com\client?  You need to run that one, select the COM
interface you want to use, e. g. "Microsoft Excel 5.0 Object Library
(1.0)" (or similar).  makepy.py then generates the necessary Python
definitions, in a file in the Python\win32com\gen_py directory (named
according to the COM interface's GUID.

After that you should be able to access Excel without too much
trouble.

There's a introduction to the Win32COM stuff, which includes a brief
description of makepy.py, in the QuickStartClientCom.html file
included in the Win32COM installation.


-- Mikael




From skip at mojam.com  Sat Jul 24 00:46:59 1999
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 23 Jul 1999 23:46:59 -0500 (CDT)
Subject: int(long(-sys.maxint-1)) fails on Linux
In-Reply-To: <000f01bed58c$4abb9e20$832d2399@tim>
References: <37992994.F4A6A30@vision.arc.nasa.gov>
	<000f01bed58c$4abb9e20$832d2399@tim>
Message-ID: <14233.17625.185090.895516@146.chicago-41-42rs.il.dial-access.att.net>

    timbot> Check out this link:

    timbot> http://www.comp.it.bton.ac.uk/burks/burks/language/asm/artofasm/artof001.htm

    timbot> Arguably the best book on assembler programming ever written!
    timbot> The chapters on x86 addressing modes and LEA will chase the dark
    timbot> clouds away.

Does the timbot have any other interesting URLs squirrelled away in its
large backside cache?

;-)

Skip




From evan at tokenexchange.com  Mon Jul 19 14:07:57 1999
From: evan at tokenexchange.com (Evan Simpson)
Date: Mon, 19 Jul 1999 13:07:57 -0500
Subject: Killing a thread
References: <7mu6a5$h61$1@nnrp1.deja.com> <7mvml3$u5k$1@nntp6.u.washington.edu>
Message-ID: <ePJk3.4$Z11.6082@news-dal.corridex.com>

This makes me wonder; Zope has a similar problem, which it mostly handles
with severe restrictions on what's available to the user (no 'while', only
'for', and no '[x]range' to generate arbitrarily large tuples).
Nevertheless, a creative malcontent could consume a lot of resources with
nested loops.

Bytecodehacks to the rescue?  <waves hands furiously> Wrap the user code in
a function (if it isn't already), find all jump-back codes, and insert a bit
of test code in front of them which bails out if ?a global variable is set?
(not sure what's best, here).  You'll take a slight performance hit, but
that probably won't matter in this context.  Perhaps you could inline a call
to time.clock and use it to limit the running time of the function.

Donn Cave <donn at u.washington.edu> wrote in message
news:7mvml3$u5k$1 at nntp6.u.washington.edu...
catlee at my-deja.com writes:
| I'm writing a progrma where some user code is executed in a restricted
| environment running in a separate thread.  Now, if the user is
| malicious, s/he can just put 'while 1: 1+1'.  I'd like to be able to
| recover from this.  Is there any way I can kill the thread that is stuck
| in the infinite loop?






From Anders.Andersen at p98.f112.n480.z2.fidonet.org  Thu Jul  1 11:05:28 1999
From: Anders.Andersen at p98.f112.n480.z2.fidonet.org (Anders Andersen)
Date: Thu, 01 Jul 1999 16:05:28 +0100
Subject: Reflection in Python?
Message-ID: <00001605@bossar.com.pl>

From: Anders Andersen <AAndersen at ACM.Org>

> I've just started programming in Python, and I could really use some
> help figuring out something.  I'd like to implement Kent Beck's unit
> tester in Python (someone posted a very nice alternative a few weeks
> ago, but Junit's approach works better for my project's needs).  How
> do I do the equivalent of what Java calls reflection?

I am not familiar with Kent Beck's unit tester, but I did write a note
a while ago about Python and reflection.  This work has now matured a
little bit more, and is a part of my Dr. thesis that will be submitted
soon.  Anyway, the note is available at:

  http://www.itek.norut.no/~anders/abstract/andersen98a.html

> Thanks, Anders Schneiderman 

Anders,

-- 
Anders Andersen - <mailto:aa at computer.org> - <http://w3.to/andersen/>



From conversy at lri.fr  Wed Jul 28 09:36:18 1999
From: conversy at lri.fr (Stephane Conversy)
Date: Wed, 28 Jul 1999 15:36:18 +0200
Subject: beep
Message-ID: <379F0751.71214802@lri.fr>

Hi all.

How to make the system beep with python ?
I couldn't find anything about that in the faq and in the doc, and I'm
almost
sure I've seen something about that...

    stef


--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr






From drew.mcdowell at msfc.nasa.gov  Fri Jul  9 14:55:03 1999
From: drew.mcdowell at msfc.nasa.gov (Drew McDowell)
Date: 9 Jul 1999 18:55:03 GMT
Subject: Python Code Repository?
Message-ID: <7m9tmr$6qg$36@rks1.urz.tu-dresden.de>

After seeing the "Brensenham line algorithm" post above, I'm curious
if there is a public "python code repository" (something akin to CPAN)
out there?  Or does Starship assume that role?

-Drew



From gubitz at netcologne.de  Sun Jul 11 18:14:15 1999
From: gubitz at netcologne.de (Hans Gubitz)
Date: 12 Jul 1999 00:14:15 +0200
Subject: Tk-Problem?
Message-ID: <lz1zeej0lk.fsf@netcologne.de>

I need some help. What am I doing wrong?

gubitz at pumuckl:~/python/apk > python
Python 1.5.1 (#1, Apr  5 1999, 00:08:45)  [GCC egcs-2.91.66 19990314 (e on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> float("1")
1.0
>>> from Tkinter import *
>>> root=Tk()
>>> button=Button(root, text="Quit", command=root.destroy)
>>> button.pack()
>>> root.mainloop()
>>> float("1")
Traceback (innermost last):
  File "<stdin>", line 1, in ?
ValueError: float() literal too large: 1


--
Hans Gubitz
email: gubitz at netcologne.de



From m.faassen at vet.uu.nl  Wed Jul 21 11:35:06 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 21 Jul 1999 15:35:06 GMT
Subject: Spreading the word about Python; ?s and ideas
References: <199907211307.IAA14980@cro.ots.ac.cr>
Message-ID: <7n4pba$u8a$1@news.accu.uu.nl>

Isidor <rodisi01 at my-deja.com> wrote:
> ...And
> that's where Python came in for me. And maybe that's where it can
> come in for a *lot* of other people. Maybe?

I hope so. I'd be wonderful if people would use Python instead of 
VBScript/Other Office stuff to program, but that's just random 
aesthetics. :)

[tweaking tools]
> An aside: Is this a Windoze only
> phenomenon? Are macs so well-designed that they don't cause this
> sort of angst?

I think such mac tweak tools exist too; but the mac interface is a bit
better integrated and had more time to grow, so it may be less necessary.

Still, there's this system (Kaleidoscope?) to totally change the way
your entire Mac GUI looks, so, yes, it exists. :)

> I am guessing that all folks using unix already
> know that they can make things be the way they want them...a unix
> user *is* a power user, by definition(?).)

I think there are two main groups of Unix users:

* Power user/developer

* Person who doesn't know that Unix is used
They just use apps that happen to run in Unix.

Though I think with the rise of Linux we may indeed see the rise of a class
of non-power unix users.

[snip]

> 1. Things that office scripters will probably need help with:

> a)     object orientation
[snip]

> b)     how do I do x in Python? In WordBasic/Excel/Access, it
>   was blah blah...

>   It strikes me that there is a key set of tasks that office
>   scripters quickly learn how to do. Many of these key tasks are
>   not available in python's built-in functions; they need to be
>   imported from separate modules. This seemed like a big deal to
>   me at first. The importance (and normal-ness ;)  ) of
>   importing needs to be explained early on.

That's indeed an interesting difference. Experienced software developers
tend to be acutely aware of the importance of modular code, and code
organisation issues in general. Python's module system is simple to use,
yet good at organizing things (due to the fact that everything is in its
own namespace, so you don't get issues like 'where is this function/class
defined' so often (and you don't have to come up with wacky names because
other names were already used).

>   An "office-
>   scripter's index" to the module documentation would be
>   invaluable. "Change directory? See os.chdir()") (See the
>   recent debate in this list about where to put os.chdir() in
>   the documentation. The whole time I was asking myself...you
>   mean I have to *load* something just to change the directory?!
>   whoa. ;)

You usually only have to load it once in your application, though, at the
beginning. :)

The idea in Python is that there is a core language that only provides
basic language functionality; control structures, lists, dictionaries, and
such. Then there are a number of built-in modules that provide some very
basic things -- to do with input and output, and such. These are the only
things Python needs to function.

The rest (including things like changing directories) can be left away,
and can be different on different platforms.

In practice the separation is not as pure as I describe it here, but it comes
close to it.

> *Whatever* the correct place for os.chdir() is, it is
>   going to be buried too deep...I suspect that "correctness" is
>   audience specific.) Some nice "office-scripter-targetted"
>   faq's could help a lot along these lines. (I'm already
>   starting my own collection, of course. When they're worth
>   anything, I'll be happy to share them.)

This sounds like a good idea to me!

> c)     lots of other really obvious things that I can't think of
>   right now 'cause I'm too tired.

> 2.     Distributing Python scripts

[snip distribution difficulties]

Many people feel the same; it's too hard to distribute Python scripts to
end users, and it's also could be easier to distribute Python extensions.

Many many issues are involved in this, and that's why the distutils-sig
(special interested group) was started, to discuss the problems, and to
come up with an infrastructure to solve them. We're slowly getting somewhere,
and input is very welcome still. Your perspective is certainly different from
lots of us, who look at things more from the computer science/software
engineering perspective. Take a look at:

http://www.python.org/sigs/distutils-sig/

Thanks for your postings; I haven't read all of them yet but your 
perspective is very interesting! How'd you find out about Python in the
first place? (or did you say somewhere in the posts? :)

Regards,

Martijn




From MHammond at skippinet.com.au  Tue Jul 13 19:17:23 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Wed, 14 Jul 1999 09:17:23 +1000
Subject: Python and COM (long)
References: <4D0A23B3F74DD111ACCD00805F31D8100DB90F1D@RED-MSG-50> <378d8ee2.10645225@news.freeserve.net>
Message-ID: <7mghas$bgo$1@m2.c2.telstra-mm.net.au>

Toby Dickenson wrote in message <378d8ee2.10645225 at news.freeserve.net>...
>Bill Tutt <billtut at microsoft.com> wrote:
>
...
>So, my conclusion is that you need to support QueryInterface for
>_both_ the IID of your dispinterface and IID_IDispatch, and that they
>must both return the same implementation of IDispatch.

I wish I had time to dig the thread up (in the pycom-dev archives) and
ensure I am not talking crap, but this also appears to be consistent with
how you must work with multiple dispatch interfaces - with fear of being
totally wrong, I seem to recall we must change win32com in that situation to
explicitely QI for the dispinterface, as a QI for IDispatch may not give us
the IDispatch we are after...

[This is still a pending win32com change too]

Mark.





From donn at u.washington.edu  Tue Jul 27 19:33:58 1999
From: donn at u.washington.edu (Donn Cave)
Date: 27 Jul 1999 23:33:58 GMT
Subject: 1.5.2 for: else:
References: <Pine.LNX.4.04.9907271612140.9849-100000@emerald.netskate.ru> <19990727142943.V4068@xs4all.nl> <slrn7ps8ej.vm0.wtanksle@dolphin.openprojects.net>
Message-ID: <7nlfl6$7te$1@nntp6.u.washington.edu>

wtanksle at dolphin.openprojects.net (William Tanksley) writes:
|On Tue, 27 Jul 1999 14:29:43 +0200, Thomas Wouters wrote:
|>On Tue, Jul 27, 1999 at 04:15:44PM +0400, Oleg Broytmann wrote:
|
|>|    The following program:
|>| ----------
|>| for a in ['a', 12]:
|>|    print a
|>| else:
|>|    print "Empty!"
|>| ----------
|
|> http://www.python.org/doc/current/tut/node6.html#SECTION006200000000000000000
|
|> 4.4 break and continue Statements, and else Clauses on Loops 
|
|> Loop statements may have an else clause; it is executed when the loop
|> terminates through exhaustion of the list (with for) or when the condition
|> becomes false (with while), but not when the loop is terminated by a break
|> statement. 

| Well, I'm suprised.  I had also expected the else clause to run on empty
| loop.  It seems so natural, in line with the meaning of a loop (loop on
| this; otherwise do that).
|
| Fortunately I've never actually used it and thus been disenchanted.

[(I'm confused? else clause _does_ execute after an empty loop, like
   for x in ():
       pass
   else:
       print 'for else'
)]

It seems to me to work the only useful way, so you almost can't
go wrong - at worst, if you mistake its meaning, you won't use it.
That may not be the same thing as "natural", but it's close.
It's one of the neat little things that make me pause for a moment
of appreciation when I get to use it.

| Understanding the current semantics seems to require understanding the
| nature of loop tests in Python.  Since I understand them I can agree that
| this makes sense, but I'm still a bit leery, because you're not 'elsing'
| on the data but rather on code inside the loop.
|
| I would _definitely_ use exceptions to handle this.

Note that "try/except" also takes an "else", which also happens to have
the only useful sense and executes when the try block executes with no
exception.  It's handy for code that should execute in the normal case
but does not itself need to be enclosed in an exception handler, so you
can focus the try block more precisely on the section that needs it.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From aaron_watters at my-deja.com  Mon Jul 26 11:07:17 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Mon, 26 Jul 1999 15:07:17 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com> <al86737ocwo.fsf@sirppi.helsinki.fi>
Message-ID: <7nhtiv$h8f$1@nnrp1.deja.com>

In article <al86737ocwo.fsf at sirppi.helsinki.fi>,
  Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:

> None of them touch my primary speed problem; the function/method
> invocation 'speed'; C++ function/method invocation seems 200+ times as
fast
> as Python's, and it really limits the proper OO nature of things,
> especially while using apply (another 4x slowdown). Deep class
hierarchies
> seem no-go, as do several other things.

This is absolutely valid.  Perhaps it is a perversion of some
kind but I simply don't use deep hierarchies, &c.  I find them
confusing too often anyway.

Another perversion
is the use of aggregate data types -- instead of wrapping an
abstraction around a number, wrap an abstraction over a sequence
or matrix of numbers, to allow aggregate operations at the lowest
level.  Sometimes I find that this sort of
perversion actually helps bring
out deep structure in the program.  Please see Python FAQ 4.7.

I think the "object oriented" approach is of fundamental importance,
but it can be overdone.  Sometimes there are other ways to think
about a problem which are also useful.

    -- Aaron Watters

===

We lose money on every unit, but make it up in volume.
   -- AT&T folklore


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From skip at mojam.com  Sun Jul 25 19:18:59 1999
From: skip at mojam.com (Skip Montanaro)
Date: Sun, 25 Jul 1999 18:18:59 -0500 (CDT)
Subject: Anti-Port scan.... (Port Monitor / Logger)
In-Reply-To: <rNHvjEdhm5Pp-pn2-vyeUbHbYW9pR@d185d18f8.rochester.rr.com>
References: <rNHvjEdhm5Pp-pn2-vyeUbHbYW9pR@d185d18f8.rochester.rr.com>
Message-ID: <14235.39649.786121.5135@dolphin.mojam.com>

    Benjamin> I'm having some activity on my modem that I can't explain, but
    Benjamin> I run OS/2....  So I can't use NUKENABBER, etc, for Windows to
    Benjamin> monitor the ports on my "server" box.

Can you run Linux?  If so, you can use tcpdump to monitor your ppp
connection.  It's the best way to look at what's really happening on your
network. (No Python content, but then there are some things with good
non-Python solutions already.)

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From dwelton at linuxcare.com  Fri Jul 30 20:52:32 1999
From: dwelton at linuxcare.com (David N. Welton)
Date: 30 Jul 1999 17:52:32 -0700
Subject: Import a lot of modules, or, import 'foo'
Message-ID: <874silabe7.fsf@linuxcare.com>

So, basically, I'd like to do this:

 tests = ['fs', 'swap']

 for t in tests:
     import t

 do more stuff on what promises to be a lot of modules

I've looked around a bit, but can't find what I need.  Maybe I'm just
going about this the wrong way, but... Hmm, now I want to know how to
do this in any case:-)

Thanks,
-- 
David N. Welton, Web Engineer, Linuxcare, Inc.
415.354.4878 x241 tel, 415.701.7457 fax
dwelton at linuxcare.com, http://www.linuxcare.com/
Linuxcare. At the center of Linux.



From mwh21 at cam.ac.uk  Tue Jul 13 08:22:14 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: Tue, 13 Jul 1999 13:22:14 +0100
Subject: ANN: Stackless Python 0.3
In-Reply-To: <378AFD45.8793D507@appliedbiometrics.com>
References: <378A40AD.B4FDB79F@appliedbiometrics.com> <m37lo5kpd0.fsf@atrus.jesus.cam.ac.uk> <378AFD45.8793D507@appliedbiometrics.com>
Message-ID: <Pine.LNX.4.10.9907131314330.18703-100000@atrus.jesus.cam.ac.uk>

On Tue, 13 Jul 1999, Christian Tismer wrote:
> Michael Hudson wrote:
> 
> > Very nice, but
> 
> > >>> apply(l.sort,())
> > Traceback (innermost last):
> >   File "<stdin>", line 1, in ?
> > SystemError: error return without exception set
> > >>> l
> > [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
> 
> Here is the patch.
> It was one of those "=" instead of "==" bugs.
> 
> Line 131 of bltinmodule.c should read now
> 
> 	if (retval == Py_UnwindToken) {
> 	-----------!!------------------
> 
> I'm updating the files right now, without changing the version number.
> 
> Thanks, Michael!
> 
> 

A pleasure!
Passes all of make test now - except for rfc822, which is odd, and happens
with both stackless and vanilla python, so presumably isn't your fault.

Now, about continuations...

this-is-posted-from-Pine-rather-than-my-normal-gnus-so-I-hope-
it-all-works-OK-ly y'rs
Michael

PS: If anyone wants a patch from latest CVS source to stackless-python
0.3, building a binary called stackless-python (on unix), drop me a line
and I'll email one to you.

-- 
Oh, very funny. Very sar-cah-stic.         http://www.ntk.net
http://www.ntk.net/doh/options.html - ho ho




From Voynow  Thu Jul  1 17:10:43 1999
From: Voynow (Voynow)
Date: Thu, 01 Jul 1999 13:10:43 -0800
Subject: wddx wrapper?
References: <%0Qe3.4034$lr3.164499@news2.jacksonville.net>
Message-ID: <930863445.26638@www.remarq.com>

If you're in a Windows environment you can use Python's COM
support to to use the COM object they offer.

Sasha Voynow

sasha at ofoto.com



**** Posted from RemarQ - http://www.remarq.com - Discussions Start Here (tm) ****



From jsight at mindspring.com  Thu Jul 22 19:54:53 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Thu, 22 Jul 1999 19:54:53 -0400
Subject: Dreaded Newbie Question
References: <K4tK+KAA26l3EwvK@epstein.co.uk>
Message-ID: <3797AF4D.5991CDBA@mindspring.com>

This group is as good for newbies as anyone else, IMO, but I believe
that there is also a Python "newbie" list.  Also, many common questions
really are answered in the FAQ.

As to your question, I'm really not 100% sure about the "right" way to
do it, but look at the stat function under
http://www.python.org/doc/current/lib/os-file-dir.html for a good
start.  Generally the lib documentation is the best place to start
looking for any stuff like this.  In this particular case you would have
been looking for section 6.1, "Miscellaneous OS interfaces".

Enjoy!

----
Jess
http://ww.biddin.com/pyPit/

Dani Epstein wrote:
> 
> Hi!
> 
> A newbie to Python (though currently coding in Delphi), I'm having a
> little difficulty tracking down stuff, although I have the tutorial and
> manuals. Where is the best place to get more info, and why is there no
> newsgroup for newbies like myself?
> 
> For the moment though my grateful thanks goes to anyone who can tell me
> how to get a file's size, and where that info is to be found in the
> references.
> 
> BTW, I'm running PythonWin 1.5.1
> --
> 
> Dani Epstein



From David.sparrman at delphi.se  Thu Jul  1 04:34:06 1999
From: David.sparrman at delphi.se (David.sparrman at delphi.se)
Date: Thu, 1 Jul 1999 10:34:06 +0200
Subject: Out of Office AutoReply: Daily Digest: comp.lang.python 1/2
Message-ID: <6935EFF15210D311896D00A0C9DAF28F0470D4@exchange.delphi.se>

Jag har f?r n?rvarande semster
?ter p? kontoret m?ndag den 25 juli

/mvh David 




From mwh21 at cam.ac.uk  Thu Jul 15 07:54:45 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: Thu, 15 Jul 1999 12:54:45 +0100
Subject: ANN: Stackless Python 0.3
In-Reply-To: <378AFD45.8793D507@appliedbiometrics.com>
References: <378A40AD.B4FDB79F@appliedbiometrics.com> <m37lo5kpd0.fsf@atrus.jesus.cam.ac.uk> <378AFD45.8793D507@appliedbiometrics.com>
Message-ID: <Pine.LNX.4.10.9907151242590.16482-100000@atrus.jesus.cam.ac.uk>

On Tue, 13 Jul 1999, Christian Tismer wrote:

> 
> 
> Michael Hudson wrote:
> 
> > Very nice, but
> 
> > >>> apply(l.sort,())
> > Traceback (innermost last):
> >   File "<stdin>", line 1, in ?
> > SystemError: error return without exception set
> > >>> l
> > [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
> 
> Here is the patch.
> It was one of those "=" instead of "==" bugs.
> 
> Line 131 of bltinmodule.c should read now
> 
> 	if (retval == Py_UnwindToken) {
> 	-----------!!------------------
> 
> I'm updating the files right now, without changing the version number.
> 
> Thanks, Michael!
> 
> 

More problems, I'm afraid.

It would seem to be impossible for a script executed from the commandline
(ie, like

stackless-python foo.py

) to exit with any exit status other than 1, except by reaching the end.

I'm sorry, I'm not explaining this very well. If a script tries to exit by
calling sys.exit, sys._exit or raising SystemExit the exit status is
always 1. os._exit works as advertised, I've just noticed.

They all work when interactive too.

Another, maybe connected phenomenon is that

echo "fudge" | stackless-python

does *not* print an error!

echo "fudge" | stackless-python -i

does!

As you can see I have built stackless python into an executable called
stackless-python. This required extensive hackery of the build process, so
that could be at fault. This makes even less sense to me.

A consequence of all this is that stackless-python cannot install itself
without error on my system (compileall.py "fails").

Ideas?

TIA
Michael

PS: Did you see stackless python got a mention on Linux Weekly News Today?
Cool.

-- 
Oh, very funny. Very sar-cah-stic.         http://www.ntk.net
http://www.ntk.net/doh/options.html - ho ho




From KFGKU at UYTI.COM  Sat Jul 24 02:32:01 1999
From: KFGKU at UYTI.COM (KFGKU at UYTI.COM)
Date: Sat, 24 Jul 1999 00:32:01 -0600
Subject: ONLY ADULTS 76883
Message-ID: <240799003201@UYTI.COM>

Free XXX:

http://207.240.225.250






y2mmW at n;



From mstenber at cc.Helsinki.FI  Mon Jul  5 02:44:02 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 05 Jul 1999 09:44:02 +0300
Subject: Is it just me, or..
References: <000017cd@bossar.com.pl>
Message-ID: <al84sjj8ukt.fsf@myntti.helsinki.fi>

"Michael P. Reilly" <Michael.P..Reilly at p98.f112.n480.z2.fidonet.org> writes:
> From: "Michael P. Reilly" <arcege at shore.net>
> Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
> : Is there a bug in the handling of pipes with the Python? Esp. in
> : non-blocking way.
> : Example:
> :  (r,w)=os.pipe()
> :  r = os.fdopen(r, "r", 0)
> :  w = os.fdopen(w, "w", 16384)
> :  w.write('foobarbaz')
> :  select.select([r],[],[]) # returns r, as it should
> :  r.read(1) # returns 'f'
> :  select.select([r],[],[]) # blocks!
> 
> : Now select _blocks_, despite there being obviously some data that at least
> : should not have been buffered elsewhere. Also, fdopened stuff's reading
> : features are bit annoying, as doing read(bigNumber) results in blocking
> : call, as opposed to recv(bigNumber)'s semantics. Thus instead of pipes
> : (=fast) I am using TCP/UNIX sockets (=slow) but they seem to work better
> : for what I need. Am I missing something really obvious?
> 
> : This is on Linux, if it matters.
> 
> At first glance, I'd say that it is the buffering in the fdopen(w, "w",
> 16384) call.  Put the bufsize to 0 or call the flush method.


Oh, yes, I was flushing it (that example was done by hand out of existing
code that is actually in use). Yet, still no dice. Also, buffer size for
"r"does not matter (0,1,N); behavior is same always. Frustrating. (Oh
well. UNIX sockets work, I suppose)

>   -Arcege

-- 

	Markus Stenberg



From arcege at shore.net  Sat Jul 10 06:20:10 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Sat, 10 Jul 1999 10:20:10 GMT
Subject: console show in Python?
References: <378626CF.9495FAB0@mry.rbd.com>
Message-ID: <7m9tnk$6qg$62@rks1.urz.tu-dresden.de>

Scott Shillcock <shillcoc at mry.rbd.com> wrote:
: I am moving to Python from Tcl/Tk.
: Is there something similar to the Tcl/Tk 'console show' command in
: Python.

: I am using Tkinter for my gui and once I call mainloop() I no longer
: have access to the Python command line. I want to be able to interact
: with my gui from the command line to test new code.

: From Tcl/Tk I could turn on and off the interactive command line with
: 'console show' and 'console hide'. That is all I want to do with Python.
: Can anyone help me with this or point me in the right direction.

On UNIX anyway, if you use the quit() method from the mainloop, then
you can restart from Python's interactive prompt by calling the
mainloop() method again.

The code for Tkinter should make this behavior consistant on all
platforms, but has anyone tested this there?

  -Arcege




From mal at lemburg.com  Tue Jul  6 11:10:45 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 06 Jul 1999 17:10:45 +0200
Subject: mxCGIPython question (was: Build the Python library into the binary)
References: <376F9518.D3252CEE@interet.com> <7kolnv$gp6$1@nnrp1.deja.com> <015e01bebe1f$202e9560$4500a8c0@thomasnotebook>
Message-ID: <37821C75.405CD9C5@lemburg.com>

Thomas Heller wrote:
> 
> > Hi
> >
> > did you have a look at mxCGIPython at
> > http://starship.skyport.net/~lemburg/mxCGIPython.html ?
> > This is a single executable not shared library with the
> > standard Python library embedded.
> >
> > Cheers
> > --
> > Florent Heyworth
> >
> Works great, except for one problem (under Windows):
> you cannot use any extension modules, because they do
> not find python15.dll (or, worse, they find the wrong
> python15.dll).
> 
> Is it possible to build a python15.dll containing the
> standard library or is it only possible to freeze executables?

The same problem seems to exist on Unix platforms. For some
reason, the symbols defined in cgipython are not accessible
to the dynamic linker on Unix. Don't know why, because 'nm cgipython'
clearly shows them as being defined...

For Windows: you can probably use the python15.dll from the
standard 1.5.2 distribution (simply place it into the same
dir as cgipython).

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   178 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From krodgers at tdyryan.com  Fri Jul  2 14:40:16 1999
From: krodgers at tdyryan.com (krodgers at tdyryan.com)
Date: Fri, 02 Jul 1999 19:40:16 +0100
Subject: Debugging embedded pgm under WinNT
Message-ID: <0000168b@bossar.com.pl>

From: krodgers at tdyryan.com

I have Python 1.5.2 embedded in an application on WinNT (4.0, SP4) using
Visual C++ 6.0 SP2.  When I compile the project in Debug mode, even
though it links correctly to the python15_d.lib file, it won't run
because it complains about "can't find python15_d.dll".  So I got the
source, compiled the python15_d.dll, put it in WINNT/system32 right next
to python15.dll, ran my program . . . and now it couldn't find
"multiarray" (my embedded Python code uses the Numeric extensions).  I
verified that sys.path was correct and included the Numeric directories.
Do I have to build the Numeric extensions under Debug mode?  Anybody
seen something like this before?  Thanks in advance . . .
--
Kevin Rodgers  krodgers at tdyryan.com  kmrodgers at my-deja.com


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From robin at alldunn.com  Mon Jul 19 02:16:51 1999
From: robin at alldunn.com (Robin Dunn)
Date: Sun, 18 Jul 1999 23:16:51 -0700
Subject: what is python?
References: <7mt81n$s6e$1@bgtnsc03.worldnet.att.net>
Message-ID: <cpzk3.1$ly.393540@news.randori.com>

>
> So what are people using python for?

Nearly everything.

> What types of problems is python
> better suited relative to, say, java?

Nearly everything.

> I do alot of server-side programming
> & xml stuff.
>

Pick a problem, any problem... Python will do at least as good of a job as
Java, and probaly better once you get to know it good.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!







From nathan.froyd at rose-hulman.edu  Wed Jul  7 19:26:57 1999
From: nathan.froyd at rose-hulman.edu (Nathan Froyd)
Date: 07 Jul 1999 18:26:57 -0500
Subject: make or similar in Python
References: <378390DD.58CA6BD5@nevex.com> <3783db76.118974654@news.erols.com>
Message-ID: <m3673woxby.fsf@froydnj.laptop.rose-hulman.edu>

kernr at mail.ncifcrf.gov (Robert Kern) writes:
> On Wed, 07 Jul 1999 13:39:42 -0400, Greg Wilson <gvwilson at nevex.com>
> wrote:
> >Hi, folks.  Has anyone ever written a dependency management/update tool
> >(like 'make') in Python?
> >I'm not looking for something that reads makefiles, and happens to be
> >written in Python; I'm looking
> >for a Python module that users import, then provide data to, in order to
> >manage dependencies.  (The
> >answer I got from the Perl crowd was 'cool idea, but no'; I'm hoping
> >I'll have more luck here.)
> >Thanks,
> >Greg (who'd really, really like to override some of the things that
> >GNU make does)
> 
> Check out
> ftp://ftp.python.org/pub/python/contrib/System/pymake-0.1-sb.tar.gz
> You will probably have to hack it up a bit to do exactly what you
> want, but it's probably a really good start.

If you're not tied down to Python (or you're looking for a coding
challenge), there's a tool written in Perl called 'cons' that purports to
eliminate make and whose 'Makefiles' are actual Perl programs. You might
take a look at that and adapt it to your needs.
-- 
Nathan  |  froydnj at rose-hulman.edu  |  http://www.rose-hulman.edu/~froydnj/

Evolution is a million line computer program falling into place by accident.



From greg.ewing at compaq.com  Thu Jul 22 17:03:34 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 23 Jul 1999 09:03:34 +1200
Subject: how do I find memory leaks?
References: <7n4skv$8do@thalamus.wustl.edu>
Message-ID: <37978726.EA30794B@compaq.com>

"Heather A. Drury" wrote:
> 
> I thought using python/vtk
> would take care of all this memory garbage for me

A reasonable expectation, but unfortunately Python
doesn't quite deliver. Garbage collection in Python
is by reference counting only, so if there are any
cycles in your garbage, it never goes away.

You have to design your program so that it either
doesn't create any cycles or ensures that they are
broken when the data is no longer needed.

There was a module called Cyclops announced here
by Tim Peters recently for finding cycles in data 
structures. Take a look back in dejanews.

Greg



From mal at lemburg.com  Wed Jul 21 13:29:47 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Wed, 21 Jul 1999 19:29:47 +0200
Subject: ImportError: Unresolved symbol
References: <199907210617.BAA26862@pheno.physics.wisc.edu>
Message-ID: <3796038B.D7EA6174@lemburg.com>

Nikolas Kauer wrote:
> 
> When I run a Python CGI script (first line #!/patjh/bin/python), I get the
> following error in the Apache error log:
> 
> [Tue Jul 20 16:58:55 1999] [error] [client ...] Premature end of script headers: /usr/share/test/apache_1.3.6/cgi-bin/db_
> driver.cgi
> Traceback (innermost last):
>   File "/usr/share/test/apache_1.3.6/cgi-bin/db_driver.cgi", line 24, in ?
>     from cgi_namespace import *
>   File "/usr/share/test/python/lib/python1.5/lib-dynload/cgi_namespace.py", line 8, in ?
>     from cgi_state import State
>   File "/usr/share/test/python/lib/python1.5/lib-dynload/cgi_state.py", line 1, in ?
>     import types, DateTime, string
>   File "/usr/share/test/python/lib/python1.5/lib-dynload/DateTime.py", line 8, in ?
>     from mxDateTime import *
> ImportError: Unresolved symbol in /usr/share/test/python/lib/python1.5/lib-dynload/mxDateTime.so: timegm
> 
> When I start an interactive Python session, and do "import DateTime", or execute the Python script on
> the command line (like tcsh> ./db_driver.cgi) no error messages appear, everything seems to work fine.
> The Web server runs on Digital Unix 4.0E.
> 
> The only solution I can currently think of would be to try to statically link mxDateTime.so ...

Looks like you are using a different LIB path for CGIs and
your interactive sessions. Here's a snippet from my mxODBC
page about CGI and shared libs:

     Running mxODBC from a CGI script

     ODBC drivers and managers are usually compiled as a shared library. When running CGI scripts
     most HTTP daemons (aka web servers) don't pass through the path for the dynamic loader (e.g.
     LD_LIBRARY_PATH) to the script, thus importing the mxODBC C extension will fail with unresolved
     symbols because the loader doesn't find the ODBC driver/manager's libs. 

     To have the loader find the path to those shared libs you can either wrap the Python script with a shell
     script that sets the path according to your system configuration or tell the HTTP daemon to set or
     pass these through (see the daemon's documentation for information on how to do this; for Apache
     the directives are named SetEnv and PassEnv). 

Note that you can also turn off the timegm API usage
by editing the Setup file in DateTime/mxDateTime and
recompiling the package.

Hope that helps.
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   163 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From news at dorb.com  Thu Jul 22 18:06:42 1999
From: news at dorb.com (Darrell)
Date: Thu, 22 Jul 1999 18:06:42 -0400
Subject: string.replace
References: <Pine.BSF.4.10.9907221446520.20655-100000@soda.CSUA.Berkeley.EDU>
Message-ID: <vCMl3.1619$u74.24916@newsr1.twcny.rr.com>

Works for me.
Upgrade your python version.
Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on w
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import sys, string
>>> line = 'cat "in" the hat'
>>> line = string.replace( line, '"', '')
>>> print line
cat in the hat
>>>

--
--Darrell
marlon <marlon at soda.CSUA.Berkeley.EDU> wrote in message
news:Pine.BSF.4.10.9907221446520.20655-100000 at soda.CSUA.Berkeley.EDU...
> can anyone tell me why this breaks?
>
> Python 1.4 (Nov  4 1997)  [GCC 2.7.2.3]
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import sys, string
> >>> line = 'cat "in" the hat'
> >>> line = string.replace( line, '"', '')
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> AttributeError: replace
> >>>
>
>
> _
>





From avoinov at raima.com  Fri Jul 23 16:46:05 1999
From: avoinov at raima.com (Alexander V. Voinov)
Date: Fri, 23 Jul 1999 13:46:05 -0700
Subject: Interactive use of Tk
References: <3798A5D7.674D@cs.bham.ac.uk>
Message-ID: <3798D48D.CE9BA6A3@raima.com>

Stuart Reynolds wrote:

> I'm trying to debug a tkinter application I have using the interactive
> prompt via XEmacs. The program runs fine first time around but when run
> again (without exiting the interactive prompt) it complains that the
> application has been destroyed. How to I stop the application from being
> destroyed?
>
> Secondly, has anyone managed to get tkinter windows running whilst still
> being able to use the interactive prompt (e.g. by starting the mainloop
> in a separate thread). I guess its possible and its just the dodgy
> Solaris threads libraries we have installed here that stop it working
> for (the jpython examples also cause the command prompt to block for
> me).
>
> Cheers,
>
> Stuart
>
> PS. The code which won't run twice:
>
> -----
> if __name__ == '__main__':
>
>     from Tkinter import *
>     frame = Frame()
>     ... some other code here ...
>     Pack.config(frame)
>     frame.mainloop()
>
> -------
>
> #Works fine:
> >>> ## working on region in file /usr/tmp/python1f5be_...
> >>>
> #Run again:
> >>> ## working on region in file /usr/tmp/pythonh2FQL_...
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "/usr/tmp/pythonh2FQL_", line 117, in ?
>     frame = Frame()
>   File
> "/bham/ums/common/pd/packages/Python/lib/python1.5/lib-tk/Tkinter.py",
> line 1406, in __init__
>     Widget.__init__(self, master, 'frame', cnf, {}, extra)
>   File
> "/bham/ums/common/pd/packages/Python/lib/python1.5/lib-tk/Tkinter.py",
> line 1084, in __init__
>     self.tk.call(
> TclError: can't invoke "frame" command:  application has been destroyed

class MyToplevel(Tix.Toplevel):
    def __init__(self):
        Tix.Toplevel.__init__(self)
        self.tk.call('load', '', 'Tix')
        self.bind("<Alt-x>", self.destroy)
        self.bind("<Control-x>", self.destroy)
        self.tk.call('wm','withdraw','.')

    def destroy(self, *args):
        self.withdraw()
        Tix.Toplevel.destroy(self)
        self.quit()

Either replace Tix with Tkinter (I didn't try it with vanilla Tkinter,
though), or go to my Starship site and pick up a fixed version of PyTix (it
seems non-maintained so you may consider me as a vice-maintainer :-):

ftp://starship.python.net/pub/crew/avv/PyTix/

Alexander





From lvirden at cas.org  Fri Jul  2 08:28:57 1999
From: lvirden at cas.org (Larry W. Virden)
Date: Fri, 2 Jul 1999 08:28:57 -0400
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
In-Reply-To: Gerrit Holl's message as of Jul  2,  8:16
       <19990702141624.A15051@humbolt.nl.linux.org> <7li9cq$inp$1@srv38.cas.org>; from lvirden@cas.org on Fri, Jul 02, 1999 at 11:56:10AM +0000
Message-ID: <199907021228.IAA13745@cas.org>

On Jul 2,  8:16, Gerrit Holl wrote:
} Subject: Re:  Python GUIs: Abandoning TkInter and welcoming wxPython?
--> On Fri, Jul 02, 1999 at 11:56:10AM +0000, lvirden at cas.org wrote:
--> > From:	lvirden at cas.org
--> > According to Mikael Lyngvig <mikael at pobox.com>:
--> > :Well, TkInter is extremely slow, but not because of the Python part.
--> > When I run the Tk test suite  on my SPARC, Tk cranks thru creating and
--> > deleting windows as fast as the eye can track.
--> 
--> I think you should also say what kind of processor you have if you're
--> talking about this:
--> 
--> "Windows NT on a Pentium IV 500 MHZ runs over 10 times faster than a
--> Linux 386 30 MHZ" (is it?)

Probably worthwhile to indicate what OS, what version of Tk (and Python
if that's factoring into the discussion), and probably amount of memory.
Another factor may be the kind of video card - 8 bit vs 32 bit vs something
else...

I know I was referring to an Ultra 5 / Solaris 2.6 / Tk any version /
128 meg of memory, 8 bit color 19" monitor.


-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.



From ivanlan at callware.com  Thu Jul  8 12:05:20 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Thu, 8 Jul 1999 16:05:20 GMT
Subject: [PSA MEMBERS] Using (physical) units and quantities
References: <1280719924-72790315@hypernet.com>
		<3784268C.340CAD46@callware.com> <7m29m0$f57$1@Starbase.NeoSoft.COM>
Message-ID: <3784CC40.A1E6DA91@callware.com>

Hi All--

Cameron Laird wrote:
> 

[bobbitt]
                     .
>                         .
> 1.  There are *two* POSIX subsystems.  There's one
>     Microsoft did with its first releases of NT.
>     It somehow passes requirements, so that MS is
>     eligible for certain US government contracts
>     (or equivalents), but no one uses it.  NO one.
>     Seriously.  Now that I think of it, I'm not
>     even sure it is compliant; it could well be
>     just that MS *says* it is.  It's a non-issue,
>     because you have in effect to disable NT to get
>     at the POSIXness.
> 
>     I'll decline this morning to give a URL for this
>     travesty.

Thanks, but no thanks;-)

[more bobbitting]

> 3.  Many people *pay* to read MS documents.  Smart
>     people, well familiar with Unix, whom I respect,
>     do so, and then enthusiastically describe the
>     experience to me.  While the phenomenon continues
>     to puzzle me, I can report multiple instances.
>     I'm not much opposed to any of this in principle,
>     but, when I've tried, the experience hasn't been
>     happy.  All the documents I've needed have been
>     frustrating.

Oh, yeah, we here at Callware pay to "read" them too.  And there *is*
information there--you just can't get there from here, is all.  With
just average mediocrity, you can spend *days* tracking down how to turn
that stupid question mark off--only to discover that you can't without
major voodoo.  You'd think that with nearly a gigabyte of documentation
they'd figure out a straightforward way of indexing it, wouldn't you? 
If you don't *have* to dig around in the cesspool, don't, is my motto.
..


> 4.  Ivan, I haven't yet solved the mystery of why
>     e-mail from you doesn't reach me.  Let's try some
>     experiments.  Please send test messages to
>       <claird at neosoft.com>
>       <cameron at lairds.com>
>       <claird at www.guardtower.net>
>       <claird at place.net>
>       <test4 at oneplace.com>.
> --
> 

OK, I've added all these addresses to this reply.

<shotgun-empiricism,-like-pork-fat,-rules>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From donn at u.washington.edu  Tue Jul  6 13:02:22 1999
From: donn at u.washington.edu (Donn Cave)
Date: 6 Jul 1999 17:02:22 GMT
Subject: problem with lists
References: <378220B1.9E0C272@daimi.au.dk>
Message-ID: <7ltcqu$r5i$1@nntp6.u.washington.edu>

Peter Posselt Vestergaard <posselt at daimi.au.dk> writes:
| It seems like I've completely misunderstood something or at least done
| something wrong. There's noway I can see why these two scripts are
| outputting nothing or more correctly only the 4 first lines which cannot
| be seen in a browser. Anybody knows why?

I don't see any reference to nydata before its insert() operation,
so I would expect a NameError exception.

nydata.insert(0, 'hej') looks in the module namespace for an "nydata",
and then looks in that object's namespace for an "insert" attribute.
I guess you may want to bind that name to an initially empty list,
nydata = [].

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu
-----------------------------
| #!/usr/local/bin/python
| import cgi
|
| print "Content-type: text/html"
| print
|
| print "<HTML><HEAD>"
| print "<TITLE> ?ndring af data </TITLE>"
| print '</HEAD><BODY BGCOLOR="#FFFFFF"><FONT FACE="Helvetica">'
|
| nydata.insert(0,'hej')
| print nydata[0]
| print 'strange'
|
| print '</BODY></HTML>'



From fredrik at pythonware.com  Fri Jul  2 04:05:47 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 2 Jul 1999 08:05:47 GMT
Subject: Mailbox, tell and read
References: <1281273761-39459133@hypernet.com>
Message-ID: <00d601bec461$b24c01b0$f29b12c2@secret.pythonware.com>

Gordon McMillan wrote:
> > First, are there platforms other than DOS/Windows/Win32 where there
> > is a difference between text and binary mode?
> 
> There's a theoretical difference on *nix - otherwise, they wouldn't 
> have made up the distinction!

who are "they?"

fwiw, the ANSI C rationale mentions Unix as the
special case here... go figure ;-)

</F>




From catlee at my-deja.com  Tue Jul 13 13:20:43 1999
From: catlee at my-deja.com (catlee at my-deja.com)
Date: Tue, 13 Jul 1999 17:20:43 GMT
Subject: Python re module
References: <199907120337.XAA06072@eric.cnri.reston.va.us> <14217.63630.458437.222314@amarok.cnri.reston.va.us>
Message-ID: <7mfsgt$ood$1@nnrp1.deja.com>

In article <14217.63630.458437.222314 at amarok.cnri.reston.va.us>,
  "Andrew M. Kuchling" <akuchlin at mems-exchange.org> wrote:
...
>        However, GvR reports that it isn't actually hanging, just
> taking a long time to finish.  I think it escapes the infinite loop
> because \1+ is actually treated the same as \1{0,INT_MAX}, so once
> it's counted up to INT_MAX (2**31-1) = 2.1 billion, it then returns.
>
>      The latest version of PCRE, v2.06, does not have a problem with
> this (more checks for repeated null-string matches have been added),
> so I think the problem will be fixed when the Python module is
> resynchronized with the latest version.

Ok, that makes sense.  Two questions: where can I get the latest
version of PCRE and how can I tell (from python) what version of PCRE I
am using?

Thanks,
Chris


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From robin at jessikat.demon.co.uk  Thu Jul 22 14:16:39 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 22 Jul 1999 19:16:39 +0100
Subject: Is anybody aware of a tree GUI widget for tkinter or Pmw?
References: <3796E213.9F326412@eng.cam.ac.uk>
Message-ID: <Qy5FLCAHA2l3Ewst@jessikat.demon.co.uk>

In article <3796E213.9F326412 at eng.cam.ac.uk>, D. Lopez-De-Ipina
<dl231 at eng.cam.ac.uk> writes
>Hi,
>
>Could anybody please tell me if there exits any TREE control widget for
>tkinter or Pmw? I know that Pmw provides a rather rich set of
>'megawidgets' but the tree control has not been implemented yet as far
>as I know.
>
>Thanks a lot,
>
>Diego
>
the BWidget set in pure tcl at http://www.unifix-online.com/BWidget/
contains a nice tree widget.
-- 
Robin Becker



From bwarsaw at cnri.reston.va.us  Wed Jul  7 13:23:56 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Wed, 7 Jul 1999 13:23:56 -0400 (EDT)
Subject: Mail-lax
Message-ID: <14211.36140.697494.282369@anthem.cnri.reston.va.us>

Uh, many of you may start to notice a flood of old email for various
mailing lists run from python.org.  I've been investigating the
complaints that some people were missing some messages.  Mailman
queues messages that it can't deliver, then wakes up every 1/2 hour
and tries to re-deliver those messages.  For some reason, the cronjob
that re-delivers wasn't enabled, possibly as far back as February!  I
have no idea why this was, and I'm manually cleaning out the queue
now, so apologies if you get innundated.  Just be aware that some of
the messages you'll be getting could be several months old.

Hopefully this will fix the delivery problems some of you have been
seeing.

-Barry



From mgushee at havenrock.com  Sat Jul 24 20:22:58 1999
From: mgushee at havenrock.com (Matt Gushee)
Date: 25 Jul 1999 09:22:58 +0900
Subject: how do I find memory leaks?
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AAC26@molach.origin.ea.com>
Message-ID: <m37lnptw65.fsf@havenrock.com>

"Stidolph, David" <stidolph at origin.ea.com> writes:

> The following is our Debug.py file.

Thanks!



From tbryan at arlut.utexas.edu  Fri Jul  9 09:45:11 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Fri, 09 Jul 1999 08:45:11 -0500
Subject: htmllib, cgi, HTMLfmt, genCGI, HTMLgen, html, Zope, ...
References: <tot2m7.su.ln@final.math.uni-bonn.de> <19990708223856.A29441@optiplex.palga.uucp>
Message-ID: <3785FCE7.59FA77EB@arlut.utexas.edu>

Gerrit Holl wrote:

> I used HTMLgen and I don't really understand why to use it. If you don't
> know HTML, you can't use it: you need to understand tags like <B>, <I>, <U>
> and such. I strongly recommend to have a look at HTML:

If you know HTML, then you can use HTMLgen to help you maintain a 
consistent appearance from page to page on a web site.  In particular, 
many users find a consistent use of navigation aides (such as next page,
previous page, and home) to be very helpful.  It is also useful for 
helping multiple people working together to produce pages with a 
consistent appearance by using the same document base classes.
It also provides a nice abstraction for programatically generated
documents (for example, from CGI).  That way, you don't have to 
output raw HTML with print statements from your program.  Again, if
several CGI scripts should produce pages with a similar appearance, 
you can factor out the common parts into a class.  Then each program 
only needs to instantiate the class and fill in the extra information.

For a site of a dozen pages, HTMLgen is definitely unnecessary.
For a site of millions of pages, I think that HTMLgen would prove
insufficient.  For a site of a few hundred pages, I would probably
consider HTMLGen, especially if python scripts will be generating 
those pages (either as CGI scripts or as a cron job).  

As a bonus, it seems that it would be simple to use the HTMLgen
document classes to generate multiple formats.  I have been thinking
about writing output methods for LaTeX output. :)

-------------------------------------------
Tom Bryan
Applied Research Laboratories
University of Texas at Austin



From mwh21 at cam.ac.uk  Sun Jul  4 12:26:08 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 04 Jul 1999 17:26:08 +0100
Subject: Another bytecodehacks!
Message-ID: <m34sjkpeji.fsf@atrus.jesus.cam.ac.uk>

It's been a while, but there's a new version of bytecodehacks on the
starship:

http://starship.python.net/crew/mwh/

Quite a lot has changed, especially under the hood. There's a
design-by-contract module, a function that will make another tail
recursive, a set of simple rationalizations/optimizations (originally
posted to this newsgroup by Corran Webster) and probably lots of stuff
I've forgotten about.

The documentation has been updated, but there are probably places
where it's out of date, or just plain wrong. If you find such a place
please do tell me!

The documentation contains what I think is quite a nice description of
one way of avoiding cycles in complex data structures (look at the
doumentation of fleshy).

Have fun, and thanks for reading!

Michael



From Fredrik.Lundh at p98.f112.n480.z2.fidonet.org  Fri Jul  2 03:57:43 1999
From: Fredrik.Lundh at p98.f112.n480.z2.fidonet.org (Fredrik Lundh)
Date: Fri, 02 Jul 1999 08:57:43 +0100
Subject: wddx wrapper?
Message-ID: <00001656@bossar.com.pl>

From: "Fredrik Lundh" <fredrik at pythonware.com>

Hank Fay <hank at prosysplus.com>:
> Does anyone know a reference to a Python wrapper for wddx (www.wddx.org)?
> It looks like a neat way to pass data around (being able to transport
> recordsets between differnet languages, in particular), and I'd like to be
> able to have Python doing some of the heavy lifting on the server side.

no, but if the protocol doesn't matter, as long as it's
light-weight, cross-platform, XML, and supported by
lots of languages, you might wish to check out:

    http://www.xmlrpc.com

implementations available for python, perl, java, tcl, etc.
and zope supports it too, btw.

</F>



From kuncej at mail.conservation.state.mo.us  Wed Jul 14 10:54:50 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Wed, 14 Jul 1999 09:54:50 -0500
Subject: Printing from a wxpython application under Windows98
Message-ID: <s78c5e7a.070@mail.conservation.state.mo.us>

>How I handle printing in Windows?  How do you set up the
>printer?

I'm also interested in the answer to this :-)

>4. Any other way?

You might look into piddle http://www.strout.net/python/piddle/ 
It can generate output in many formats including pdf and postscript.
A wxPython backend for piddle is also under development.

  --Jeff




From nospam at this.address  Mon Jul 19 22:19:51 1999
From: nospam at this.address (Michael Novello)
Date: Mon, 19 Jul 1999 19:19:51 -0700
Subject: Python and the Web
Message-ID: <3793dd72.0@hoh.olympus.net>

Can anyone tell me where to find information on using python to grab form
data and create dynamic web pages?

Thanks,

Michael





From nadim_ghaznavi at my-deja.com  Thu Jul 29 14:25:20 1999
From: nadim_ghaznavi at my-deja.com (nadim_ghaznavi at my-deja.com)
Date: Thu, 29 Jul 1999 18:25:20 GMT
Subject: shelve portability problem
References: <eZUl3.1080$wm4.135734@news.uswest.net> <7nirjd$n52$1@m2.c2.telstra-mm.net.au>
Message-ID: <7nq6a7$vka$1@nnrp1.deja.com>

I'm having a problem moving a Python shelve ODB from Solaris to Linux.
Here's some sample code to demo the problem:

On the Solaris Platform:

# python
Python 1.5.2 (#1, Jun  5 1999, 15:51:52)  [GCC 2.8.1] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import shelve
>>> friends = ['Jane','Jan','Simon']
>>> odb = shelve.open('odbfile')
>>> odb['foo']=friends
>>> odb.close()

Which leaves me with two files:
# ls -l
-rw-r--r--   1 root     other          0 Jul 29 13:50 odbfile.dir
-rw-r--r--   1 root     other       1024 Jul 29 13:57 odbfile.pag

To confirm that all is well, lets have a look at the data again:
On Solaris:

# python
Python 1.5.2 (#1, Jun  5 1999, 15:51:52)  [GCC 2.8.1] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import shelve
>>> samedb = shelve.open('odbfile')
>>> buddies = samedb['foo']
>>> buddies
['Jane', 'Jan', 'Simon']
>>> samedb.close()

As expected, the data is still in the odb.  Next we'll move the
odbfile.dir and odbfile.pag to the Linux platform (I used FTP, nothing
strange there).

# ls -l
-rw-r--r--   1 root     root            0 Jul 29 13:59 odbfile.dir
-rw-r--r--   1 root     root         1024 Jul 29 13:59 odbfile.pag
# python
Python 1.5.1 (#1, Mar 21 1999, 22:49:36)  [GCC egcs-2.91.66 19990314/Li
on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>  import shelve
>>> nodb = shelve.open('odbfile')
>>> nodb.keys()
[]
>>> nodb.close()
>>>
# ls -l
-rw-r--r--   1 root     root         4096 Jul 29 14:00 odbfile
-rw-r--r--   1 root     root            0 Jul 29 13:59 odbfile.dir
-rw-r--r--   1 root     root         1024 Jul 29 13:59 odbfile.pag

What seems to be happening is that Solaris is storing the data in an old
DBM style database, and Linux is using a new gdbm format.  When I try
opening the old Solaris ODB, python incorrectly opens a new gdbm ODB
instead.

A couple of things to note:  I do have the free gdbm package installed
on my solaris platform (even though Python isn't using it):

# pkginfo GNUgdbm
application GNUgdbm        gdbm

I know I can work around this by using pickle and a specific DB module,
but does anyone know how I can get it working using the shelve module?

	Nadim




Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From mstenber at cc.Helsinki.FI  Sun Jul  4 14:19:28 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 04 Jul 1999 21:19:28 +0300
Subject: Is it just me, or..
Message-ID: <al8btds8ehb.fsf@myntti.helsinki.fi>

Is there a bug in the handling of pipes with the Python? Esp. in
non-blocking way.

Example:
	(r,w)=os.pipe()
	r = os.fdopen(r, "r", 0)
	w = os.fdopen(w, "w", 16384)
	w.write('foobarbaz')
	select.select([r],[],[]) # returns r, as it should
	r.read(1) # returns 'f'
	select.select([r],[],[]) # blocks!

Now select _blocks_, despite there being obviously some data that at least
should not have been buffered elsewhere. Also, fdopened stuff's reading
features are bit annoying, as doing read(bigNumber) results in blocking
call, as opposed to recv(bigNumber)'s semantics. Thus instead of pipes
(=fast) I am using TCP/UNIX sockets (=slow) but they seem to work better
for what I need. Am I missing something really obvious?

This is on Linux, if it matters.

-Markus

-- 
	This message was written in, or processed by a Linux system. 
	Linux, as in Linus' *NIX, not Lignux, as in RHS' arrogance.




From walrus at ans.net  Mon Jul 26 16:51:51 1999
From: walrus at ans.net (michael shiplett)
Date: Mon, 26 Jul 1999 20:51:51 GMT
Subject: Editors for Macintosh 
In-Reply-To: Your message of Mon, 26 Jul 1999 13:02:32 -0600.
             <379CB0C8.FA84F91C@callware.com> 
References: <379CB0C8.FA84F91C@callware.com>
Message-ID: <199907262051.QAA25501@fuseki.aa.ans.net>

"ivl" == Ivan Van Laningham <Ivan> writes:

ivl> Are there any other good editors (i.e., Python-aware) besides
ivl> Alpha out there for the Macintosh? I don't want to _use_ one, I
ivl> just want to know if there are any that people recommend.

Regarding the Mac download from python.org, Jack Jansen's MacPython
Page [1] states

    Also included is an integrated development environment by Just van
    Rossum, which includes an editor, debugger and class browser all
    closely integrated.

I've found this IDE sufficient for Python development on the Mac.

michael


1. http://www.cwi.nl/~jack/macpython.html



From da at ski.org  Wed Jul  7 18:18:16 1999
From: da at ski.org (David Ascher)
Date: Wed, 7 Jul 1999 15:18:16 -0700 (Pacific Daylight Time)
Subject: Binary pickle portability
In-Reply-To: <Pine.OSF.4.05.9906292018360.28230-100000@pinax.Mathematik.Uni-Bielefeld.DE>
Message-ID: <Pine.WNT.4.04.9907071517320.248-100000@rigoletto.ski.org>

On Tue, 29 Jun 1999, Olaf Delgado wrote:

> On Tue, 29 Jun 1999, Olaf Delgado wrote:
> 
> > Lately I observed strange things when trying to unpickle a binary pickle
> > (including floating point numbers) produced on a different architecture.
> 
> Sorry, that was too quick.
> 
> After looking a bit closer, I found that the problem is probably related
> to the Numeric module (i.e., pickling matrices) and doesn't depend on
> whether binary or ordinary mode is used. So please don't bother now!
> Still, I am interested in whether people use binary pickles.

There is a bug in pickling of NumPy arrays, which I expect will be fixed
in the next release.  Binary pickles works across byte-order and byte-size
platforms (when it's bug-free =).

--david




From andrew at andrewcooke.free-online.co.uk  Thu Jul  1 08:54:04 1999
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Thu, 01 Jul 1999 12:54:04 GMT
Subject: theads & global namespaces
References: <377A5C88.32A1FAB6@notes.cba.ufl.edu>
Message-ID: <7lfod8$cjp$1@nnrp1.deja.com>

In article <377A5C88.32A1FAB6 at notes.cba.ufl.edu>,
  andersdt at notes.cba.ufl.edu wrote:
> Greetings all,
>
> I am trying to share globals between threads.  Is is possible for me
to
> share globals between threads that call functions from different
> modules?  Thus far I have had little success.  I ask this question
> because  I would prefer not to put all my code in one module.

Hi,

I may be teaching my granny to suck eggs here, so ignore this if it's
obvious, but in general you should not spin a thread while waiting for
another thread to do something (this doesn't fix your global problem,
it's just advice about multi-threaded programming).

All my multi-threaded experience is with java, but the classes in
Python are similar, so I'll have a go at explaining how it "should" be
done and someone can point out my errors...

(Incidentally, whether or not your program works as it is - assuming
the global access thing is fixed - will depend on the platform you
runit out, and the threads implementation used.  Using conditions,
it will run on any platform).

from threading import *

globalX = 0

def test():
  c = Condition()
  c.acquire() # get hold of the condition
  t = Thread(target=XChanger, argcs=c)
  t.start()
  c.wait() # this halts this thread and releases c
  print globalX
  c.release()

class XChanger:
  def run(cond): # cond is c, passed in from the args in Thread
    cond.acquire() # waits for c to be released (in c.wait() above)
    globalX = globalX + 1
    cond.notifyAll() # signal the other thread to restart when...
    cond.release() # ...we release this

print globalX # should print 0
test() # should print 1 inside test()
print globalX # should print 1

(This is only one way of doing things - it may not even be best, but
it shows what I'm thinking of).  Doug Lea has written a very good
book about threads, but it discusses Java.  Because the Python threading
module is similar, you may find it useful (especially if you already
know Java).

There's also a completely different approach to using threads, which
is based on a (fairly ancient) book by Hoare called Communicating
Sequential Processes.  It is more elegant, but not as popular (it's
used in occam, and there is a Java library to support it, but nothing
for Python that I know of - ask me if you're interested and I can ask
on a mailing list I am subscribed to).

Cheers,
Andrew


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From hschroeder at gmx.net  Fri Jul 30 07:25:12 1999
From: hschroeder at gmx.net (Henning Schroeder)
Date: Fri, 30 Jul 1999 13:25:12 +0200
Subject: Problem compiling Python on Suse 6.1
References: <37A01B2B.2CF78A89@gmx.net>
Message-ID: <o22sn7.co.ln@odin.henning-schroeder.de>

Tomaz Ficko <tomaz at gmx.net> wrote:
> I have a problem compiling Python 1.5.2 with Tk and readline support.
> I'm using Suse 6.1 (evaluation iso):
> gcc version egcs-2.91.66 19990314 (egcs-1.1.2 release)
> Tcl/Tk 8.0.5
> bash-2.02
I had also some problems with TK and Readline on SuSE 6.0.
SuSE renamed some files in order to be compatible with older versions
They renamed 
 /usr/X11R6/include/tk.h to /usr/X11R6/include/tk8.0.h
 /usr/X11R6/lib/tkConfig.sh to /usr/X11R6/tkConfig8.0.sh
 /usr/X11R6/lib/libtk.a to /usr/X11R6/lib/libtk8.0.a
 /usr/X11R6/lib/libtk.so to /usr/X11R6/lib/libtk8.0.so
 /usr/include/tcl.h to /usr/include/tcl8.0.h
 /usr/lib/tclConfig.sh in /usr/tclConfig8.0.sh
 /usr/lib/libtcl.a to /usr/lib/libtcl8.0.a
 /usr/lib/libtcl.so to /usr/lib/libtcl8.0.so
See also /usr/doc/packages/tk8.0/README.SuSE and 
/usr/doc/packages/tcl8.0/README.SuSE.

So there are two ways to fix this. Either you rename your files which would
break other installed programs or you patch the source :-). For the second
way look into _tkinter.c and tkappinit.c and change some lines containing
tcl.h and tk.h into tcl8.0.h and tk8.0.h
Addionally look into Modules/Setup to make sure that you link tkinter with
the right libraries (-ltk8.0 -ltcl8.0)

To have readline support (and ncurses btw) add -L/usr/lib/termcap everywhere 
you find -ltermcap in your Modules/Setup. This seems SuSE specific, too.


Still Problems? Mail me.
  Henning 





From G..David.Kuhlman at p98.f112.n480.z2.fidonet.org  Fri Jul  2 14:50:54 1999
From: G..David.Kuhlman at p98.f112.n480.z2.fidonet.org (G. David Kuhlman)
Date: Fri, 02 Jul 1999 19:50:54 +0100
Subject: Installing Python/XML on Windows?
Message-ID: <0000168c@bossar.com.pl>

From: dkuhlman at netcom.com (G. David Kuhlman)

Jonathan W Hendry (jhendry at shrike.depaul.edu) wrote:
> Does any documentation exist on this? How do I set up the
> paths so that Python can use all the various components of
> the package?
> 
> Thanks,
> 
> - Jon


Here is a message that should help.  You should also be able to
find it in the comp.lang.python archives at DejaNews.

  - Dave

==============================


From: digitome at iol.ie (Sean Mc Grath)
Newsgroups: comp.lang.python
Subject: Re: XML-toolkit for Windows - binaries available?
Date: Wed, 16 Jun 1999 19:37:35 GMT
Organization: Ireland On-Line
Lines: 33
Message-ID: <3767fc51.1547455 at news.iol.ie>
References: <376701a3.72288277 at newssrv> <wkzp208udt.fsf at ifi.uio.no>
NNTP-Posting-Host: dialup-011.ballina.iol.ie
X-Newsreader: Forte Free Agent 1.11/32.235
Xref: ix.netcom.com comp.lang.python:61665

On 16 Jun 1999 09:41:34 +0200, Lars Marius Garshol
<larsga at ifi.uio.no>
wrote:

>
>* Lee Borkman
>|
>| Can anyone help me out in installing from the standard
distribution?
>| I don't have a C-compiler on my system, but I do have
nmake/dmake.
>
>I think the web page is out of date, since the /xml/windows
directory
>in the CVS does contain the .dll files for pyexpat and sgmlop (and
has
>done for quite a while). So just download the ZIP file and they
should
>be there.
>
>If not, send me an email and I'll send them to you.
>
The binaries are there but a bit of fiddling is required to
get things working. (This is from memory but here goes)

The pyexpat and sgmlop dlls have to be copied into the
xml/parsers subdirectory. You also need to edit the
__init__.py file to include 'pyexpat' and 'sgmlop' in
the __all__ string.

Pyexpat requires xmlparse.dll. This dll needs to be
on your path.

I *think* that is what I did to get it all working:-)

Hope it helps,
Sean





From robin at alldunn.com  Thu Jul 15 00:51:40 1999
From: robin at alldunn.com (Robin Dunn)
Date: Wed, 14 Jul 1999 21:51:40 -0700
Subject: wxPython Form editor
References: <378c488e@news.xtml.co.uk> <378c8b3c@news.xtml.co.uk>
Message-ID: <tNdj3.9$oe.5111882@news.randori.com>

Nick Trout <nick@#spam_me_not#videosystem.co.uk> wrote in message
news:378c8b3c at news.xtml.co.uk...
> Is there a wxPython mail group/news group?
>

Yep, you can subscribe at
http://starship.python.net/mailman/listinfo/wxpython-users


>
> Nick Trout <nick at NOSPAMvideosystem.co.uk> wrote in message
> news:378c488e at news.xtml.co.uk...
> >Has anybody written a dialog/form editor for wxPython?
> >


Not that I know of.  I'd sure like to see one get contributed though, (so I
don't have to do it! :-)


--
Robin Dunn
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!
Try http://AllDunn.com/laughworks/ for a good laugh.






From greg.ewing at compaq.com  Wed Jul 14 16:58:08 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 15 Jul 1999 08:58:08 +1200
Subject: Recursive method
References: <1280168234-10452044@hypernet.com> <378CA422.B2E99CB7@eml.villa-bosch.de>
Message-ID: <378CF9E0.C6DD4B63@compaq.com>

Ralph Gauges wrote:
> 
> class Test:
>         def TestMethod(self):
>                 print "ABC"
> 
>         TestMethod(self)        # <- this doesn't work, because
> there is no 'self'
>         self.TestMethod()       # <- this doesn't work, because
> there is no 'self'

The reason there's no 'self' here is simply
that the code is not being executed in an 
environment where a name 'self' is defined.

Code which is inside a class definition but
not inside a method is executed when the class
is defined. At that time there isn't any instance
of the class that could be called 'self', so
it doesn't make sense to want to have 'self'
defined then.

This applies to both Python and JPython,
by the way. The code you wrote wouldn't work
in Python either.

If you want initialisation code which is only
executed once, the best place to put it is at
the module level, outside of any class definition.
Then you can make recursive calls, etc. without
any problem (as long as you don't nest function
definitions).

On the other hand, if you want initialisation code
executed for each instance (e.g. if the class
corresponds to a window, and you want a new
window for each instance of the class) then
the code has to go in the __init__ method, or
other methods that it calls. Again, methods can
call themselves or each other without problem.

Keep in mind that code in the class body is only
executed *once*, not every time an instance is
created.

Hope that helps,
Greg



From Mike.C..Fletcher at p98.f112.n480.z2.fidonet.org  Fri Jul  2 13:29:17 1999
From: Mike.C..Fletcher at p98.f112.n480.z2.fidonet.org (Mike C. Fletcher)
Date: Fri, 02 Jul 1999 18:29:17 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001681@bossar.com.pl>

From: "Mike C. Fletcher" <mfletch at vrtelecom.com>

None of the demonstrations I've seen are using a "grid" method.  There is
one which looks similar, the "header" demo, but that is a set of "list"
widgets standing next to each other and a sitting below a header widget.  To
create a real grid, you would need to do some (likely C++) programming.  You
would probably also want to make such thing allow for editing in-field, and
have the capability of updating directly from an application table to the
on-screen representation.  It should be possible to create such a thing in
Python by dropping data into a matrix of entry fields in response to
scrolling/moving etc., but a grid object generally wants a high-performance
solution with all sorts of funky options.

A "rich" text editor is evident in the Fox IDE, FXEditor, a separate project
from the Core Fox libraries, though I have not yet downloaded that project
to test it.  Something similar should likely become part of a base Fox
distribution if it is to become a widely deployed Python GUI.

Fox has a basic GUI editor under development (it is a separate project
again), though my tests with it so far have been unsuccessful, as the
application exits silently about every-other time I click on a widget.

Neither tkinter nor Fox has any perceptible "slowness" on my system, so I
cannot comment in that direction.

I'm continuing my experiments, will continue to report as time permits.

Enjoy,
Mike


Alexander V. Voinov <avv at quasar.ipa.nw.ru> wrote in message
news:377BF583.6846FBF9 at quasar.ipa.nw.ru...
...
> Did you look at the demos for the underlying toolkit? Some of them
presents a
> 'table' which is probably less rich than that of wxWindows, but it does
exist.
>
> What it really lacks is a 'richtext'/html widget, sort of which is present
and
> actively developed in wxWindows. As I understand it's just about to become
> available in wxPython.
...




From fredrik at pythonware.com  Thu Jul 15 18:12:54 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Jul 1999 22:12:54 GMT
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net>
Message-ID: <002901becf0f$3046fbd0$f29b12c2@secret.pythonware.com>

mary jeanette <marjea at uswest.net> wrote:
> I decided to learn python because I heard it was a very easy language to
> learn. So I downloaded it and bought a couple of very expensive
> paperback books from o'reilly only to discover that they are written for
> unix.

umm.  not entirely true.  afaik, most examples in
both Learning Python and Programming Python
works on all platforms.

> In other words, typing in the examples from the book into the
> interpretor only yields various syntax errors.

strange.  guess you have to tell us what you typed,
and what errors you got, before we can help you
figure out what's wrong.

> I have no idea what I am supposed to change in order to make the
> examples work, and apparently there is nowhere to turn in order
> to find out.

well, you found this newsgroup, didn't you? ;-)

if you post the errors, we can help you find out.
but a better way might be to join the Python
Tutor mailing list.  see:

    http://www.python.org/mailman/listinfo/tutor

    "This list is for folks who want to ask questions
    regarding how to learn computer programming
    with the Python language.

    Python (http://www.python.org) is a programming
    language which many feel is a good first language,
    because it makes it easy to express the fundamental
    concepts of programming such as data structures
    and algorithms with a syntax which many find easy
    to read and write. 

    Folks interested in learning about programming
    with Python are encouraged to join, as are folks
    interested in helping others learn. While the list
    is called tutor, anyone, whether novice or expert,
    can answer questions. 

    If individuals wish to start off-line conversations
    about a particular concept and become one-on-
    one tutor/tutee, that's fine. If either party wants
    to summarize what they learned for others to
    benefit, that's fine too."

hope this helps!

</F>




From tomaz at gmx.net  Thu Jul 29 05:13:15 1999
From: tomaz at gmx.net (Tomaz Ficko)
Date: Thu, 29 Jul 1999 11:13:15 +0200
Subject: Problem compiling Python on Suse 6.1
Message-ID: <37A01B2B.2CF78A89@gmx.net>

I have a problem compiling Python 1.5.2 with Tk and readline support.
I'm using Suse 6.1 (evaluation iso):
gcc version egcs-2.91.66 19990314 (egcs-1.1.2 release)
Tcl/Tk 8.0.5
bash-2.02

Can somebody help me.

Here is what I get when I run 'make' ( this is only the part with
errors).

make[1]: Entering directory `/home/tomaz/src/Python-1.5.2/Modules'
gcc  -Xlinker -export-dynamic python.o \
	  ../libpython1.5.a -lreadline
-L/usr/lib/libtermcap.so.2                 -L/usr/X11/lib -ltk8.0
-ltcl8.0 -L/usr/X11R6/lib -lX11        -lieee -ldl  -lm  -o python 
../libpython1.5.a(_tkinter.o): In function `Tkapp_CreateFileHandler':
/home/tomaz/src/Python-1.5.2/Modules/./_tkinter.c:1456: undefined
reference to `Tcl_GetFile'
../libpython1.5.a(_tkinter.o): In function `Tkapp_DeleteFileHandler':
/home/tomaz/src/Python-1.5.2/Modules/./_tkinter.c:1483: undefined
reference to `Tcl_GetFile'
../libpython1.5.a(_tkinter.o): In function `EventHook':
/home/tomaz/src/Python-1.5.2/Modules/./_tkinter.c:1912: undefined
reference to `Tcl_GetFile'
/usr/lib/libreadline.a(display.o): In function `rl_redisplay':
display.o(.text+0xb6e): undefined reference to `tputs'
/usr/lib/libreadline.a(display.o): In function `update_line':
display.o(.text+0x10c2): undefined reference to `tputs'
/usr/lib/libreadline.a(display.o): In function
`_rl_move_cursor_relative':
display.o(.text+0x1399): undefined reference to `tputs'
/usr/lib/libreadline.a(display.o): In function `_rl_move_vert':
display.o(.text+0x143f): undefined reference to `tputs'
display.o(.text+0x1476): undefined reference to `tputs'
/usr/lib/libreadline.a(display.o)(.text+0x1948): more undefined
references to `tputs' follow
/usr/lib/libreadline.a(display.o): In function `insert_some_chars':
display.o(.text+0x19ea): undefined reference to `tgoto'
display.o(.text+0x19f7): undefined reference to `tputs'
display.o(.text+0x1a27): undefined reference to `tputs'
display.o(.text+0x1a52): undefined reference to `tputs'
display.o(.text+0x1a7f): undefined reference to `tputs'
/usr/lib/libreadline.a(display.o): In function `delete_chars':
display.o(.text+0x1ab1): undefined reference to `tgoto'
display.o(.text+0x1abd): undefined reference to `tputs'
display.o(.text+0x1af2): undefined reference to `tputs'
/usr/lib/libreadline.a(display.o): In function `cr':
display.o(.text+0x1bf5): undefined reference to `tputs'
/usr/lib/libreadline.a(display.o): In function
`_rl_redisplay_after_sigwinch':
display.o(.text+0x1c2c): undefined reference to `tputs'
display.o(.text+0x1c4f): undefined reference to `tputs'
/usr/lib/libreadline.a(display.o)(.text+0x1c79): more undefined
references to `tputs' follow
/usr/lib/libreadline.a(terminal.o): In function `_rl_get_screen_size':
terminal.o(.text+0x84): undefined reference to `tgetnum'
terminal.o(.text+0xda): undefined reference to `tgetnum'
/usr/lib/libreadline.a(terminal.o): In function `get_term_capabilities':
terminal.o(.text+0x1b3): undefined reference to `tgetstr'
/usr/lib/libreadline.a(terminal.o): In function `_rl_init_terminal_io':
terminal.o(.text+0x26b): undefined reference to `tgetent'
terminal.o(.text+0x35c): undefined reference to `PC'
terminal.o(.text+0x366): undefined reference to `BC'
terminal.o(.text+0x370): undefined reference to `UP'
terminal.o(.text+0x3be): undefined reference to `tgetflag'
terminal.o(.text+0x3cf): undefined reference to `tgetflag'
terminal.o(.text+0x41d): undefined reference to `tgetflag'
terminal.o(.text+0x42e): undefined reference to `tgetflag'
/usr/lib/libreadline.a(terminal.o): In function `_rl_backspace':
terminal.o(.text+0x646): undefined reference to `tputs'
/usr/lib/libreadline.a(terminal.o): In function `ding':
terminal.o(.text+0x6df): undefined reference to `tputs'
/usr/lib/libreadline.a(terminal.o): In function `_rl_enable_meta_key':
terminal.o(.text+0x73e): undefined reference to `tputs'
/usr/lib/libreadline.a(terminal.o): In function `_rl_control_keypad':
terminal.o(.text+0x772): undefined reference to `tputs'
collect2: ld returned 1 exit status
make[1]: *** [link] Error 1
make[1]: Leaving directory `/home/tomaz/src/Python-1.5.2/Modules'
make: *** [python] Error 2




From jbkerr at sr.hp.com  Tue Jul  6 18:13:38 1999
From: jbkerr at sr.hp.com (Jim Kerr)
Date: Tue, 06 Jul 1999 15:13:38 -0700
Subject: problem with win32 install
Message-ID: <37827F92.3BC0BAED@sr.hp.com>

  I just tried installing the win32all package, and the installation
mysteriously failed. I don't remember the exact phrasing, but
something
to the effect of 

  AXScript and Python COM registration failed: 
    a device attached to the system is not functioning

  Hoping this was just a bogus error message, I tried loading a win32
module:

-- [C:/MKS] python
Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import win32com.client
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "C:\Program Files\Python\win32com\client\__init__.py", line 8,
in ?
    import dynamic, CLSIDToClass, pythoncom
  File "C:\Program Files\Python\win32com\client\dynamic.py", line 22,
in ?
    import pythoncom
ImportError: DLL load failed: A device attached to the system is not
functioning.
>>>

  The installation was done on an HP laptop running win95, with an
external floppy
drive attached. I can't figure out what 'attached device' this message
is referring
to.

  Any help in figuring this out would be much appreciated!
	
					-Jim

Jim Kerr
Hewlett-Packard
Santa Rosa Systems Division
1400 Fountaingrove Pkwy, MS 3USZ
Santa Rosa, CA 95403
Phone: 707-577-5575
Telnet: 577-5575



From hank at prosysplus.com  Thu Jul  1 07:47:45 1999
From: hank at prosysplus.com (Hank Fay)
Date: Thu, 1 Jul 1999 07:47:45 -0400
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <3773c3a0.102825445@news.oh.verio.com> <xPPc3.535$or1.22866@typhoon-sf.snfc21.pbi.net> <xJRc3.292$Of3.67094@news.shore.net> <7laovo$c4s$1@Starbase.NeoSoft.COM> <377A24B7.AD6FAA1A@cfdrc.com>
Message-ID: <7DIe3.4020$lr3.161937@news2.jacksonville.net>

The current reference is: http://www.cfdrc.com/FOX/fox.html

Hank Fay

--
http://www.prosysplus.com
Lyle Johnson <jlj at cfdrc.com> wrote in message
news:377A24B7.AD6FAA1A at cfdrc.com...
> Just to add fuel to the fire, we now have a Python interface to FOX, a
> cross-platform GUI toolkit:
>
>     http://www.cfdrc.com/pub/FOX/fox.html
>
> The latest version of the Python wrappers can be picked up at:
>






From pduffin at mailserver.hursley.ibm.com  Thu Jul 29 09:16:26 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Thu, 29 Jul 1999 14:16:26 +0100
Subject: Why not standard ? (Re: Building a shared libpython)
References: <Pine.GSO.4.03.9907281701150.15012-100000@spica> <001001bed938$b6b34330$f29b12c2@secret.pythonware.com> <37A02B7E.4083@mailserver.hursley.ibm.com>
Message-ID: <37A0542A.7EA4@mailserver.hursley.ibm.com>

Paul Duffin wrote:
> 
> Fredrik Lundh wrote:
> >
> > Carlos A M dos Santos <nospamplease at inf.ufrgs.br> wrote:
> > > I'm installing Python 1.5.2 and noted that there is no configure option to
> > > buld a shared libpython.  How could I do it?
> >
> > www.python.org => search => newsgroup search for
> > "shared libpython" brings up lots of hits.  start with
> > these:
> >
> >     http://www.deja.com/=dnc/getdoc.xp?AN=488431463
> >     http://www.deja.com/=dnc/getdoc.xp?AN=398994871
> >
> 
> Why is this not available with the standard distribution ?
> Could someone who knows explain why, or someone who can add it to the
> standard distribution.
> 

Fredrik kindly pointed me at 
	http://www.deja.com/=dnc/getdoc.xp?AN=398994871
again, (I did not look at it because I assumed it was a "how to").
In it Guido gives his reasons for why he considers a shared library
version of Python a low priority.

GvR writes:
> It's just a nightmare to get this to work on all different platforms.
> Shared library portability is a pain.

I do not think that it is as hard as all that; Python already supports 
loadable extensions (which I presume are shared objects) so all that
is left is dynamic linking / binding, and shared library issues.

On Unix systems Python either uses the OS backlinking mechanism or grows
its own which means that it will cope whether Python is a shared object
or not.

Creating executables, either requires that Python is a shared library
(which has a few issues around names etc) or that the executable 
dynamically loads Python as a shared object. Neither is really that
hard to do although it is probably a bit fiddly.

> In practice, few applications embed Python -- it's much more common to

Few applications embed Python because it is not very easy to do so.
Embedding Python in executables is probably not that hard but we are
talking about embedding Python dynamically.

> have Python extensions, which already are shared libraries.  Also,

There are lots of things which you cannot do if Python is not a shared
library.

    Dynamically embed in other scripting languages.
	Python (and Perl and others) have all benefited an awful lot
	from being able to dynamically load Tk. I am sure that Python
	has a lot to offer other scripting languages but can't simply
	because it is not dynamically loadable.

	See minotaur at http://mini.net/pub/ts2/minotaur.html for
	details on just how this could work.

    Browser plugins.

> serious embedders often want total control over which Python version
> and configuration they use so they wouldn't want to use a standard
> shared library anyway.  So while this motivation is nice in theory, I

I can still have total control over the configuration by simply 
renaming the library that I built to something else.

> doubt that it will save much in practice.  (Hence the low priority I
> give to making a shared library.)

Over the last couple of months I have seen probably one request a week 
for this.

At the very least add an option to build a libpython.a with Position
Independent Code so that it can be linked into shared objects.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From fredrik at pythonware.com  Thu Jul 15 13:36:35 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Jul 1999 17:36:35 GMT
Subject: MIME coding problems
References: <7ml3mj$l37$1@nnrp1.deja.com>
Message-ID: <004d01becee8$96b76b10$f29b12c2@secret.pythonware.com>

<sennomo at my-deja.com> wrote:
> I am having problems with MIME encoding and decoding.  The following
> code is supposed to take a binary file, base64-encode it, then decode
> the encoded form, producing a copy of the original file.  Can anyone
> explain to me why it doesn't work?
> 
> ------------------------------------------
> import mimetools
> 
> # open a binary file for reading
> input_file = open("any.gif", 'r')

the comment doesn't match the code. try:

input_file = open("any.gif", 'rb')

('b' for binary, see the library reference for
more information on the "open" function).

</F>




From philh at vision25.demon.co.uk  Tue Jul 13 17:04:44 1999
From: philh at vision25.demon.co.uk (Phil Hunt)
Date: Tue, 13 Jul 99 21:04:44 GMT
Subject: XML-RPC
Message-ID: <931899884snz@vision25.demon.co.uk>

Does anyone know if an XML-RPC server exists for Python?

(XML-RPC is a protocol for making remote procedure calls, where
the parameters are sent in XML. The XML-RPC spec is at www.xmlrpc.com
There's a python XML-RPC client at www.pythonware.com/products/xmlrpc/
but I'm not aware of a Python server).

-- 
Phil Hunt....philh at vision25.demon.co.uk




From neelk at brick.cswv.com  Sun Jul 18 19:12:30 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 18 Jul 1999 18:12:30 -0500
Subject: Strange side effect from reading a list of numbers from argv
References: <O3sk3.6047$Sg6.52046@news1.rdc1.sdca.home.com>
Message-ID: <7mtn0u$8sr$1@brick.cswv.com>

In article <O3sk3.6047$Sg6.52046 at news1.rdc1.sdca.home.com>,
Harold Gottschalk <heg at softbrew.com> wrote:
>I just wrote this program to become familiar with python and I do not
>understand why it works with a list I create with in the program and not
>when I use sys.argv[1:].
>
>I am sure there is some nuance I am missing, but an explanation to its
>behavior would be helpful.

The elements of sys.argv are *strings*, and strings compare 
ascii-betically:

>>> "3" > "2"
1

>>> "3" > "20"
1

Just like "bat" > "ant", and "bat" > "anthropologist". 

To convert elements of sys.argv (or any string) into numbers, use the
int() or float() built-in functions, depending on whether you want an
integer or floating-point number.

>>> int("3") > int("2")
1

>>> int("3") > int("20")
0

More like what you wanted, right? 

It would be nice if the comparison between your number and the string
had raised an exception, but that's not how Python comparisons work --
the language reference documents this in Section 5.9:

  The operators "<", ">", "==", ">=", "<=", and "!=" compare the values
  of two objects. The objects needn't have the same type. If both are
  numbers, they are coverted to a common type. Otherwise, objects of
  different types always compare unequal, and are ordered consistently
  but arbitrarily.

Nonetheless, this is IMHO a misfeature. We'll just have to live with
a language that's 95% perfect. :)


Neel



From kperacles at geocities.com  Sat Jul 10 21:38:54 1999
From: kperacles at geocities.com (Karl Putland)
Date: Sat, 10 Jul 1999 20:38:54 -0500
Subject: Python and COM?
Message-ID: <7m9to7$6qg$82@rks1.urz.tu-dresden.de>

For a COM server implemented in python, can a type library be generated?

I would like to be able to distribute some of the work I've done, but most of the people that would use it would be using VB/VBA so I would like expose the Python objects as COM objects and generate a type library so that they can take advantage of the QuickInfo and parameter hints.

TIA

Karl Putland
kperacles at geocities.com


From skip at mojam.com  Thu Jul 29 14:55:18 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 29 Jul 1999 13:55:18 -0500 (CDT)
Subject: shelve portability problem
In-Reply-To: <7nq6a7$vka$1@nnrp1.deja.com>
References: <eZUl3.1080$wm4.135734@news.uswest.net>
	<7nirjd$n52$1@m2.c2.telstra-mm.net.au>
	<7nq6a7$vka$1@nnrp1.deja.com>
Message-ID: <14240.41637.859078.733359@dolphin.mojam.com>

    nadim> I'm having a problem moving a Python shelve ODB from Solaris to
    nadim> Linux.
	...
    nadim> What seems to be happening is that Solaris is storing the data in
    nadim> an old DBM style database, and Linux is using a new gdbm format.

    nadim> I know I can work around this by using pickle and a specific DB
    nadim> module, but does anyone know how I can get it working using the
    nadim> shelve module?

Make sure the same database libraries are available on the two platforms.
Apparently, gdbm isn't available on your Solaris system.  Shelve uses
anydbm, which runs through a canned list of potential modules.  It uses the
first one it finds.  It might be possible to also fiddle the order of the
anydbm._names list to guarantee whatever you have in common on the two
platforms is found first.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From olipt at mayo.edu  Thu Jul 29 05:44:22 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Thu, 29 Jul 1999 04:44:22 -0500
Subject: passing data to Fortran in Python
In-Reply-To: <7nov55$5g6$1@nnrp1.deja.com>
References: <7nov55$5g6$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.10.9907290436070.12075-100000@us2.mayo.edu>

There are two solutions that I know of.  This month two programs were
released that can autogenerate C-extension modules to call (fairly)
arbitrary Fortran code.  In no particular order they are.

PyFort by Paul Dubois
  ftp://ftp-icf.llnl.gov/pub/python/dubois {pyfort.tar, pyfort.pdf}

f2py  by  Pearu Peterson
  http://koer.ioc.ee/~pearu/python/f2py/ {f2py.py}

   This is version 0.61 to get version 0.71 (what I use) you need to check
    it out from the CVS server.  Pearu can help you, there.

You should look at the matrix-sig archives for the month of July to see
posts regarding them.


- Travis






From bduncan1 at san.rr.com  Wed Jul 21 17:07:59 1999
From: bduncan1 at san.rr.com (Bruce Duncan)
Date: Wed, 21 Jul 1999 14:07:59 -0700
Subject: ISP CGI  Python Support
Message-ID: <BFql3.4593$2D1.40042@newsr2.san.rr.com>

Greetings,

I would like to see a short document that I can send to my Web hosting ISP
to convince them to support CGI programming in Python.  Some ISP's do,
but others don't.  Because this may be of common concern, perhaps such
a Python "pitch" can be stored at python.org in addition to a list of
providers
that do support Python CGI.

Comments,

-bsd-

bsd at san.rr.com





From nick.bower at ssec.wisc.edu  Sat Jul 10 19:12:42 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sat, 10 Jul 1999 18:12:42 -0500
Subject: Simple Qu: Changing directory
Message-ID: <7m9to1$6qg$75@rks1.urz.tu-dresden.de>

I feel really dumb asking this question, but how do I change the current
working directory?  For browsing the file system I mean.

I've found exec commands, but these would surely be device dependent and
there doesn't seem to be anything obvious in the os or os.path
documentation.  i looked through the tutorials too.

thanks in advance, nick





From gherron at aw.sgi.com  Wed Jul 21 20:45:43 1999
From: gherron at aw.sgi.com (Gary Herron)
Date: Thu, 22 Jul 1999 00:45:43 GMT
Subject: Tkinter question
References: <37965ED6.A4238A27@anisci.com>
Message-ID: <379669B7.BF4EDB37@aw.sgi.com>

Joel Davis wrote:
> 
> I've got a tkinter canvas, and I want to get it's width.... I call
> 
> w = int( thecanvas.cget("width") )
> 
> this works fine.. except when I resize the window, this still returns
> the width that the canvas was when it was created.. the canvas is
> getting resized (it has a visible background color)..
> 
> any ideas?
> 
> joel
> 
> --
> ----------------------------------------------------------------
>                                                 joeld at anisci.com

Use

  thecanvas.winfo_width()

This works for any widget, not just canvas.


Here's a list of all the winfo_... methods, taken from Tkinter.py:

	def winfo_atom(self, name, displayof=0):
	def winfo_atomname(self, id, displayof=0):
	def winfo_cells(self):
	def winfo_children(self):
	def winfo_class(self):
	def winfo_colormapfull(self):
	def winfo_containing(self, rootX, rootY, displayof=0):
	def winfo_depth(self):
	def winfo_exists(self):
	def winfo_fpixels(self, number):
	def winfo_geometry(self):
	def winfo_height(self):
	def winfo_id(self):
	def winfo_interps(self, displayof=0):
	def winfo_ismapped(self):
	def winfo_manager(self):
	def winfo_name(self):
	def winfo_parent(self):
	def winfo_pathname(self, id, displayof=0):
	def winfo_pixels(self, number):
	def winfo_pointerx(self):
	def winfo_pointerxy(self):
	def winfo_pointery(self):
	def winfo_reqheight(self):
	def winfo_reqwidth(self):
	def winfo_rgb(self, color):
	def winfo_rootx(self):
	def winfo_rooty(self):
	def winfo_screen(self):
	def winfo_screencells(self):
	def winfo_screendepth(self):
	def winfo_screenheight(self):
	def winfo_screenmmheight(self):
	def winfo_screenmmwidth(self):
	def winfo_screenvisual(self):
	def winfo_screenwidth(self):
	def winfo_server(self):
	def winfo_toplevel(self):
	def winfo_viewable(self):
	def winfo_visual(self):
	def winfo_visualid(self):
	def winfo_visualsavailable(self, includeids=0):
	def winfo_vrootheight(self):
	def winfo_vrootwidth(self):
	def winfo_vrootx(self):
	def winfo_vrooty(self):
	def winfo_width(self):
	def winfo_x(self):
	def winfo_y(self):

-- 
Dr. Gary Herron <gherron at aw.sgi.com>
206-287-5616
Alias | Wavefront
1218 3rd Ave, Suite 800, Seattle WA 98101



From ivanlan at callware.com  Thu Jul  8 00:18:20 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Thu, 8 Jul 1999 04:18:20 GMT
Subject: [PSA MEMBERS] Using (physical) units and quantities
References: <1280719924-72790315@hypernet.com>
Message-ID: <3784268C.340CAD46@callware.com>

Hi All--

Gordon "The Great and Terrible" McMillan wrote:
> 
> Alexander V. Voinov wrote:
> 
> > Please explain me this statement: Are there any versions of Windows
> > with case-censitive filesystems? FAT32? Is it possible to switch
> > this on and off for the existing versions: NT 4.0, 95, 98?
> 
> NTFS (and, I think, FAT 32) are case preserving but case insensitive.

Of sorts.  Bill will capitalize filenames and directory names at seeming
random.  If you put capitals in the middle of names, they'll be
preserved.  Mostly.  Running Bill95 on FAT16 makes it act like
FAT32/NTFS, i.e., if you insist enough times you can get the case
preserved.  Sometimes you can keep the case intact when you transfer to
Unix and back.  Sometimes you can't.

> FAT 16 doesn't even preserve case. Command.com (16 bit) trashes case
> (and long file names), while Cmd.exe (32 bit) doesn't.
> 

OS/2 (and I suspect) OS/3 (HPFS) act like FAT32 etc, but a lot more
consistently.  If anyone could stand to run OS/3, we could find out.

> NT has a POSIX (har har) subsystem which will give you case
> sensitivity. But none of the tools work, so you're dead in the water.
> 

Oh, I didn't know that.  So, how does anyone know it exists?  Rumour? 
It's documented but not implemented?  You mean people actually *read*
MSDN docs????

<not-worth-the-bits-they're-printed-on>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From hanche at math.ntnu.no  Wed Jul  7 17:12:33 1999
From: hanche at math.ntnu.no (Harald Hanche-Olsen)
Date: 07 Jul 1999 23:12:33 +0200
Subject: Emacs: highlighting "self" tweak
References: <7luqpa$3s9$1@nnrp1.deja.com> <19990707133146.F3AD4301CF2@bireme.oratrix.nl>
Message-ID: <pcoiu7w17we.fsf@thoth.math.ntnu.no>

+ Sjoerd Mullender <sjoerd at oratrix.nl>:

| You can use this piece of magic in your .emacs file (tested on XEmacs
| 21.1):
| 
| (add-to-list 'after-load-alist
| 	     (cons "python-mode"
| 		   '((add-to-list (quote python-font-lock-keywords)
| 				  (quote ("\\b\\(self\\)[ .,]?"
| 					  1 font-lock-type-face))))))

which could be written more briefly as

(eval-after-load 
 "python-mode"
 '((add-to-list
    'python-font-lock-keywords
    '("\\b\\(self\\)[ .,]?" 1 font-lock-type-face))))

-- 
* Harald Hanche-Olsen     <URL:http://www.math.ntnu.no/~hanche/>
- "There arises from a bad and unapt formation of words
   a wonderful obstruction to the mind."  - Francis Bacon



From Darrell at p98.f112.n480.z2.fidonet.org  Thu Jul  1 07:14:38 1999
From: Darrell at p98.f112.n480.z2.fidonet.org (Darrell)
Date: Thu, 01 Jul 1999 12:14:38 +0100
Subject: Installing Python/XML on Windows?
Message-ID: <0000162e@bossar.com.pl>

From: "Darrell" <news at dorb.com>

Check the doc for the details but in short.
Use site.pth which for me on windows is in my python directory.
It looks something like this.

d:/python/site-packages
d:/python/site-packages/other
d:/python/site-packages/Aycock
d:/python/site-packages/kj

In this same vein I use this to avoid adding each one of these to my system
path.

#!/usr/bin/env python
"""
The .py you want to run is in the python path but not
in your shell path.

I know everyone but me already has this problem solved
in a better way than this. If so please show me the way.

Assuming this is runIt.py then
runIt.py someScript.py arg1 arg2
"""
import sys
import imp

fp, pathname, description = imp.find_module(sys.argv[1])
sys.argv=sys.argv[2:]
md=imp.load_module('__main__', fp, pathname, description)



--
--Darrell
Jonathan W Hendry <jhendry at shrike.depaul.edu> wrote in message
news:377bab1c at news.depaul.edu...
> Does any documentation exist on this? How do I set up the
> paths so that Python can use all the various components of
> the package?
>
> Thanks,
>
> - Jon




From Brian at digicool.com  Fri Jul 30 10:23:49 1999
From: Brian at digicool.com (Brian Lloyd)
Date: Fri, 30 Jul 1999 14:23:49 GMT
Subject: NT select.select?
Message-ID: <613145F79272D211914B0020AFF6401914DC19@gandalf.digicool.com>

> Is there some low limit on maximum number of sockets you can 
> have in the
> Python-NT's select call? A program that happens to work 
> perfectly on Linux
> seems to die on NT around 64(?) sockets to the 'too many file 
> descriptors
> in call' error.
> 
> Any portable ways to bypass it?
> 
> -Markus

Hi Markus,

It turns out that NT has a default 64 fd limit on arguments to
select(). The good news is that you can actually bump the limit up 
to whatever number you want by specifying a define when compiling 
python15.dll.

If you have the ability to rebuild your python15.dll, you can add
the define:

FD_SETSIZE=1024

to the preprocessor options for the python15 project to raise the
limit to 1024 fds.

The default 64 fd limit is too low for anyone trying to run
an async server that handles even a modest load, so I've 
submitted a bug report to python.org asking that the define
above find its way into the next python release...


Brian Lloyd        brian at digicool.com
Software Engineer  540.371.6909              
Digital Creations  http://www.digicool.com 





From darko.cvagic at arcus.tel.hr  Wed Jul 21 11:42:55 1999
From: darko.cvagic at arcus.tel.hr (Darko Cvagic)
Date: Wed, 21 Jul 1999 17:42:55 +0200
Subject: gadfly RDBMS ???
References: <Pine.LNX.4.04.9907211629470.12211-100000@emerald.netskate.ru> <3795C21F.A0F13648@onera.fr>
Message-ID: <3795EA7F.4A452490@arcus.tel.hr>


Marc POINOT wrote:
> 
> Oleg Broytmann wrote:
> 
> > On Wed, 21 Jul 1999, Darko Cvagic wrote:
> > > What does it means that GADFLY is Relational ?
> >
> >    Because Gadfly is almost Yet Another SQL Engine (not that I am against
> > Gadfly or SQL).
> >
> > > How can I make relations between tables in database?
> >
> > SELECT t1.bor, t1.belt, t2.der
> >    FROM table1 t1, table2 t2
> > WHERE t1.plux = t2.flax
> >
> 
> SQL is a query langage, OSQL is an extension to OODBMS (Object oriented).
> 
> The "relationnal" model is an old model of data [1] but it as a theorical
> background
> with set theory, and it is widely spread in the industry.
> Say you have a set of tables, each table can be viewed as
> the subset of the cartesian product of to sets.
> 
> For more info, I think you can read some good books about
> DBMS history, such as [2].
> 
> [1] The Entity-Relationship model - toward a unified view of data
> P.P.-S. Chen, MIT, Readings in Database systems, M.Stonebraker,
> Morgan Kaufmann, 1988
> [note: paper is from ACM 1976]
> 
> [2] Principle of database systems - J. D. Ullman, Computer
> science press1982
> 
> Marcvs [alias Hum, I'm quite pleased about this pedantic and obscure answer. ]

Thanks, I think that I'm exchange terms like RELATION and LINK
So, Gadfly IS RELATIONAL, but I can't make something like:
	CREATE LINK order_customer
	FOREIGN KEY order(cno)
	REFERENCES customer
	ON DELETE RESTRICT

I can't do it in gadfly, am I wright?  
What is name for construction like this? 
I was thought that it is RELATIONAL , I was wrong!



From fredrik at pythonware.com  Mon Jul 26 17:59:29 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 21:59:29 GMT
Subject: Python and Red Hat Linux 6
References: <379CD547.22A10A50@postoffice.pacbell.net>
Message-ID: <00ab01bed7b2$24db1280$f29b12c2@secret.pythonware.com>

Rob <vaton at postoffice.pacbell.net> wrote:
> When programming in Perl or Python, I am using a header like the
> following on an executable file to cause the shell (bash or csh) to run
> the scripts as Perl/Python files:
> #!/usr/bin/perl   for Perl   and #!/usr/bin/python   for Python. I've
> checked and the compilers are located in the appropriate directories. I
> keep getting this when I run scripts:
> bash: <script name>: command not found.

did you try:

$ ./scriptname

if this works, check your PATH settings (having "."
in the path is usually considered a bad idea, but it's
your computer ;-)

</F>




From markusk at bidra241.bbn.hp.com  Fri Jul 23 10:54:33 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 23 Jul 1999 16:54:33 +0200
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have code.)
References: <000601bed335$893916e0$a72d2399@tim> <p5hfmymh9z.fsf@bidra241.bbn.hp.com> <Pine.LNX.4.10.9907211223270.11985-100000@atrus.jesus.cam.ac.uk> <p5n1woncm0.fsf@bidra241.bbn.hp.com> <7n8fnr$2al$1@vvs.superst.iae.nl> <p5k8rrn5o0.fsf@bidra241.bbn.hp.com> <379854CD.8B953362@jumpnet.com>
Message-ID: <p57lnrmn6e.fsf@bidra241.bbn.hp.com>

Tim Olson <tim at jumpnet.com> writes:

> Markus Kohler wrote:
> > 
> > Sorry my Smalltalk example is wrong.
> > I must have coded for to long in this poor language called C ...
> > I must look like this :
> > 
> > do: aBlock
> >  a do:[:each | aBlock value: each]
> >  b do:[:each | aBlock value: each]
> > 
> > more complicated but still very easy to understand.
> 
> No, your original code was fine:
> 
> 	do: aBlock
>  		a do: aBlock. 
>  		b do: aBlock. 
> 
> No need to wrap the incoming block in yet another block.
> 
> You could also write it as:
> 
> 	do: aBlock
> 		a , b do: aBlock.
> 

Ah that's cool, 
I didn't have the time to try it out ...
I saw code that used the #value: message
but, of course it has to be only used if you want to do addiotional
things for each element. 

It seems to be  really time to get back to Smalltalk for me ...

Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From Gordon.McMillan at p98.f112.n480.z2.fidonet.org  Thu Jul  1 04:01:02 1999
From: Gordon.McMillan at p98.f112.n480.z2.fidonet.org (Gordon McMillan)
Date: Thu, 01 Jul 1999 09:01:02 +0100
Subject: Mailbox, tell and read
Message-ID: <00001624@bossar.com.pl>

From: "Gordon McMillan" <gmcm at hypernet.com>

Paul Prescod wrote:

> Gordon McMillan wrote:
> > 
> > How 'bout opening in binary, and munging line endings yourself just
> > before returning the message? 
> 
> First, are there platforms other than DOS/Windows/Win32 where there
> is a difference between text and binary mode?

There's a theoretical difference on *nix - otherwise, they wouldn't 
have made up the distinction!

> I didn't want to munge
> things myself because I'm not sure if what is the "right" munge on
> the Mac, for instance. Can I just do a search and replace for CR/LF
> -> CR on all platforms, all of the time?

On Windows, Notepad (and the underlying default text widget) require
\r\n, but newer widgets / editors will be happy with \n. On *nix,
most tools will show a \r as a noise character. No idea how
forgiving Mac tools are.

AFAIK, we have
 *nix -> \n
 Windows -> \r\n
 Mac -> \r
 
However, since you're dealing with messages created who-knows-where 
and run through who-knows-what transformations, you may well have to 
munge no matter what.


- Gordon




From nathan at islanddata.com  Wed Jul 28 14:36:03 1999
From: nathan at islanddata.com (Nathan Clegg)
Date: Wed, 28 Jul 1999 11:36:03 -0700 (PDT)
Subject: global and import
Message-ID: <XFMail.990728113603.nathan@islanddata.com>

I'm having a little trouble understanding how global and import interact. 
Currenly I have a script that includes a couple of global variables that
allow for a small amount of communication between multiple threads and
signal handlers.  I would like to move some of the functions to separate
modules for several reasons, but need to maintain just the two global
variables between them.  Can this be done?


----------------------------------
Nathan Clegg
 nathan at islanddata.com





From greg.ewing at compaq.com  Wed Jul 28 17:16:11 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 29 Jul 1999 09:16:11 +1200
Subject: String2List: odd function
References: <199907272152.QAA01065@cro.ots.ac.cr>
Message-ID: <379F731B.C2C39989@compaq.com>

Isidor wrote:
> 
> However, I suspect that this use is not among the common uses for
> which exec and eval were designed.

Using eval for something like this is probably overkill.
Eval and exec will accept *any* valid Python expression
or statement, respectively - which can do some nasty
things to the internals of your program if you're not
careful what you feed into them!

This might be quite all right, or it might not,
depending on how bad the consequences will be if
the program blows up due to someone putting in
weird (or even malicious) input.

So there is a use for the other, more laborious
solutions that have been posted. They have the
advantage that they will only accept a certain
subset of Python expressions, and can't cause any
damage if the input doesn't conform to that
subset.

Greg



From jmaskell at saskpublishers.sk.ca  Tue Jul 20 12:28:41 1999
From: jmaskell at saskpublishers.sk.ca (Jason Maskell)
Date: Tue, 20 Jul 1999 10:28:41 -0600
Subject: Python suitability questions
Message-ID: <0CCE781E44F2C872.F528377A62B41DCA.73CAA63AD6E79314@lp.airnews.net>

Well, after a day or two of reading lots of newsgroup posts and
documentation, I'm still not quite certain if Python is suitable for the
project I have in mind. I'm rather reluctant to jump in and start fiddling
with it without having some idea whether it's what I'm looking for or not. A
quick outline of the project:

A collectible card game framework. As some may know, CCG's are rather
complex, so to get the flexibility I need, I'm going to need to have the
non-interface/graphics stuff be a scripting language. What I was envisioning
was a model where all cards had base classes that were derived for specific
cards (for the Magic enabled: Swamps and Forests would both be derivations
of Basic Lands). And of course I need some sort of internal messaging
system, and the scripting language has to be comfortable being completely
embedded in my C++ framework, and having really no system access at all. I
know that Python can handle the OO stuff, but is there any kind of internal
object messaging, or do I roll my own? How flexible can I lay out my source,
is it as stiff as say, C?

Any help would be appreciated. I'm fairly convinced that it's what I'm
looking for, I just need some outside confirmation. ;>

Jason






From lvirden at cas.org  Mon Jul  5 06:51:07 1999
From: lvirden at cas.org (lvirden at cas.org)
Date: 5 Jul 1999 10:51:07 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <7li9t6$j2k$1@srv38.cas.org> <30CF6C97EED47EAD.8A8612E26FB53A0A.B0666516E77A0F15@lp.airnews.net> <377CDFD7.3103B1AE@equi4.com>
Message-ID: <7lq2mr$ob3$1@srv38.cas.org>

According to Jean-Claude Wippler  <jcw at equi4.com>:
:Sales != installed base.
:Also: Sales of one vendor vs. combined sales of N.

On the other hand, 

Previous Sales != installed base

either.  Homes tend to accumulate machines (I know we do - we've currently
got 4 old machines and are about to inherit someone else's cast off machine
as number 5...).

Businesses on the other hand tend to replace machines in varying patterns.
Some accumulate, some stagnate, some have rapid turnover, etc.  So
it's going to be pretty tough to determine installed base.

The bottom line is the original poster claimed that "macs are not very
common" - meaning I assume that in his sphere of operation he seldom knows
of anyone using a Mac.  Others have varying differing experiences.
It's up to people who are using a platform to let the software producers
(whether free or commercial) know if they need software on their platform.

-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.



From tim at jumpnet.com  Fri Jul 23 07:41:01 1999
From: tim at jumpnet.com (Tim Olson)
Date: Fri, 23 Jul 1999 07:41:01 -0400
Subject: Loops and stuff (was Re: Python and Boehm-Demers GC, I have code.)
References: <000601bed335$893916e0$a72d2399@tim> <p5hfmymh9z.fsf@bidra241.bbn.hp.com> <Pine.LNX.4.10.9907211223270.11985-100000@atrus.jesus.cam.ac.uk> <p5n1woncm0.fsf@bidra241.bbn.hp.com> <7n8fnr$2al$1@vvs.superst.iae.nl> <p5k8rrn5o0.fsf@bidra241.bbn.hp.com>
Message-ID: <379854CD.8B953362@jumpnet.com>

Markus Kohler wrote:
> 
> Sorry my Smalltalk example is wrong.
> I must have coded for to long in this poor language called C ...
> I must look like this :
> 
> do: aBlock
>  a do:[:each | aBlock value: each]
>  b do:[:each | aBlock value: each]
> 
> more complicated but still very easy to understand.

No, your original code was fine:

	do: aBlock
 		a do: aBlock. 
 		b do: aBlock. 

No need to wrap the incoming block in yet another block.

You could also write it as:

	do: aBlock
		a , b do: aBlock.

which might be even clearer.

	-- Tim Olson



From neelk at brick.cswv.com  Sat Jul 17 10:09:05 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 17 Jul 1999 09:09:05 -0500
Subject: Python 1.5.2 + Linux 2.0.36 == Bad Combo?
References: <slrn7ou0ru.3sp.kas@titan.magnetic-ink.dk>
Message-ID: <7mq2q1$7q8$1@brick.cswv.com>

In article <slrn7ou0ru.3sp.kas at titan.magnetic-ink.dk>,
Klaus Alexander Seistrup <kas at maps.magnetic-ink.dk> wrote:
>
>  *  DejaGrabber
>     -- The program connects without problem but seems to only retrieve
>     headers, not the full articles.
>     Haven't contacted the author yet.

I wrote DejaGrabber, and when I looked at the documentation, I think
you might have been bitten by a bug in DejaGrabber's documentation
rather than a bug in your system. 

To look at an article, you need to use the Article.message_text() 
method -- the __str__ method simply returns the header.

The corrected version of the example in the docstring should look 
like this:

  >>> from DejaGrabber import *
  >>> d = DejaGrabber(group='comp.lang.python', author='Tim Peters')
  >>> l = d.get_messages(4)
  >>> print l
  [99/06/23  30 comp.lang.python          RE: Python Exes!      Tim Peters
  , 99/06/18  28 comp.lang.python RE: Newbie: Truth values (th      Tim Peters
  , 99/06/18  28 comp.lang.python       RE: DATE ARITHMETIC      Tim Peters
  , 99/06/18  28 comp.lang.python     RE: NameError Problem      Tim Peters
  , 99/05/18  25 comp.lang.python     RE: while (a=b()) ...      Tim Peters
  ]
  >>> print l[0].message_test()
  [SunitJoshi]
  > Could someone please point me to the URL where Gordon Mcmillians,
  > installer is located.
  
  Start at
  
      http://www.mcmillan-inc.com/python.html
  
  would-be-more-specific-but-the-server-appears-to-be-resting-ly y'rs  - tim

Before, I had "print l[0]" -- this would invoke the __str__ method, 
which would just print out the header. 

Please accept my apologies for any inconvenience. Also, I've written a
command-line driver for DejaGrabber, which might be more useful. It has
the very clever name "newssearch.py", and you can get it at:

    www.sff.net/people/neelk/free-software/newssearch.py


Neel



From simngolda at voyager.net  Fri Jul  2 20:16:13 1999
From: simngolda at voyager.net (Sim & Golda Zacks)
Date: Fri, 02 Jul 1999 20:16:13 -0400
Subject: Python supporting web hosting?
Message-ID: <7ljkim$pmo@newsops.execpc.com>

I am looking for a web hoster that supports Python 1.5 and also has a
database with a python interface (MySQL, for example). If you know of
one with affordable pricing please let me know. Currently I'm with an
ISP with Python 1.4 and they say they'll upgrade but not for some time,
and they have MySQL but only with the PERL interface installed, and they
will think about the Python one. And I'm getting sick of a lot of other
things. I just need a good hosting site, that preferably has anonymous
reseller accounts.

Thanks
Sim




From gmcm at hypernet.com  Fri Jul 23 14:37:39 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 23 Jul 1999 13:37:39 -0500
Subject: Python documentation available as ZIP files
In-Reply-To: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
Message-ID: <1279371240-26729947@hypernet.com>

Fred L. Drake wrote:

> ... but be aware that, even with maximal
> compression, the HTML ZIP archive is substantially larger than the
> .tgz equivalent.  This is likely caused by the ZIP format having two
> directory entries for each file instead of one; with a lot of small
> files, this can make a difference.

Nah, it's because a zip file is (conceptually) a tar of a bunch of 
gzips, while a tgz is a gzip of a tar. Compression ratios get better 
on longer inputs. But "random" access is a lot easier on the former.

- Gordon



From ivnowa at hvision.nl  Tue Jul 13 17:18:40 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Tue, 13 Jul 1999 23:18:40 +0200
Subject: Recursive method
References: <378B499B.4301F893@eml.villa-bosch.de>
Message-ID: <378BAD30.7D1A5196@hvision.nl>

Ralph Gauges wrote:
> 
> Hello,
> 
> I am trying to generate a JTree (in JPython) out of some
> Data I have. The data I have is structured in classes which
> have a list of classes that have a list of classes and so
> on. Now I want to call the makenodes method with the top
> level class and have the function follow the tree
> recursively (see code below). Unfortunately all I get is a
> 'NameError' for makenodes, when the function is supposed to
> call itself. This seems to be a problem with the namespace,
> but I have no idea, how to solve this. Any help would be
> appreciated.
> Thanks a lot in advance
> 
> Ralph
> 
>         def makeTree():
>                 .
>                 .
>                 .
>             def makenodes(C):
> 
> N=javax.swing.tree.DefaultMutableTreeNode(C.Name)
>                 Children=C.getChildren()
>                 for Child in Children:
>                     N.add(makenodes(Child))
>                 return N
> 
>             for Class in M.Classes:
>                 ClassNode.add(makenodes(Class))
>                  .
>                  .
>                  .

You have a nested function here. (I say function, because it does not
look like a method... it's missing the self
argument! But maybe you didn't post the real code.) 

Anyway, makenodes is nested within makeTree, and that's the reason it
won't find itself... scope in Python is not
like C's or Pascal's. If makenodes were 'visible' from somewhere (e.g.
it would be global, accessible in a module, 
or directly in the class definition, then it would be no problem.
Recursive functions are usually no problem,
except in this case. *frown*  (This might be a FAQ... i'm too lazy to
look it up though. :)

So, to solve this, you could try a different approach... maybe

(you said it's a method, right?)

class Blahblah:
    ...
    def makeTree(self):
         ....
    def makenodes(self, C):
         ....

I don't know if this is possible in your case, but it could be a
solution.

Good luck,



From adv at langdale.com.au  Sat Jul 24 22:44:34 1999
From: adv at langdale.com.au (Arnold de Vos)
Date: Sun, 25 Jul 1999 02:44:34 GMT
Subject: HTTP 1.1 anyone?
References: <37997331.0@pink.one.net.au> <004301bed5b2$769dca60$0301a8c0@mnot.net>
Message-ID: <002701bed647$a0cca1e0$ef01a8c0@lakes.com.au>

Hi,

I'm mainly interested in the persistent connections and request pipelining.

- Arnold



----- Original Message -----
From: Mark Nottingham <mnot at pobox.com>
To: Arnold de Vos <adv at langdale.com.au>
Cc: <python-list at cwi.nl>
Sent: Saturday, July 24, 1999 6:56 PM
Subject: Re: HTTP 1.1 anyone?


> Which part(s) of it? Are you looking for just one feature, or everything
> (everything being a pretty tall order)?
>
>
> ----- Original Message -----
> From: Arnold de Vos <adv at langdale.com.au>
> Newsgroups: comp.lang.python
> To: <python-list at cwi.nl>
> Sent: Saturday, July 24, 1999 6:02 PM
> Subject: HTTP 1.1 anyone?
>
>
> > Is anyone working on an HTTP 1.1 implementation in Python?   I am
> > considering creating variants of BaseHTTPServer and httplib to handle
the
> > 1.1 protocol.  But I would not like to re-invent the wheel.
> >
> > Thanks,
> > Arnold
> >
> > ---
> > Arnold deVos
> > Langdale Consultants
> > adv at langdale.com
> >
> >
> >
> >
> >




From fredrik at pythonware.com  Tue Jul 27 16:33:02 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 27 Jul 1999 20:33:02 GMT
Subject: [Tkinter] Inheriting from Toplevel
References: <7nl2h8$keo$1@nnrp1.deja.com>
Message-ID: <001d01bed86f$39de44f0$f29b12c2@secret.pythonware.com>

dwolff567 at my-deja.com wrote:
> I am having some fairly weird problems with my Tkinter app.  I am
> spawning a new Toplevel window where I display a graph in a
> canvas instance.  However, when I send the data to the widget to
> be plotted, the whole thing freezes.

I'd look for a widget that is created without a master, and
displayed using a call to "grid"...

if you don't use grid at all, you probably have to post more
code...

</F>




From m.faassen at vet.uu.nl  Thu Jul 22 10:09:58 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 22 Jul 1999 14:09:58 GMT
Subject: Fw: ANNOUNCE: Mailing list for Australian Python Users
References: <007201bed42e$c4316c30$f29b12c2@secret.pythonware.com> <wkogh56j2o.fsf@ifi.uio.no>
Message-ID: <7n78nm$lic$1@news.accu.uu.nl>

Lars Marius Garshol <larsga at ifi.uio.no> wrote:

> * Fredrik Lundh
> | 
> | python-se, n?gon?  Link?ping?  :-)

> Tjah. Si Oslo, s? kan vi bli en del. :-)

Utrecht is veel beter denk ik, Python-NL en zo..

An-large-percentage-of-pythoneers-speaks-Dutch-anyway-ly yours,

Martijn




From fredrik at pythonware.com  Wed Jul 28 17:36:22 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Jul 1999 21:36:22 GMT
Subject: 1.5.2 or: else:
References: <1279007788-11957995@hypernet.com> <379F6D43.3531FE0@compaq.com>
Message-ID: <003a01bed942$b64e3440$f29b12c2@secret.pythonware.com>

Greg Ewing <greg.ewing at compaq.com> wrote:
> Not if you go about it the right way. I always
> put such loops in a procedure of their own,
> and use return:
> 
>   def find_the_holy_grail(stuff):
>     for x in stuff:
>       if is_holy_grail(x):
>         return x
>     return None

hey, didn't your MOM tell you to only use a
SINGLE return statement per function, and
put that at the END of the function body? ;-)

> I wouldn't mind if there weren't any break - and
> without break, there would be no need for the
> weird else.

well, I think it's time for a break...

</F>




From MHammond at skippinet.com.au  Sat Jul 10 04:54:12 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Sat, 10 Jul 1999 18:54:12 +1000
Subject: idle
References: <1tBh3.875$uZ.2396@news.flash.net>
Message-ID: <7m9tni$6qg$60@rks1.urz.tu-dresden.de>

I believe the lastest version of IDLE can only be found via CVS -
http://www.python.org/download/cvs.html

Mark.

Tim Hamza wrote in message <1tBh3.875$uZ.2396 at news.flash.net>...
>Have there been any updates to idle.  I thought I saw something about a 0.5
>version, but I haven't been able to find anything at python.org.







From casado at ogl.llnl.gov  Fri Jul  2 13:17:38 1999
From: casado at ogl.llnl.gov (Martin Casado)
Date: Fri, 2 Jul 1999 10:17:38 -0700 (PDT)
Subject: C shared object files
Message-ID: <199907021717.KAA03837@ogl.llnl.gov>

  Python Guru's,

   I am trying to get python to dynamically load a C object module, and 
first off I can't figure out how to compile myCollProggy.c into a 
shareable object file, and secondly I was curios if there is a way to check
at runtime if a module is on the system or not.  

  I have made arrangements with the proper authorities and there is a veritable
wealth of Karma points waiting for anyone who can help me :-)  Hehe thanks
very much in advance.

			-Martin 



From fredrik at pythonware.com  Fri Jul 16 10:05:11 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 16 Jul 1999 14:05:11 GMT
Subject: [XML-SIG] sgmllib has problems with dots in tag names
References: <19990716125230.A14994@sz-sb.de> <14223.12495.910694.301072@weyr.cnri.reston.va.us> <19990716153323.A23945@sz-sb.de>
Message-ID: <00c301becf94$386efda0$f29b12c2@secret.pythonware.com>

Andreas Jung <ajung at sz-sb.de> wrote:
> On Fri, Jul 16, 1999 at 09:17:03AM -0400, Fred L. Drake wrote:
> > 
> >   Ok, I've poked at the standard sgmllib a bit to see what the problem
> > is.  The parser is recognizing the start and end tags.  Once
> > recognized, it is looking for the handler methods start_*() / end_*()
> > or do_*().  Since there's a dot in the name, these methods are not
> > defined, and the unknown_*tag() methods are called instead of the
> > handle_*tag() methods.
> 
> 
> >   It should be easy to override the unknown_*tag() methods to use a
> > table-based dispatcher or performs some form or name mangling, then
> > passes known tags through to the handle_*tag() methods or whatever.
> > This seems to be the easiest way to deal with the situation in the
> > short term.
> 
> That's a solution that works with the standard sgmllib from the Python
> distribution. However this solution does not work with sgmllib from 
> xml.parsers. I'm not sure if tag names with dots in their names are valid in XML 
> or not. So this might explain the different behaviour however I don't think
> that's the reason. Maybe I'll find the real reason over the weekend.

iirc, the sgmllib in xml.parsers is an extended version
which uses the sgmlop accelerator, if installed.  it's
based on 1.5.1's standard sgmllib, which did not accept
dots in tag names.  I suppose nobody's gotten around to
fix it...

(btw, the sgmlop accelerator recognizes dots just fine,
so one way to get around this problem is to install the
accelerator...)

</F>




From mal at lemburg.com  Thu Jul 29 09:08:38 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 29 Jul 1999 15:08:38 +0200
Subject: ANN: mxTextTools Package - Version 1.1.0
Message-ID: <37A05256.E18A9380@lemburg.com>

ANNOUNCING:

                   mxTextTools Version 1.1.0

          A Python extension package providing fast text
                 scanning and processing tools


WHAT IT IS:

mxTextTools is an extension package for Python that provides 
* a fast text search type (using a modified Boyer-Moore algorithm)
* a finite state machine for marking and analysing text
  (the Tagging Engine)
* a set of specialized text processing functions
* a set of functions that work with character sets

Applications include parsing structured text, finding and extracting
text (either exact or using translation tables) and recombining
strings to form new text. Everything was implemented having high
performance in mind.

WHAT'S NEW ?

The 1.1.0 release adds several new text processing functions to
the already large set: splitlines(), countlines(), suffix(), prefix(),
str2hex() and hex2str(). Many existing functions/objects were
enhanced to work in a more general context.

All the details are available at:

        http://starship.skyport.net/~lemburg/mxTextTools.html#History

This release also includes a precompiled Windows DLL so that you can
use the package right away. Installing on that platform boils down
to a simple unzip in the \Python\Lib directory.

WHERE CAN I GET IT ?

The full documentation and instructions for downloading and installing
can be found at:

        http://starship.skyport.net/~lemburg/mxTextTools.html

WHAT DOES IT COST ?

It comes with a Python-type license, but is free otherwise.

WHERE CAN I GET SUPPORT ?

I am offering commercial support for this package through
Python Professional Services Inc. (http://www.pythonpros.com).
Look on their support pages for details or contact me directly.

REFERENCE:

<P><A HREF="http://starship.skyport.net/~lemburg/mxTextTools.html>
mxTextTools 1.1.0</A> - fast text processing tools. (29-Jul-1999)

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   155 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From greg.ewing at compaq.com  Tue Jul 13 19:18:13 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 14 Jul 1999 11:18:13 +1200
Subject: My first Tkinter app. Did I do good?
References: <danielt3-0807991903200001@1cust103.tnt9.clearwater.fl.da.uu.net>
Message-ID: <378BC935.F31261DB@compaq.com>

"Daniel T." wrote:
> 
> I would love to hear from experienced Python programmers. Does it look
> like I know what I'm doing?

Looks fine to me. Works, too!

Greg



From amk at mira.erols.com  Fri Jul  9 19:38:32 1999
From: amk at mira.erols.com (A.M. Kuchling)
Date: 9 Jul 1999 23:38:32 GMT
Subject: Python Code Repository?
References: <378645AA.5430A026@msfc.nasa.gov> <19990709163127.A1964@toast.internal>
Message-ID: <7m9tn8$6qg$50@rks1.urz.tu-dresden.de>

On Fri, 9 Jul 1999 16:31:27 -0400, jam <jam at newimage.com> wrote:
>does anyone have any further suggestions or ideas on how we can get this
>rolling?

Michel Pelletier (the one at Digital Creations) was working on a
ZTrove at one point, I believe.  (For information on the whole idea of
Trove, see http://www.tuxedo.org/~esr/trove/).  I don't know what's
happened to ZTrove; anyone from DC care to comment?

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
    "If you had stayed with us, we could have given you life until death."
    "Don't I get that anyway?"
    -- Stheno and Lyta Hall, in SANDMAN #61: "The Kindly Ones:5"




From chad at vision.arc.nasa.gov  Thu Jul 22 22:51:13 1999
From: chad at vision.arc.nasa.gov (Chad Netzer)
Date: Thu, 22 Jul 1999 19:51:13 -0700
Subject: int(long(-sys.maxint-1)) fails on Linux
References: <000001bed49f$c052e820$2c2d2399@tim>
Message-ID: <3797D8A1.E6D95CCB@vision.arc.nasa.gov>

Tim Peters wrote:

> Chad, let's break this down to see where the overflow is coming from.  Try
> this line by line and post your results:
>
> >>> import sys
> >>> sys.maxint
> 2147483647
> >>> -sys.maxint
> -2147483647
> >>> -sys.maxint-1
> -2147483648
> >>> long(-sys.maxint-1)
> -2147483648L
> >>> int(long(-sys.maxint-1))
> -2147483648
> >>>
>

>>> import sys
>>> sys.maxint
2147483647
>>> -sys.maxint
-2147483647
>>> -sys.maxint-1
-2147483648
>>> long(-sys.maxint-1)
-2147483648L
>>> int(long(-sys.maxint-1))
Traceback (innermost last):
  File "<stdin>", line 1, in ?
OverflowError: long int too long to convert
>>>

This is on a Pentium III system running straight Mandrake 6.0.  I think, as Christian
and Oliver pointed out, it is pgcc related.  I tried it on a Mandrake 5.3 system (using
the normal gcc, not pgcc, and it worked correctly)

For kicks, I fired up the debugger, and traced through the PyLongASLong call to see
what was happening.  The problem is at the line:

if ((long)x < 0 && (sign > 0 || (x << 1) != 0))

Even though it should evaluate to 0 (and appears to from the debugger's point of view),
it incorrectly proceeds to the "goto overflow" statement.  I looked at the assembly language
output from a working version and a non-working version (both 1.5.2 on Pentiums), and
found that the working version tends to use "cmpl", "testl", etc. and the non-working version
doesn't.  Here is what I mean, the gist of the code is apparent, even though I'm no wizard at
386 asm (I'll provide some comments):

C code:
   if ((long)x < 0 && (sign > 0 || (x << 1) != 0))
            goto overflow;
    return (long)x * sign;


working asm:
<PyLong_AsLong+140>:  cmpl   $0x0,0xfffffffc(%ebp)                      ; if x<0
<PyLong_AsLong+144>:  jnl    0x806adbc <PyLong_AsLong+172>
<PyLong_AsLong+146>:  cmpl   $0x0,0xfffffff0(%ebp)                      ; if sign > 0
<PyLong_AsLong+150>:  jg     0x806adb8 <PyLong_AsLong+168>
<PyLong_AsLong+152>:  movl   0xfffffffc(%ebp),%esi
<PyLong_AsLong+155>:  leal   0x0(,%esi,2),%eax                               ; x<<0 (?)
<PyLong_AsLong+162>:  testl  %eax,%eax
<PyLong_AsLong+164>:  jne    0x806adb8 <PyLong_AsLong+168>
<PyLong_AsLong+166>:  jmp    0x806adbc <PyLong_AsLong+172>
<PyLong_AsLong+168>:  jmp    0x806adc8 <PyLong_AsLong+184>  ; goto overflow
<PyLong_AsLong+170>:  movl   %esi,%esi                                           ; return (long)x * sign;
<PyLong_AsLong+172>:  movl   0xfffffffc(%ebp),%edx
<PyLong_AsLong+175>:  imull  0xfffffff0(%ebp),%edx
<PyLong_AsLong+179>:  movl   %edx,%eax
<PyLong_AsLong+181>:  jmp    0x806ade4 <PyLong_AsLong+212>

not working:
<PyLong_AsLong+150>: cmpl   $0x0,0xfffffffc(%ebp)
<PyLong_AsLong+154>: jge    0x806c3f2 <PyLong_AsLong+178>
<PyLong_AsLong+156>: cmpl   $0x0,0xfffffff0(%ebp)
<PyLong_AsLong+160>: jg     0x806c3f0 <PyLong_AsLong+176>
<PyLong_AsLong+162>: mov    0xfffffffc(%ebp),%eax
<PyLong_AsLong+165>: add    %eax,%eax                                          ;  x<<1
<PyLong_AsLong+167>: test   %eax,%eax
<PyLong_AsLong+169>: jne    0x806c3f0 <PyLong_AsLong+176>
<PyLong_AsLong+171>: jmp    0x806c3f2 <PyLong_AsLong+178>
<PyLong_AsLong+173>: lea    0x0(%esi),%esi
<PyLong_AsLong+176>: jmp    0x806c400 <PyLong_AsLong+192> ; goto overflow
<PyLong_AsLong+178>: mov    0xfffffffc(%ebp),%edx                     ; return (long)x * sign;
<PyLong_AsLong+181>: imul   0xfffffff0(%ebp),%edx
<PyLong_AsLong+185>: mov    %edx,%eax
<PyLong_AsLong+187>: jmp    0x806c41a <PyLong_AsLong+218>

If nyone who knows more about assembly, I would mind hearing where the problem
is, and to know if it is worth submitting a bug-report.

It sounds as if Oliver's very-soon-to-be-released versions of the packages may have
fixed the problem w/ compiler flags.  I'm anxious to try it out (Oliver, thanks for all
the effort, BTW.  I got impatient waiting for your new releases, and have been making
my own based on you specs and packages...  LOT of work getting everything just right!
I figured you were busy, as we all are, but I'm happy to see that you'll be releasing soon.
I will definately join the python4linux mailing list, as I would like to help out in the
future)

Thanks for the help everyone.

Chad Netzer
chad at vision.arc.nasa.gov





From MHammond at skippinet.com.au  Wed Jul 21 19:39:50 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Thu, 22 Jul 1999 09:39:50 +1000
Subject: ANNOUNCE: Mailing list for Australian Python Users
Message-ID: <7n5ll7$afv$1@m2.c2.telstra-mm.net.au>

I have (well, Chris did - thanks!) created a mailing list for Australian
users of the Python language.  This mailing list should initially be a forum
for anything uniquely of interest to Aussie Python users.  First order of
business should be a contest for which city can get the first informal
Python users meeting together - I think that Melbourne should be able to get
something happening pretty soon :-)

Subscription and other list information can be found at
http://starship.python.net/mailman/listinfo/python-au

Mark.






From gmcm at hypernet.com  Fri Jul  9 13:54:50 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 9 Jul 1999 12:54:50 -0500
Subject: win32 python as a dll?
In-Reply-To: <378609F9.99338FD@enews.nrl.navy.mil>
Message-ID: <1280583297-81012085@hypernet.com>

 Scott Wolford wrote:

> Mark Hammond wrote:

[debug python builds on Windows]
 
> How about:
> python.lib/dll and msvcrt.lib/dll
> python.lib/dll and msvcrtd.lib/dll
> python_d.lib/dll and msvcrt.lib/dll
> python_d.lib/dll and msvcrtd.lib/dll

Debug python linked to non-debug msvcrt makes sense - means you can 
load non-debug extension modules (see below). 

> > It is also worth noting that Gordon pointed this out to you before this
> > particular thread, and you still had the problem.  Im not sure that
> > documenting it would be effective in all cases.
> 
> He said to use the Multithreaded DLL. Unfortunately, msvcrt.dll and
> msvcrtd.dll are *both* the Multithreaded DLL; one is debug the other
> is not. It never occured to me (well, it finally did) that it
> mattered whether msvcrt.dll or msvcrtd.dll were used. I assumed that
> the semantics of the two DLLs were essentially the same, just the
> implementation differed slightly. I'm pretty sure (though I haven't
> tested it, maybe you'll know) that you can mix and match optimized
> and debug code pretty easily under Unix.

To beat a dead horse into an unrecognizable pulp:

That's the whole point - they are *different* dlls. So a file open in 
one will not be found in the other. Kaboom. Nothing to do with 
semantics - it's just where buffers, files and other statics / 
globals live.

You can mix debug and optimized in Windows, too - you'll just end up 
stepping thru unintelligable assembler when you hit non-debug code.

[And for the masochistically curious - the reason the runtime lib is 
specified to the MSVC compiler and not the linker is that the 
compiler generates slightly different code depending on whether the c 
runtime is in a static lib or in a dll.]

- Gordon



From mstenber at cc.Helsinki.FI  Thu Jul 29 03:23:09 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 29 Jul 1999 10:23:09 +0300
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com> <379f06e9@news.xtml.co.uk>
Message-ID: <al8btcvudgi.fsf@sirppi.helsinki.fi>

"Nick Trout" <nick@#spam_me_not#videosystem.co.uk> writes:
> >> Thus I've had great deal of interest in getting the speed up; however,
> >> speeding up dynamic interpreted languages is nontrivial, especially
> >> languages as dynamic as Python.
> Are there any pieces of test code anywhere which could be used as a good
> test for Pythons speed (encompasing all of its language features) so that if
> anyone tinkered with the bytecode execution unit they could measure the
> speed improvement. It would be useful if it also tested correctness to
> ensure no mistakes?!

There is PyStone (comes along with distribution) and PyBench (which I
haven't tried). I myself use a performance testing framework and ton of
small examples to judge whether things speed up or not.

> N

-Markus



From Gordon.McMillan at p98.f112.n480.z2.fidonet.org  Sat Jul  3 22:43:49 1999
From: Gordon.McMillan at p98.f112.n480.z2.fidonet.org (Gordon McMillan)
Date: Sun, 04 Jul 1999 03:43:49 +0100
Subject: Tkinter._test() problem
Message-ID: <000016bf@bossar.com.pl>

From: "Gordon McMillan" <gmcm at hypernet.com>

Per Kistler writes:
> 
> I installed py1.5.2 with tcl/tk.8.1 on SuSE6.1 Linux.

That's your problem. Either drop back to tcl/tk 8.0, or may try 
Dieter Maurer's patches, posted June 1.


- Gordon




From tim_one at email.msn.com  Wed Jul 28 22:27:55 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 28 Jul 1999 22:27:55 -0400
Subject: Memory leak in Tkinter.py ?
In-Reply-To: <379F3B68.4D533F3@weihenstephan.org>
Message-ID: <000c01bed969$f725ddc0$71a22299@tim>

[Dr. Peter Stoehr]
> ...
> The tag_bind method of the Canvas class maintains a dict called
> self._tagcommands to store lists of function-ids returned from the _bind
> call.
> As far as I can see, this information not used by any other method.
> Furthermore, the lists grow with every call of tag_bind and the memory
> is never released.

Your analysis looks good to me, although the memory will be released when
the owning  Canvas instance goes away (which it never did in your test
program).

> ...
> Is it okay to remove the
> 	if sequence and func and res:
> block or did I miss something?

I think removing that block is fine.  The

	_tagcommands = None

at the start of the Canvas class should be removed too, then.

agreeably y'rs  - tim





From esr at thyrsus.com  Fri Jul 30 11:39:31 1999
From: esr at thyrsus.com (Eric S. Raymond)
Date: Fri, 30 Jul 1999 11:39:31 -0400
Subject: Need help with TCP/IP client access from Windows
In-Reply-To: <1278785133-25350622@hypernet.com>; from Gordon McMillan on Fri, Jul 30, 1999 at 08:25:58AM -0500
References: <1278839420-22085362@hypernet.com>; <19990730005211.T7149@thyrsus.com> <1278785133-25350622@hypernet.com>
Message-ID: <19990730113931.D17721@thyrsus.com>

Gordon McMillan <gmcm at hypernet.com>:
> Probably the same directory where you put your script. (The Win31 
> convention of putting all ini files in the Windows directory under 
> <appname>.ini is now, thank goodness, passe).

OK.  Will giving the filename ".\\intriguer" work for this?  And, given
that the filename is going through the C library, will leaving it as 
"./intriguer" work?

(Sorry for the newbie questions, but I have zero Windows-programming
experience.)
-- 
		<a href="http://www.tuxedo.org/~esr">Eric S. Raymond</a>

Never trust a man who praises compassion while pointing a gun at you.



From aa8vb at yahoo.com  Fri Jul  2 09:49:28 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Fri, 2 Jul 1999 09:49:28 -0400
Subject: newbie question about tkFileDialog.py
In-Reply-To: <3BE7847CF88DD211893100A0C9D43E1A0259E1ED@hqmail2.clarify.com>; from Dinesh Shanbhag (c) on Mon, Jun 28, 1999 at 06:01:48PM -0700
References: <3BE7847CF88DD211893100A0C9D43E1A0259E1ED@hqmail2.clarify.com>
Message-ID: <19990702094928.A129582@epa.gov>

Dinesh Shanbhag (c):
 |Did not find anything like tk_getOpenFile in the Tcl/Tk man pages for
 |selecting Directories...
 |
 |Am i missing something obvious ?

I don't think so.  I looked for the same and didn't find it either.

Randall



From nick.bower at ssec.wisc.edu  Fri Jul  9 15:50:21 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Fri, 09 Jul 1999 14:50:21 -0500
Subject: Grail won't work
Message-ID: <3786527C.6230C4D2@ssec.wisc.edu>

I just installed Grail 0.6 to take a look.  I have Py1.5 and Tk8 but it
won't run:

Traceback (innermost last):
  File "d:\grail-0.6\grail.py", line 499, in ?
    main()
  File "d:\grail-0.6\grail.py", line 150, in main
    from Browser import Browser
  File "D:\grail-0.6\Browser.py", line 26, in ?
    _mydir = os.path.dirname(grailutil.abspath(__file__))
  File "D:\grail-0.6\utils\grailutil.py", line 22, in abspath
    if not isabs(path):
NameError: isabs

What does this mean?  Cheers, Nick.




From Robert.Kern at p98.f112.n480.z2.fidonet.org  Fri Jul  2 18:41:41 1999
From: Robert.Kern at p98.f112.n480.z2.fidonet.org (Robert Kern)
Date: Fri, 02 Jul 1999 23:41:41 +0100
Subject: GUI's etc: PyGtk on Windows
Message-ID: <00001696@bossar.com.pl>

From: kernr at mail.ncifcrf.gov (Robert Kern)

On Fri, 2 Jul 1999 13:14:42 -0500, Travis Oliphant <olipt at mayo.edu>
wrote:

>
>Previously I mentioned that Gtk works on any system with an X server,
>which is true.
>
>I also want to point out that Gtk is a cross-platform solution
>because there is a port of the GDK (the underlying windowing calls that
>Gtk uses) for Win32.  You can find a link at http://www.gtk.org
>
>From what I've read it looks like Gtk was intentionally written on top of
>GDK so that a port of GDK would make Gtk available.  The person who wrote
>the GDK port to Win32 did it to get the GIMP to run on Windows.
>There are screenshots of the result available on the site.
>
>So if someone develops mainly for X and just wants to make sure that it is
>not impossible to run on Windows, you can use PyGTK.  In fact a great
>program using GTK can be a strong motivator to get the ports of GDK going.

Last time I tried to compile PyGTK on windows (about a week ago with
the June 21 snapshot), there were some symbols missing from the
libraries.

Soon, perhaps.

>Just some thoughts,
>
>Travis

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From jepler at inetnebr.com  Tue Jul 27 11:10:13 1999
From: jepler at inetnebr.com (Jeff Epler)
Date: Tue, 27 Jul 1999 15:10:13 GMT
Subject: strptime: unpleasant surprise (py 1.5.2 on Linux)
References: <m3aesltwg8.fsf@havenrock.com>
Message-ID: <slrn7prja5.b0h.jepler@craie.inetnebr.com>

On 25 Jul 1999 09:16:55 +0900, Matt Gushee
 <mgushee at havenrock.com> wrote:
>Hi, Folks--
>
>I'm working on some date-related functions, and wanted to see what
>kind of exception 'strptime' would raise if I fed it a European-style
>date string with a US format. But instead of an exception, I get:

This may be a Linux libc bug.  It's not clear what the behaviour is (should
be) on invalid dates from the manpage.  It's probably hard or impossible
for python to check the format of the date before passing it to the OS's
strptime, since %x uses the locale's format and any argument is potentially
valid for %x in some locale.

Your best bet might be to bug the glibc folks with a bug report, since the
analagous C program also segfaults:

#include <time.h>

main() {
	struct tm tm;
	strptime("31/7/99", "%x", &tm);
}

(gdb) run
Starting program: /usr/local/home/jepler/a.out 

Program received signal SIGSEGV, Segmentation fault.
0x40056eba in __strncasecmp (s1=0x4009ebef "Sunday", s2=0x0, n=6)
    at ../sysdeps/generic/strncase.c:41
../sysdeps/generic/strncase.c:41: No such file or directory.
(gdb) where
#0  0x40056eba in __strncasecmp (s1=0x4009ebef "Sunday", s2=0x0, n=6)
    at ../sysdeps/generic/strncase.c:41
#1  0x4006c3d3 in strptime_internal (buf=0x0, 
    format=0x4009eb20 "%a %b %e %H:%M:%S %Y", tm=0xbffff62c, 
    decided=0xbffff614) at strptime.c:261
#2  0x4006c890 in strptime_internal (buf=0x804854f "31/7/99", 
    format=0x804854c "%x", tm=0xbffff62c, decided=0xbffff614) at strptime.c:368
#3  0x4006de30 in strptime (buf=0x804854f "31/7/99", format=0x804854c "%x", 
    tm=0xbffff62c) at strptime.c:736
#4  0x80484f1 in main () at tst.c:5

Jeff
-- 
\/ http://www.freshmeat.net/                     Jeff Epler jepler at inetnebr.com
You may easily play a joke on a man who likes to argue -- agree with him.
		-- Ed Howe



From fredrik at pythonware.com  Wed Jul 14 21:27:50 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Jul 1999 01:27:50 GMT
Subject: Terminating Python...
References: <BFC0587634C3D111AACB00805FA766FA02C1C652@exchny12.ny.smb.com> <378CFDDF.489CDCEF@compaq.com> <7mj8s9$op2@newsops.execpc.com>
Message-ID: <004901bece61$41270b60$f29b12c2@secret.pythonware.com>

Sim & Golda Zacks <simngolda at voyager.net> wrote:
> I use raise SystemExit which is documented in "Internet Programming with
> Python"

in which way is typing "raise SystemExit<Return>" better
than typing "<Control-D>" (which is documented in the
Python Tutorial) ?

</F>




From bennettt at am.appstate.edu  Sun Jul 18 22:44:59 1999
From: bennettt at am.appstate.edu (TMGB)
Date: Sun, 18 Jul 1999 22:44:59 -0400
Subject: cgi - push a page from a select menu
Message-ID: <3792912B.86C3584E@am.appstate.edu>

I am trying to use python to receive a url from a form select menu and
then send that page back to the browser.  All three of my attempts so
far leave this error message on the web server: 

[18/Jul/1999:22:06:34] failure: for host ....library.appstate.edu trying
to POST /CGI-NT_Python/goto.pyc, cgi-parse-output reports: the CGI
program D:\...(path to python)\Python\python.exe did not produce a valid
header (program terminated without a valid CGI header (check for core
dump or other abnormal termination) 

I also get this on a web page for the third attempt: AttributeError:
open_http 

The python paths are correct, I have another python form using these
paths.  I would think that the urllib would send the url requested with
its header( if I use it correctly).  I tried printing a header and still
get this error.  The URL of the page with the selector is:

http://www.library.appstate.edu/saat/forms/test/


# Author: Thomas McMillan Grant Bennett
# webmaster at www.library.appstate.edu
# Python script to accept a url from a form and go to that URL
# Filename: goto.py

import cgi
import urllib

form = cgi.FieldStorage()
form_ok = 0
if form.has_key("select"):

# third attempt
#   html_page=urllib.open_http(form["select"].value)
#   print html_page

# second attempt  
#   html_page=urllib.urlretrieve(form["select"].value)
#   print html_page

# first attempt
#   print(read(urllib.urlopen(form["select"].value)))
   


----------------------------------------------------------------------
Thomas McMillan Grant Bennett           Appalachian State University
Computer Consultant II                  University Library
bennettt at am.appstate.edu               
http://www.library.appstate.edu/admin/
Voice:  828 262 6587			FAX:    828 262 3001

Windows 95 is a 32-bit extension to a 16-bit patch for an 8-bit
operating system that was originally coded for a 4-bit microprocessor.
 - Chris Dunphy     Boot Magazine



From wolfgang.grafen at de.bosch.com  Mon Jul 26 10:40:48 1999
From: wolfgang.grafen at de.bosch.com (Wolfgang Grafen)
Date: Mon, 26 Jul 1999 16:40:48 +0200
Subject: seqdict - 2 sequential dictionaries for python
References: <mt2.0-27614-932479015@news.informatik.uni-bonn.de> <379C51CC.B496B966@de.bosch.com>
Message-ID: <379C7370.14C69BB@de.bosch.com>

Wolfgang Grafen wrote:
> 
> 26.7.1999
> 
> Package seqdict
> 
> seqdict - single value dictionary
> msedict - multiple value dictionary

oops, should have gone to comp.lang.python.announce

Sorry for that...

Wolfgang
-------------- next part --------------
A non-text attachment was scrubbed...
Name: wolfgang.grafen.vcf
Type: text/x-vcard
Size: 329 bytes
Desc: Card for Wolfgang Grafen
URL: <http://mail.python.org/pipermail/python-list/attachments/19990726/6316d374/attachment.vcf>

From aahz at netcom.com  Thu Jul  1 21:07:28 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 2 Jul 1999 01:07:28 GMT
Subject: please help with simple regex
References: <7lh1h6$t84$1@nnrp1.deja.com>
Message-ID: <7lh3cg$5c8@dfw-ixnews4.ix.netcom.com>

In article <7lh1h6$t84$1 at nnrp1.deja.com>,  <viofis at my-deja.com> wrote:
>
>I have a string, for example, myString="the Potato".
>1. Convert it to all lowercase ("the potato")

import string
myString = string.lower(myString)

>2. Match it against a list of strings - "potato,tomato,corn,etc...",
>extract "potato" (if it matches with something on the list), then save
>into myNewString.

If you mean literally a list of strings, here's one approach:

stringList = [ 'potato', 'tomato', 'corn' ]
if myString in stringList :
	myNewString = myString

Of course, a better (more efficient) approach equivalent to above is

stringDict = { 'potato':None, 'tomato':None, 'corn':None }
if stringDict.has_key(myString) :
	myNewString = myString

(I'm pretty sure that's not what you want, but Python's regex is a much
smaller hammer than in Perl, and you should get used to trying out other
approaches.)
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From olli at rhein-zeitung.de  Thu Jul 22 08:37:20 1999
From: olli at rhein-zeitung.de (olli at rhein-zeitung.de)
Date: 22 Jul 1999 12:37:20 GMT
Subject: int(long(-sys.maxint-1)) fails on Linux
References: <3796DC4F.224795DA@vision.arc.nasa.gov> <379706FE.6C5351F@appliedbiometrics.com>
Message-ID: <7n73a0$fll$2@news.rhein-zeitung.DE>

> 1) The mandrake compiler is buggy

That is the problem. Don't be to aggressive with the compiler options and
don't use the build.pl script. The results are just a little bit better, but a
lot of problems are caused by the architecture options. ;-)))

> For 3), Oliver is guilty, and he deserves so many hugs about
> his RPM stuff that this is just something that happens.

Thanks. I guess the problem above is the reason. This time I am not
responsible. ;-)))

Bye, Oliver Andrich

-- 
Oliver Andrich, RZ-Online, Schlossstrasse Str. 42, D-56068 Koblenz
Telefon: +49-261-39210-40 Fax: -33 |       http://rhein-zeitung.de 
                                   |            http://andrich.net
GMC d- s++:++ a- C++ UL++++ P-- L+++ E--- W++ N+ w-- O- M-- 
V- PS++ PE Y+ PGP+ t 5+ X R+ tv+ b++ DI++ D- G e h- r z+



From fdrake at cnri.reston.va.us  Wed Jul 21 10:51:42 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Wed, 21 Jul 1999 14:51:42 GMT
Subject: uu.py - get result of uuencode() in a string
In-Reply-To: <3795CDC1.40F5E36@freemail.nl>
References: <3795CDC1.40F5E36@freemail.nl>
Message-ID: <14229.56958.548904.917898@weyr.cnri.reston.va.us>

Richard van de Stadt writes:
 > Is it possible to get the output of the uuencode() function
 > into a string, instead of a file or stdout?

  The output file can be a StringIO instance; a convenience function
to handle this could be written as (untested):

------------------------------------------------------------------------
import StringIO
import uu

def uu2string(data, mode=None):
    outfile = StringIO.StringIO()
    infile = StringIO.StringIO(data)
    uu.decode(infile, outfile, mode)
    return outfile.getvalue()
------------------------------------------------------------------------


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From cjw at connection.com  Wed Jul 28 12:24:41 1999
From: cjw at connection.com (Colin J. Williams)
Date: Wed, 28 Jul 1999 12:24:41 -0400
Subject: Incompatibility.py
References: <000a01bed7ce$8b320ae0$eea22299@tim>
Message-ID: <379F2EC8.DBE6F44B@connection.com>

Incidentally, PTUI returns:The name is:__builtins__

Colin W.

Tim Peters wrote:

> [Colin J. Williams]
> > IDLE returns:                        The name is: Incompatibility
>
> [Mark Hammond]
> > Not for IDLE 0.5 it doesnt.  Are you sure you are not using
> > "Import Script".
> > ...
> > The behaviour you describe sounds like you are simply confusing
> > "Run Script" with "Import Script".
>
> Except that "Run Script" doesn't exist in the released IDLE -- you have to
> get the CVS version for that.
>
> Since Guido doesn't have anything else to do <wink>, I vote he fixes it so
> breakpoints can be set under Windows again, then put out a new IDLE release
> + updated Lib/pyclbr.py.  The CVS is version is gads better in several key
> ways.
>
> release-early-&-often-ly y'rs  - tim





From tismer at appliedbiometrics.com  Tue Jul 20 08:30:19 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 20 Jul 1999 12:30:19 GMT
Subject: Python and Boehm-Demers GC, I have code.
References: <000901becffa$b0f25e00$91a22299@tim>
		<p5k8rvmymj.fsf@bidra241.bbn.hp.com> <Pine.LNX.4.10.9907201250340.6803-100000@atrus.jesus.cam.ac.uk>
Message-ID: <37946BDB.CB4CEDA9@appliedbiometrics.com>


Michael Hudson wrote:
> 
> On 20 Jul 1999, Markus Kohler wrote:
> [snippety snip]
> >
> > Is there anything that would prevent someone to reimplement loops such
> > that only one variable is allocated ?
> >
> 
> Yes. The fact that ints are immutable. Consider:
> 
> for i in xrange(10):
>     if i == 2:
>         j = i
> 
> if the loop only allocated one variable, then at the end of the loop
> 'print j' would give the answer '9'!
> 
> I guess this could be worked around by inspecting refcounts and such...
> but not easily.

FYI, stackless Python already saves the internal loop count into
a counter object. The explicit creation of "i" is still there,
of course. But you might know that inteegr allocation is completely
handled through a cache, and I see not much benefit coming from
making my mutable integer object public.
This look to cheap for me, to be replaced by the hair of extra
refcount tracking.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From bwarsaw at cnri.reston.va.us  Mon Jul 19 12:54:50 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Mon, 19 Jul 1999 12:54:50 -0400 (EDT)
Subject: Are postings to comp.lang.python.announce mirrored here?
References: <37933DB1.47814E84@cfdrc.com>
Message-ID: <14227.22618.779710.843912@anthem.cnri.reston.va.us>

>>>>> "LJ" == Lyle Johnson <jlj at cfdrc.com> writes:

    |  1. For some reason I'd convinced myself that posts
    |     to comp.lang.python.announce were automatically
    |     copied here; not true?

Not true.  But remember that c.l.py.a is also mirrored on
python-announce-list at python.org.  You can subscribe via

    http://www.python.org/mailman/listinfo/python-announce-list

    |  2. Is it considered bad form to cross-post new S/W
    |     announcements to both lists? Or is that the Right
    |     Thing to do?

It's probably okay (as long as it's a real crosspost and not two
separate posts).  The thing /not/ to do is post announcements only to
c.l.py.  The `Recent Announcements' part of the Python home page, and
the corresponding Netscape channel, is only culled from c.l.py.a.

-Barry



From kistler at fnmail.com  Sun Jul  4 16:45:24 1999
From: kistler at fnmail.com (Per Kistler)
Date: Sun, 04 Jul 1999 22:45:24 +0200
Subject: Tkinter._test() problem (THANKS!!)
References: <1281033721-53904975@hypernet.com>
Message-ID: <377FC7E4.DE48CC28@fnmail.com>

Hi Gordon

Many thanks! I could now install py152 and tcl/tk8.0.5 
not only on Linux but also on IRIX6.5 :-))

After Perl/Tk and Tcl/Tk I'd like now to learn about Python:-)

Regards, per.

Gordon McMillan wrote:
> 
> Per Kistler writes:
> >
> > I installed py1.5.2 with tcl/tk.8.1 on SuSE6.1 Linux.
> 
> That's your problem. Either drop back to tcl/tk 8.0, or may try
> Dieter Maurer's patches, posted June 1.
> 
> - Gordon

-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net
------------------------------------------------------------



From billtut at microsoft.com  Fri Jul 23 07:44:07 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Fri, 23 Jul 1999 11:44:07 GMT
Subject: COM bites - help please
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F9C@RED-MSG-50>


> From: Karl Putland [mailto:kperacles at geocities.com]

> I've got a couple of pyCOM things that are kicking my @$$! 
> 
> 1.  Trying to wrap a list of instances into a 
> util.Collection.  All instances have been prepped with 
> _public_methods_ and _public_attrs_ and have been wrapped 
> with a call to win32com.server.util.wrap(field) Trying to 
> access the Collection from a com instance produces the 
> following error.
> 
[...]
> >>> c1.fields        # This is a util.Collection(list of 
> GMField instances)
> Traceback (innermost last):
>   File "<pyshell#56>", line 1, in ?
>     c1.fields
>   File "D:\Python\win32com\client\dynamic.py", line 394, in 
> __getattr__
>     raise pythoncom.com_error, details
> com_error: (-2147467259, 'Unspecified error', None, None)
> 

Not having the source makes it more difficult to help here...
If memory serves given a list of Python instances that can be wrapped as COM
objects, the way to make a collection is:

collection = util.Collection(map(lambda x: util.wrap(x), list)

> 3. PyUnicode objects bit me in the ass when strings are 
> passed into methods and manipulated.
> class Foo:
>     <COM stuff>
>     def bar(self, somestring):
>         somestring = string.upper(somestring)
>         return somestring
> 
> >>>foo = win32com.client.Dispatch("Python.Foo")
> >>>foo.bar("Hello World")
> 

PyUnicode strings have an upper and a lower method so this becomes:
def bar(self, somestring):
	somestring = somestring.upper()
	return somestring

> Some traceback about TypeError PyUnicode.
> 
> Causing a need to change bar to
>     def bar(self, somestring):
>         somestring=str(somestring)
>         somestring = string.upper(somestring)
>         return somestring
> This change needs to be made in ALL methods that accept and 
> manipulate strings???
> 

This is the bane of Unicode, the string module can't handle them.
You either have to pass Unicode into functions that like Unicode, or convert
the Unicode into a normal Python string. Not much else you can do about
this.
It's been rumored that Unicode support will get better in Python 1.6.

> Is there a better method than COM for allowing someone who 
> uses VB access to Python objects, short of converting all of 
> the VB users to Python or resorting to reimplementing in VB 
> or C++?  

Using Python is good. :) But yes, this is the easiest way for Python and VB
to talk to each other.

> The COM interface I've been able to expose is 
> sufficient for now but does not allow some of the access that 
> I have in Python.  Maybe that's good in the end?
> 

I think so. :)

Hope this helps out a little bit...

Bill



From Gaetan_Corneau at baan.com  Wed Jul 14 13:31:54 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 14 Jul 1999 17:31:54 GMT
Subject: Redirect stdout to file
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF47B@ex-quebec-u1.baan.com>

Hello,

I'm in Barneveld Holland right now (hey, maybe I'll go visit the famous
Stichting Mathematisch Centrum ;) and I would like to redirect output from
the print statement to a text file. I think a saw a message about that some
time ago, but I can't find it right now.

Anyone could help?

Thanks,
______________________________________________________
   Gaetan Corneau
   Software Developer (System integration Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From fdrake at cnri.reston.va.us  Wed Jul 21 16:48:03 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Wed, 21 Jul 1999 20:48:03 GMT
Subject: Python Mode for Alpha ?
In-Reply-To: <lbzr9m1etb7.fsf@morgoth.uio.no>
References: <lawter-2107990058210001@client-151-197-119-26.bellatlantic.net>
	<14229.57251.996871.502133@anthem.cnri.reston.va.us>
	<14229.58766.234709.815774@weyr.cnri.reston.va.us>
	<o9vhbevwgh.fsf@panix3.panix.com>
	<14229.61952.83320.458319@weyr.cnri.reston.va.us>
	<lbzr9m1etb7.fsf@morgoth.uio.no>
Message-ID: <14230.12803.340986.633811@weyr.cnri.reston.va.us>

I wrote:
 > |   I've no idea; I've not used a Macintosh as a development environment 
 > | for at least four years!

And Rolf Marvin B?e Lindgren responded:
 > well shame to you!

Rolf,
  I'll be glad to spend time learning about current Mac systems and
developing neat stuff on them if you send along a machine and salary
with which I can support my family.  ;-)  I'm not (horribly)
prejudiced!


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From Michael.P..Reilly at p98.f112.n480.z2.fidonet.org  Thu Jul  1 16:13:59 1999
From: Michael.P..Reilly at p98.f112.n480.z2.fidonet.org (Michael P. Reilly)
Date: Thu, 01 Jul 1999 21:13:59 +0100
Subject: Python GUIs: Summary and Conclusion
Message-ID: <00001634@bossar.com.pl>

From: "Michael P. Reilly" <arcege at shore.net>

Mikael Lyngvig <mikael at pobox.com> wrote:
: On Tue, 29 Jun 1999 09:27:57 +0200, Gerrit Holl
: <gerrit.holl at pobox.com> wrote:

:>I think one could also make an application "multiple-interface". The user can
:>choose: svgalib (any support in python, by the way?), curses, commandline,
:>tk, qt/kde or some of the other stuff. The developer can put the interface
:>in a seperate class or even in a seperate module (I'll do with my discoverb
:>project (see signature)), so the user can exactly download what he want. I
:>believe KDE users would like KDE applications...

: I think that would be a good idea, but I don't think it is directly
: related to the issue at hand.  Such an API would certainly be slower
: than the actual implentation and the problem is, IMHO, that the actual
: implementation (of Tk) is too slow to be really interesting.

: As a matter of fact, I've always wished that I could make my apps so
: they took an option, say -ui:name, what would allow the user to
: specify which kind of user interface it had:

:    bui - batch user interface; Unix style command-line interface.
:    cui - console user interface; Unix curses interface.
:    gui - graphical user interface; X/Windows style interface.

: That way the user could use the application in any way he/she like,
: without having to resort to Win32 COM (automation server) APIs etc.

Been there, done that.  My pyirc client (alpha release) was designed
for just about any user-interface, and comes with three: dummy (no
input or output, just batch-mode)), curses and Tkinter.  You just can't
switch between them at runtime, ;)  I've done it for other apps too.

  -Arcege



From graham at sloth.math.uga.edu  Mon Jul 19 12:12:18 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 19 Jul 1999 16:12:18 GMT
Subject: Stackless & String-processing
References: <000a01bed19c$edcea9c0$f0a02299@tim>
Message-ID: <7mvip2$cpj$2@cronkite.cc.uga.edu>

[Graham Matthews]
: > You have totally misunderstood what I said.
[Fernando Pereira]
: > One last try. I understood your point from the beginning. However, you
: > did not understood *mine*.
Tim Peters (tim_one at email.msn.com) wrote:
: Trust me on this one, fellows:  you each understand the other, and have for
: several (interesting) msgs already.  Graham has an unusual way of expressing
: his appreciation for a good discussion

Tim I don't think that we are agreeing.: Fernando seems to want to
view combinators as parsing constructs and hence says things like
"the combinators don't have to do much if you are using automata
based black boxes" (to paraphrase). But combinators are more than
parsing constructs -- they are sequencing constructs used for a wide
variety of tasks where automata techniques are not available.

graham

-- 
            The girls is all salty, the boys is all sweet, 
                      the food ain't too shabby 
                    an' they piss in the streets
    In France, way down in France, way on down, way on down in France



From mnot at pobox.com  Fri Jul  9 22:52:32 1999
From: mnot at pobox.com (Mark Nottingham)
Date: Sat, 10 Jul 1999 02:52:32 GMT
Subject: Python webcrawler?
References: <99Jul9.164357pdt."3569"@watson.parc.xerox.com>
Message-ID: <055001beca7f$42756600$0301a8c0@mnot.net>

I started writing a crawler, but never finished. However, the engine was
used in the Cacheability Engine -- http://www.mnot.net/cacheability/

It might be useful to rip the core out as a start for one if you don't have
anything else.

Cheers,




From aa8vb at yahoo.com  Fri Jul  9 07:55:04 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Fri, 9 Jul 1999 07:55:04 -0400
Subject: .pyo's without .pyc's?
In-Reply-To: <199907081309.JAA01063@eric.cnri.reston.va.us>; from Guido van Rossum on Thu, Jul 08, 1999 at 09:09:28AM -0400
References: <19990526140137.A66609@vislab.epa.gov> <374C6C6E.1D07326B@lemburg.com> <19990708074113.A61389@epa.gov> <37849A3F.28C0EF21@lemburg.com> <199907081309.JAA01063@eric.cnri.reston.va.us>
Message-ID: <19990709075504.A59930@epa.gov>

Guido van Rossum:
 |>Randall Hopper
 |>| I'm trying to cook a minimalist Python install, and I find that python -O
 |>| won't run without .pyc's in-place, even when there are pyo's.
 |>
 |> exceptions.py is imported prior to all other modules at a time where
 |> the switch from .pyc to .pyo hasn't been done yet
 |
 |Seems like this would be the fix, right?

It's working fine now.  Thanks.

Randall



From ivanlan at callware.com  Thu Jul 15 18:02:52 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Thu, 15 Jul 1999 22:02:52 GMT
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net>
Message-ID: <378E5A8C.C6FB16AF@callware.com>

Hi Mary--

mary jeanette wrote:
> 
> I decided to learn python because I heard it was a very easy language to
> learn. So I downloaded it and bought a couple of very expensive
> paperback books from o'reilly only to discover that they are written for
> unix. In other words, typing in the examples from the book into the
> interpretor only yields various syntax errors. I have no idea what I am
> supposed to change in order to make the examples work, and apparently
> there is nowhere to turn in order to find out. Q: Am I supposed to know
> DOS as a prerequisite? The idea of having to learn DOS is appalling.
> Should I try partitioning my hard drive and installing Linux or freeBSD?
> Perhaps I should learn Perl instead, I have seen books on Perl written
> for Win32 systems.   Thanks for your advice.

The example Python code should work just fine in the interpreter,
regardless of whether the OS is UNIX or DOS or Windows or Mac.  Only a
few small parts of Python are platform-specific enough to cause problems
if not run on UNIX.  *Please* don't give up on Python so easily!  You
would regret learning Perl instead, although it is true that if you
learn Perl, you'll have no trouble learning Python!

Why don't you post some example code that you tried on your Windows/DOS
platform that caused syntax errors?  There are hundreds of nice people
here who would *love* to help you get started, but we can't do it unless
you give us a little more to go on.  If you post code, we can help.  If
you're not specific enough about your problems, we can only lend a
sympathetic ear.

<no-no-not-the-comfy-perl!>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From philen at earn.aa.ans.net  Fri Jul  9 00:35:30 1999
From: philen at earn.aa.ans.net (Phillip Lenhardt)
Date: Fri, 09 Jul 1999 04:35:30 GMT
Subject: Example Code: Bresenham Line Algorithm
Message-ID: <slrn7oav0i.mct.philen@earn.aa.ans.net>

If there is some python algorithm repository this
should go in, let me know. Otherwise, enjoy!

def Brensenham_line(x,y,x2,y2):
    """Brensenham line algorithm"""
    steep = 0
    coords = []
    dx = abs(x2 - x)
    if (x2 - x) > 0: sx = 1
    else: sx = -1
    dy = abs(y2 - y)
    if (y2 - y) > 0: sy = 1
    else: sy = -1
    if dy > dx:
        steep = 1
        x,y = y,x
        dx,dy = dy,dx
        sx,sy = sy,sx
    d = (2 * dy) - dx
    for i in range(0,dx):
        if steep: coords.append((y,x))
        else: coords.append((x,y))
        while d >= 0:
            y = y + sy
            d = d - (2 * dx)
        x = x + sx
        d = d + (2 * dy)
    coords.append((x2,y2))
    return coords



From fdrake at cnri.reston.va.us  Wed Jul 21 11:21:50 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Wed, 21 Jul 1999 15:21:50 GMT
Subject: Python Mode for Alpha ?
In-Reply-To: <14229.57251.996871.502133@anthem.cnri.reston.va.us>
References: <lawter-2107990058210001@client-151-197-119-26.bellatlantic.net>
	<14229.57251.996871.502133@anthem.cnri.reston.va.us>
Message-ID: <14229.58766.234709.815774@weyr.cnri.reston.va.us>

>>>>> "JL" == John Lawter <lawter at cse.buffalo.edu> writes:
    JL> Is there a more recent Python mode for Alpha than the 6.5 era
    JL> one ?  I have searched around the python.org site, but their
    JL> copy is missing.  Any help would be appreciated, as I don't
    JL> look forward to having to write my own.

Barry A. Warsaw writes:
 > Are you asking about python-mode.el for Emacs?  If so, the one at
 > http://www.python.org/emacs/python-mode/ will work for any modern
 > version of Emacs or XEmacs.  If not, I'm not sure I can help ya.

  No, he's asking about an editing mode for Alpha, a Mac editor.  I
remember having one when I used a Macintosh, but that was almost 5
years ago.  I have no idea where to get such a beast these days.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From S.I.Reynolds at cs.bham.ac.uk  Wed Jul 14 09:47:08 1999
From: S.I.Reynolds at cs.bham.ac.uk (Stuart Reynolds)
Date: Wed, 14 Jul 1999 14:47:08 +0100
Subject: Pickling C objects
References: <378B27A9.4201@cs.bham.ac.uk> <14219.39612.343183.547443@weyr.cnri.reston.va.us>
Message-ID: <378C94DC.2663@cs.bham.ac.uk>

Fred L. Drake wrote:
> 
> Stuart Reynolds writes:
>  > I'm writing some new Python classes with a C implementation. How do you
>  > make a C object pickleable?
> 
>   Take a look at the source for the parser module in the standard
> source distribution; it uses the copy_reg module to register a
> mechanism to support pickling.  It worked when I wrote it, but I
> haven't tried it lately.  (A good candidate for a regression test!)
> 

Groovy. You don't even need to worry about whether the object is written
partly in C as long as you can recreate the C bits from Python (e.g. via
the constructor call).


---------

import copy_reg


class Saveable:
    def __init__(self, a, b):
	self.a = a
	self.b = b

    def __repr__(self):
	"""Returns a string which may be evaluated
	in order to reconstruct this object.
	"""
	ser = self.__class__.__name__      \
	      +'('                         \
 	      +   repr(self.a) + ','       \
	      +   repr(self.b)             \
	      +')'
	return ser




def __Saveablesaver__(object):
    """Returns a serialised version of object"""
    return repr(object)


def __Saveableloader__(serialised):
    """ Unserialised `serialised' and return
    a new Saveable object """
    return eval(serialised)





#Register the Saveable pickling functions
copy_reg.pickle( Saveable, __Saveablesaver__, __Saveableloader__ )



if __name__ == '__main__':
    """ Test the serialisation of Saveable  """
    #import pickle
    import cPickle
    pickle = cPickle

    sm = Saveable( 1, ['a','b',(4,5,6)] )
    
    #View the serialised form
    print 'Saving: ', repr(sm)


    pickle.dump( sm, open('deleteme','w') )
    obj = pickle.load( open('deleteme','r') )

    print 'Loaded: ', repr(obj)

------
Should output the following:

Saving:  Saveable(1,['a', 'b', (4, 5, 6)])
Loaded:  Saveable(1,['a', 'b', (4, 5, 6)])



From sreekantk at freenet.co.uk  Sun Jul 18 19:20:10 1999
From: sreekantk at freenet.co.uk (sreekant kodela)
Date: Mon, 19 Jul 1999 00:20:10 +0100
Subject: python?
References: <378DEB78.34747906@hou.valmet.com>
Message-ID: <3792612A.57801B43@freenet.co.uk>

This is an adult site. Subject header is misleading, saying python.

Regards
sreekant

Rod Haper wrote:

> http://207.240.225.250




From haper at ibm.net  Sat Jul 31 12:03:23 1999
From: haper at ibm.net (Rod Haper)
Date: Sat, 31 Jul 1999 16:03:23 +0000
Subject: How to find out the pid of a running program?
References: <7nucu2$11ko@enews1.newsguy.com>
Message-ID: <37A31E4B.2CD53D40@ibm.net>

Peter Torstenson wrote:
> 
> I'd like tknow if there is a way of getting the pid of a program running,
> using python commands?

Use the getpid() function from the posix module which is imported
automatically when you import the os module.  Don't import the posix
module directly.  Do it via the os module since this will afford you
code better portablility.  Refer to the Python Library Reference for
more info.



From dont at try.it  Mon Jul  5 11:06:54 1999
From: dont at try.it (dont at try.it)
Date: Mon, 5 Jul 1999 17:06:54 +0200
Subject: CGI under NT
Message-ID: <3780ca11@news2.isys.net>

Hallo everybody,

I want to work with CGI under Windows NT 4. The CGI Script should work with
the IIS 4.

Until now I wrote a script which takes alle parameters of the forms out of
"QUERY_STRING". There is my problem. If I press the submit-button I get a
Message that the script has problems to read the system-variable (because it
is not there!), but if I call the script with the parameters directly
(..../script.py?value=100&browser=netscape etc.) the script works.

Can anyone helps me or give me a hint to more information on the web?

Thanks a lot
geetings Moritz





From nascheme at ucalgary.ca  Fri Jul 16 15:53:02 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 16 Jul 1999 19:53:02 GMT
Subject: Python and Boehm-Demers GC, I have code.
Message-ID: <7mo2iu$iqg@ds2.acs.ucalgary.ca>

I have added the Boehm-Demers garbage collector to Python and it
works well!  With the default set of extension modules all tests
pass.  Also I have done some benchmarks using pystone and
pybench.  The GC version of Python is just as fast as the regular
version!  I was quite amazed.  Here are the pybench results:


PYBENCH 0.5

Benchmark: free3 (rounds=3, warp=10)

Tests:                              per run    per op.    diff *
------------------------------------------------------------------------
          BuiltinFunctionCalls:     746.83 ms    2.93 us  -11.90%
           BuiltinMethodLookup:    1139.33 ms    1.09 us   -1.07%
               CreateInstances:    1006.83 ms   11.99 us   -9.09%
                  DictCreation:     847.17 ms    2.82 us  -21.66%
                    IfThenElse:    1357.67 ms    1.01 us   -1.84%
                   ListSlicing:    1116.33 ms  127.58 us  +38.30%
                NestedForLoops:     970.00 ms    1.29 us   -1.40%
          NormalClassAttribute:    1126.50 ms    0.94 us   -1.46%
       NormalInstanceAttribute:    1086.33 ms    0.91 us   -2.75%
           PythonFunctionCalls:    1083.50 ms    3.28 us   +3.63%
             PythonMethodCalls:     837.83 ms    5.59 us   +1.23%
                     Recursion:     981.50 ms   39.26 us  +16.37%
        SimpleDictManipulation:     934.67 ms    1.56 us   -7.96%
         SimpleFloatArithmetic:    1112.83 ms    1.01 us   +4.95%
       SimpleIntegerArithmetic:    1000.83 ms    0.76 us   -2.48%
        SimpleListManipulation:    1213.50 ms    2.25 us  +36.74%
                    SmallLists:    1132.67 ms    2.22 us  -15.34%
                   SmallTuples:    1198.50 ms    2.50 us   -3.44%
         SpecialClassAttribute:    1132.83 ms    0.94 us   -1.51%
      SpecialInstanceAttribute:    1299.67 ms    1.08 us   -2.44%
                  TupleSlicing:    2027.17 ms    3.86 us   -5.52%
------------------------------------------------------------------------
            Average round time:   26870.00 ms              -1.39%

*) measured against: py152 (rounds=10, warp=10)


I started with Sam Rushing's patch and modified it for Python
1.5.2c.  To my surprise, removing the reference counting
(Py_INCREF, Py_DECREF) actually slowed down Python a lot (over
two times for pystone).  I not exactly sure why this is.  Maybe
someone can enlighten me.  I will try to do some more testing.

I have to fix up modules which use other mallocs (readline and
_tkinter for sure).  Guido, will a patch be accepted which
changes Python code to use Py_Mem_* instead of the libc names
where approprate?  I will have to go through the extension
modules and fix them by hand.  Also, what about a patch that
makes using garbage collection a configure option.  The changes
are quite minor.  I don't know autoconf but maybe I can figure it
out.

With regards to finalization, __del__ methods are not called from
the garbage collector.  If you create reference loops then too
bad for you.  The garbage collector supports finalization so I
may fix this later.

A preliminary patch is attached.  The patch assumes libgc.a and
gc.h are installed somewhere and that you have already run
configure.  The Makefile diffs are probably okay only for a Linux
box.  I must learn autoconfig. :(


    Neil


diff -u -r Python-1.5.2c1/Include/mymalloc.h Python-gc/Include/mymalloc.h
--- Python-1.5.2c1/Include/mymalloc.h	Fri Dec  4 11:48:10 1998
+++ Python-gc/Include/mymalloc.h	Fri Jul 16 03:49:55 1999
@@ -87,6 +87,18 @@
 #define _PyMem_EXTRA 0
 #endif
 
+#ifdef Py_USE_GC
+#include <gc.h>
+#define malloc(n) GC_MALLOC(n)
+#define calloc(m,n) GC_MALLOC((m)*(n))
+#define realloc(o,n) GC_REALLOC(o,n)
+#ifdef Py_USE_GC_FREE
+	#define free(n) GC_FREE(n)
+#else
+	#define free(n)
+#endif /* Py_USE_GC_FREE */
+#endif /* Py_USE_GC */
+
 #define PyMem_NEW(type, n) \
 	( (type *) malloc(_PyMem_EXTRA + (n) * sizeof(type)) )
 #define PyMem_RESIZE(p, type, n) \
diff -u -r Python-1.5.2c1/Makefile Python-gc/Makefile
--- Python-1.5.2c1/Makefile	Fri Jul 16 04:08:43 1999
+++ Python-gc/Makefile	Fri Jul 16 04:08:37 1999
@@ -143,7 +143,7 @@
 WITH=		
 
 # Compiler options passed to subordinate makes
-OPT=		-g -O2
+OPT=		-g -O2 -DPy_USE_GC
 
 # Subdirectories where to run make recursively
 SUBDIRS=	Parser Objects Python Modules
diff -u -r Python-1.5.2c1/Modules/Makefile.pre Python-gc/Modules/Makefile.pre
--- Python-1.5.2c1/Modules/Makefile.pre	Fri Jul 16 04:08:44 1999
+++ Python-gc/Modules/Makefile.pre	Fri Jul 16 04:16:06 1999
@@ -28,7 +28,7 @@
 SGI_ABI=	
 
 DEFS=		-DHAVE_CONFIG_H
-LIBS=		-lieee -ldl 
+LIBS=		-lieee -ldl -lgc
 LIBM=		-lm
 LIBC=		
 
diff -u -r Python-1.5.2c1/Objects/Makefile Python-gc/Objects/Makefile
--- Python-1.5.2c1/Objects/Makefile	Fri Jul 16 04:08:43 1999
+++ Python-gc/Objects/Makefile	Fri Jul 16 04:09:17 1999
@@ -13,7 +13,7 @@
 RANLIB=		ranlib
 AR=		ar
 
-DEFS=		-DHAVE_CONFIG_H
+DEFS=		-DHAVE_CONFIG_H -DPy_USE_GC_FREE
 
 
 # === Other things that are customizable but not by configure ===
diff -u -r Python-1.5.2c1/Parser/Makefile Python-gc/Parser/Makefile
--- Python-1.5.2c1/Parser/Makefile	Fri Jul 16 04:08:43 1999
+++ Python-gc/Parser/Makefile	Fri Jul 16 04:08:52 1999
@@ -14,7 +14,7 @@
 AR=		ar
 
 DEFS=		-DHAVE_CONFIG_H
-LIBS=		-lieee -ldl 
+LIBS=		-lieee -ldl -lgc
 
 
 # === Other things that are customizable but not by configure ===



From billtut at microsoft.com  Tue Jul 13 11:23:19 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Tue, 13 Jul 1999 15:23:19 GMT
Subject: making a Windows NT service out of a python program
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB90F24@RED-MSG-50>


> From: Christian Tismer [mailto:tismer at appliedbiometrics.com]
> 
> Bill Tutt wrote:
> > 
> > Take a look at the stuff in win32\demos\service\ of your 
> friendly latest
> > Win32 extension install tree.
> > 
> > It shows you how to wrap Python code into an NT service 
> quite easily.
> > 
> > The code in win32\lib\win32serivceutil.py appears to do 
> almost everything
> > you would ever want to do with a service. Very cool stuff...
> 
> This is all great stuff if you have understood it already.
> For me, it is still unclear what makes up a service, and
> what I need at the minimum to turn a Python program into
> an NT service.
> 
> I need nothing more than to run Medusa as a service. Did
> anybody do this before? Do I need special calls at all?
> 

Blech, well I was working on the minimal code to turn a subclass of
SocketServer into an NT service, when I discovered low and behold, the
WSAEventSelect function isn't exposed anywhere to Python. (WSAEventSelect
allows you to call WaitForMulitpleObject/WaitForSingleObject on a socket)

On top of which, WSAEventSelect makes the socket non-blocking. Blech.

So, hopefully I'll have something nice before I fly to Iowa on Friday.
But what I have will require a rebuild of the win32 extensions. :(

Bill



From stephensong at iname.com  Fri Jul  2 08:10:46 1999
From: stephensong at iname.com (Gary Stephenson)
Date: Fri, 02 Jul 1999 22:10:46 +1000
Subject: Tuples cf. Expanded objects??
References: <377bbde1@news.xtml.co.uk> <7lh375$5ea$1@news.udel.edu>
Message-ID: <377CAC46.508D502F@iname.com>

>
> There is probably something in the the FAQ.  BAsically, tuples are stored
> more compactly and can be used as dictionary keys (because they ARE
> readonly).
>

How similar are Python's "Immutable Tuples"   to Eiffel's "Expanded objects".  If
one
was to try to implement  immutable tuples in another language, would an "expanded
objects" implementation be the best, or even appropriate?

many tias

gary




From Giambattista.Vico at p98.f112.n480.z2.fidonet.org  Thu Jul  1 10:43:17 1999
From: Giambattista.Vico at p98.f112.n480.z2.fidonet.org (Giambattista Vico)
Date: Thu, 01 Jul 1999 15:43:17 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001641@bossar.com.pl>

From: maedel at nospam_erols.com (Giambattista Vico)

I'd like to hear what people have to say in this thread if we consider
freezing or otherwise creating a standalone app, even if only for Unix and
Win. Has anyone found a Python tool or combination of tools that solves
this problem?

--tom



From stidolph at origin.ea.com  Tue Jul  6 13:13:48 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Tue, 6 Jul 1999 17:13:48 GMT
Subject: AttributeError - a dumb question
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B4179301E10FAA@molach.origin.ea.com>

Python is case sensitive:

class F:
    def __init__(self):

        self.x = self.KryLys() #note the upper case L

    def KryLys(self):
        l = [1,2,3,4,5]
        return l
    

l = F()
-----Original Message-----
From: Johann Spies [mailto:jhspies at futurenet.co.za]
Sent: Tuesday, July 06, 1999 11:24 AM
To: python-list at cwi.nl
Subject: Re: AttributeError - a dumb question


Hallo Oleg,

Thanks for your reply.  Your answer helped me to back to the original
problem I had:

class F:
    def __init__(self):

        self.x = self.Krylys()

    def KryLys(self):
        l = [1,2,3,4,5]
        return l
    

l = F()

Now this assignment fails:

Traceback (innermost last):
  File "<stdin>", line 11, in ?
  File "<stdin>", line 4, in __init__
AttributeError: Krylys

I hope you can help me with this one.

Johann

 --------------------------------------------------------------------------
| Johann Spies                                 Windsorlaan 19              |
| jhspies at futurenet.co.za                3201 Pietermaritzburg             |
| Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
 --------------------------------------------------------------------------

     "Ask, and it shall be given you; seek, and ye shall
      find; knock, and it shall be opened unto you."        
                                  Matthew 7:7 






From kernr at mail.ncifcrf.gov  Wed Jul 21 17:46:18 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Wed, 21 Jul 1999 21:46:18 GMT
Subject: Scientific Plotting?
References: <19990721010203.96101.qmail@hotmail.com> <37952384.14766416@news.erols.com> <3795c243.764244896@news.ufl.edu>
Message-ID: <37963f6d.2239819@news.erols.com>

On Wed, 21 Jul 1999 13:03:52 GMT, lazrnerd at ufl.edu (Craig Schardt)
wrote:

>On Wed, 21 Jul 1999 01:43:48 GMT, kernr at mail.ncifcrf.gov (Robert Kern)
>wrote:
>
>>On Tue, 20 Jul 1999 18:02:01 PDT, "Michael Huster"
>>
>>[snip]
>>
>>>Gnuplot: Runs gnuplot through a pipe. All of the data (upto many MB) also 
>>>has to go through the pipe. Unsat. I want to plot efficiently.
>>
>>The Gnuplot module as updated by Michael Haggerty
>>(http://monsoon.harvard.edu/~mhagger/download/) puts the data into
>>temporary files.  Pipes are used for commands only.  If you haven't
>>used it recently, you might want to try it again.
>>
>
>I have modified Michael Haggerty's Gnuplot.py to work under Windows.
>The Windows version of Gnuplot doesn't support pipes but there is a
>helper application, pgnuplot, which works around this problem and
>provides a pipe interface. The combination works well on NT and 98
>with PythonWin.

I've done the same thing, but I don't think I required PythonWin (as
in, Mark Hammond's win32 extensions).

>If anyone is interested in these programs, please let me know.
>
>-craig.
>lazrnerd at ufl.edu

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From martinp at mincom.com  Wed Jul 28 22:45:13 1999
From: martinp at mincom.com (Martin Pool)
Date: Thu, 29 Jul 1999 12:45:13 +1000
Subject: 1.5.2 for: else:
References: <1278996512-12636218@hypernet.com> <slrn7pun42.2ro.wtanksle@dolphin.openprojects.net>
Message-ID: <379FC039.73C436FC@mincom.com>

William Tanksley wrote:
> 
> On Tue, 27 Jul 1999 21:43:00 -0500, Gordon McMillan wrote:
> >William Tanksley wrote:
> >> On Tue, 27 Jul 1999 18:35:10 -0500, Gordon McMillan wrote:
> 
> >> >Maybe it violated your expectations, but it's much more valuable than
> >> >what you expected. After all, testing for an empty sequence is a
> >> >no-brainer. But finding a match in a list, and testing whether you
> >> >fell off the end without finding one, is (without the else clause) a
> >> >much messier proposal.

If nowhere else, I would have liked something like this in Java for
testing exception-handling code.  With this, I can write:

  >>> try:
  ...   print 0/0
  ... except ZeroDivisionError:
  ...   print "caught exception as expected"
  ... else:
  ...   print "Oops!  Exception didn't happen"

whereas in Java I have to write

  boolean caught = false;
  try {
    int a = 0 / 0;
  } catch (ArithmeticException a) {
    out.println("caught as expected");
    caught = true;
  } 
  if (!caught)
    out.print("exception didn't happen");

I doubt if I'll often use for: else or try: else, and perhaps they could
have clearer names, but I reckon they're nice to have.

-- 
 /\\\  Mincom | Martin Pool          | martinp at mincom.com
// \\\        | Software Engineer    | Phone: +61 7 3303-3333
\\ ///        | Mincom Ltd.          | 
 \///         | Teneriffe, Brisbane  | Speaking for myself only



From carlos.velasquez at ssmb.com  Wed Jul 14 14:36:14 1999
From: carlos.velasquez at ssmb.com (Velasquez, Carlos)
Date: Wed, 14 Jul 1999 14:36:14 -0400
Subject: Terminating Python...
Message-ID: <BFC0587634C3D111AACB00805FA766FA02C1C652@exchny12.ny.smb.com>

Hello,

I just downloaded and built Python v1.51 for the Sun Solaris platform. All
the tests went well and I created some simple scripts to verify its
operation. One question is, how do I exit python? The only thing that works
is <Ctrl> Z but that just stops the job. if I execute Python again and Ctrl
Z again, and if I do a ps -ef to see all processes, there are 2 python
instances.

Is there a correct way to exit from Python?

______________________________________________________________________
C.F. Velasquez
Asst. Vice President
Global Database and Technology Division
Salomon Smith Barney, Inc.




From herzog at online.de  Thu Jul 15 06:20:54 1999
From: herzog at online.de (Bernhard Herzog)
Date: 15 Jul 1999 12:20:54 +0200
Subject: Getters and Setters
References: <000701bece99$6dbeede0$51a22299@tim>
Message-ID: <m3aesy43jt.fsf@greebo.nodomain.de>

"Tim Peters" <tim_one at email.msn.com> writes:

> Oops!  Almost forgot the code!  Dump _Getter.  Dump _Setter.  Have the mixin
> class build the obvious function for what you're after, and add that
> directly to self as an attribute.  Then there's no mixin overhead at all
> after the first time a set/get method is invoked for an instance.  Downside:
> generates many function objects.  Possible alternative:  plug synthesized
> get/set functions into the class object instead; a little slower, but less
> proliferation of function objects.

Hmm. Unfortunately,...

> class GetterSetter:
>     def __getattr__(self, name):
>         if name[:3] == "get":
>             def getter(d=self.__dict__, key=name[3:]):
>                 return d[key]
>             setattr(self, name, getter)
>             return getter

..this introduced a circular reference.

So, putting the functions into the class would be better:

              def getter(self, key=name[3:]):
                  return getattr(self, key)
              setattr(self.__class__, name, getter)
              return getattr(self, name)

>         elif name[:3] == "set":
>             def setter(value, d=self.__dict__, key=name[3:]):
>                 d[key] = value
>             setattr(self, name, setter)
>             return setter

Same here.

>         else:
>             raise AttributeError(name)

[snipped rest of code]




-- 
Bernhard Herzog	  | Sketch, a python based drawing program
herzog at online.de  | http://www.online.de/home/sketch/



From donn at u.washington.edu  Thu Jul 22 12:30:14 1999
From: donn at u.washington.edu (Donn Cave)
Date: 22 Jul 1999 16:30:14 GMT
Subject: cin
References: <3srl3.1302$Gj8.191985152@newsa.telia.net>
Message-ID: <7n7gum$qlm$1@nntp6.u.washington.edu>

"Joacim L?wgren" <fire at netlimit.com> writes:
| I'm new to Python and I'm wondering what the C++ "cin" (or C
| "scanf()")equivalent in Python is. Doesn't seem there's a quick function for
| this.

Think not, in the standard distribution.  A few years back someone
contributed a C module (Steve Clift, from Tasmania, 1995), and you can
find it under contrib/Misc on ftp.python.org.  Seems to me there may
have been other efforts along these lines, too.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From Isao.Akiyama at unisys.co.jp  Thu Jul 15 06:29:09 1999
From: Isao.Akiyama at unisys.co.jp (Isao AKIYAMA)
Date: Thu, 15 Jul 1999 19:29:09 +0900
Subject: scoping assertion
References: <37839491.3F7C1F6@mikemccandless.com>
		 <3783C878.B8C3790B@bioreason.com> <19990708211651.A17658@optiplex.palga.uucp> <7cQP0BAYKRh3Ew6K@jessikat.demon.co.uk>
Message-ID: <378DB7F5.1A48506A@unisys.co.jp>

> I use
> try:
>         raise None
> except:
>         ....

Wow!

>>> import sys
>>> try: raise None
.. except: print sys.exc_info()
..
(<class exceptions.TypeError at 89a4e0>, <exceptions.TypeError instance
at 8cc17
0>, <traceback object at 8cc150>)
>>>

'raise None' actually raises a TypeError ;-)

Isao AKIYAMA




From lvirden at cas.org  Fri Jul  2 07:56:10 1999
From: lvirden at cas.org (lvirden at cas.org)
Date: 2 Jul 1999 11:56:10 GMT
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <slrn7nai1j.8ll.zeitlin@seth.lpthe.jussieu.fr> <64F6457BE0585BB3.1CF2F67BC8A680BD.A19A9FEE23BC9E58@lp.airnews.net>
Message-ID: <7li9cq$inp$1@srv38.cas.org>

According to Mikael Lyngvig <mikael at pobox.com>:
:Well, TkInter is extremely slow, but not because of the Python part.
:Even the Tk widget tour, included with Tk, is so slow that you almost
:think "Please, draw the next character - you can do it." whenever a
:window with some text needs to be redrawn.

When I run the Tk test suite  on my SPARC, Tk cranks thru creating and
deleting windows as fast as the eye can track.

However, there are reports that the Windows version of Tk, because of its
use of the native window widgets, is noticably slower than the X
implementation.  Is that the platform to which you refer?

-- 
<URL: mailto:lvirden at cas.org> Quote: Saving the world before bedtime.
<*> O- <URL: http://www.purl.org/NET/lvirden/>
Unless explicitly stated to the contrary, nothing in this posting
should be construed as representing my employer's opinions.



From genglerb at no_spam.uwplatt.edu  Fri Jul 23 21:34:52 1999
From: genglerb at no_spam.uwplatt.edu (Brett Gengler)
Date: Fri, 23 Jul 1999 20:34:52 -0500
Subject: Tk and a little headache (installation troubles)
Message-ID: <3799183B.28530914@no_spam.uwplatt.edu>

For some strange reason I'm having trouble getting the tk libray to work
with python.  I've installed tk & tcl 8.2 with a standard install.  I've
also sucessfully (or so I thought) re-compiled Python after editing the
Modules/Setup file.  I am able to import _tkinter but if I try to kick
up a quick window with say:
from Tkinter import *
widge1 = Label(None, text="Please, Please work! ")

I get a traceback error(in quotes):

"Traceback (innermost last):
    File "<stdin>", line 1, in ?
    File "/usr/local/Python-1.5.2/Lib/lib-tk/Tkinter.py", line 1410, in
__init__
        Widget.__init__(self, master, 'label',cnf, kw)
    File "/usr/local/Python-1.5.2/Lib/lib-tk/Tkinter.py", line 1410, in
__init__
        BaseWidget._setup(self, master, cnf)
    File "/usr/local/Python-1.5.2/Lib/lib-tk/Tkinter.py", line 1410, in
__init__
        _default_root = Tk()
    File "/usr/local/Python-1.5.2/Lib/lib-tk/Tkinter.py", line 1410, in
__init__
        self.tk = _tkinter.create(screenName, baseName, className)
TclError: Can't find a useable init.tcl in the following directories:


This probably means the Tcl wasn't installed properly. "

Okay, well after reinstalling everingthing just for kicks I began trying
to figure out a way for Tcl to find init.tcl.
in my path(in 'profile') I added "export TCL_LIBRARY =
/usr/local/lib/tcl8.2 "  but that didn't seem to work.  I thought
that just maybe it might have something to do with the Setup file having
the wrong path. "TKPATH :=lib-tk" is what it's set to.   I'm still
fairly new to Linux (Slackware 4.0) so I'm sure this is some dumb newbie
error <wink>  Any ideas what I'm doing wrong?























From avv at quasar.ipa.nw.ru  Sat Jul 24 16:33:11 1999
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Sat, 24 Jul 1999 13:33:11 -0700
Subject: Python documentation available as ZIP files
References: <1279303589-30799085@hypernet.com>
Message-ID: <379A2307.7201C0D1@quasar.ipa.nw.ru>

Gordon McMillan wrote:

> Most likely, you're not downloading properly <wink>. You're supposed
> to use a browser, and let it choose a sensible name like
> RockyAndBullWinkle_tar.gz.
>
> > If you want to preach the benefits of .tar.gz I'd reccomend
> > preaching about .tar.bz2... And that way you can try educating lots
> > of Unix users as well as the Windows gang <wink>.
>
> Tee hee. Use it on my Linux box. Someone pointed me to a Windows
> port. The readme says "can't run in a 16 bit subsystem", but the port
> *is* an MSDOS executable (not a 32 bit console executable). It did
> manage to decompress, but mangled file names beyond recognition.

There is nothing better is that case than to have a choice. Even
ftp://ftp.kernel.*.org present both .gz and .bz2. My experience is that
bzip2 is extremely good for text files, but for some sorts of binary
files (thousands of double's written by a Fortran program) bzip2 is even
worse than zip. I even planned to send such a package to the maintainer
of bzip2, but got too lazy...

Alexander





From Drew.Whitehouse at p98.f112.n480.z2.fidonet.org  Thu Jul  1 02:23:08 1999
From: Drew.Whitehouse at p98.f112.n480.z2.fidonet.org (Drew Whitehouse)
Date: Thu, 01 Jul 1999 07:23:08 +0100
Subject: Python in the public eye !
Message-ID: <000015b0@bossar.com.pl>

From: Drew Whitehouse <Drew.Whitehouse at anu.edu.au>



Anyone who visits the Powerhouse Museum (Sydney, Australia) might be
interested to know that the "Wedge" virtual environment exhibit in the
new "Universal Machine" permanent exhibition which opened last night
is Python powered.  The whole program for the exhibit was put together
over the last fortnight on top of our in house "pSpace" VE system
which use python as it's embedded language. I got to watch primary
school kids using it for the first time today and it was a real
buzzzz. Thanks Guido and Co !!!

	http://www.phm.gov.au/universal/
	http://anusf.anu.edu.au/pSpace/


-Drew

-- 
;; mailto:Drew.Whitehouse at anu.edu.au   http://anusf.anu.edu.au/~drw900/
;; Viz programmer, Australian National University Supercomputer Facility



From gmcm at hypernet.com  Tue Jul  6 00:52:54 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 5 Jul 1999 23:52:54 -0500
Subject: file.tell problems (1.5.2/Win32)
In-Reply-To: <37808C26.1693E239@phoenix-edv.netzservice.de>
Message-ID: <1280889249-62599556@hypernet.com>

Holger Jannsen wrote:
> 
> Greg Ewing schrieb:
> > 
> > Joshua Rosen wrote:
> > >
> > > This leads me to believe that `append mode' is intended to `append' at the
> > > *current* point in the file, which should, for consistency's sake, start at 0.
> 
> I think there is the answer: "Some Unix-Systems" properly would
> append the write-string directly at the current file position. 
> 
> > I just tried an experiment (on Windows) using "a+"
> > mode. It seems that f.tell() is telling the truth,
> > but the truth is strange.
> > 
> > What "a"/"a+" modes seem to mean is "open the
> > file positioned at 0, but if the first operation
> > is a write, go to the end of file before doing
> 
> I think even this is `bizarre`!
> Because you could 1st work with "read(1)" or readline()
> on the file, getting correct fileposition (tell), but then a
> write even wouldn't work like expected!;-( It doesn't work!!
> Take the test:
> Create a small file with e.g. 3 lines a 10 characters.
> >>> f=open('test.txt','a+')
> >>> f.tell()
> 0
> >>> f.read(12)
> '111111111111'
> >>> f.write("sssss")
> >>> f.tell()
> 17
> >>> f.close()
> You won't even see the obviosly written string inside the file!! I
> think, here's something wrong!?
> 
> > it". Which seems quite bizarre to me - wouldn't
> > it be simpler to just go to the end of the
> > file on opening?

The MSVC help text says """\
When a file is opened with the "a" or "a+" access type, all write
operations occur at the end of the file. The file pointer can be
repositioned using fseek or rewind, but is always moved back to the
end of the file before any write operation is carried out. Thus,
existing data cannot be overwritten.
"""

This appears to say the same as the (Linux) man page for "open" when 
it discusses "O_APPEND". And the MSVC help text makes it clear that 
"a" and "a+" in fopen correspond to the use of "O_APPEND" in the 
flags to open.

The man page for "fopen" notes that """\
Note that ANSI C requires that a file poistioning function intervene 
between output and input, unless an input operation encounters 
end-of-file. (If this condition is not met, then a read is allowed to 
return the result of writes other than the most recent.)
""" 

But it never explicitly discusses whether writes can take place at 
anything other than end of file. I have only used "a+" when I wanted 
that behavior, and "r+" when trying to do something as you do above.

So the only deviation I detect in Windows behavior is that an ftell() 
on a just opened non-empty "a+" file returns 0 instead of 
end-of-file.

> You've seen: I could even read on a 'a+'-mode-opened file
> getting new file-position.
> And there are 'some Unix-OSs' that do the IMHO right thing to 
> append directly at the current file-position without jumping
> to the end of file.
> 
> Ciao,
> Holger
> [Python1.52, WinNT4.0]
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list

- Gordon



From kas at maps.magnetic-ink.dk  Fri Jul 23 00:52:40 1999
From: kas at maps.magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Fri, 23 Jul 1999 06:52:40 +0200
Subject: Python, Linux 2.0, GCC 2.7.2.3 and threading
References: <7n8m6a$c7f$1@nnrp1.deja.com>
Message-ID: <slrn7pft8o.cmp.kas@titan.magnetic-ink.dk>

Preston Landers <mithy at my-deja.com> wrote:

> A friend of mine posted here recently describing a problem he was
> having running my program Pagecast under Linux 2.0.  Pagecast is
> a Python program that uses several threads.  He was using Linux
> 2.0.36, glibc 2.0.7, and Python 1.5.2.  I can't confirm at this
> time what compiler version he was using (to compile Python) but
> I think it was probably GCC 2.7.2.3.

It was indeed gcc 2.7.2.3.  I will try and recompile Python using egcs and
see if that works, thanks for the hint.


  //Klaus

-- 
???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????



From aahz at netcom.com  Tue Jul 27 19:21:11 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 27 Jul 1999 23:21:11 GMT
Subject: newbie namespace question
References: <m3u2qp9981.fsf@ghostplanet.dok.org>
Message-ID: <7nlet7$pj2@dfw-ixnews10.ix.netcom.com>

In article <m3u2qp9981.fsf at ghostplanet.dok.org>,
Chris Green  <sprout at dok.org> wrote:
>
>How can a subfunction of a method of a class access the other members
>of a class? I'm playing with pygtk and more or less munging what was
>already in the test gtk example.
>
>class TimeCalc:
>	def __init__:
>		[ snip ]
>		self.make_list()
>	def make_list():
>		s=[0 , 0]
>
>		def select_clist(self, row, col, event, selected=s):
>			print self.get_text(row,col)	
>               self.m_clist.connect("select_row", select_clist)
>	def grok_data():
>		pass
>
>Ok. Excuse the bad example but when select_clist is called, how can I
>access the previous self so that I can use self.grok_data().

If m_clist.connect() is *not* passing "self" to select_clist(), I don't
think there's anything you can do.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From greg.ewing at compaq.com  Thu Jul  1 17:05:46 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 02 Jul 1999 09:05:46 +1200
Subject: total idiot question: +=, .=, etc...
References: <000d01bec11f$8f966e20$e19e2299@tim> <377847EE.D5B97A23@compaq.com> <5l67476yd2.fsf@eric.cnri.reston.va.us>
Message-ID: <377BD82A.F556CA@compaq.com>

Guido van Rossum wrote:
> 
> Oh, but that check *is* being made!

No, it's not. The check I'm talking about would
cause the call to A.foo in class C below to
fail. (And it doesn't -- I've just tried it.)

class A:
    def foo(self):
        print "foo of A"
        
class B(A):
    def foo(self):
        print "foo of B"
        A.foo(self)
        
class C(B):
    """Fred Coder originally wrote this
    so it was C(A), then changed it and
    forgot to update the method."""
    def foo(self):
        print "foo of C"
        A.foo(self)

c = C()
c.foo()

Do you see what I mean now?
Greg



From timo at prompt.deANTISPAM  Thu Jul  1 05:00:52 1999
From: timo at prompt.deANTISPAM (Timo Rainio)
Date: Thu, 1 Jul 1999 11:00:52 +0200
Subject: [Q] dbase-file/clipper-index read-module
Message-ID: <930819652.351193@services>

hello world,

anyone around who knows about a Python-module for dbase-file (.dbf) access?
support for clipper-style NTX-index structures would be great. I only want
to _READ_ clipper-files here.

TYIA!

.Timo





From greg.ewing at compaq.com  Wed Jul 28 19:28:17 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 29 Jul 1999 11:28:17 +1200
Subject: global and import
References: <XFMail.990728113603.nathan@islanddata.com>
Message-ID: <379F9211.E65CD824@compaq.com>

Nathan Clegg wrote:
> 
> I would like to move some of the functions to separate
> modules for several reasons, but need to maintain just the two global
> variables between them.  Can this be done?

Yes, but you have to be careful how you refer to the
global variables from other modules. Always import
the name of the module and use modulename.varname --
don't use the "from" form of input. E.g.

# Module Foo
fred = 42

# Module Blarg
import Foo
print Foo.fred
Foo.fred = 17
print Foo.fred

Note that this will NOT work:
# Module Blarg
from Foo import fred
fred = 42 # BAD

That will create a new variable called fred in
module Blarg and assign to that, not the one
in Foo.

A safer way to deal with variables in other modules
is not to munge them directly, but to import accessor
functions, e.g.

# Module Foo
def get_fred():
  return fred

def set_fred(x):
  global fred
  fred = x

# Module Blarg
from Foo import get_fred, set_fred
print get_fred()
set_fred(17)
print get_fred()

Encapsulating access to globals in this way makes
your code more flexible. You can change your mind
about which module fred is in, or even whether it's
a module level variable or kept somewhere else,
without having to track down all the explicit
references to Foo that you would have scattered
about otherwise.

Greg



From hinsen at cnrs-orleans.fr  Fri Jul 16 04:20:08 1999
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 16 Jul 1999 10:20:08 +0200
Subject: Pickling C objects
References: <378B27A9.4201@cs.bham.ac.uk>
Message-ID: <m31ze9gg5j.fsf@chinon.cnrs-orleans.fr>

Stuart Reynolds <S.I.Reynolds at cs.bham.ac.uk> writes:

> I'm writing some new Python classes with a C implementation. How do you
> make a C object pickleable?

Check the copyreg module. You have to provide code to create a string
representation from your object and to recreate an object from a string
representation, that's all.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------



From greg.ewing at compaq.com  Mon Jul 19 18:29:47 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 20 Jul 1999 10:29:47 +1200
Subject: Killing a thread
References: <7mu6a5$h61$1@nnrp1.deja.com> <7mvml3$u5k$1@nntp6.u.washington.edu> <ePJk3.4$Z11.6082@news-dal.corridex.com> <7n04dm$8ca$1@nnrp1.deja.com>
Message-ID: <3793A6DB.731BDB83@compaq.com>

catlee at my-deja.com wrote:
> 
> - Is it conceivable that a user could make write some resource
> intensive code without using any kind of jump operations?

Yes!

def fib(n):
    return {
        0: lambda: 1,
        1: lambda n=n: fib(n-1) + fib(n-2)
    }[n>=3]()

Although if you include function calls as well 
as jumps, you're probably safe (until Tim comes
up with an even more perverted counterexample).

Greg



From simngolda at voyager.net  Sun Jul 18 01:23:49 1999
From: simngolda at voyager.net (Sim & Golda Zacks)
Date: Sun, 18 Jul 1999 01:23:49 -0400
Subject: upload file with cgi
Message-ID: <7mro7g$kr5@newsops.execpc.com>

I am uploading a file using the cgi module and I can not figure out how
to get the name of the file.
The code I am using is
form=cgi.FieldStorage()
pic=form["picture"]
Now when I print pic it gives me the name of the File  and the actual
file.
When I get pic.value it is the actual picture.
pic.key doesn't give me anything and I have absolutely no clue how to
get the filename.

Please help.

Sim




From clgonsal at kami.com  Fri Jul  9 07:04:51 1999
From: clgonsal at kami.com (C. Laurence Gonsalves)
Date: Fri, 09 Jul 1999 11:04:51 +0000
Subject: VIM and [Python] block jumping
References: <3781347D.CB2FBAB3@starvision.com> <Pine.SOL.4.05.9907051548470.13052-100000@saturn.radsci.uci.edu> <37815B80.B4B51F94@starvision.com>
Message-ID: <3785D753.DB2FF1ED@kami.com>

Angus MacKay wrote:
> 
> I am not looking for indentation but just code navigation.
> 
> when code uses identifiers for blocks (such as "()" in lisp, "{}" is
> most sane languages, "#if/#endif" for C preproc ...) vim can navigate
> blocks very easily (with [{/]} for "{}", [#/]# for "#if/#endif" ...)
> 
> with python this becomes harder. is there any way to do it? (I think I
> could write a function for it but I thought I'd see if it had already
> been done)

That's a neat idea. I just wrote this up, so it hasn't been tested a
lot, but it seems to work. The way I implemented this requires that you
have python support compiled into VIM. It's probably possible to
reimplement this as a pure-VIM script, but the python version is
probably orders of magnitude faster (not to mention being easier to
write). :-)

(I had once implemented the C-style [{ ]} motion keys as vim-scripts
before I realized that VIM already had built-in support for that. They
were quite slow.)

First, make a file called "vimMotion.py":

## begin vimMotion.py ##

import vim
import string
import re

# returns a tuple containing the indentation, and the line number where
we got
# that indentation. That can be different than lineNo if lineNo refers
to an
# empty (ie: only whitespace) line.
def getIndent(lineNo):
    for line in vim.current.buffer[lineNo:]:
        result = getIndentQuick(line)
        if result>=0:
            return (result,lineNo)
        lineNo = lineNo + 1
    return (0,len(vim.current.buffer)-1)
        
# grabs the initial whitespace and the stuff after it into two groups
indentRE = re.compile( r"([\t ]*)(.*)" )

# returns the indentation for a line, or -1 if it can't be determined
(ie:
# only whitespace)
def getIndentQuick(line):
    match = indentRE.match(line)
    if len(match.group(2)) == 0:
        # if the line is just whitespace, we don't know the indentation
        return -1
    else:
        # make sure we expand tabs like python does...
        return len(string.expandtabs(match.group(1)))
        
# if dir is positive, goes down, negative, goes up. Other than that, the
value
# is irrelevant
def blockMotion( dir ):
    lineNo0 = vim.current.window.cursor[0]-1
    indent0,lineNo = getIndent(lineNo0)
    newLine = lineNo0

    if indent0 == 0:
        # this is an optimization
        if dir>0:
            newLine = len(vim.current.buffer)-1
        else:
            newLine = 0
    elif dir>0:
        indent = indent0
        numLines = len(vim.current.buffer)
        while lineNo<numLines and indent >= indent0:
            lineNo = lineNo + 1
            indent,lineNo = getIndent(lineNo)
        newLine = lineNo
    else:
        indent = indent0
        lineNo = lineNo0
        while lineNo>0 and indent >= indent0:
            lineNo = lineNo - 1
            qIndent = getIndentQuick(vim.current.buffer[lineNo])
            if qIndent>=0:
                indent = qIndent
        newLine = lineNo
    # move the cursor
    vim.current.window.cursor = (newLine+1,0)

## end vimMotion.py ##

Then add the following three lines to your .vimrc:

pyfile vimMotion.py
nmap [{ :python blockMotion(-1)<CR>^
nmap ]} :python blockMotion(1)<CR>^

You should probably just execute these when editting a python file.

Note that python is a little bit funny as to where blocks can end,
because multiple blocks can all end at exactly the same position. That
means that doing ']}' can potentially (and often does) take you out of
several blocks. In a language with braces, ]} will always take you out
of one block at a time.

The other tricky part was figuring out the indentation of a line. The
two things that make this tricky are tabs (but string.expandtabs made
that pretty easy), and the fact that python seems to ignore
empty/whitespace-only lines when figuring out block levels. The "right"
thing to do seems to be to scan forward until we find a line with some
non-whitespace on it.

Let me know if you have any problems.

(note: this was posted to both comp.lang.python and comp.editors, since
people in both those groups might be interested in this)

-- 
  C. Laurence Gonsalves              "Any sufficiently advanced
  clgonsal at kami.com                   technology is indistinguishable
  http://www.cryogen.com/clgonsal/    from magic" -- Arthur C. Clarke



From tbryan at arlut.utexas.edu  Sat Jul  3 11:34:19 1999
From: tbryan at arlut.utexas.edu (Thomas Bryan)
Date: Sat, 03 Jul 1999 10:34:19 -0500
Subject: Problem with informixdb and python1.5.2
References: <7ljsbc$srn$1@news.harvard.net>
Message-ID: <377E2D7B.BEBC065C@arlut.utexas.edu>

Ken Liu wrote:
> 
> I've been using the informixdb module (version 1.2a1) with python1.5.1 for a
> while now with no problems.  Just a few days ago we installed python1.5.2,
> and now informixdb won't work any more.  when I attempt to construct a
> Connection object, I get an error about "unable to get name of current
> working directory."  Here's a log of my interaction with the interpreter

Do you now have both python 1.5.1 and python1.5.2 installed?

> which does not make any sense.  Our Informix setup has not changed for the
> last few months, and the informixdb modules continues to work great 
> under python1.5.1.
> 
> Does anyone have any suggestions on what might be wrong?

informixdb might be getting confused about which Python 
version/libraries it should be using.  I'm not sure how to 
determine whether that's the problem, but I know that others
have had problems with various packages when they have two 
versions of Python installed at the same time.

Have you tried recompiling informixdb with the 1.5.2 header files?  
I don't think that too much has changed between 1.5.1 and 1.5.2, but 
it couldn't hurt to try.



From nadim_ghaznavi at my-deja.com  Thu Jul 29 16:54:21 1999
From: nadim_ghaznavi at my-deja.com (nadim_ghaznavi at my-deja.com)
Date: Thu, 29 Jul 1999 20:54:21 GMT
Subject: shelve portability problem
Message-ID: <7nqf1n$68e$1@nnrp1.deja.com>

First of all, I'd like to apologize for posting this twice, I
accidently posted this as part of the "COM bites - help please" thread!
Here it is again:

I'm having a problem moving a Python shelve ODB from Solaris to Linux.
Here's some sample code to demo the problem:

On the Solaris Platform:

# python
Python 1.5.2 (#1, Jun  5 1999, 15:51:52)  [GCC 2.8.1] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import shelve
>>> friends = ['Jane','Jan','Simon']
>>> odb = shelve.open('odbfile')
>>> odb['foo']=friends
>>> odb.close()

Which leaves me with two files:
# ls -l
-rw-r--r--  1 root    other         0 Jul 29 13:50 odbfile.dir
-rw-r--r--  1 root    other      1024 Jul 29 13:57 odbfile.pag

To confirm that all is well, lets have a look at the data again: On
Solaris:

# python
Python 1.5.2 (#1, Jun  5 1999, 15:51:52)  [GCC 2.8.1] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import shelve
>>> samedb = shelve.open('odbfile')
>>> buddies = samedb['foo']
>>> buddies
['Jane', 'Jan', 'Simon']
>>> samedb.close()

As expected, the data is still in the odb.  Next we'll move the
odbfile.dir and odbfile.pag to the Linux platform (I used FTP, nothing
strange there).

# ls -l
-rw-r--r--  1 root    root           0 Jul 29 13:59 odbfile.dir
-rw-r--r--  1 root    root        1024 Jul 29 13:59 odbfile.pag
# python
Python 1.5.1 (#1, Mar 21 1999, 22:49:36)
[GCC egcs-2.91.66 19990314/Li on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import shelve
>>> nodb = shelve.open('odbfile')
>>> nodb.keys()
[]
>>> nodb.close()
>>>
# ls -l
-rw-r--r--  1 root    root        4096 Jul 29 14:00 odbfile
-rw-r--r--  1 root    root           0 Jul 29 13:59 odbfile.dir
-rw-r--r--  1 root    root        1024 Jul 29 13:59 odbfile.pag

What seems to be happening is that Solaris is storing the data in an old
DBM style database, and Linux is using a new gdbm format.  When I try
opening the old Solaris ODB, python incorrectly opens a new gdbm ODB
instead.

A couple of things to note:  I do have the free gdbm package installed
on my solaris platform (even though Python isn't using it):

# pkginfo GNUgdbm
application GNUgdbm        gdbm

I know I can work around this by using pickle and a specific DB module,
but does anyone know how I can get it working using the shelve module?

                  Nadim


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From gregm at iname.com  Thu Jul  8 22:22:42 1999
From: gregm at iname.com (Greg McFarlane)
Date: Fri, 9 Jul 1999 12:22:42 +1000
Subject: Problem with BLT busy cursor on NT
In-Reply-To: <19990707173207.62192@nms.otc.telstra.com.au>; from Greg McFarlane on 07 Jul 1999 at 05:32:07PM
References: <19990707173207.62192@nms.otc.telstra.com.au>
Message-ID: <19990709122242.35379@nms.otc.telstra.com.au>

For those interested, here is the reply from George Howlett, author of
BLT, which he posted to comp.lang.tcl.  The problem is with cursors in
Tk under NT and is not specific to BLT.

>From George A. Howlett <gah at cadence.com>:
> Use "update" instead of "update idletasks". 
>  
> The transparent busy window needs to be mapped, raised, and resized
> each time you "hold" a widget busy.  Remember also that Windows,
> unlike X11, doesn't manage cursors on an individual window basis.  Tk
> emulates this behavior by watching Enter/Leave events in Tk windows. 
> A further compilication is that Windows doesn't immediately change the
> cursor when it's set.  Instead, it waits for the mouse to be moved or
> the next system call.

In python, the answer is:
> Use update() instead of update_idletasks(). 

On 7 Jul, Greg McFarlane wrote:
> Has anyone noticed that, under some circumstances, the BLT busy cursor
> is not displayed under NT?  The (tcl) code below demonstrates this. 
> (The equivalent python has the same problem, so it is not a problem
> with python, but someone on that mailing list may know that answer.)
> 
> Clicking on the top "Show busy" button changes the text of the button
> as well as displaying the busy cursor for two seconds, as expected. 
> Under Unix, clicking on the second button does the same thing, but
> under NT, it only changes the text of the button for two seconds and
> does not display the busy cursor.
> 
> It seems that, under NT, to actually get the busy cursor displayed,
> the interpreter has to return to the main loop after a call to
> blt::busy.  If there is a long lasting calculation or a blocked read,
> etc, then the cursor will not be not changed.  I would have thought
> that the call to "update idletasks" would have been enough to flush
> out the cursor change.
> 
> Has anyone seen this before?  Is it a bug in BLT or a limitation
> of NT?
> 
> ==================================================================
> package require BLT
> 
> proc startBeingBusy {} {
>     .b1 configure -text "I am busy..."
>     ::blt::busy hold .
>     update idletasks
>     after 2000 stopBeingBusy
> }
> 
> proc stopBeingBusy {} {
>     ::blt::busy release .
>     .b1 configure -text "Show busy"
> }
> 
> proc beBusy {} {
>     .b2 configure -text "I am busy..."
>     ::blt::busy hold .
>     update idletasks
>     after 2000
>     ::blt::busy release .
>     .b2 configure -text "Show busy"
> }
> 
> button .b1 -text "Show busy" -command startBeingBusy
> pack .b1 -padx 40 -pady 40
> 
> button .b2 -text "Show busy" -command beBusy
> pack .b2 -padx 40 -pady 40
> ==================================================================

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From pvelikho at cs.ucsd.edu  Fri Jul  9 01:06:53 1999
From: pvelikho at cs.ucsd.edu (Pavel Velikho)
Date: Thu, 08 Jul 1999 22:06:53 -0700
Subject: PyApache and Apache 1.3.6 anyone ???
References: <19990708221333.B16851@hoffleit.org>
Message-ID: <3785836D.3BE3EBC7@cs.ucsd.edu>

Gregor Hoffleit wrote:

> Is anybody successfully running PyApache on an Apache 1.3.6 ?
> Preferably on a Linux system ? When I build PyApache and load it from
> httpd.conf, /usr/sbin/apache just doesn't start up. It doesn't give
> any error or what, it just starts up and quits immediately.

>
> Any experiences ?

Did you check the error log? If it is a configuration error it will
usually keep quiet about
it and just write it in the log file.
Hope this helps

>
> Thanks in advance,
>
>         Gregor

Pavel




From mnot at pobox.com  Mon Jul 19 00:37:40 1999
From: mnot at pobox.com (Mark Nottingham)
Date: Mon, 19 Jul 1999 04:37:40 GMT
Subject: cgi - push a page from a select menu
References: <3792912B.86C3584E@am.appstate.edu>
Message-ID: <007601bed1a0$6fc0ee40$0301a8c0@mnot.net>

Try:

import cgi

form = cgi.FieldStorage()
if form.has_key("select"):
    print "Location: %s" % (form['select'].value)
    print
else:
    print "Content-type: text/html"
    print
    print "<HTML><BODY><H1>What?</H1></BODY></HTML>"



----- Original Message ----- 
From: TMGB <bennettt at am.appstate.edu>
Newsgroups: comp.lang.python
To: <python-list at cwi.nl>
Sent: Monday, July 19, 1999 12:44 PM
Subject: cgi - push a page from a select menu


> I am trying to use python to receive a url from a form select menu and
> then send that page back to the browser.  All three of my attempts so
> far leave this error message on the web server: 
> 
> [18/Jul/1999:22:06:34] failure: for host ....library.appstate.edu trying
> to POST /CGI-NT_Python/goto.pyc, cgi-parse-output reports: the CGI
> program D:\...(path to python)\Python\python.exe did not produce a valid
> header (program terminated without a valid CGI header (check for core
> dump or other abnormal termination) 
> 
> I also get this on a web page for the third attempt: AttributeError:
> open_http 
> 
> The python paths are correct, I have another python form using these
> paths.  I would think that the urllib would send the url requested with
> its header( if I use it correctly).  I tried printing a header and still
> get this error.  The URL of the page with the selector is:
> 
> http://www.library.appstate.edu/saat/forms/test/
> 
> 
> # Author: Thomas McMillan Grant Bennett
> # webmaster at www.library.appstate.edu
> # Python script to accept a url from a form and go to that URL
> # Filename: goto.py
> 
> import cgi
> import urllib
> 
> form = cgi.FieldStorage()
> form_ok = 0
> if form.has_key("select"):
> 
> # third attempt
> #   html_page=urllib.open_http(form["select"].value)
> #   print html_page
> 
> # second attempt  
> #   html_page=urllib.urlretrieve(form["select"].value)
> #   print html_page
> 
> # first attempt
> #   print(read(urllib.urlopen(form["select"].value)))
>    
> 
> 
> ----------------------------------------------------------------------
> Thomas McMillan Grant Bennett           Appalachian State University
> Computer Consultant II                  University Library
> bennettt at am.appstate.edu               
> http://www.library.appstate.edu/admin/
> Voice:  828 262 6587 FAX:    828 262 3001
> 
> Windows 95 is a 32-bit extension to a 16-bit patch for an 8-bit
> operating system that was originally coded for a 4-bit microprocessor.
>  - Chris Dunphy     Boot Magazine
> 
> 
> 




From avoinov at raima.com  Mon Jul 26 13:27:43 1999
From: avoinov at raima.com (Alexander V. Voinov)
Date: Mon, 26 Jul 1999 10:27:43 -0700
Subject: Configure and cross-compiler
References: <379C6E3E.78CEB6DD@onera.fr>
Message-ID: <379C9A8E.66FF66A9@quasar.ipa.nw.ru>

Marc POINOT wrote:

> I've compiled Python for a  NEC super-computer,
> using a cross compiler.
>
> The configure script failed for the tests, it
> leaves with a exit 1 emergency-escape-do-not-use-elevators.
> Maybe a slight modification can allow
> to produce Python with less verification, at least
> for cross-compilation.
>
> I've replaced the corresponding exit 1 with "true".

I did the same when compiled Python for mingw32 using a crosscompiler.
As memory serves me, I even posted the patches I needed to apply.

Alexander





From sdossey at dolphin.openprojects.net  Sat Jul 10 19:13:02 1999
From: sdossey at dolphin.openprojects.net (Scott Dossey)
Date: Sat, 10 Jul 1999 23:13:02 GMT
Subject: Dreams.
Message-ID: <7m9to0$6qg$74@rks1.urz.tu-dresden.de>

Okay, I am new to the newsgroup, but I thought I'd post something
interesting that I've just started working on, and get some feedback.  I've
been toying around with the idea of implementing "dreams" into the Python
source code.  

What are dreams, you say?  Look at the following python code fragment for an
idea, first of all, of how they are put into code:

#START PSEUDOCODE FRAGMENT
import sys
import dreams

aDream=dreams.dream()
try:
	#Insert code here.
except: 
	aDream.wakeup()
	sys.exit(1)

aDream.realize()
#END PSEUDOCODE FRAGMENT

The code, at the location denoted by "#Insert code here." would run, but if
an exception was thrown, it will have been as if the code had "happened in a
dream".  No variables will have been changed/deleted.  Nothing will have
changed from before the "dream" in terms of Python's variable state.
Ideally, (but this is much harder), all file changes, all things that are
"possible" to return to their original state, would be returned as well.
(You can't take back network traffic).  

Dream.wakeup() would restore to the state before the dream, and continue
execution from that line forward.  Dream.realize() would commit to the
dream, stop saving change-state, and delete the change state it had already
recorded. 

Before you start telling me about the numerous interaction problems this
will cause, and how impossible it will be to implement, think of what it
does to certain types of error handling!  You don't have to back out changes
when your change to a transactional database goes awry midway through the
change.  Even though functionality would be limited, (not all things could
or would be backed out), any automatic handling (even partial) of these
types of problems could be a huge blessing for certain types of programming.

Here's how I was thinking about implementing them (If you are not familiar
with python's C implementation this may be unfathomable to you):

The easiest way to do it that I can see is the following:
	      
	Step One:  Put a wrapper around each variable and function pointer
		in any PyObject and it's contained structures (IE,
		PyTypeObject), so that any reference to any object can be 
		"caught".  This would involve tweaks all over the python
		codebase to deal with the new way of accessing variables.
		I'll probably do this by changing the name and method of 
		access for everything inside a PyObject and letting the
		compiler find all the millions of references, then fixing these
		up to use the new wrappers.  It's too bad there are not access
		methods for all these things already, rather than being 
		indexed directly.

        Step Two:  Call pickling code inside those wrappers to save the
		state of the object before and after the calls.  Write the
		diffs out to the dream code so it can store it in a "dreamlog".
	        Deletions of objects will similarly "pickle" the objects and
 		store them in the dreamlog.  If a pickle of an object before
		it is referenced doesn't work, then
		"abort" the dream by restoring the state and
		throwing a "DreamException".
		
	Step Three:  Fix up more objects so they can be pickled, so that
		dreams are mode useful.  IE, make files pickleable, modules
		pickleable, text screen state and TK graphical screen state
		pickleable and restoreable, etc. 

	Step Four:  Major debugging, testing, and module usage documentation.
	
I'd probably include dreaming as a compile time option of python, since
support for it will cause some pretty significant code slowdown of python,
even when not "dreaming".

However, there are a couple of major problems with my approach:
1.	They break every C module known to man in a way that's not too hard
to fix, but will have to be done for every C module that is going to support
dreams.

2.  	Python will be slowed down.  Every call/variable reference 
	will have a little more overhead when not dreaming, A LOT when 
	dreaming.

3.	It's very invasive.

I'll take feedback in terms of ideas, problems, feasibility statements,
flames,  whatever.  Most of all, I want help!  I think this would add a very
interesting and useful feature to Python (as perhaps, a compile time option).
Guido, I know you're out there, your feedback would be most appreciated,
along with the feedback of any other Python maintainers.

-Scott Dossey
sdossey at armored.net, sdossey at openprojects.net




From sreekantk at freenet.co.uk  Tue Jul 20 14:56:02 1999
From: sreekantk at freenet.co.uk (sreekant kodela)
Date: Tue, 20 Jul 1999 19:56:02 +0100
Subject: Python Performance
Message-ID: <3794C642.AB6A3E5D@freenet.co.uk>

Dear all

I finally completed my website. It is fully functional and I used
PostgreSQL, Python and Apache for that. Before I actually put the thing
on the net,

- Is there anyway I can improve the performance of python code. It is
vital because almost all of the website is dynamically generated except
intial welcome / logon screens etc and images. PostgreSQL being slow
doesn't make things easier either.

I know there is Pyapache. I am looking for it's binary. Is there any
other way of doing it faster ?

Help is much appreciated
Thanks
sreekant





From ralph.gauges at eml.villa-bosch.de  Wed Jul 14 01:34:07 1999
From: ralph.gauges at eml.villa-bosch.de (Ralph Gauges)
Date: Wed, 14 Jul 1999 07:34:07 +0200
Subject: Recursive method
References: <378B499B.4301F893@eml.villa-bosch.de> <378BAD30.7D1A5196@hvision.nl>
Message-ID: <378C214F.3CF9CDD3@eml.villa-bosch.de>

Hans Nowak wrote: 
> You have a nested function here. (I say function, because it does not
> look like a method... it's missing the self argument!



Actualy it is a method, but JPython doesn't seem to have a
self. I was missing that as well.



> 
> So, to solve this, you could try a different approach... maybe
> 
> (you said it's a method, right?)
> 
> class Blahblah:
>     ...
>     def makeTree(self):
>          ....
>     def makenodes(self, C):
>          ....
> 



I tried that as well, but I got the same error message. Only
after I moved the makenodes outside the class it worked.
This might also be related to the missing 'self' in JPython
since I can not call 'self.makenodes'.
Since moving makenodes out of the class is actualy not what
I want, I was wondering, if anybody has had a similar
problem and found a solution.
 


Thanks for the answers. Although it is still not, what I
want it to be, it works and I can go on, till I (or someone
else) find a better solution.

Ralph



From Tim.Peters at p98.f112.n480.z2.fidonet.org  Sat Jul  3 06:38:05 1999
From: Tim.Peters at p98.f112.n480.z2.fidonet.org (Tim Peters)
Date: Sat, 03 Jul 1999 11:38:05 +0100
Subject: ordered threading
Message-ID: <000016b1@bossar.com.pl>

From: "Tim Peters" <tim_one at email.msn.com>

[Robin Becker]
> I have an unbounded number of worker threads which occasionally require
> access to a unique server thread. I can easily mutex access to the
> server thread, but is there any obvious/easy way to ensure that the
> blocked workers wake up in the right order?

No threading system on earth allows specifying the order in which blocked
threads will wake up, although some systems (not Python's) allow you to
*influence* the order via assorted "priority" gimmicks.

You could easily enough build an OrderedLock class whose acquire and release
methods implement whatever scheme you want, hiding the unordered nature of
an internal mutex under a layer of serial-number cruft.  But then why not do
these things serially in a single thread to begin with?

If you have frequent rigid ordering requirements, threads are likely the
wrong approach.

> Also if blocked can a thread be killed in any way?

No, not short of killing the whole program.  A thread blocked on a condition
that will never become true is a symptom of confused application design, not
of a weakness in the thread model <wink>.

expand-the-condition-to-include-"or-i-should-die-now"-ly y'rs  - tim






From arcege at shore.net  Thu Jul 15 17:28:20 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 15 Jul 1999 21:28:20 GMT
Subject: Terminating Python...
References: <BFC0587634C3D111AACB00805FA766FA02C1C652@exchny12.ny.smb.com> <378CFDDF.489CDCEF@compaq.com> <7mj8s9$op2@newsops.execpc.com> <004901bece61$41270b60$f29b12c2@secret.pythonware.com> <378DD196.BAB576C6@njeb.att.com> <87u2r5lj96.fsf@meryllon.death-gates.fr>
Message-ID: <Unsj3.135$ui4.39372@news.shore.net>

Sebastien Tanguy <sebastien.tanguy at wanadoo.fr> wrote:
:>>>>> "AF" == Alex Flinsch <aflinsch at njeb.att.com> writes:

: AF> Fredrik Lundh wrote:

:>> in which way is typing "raise SystemExit<Return>" better
:>> than typing "<Control-D>" (which is documented in the
:>> Python Tutorial) ?
:>> 

: AF> "raise SystemExit" could be embedded in some python code, while a
: AF> <Control-D> could not be.  "raise SystemExit" would work on any
: AF> OS, no need to check if we are running windows or something else.

: and what about :

: import sys
: sys.exit(0)

: Doesn't it work on every platform ?

FYI: "sys.exit(N)" is just "raise SystemExit, N".

  -Arcege




From tre17 at pc142.cosc.canterbury.ac.nz  Thu Jul 15 19:16:53 1999
From: tre17 at pc142.cosc.canterbury.ac.nz (Timothy R Evans)
Date: 16 Jul 1999 11:16:53 +1200
Subject: Oneliners?
References: <378D99C4.BDACE15B@serop.abb.se>
Message-ID: <c75iu7ljyfu.fsf@pc142.cosc.canterbury.ac.nz>

"Thomas S. Strinnhed" <thstr at serop.abb.se> writes:
> 
> I heard there was a webserver in one line of Perl,
> does Python match that (or two or three...)?
> 

Shouldn't be too hard, if you count:

import SimpleHTTPServer

as only one line!

--
Tim Evans



From wtanksle at dolphin.openprojects.net  Wed Jul  7 15:39:43 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Wed, 07 Jul 1999 19:39:43 GMT
Subject: What's ZOPE anyway?
References: <3scg3.1267$B95.91660@typ42b.nn.bcandid.com> <7m064a$921@journal.concentric.net>
Message-ID: <slrn7o7b7u.6la.wtanksle@dolphin.openprojects.net>

On 07 Jul 1999 11:26:18 PDT, Christopher Petrilli wrote:
>Jayr Magave <jmagave at amazon.com.br> wrote:
>> I went through www.zope.org and read everything I could, and found the
>> following definition:

>from an architecture perspective is Apple's Web Objects.   JUnfortunately,
>a 30 word snippet doesn't provide any indication of the power.  I have
>taught 3 day classes only as an "introduction" to the concepts, hardly
>expert. There's just that much there.

...and what's more is that it's EASY to use.  When I started I didn't know
much HTML (only the basics).  Thanks to Zope, I can learn bit by bit.

Go ahead, download the version for your platform and install it.  It's
trivial.  Then point your web browser at the address indicated in the
instructions, and start playing with it.

>Chris

-- 
-William "Billy" Tanksley



From templon at jlab.org  Wed Jul 28 23:25:00 1999
From: templon at jlab.org (Jeff Templon)
Date: 28 Jul 1999 23:25:00 -0400
Subject: where is nsremote?
Message-ID: <yhs9080jfxv.fsf@haplix8.jlab.org>

Hi,

I found a tool called nsremote by Scott Hassan that does something I
want.  However, the link is broken on the python www site.  Does anyone
know where I can download this?

				JT



From invalid.address at do.not.email  Wed Jul 14 18:44:18 1999
From: invalid.address at do.not.email (guppy)
Date: Wed, 14 Jul 1999 22:44:18 GMT
Subject: wxPython Form editor
References: <378c488e@news.xtml.co.uk> <378c8b3c@news.xtml.co.uk>
Message-ID: <378d128f.24084502@news.bctel.ca>

On Wed, 14 Jul 1999 14:13:34 +0100, "Nick Trout"
<nick@#spam_me_not#videosystem.co.uk> wrote:

>Is there a wxPython mail group/news group?

There is, and I'm reasonably sure that you can sign up through the Website
or by checking out the documentation.  Must be possible one of those ways,
'cause the mailing list is active and growing...


-- 
MSpeak: in-no-va-tion, n. 1. extending or modifying
existing standards in undocumented ways. 2. creating
new, undocumented standards. --de-com-mod-it-ize, v.



From jsight at mindspring.com  Sun Jul 11 15:40:52 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Sun, 11 Jul 1999 15:40:52 -0400
Subject: Python Code Repository?
References: <378645AA.5430A026@msfc.nasa.gov> <19990709163127.A1964@toast.internal> <7m9tnb$6qg$54@rks1.urz.tu-dresden.de>
Message-ID: <3788F344.A62F24D@mindspring.com>

Will Ware wrote:
> 
> jam (jam at newimage.com) wrote:
> : I don't think anything as nice as CPAN exists for python ...
> : I know enough about python and cgi scripting and sql databases to come
> : up with something pretty quick, and all we'd need is submissions. there
> : might already be some examples of code snippets out on the net in various
> : places..
> : does anyone have any further suggestions or ideas on how we can get this
> : rolling?
> 
> There is alot of merit to the idea, and probably if you implement it,
> getting contributions won't be a problem. It might be worthwhile to
> set up a voting system that allows people to rate contributions for
> interesting-ness. You might look at the slashdot.org moderation scheme,
> that guy has thought long and hard about the problem of rating content
> in a way that is both fair and informative.

The pyPit - http://www.biddin.com/pyPit/

Ok, I've put up a submission page for a CPAN style system on my site. 
:-)  I've decided at least in the short-term that I will use a Gadfly
database for the actual libraries, and will maintain mirrors of all
submitted software locally.  There will probably be "Freshmeat-style"
product pages and comments pages in the future, but that will have to
wait.  Wait on me to code them that is.  :-)

Anyway, the current stuff is very simplistic, but if you guys start
submitting things, I'll start taking them.  Also, if anyone has a Python
project, but needs webspace, just let me know.  I have a fairly good bit
of extra space right now, so I should be able to take quite a few.  :)

Enjoy!

-- 
---------------
Jesse D. Sightler
http://www3.pair.com/jsight/

"Do not use a hatchet to remove a fly from your friend's forehead." 
      - Chinese Proverb



From aahz at netcom.com  Sat Jul 10 19:21:58 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 10 Jul 1999 23:21:58 GMT
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu>
Message-ID: <7m8kim$p9f@dfw-ixnews3.ix.netcom.com>

In article <3787D36A.815B7C29 at ssec.wisc.edu>,
Nick Bower  <nick.bower at ssec.wisc.edu> wrote:
>
>I feel really dumb asking this question, but how do I change the current
>working directory?  For browsing the file system I mean.

os.chdir()
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From ralph.gauges at eml.villa-bosch.de  Tue Jul 13 10:13:47 1999
From: ralph.gauges at eml.villa-bosch.de (Ralph Gauges)
Date: Tue, 13 Jul 1999 16:13:47 +0200
Subject: Recursive method
Message-ID: <378B499B.4301F893@eml.villa-bosch.de>

Hello,

I am trying to generate a JTree (in JPython) out of some
Data I have. The data I have is structured in classes which
have a list of classes that have a list of classes and so
on. Now I want to call the makenodes method with the top
level class and have the function follow the tree
recursively (see code below). Unfortunately all I get is a
'NameError' for makenodes, when the function is supposed to
call itself. This seems to be a problem with the namespace,
but I have no idea, how to solve this. Any help would be
appreciated.
Thanks a lot in advance


Ralph



	def makeTree():
                .
                .
                .
            def makenodes(C):
               
N=javax.swing.tree.DefaultMutableTreeNode(C.Name)
                Children=C.getChildren()
                for Child in Children:
                    N.add(makenodes(Child))
                return N
            
            for Class in M.Classes:
                ClassNode.add(makenodes(Class))
                 .
                 .
                 .



From hochberg at psn.net  Mon Jul 19 15:42:32 1999
From: hochberg at psn.net (Tim Hochberg)
Date: Mon, 19 Jul 1999 13:42:32 -0600
Subject: Jpython Question
References: <Pine.GSO.4.05.9907191525580.20433-100000@allegro>
Message-ID: <k3Mk3.1333$J3.30561@news15.ispnews.com>

>I had a quick question on Jpython and I'm not sure if this is the right
>place to post, but I figured I would try anyway...


You might want to subscribe to jpython-interest mailing list. (See
www.jpython.org for subscription information)

>I was just wondering if jpython can instantiate java classes and invokes
>methods on them...any help would be welcome...


Yes and yes. I'm always amazed by just how well the Java/Python integration
works in JPython (Kudos to Jim and Barry).

-tim





From paul at prescod.net  Wed Jul 14 01:10:45 1999
From: paul at prescod.net (Paul Prescod)
Date: Wed, 14 Jul 1999 05:10:45 GMT
Subject: XML-RPC
References: <931899884snz@vision25.demon.co.uk>
		<378BBD4D.9DA43FE5@prescod.net> <slrn7ont1m.i9e.bwinton@tor.dhs.org>
Message-ID: <378C1BD5.448C8F5@prescod.net>

Blake Winton wrote:
> 
> On Tue, 13 Jul 1999 22:27:25 GMT, Paul Prescod <paul at prescod.net> wrote:
> >Phil Hunt wrote:
> >> Does anyone know if an XML-RPC server exists for Python?
> >> There's a python XML-RPC client at www.pythonware.com/products/xmlrpc/
> >> but I'm not aware of a Python server).
> >I think that you want to look at Zope.
> 
> I think that's a bit misleading...  The XML-RPC client also contains an
> implementation of an XML-RPC server.  You don't need to bring Zope into
> the picture at all.

Sorry, I interpreted the question as "how do I publish Python objects
through XML-RPC". I'm fuzzy on what a "RPC server" is. Here's my
interpretation: as a generic FTP server serves files, so an RPC server
should serve callable object return codes and exceptions? Zope would do
that. xmlrpcserver is more like an "FTP server shell" which doesn't serve
files but can be extended to...a BaseRPCServer like BaseHTTPServer.

But I admit that the distinction between shell and server is blurry. Now
the questioner has the information necessary to make his decision.

-- 
 Paul Prescod  - ISOGEN Consulting Engineer speaking for only himself
 http://itrc.uwaterloo.ca/~papresco

To me programming is more than an important practical art. It is
also a gigantic undertaking in the foundations of knowledge.
	- Grace Hopper



From cs at ctzen.com  Fri Jul  9 11:18:37 1999
From: cs at ctzen.com (chacs)
Date: Fri, 09 Jul 1999 15:18:37 GMT
Subject: Help! File Locking
References: <37856BF1.BD046003@freenet.co.uk>
Message-ID: <7m9tmf$6qg$27@rks1.urz.tu-dresden.de>

Hi All,

Need some help in file locking.

x1.py:
  import fcntl, FCNTL
  f = open('myfile', 'r+')
  fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
  input()
  x = str(int(f.readline()) + 1)
  f.seek(0)
  f.write(x)
  fcntl.flock(f.fileno(), FCNTL.LOCK_UN)
  f.close()

x2.py:
  import fcntl, FCNTL
  f = open('myfile', 'r+')
  fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
  x = str(int(f.readline()) + 1)
  f.seek(0)
  f.write(x)
  fcntl.flock(f.fileno(), FCNTL.LOCK_UN)
  f.close()

Both script are the same except x1 has an additional input() statement.

Basically, what the script does is to open a file, lock it and update
the first line to the next higher number.

My problem is (assume file content is "0"):

1. in shell 1, run x1.py, x1 got lock and block at input().
2. in shell 2, run x2.py, x2 block at flock().
3. in shell 1, input something to unblock input().
4. x1 update "0" to "1", unlock and exit.
5. x2 gets lock, BUT instead of getting the new file content (i.e. "1")
written by x1, it got the old content and update the file to "1" again
(instead of reading "1" and update to "2").

Any idea what is wrong ?

Thanx in advance.

C S Chang
cs at ctzen.com


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From slj at pixar.com  Thu Jul 15 15:01:11 1999
From: slj at pixar.com (Steve Johnson)
Date: Thu, 15 Jul 1999 12:01:11 -0700
Subject: [Python-Help] Calling superclass method with variable args?
References: <378E29DD.1613913E@pixar.com> <378E2CC3.D9A5E187@digicool.com>
Message-ID: <378E2FF7.A4110373@pixar.com>

Hi Jim,

You're a God!  I thought I'd tried that, but trying it now, it does indeed
seem to work.

I don't understand why morekeyargs doesn't become a single dictionary
argument to Task.__init__,.  I'll read the FAQ.

Thanks very much!

Steve

> Steve Johnson wrote:
> >
> > Hi,
> >
> > I'm a long-time C++ programmer, but I'm really new to Python, so please
> > forgive me if this is a dumb question with a simple or obvious answer.
> >
> > A fundamental idea in OOP is the ability to have a subclass augment the
> > behavior of a superclass's method by redefine that method, and then
> > calling the superclass's version of the method before doing anything
> > else.  The key to this idea, and to polymorphism in general, is that the
> > two methods have the same argument signature so that the superclass's
> > method gets passed the same set of arguments from the caller whether the
> > subclass overrides the method or not.
> >
> > If these methods have a static argument list, this is an easy thing to
> > do in Python. Here's an example:
> >
> > class Task:
> >     def __init__(self,a,b,c):
> >         <whatever>
> >
> > class Job(Task):
> >     def __init__(self,a,b,c):
> >         Task.__init__(self,a,b,c)
> >         <whatever>
> >
> > aJob = Job(x,y,z);
> >
> > But what if the method takes unnamed positional and keyword arguments.
> > Then, how do you pass these along to the superclass's version of the
> > method such that it gets the arguments in the same form that the
> > subclass's method did?  Here's the code in question:
> >
> > class Task:
> >     def __init__(self,a,b,c,*moreargs,**morekeyargs):
> >         <whatever>
> >
> > class Job(Task):
> >     def __init__(self,a,b,c, *moreargs, **morekeyargs):
> >         Task.__init__(self,a,b,c,?????)
> >         <whatever>
> >
> > aJob = Job(x,y,z,m1,m2,m3,m4,m5);
> >
> > My question really is...what do I replace the ????? (or that whole line)
> > with to get this fundamental OOP behavior?
> >
> > This doesn't work:
> >
> >         Task.__init__(self,a,b,c,moreargs,morekeyargs)
> >
> > because now Task.__init__ will see exactly 5 arguments, instead of some
> > arbitrary number.  This almost works:
> >
> >         apply(Task.__init__,(self,a,b,c) + moreargs)
> >
> > but the extra keyword arguments don't get passed along. So, I took the
> > idea one step further and tried to introduce the keyword arguments back
> > into the parameter list that gets passed to apply:
>
> This is addressed by FAQ 4.76:
>
>   http://www.python.org/doc/FAQ.html#4.76
>
> you want: apply(Task.__init__,(self,a,b,c) + moreargs, morekeyargs)
>
> Jim
>
> --
> Jim Fulton           mailto:jim at digicool.com   Python Powered!
> Technical Director   (888) 344-4332            http://www.python.org
> Digital Creations    http://www.digicool.com   http://www.zope.org
>
> Under US Code Title 47, Sec.227(b)(1)(C), Sec.227(a)(2)(B) This email
> address may not be added to any commercial mail list with out my
> permission.  Violation of my privacy with advertising or SPAM will
> result in a suit for a MINIMUM of $500 damages/incident, $1500 for
> repeats.




From nick at #spam_me_not#videosystem.co.uk  Thu Jul 29 06:25:30 1999
From: nick at #spam_me_not#videosystem.co.uk (Nick Trout)
Date: Thu, 29 Jul 1999 11:25:30 +0100
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com> <379f06e9@news.xtml.co.uk> <wkk8rkx77y.fsf@ifi.uio.no>
Message-ID: <37a02a70@news.xtml.co.uk>

>| Are there any pieces of test code anywhere which could be used as a
>| good test for Pythons speed (encompasing all of its language
>| features) so that if anyone tinkered with the bytecode execution
>| unit they could measure the speed improvement. It would be useful if
>| it also tested correctness to ensure no mistakes?!
>
>My xmlproc XML parser might be a good test of string processing at
>least, and I have regression and speed testing harnesses for it.
>However, it's already been optimized by hand, so it may not be the
>best choice.

Thanks Lars, I was thinking I wouldn't mind having a tinker with the source
when I get some time :-). In Python/Libs/test there are a load of system
test fns to test correctness. Ideally I assume test of Pythons speed would
some algorithmic test involving all of its language features? Your parser
may be a start. Where can it be found?

N









From MHammond at skippinet.com.au  Mon Jul 26 19:53:43 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Tue, 27 Jul 1999 09:53:43 +1000
Subject: [PSA MEMBERS] Python Conferences
References: <199907230917.0631499.9@summit-research.com> <3798845C.BC16931A@callware.com>
Message-ID: <7nis9v$o3k$1@m2.c2.telstra-mm.net.au>

Ivan Van Laningham wrote in message <3798845C.BC16931A at callware.com>...

>makes me wonder how many others are in the same boat as I am. ...  And
>as far as international conferences go, I would not be able to attend
>conferences outside continental North America, on a cost basis alone.
>
>I'm sorry to be such a grump; it's just that I think if fees were lower
>it might attract interesting contributors and fresh young minds to
>Python cons, and that would be good for Python.  But maybe you're right,
>the flight/hotel may well outweigh the fees.

Maybe you really just want a local Users Group.  It is important that at
least some Python conferences attract international attendees, otherwise you
will end up with a North America Python Workshop - in no-ones interest,
include the North Americans.

I certainly would not pay around $2000 in airfares and accomodation to
attend a $40.00 conference (assuming the facilities were worth the $40.00,
and not subsidised elsewhere)

In other words, there are Python users all over the world in the same
position as you - except you are lucky in that the International Python
Conferences are hosted in North America.  I would consider it unfair in the
extreme to cater for the "poor" North American Python user without catering
for the poor international user.  Maybe some funds would be better spend on
CD or Video of the proceedings so that people who can not afford to attend
could still view it, whereever they happen to reside.  Thus, for the $40.00,
both you _and_ the International Python user, could get value for money.

All due respect, but in my opinion, it is the people who _can_ afford the
existing conferences that are important to Python obtaining mainstream
acceptance.  We need company sponsored individuals and professional
consultants to arrive, and spread the Python word.  It is often these people
that seed Python into interesting places, rather than the people who would
attend $40.00 conferences hosted in dormitories.

Mark.






From MHammond at skippinet.com.au  Tue Jul  6 20:01:20 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Wed, 7 Jul 1999 10:01:20 +1000
Subject: problem with win32 install
References: <37827F92.3BC0BAED@sr.hp.com>
Message-ID: <7lu592$cs7$1@m2.c2.telstra-mm.net.au>

Check out http://statship.python.net/crew/mhammond/win32/  There is a
section on Installation Problems, and this one is explicitely covered.

Hope this helps...

Mark.

Jim Kerr wrote in message <37827F92.3BC0BAED at sr.hp.com>...
>  I just tried installing the win32all package, and the installation
>mysteriously failed. I don't remember the exact phrasing, but
>something
>to the effect of
>
>  AXScript and Python COM registration failed:
>    a device attached to the system is not functioning
>
>  Hoping this was just a bogus error message, I tried loading a win32
>module:






From guido at cnri.reston.va.us  Thu Jul 15 18:52:01 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 15 Jul 1999 18:52:01 -0400
Subject: Case mismatch on NT
References: <378CCB4A.BDAD6FC7@bigfoot.com>
Message-ID: <5loghdcyr2.fsf@eric.cnri.reston.va.us>

Hoon Yoon <hyoon at bigfoot.com> writes:

>  I realize that this has been discussed before. And you might think the
> solution is  os.environ["PYTHONCASEOK"] = "hotdog" but try this one.
>  There is dialog.py sitting in pythonwin/.../mfc directory and Dialog.py
> in lib-tk directory. If I have both directory in my Pythonpath, I cannot
> call the one sitting at the second directory with out encountering the
> dreaded "Case Mismatch" error.
>  I feel that "Case Mismatch" should only occur after all directories has
> been searched to find the 'right' case import. Not after finding the
> first non-matching one. Is there a way to set environ to do what I want?
>  If not, can anyone tell me how to get around this prob or tell me if
> what I am proposing makes sense or not?

This is intentional -- it prevents people from *relying* on the case
matching to make their program work.  (The case sensitivity is there
to make your program fail if you typed "import String" when you meant
"import string".)

Two workarounds: (1) make a copy of the second one with a different
name, and use that; (2) rearrange your sys.path so that the one you
want to use comes first.  Neither might work if you have some library
module that uses one and some other library module that uses the other...

--Guido van Rossum (home page: http://www.python.org/~guido/)



From mwh21 at cam.ac.uk  Sat Jul 17 05:58:07 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 17 Jul 1999 10:58:07 +0100
Subject: Snippet: bitcount()
References: <000601bed02a$2235af20$06a22299@tim>
Message-ID: <m33dynlhsg.fsf@atrus.jesus.cam.ac.uk>

"Tim Peters" <tim_one at email.msn.com> writes:
> A curious thing in Python is that you can't march over the bits of a long in
> linear time without resorting to hex() or marshal.dumps.  This really
> complicates lookup methods, which *can* work in O(B) time straightforwardly
> in C.

I remember being quite aggravated by this some time back, when I was
trying to implement a reasonably fast sqrt routine for Python longs
that worked even when they were too big to be converted into a float.

I was using the time honoured method 

x -> (x+a/x)/2

which works pretty well, but to find an initial guess I thought I'd
bitshift the number by half it's (bit) length. Finding the length
seemed to be impossible, without doing something like
(len(hex(a))-1)/4, which strikes me as silly, not to mention very
memory wasteful.

Mini-proposal: len(a) for a long returns the length of the number in
radix-256.

Cheers,
Michael

-- 
Oh, very funny. Very sar-cah-stic.         http://www.ntk.net
http://www.ntk.net/doh/options.html - ho ho



From wtanksle at dolphin.openprojects.net  Wed Jul  7 16:20:29 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Wed, 07 Jul 1999 20:20:29 GMT
Subject: What's ZOPE anyway?
References: <3scg3.1267$B95.91660@typ42b.nn.bcandid.com> <7m064a$921@journal.concentric.net> <slrn7o7b7u.6la.wtanksle@dolphin.openprojects.net> <7m0bae$11l@dfw-ixnews4.ix.netcom.com>
Message-ID: <slrn7o7dkd.6la.wtanksle@dolphin.openprojects.net>

On 7 Jul 1999 19:54:54 GMT, Aahz Maruch wrote:
>William Tanksley <wtanksle at dolphin.openprojects.net> wrote:

>>...and what's more is that it's EASY to use.  When I started I didn't know
>>much HTML (only the basics).  Thanks to Zope, I can learn bit by bit.

>It's a little more difficult if you're trying to do Real Work while
>you're learning.

Certainly true.  That's how I worked (building a website in Zope while
learning HTML), but I had a person who knew HTML well working with me
(although we didn't meet during the production of the website).

It's really nice being able to see the effects of experiments on a live
website.

>                      --- Aahz (@netcom.com)

-- 
-William "Billy" Tanksley



From skip at mojam.com  Thu Jul 22 19:22:01 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 22 Jul 1999 18:22:01 -0500 (CDT)
Subject: Maintainability (was Re: Stackless & String-processing)
In-Reply-To: <7n82u1$jid$1@nntp.Stanford.EDU>
References: <001101bed3f2$5e8874a0$642d2399@tim>
	<7n675p$qsk@dfw-ixnews7.ix.netcom.com>
	<14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>
	<7n82u1$jid$1@nntp.Stanford.EDU>
Message-ID: <14231.42621.76462.103289@96.chicago-25-30rs.il.dial-access.att.net>

    Amit> import re
    Amit> from string import *  # Great whenever you're doing string processing
    Amit> re.compile('abc')

Two points.  One, the string module wasn't designed to be imported using
"from string import *".  In my previous message I mentioned the types and
Tkinter modules.  I'm not aware of any others that were explicitly designed
for that.  Two, the bug that placed an object called "re" in the string
module at module scope was fixed several months ago.

    % python
    Python 1.5.2+ (#28, Jun 17 1999, 15:24:35)  [GCC 2.7.2.3] on linux2
    Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
    >>> import string
    >>> dir(string)
    ['_StringType', '__builtins__', '__doc__', '__file__', '__name__',
    '_apply', '_float', '_idmap', '_idmapL', '_int', '_long', 'atof',
    'atof_error', 'atoi', 'atoi_error', 'atol', 'atol_error', 'capitalize',
    'capwords', 'center', 'count', 'digits', 'endswith', 'expandtabs',
    'find', 'hexdigits', 'index', 'index_error', 'join', 'joinfields',
    'letters', 'ljust', 'lower', 'lowercase', 'lstrip', 'maketrans',
    'octdigits', 'replace', 'rfind', 'rindex', 'rjust', 'rstrip', 'split',
    'splitfields', 'startswith', 'strip', 'swapcase', 'translate', 'upper',
    'uppercase', 'whitespace', 'zfill']

Look, Ma! No re!

    Amit> I think the system modules should define all internal variables
    Amit> with a '_' prefix! :-)

Again, the string.re thing was a bug.  Most system modules do follow this
convention.

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758




From Markus_Ewald at gmx.net  Mon Jul 19 10:31:49 1999
From: Markus_Ewald at gmx.net (Markus Ewald)
Date: Mon, 19 Jul 1999 16:31:49 +0200
Subject: Problems with PyRun_SimpleFile()
Message-ID: <379336D5.799D83E0@gmx.net>

Excuse this dumb question, but how can I use PyRun_SimpleFile() ?
I'm using VC6 on Win98, where I've declared some simple functions for
testing.

Well -- everything works fine when I use PyRun_SimpleString().

If I put the same lines I gave PyRun_SimpleString() in a File and then
use PyRun_SimpleFile() to execute it, I get an access violation.

I did everything exactly as I've seen in Modules\Main.C (except that I'm
using CPP), what's wrong ?
(I've hung a minimal, commented version on this mail, if anyone wants to
take a look at it)

-Markus-


-------------- next part --------------
A non-text attachment was scrubbed...
Name: Main.CPP
Type: application/x-unknown-content-type-cppfile
Size: 1517 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19990719/c89e545f/attachment.bin>

From thstr at serop.abb.se  Fri Jul  2 09:13:24 1999
From: thstr at serop.abb.se (Thomas S. Strinnhed)
Date: Fri, 02 Jul 1999 15:13:24 +0200
Subject: Why does shelve make such large files?
References: <19990701211450.A6827@optiplex.palga.uucp> <377BF01D.16F8AA33@cup.hp.com>
Message-ID: <377CBAF4.162E1659@serop.abb.se>

Hi

Ovidiu Predescu wrote:
>[...]
> 
> The shelve module uses DBM which is like a small database that allows
> you to store objects and search it for a given key. DBM allows you to
> store millions of objects and search for them later without requiring
> you to load all the file in memory first.
> 
> The pickle module on the other hand is serializing the objects with the
> purpose of deserializing them _all_ from the file later. Pickle does not
> offer you any way to search for data based on a key, you have to do this
> yourself after the objects have been created from the file. This is
> opposed to the way shelve handles this, all the key accesses and
> insertions in a shelve object are actually reads or writes to or from
> the DBM file.
> 
> And to answer your question, DBM is creating these big files because of
> the way it manages the database. The data in the database file could
> have gaps as a result of multiple insertions and deletions. Pickle's
> data in files is a simple representation of the objects that were
> written and there is no way to update the file other than rewriting it
> entirely.
> 
> --
> Ovidiu Predescu <ovidiu at cup.hp.com>
> http://www.geocities.com/SiliconValley/Monitor/7464/

So, just to make shure I follow, in short terms 
 * pickle makes "persistant objects" flushed to a file 
 * shelve + dbm is in fact a (relational) database storing
   (arbitary**) objects in some hash-searchable order

(**) By arbitary I mean _any_ kind of object, or do they need to be
     of the same class?? (In order for searching to work)

About usage: shelves when I need to search and pickle when I just need
to save my objects?? Are there other considerations based on number and
size of objects?

Best regards
 -- Thomas S. Strinnhed, thstr at serop.abb.se



From Fredrik.Lundh at p98.f112.n480.z2.fidonet.org  Thu Jul  1 05:21:42 1999
From: Fredrik.Lundh at p98.f112.n480.z2.fidonet.org (Fredrik Lundh)
Date: Thu, 01 Jul 1999 10:21:42 +0100
Subject: Tkinter: why does an empty frame not shrink?
Message-ID: <00001602@bossar.com.pl>

From: "Fredrik Lundh" <fredrik at pythonware.com>

Greg McFarlane wrote:
> If you have several child widgets packed in a frame and you unpack the
> children one by one, the frame will gradually get smaller, except when
> the last child is unpacked, in which case the frame does not resize. 
> Does anyone know why the frame does not shrink to 0x0 when you unpack
> the last child?

when geometry propagation is on (as it is by default), the
packer sets the size of the geometry master to exactly fit
the child widgets.  if there are no children, the packer
simply keeps the current size.

> Any solution?

why not just insert an extra widget that you never delete?

from Tkinter import *

root = Tk()
root.geometry("400x400")

frame = Frame(background="red")
frame.pack(fill=X)

Frame(frame).pack()

for i in range(5):
    button = Button(frame, text="Unpack me")
    button.configure(command=button.pack_forget)
    button.pack()

mainloop()

</F>



From guido at cnri.reston.va.us  Thu Jul 15 18:33:05 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 15 Jul 1999 18:33:05 -0400
Subject: History of MAGIC number ?
References: <19990714114525.A739@hoffleit.org>
Message-ID: <5lr9m9czmm.fsf@eric.cnri.reston.va.us>

Gregor Hoffleit <flight at mathi.uni-heidelberg.de> writes:

> I'd like to submit a patch to the maintainer of the Linux file command
> so that file recognizes Python .pyc and .pyo files.
> 
> .pyc and .pyo files are quiet easily recognizable on their magic
> number, a long word at the start of the file. The format of the long
> word is 32bit little endian, with the higher 16 bits set to 0x0a0d and 
> the lower 16 bits set to a pyc version number (20121 at the moment
> in Python 1.5.2).
> 
> Now if I want to write a robust test method for file I could either
> test for 0x0a0d in position 2 (which is not very rigid), or I had to
> list the explicit versions, like
> 
>   0  lelong   0x0a0d4e99    compiled Python code, magic version 20121
> 
> Now the problem is that this verion number has changed quite a few
> times in Python's history, but I can't find a list of the ancient
> values anywhere (Misc/HISTORY only lists the times changes took place, 
> but not the values).
> 
> Btw, should I say "Python bytecode" or "compiled Python code" ?

I would say Python bytecode; another acceptable term is compiled
Python module.

The history is in the CVS tree: http://www.python.org/download/cvs.html

--Guido van Rossum (home page: http://www.python.org/~guido/)



From S.I.Reynolds at cs.bham.ac.uk  Fri Jul 23 13:26:47 1999
From: S.I.Reynolds at cs.bham.ac.uk (Stuart Reynolds)
Date: Fri, 23 Jul 1999 18:26:47 +0100
Subject: Interactive use of Tk
Message-ID: <3798A5D7.674D@cs.bham.ac.uk>

I'm trying to debug a tkinter application I have using the interactive
prompt via XEmacs. The program runs fine first time around but when run
again (without exiting the interactive prompt) it complains that the
application has been destroyed. How to I stop the application from being
destroyed?

Secondly, has anyone managed to get tkinter windows running whilst still
being able to use the interactive prompt (e.g. by starting the mainloop
in a separate thread). I guess its possible and its just the dodgy
Solaris threads libraries we have installed here that stop it working
for (the jpython examples also cause the command prompt to block for
me).



Cheers,

Stuart

PS. The code which won't run twice:

-----
if __name__ == '__main__':


    from Tkinter import *
    frame = Frame()
    ... some other code here ...
    Pack.config(frame)
    frame.mainloop()
    
-------


#Works fine:
>>> ## working on region in file /usr/tmp/python1f5be_...
>>> 
#Run again:
>>> ## working on region in file /usr/tmp/pythonh2FQL_...
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/tmp/pythonh2FQL_", line 117, in ?
    frame = Frame()
  File
"/bham/ums/common/pd/packages/Python/lib/python1.5/lib-tk/Tkinter.py",
line 1406, in __init__
    Widget.__init__(self, master, 'frame', cnf, {}, extra)
  File
"/bham/ums/common/pd/packages/Python/lib/python1.5/lib-tk/Tkinter.py",
line 1084, in __init__
    self.tk.call(
TclError: can't invoke "frame" command:  application has been destroyed



From fredrik at pythonware.com  Fri Jul 30 10:35:48 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 30 Jul 1999 14:35:48 GMT
Subject: Libraries needed, please help!!!
References: <7ns6o5$jo2@news.go.dlr.de>
Message-ID: <004e01beda98$d14f9390$f29b12c2@secret.pythonware.com>

Sven Drescher <Sven.Drescher at dlr.de> wrote:
> I wanted to embed Python into C. I looked for the libraries to link them to
> my program. But I haven't the libraries. I looked for it in the Internet and
> everywhere, but never found.
> 
> Please give me a hint, where I can download the files:
> libModules.a
> libPython.a
> libObject.a
> libParser.a
> 
> Thank you very much!!!

assuming you're on unix,

1) get sources from www.python.org
2) unpack
3) type "./configure"
4) type "make"

ta da!

</F>




From simngolda at voyager.net  Thu Jul 15 00:21:52 1999
From: simngolda at voyager.net (Sim & Golda Zacks)
Date: Thu, 15 Jul 1999 00:21:52 -0400
Subject: File Upload in cgi
Message-ID: <7mjnf6$7eo@newsops.execpc.com>

Code posted below
I am trying to upload a file to my cgi-script and I seem to be failing
miserably. In my html page the item is of input type file and I select a
file and hit submit and when I access it with the FieldStorage() and
then print it it says it's not a file. I am basically copyng the example
in the cgi module documentation.  it prints out
MiniFieldStorage('picture',
'C:\\TEMP\\WebPage\\baby3.gif')
and then it prints not a file. The file isn't on the server either. Can
someone tell me what I'm doing wrong?

import cgi

def main():
        print "Content-type: text/html"
        print
        print "Testing what happens for a picture"
        form=cgi.FieldStorage()
        pic=form["picture"]
        print pic
        if pic.file:
                l=pic.file.readlines()
                g=open('newpic.gif','wb')
                g.writelines(l)
                g.close()
                print 'wrote file'
        else:
                print 'not a file'

if __name__=="__main__":
        main()






From gmcm at hypernet.com  Wed Jul 28 08:46:04 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 28 Jul 1999 07:46:04 -0500
Subject: Designing questions
In-Reply-To: <379EAB6B.2AD7E3C0@worldnet.fr>
Message-ID: <1278960329-14812603@hypernet.com>

Pierrette writes:

> I'm writing  a application wich contains abouth 50 inputscreen
> linked to db-tables. Each inputscreen has 2 classes - 1 GUI class
> wich  construct the screen. - 1 db class wich  do the db acces.

I remember you!
 
> All the gui-classes are done at the moment, so i have the prototype
> of the application. And here's my problem : The source code becomes
> realy long, wich gives a long time to load it. I was thinking of
> spliting it in modules (1 module for each screen wich contains the
> gui-class and the db-class) and then importing it when i needed it,
> this is when the user makes the choice in the main-menu. The
> question i have is how to 'unload' a module when i don't need it
> anymore ? If i don't unload a module i would  have as many modules
> in memory as the user has made choises in the main menu. I think
> this could become a memory problem especailly since i'm using gadfly
> as db, wich allso holds the tables in memory.

Well, you don't unload modules in general, and I bet it won't be a 
problem. You will gain a lot of speed, because for imported modules, 
Python will create a .pyc the first time it's imported, and use that 
(if the .py file hasn't changed) on subsequent imports.

In other words, one way to speed up loading is to create a very small 
main script:
-----------------------------
import myrealstuff
myrealstuff.run()
-----------------------------

This script will be interpreted each time, but "myrealstuff" will get 
pre-compiled.
 
Breaking up your script is a good idea for maintainability, too. But 
rather than just chop it in pieces, I'd look carefully to see if all 
those GUI classes don't have some things in common that can be 
factored into a base class. Then each GUI class can inherit from 
that, and just add the stuff that's particular to that screen. 

You could use the same idea on the db classes, too, but I'd guess 
that most of your code is GUI code. With gadfly, make sure you use 
the form of SQL that uses "?"s to mark the variables. That will let 
gadfly precompile the SQL statements.

> I really would appreciate some hints and comments here from
> experienced guys who allready wrote BIG real world applications in
> PYTHON.

That's me. Big guy, big Python <2.0 wink>

- Gordon



From mithy at my-deja.com  Wed Jul  7 11:17:19 1999
From: mithy at my-deja.com (Preston Landers)
Date: Wed, 07 Jul 1999 15:17:19 GMT
Subject: Emacs: highlighting "self" tweak
References: <7luqpa$3s9$1@nnrp1.deja.com> <19990707133146.F3AD4301CF2@bireme.oratrix.nl>
Message-ID: <7lvr1i$fai$1@nnrp1.deja.com>

Very cool Sjoerd, thanks a lot!

I'm not very knowledgable about elisp, so maybe you all can forgive my
naive approach to modifying python-mode.el.  Good thing there are a few
adults around here to mind my manners.

Thanks again!

---Preston

In article <19990707133146.F3AD4301CF2 at bireme.oratrix.nl>,
  Sjoerd Mullender <sjoerd at oratrix.nl> wrote:

> You can use this piece of magic in your .emacs file (tested on XEmacs
> 21.1):
>
> (add-to-list 'after-load-alist
> 	     (cons "python-mode"
> 		   '((add-to-list (quote python-font-lock-keywords)
> 				  (quote ("\\b\\(self\\)[ .,]?"
> 					  1 font-lock-type-face))))))
>
> This changes the python-font-lock-keywords variable right after
> python-mode is loaded.

--
||  Preston Landers <mithy at my-deja.com>  ||
||        http://askpreston.com          ||


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From phd at emerald.netskate.ru  Fri Jul 23 09:16:02 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Fri, 23 Jul 1999 17:16:02 +0400 (MSD)
Subject: how to access ports ?
In-Reply-To: <7n9o8o$41e@zam201.zam.kfa-juelich.de>
Message-ID: <Pine.LNX.4.04.9907231714040.30978-100000@emerald.netskate.ru>

On 23 Jul 1999 r.skowronek at fz-juelich.de wrote:
> Well, how do I access arbitrary ports?
> I managed to read and write to serial ports easy, but I would like to access
> an mca card at addresses 0x220 - 0x31f.
> I checked the docs but did not see any reference to hardware access - or did
> I not look in the right places?

   Hardware access is too hardware/software/OS dependent, generic langauage
like Python cannot handle it in portable way.
   If you can program in C, and if you agree to learn "Python Extending and
Embedding" - write extension module for your particular task in C.

> Thanks for any hints.
> 
> Rolf

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From Gordon.McMillan at p98.f112.n480.z2.fidonet.org  Fri Jul  2 09:44:55 1999
From: Gordon.McMillan at p98.f112.n480.z2.fidonet.org (Gordon McMillan)
Date: Fri, 02 Jul 1999 14:44:55 +0100
Subject: Mailbox, tell and read
Message-ID: <0000166c@bossar.com.pl>

From: "Gordon McMillan" <gmcm at hypernet.com>

Fredrik Lundh wrote:
> Gordon McMillan wrote:
> > > First, are there platforms other than DOS/Windows/Win32 where there
> > > is a difference between text and binary mode?
> > 
> > There's a theoretical difference on *nix - otherwise, they wouldn't 
> > have made up the distinction!
> 
> who are "they?"

Whatever AT&T employee it was. AFAIK, it predates any glimmer of an 
idea of a "PC".

> fwiw, the ANSI C rationale mentions Unix as the
> special case here... go figure ;-)

I have a vague recollection that it can make a difference on AIX. At 
least I remember fixing something by changing someone's "r" to "rb", 
but I've completely forgotten the context.

- Gordon



From pmagwene at uchicago.edu  Thu Jul 15 18:13:09 1999
From: pmagwene at uchicago.edu (Paul M.)
Date: Thu, 15 Jul 1999 22:13:09 GMT
Subject: I envy newbies
References: <378E3481.EE31C97@uswest.net>
Message-ID: <FExMpA.Juv@midway.uchicago.edu>

Me thinks thou has not read the books, nor the copious documentation,
nor the website, very carefully.

Once you've gotten through the install process (you can handle that,
yes?  All it takes is clicking a little icon on your screen), you can
test the INTERPRETTER (which will appear in a DOS box, but it NOT a
DOS command line) by clicking on the little icon in your Python
install directory that says "python.exe" or by doing the same thing
from your Start>Programs>Python window.

I qoute from the web page:

--------------<see
http://www.python.org/download/download_windows.html
>--------------------------
"Once you have downloaded an installer, you must run it as a program.
To do this, double click on its icon. The installer will guide you
through the installation process.

You can verify that the interpreter works by double-clicking on its
icon. A DOS style console window should pop up with a copyright notice
and a ">>>" prompt. You can now run Python statements, e.g.

>>> 12*12
144
>>> import string
>>> string.split("hello world")
['hello', 'world']
>>>
Type Control-Z or click the window's close button to exit.

Important: if the "import string" statement above doesn't work, you
are in trouble. Most likely, the registry hasn't been set up
correctly. The cause of this problem is probably that you messed with
the installation after running the installer. Try uninstalling and
reinstalling.
----------------------------------------------------------------------
----------------------------------


If you're flustered by anything even remotely resembling a command
window I believe (but am not sure) that the latest installer even puts
a shortcut to IDLE (a nice tk driven interpretter/programming
environment) in your Python folder on your Start>Programs>Python
window.

It might help to know at least something about the OS you're using
(e.g., to set your path to include the Python install directory so you
can access the interpretter from any command line) - even if you do
find it appalling.  However, the standard install makes it incredibly
easy, even for the clueless.

If all this is STILL too obscure, then yes, perhaps you should learn
Perl, which I hear is known for its extreme readability and ease of
use ;)

--Paul


mary jeanette wrote in message <378E3481.EE31C97 at uswest.net>...
>I decided to learn python because I heard it was a very easy language
to
>learn. So I downloaded it and bought a couple of very expensive
>paperback books from o'reilly only to discover that they are written
for
>unix. In other words, typing in the examples from the book into the
>interpretor only yields various syntax errors. I have no idea what I
am
>supposed to change in order to make the examples work, and apparently
>there is nowhere to turn in order to find out. Q: Am I supposed to
know
>DOS as a prerequisite? The idea of having to learn DOS is appalling.
>Should I try partitioning my hard drive and installing Linux or
freeBSD?
>Perhaps I should learn Perl instead, I have seen books on Perl
written
>for Win32 systems.   Thanks for your advice.
>




From gmcm at hypernet.com  Tue Jul 13 23:39:49 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 13 Jul 1999 22:39:49 -0500
Subject: Running 2 gui libs simultaneously
In-Reply-To: <378BEC9C.271A69EB@compaq.com>
Message-ID: <1280202722-8377598@hypernet.com>

Greg Ewing wrote:

> Gordon McMillan wrote:
> > 
> > You don't have to use Tk's mainloop. You can call update() in a loop.
> > Perhaps wxWindows has something similar, or you can call Tk's update
> > from within some idletask in wxWindows.
> 
> I don't fancy your chances of getting that to work.

As I said before, I have no intention of trying.

Or maybe you meant to reply to someone else?

- Gordon



From tim_one at email.msn.com  Sat Jul 17 14:30:33 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Jul 1999 14:30:33 -0400
Subject: I envy newbies
In-Reply-To: <3790884A.A58864B@callware.com>
Message-ID: <000101bed082$747b0de0$27a22299@tim>

[Ivan, clinging to an illicit affair with autoexec.bat]

I'm not arguing about this anymore -- it's too easy <wink>.

If you're going to tell people to edit the file, just be sure to cover all
the problems they can face.  Here's one you're not expecting <wink>:
autoexec.bat may not exist!  Win9*/NT don't need it, and not all
installations have one.

I think the best advice you can give a newbie regardless is to force the
install to put Python in a directory *they* choose, and one without spaces
in its name.  Like C:\Python.  This is easier than telling them how to find
python.exe in the (ill-defined) "default" location, and will simplify later
tasks whether or not they muck with autoexec.bat.

> Usually, these things modify the path by simply adding a line like this:
>
> 	PATH=L:\BC5\BIN;%PATH%

That's cute:  they rely on a bug in the PATH cmd <wink>?  The "=" should be
a blank, or "set " should preface this, although as given it works by
accident.

> which does get around the simple line length limitation.  The actual
> length of the path variable in the environment seems to not suffer the
> 255 character limits that W3.1 had; I was pushing or exceeding that
> limit with 3.1, but I have no such problems now.

Then, Windows being the model of consistency that it is, I'm sure nobody
else could possibly have a problem with this anymore either <wink>.

> ...
> If newbies are *ever* going to use Python, the path has to be correct.

I've lived happily for years without it on my path.  There's also a DOS-free
alternative for newbies:  they can start Python via Start->Run, typing
"python whatever.py" or plain "python" into the dialog.  python.exe doesn't
need to be on the PATH for that to work (unlike a DOS box, Start->Run also
looks in the registry under

    HKLM\SOFTWARE\Microsoft\Windows\CurrentVersion\App Paths\

and the Python installer does register the full path to python.exe there).
MS seemed to be on to a good idea here, but shipped before it was finished.

> ...
> I shall.  I've got a new system that I can afford to mess up.
>
> Does or does not Win32All modify your path?

I certainly hope not <wink>.

[understandable whining about other Windows install programs]
> ...
> Hacking the autoexec.bat is safer than dumping DLLs into the system32
> directory.

It depends, of course.  Apps written for Win95 (or beyond) can also use the
App Paths registry subtree (see above) to fiddle the PATH to be anything it
wants each time the app is started.  This doesn't have to be done in
autoexec.bat anymore, and indeed *shouldn't* be done there.  Apps
overwriting shared DLLs in the System directory are simply braindead, or
haven't been dusted off since their pre-95 versions.

> Have you tried installing AOL4.0 by downloading it recently?

Not recently -- not ever <wink>.

> Not only does it overwrite all your TCP/IP DLLs with old gunk
> that f**ks every other form of networking you've got, but IT WON'T
> BELIEVE THAT IT HAS ALREADY SCREWED YOU UP, so it keeps on trying to
> re-copy the bad DLLs to your system32 directory every time you re-boot.

Things that can cause this include a rogue program in your Programs->Startup
group, or a WININIT.INI file (on Win95) somewhere on your boot drive.  You
should whine at AOL, though -- they presumably wrote their install program.

> Honestly, I can't tell them, "cd to your Python directory for every
> exercise in this book."  I'd rather tell my readers to edit their
> autoexec.bat than tell them how to cd to a directory with spaces in the
> path.  Bleaugh;-(>

As at the start, tell them to install into a directory that's easy to get
at.  You're presumably going to have to tell them how to bring up a DOS box
too.  Then they're going to have a DOS box icon on their desktop, right?
Right-click that icon, select Properties, select the Program tab, enter the
name of the Python directory into the "Working" line, click OK.  Voila.
Their DOS box icon will now start a DOS box already cd'd to the Python
directory, and autoexec.bat hasn't been touched.  Easier than editing a
file, and they can't hurt anything no matter how badly they screw up this
step.

although-later-they-can-easily-delete-their-python-directory-by-
    mistake<wink>-ly y'rs  - tim





From simngolda at voyager.net  Wed Jul 14 00:10:47 1999
From: simngolda at voyager.net (Sim & Golda Zacks)
Date: Wed, 14 Jul 1999 00:10:47 -0400
Subject: Gadfly accessing results
Message-ID: <7mh2ed$hrb@newsops.execpc.com>

When I do a Select query in Gadfly how would I then be able to use
individual fields from the results? In Delphi (my primary programming
language) I would use a query and then call
query.fieldbyname('fielname').value and I have access to all the fields
that were selected. How would I do this in Gadfly? For example my select
is for 5 fields and it returns 2 records. I want to be able to get each
field individually from each record

Thanks
Sim




From AKiss at GEOMETRIA.hu  Thu Jul 22 07:30:31 1999
From: AKiss at GEOMETRIA.hu (Kiss, Arpad)
Date: Thu, 22 Jul 1999 13:30:31 +0200
Subject: Is anybody aware of a tree GUI widget for tkinter or Pmw?
Message-ID: <01BED446.5FB7C1B0@exchange1.geometria.hu>


D. Lopez-De-Ipina <dl231 at eng.cam.ac.uk> wrote in message
news:<3796E213.9F326412 at eng.cam.ac.uk>...
> Hi,
> 
> Could anybody please tell me if there exits any TREE control widget for
> tkinter or Pmw? I know that Pmw provides a rather rich set of
> 'megawidgets' but the tree control has not been implemented yet as far
> as I know.
> 
> Thanks a lot,
> 
> Diego
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 
I have one: http://starship.python.net/crew/arpadk
Cheers,
Arpad



From parkw at better.net  Thu Jul  8 06:36:05 1999
From: parkw at better.net (William Park)
Date: Thu, 8 Jul 1999 06:36:05 -0400
Subject: Decent Tutorial
In-Reply-To: <000015af@bossar.com.pl>; from Newbie on Wed, Jun 30, 1999 at 01:55:49AM +0100
References: <000015af@bossar.com.pl>
Message-ID: <19990708063605.B2111@better.net>

On Wed, Jun 30, 1999 at 01:55:49AM +0100, Newbie wrote:
> From: "Newbie" <8499 at greenhead.ac.uk>
> 
> 
> Where can I find a decent tutorial which is 
> for beginners to Python, preferably including 
> some documented source code and avaliable for 
> download.
> 
> Thanks.
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list

http://www.python.org/doc/...



From kaine_the_bloody at my-deja.com  Wed Jul 14 16:17:25 1999
From: kaine_the_bloody at my-deja.com (Kaine)
Date: Wed, 14 Jul 1999 20:17:25 GMT
Subject: Python and Paradox
References: <7m2mrj$hes$1@nnrp1.deja.com>
Message-ID: <7mir86$r49$1@nnrp1.deja.com>

In article <7m2mrj$hes$1 at nnrp1.deja.com>,
  Kaine <kaine_the_bloody at my-deja.com> wrote:
> I am looking for a way to access Paradox tables from Python without
> using ODBC. Does anyone have any ideas?
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.
>
Thanks guys. So unless I really want to hack the answer is no. I can
think of a number of other things I would rather hack. :) Anyone know
if Paradox is going to get a COM interface anytime soon?


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From too at pk.highway.ne.jp  Tue Jul 27 02:54:54 1999
From: too at pk.highway.ne.jp (too at pk.highway.ne.jp)
Date: Tue, 27 Jul 1999 15:54:54 +0900
Subject: glob and directory
Message-ID: <379D57BE.53DA9B37@pk.highway.ne.jp>

Hello.

I would like each output of "glob.glob()" to be
added "/"
if returned value is directory.

That is...

>>> import glob
>>> print glob.glob('*')
['file1', 'file2', 'file3', 'dir1', 'dir2']

===> ['file1', 'file2', 'file3', 'dir1/', 'dir2/'] 

Is it possible ?

If not, is there any way to determine
if the given name is directory or not ?

Thanks for reading.

Thooney Millennier.



From conversy at lri.fr  Wed Jul 28 10:38:23 1999
From: conversy at lri.fr (Stephane Conversy)
Date: Wed, 28 Jul 1999 16:38:23 +0200
Subject: beep
References: <379F0751.71214802@lri.fr> <379F1B1A.2199B575@inrialpes.fr> <379F0EF9.54CDDBEC@lri.fr> <379F119E.432EA0BA@onera.fr>
Message-ID: <379F15DF.62F4D6F6@lri.fr>

Marc POINOT wrote:

> Stephane Conversy wrote:
>
> > Vladimir Marangozov wrote:
> >
> > > Perhaps printing the standard escape sequence '\a' (Alert) would do?
> > >
> > >
> >
> > is it portable ? (ie Mac OS)
> >
>
> As a matter of fact, <Klung> is much more heavy than <Bip>.
>

ok, I saw it in Tkinter: it's <anywidget>.bell()

Thank you all.

    stef

--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr






From Tim.Peters at p98.f112.n480.z2.fidonet.org  Sat Jul  3 06:37:59 1999
From: Tim.Peters at p98.f112.n480.z2.fidonet.org (Tim Peters)
Date: Sat, 03 Jul 1999 11:37:59 +0100
Subject: Slicing beyond list --  bug or featurette?
Message-ID: <000016b0@bossar.com.pl>

From: "Tim Peters" <tim_one at email.msn.com>

[Pierre Fortin]
> Just ran into a situation which can be summarized as follows:
>
> list = [ 0,1,2,3,4 ]
>
> list[<0..4>] returns the appropriate item
>
> list[m] where m is >= len(list) expectedly fails
>
> list[2:50] returns [ 2,3,4 ]
> list[20:] or list[20:30] returns []
>
> Can I count on this, or should I code try/except for the day it's
> fixed?

It's guaranteed by footnote 2 to the table in the Library Ref's section
2.1.5 (Sequence Types).  I think the Language Ref leaves it open to debate,
though (if anything ever changes here, it's almost certainly that the
Language Ref will get more words blessing the current implementation).

> This 'feature' reduces the error situations; but is it correct/intended?

All three <wink>.  Slicing returns a subsequence, and the possibility to
return an empty subsequence works out very smoothly in practice.

More dubious is that negative indices index "from the right end":

    x[-1] == x[len(x)-1]

This is too handy to give up, but in my experience *has* let some errors
slip by; while I can't recall any case where

    x[insane_1 : insane_2] == []

didn't simply handle an expected endcase gracefully.  Doesn't mean we
couldn't construct one, just saying I've never seen one occur naturally.

BTW, in some older code you may see the strange idiom

    prefix[sys.maxint:] = suffix

or more likely

    prefix[HUGE:] = suffix

in a loop where HUGE = sys.maxint is computed once outside the loop.  In
1.5.2 it's spelled

    prefix.extend(suffix)

the-set-of-all-i-less-than-2-where-i-greater-than-100-is-just-empty-ly
'rs  - tim






From robin at alldunn.com  Thu Jul 15 00:45:16 1999
From: robin at alldunn.com (Robin Dunn)
Date: Wed, 14 Jul 1999 21:45:16 -0700
Subject: newbie questions about sys.path, wxpython, IDLE and pythonwin
References: <378B7A69.12874AAC@gnet.tn>
Message-ID: <sNdj3.8$oe.5111882@news.randori.com>

MB <ritm at gnet.tn> wrote in message news:378B7A69.12874AAC at gnet.tn...
>  Then I
> tried the same thing through the IDLE GUI and received this error:
> ------------------
> Traceback (innermost last):
>   File "C:\PROGRA~1\PYTHON\TOOLS\IDLE\ScriptBinding.py", line 131, in
> run_module_event
>     execfile(filename, mod.__dict__)
>   File "C:\Program Files\Python\wxPython\demo\demo.py", line 3, in ?
>     import Main
> ImportError: No module named Main
> -----------------
>
> Obviously it was a problem of path.

Since demo.py imports things in the same directory where demo.py lives, it
should be found by default.  Perhaps IDLE is masking this path feature
somehow?  What is the CWD when executing from an edit buffer?

> So I added the directory an its
> subdirectories to sys.path and tried again. I received this
>
> ---------------------------------
> Traceback (innermost last):
>   File "C:\PROGRA~1\PYTHON\WXPYTHON\DEMO\Main.py", line 186, in
> OnSelChanged
>     self.GetDemoFile('Main.py')
>   File "C:\PROGRA~1\PYTHON\WXPYTHON\DEMO\Main.py", line 221, in
> GetDemoFile
>     except IOException:
> NameError: IOException
> Traceback (innermost last):
>   File "C:\PROGRA~1\PYTHON\WXPYTHON\DEMO\Main.py", line 186, in
> OnSelChanged
>     self.GetDemoFile('Main.py')
>   File "C:\PROGRA~1\PYTHON\WXPYTHON\DEMO\Main.py", line 221, in
> GetDemoFile
>     except IOException:
> NameError: IOException
> ---------------------------------


The demo expects that its current working directory is where the demo.py
lives.  It tries to open files there.

>
> Other thing. I examined demo.py through the debugger under pythonwin I
> found this statement:
> 'import miscc'
> The interpreter import this module but is not able to locate it. When I
> type >>> miscc it says that is a built-in module. Are there modules that
> are not python files ? What is a built-in module ? I didn't find its
> definition in the docs.

The miscc module, and about a dozen others are Python extension modules
written in C++.  Most of them live in wxc.pyd and are initialized when the
wxc module is imported.  Since it is not written in Python, it's a bit hard
for the debugger to trace into it.  You can read more about extension
modules in the "Extending and Embedding" document at
http://www.python.org/doc/  Also, the wxPython sources can be downloaded and
you can browse the C++ code to your hearts content.  On the other hand since
I (the author) don't even like reading it you might not want to.  It's
generated by SWIG and is fairly ugly.


> Last question. About the  syntaxic ':' at the end of a line when it
> begins with def, class,if ,when,etc... Is it absolutely necessary for
> the language ?

Yep.  The ":" is what keeps the def, class, if, when, etc. from falling off
the end of the line.  <grin>


--
--
Robin Dunn
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!
Try http://AllDunn.com/laughworks/ for a good laugh.









From Hank.Fay at p98.f112.n480.z2.fidonet.org  Thu Jul  1 08:19:25 1999
From: Hank.Fay at p98.f112.n480.z2.fidonet.org (Hank Fay)
Date: Thu, 01 Jul 1999 13:19:25 +0100
Subject: wddx wrapper?
Message-ID: <00001636@bossar.com.pl>

From: "Hank Fay" <hank at prosysplus.com>

Does anyone know a reference to a Python wrapper for wddx (www.wddx.org)?
It looks like a neat way to pass data around (being able to transport
recordsets between differnet languages, in particular), and I'd like to be
able to have Python doing some of the heavy lifting on the server side.

thanks,

Hank Fay

--
http://www.prosysplus.com




From tim_one at email.msn.com  Sat Jul 17 03:58:19 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Jul 1999 03:58:19 -0400
Subject: Snippet: bitcount()
In-Reply-To: <slrn7p09o9.ahf.kas@titan.magnetic-ink.dk>
Message-ID: <000601bed02a$2235af20$06a22299@tim>

[Klaus Alexander posts a pretty bit-counter, taking time proportional to
 the number of 1 bits for ints, and for N-bit longs taking time
 proportional to the product of N and the number of 1 bits]

[Egene Leitl]
> Ouch, ouch. That was really _inefficient_.  At least you could nybble
> or byte-wise lookup,

This isn't C or assembler -- the speed difference for ints in Python isn't
something to get excited about.  Klaus's method has a long pedigree, and is
the method of choice "even in C" when dealing with sparsely populated large
bit vectors.

> and there should be lots of bigmask magic which does in in O(logN).
> Consider bitcounting L's.

There are general methods that work in log2(B) operations where B is the
number of bits in the long.  But each primitive operation (such as a simple
shift) on a B-bit long takes time proportional to B, so the overall time is
O(B * log2(B)).  The worst case of Klaus's method (all bits are set) is
O(B**2).  For B large enough, that's indeed a huge difference -- but unless
B is at least a few hundred bits, the overhead of controlling the advanced
methods cancels the reduction in operations.  (BTW, I'm lying a little here:
the fancy shift-&-mask methods continually cut the number of bits in half,
so that they're really O(B) "in theory" -- and in practice too, for B very
large.)

A curious thing in Python is that you can't march over the bits of a long in
linear time without resorting to hex() or marshal.dumps.  This really
complicates lookup methods, which *can* work in O(B) time straightforwardly
in C.

You can find more than you can bear to read <wink> about this by searching
for "popcount" in DejaNews.

[back to Klaus]
> ...
> -- unless I decide to count several bits at a time, but
> I wouldn't know how to construct such a thing.

Oh, sure you do <wink>:

count[0] = 0
count[1] = 1
count[2] = 1
count[3] = 2
...
count[255] = 8

Now you can get the popcount of any one-byte int with one table-lookup.  And
the popcount of any N-byte int or long with N table lookups and N-1 adds.
Put 2**16 entries in the table to cut the work in half.

repeat-as-needed<wink>-ly y'rs  - tim





From c.evans at clear.net.nz  Mon Jul 19 05:15:40 1999
From: c.evans at clear.net.nz (Carey Evans)
Date: 19 Jul 1999 21:15:40 +1200
Subject: problem with smtplib.sendmail
References: <931731872.16311@www.remarq.com>
Message-ID: <871ze5ov9f.fsf@psyche.evansnet>

kevinsl <kevinsl at yahoo.com> writes:

> fromaddr='test at test.com'
> toaddr='list-subscribe at test.com'
> msg=''
> mailserver=smtplib.SMTP('localhost')
> mailserver.sendmail(fromaddr,toaddr,msg) #error occurs here
> mailserver.quit()
> 
> localhost is running qmail 1.03.

qmail is a little more picky than some other mailers, and an empty
message is *not* a valid RFC822 message.

Try adding a minimum number of headers, e.g.,

msg='To: %s\n\n' % toaddr
[...]

Actually, that's not a valid RFC822 message either, but I think qmail
will accept it...

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

			 print [<>]->[rand$.]



From da at ski.org  Wed Jul 21 01:31:04 1999
From: da at ski.org (David Ascher)
Date: Tue, 20 Jul 1999 22:31:04 -0700 (Pacific Daylight Time)
Subject: Scientific Plotting?
In-Reply-To: <19990721010203.96101.qmail@hotmail.com>
Message-ID: <Pine.WNT.4.05.9907202134080.152-100000@david.ski.org>

On Tue, 20 Jul 1999, Michael Huster wrote:

> I am also trying to find an adequate *object* based plotting package. I am 
> writing some signal processing software and want to code plotting into the 
> classes I am defining.

[...]

> I am kind of bummed. If we wrap intelligent python wrappers around 
> something, we will post it.

Not that it's in a state to be useful at present, but I feel I must point
you to my back-burner project, Snow.  While it's been useful enough for me
to make figures for several papers and a couple of posters, it's not
stable enough for others to use effectively.  Hopefully, it will be
someday.

   http://starship.python.net:9673/crew/da/Snow

--david






From survey at ipro.com  Tue Jul  6 12:28:36 1999
From: survey at ipro.com (I/PRO Survey Research)
Date: Tue, 06 Jul 1999 09:28:36 -0700
Subject: IGNORE: Test of posting software
Message-ID: <37822EB4.7D0E602E@ipro.com>

This is a test.



From conrad at srl.caltech.edu  Fri Jul  9 20:34:04 1999
From: conrad at srl.caltech.edu (Conrad Steenberg)
Date: Fri, 09 Jul 1999 17:34:04 -0700
Subject: Printing returned object in embedded app
Message-ID: <378694FC.D857F68A@srl.caltech.edu>

Hi

I'm trying to construct an interactive front-end for Python, using GTK.

My C program passes strings to PyRun_String, and obtains the returned object.
Now in the normal interactive Python on a terminal, a command like
1+1
would print out the value, but using my method it does not. Using
print(1+1) works like a charm. However, when I try to have the returned object
printed using eg. 
PyObject_CallMethod(object,"print","") 
I get an error, "AttributeError: print". Can anybody tell me what I'm doing
wrong? Or how to get the result printed automatically like in interactive
Python?

Thanks!

Conrad
--



From eugene.leitl at lrz.uni-muenchen.de  Fri Jul 16 20:00:34 1999
From: eugene.leitl at lrz.uni-muenchen.de (Eugene Leitl)
Date: Fri, 16 Jul 1999 17:00:34 -0700 (PDT)
Subject: Snippet: bitcount()
In-Reply-To: <slrn7oukop.68v.kas@titan.magnetic-ink.dk>
References: <slrn7oukop.68v.kas@titan.magnetic-ink.dk>
Message-ID: <14223.51006.317025.189165@lrz.de>

Ouch, ouch. That was really _inefficient_. At least you could nybble
or byte-wise lookup, and there should be lots of bigmask magic which
does in in O(logN). Consider bitcounting L's. 

Klaus Alexander Seistrup writes:
 >   G'day mates.
 > 
 > Here's a snippet for the repository:
 > 
 > # The function bitcount() will count the number
 > # of bits in a number, iterating at most as many
 > # times as the number of bits.
 > #
 > def bitcount(x):
 >     n = 0
 >     while x:
 >         x = x & (x-1)
 >         n = n + 1
 >     # end while
 >     return n
 > # end def bitcount
 > 
 > 
 > Cheers,
 > 
 >   //Klaus
 > 
 > -- 
 > ???[ Magnetic Ink ]?????????????????????????????????????????????????? ><> ???
 > ???[ http://www.magnetic-ink.dk/ ]???????????????????????????????????????????
 > 
 > -- 
 > http://www.python.org/mailman/listinfo/python-list



From Timo.Rainio at p98.f112.n480.z2.fidonet.org  Thu Jul  1 09:00:52 1999
From: Timo.Rainio at p98.f112.n480.z2.fidonet.org (Timo Rainio)
Date: Thu, 01 Jul 1999 14:00:52 +0100
Subject: [Q] dbase-file/clipper-index read-module
Message-ID: <00001601@bossar.com.pl>

From: "Timo Rainio" <timo at prompt.deANTISPAM>

hello world,

anyone around who knows about a Python-module for dbase-file (.dbf) access?
support for clipper-style NTX-index structures would be great. I only want
to _READ_ clipper-files here.

TYIA!

.Timo




From holger at phoenix-edv.netzservice.de  Mon Jul 12 12:25:21 1999
From: holger at phoenix-edv.netzservice.de (Holger Jannsen)
Date: Mon, 12 Jul 1999 16:25:21 GMT
Subject: Win32: python-network-installation?
References: <378349DE.7ED68819@phoenix-edv.netzservice.de> <37835B30.BFCC4A05@phoenix-edv.netzservice.de> <7m1led$97e$1@m2.c2.telstra-mm.net.au>
Message-ID: <378A16F1.9C06EA7A@phoenix-edv.netzservice.de>

Hi Mark,

Mark Hammond answered:
> 
> Holger Jannsen wrote in message
> <37835B30.BFCC4A05 at phoenix-edv.netzservice.de>...
> 
> >To reach Python-dlls I copied the dlls '\WinNT\System32\PythonCOM15.dll'
> >and '\WINNT\System32\Python15.dll' to a directory '\\as1\python\dll'
> >onto the server.
> 
> Dont forget pywintypes15.dll.

Ouch, yeah! Thanx;-)

> >3)only '\\as1\python\pythonwin\pythonwin.exe' occurs a
> > 'Microsoft Visual C++ Runtime Library'-RuntimeError:
> > Program: \\as1\python\pythonwin\pythonwin.exe
> > abnormal program termination
> >
> >What do you think, does I forget any dll? Why does that error (3)
> >occur? (Perhaps I need that pythonwin.exe to check some
> >scripts when sitting at a client...?!)
> 
> That error normally means 2 Python DLLs were found.  My guess is either
> pythoncom15.dll or pywintypes15.dll.

It was because of pywintypes15.dll missing. Now I only get module 
'win32ui.pyd' missing. Check that tomorrow, I think...

 
> The more I think about it, the problem is almost certainly the fact that
> win32api.pyd, win32ui.pyd etc all use pywintypes15.dll - not as a Python
> module, but as a regular DLL.  So assuming you still have pywintypes15.dll
> in your system directory, the DLL load process will be:
> 
> * Python does "import win32api" - part of the NT process finds and loads
> pywintypes15.dll, probably from the system directory.
> * Python does "import pywintypes" - Python searches for Pywintypes, consults
> the registry and finds it listed somewhere else.  Python loads this DLL.
> 
> Things then go boom very quickly.
> 
> I would put all the DLLs (not necessarily .pyds, but python15.dll,
> pywintypes15.dll, pythoncom15.dll) and all the .EXEs (python.exe,
> pythonw.exe and pythonwin.exe) into the same directory.  Then seek out and
> destroy all other copies of these files from your machine.
>
> >Are there other (better) ways to get a network installation for
> Win32-Clients?
> 
> Nope, but feel free to help :-)

I think to make that easy registry-changes is a good way..;-)
Don't like to spent more time than usual on it.
 
> >[HKEY_CURRENT_USER\Software\Python\Python for Win32]
> ALL of these HKCU\Software\Python\Python for Win32" entries should go.  the
> installer doesnt create them - Pythonwin does when it shuts down for the
> first time.  Unless there are specific preferences you want to default for
> these users, dont add them.

Ok. Checked!
 
> >[HKEY_LOCAL_MACHINE\SOFTWARE\Classes\Python.File\AutoRegister]
> >@="\\\\as1\\python\\DLL\\PythonCOM15.dll"
> 
> Dunno where this came from - the installer didnt add it.

Mmmh. Will check tomorrow...
 
> >[HKEY_LOCAL_MACHINE\SOFTWARE\Python\PythonCore\1.5]
> 
> You are missing all the important ones - PythonPath and Modules keys are
> critical for the win32 extensions to work.

Yes, could be. It was just a fast run. Have to look some further...
 
> Note the installer sources are on my starship page (in the "buildtools"
> archive).

Mmmh? Loaded all sources. Couldn't find a special souce/install-file where
I could take all registry-items from.;-( 
 
> Also note you can become a registered user :-)
> http://starship.python.net/crew/mhammond/registration

;-))I'm a poor programmer...;-))))


Ciao,
Holger



From j.l.gomez-dans at sheffield.ac.uk  Thu Jul  8 07:54:43 1999
From: j.l.gomez-dans at sheffield.ac.uk (Jose L Gomez Dans)
Date: 8 Jul 1999 11:54:43 GMT
Subject: Python and Fortran code
Message-ID: <7m23i3$ktr$1@bignews.shef.ac.uk>

Hi!
       	I have a quetion on using python to do some simple processing on
some data I pre-process with some Fortran code. I wanted to be able to call
this subroutine from within python, as it is already "tried and tested" and
optimized. Basically, my subroutine looks like this:

	subroutine ex1(exdata)
	dimension exdata(256,200)
	print*,'Doing some processing in FORTRAN'
 C      We said we were doing some processing, so we might as well do it :)
 C	Processing of exdata
 	return
	end
	
	This subroutine would need to get input from the user and would also
need to print stuff out for the user to see several parameters as the
software runs. I'd like to pass exdata as a Python object (NumPy array
object), and then recover it with the results. I wonder if I could somehow
do that compiling this subroutine as an object file and then calling it from
python (not that i know how to do it, but with scilab, it works :D).

	If it's of any use, I am using Debian GNU/Linux and g77. And using
loads of complex numbers :))))

	Thanks in advance!
	Jose


-- 
Jose L Gomez Dans			PhD student
					Radar & Communications Group
					Department of Electronic Engineering
					University of Sheffield UK



From pereira at research.att.com  Thu Jul 15 22:55:12 1999
From: pereira at research.att.com (Fernando Pereira)
Date: Fri, 16 Jul 1999 02:55:12 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com>
Message-ID: <150719992255127735%pereira@research.att.com>

In article <7mlvju$6v3$1 at brick.cswv.com>, Neel Krishnaswami
<neelk at brick.cswv.com> wrote:

> In article <7mkrrt$a81$1 at cronkite.cc.uga.edu>,
> Graham Matthews <graham at sloth.math.uga.edu> wrote:
> >
> >[Parser Combinators] 
> >
> >They are something like what Neel is talking about (as the name
> >"parser combinators" suggests you write small "string parsers" as
> >functions, and then have "parser combinators" that stitch these
> >together). Note that in this parser combinator approach there is no
> >distinction between the code for "the pattern" and the code for "the
> >processing" -- they are all written in the same language
> >(Clean/Haskell in my case). 
<snip>
> This really seems like a clean and straightforward way of doing
> things.  My experience with recursive descent parsers has been pretty
> bad, but now I suspect that this is because I wasn't organizing my
> work in a systematic enough fashion. 
Clean and straightforward it may be, but not efficient, and subject to
termination problems on left-recursive grammars. Techniques for
compiling regular expressions and deterministic context-free grammars
into efficient pattern-matching/parsing automata do not extend readily
to programs in a general purpose declarative (functional or logic)
programming language. Certain clever things can be done (in particular
see "The Functional Treatment of Parsing," by Rene Leermakers,
Kluwer Academic Publishers, 1993), but, as far as I know, nothing beats
the automata-theoretic optimization techniques for pattern matching and
parsing, of which the typical modern compiler for regular expressions
is just the best-known instance.

-- F



From thomas at xs4all.nl  Tue Jul 27 08:29:43 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Tue, 27 Jul 1999 14:29:43 +0200
Subject: 1.5.2 for: else:
In-Reply-To: <Pine.LNX.4.04.9907271612140.9849-100000@emerald.netskate.ru>; from Oleg Broytmann on Tue, Jul 27, 1999 at 04:15:44PM +0400
References: <Pine.LNX.4.04.9907271612140.9849-100000@emerald.netskate.ru>
Message-ID: <19990727142943.V4068@xs4all.nl>

On Tue, Jul 27, 1999 at 04:15:44PM +0400, Oleg Broytmann wrote:

>    The following program:

> ----------
> for a in ['a', 12]:
>    print a
> else:
>    print "Empty!"
> ----------

http://www.python.org/doc/current/tut/node6.html#SECTION006200000000000000000

[..]
4.4 break and continue Statements, and else Clauses on Loops 

Loop statements may have an else clause; it is executed when the loop
terminates through exhaustion of the list (with for) or when the condition
becomes false (with while), but not when the loop is terminated by a break
statement. 
[..]

So, 

>>> def do_a(list):      
...     for a in list:
...         if not a:
...             break
...         print a
...     else:
...         print "No FALSE items in list!"
...

>>> do_a(['a', 12])
a
12
No FALSE items in list!

>>> do_a(['a', 12, []])
a
12

>>> do_a(['a', 12, 0])
a
12

Else clauses get executed on normal loop exit, even if the loop was 'empty'.
It does not get executed when you 'break' out of a loop.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From sreekantk at freenet.co.uk  Thu Jul 22 18:03:16 1999
From: sreekantk at freenet.co.uk (sreekant kodela)
Date: Thu, 22 Jul 1999 23:03:16 +0100
Subject: Q: anyone running Python on OS X server?
References: <004401bed44f$1a5c6400$f29b12c2@secret.pythonware.com>
Message-ID: <37979524.C9F4F6BC@freenet.co.uk>

Do you mean Linux / Xfree ?
If that is what you asked then answer is Yes.

Regards
sreekant

Fredrik Lundh wrote:

> just curious...
>
> </F>




From Victor.S..Miller at p98.f112.n480.z2.fidonet.org  Fri Jul  2 09:46:57 1999
From: Victor.S..Miller at p98.f112.n480.z2.fidonet.org (Victor S. Miller)
Date: Fri, 02 Jul 1999 14:46:57 +0100
Subject: python -c not working
Message-ID: <00001693@bossar.com.pl>

From: victor at idaccr.org (Victor S. Miller)

I built Python on various of our systems (1.5b2) (Solaris 2.7 and AIX
4.x), and everything seemed to be ok, until today when I was trying to
install Numerical Python.  I tracked down the error to the fact that

python -c

doesn't work

for example

python -c 'print 1'

gets no output, and

python -c 'import path; print sys.path'

get a syntax error at import.  Does anybody have a clue as to what can
be wrong?

I built it in without such problems on another set of Solaris 2.6
systems.
-- 
Victor S. Miller     | " ... Meanwhile, those of us who can compute can hardly
victor at idaccr.org    | be expected to keep writing papers saying 'I can do the
CCR, Princeton, NJ   | following useless calculation in 2 seconds', and indeed
    08540 USA        | what editor would publish them?"  -- Oliver Atkin



From gmcm at hypernet.com  Wed Jul 21 09:26:10 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 21 Jul 1999 08:26:10 -0500
Subject: Tcl/Python: Who was first?
In-Reply-To: <16E7930929E7C7E5.430278EC189FFB38.BB6793F4819F7C95@lp.airnews.net>
Message-ID: <1279562730-15211949@hypernet.com>

Cameron wrote:

> ... Barry
> has Mailman and JPython and starship under con-
> trol, ...

But at such a cost
Now speaks only haiku
And eats living rodents


- Gordon



From godzilla at netmeg.net  Wed Jul 21 08:56:34 1999
From: godzilla at netmeg.net (Les Schaffer)
Date: Wed, 21 Jul 1999 12:56:34 GMT
Subject: How fast can we multiply?
References: <000401bed335$862ee420$a72d2399@tim>
Message-ID: <m2673etb3s.fsf@netmeg.net>

Tim 'multiply-this!' Peters rejoiced so:

> Turns out it's possible to multiply two 2x2 matrices using 7
> multiplies.  Since matmult can be "blocked", a larger matrix can be
> treated as a 2x2 array of smaller submatrices, and the same trick
> recursively applied.  In the end you get a method with runtime
> O(n**log2(7)) ~= O(n**2.81) instead of O(n**3).

okay, THAT stuff...

Numerical Recipes attributes this to Strassen -- i've seen some stuff
on the web on it too:

http://www.cs.engr.uky.edu/~lewis/cs-alg/materials/Strassen.html
http://www.compapp.dcu.ie/~aileen/research/matmult4.html
http://www.cs.duke.edu/~luoma/strass/strass.html

i'll grab a look at Knuth ....

> The intmult trick is less work to code, and yields a bigger payoff
> earlier.

yeah, and N has to be large enough to justify using strassen ... 

les



From c.rafferty at formulae.co.uk  Mon Jul 19 03:48:11 1999
From: c.rafferty at formulae.co.uk (Colm Rafferty)
Date: Mon, 19 Jul 1999 08:48:11 +0100
Subject: Python/Gadfly and SQL wildcards
Message-ID: <001d01bed1bb$0cecfd20$a95d38d4@default>

Dear all,
I'm new to list.
I'm looking at gadfly to use SQL with python.
Does anyone know how to implement the wildcard in gadfly SQL?
e.g.
Select * from table where fieldname = 'A*'
or
Select * from table where fieldname like 'A*'

* does not work!
Though it does in other SQL languages

Thanks for any help

Colm Rafferty



From arcege at shore.net  Thu Jul  1 11:17:40 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 01 Jul 1999 15:17:40 GMT
Subject: Why must path be in args in os.execv?
References: <377A2590.D95D72FE@phoenix-edv.netzservice.de>
Message-ID: <oELe3.511$6M6.157342@news.shore.net>

Holger Jannsen <holger at phoenix-edv.netzservice.de> wrote:
: High,

: there's again something I don't understand. Hope you do.;-)

: I'd like to start a windows-executable with some arguments.
: Ok. 'os.execv' or anything like that is the thing.

: But if I don't repeat the path with the arguments, the execution fails.
: Better: Something happens, but not the correct execution.

: For testing I constructed an easy batch named test.bat with only one command:
: echo %0 %1 %2 %3 %4 %5 >result.log

: Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
: Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
:>>> import os
:>>> os.chdir("c:\\mypython")
:>>> os.execv("test.bat",["test.bat","1","2","3","4"])

: When I start with:
:>>> os.execv("test.bat",["1","2","3","4"])
: only the interpreter ends. No Log-File has been written...

: I really spent a long time since I found out. Perhaps the documentation
: have to be changed?
:>>>
: execv (path, args) 
: Execute the executable path with argument list args, replacing the current process (i.e.,
: the Python interpreter). The argument list may be a tuple or list of strings.
: Availability: Unix, Windows. 
:>>>

os.execv is a construct originating on UNIX, where the program name
does not always reflect the program file's name (with reason).  It is
common to create an executable that can react differently when called
with a different name.  As a simple example, gzip, gunzip and gzcat are
all the same physical file on a UNIX box, but called with those names,
it will compress, uncompress and uncompress-and-stream-output,
respectively.

Including the program name in the execv call allows this from a calling
process.

But in most cases, you can just use the program file:
  def myexec(program, *args):
    if len(args) == 1 and type(args[0]) == type(args):
      args = args[0]
    os.execv(program, (program,) + args)

This places the program filename in the arg0 position (see section 6.1.5
"Process Management" in the Python Library Reference.

  -Arcege




From news at dorb.com  Fri Jul  2 14:03:30 1999
From: news at dorb.com (Darrell)
Date: Fri, 2 Jul 1999 14:03:30 -0400
Subject: Tuples & lists
References: <377bbde1@news.xtml.co.uk> <7lh375$5ea$1@news.udel.edu>
Message-ID: <2c7f3.4658$U6.26365@newsr2.twcny.rr.com>

Did an experiment on NT and lists and tuples took the same amount of space.
Humm ? Not what I expected.

--
--Darrell
Terry Reedy <tjreedy at udel.edu> wrote in message
news:7lh375$5ea$1 at news.udel.edu...
> In article <377bbde1 at news.xtml.co.uk>, nick at NOSPAMvideosystem.co.uk
says...
>
> >Why both having tuples and lists? As far as I can see there're the same,
but
> >tuples are read only..? Have I missed something in the tutorial?
>
> There is probably something in the the FAQ.  BAsically, tuples are stored
> more compactly and can be used as dictionary keys (because they ARE
> readonly).
>
> TJR
>





From Darrell at p98.f112.n480.z2.fidonet.org  Thu Jul  1 05:48:09 1999
From: Darrell at p98.f112.n480.z2.fidonet.org (Darrell)
Date: Thu, 01 Jul 1999 10:48:09 +0100
Subject: Name of function within the function?
Message-ID: <0000162a@bossar.com.pl>

From: "Darrell" <news at dorb.com>

import sys, string, os

def myDirExcept(howFarBack):
        """
        Get the directory of the module running at "howfarBack "
        in the stack frames
        This value is compiled into the .pyc file
        This can be a problem.
        This name will change depending on where python was run from
        when the pyc file was created
        Unix can store a relative path
        Windows stores an absolute path
        """
        try:
                raise ZeroDivisionError
        except ZeroDivisionError:
                f = sys.exc_info()[2].tb_frame

        for i in range(howFarBack):
                f = f.f_back

        t = f.f_code
        fname = t.co_filename
        p = string.split(fname, os.sep)
        fname = string.join(p[:-1], os.sep)
 print t.co_name, fname
        return fname


myDirExcept(0)

###################################

E:\ace\v2\code\comp>junk2.py
myDirExcept E:\ace\v2\code\comp


--
--Darrell
Dula, Debbie (EXCHANGE:RICH2:2C25) <violet at americasm01.nt.com> wrote in
message news:377B8C2F.80F44F9 at americasm01.nt.com...
> I've looked and looked, but can't find the answer to this ... I'm
> sure it's right in front of me ...
>
>
> I want the current function's name to be output in my error
> messages, but can't figure out how a function can get its own name
> (without already knowing it).  For example,
>
> def f1():
> print "Error in function", f1.__name__
>
> is not really useful for what I need as I might as well print "Error
> in function f1".  Is there any generic way I can get the current
> function's name?
>
> Thanks,
> Debbie




From fire at netlimit.com  Mon Jul 19 17:06:21 1999
From: fire at netlimit.com (Joacim Löwgren)
Date: Mon, 19 Jul 1999 21:06:21 GMT
Subject: SMTP problem
Message-ID: <hrMk3.1158$Gc6.192154624@newsb.telia.net>

Hi,

I copied this straight from the Python SMTP example, however it doesn't seem
it's working. I'm not quite sure what ^D does and if it is different on
Unix/Windows as I think this is the problem with the script.

_________________________________

import string, sys
import smtplib

def prompt(prompt):
    sys.stdout.write(prompt + ": ")
    return string.strip(sys.stdin.readline())

fromaddr = prompt("From")
toaddrs  = string.splitfields(prompt("To"), ',')
print "Enter message, end with ^D:"
msg = ""
while 1:
    line = sys.stdin.readline()
    if not line:
        break
    msg = msg + line
print "Message length is " + `len(msg)`

server = smtplib.SMTP('localhost')
server.set_debuglevel(1)
server.sendmail(fromaddr, toaddrs, msg)
server.quit()





From dkuhlman at netcom.com  Fri Jul  2 14:50:54 1999
From: dkuhlman at netcom.com (G. David Kuhlman)
Date: 2 Jul 1999 18:50:54 GMT
Subject: Installing Python/XML on Windows?
References: <377bab1c@news.depaul.edu>
Message-ID: <7lj1me$ms6@dfw-ixnews5.ix.netcom.com>

Jonathan W Hendry (jhendry at shrike.depaul.edu) wrote:
> Does any documentation exist on this? How do I set up the
> paths so that Python can use all the various components of
> the package?
> 
> Thanks,
> 
> - Jon


Here is a message that should help.  You should also be able to
find it in the comp.lang.python archives at DejaNews.

  - Dave

==============================


From: digitome at iol.ie (Sean Mc Grath)
Newsgroups: comp.lang.python
Subject: Re: XML-toolkit for Windows - binaries available?
Date: Wed, 16 Jun 1999 19:37:35 GMT
Organization: Ireland On-Line
Lines: 33
Message-ID: <3767fc51.1547455 at news.iol.ie>
References: <376701a3.72288277 at newssrv> <wkzp208udt.fsf at ifi.uio.no>
NNTP-Posting-Host: dialup-011.ballina.iol.ie
X-Newsreader: Forte Free Agent 1.11/32.235
Xref: ix.netcom.com comp.lang.python:61665

On 16 Jun 1999 09:41:34 +0200, Lars Marius Garshol
<larsga at ifi.uio.no>
wrote:

>
>* Lee Borkman
>|
>| Can anyone help me out in installing from the standard
distribution?
>| I don't have a C-compiler on my system, but I do have
nmake/dmake.
>
>I think the web page is out of date, since the /xml/windows
directory
>in the CVS does contain the .dll files for pyexpat and sgmlop (and
has
>done for quite a while). So just download the ZIP file and they
should
>be there.
>
>If not, send me an email and I'll send them to you.
>
The binaries are there but a bit of fiddling is required to
get things working. (This is from memory but here goes)

The pyexpat and sgmlop dlls have to be copied into the
xml/parsers subdirectory. You also need to edit the
__init__.py file to include 'pyexpat' and 'sgmlop' in
the __all__ string.

Pyexpat requires xmlparse.dll. This dll needs to be
on your path.

I *think* that is what I did to get it all working:-)

Hope it helps,
Sean






From aaron_watters at my-deja.com  Wed Jul 28 13:29:14 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Wed, 28 Jul 1999 17:29:14 GMT
Subject: Gadfly DBase and CGI
References: <7nmkpg$cjl$1@gruvel.une.edu.au>
Message-ID: <7nnel0$6e4$1@nnrp1.deja.com>


>
> Has anyone ever done this?  I keep getting an error in the script
every time
> I go to create a new connection to an existing database.  What
permissions
> must I have (755)?

Usually, you should be running the database in client/server
mode, but there are exceptions, and in the exceptional cases
you should add some "mutual exclusion" trick not provided by gadfly
per se to make sure two cgi scripts never attempt to modify the
same database at once.

In any case the user that opens the database (the one that starts
the server or [exceptionally] the cgi user) should have the ability
to modify the database files and the database directory, if the
database will be modified.

Good luck!  -- Aaron Watters http://www.chordate.com

===

Never attribute to malice what can be explained by incompetence.
   --folklore


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From awrede at home.com  Tue Jul 13 18:03:04 1999
From: awrede at home.com (Andreas E. Wrede)
Date: Tue, 13 Jul 1999 22:03:04 GMT
Subject: problem with SocketServer -- address already in use
References: <931899928snz@vision25.demon.co.uk>
Message-ID: <378BB798.9C86451E@home.com>

Phil Hunt wrote:
> 
> I'm trying to use SocketServer.py to write a TCP server. I ran
> my program with ``python serv.py'', which went into an infinite
> loop (which is what it should do). I exited my program with ^Z,
> and tried to start it up again. I got an error message:
> 
> socket.error: (98, 'Address already in use')
> 
> which I presume is because something thinks the old invokation of
> my program is still using port 7070.
> 
> How do I clear the port, before I start up my server?

The sockets stays allocated and in use for some timeout period after you
close it. 
I solved this one by setting the SO_REUSEADDR option thru overridding
server_bind:

def server_bind(self):
    self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR,1)
    SocketServer.TCPServer.server_bind(self)


-- 
    - aew



From emile at fenx.com  Fri Jul 30 22:31:01 1999
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 30 Jul 1999 19:31:01 -0700
Subject: os.name
References: <1278785134-25350540@hypernet.com> <V5ho3.4340$VL2.463886@news.direcpc.com> <001e01beda8e$2111d2e0$f29b12c2@secret.pythonware.com>
Message-ID: <Ggto3.4361$VL2.467372@news.direcpc.com>

By the way, in windows at a command prompt:

95,98:    net config
nt:       net config [server|workstation]

gives a line that starts : Software version

I thought I'd key in it a script real quick, but I don't have
enough time now to wrap it up.  Maybe it will get you started.


def getwinver():
 import os
 winver = None
 test = os.system('net config > xxconfig.net')
 if test == 0:
  f = open('xxconfig.net', 'r')
  config = f.readlines()
  f.close()
  for line in config:
   print line
   if line[:16] == "Software version":
    winver = line
 else:
  pass # test value is error code
 return winver



--

Emile van Sebille
emile at fenx.com
-------------------


Fredrik Lundh <fredrik at pythonware.com> wrote in message
news:001e01beda8e$2111d2e0$f29b12c2 at secret.pythonware.com...
> Emile van Sebille <emile at fenx.com> wrote:
> > >>> os.name
> > 'nt'
> > >>> os.environ('OS')
> > Traceback (innermost last):
> >   File "<pyshell#8>", line 1, in ?
> >     os.environ('OS')
> > AttributeError: no __call__ method defined
>
> Gordon asked you to try os.environ['OS'],
> not os.environ('OS') ;-)
>
> </F>
>





From max at rightworks.com  Wed Jul  7 02:51:02 1999
From: max at rightworks.com (max at rightworks.com)
Date: Wed, 07 Jul 1999 06:51:02 GMT
Subject: rfc822 from 1.5.2 fails when called from cgi
Message-ID: <7lutck$4jd$1@nnrp1.deja.com>

I have a cgi app that is failing with 1.5.2 version of rfc822.  When I
post a multi-part form, sometimes, I get the failure shown below. It is
interesting that running the same data doesn't always fail. I have tried
to put trace in, and it varies which iteration thru readheaders it fails
on.  I have tried to trace with print stmts (can't always run debugger
inside of cgi) and can't figure it out.
    form = FieldStorage()
  File "G:\Python\Lib\cgi.py", line 869, in __init__
    self.read_multi(environ, keep_blank_values, strict_parsing)
  File "G:\Python\Lib\cgi.py", line 954, in read_multi
    headers = rfc822.Message(self.fp)
  File "G:\Python\Lib\rfc822.py", line 92, in __init__
    self.readheaders()
  File "G:\Python\Lib\rfc822.py", line 137, in readheaders
    startofline = tell()
IOError: (0, 'Error')

the file it it parsing looks something like:

Content-type: multipart/form-data;
boundary=---------------------------119851875720143
Content-Length: 565

-----------------------------119851875720143
Content-Disposition: form-data; name="SI"

0000000S
-----------------------------119851875720143
Content-Disposition: form-data; name="command_file";
filename="D:\ProcureWorks\WebDriver\t.txt"
Content-Type: text/plain

# Adding 1 User
*delimiter ","
*userdata username,last_name,first_name,e_mail,EMPLOYEE_ID
ruipeng2,Mai,Ruipeng,ruipeng at rightworks.com,18


-----------------------------119851875720143
Content-Disposition: form-data; name="verbose"

on
-----------------------------119851875720143--

If I run with the ver 1.5.2 of rfc882 it works fine.
any help appreciated.
max


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From herzog at online.de  Thu Jul 29 05:55:47 1999
From: herzog at online.de (Bernhard Herzog)
Date: 29 Jul 1999 11:55:47 +0200
Subject: String2List: odd function
References: <199907272152.QAA01065@cro.ots.ac.cr> <379F731B.C2C39989@compaq.com> <001701bed947$0d4fedc0$f29b12c2@secret.pythonware.com>
Message-ID: <m3iu73kcf0.fsf@greebo.nodomain.de>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> well, last time I checked, eval didn't accept statements.

True, but if the eval'ed code has access to the normal builtins, you can
sneak statements into the expression with compile:

>>> eval('''eval(compile("for i in range(10): print i", "", "exec"))''')
0
1
2
3
4
5
6
7
8
9
>>> 


-- 
Bernhard Herzog	  | Sketch, a python based drawing program
herzog at online.de  | http://www.online.de/home/sketch/



From graham at sloth.math.uga.edu  Tue Jul 20 12:24:08 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 20 Jul 1999 16:24:08 GMT
Subject: Stackless & String-processing
References: <000601bece8f$b549a420$51a22299@tim> <7mkrrt$a81$1@cronkite.cc.uga.edu> <7mlvju$6v3$1@brick.cswv.com> <150719992255127735%pereira@research.att.com> <170719990020380651%pereira@research.att.com> <7mqaht$r27$2@cronkite.cc.uga.edu> <170719991835331004%pereira@research.att.com> <7mssne$m8r$1@cronkite.cc.uga.edu> <180719992337546352%pereira@research.att.com> <7mvika$cpj$1@cronkite.cc.uga.edu> <200719990035531711%pereira@research.att.com>
Message-ID: <7n27r8$mm4$1@cronkite.cc.uga.edu>

<graham at sloth.math.uga.edu> wrote:
> What your combinators do depends on your application. If all you are
> doing is parsing using primitive automaton based parsers then your
> combinators won't have much to do. But if you are doing more than
> that then they will have more to do, and hence be more complex. For
> example you can use the same combinators you define for parsing to do
> type checking, model state, etc -- things I would not want to do with
> an automaton. That is why I said that combinators are really sequencing
> constructs.
Fernando Pereira (pereira at research.att.com) wrote:
: *What* I've been saying is that there is a generality-efficiency
: tradeoff, and that the monadic approach goes for generality at the
: expense of efficiency, by precluding optimizations in those very
: important cases in which combinators are used for parsing. 

And I am saying that this claim that generality implies loss of
efficiency is a red herring. If you are writing a program that uses
combinator based parsing there is nothing to stop you using automaton 
based techniques in your program. You just add the appropriate
automaton based primitives (eg. for regular expressions). You
might end up adding in so many automaton based primitives that
you don't need to do much with your combinators, but whether or
not that happens depends entirely on your application and your
needs. The point is that combinator based parsing doesn't prevent
you from using other more efficient techniques as and where
appropriate!

Have you ever written a combinator based parsing program?

<graham at sloth.math.uga.edu> wrote:
> Your point appears to be that if you were programming only using automaton
> for parsing you wouldn't miss the above optimisation, but that somehow
> using combinators means you would miss the above optimisation. You
> and I must program differently then. If I were using automaton, even
> in the context of also using combinators, I would be aware of the
> potential for such optimisations and look for them.
Fernando Pereira (pereira at research.att.com) wrote:
: You might be, by looking at the code and changing it by hand. But a
: program would not, because in general the optimizable skeleton of
: sequence, alternation and Kleene closure combinators is interspersed
: with arbitrary functional code.

Could you please point out to me where I claimed that "a program could
optimise your combinator program". If you can't find such a quote then
what is your point?

Fernando Pereira (pereira at research.att.com) wrote:
: But from my readings (those mentioned in the thread
: and others) as well as your comments, the whole point of monadic
: combinators for parsing is to interweave the grammatical structure with
: other computations, thus making the underlying grammatical structure
: less accessible to optimization. As I suggested above, it would be
: interesting to define classes of monadic programs that still allow (an
: appropriate extension of) those optimizations. But AFAIK that has not
: been done.

It hasn't been done, it won't likely be done, and I never said it
could be done. So what is your point?

graham
-- 
            The girls is all salty, the boys is all sweet, 
                      the food ain't too shabby 
                    an' they piss in the streets
    In France, way down in France, way on down, way on down in France



From slj at pixar.com  Thu Jul 15 14:35:10 1999
From: slj at pixar.com (Steve Johnson)
Date: Thu, 15 Jul 1999 11:35:10 -0700
Subject: Calling superclass method with variable args?
Message-ID: <378E29DD.1613913E@pixar.com>

Hi,

I'm a long-time C++ programmer, but I'm really new to Python, so please
forgive me if this is a dumb question with a simple or obvious answer.

A fundamental idea in OOP is the ability to have a subclass augment the
behavior of a superclass's method by redefine that method, and then
calling the superclass's version of the method before doing anything
else.  The key to this idea, and to polymorphism in general, is that the
two methods have the same argument signature so that the superclass's
method gets passed the same set of arguments from the caller whether the
subclass overrides the method or not.

If these methods have a static argument list, this is an easy thing to
do in Python. Here's an example:

class Task:
    def __init__(self,a,b,c):
        <whatever>

class Job(Task):
    def __init__(self,a,b,c):
        Task.__init__(self,a,b,c)
        <whatever>

aJob = Job(x,y,z);

But what if the method takes unnamed positional and keyword arguments.
Then, how do you pass these along to the superclass's version of the
method such that it gets the arguments in the same form that the
subclass's method did?  Here's the code in question:

class Task:
    def __init__(self,a,b,c,*moreargs,**morekeyargs):
        <whatever>

class Job(Task):
    def __init__(self,a,b,c, *moreargs, **morekeyargs):
        Task.__init__(self,a,b,c,?????)
        <whatever>

aJob = Job(x,y,z,m1,m2,m3,m4,m5);

My question really is...what do I replace the ????? (or that whole line)
with to get this fundamental OOP behavior?

This doesn't work:

        Task.__init__(self,a,b,c,moreargs,morekeyargs)

because now Task.__init__ will see exactly 5 arguments, instead of some
arbitrary number.  This almost works:

        apply(Task.__init__,(self,a,b,c) + moreargs)

but the extra keyword arguments don't get passed along. So, I took the
idea one step further and tried to introduce the keyword arguments back
into the parameter list that gets passed to apply:

def reconstruct(attribs):
    r = ()
    for key in attribs.keys():
        r = r + (key + "=" + attribs[key]",)
        print r
    return r

class Task:
    def __init__(self,a,b,c,*moreargs,**morekeyargs):
        <whatever>

class Job(Task):
    def __init__(self,a,b,c, *moreargs, **morekeyargs):
        apply(Task.__init__, (self,a,b,c) + moreargs +
reconstruct(morekeyargs))
        <whatever>

aJob = Job(x,y,z,m1,m2,m3,m4,m5);

but Task.__init__() doesn't recognize the keyword parameters as such,
instead it thinks they are string arguments of the form "foo=bar".

Can anyone tell me how to get the classic OOP behavior I'm looking for
here?

TIA,

Steve Johnson
Pixar Animation Studios





From petrilli at trump.amber.org  Wed Jul  7 14:26:18 1999
From: petrilli at trump.amber.org (Christopher Petrilli)
Date: 07 Jul 1999 11:26:18 PDT
Subject: What's ZOPE anyway?
References: <3scg3.1267$B95.91660@typ42b.nn.bcandid.com>
Message-ID: <7m064a$921@journal.concentric.net>

Jayr Magave <jmagave at amazon.com.br> wrote:
> I went through www.zope.org and read everything I could, and found the
> following definition:

> "Zope is an Open Source application server and portal toolkit used for
> building high-performance, dynamic Web sites. "

> ... and the rest of the page goes on to describe ZOPE.

Unfortunately today, it's very difficult to avoid "hype" and just describe
what something is.  An "application server" is a foundation for building
an application.  In the case of Zope, this includes:

	* Full concurrent object database with undo and versions
	* templating language
	* URL -> Object conversion
	* Management environment

Some people have compared it to Cold Fusion, but the reality is that Cold
Fusion is a much more limited conceptual model.  The only real competitor
from an architecture perspective is Apple's Web Objects.   JUnfortunately,
a 30 word snippet doesn't provide any indication of the power.  I have
taught 3 day classes only as an "introduction" to the concepts, hardly
expert. There's just that much there.

Chris
-- 
| Christopher Petrilli                      ``Television is bubble-gum for
| petrilli at amber.org                          the mind.''-Frank Lloyd Wright



From news at dorb.com  Thu Jul  1 17:08:57 1999
From: news at dorb.com (Darrell)
Date: Thu, 1 Jul 1999 17:08:57 -0400
Subject: Compiling Python for Win32
References: <7lgjij$180bi_004@news.sentex.ca>
Message-ID: <ZPQe3.4335$U6.23535@newsr2.twcny.rr.com>

> What I would like is if somebody who has successfully built Python 1.5.2
> using MSVC++ could do the "Export Makefile" command and send me the
> result. A simple text list of the individual source files that make all
> of the Windows versions of the various python executables, libraries and
> DLLs, along with accompanying defines would also be welcome, and either
> will earn somebody my undying thanks. :-)
>
> Thanks in advance, eh!
>

Ok. I'm waiting for the undying thanks part.
Hope this works for you.
--
--Darrell


begin 666 makes.zip
M4$L#!!0````(`+:(X29V[[%WL 4```@7```)````8G-D9&(N;6%K[5A[;]LV
M$/^[!OP=KFI0Q, BY8$5G5 #=6PY<"L_8,5+.Q at P9)&V.4NBH$<28]UWWTFB
M9,G/95FZ#5A@!"3OP;OCW?THOH$NLWP>\%D(+7I/;>Y1'XPP(HS##76I;X:4
M0*_;^*Q!F]GT![@V`USA+DP#0J8R";QJY76G#=+):;-]4Y. at 7@=)JE9P4D]8
MX SNF'MUB1M,HSDR=S7#:-QHT.-@<7?&YA'NPE!CX%&+S1 at E,O+.S, at .F3N'
MD,,./3(JTGHMW+A:V3#@-1I0EAA2FZ+9$KQ]>X at M42P5+.RX]Z;-R(:9N8:U
MO069KSP"R\R<68&Y(?VPH"[XD>O&OB6!+0A/5T#HC"6T<$'!,?%T`'>+`QXO
M6-QQ3)> S5PJ0YO[0!]-Q[.I6M!2C'%R<LI,N"H[YE**]=6/N5X8#G at 0L*F-
MFR\XLV@`,]RV[)3I[S- at WTF<3K,\DE+*Z>/[=S5HK5S38=:9SMPEZ&SJF_Y*
MJNW7E]K]E[7%230<]H?0<('M/&P6E(]Y*^?ZALAYW)CPAV#2N\4X]D:Z7H_9
M=2/>)9FZD5U2T!P,ZI8MTT=:K71O];K#B)@-C6;=M])QNE&YO';'-.;MCVY;
MG6%='J.MG9X8A]3QSOR4JUIY`]=TSEQH1D'('>C&*1:@9!2VF)]*O@$-<VR#
M(?=XJ#5'0T,3MIQ+B3,-70<UB4=B06V<YINW(HE=62!R1F\5+KA[\>-V9AS0
ML1G[LB47L25-76OT0-VK/Z%+L+9("!34OSK[B'TOH/$.:0R%)0XGD4UE/OU5
MVLUU;[T[EQF9'B)[6^2RM_31.TBWV6'Y8_JS:*X)4NP]X!^;@<M#["@L" N"
MRLEIG+_(YRP)\PN43)%P<*\B0=^A*).,%;5_.J^369KU6!N3P;#_J:ZXW.9S
M#DJW!<K=%2 at W7T#YA8'2OP2E`Y(LCSNN94<$,T?,!\U\B+_$ZS';X,D)'O?#
M`K4%4J^E78]NTO%=IW=UF0XG.&[U[PR<M;W-S) ]:X&$KVA;FQ>(XYB;;"Z@
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M8+FC`RI"?IAV]3SN_>M/XFP.]*&,GNV6M3W1)HIE7K2SV*DE.#EM:>U)NZ-K
M-1P7MJ^EY9RMU>#CAP_Q4LZ4Y,:V4,PE.NP1[,KN'8>0BZ3WQN^)6Q/R!.02
M=Y"#.IZ#7(G^?SMNH;='D LYF+T\PG$$W>*8'K7COXMP1$"<(V"N at S!'7 at CF
M)O\HS$W^A[GO"G,K_&S;B7.BHO8 at G<M)^B&>@%K7^+DYO)5V`Z HO+T0*(H[
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MJOZ,-:VRDAU_`%!+`P04````" "VB.$FB(SYA^,%``"7& ``# ```%]T:VEN
M=&5R+FUA:^U8ZV_:2!#_7"3^AZD;54&Z0!ZZJK5JJ01,SJUY"(/2GI"0L1?8
MP_9:?B1!U_O?;]9>&_-.+FW5.QT?HO7.>W9F?IM]!6UJ!2QDTPB:Y(XXS"<!
M&%%L4P8WQ".!&1$;.NWZ)Q5:U"&_P+49X@[S8!PMJ!>1H&J'?KGT4FN!='+:
M:-U4)% 4D*1R"3^4C O.X)9Z5Y=H9A+/D+^M&D;]1H4.`XMY4SJ+T19%O:%/
M+#JEQ*XB[]2,G8AZ,X at 8[%9515UJIXGFRZ4--UZB&UM"?>(0#$&"UZ^/<";J
MI8*KFG=G.M3>\#=7LG*\(/.%Q6"9651+,#>D[^?$@R#V/!YDDN>"\&0)-IG2
MA!;-";@F'A:@-9Y_OF$QUS4]&QSJD2JT6 #DP71]A\@%+<5D)P=9FZZBK;KF
M0N(JE4<DH+#LL3"D$P==F#-JD1"F:'P]-#/8Y\:!4SF=9/4EI933A[=O*M!<
M>J9+K3.=>@O0Z20P at Z54.:@R]?X?*^1EU>]W^U#W at .X\>!JN'_E6%78-T0MH
MV&;WX;@SP&QVAKJN<';=X%:23R]VUA0T>CW%<JKD at 91+[8&NN-067WVCH016
MNDX-K;?=WLQR]NYPT-3Z2G6$[FH=L8Z(ZY\%*5>Y] JNR8QZT(C#B+G0YA47
MHF0<-6F02KX"%4MN at R$/NJ\VAGU#%>Z<2TD\=5T'.4E)XD%EE)>?O[03U[)T
MY+S^,IHS[^+7[1(YK&;S$-;]N>#^-'2UW@%YKXF$+L'**2%04/_B[ /.QI!P
M"VDF"\ZPR1_2;I9H8?H^]G-T at .?.>G->I?;D$-G?(F]EA#SXQU@<>E3+(PQE
MJ5_1))XJP!^=@L<BG$DTC JRM9-37O7(YRYL&A0HF2(1[%Y%@KY#42;)%;7>
MG2OV-.T5[*AQK]_]J-0\YK 9 at UJ[";7;*ZC=?(;:[Q1JW4NH:2!5JR/-LYS8
MQDH3W[U&NFS(HU[ 9H'I)E at 8C@:6,Z(Y<Q.D3E.]'MZDZUNM<W69+L>X;G9O
MC8PP^&U<[_6TCC; G9:_JX9\:XZT+^A;BQ7H(RY@;VY@)!8DDV(CQJ)+[B):
M^GCDE^<8-$/"4,?=^Z3PRZ5KHW%UJ4Q""_& I"E+ML8MO7YCK#2R726 8E*F
M8VQ<]PT%1E at UY9*N=3ZA6H2G1:HSW1!*XY $5Y>\#&%!`H\XXF-F4[%"[T*?
M,2?Y0+"SG9F at F/:=Z6=LX9PXF3!SR&IEQI'XB&-JI[OVQ,H8<&GYXB.RG+?G
MZ6HA%GG0'$)DZ?SA at ER\.\<?YBV,)^$RQ-DIWZ?#'6JVXT -RR$@+O$BTY$]
M%,4&DO<U%HIPC,*C,:TY(KBL7;U]@RF.H]TBV&AHP.6'N).!]S/4\*]O1G-Y
M=[FF/#@<6!#)?!+E@)$?3O?ZHSC ?<-MD[@^UC)JYE\V9L6\*4Z*C>"*$"'!
MR6E3;8U;FJY6<%UPKI(.A6RO`A_>O^=;.5-27MM"G$L,]>.XF5U^#J&FG=YE
M?S!FCNTGH*:X!1U3\QS43$S\%S 3,W(<-9&).HOC3,?!E1_ 8WSZ]P*L+1#6
M%2BK(<K:QU#6EK=Q=?SSX>KX?UQ]%*[B<48I[HQ6$%O<_%9HN\1_0O?!K6BU
MIP&N:+U#D"OZ/+&+54#DD/C13PS!^5S:"<)9P#\(AG/,J5I_YH'\Q?V7TZF$
M^-SKY6K3SZ014>=[R!15+=]_GOS#P_/DB]+A)'B6]T^77_?^J?(YXG>Z8_7S
M0.UWZCJ.MIXA7L4NI)1#_:P9`^-T]6YD$S]Y+]$Z#7W8+#S ))35/2!_4+DU
M`_Z0)?/G,(XU4^SD'5*B*@KED;OXB%<.^/KUT?>Z)SZ>''B1D"%)L87Q5"61
M;?[25H%:NA"=4,,PJJO[\.KI;;_J-<4'G!#WK^_HB;@2*A>07NTV/'ORM7KO
M7?7;QR 4'\UE\2;[W;PXFL>\\(WNL-]0%;P;M9D=<_C(N\5:O\=M8(.,OJ;"
ME?4KVPLQ%-:&0<::]MJVT16T;%K=`)WGF%T+_&]02P,$% ````@`MHCA)B;L
M2Q:I!0``Y!8```H```!P87)S97(N;6%K[5CIC]I&%/\<)/Z'%R>*%JEK[Z%&
MJ15+8<&L2,PAO&B3"@D9>X IML?RL0MJ^K_WC3TVYNYVD[25R@<TQSM^[\T[
MQO,*.M0.6<2F,33)`W%90$(PX\2A#&Z)3T(K)@YT._5/.K2H2WZ"&RO"%>9#
M8(41"64G"JJ5E^T62*_/&JW;F at 2:!I)4K>!$RVC@'.ZI?WV%*B;)#*D[NFG6
M;W7H,K"9/Z6S!/50E!D%Q*932AP9::=6XL;4GT',8)\@&27IW2:JKE:V(+Q$
M"%LL`^(2A"[!FS='Z5+14 at EDVW^P7.IL(2U$K"&7>+ZP!&PKMV<%UA;WXYSX
M$":^S\U+O5MBGJS (5.:[L5S`IZ%1P2HC7N=+]C,\RS?`9?Z1(86"X$L+2]P
MB5J24G9S>GS*-+=5]JR%Q 5J)XTO#?LLBNC$1?5S1FT2P105;YIEA8<@'#R-
MLTD>3U*V<[9\][8&S95O>=0^-ZB_`(-.0BM<2;4C`C/D?UL<#Z7!H#> N@]T
M[X'3:/.H=R*O9XK81\4.>XS&W3OT9'=H&!HG-TRN)9WZB;LAH-'O:[8KDR6I
M5CIWAN911\P&9D,+[6R<*=I,LP->Y<2]X5VS/=#D$8)M=\4X)EYP'F94U<HK
MN"$SZD,CB6+F08?'682<2=RD8<;Y"G0,M"V"PN2!WA at .3%V N9!2:^J&`6KJ
MD!1!;22"+E at Y*;#<%05EL(KGS+_\>3<XC at G9=O\FEDN.I6'H]2ZH!Q6D^Q*L
M(0F&DO at 7YQ^P!D:$:\B\F$/QF).X1&:3WZ3]9 _VVPN9.I-CV\'.]I:]9!D<
M)W#I"0DG5>0N7>](W 6 /SH%G\5876 at 4ESB5UV<\CI'.6S at T+.WD@H21!P6)
M_3V"<DXNJ/7+A>9,L^C''!GW![V/FN(SE\T8*)TF*/?7H-Q^!N57"DKO"I0V
M2+(\:ONVFS at 8/V+>;V3#ACKJAVP66E[:RZ+1G>V.:$'<Q-1M=Z^OLF&WJ=\,
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M\,+AO9N/XE".E9R<H% H2IRH!^5LWL!:+LP2O#YKZJUQJVWH-1R7`-2RM,W7
M:O#A_7N^5!"E8;'+Q*E$.3W5J_*KQK%.Y63WQ1_:I\;.$SJ5N'4<%_*<3I4J
M^/?W*;3W5*="$NHN3I&<:F?<L:>Q_'=;FB-ZFB?Z6AO[FO--^]KX'^AKX__[
MV at _H:RO\)MO?V$36/*6UB2PZW-Q$LJ8:\5")&I$@+C4[41J_7ZLKJL6>9I>C
M_T'MKJCPLOU[8<D?'+V:50OL at _U^(3:;IFF",M]#+DBV@^!Y_,OE\_C+W-$D
M?!;ZI_-OHG\J?]%?N[VQ_OE.'W3K!A:>OBG>>2ZEC$+_W#;OS+/\+<0A0?H2
MT.XVC&&S>%A(U]<=MW at HN+="_C"C\N<=7O^GF(X[/"(>2H%1@#OY[0Y?O_[%
M>].3G at ..?&6KD+K51CMD27B8OQC50,D&(OH5-$!>WS1+CT@'16\(/@)"W'"^
M(Q)QZ=(N(;L\;2%[XI7UX$WPVUL@!)_T9/FF^-U0G/1B$?!F;SAHZ!K>4CII
M*8\V"[N]>:O:4_55Q)P)J6U>HEZ(<K!1!G+2+-<VD/P)4$L#!!0````(`+:(
MX2;^[:04S 4``(\9```*````<'ET:&]N+FUA:]U8;6_:2!#^7"3^P]2-JB!=
M[(1<HQXJ4L&8B!8,PHG2GB(A8R^P%^.U_)(FNMY_O[%W;6SS5IJFU1V?UKLS
ML\_,SLXS["L84,MG`9N%T"'WQ&$>\<$((YLRN"0N\<V0V* /6A\UZ%*'_ 9M
M,\ 9YH+W&"Z8*]N!5ZV\['5!.CI6NY<U"9I-D*1J!3^:7 9.X(:ZYW7<8AK-
M47J@&4;K4 at .=@<7<&9U'N ]%P< C%IU18LLH.S,C)Z3N'$(&FPS):$G3.[AU
MM5*"\!(AE%3&Q"$(78+7KW?*)::E',B>>V\ZU"XAS4RL(.=T/K,(+#/UYQ',
MDO:7!7'!CUPW=B^);DYY^@@VF=%D+5P06)IX1("[Q5&/)RRV7)JN#0YUB0Q=
MY at -Y,)>>0QHY*_DP)\>GS%)?Y:5Y)\4&FWN=SPU'+ CHU,'M%XQ:)( 9;EQT
MR_2W0=AZ&L?3-)\DOG+\\/:B!BIS`X9[M3S/H59B7:KML,>!?Z^U.)'&X^$8
M6B[0C<=-@^)!K^7=T!"9C_O:[$LPT:\PCOIUO]^,Q?M&O$ORZ49.P8 Z&C4M
M1R8/I%H9&VK3M_B8FRY>JRU1C(6'UU>=WK at IWR*\GB[&(5EZ)SZ7JE9>09O,
MJ0MJ%(1L"8,XKP+4C,(.];GF*] PL4H"F9-C3;T>&YH`<RHE^%O]/C22$"0(
M:K<BR=")!%CJ?";)U\_>K"?#+B/E@!>QG,58U+[6TJ&Q=8-D78(5)*&0,__B
MY#W6O(#$._ H9E#8]"]IE\ $L<H^";8(W5L7IS*UI[N6O;7EC>'8(< M5"NK
M%2GV#_!'9^"R$$L%#<*<IG)T'*<ERBWO;.KG5E)#`N)60V)]@Z%44Z3Y9#0>
M?F at J+G/8G($RZ(!R<P[*Y2=0_J2@#.N@]$"2Y=N>:SF1C0DAOD<J#CMXMWKZ
M>9T/]8[6OK[DXXDZU(UA7Q-?@[9JX+#KK9^A9RUPY3-NV&6YU=M8W"Y/(#P+
MDCLI@#MP^O#[Z1]82=F:Z>ST0;%7Z*J5MJ&>UYO3P,*:2_C%3J8FW7[KTE@%
M at ZV?)2IE%B9&>VPTX18/OUKI]_2/:!3+_QVWR">$R3OBN\0YK\L.G4(4$%\,
MYS85HR_4#3S&G.0#R<1VYF+%M.]-+Q4+%L1)[6#U7(W,*$S-1]3FL_;42 at 5P
M:'GB(_,OB*;!8X#UJ&&)8JQ0U_+)DKBAZ31<%,'<;6S.:(QI7-]!02I<(.DU
M>N=O+S!F4;A!(;XCH.#FGADN&E)67+(P#=L?1"@W7_+-2ZL#%LO%;<_>Q/Z6
M[EX!4KY&2G!TW-&ZDVZOK]5PG$-6XY<LG:O!^W?OXJE,*#GE=:582E2V?;21
MLOPNTK!YJ_93*6-B\TAMO at K?QB6B$?@NVT^AF&3?YR48.9CZOYR!>+2V"X at X
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MLH[(AUQB9.#V_L&'KU^_L:,[Z,U at QU_Q!B1AM= /61(1CI^1:J#P@<A^!1V0
M5SUP^65ID^F"X1T at 1#?UC$A$@]<\`]ZHE9 =V$QO;49_O =9-=T3R7Q7^FPH
M]D8Q2WAC>#U6M29V#P-F1PX)THIO'9RZ^7ZNP!L-=(OO4RLV;B]$Q2A4BE14
M7,?##GP;ABTMRQ.A;8CB2"WPX@^)XHIB=\'%IA$SY=L;1IR<WHZ2A9,S^8U<
MG]9Y&YQ[+/GAA_$\KA1ZWSVN3;:YEN> U<>_4$L#!!0````(`+:(X2:YX_&P
MJ0T``+UU```,````<'ET:&]N,34N;6%K[5QK;QLW%OU<`_X/K!(4-K"1'&=;
MM$($U+'EK%I9,B0;21<&A-$,)4\]+\S#L;K=_[YWWB0O1=H)G>UBE0_!6.>2
MA^\AS[V<%^3"M>,P"5<I.:/WU LC&I-YFCEN2-[3@,962ATRN3CY=4C.78_^
MC;RS$O at E#$BT26_#X/7W72>)]O>^'9V3SLN#T_/WAQTR&)!.9W\/_AC45N05
M^> &;XZ!9IFMP?YB.)^?O!^224CL,%BYZPRX7, at WB:CMKESJ=,%V965>Z at 9K
MDH9$GE47\AI.SH!^?T\HQK=0#)1H1CT*5>B0[[[36!;9=YBBCH)[RW,=H;Q-
M)FW!F32_A1FQK;I6&V()J3_=TH#$61#DE2S:F4F\W!"'KMP"2V\I\2WH+ )L
M>?OG/]BA[UN!0SPWH%UR'L:$/EA^Y-$^DPO;V$5']E9M;;N^==?)LQP\H@&8
MQ\LP2=RE!T6X#5V;)F0%Y'S5K'A;,12]<K"LQU>G1 X>?OSAD)QM`LMW[5=C
M-[@C8W<96_&F<ZC,LBS]9V>8#ZO9;#HC)P%QI1WO)GR7HU$XG5=S`8B=\%.R
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MA=,X#D)MZX%5&"<*@Q6TG+8L*R^T]"5>P4*5>MHB at 5E*DU1O%UO^(XH6AW_0
M0&60!?HAM*:I%:]5[006R\SU'#=8A6JS:CBJ%H;"*MK$[OI64RH_=7U5(X%)
M&&GRB&":J9:7W,2S4GBY^FJK>QHG\%)2&4&?P59"62 P\:WXM<+$]:TU5:Z(
MKA^%L8JE-%#/H-+&42UD;J ?]IZ;I$'HJ'HI-]'G$P9KK9%ON:KVAX9-;I5+
M\Z/67-_YWE;#^BQ at 1E5=K;:Z#1U]K4,GR2)-EY=%TN>5W-NQ?O7Q-S&UG&+/
MJS*"74(:9JH&#^@G+9EF^&CKE)^H8(%6M70$(T.Y)A4&:7BG,K%C_<XI"A/W
M06^U@=> $L_)E 9)"N=&I46^-<R7@)@F2ALX&BGRB:U at K1]6,5U3?;4+JTC5
M#?%ZZ?IK?48A=+?6*H$-JJT?ZXF[#BS]BSOQX!"F;8 at DM._H(SC#+'#H@\JB
MV$CI^=(XC/1T:9Q!2U!_J9P$E9DVMTVBM4EO\P5$:Z#/!S8`>J/PC@;N'\JZ
MY><9NK1LU0Q/L^@12VBZB?1&]_8/1UW76:K@:"O\R0V*$:(@V&3WL%^'/8FL
M%%O.ATH3F)DZ$\\5"XQ,RCKM[[58)S^,$OCGKD at 0IH0^P': 2=M[>9 ?T,'.
MOW/<F$'JC*HJ;\VHPB49U2GSC,Y_.AHXJ_)8#X?_Q>5L^LN@%X1>N Y)[^*,
M]#Z\(;WW'TGOGR[I38]);T0ZW>[-*+"]S*&=^N_+4W@\(YW)V?#=]?OR^<-H
M\N:X?%S \]GTP[S\ZWH^7)R-%\./E]/9%?QT'C%=V#::?0O8;T!]'C+X39[ 
M$7^ @MJDT"R$*K!E\N]@E$)W'1]!G4(`KL?PZZ="KB at 4CBJM1XX>_G[T$^FM
M0E2P^MU!>J[0$$[+M+_W;G[ZYGBP3. \>T?+YBU^6IR/3][/V^*%LN$"R9H\
M%O-WL_F W, (V]\;CR:_0K:P`[DK\RQ_J#*]HW% O3?'^8 at D&;S9J\>UXU9/
M^0R*PM K_H SB>.M*\1R[JVH-DMNJ5?G$WJT?;*RM,X^<YWR5V=IUP;P:$<U
M5;[F5\\>G*,H;)OK,OA^\=0T0JZ4]3M'#Z_IZZ/\'S1FDBUA04VIW_]4:EC0
MOIX'C1[D^PT:I);7#R IS*S^MAD'27(I#OK=LF]AT]8?O?GQAYS5*>5=*$._
M`_]!8_?"+)7FDR\1P.I'A?&6V4]Z\']^HNIW&CVQZ9OINU^J_F/&$I:<1 M.
M;$(@(S,A3!"8$,Y(2R+&B4H8;.0D! E"$L)%"4DT:,4CA BR$<*%PPO"56D9
MD4B&\?(0MFB$(00QDA#".#%(1 49",&-`"0BHO0CP6O11X0$N0?!@M"#<%'B
MD1@(X at ZR$&0=C#>"#H)X*4>$61%'@O'RC<2 $VZD."O92 Q:L48"UC*-!&H%
M&AG(2C,2G!5E$,S*,3*P$6)$D)5@,-:<Q.70MG'+"2X(#%0#CQ-99* J+2^L
MB' CJ6"@%5,PIER)&@%%`JB2":*)!.?E$F3 "R42.%/.?R2.( ->%L$P*XB(
M*"^%('1;]RK*RPD?(LA('E*H%CL0R,L<"!8$#H0WT at 9&&E$#0ZV<@3%6I!!1
M49Y N"!,2/%(WDQ(C$ &@@PAXDB 0 :B]( ,!-$!X:+<@ U:H0%AHL0@,>#$
M!0G.RPI;#!0Y<%*"B#(B at AQ2I>6%`P1SD@%".;$`H8),@'!>(!!A[G O at L*Q
M7CX96E6//6 at +6WG64=PA+P_.AN>+\]%X> C/S(;]L#Q3U[\=DI_?OLU_:HR*
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MS, at H8:8KR0II,F)!9C-4X=-:H,.4K6YG>"A5LA]B9.5 PY2-FHA(>9W1;*<R
M,J7(*RJ8QKJSU#X17:N(FFU95E 529'8:G225E*MA+11<,WV)B,`BZ2B-FR8
MF)&6$;,H.YOM7DZU1MQ(TC9.SBCB$G91+S?<[(S<CLA%*=[HT*Z$?$S:ZON&
MJ]JZ!Q"IX#DP6M':[R"2<OX(LX.*<V=(> 5?A^GJ-IX2"37O13'/7/M at I-2<
M@\8T=^G>D? R?A_3G+G72,+8.).,O7$;-Y2$C?%/F:Y?X]V2T7*>+]/,M=],
M0LRYU,SRU at XYQ,IYZ at QQEJ?HQL\GXVP=@&87J-I_*')R?D6C35MY)3%AZPTS
M-EE:-V=!]\3/D.8<\P^CJ]-_#.>?_9&3ZK'@Z?P5OGDB:?7'[]8;P9UM&ZX/
M&I'WJ9?:MK7VTZ/@<)/_EX/BOEJ3-PKVEEGG8/F(]_6;W>BUH0*(-7B.\VBY
MCC8A""(I'YM at MJI,:(.,UGQE&^(V: (1"_$49M\<13B&2-G&:!A]9]01'IB.
MB?PP7;VM"N3S") -<1Z2 at BB;.!7C9-LJR0? &)VA;/@,XA5#:\Q.%RXR1\(M
MA.V8'<5MT ]B%N*!#%>ZZ,AME69!L]LO+DH)\:(0)K-CK V 0LQ";)393FXB
MJS M%W-E=B*W(5LBK1#,9;21I:_9YWK%;AF_S_>&:X+6$"47S6:T1:M8.)&1
M#9$SSY<'V$D9F\@[LPW+!.XA6B&FSS Q$Q*(F,5P0:/+0AELB#C;$$3#%2TC
M&#%?&]AHN'IE6"0F9,(EC;U7RL-8$5OV.>=ZH81-7OFE(E499_/30_*9=Y/@
M;[!95NWUZG7W^^[Q\K@\KS)?`A5J_>1#]%^P:HMM5=M^6"T9\T!:/)YJY!E>
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M6O$^SY<0<]O[_P!02P,$% ````@`MHCA)J =_J2R!0``0Q@```L```!P>71H
M;VYW+FUA:^U8ZV^;2!#_7$O^'Z:TJF+I`K9SK5I4I'-L'+GU2R96VE,DA&%M
M[P6SB$?BZ'K_^PVP8,"O6KGT'CH^+3NOW\S.S at R\@@$U/>:S>0`=<D]LYA(/
MM""T*(,KXA#/"(@%PT'KLPI=:I.?X-+P<8<YX#X&2^8\B);O5BLO>UT07I^U
MNU<U`10%!*%:P1>%,\$YW%#GHHE&9N$"V0>JIK6N5!@R,)DSIXL0+5'4ZKO$
MI'-*+!%YYT9H!]190,!@IR815:G##AJO5DH at 7B*(LLR$V 3A"_#FS6'&6+F0
MP]ES[@V;6B6PF8X-ZIS,5Q:"::0N/8)1DGY8$@>\T'$B#^,0YX1GCV"1.8UI
MP9+ RL!S`K06A3[:,-EJ93 at 6V-0A(G29!V1MK%R;R#DM^4C'9RC-,V?%E7$G
M1!J5X^[GEF/F^W1F(X EHR;Q88ZFBXX9WCX0^T_D;);FE9!0SM;OW]6 at Y;HV
M-6.U0NV0H at 3RJ6JB])E,1A-H.4!W'C'UBX>[E6TCC6<\VK/8 at Z\/KS%RPVF_
MKT3L?2VR$K\ZH5U0T!Z/%=,6R9I4*X/KOK*B%G^;:&W%,Y-U8JAXN?9%,>(>
M3:\[O8DBWB+:WI"O`[)RS[V$JUIY!9=D01UHAW[ 5C"(4LM'R3#H4"^1? 4J
MYE:)(?-YHK:G$TWE:.I"[$ZKWP<YCDB,H':;YAFZ$2-+ at Y&Q)@R-M]O9<%!+
M^02*:!H1FG9?;0U!WFLAI at NPP<0%<NI?G/^"U<\GD84DCBD6':&('O&%W4SW
MYKNZ2*W9(;*[EXPP!P9U1#;[K<2Q.R"'.!(KU<J&)$0N`CYT#@X+L&10/\B)
M2J_/HE1%OM6=1;T<)57$<>Y5Q.D[%*62//7U\63T29$<9K,%`VG0`>GF`J2K
M+R#]2D$:-4'J at 2"*MSW'M$,+DX*_C]NX[.!]ZPTOFLERV%$OIU?)6L?]SNA&
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MCQ^CK8PI/OQMH8B+E\FC;2@=&PXU(2L9`']L"]*M$YH0GR2.:'E*$XHM_ =:
M4!:0PSS4OCO*\^]J51;O52O>KWK8KZP3^I7^=_0K_?GZE?Y_OWIZOWK$;ZD]
M#8M?D)-:%K^<L4H\;R+[Q UR+8R7N6=O8+IUH(6E(']0$\N*MFC^GCGR1^2=
MG-0([&[C<:8V>8UO`^K\"*DBT73=I\FOUT^3STO[,^])Z$^7+Z(_53YKF<.1
MKGZY5B?#5A\+R%CC_V(:0L*A?NEIU]I9]KO"(F[\[=X;MOO3SN870$S8M-'L
MV_[&\**_)W+T#R:J^W.\=-M"/"5RN9'A._Z]#=^^?>]$=-I'_(%/8QGBX)KH
MBRCP.$>_=FH@)0M^!R3T0=R,D>6?/;M4%Q0?`,%'EV=$PJ<II0')5%1"=NHX
MNG?(^^M=R,KJD5#F9\!G0W$TC%G2:Z/II*TJ\;A2J.SFZ<F;'[ZVNH2,OB7&
M:L5)ZP7.T%H;G?S.L8."@)NSVW%,.&^(;\7FK)E,6[E/Z,P:O]HG9LX_PA=]
MGR^[CR]MVF8)?J&9'\(>5_%"]2X:+IC^$U!+`P04````" "WB.$FWJS4\L8%
M``"8%@``" ```'IL:6(N;6%K[5A[;]I($/^[2'R'J5M50;K8>;15SZJE$C"(
MUCR$@]*>D)"Q%]C#]EI^)*'M??>;M=?&0"#)Y=J[DRY*HGW,>V?FM^L7T*5V
MR"(VBZ%)KHG+`A*"&2<.9= F/@FMF#C0Z]8_Z="B+OD%+JP(5Y@/7UTZE9TH
MJ%:>=UH at O3QJM-HU"30-)*E:P8G&*> 8KJA_?H;BI\D<:;NZ:=;;.O08V,R?
MT7F".BC*BP)BTQDECHRT,RMQ8^K/(6:P*T9&.7JOB6JKE2WUSU']!L.0N 1-
MEN#5JP-4J5BI9%['O[9<ZFS96 A8&UOB^<(2L*W<DQ586]PW"^)#F/@^=RR-
M:8EYN@*'S&BZ%R\(>!8>#* V'FN^8#//LWP'7.H3&5HL!')K>8%+U)*4<H#3
M0U-FF:>R9RTE+DZ[Q_'2<,"BB$Y=5+U at U"81S%#IIDM6N$_]GE,XFN;Y(V4[
M1[?OWM:@N?(MC]K'!O678-!I:(4KJ;977&;U7Q;&TV<X[ ^A[@.]\Z!IM'G$
M.]G6-T6NHV*'W423WB5&L3<R#(V3&R;7DD[]Q-T0T!@,--N5R2VI5KJ7AN91
M1\R&9D,+[6R<*=HLJSLCRDG[H\MF9ZC)8S2UTQ/CF'C!<9A152LOX(+,J0^-
M)(J9!UV>71%R)G&3AAGG"] QO;8("H>'>F,T-'5ARHF4^E(W#%#3<*06U,9I
MJ at 4K)S4K#T-!%ZSB!?-/W^RFQ7X1VX'?M..4V]$P]'H/U+WBTWT)U at 8)AI+X
M9\<?L-=%A&O((E@;7]MO3V3J3*4#V\'>;>Z&QYS$)3*;_KY%M.$KN0T.;>/?
MH>U="^X,Y'I=XHX#_M 9^"S&+D*CN,2GO#SB>8MTWM*A86DG%R1<W"M([-\A
M*.?D at EJ_GFC.+,MVK(G)8-C_J"D^<]F<@=)M at G)U#DK[,RB_45#Z9Z!T0)+E
M<<>WW<3!G!'S0:,8XB]W^O3D-7:J:QO7FR#UFOK%J)V-KSJ]\[-L.,%QLW]E
MXJP5;)V;'- at +7/^"NENLM#?FQ,[V`EIJ0UK+6SZ4E7O+>!6 at Z+,3=(KAQLC 
MU9LT3:N5"[-Q?J9-(QN[-,E"DBY-6D:];:XELNWC118IYY^8%T-3 at S%F0[5B
M='J?4"2"Q3*3ERT(@4E$PO,SGEJP)*%/7#&9.U2,T+(H8,Q-)P@]CCL7.Y9S
M;04Y6;0@;L[,7+(>64DL)DE"G6S5F=HY`0[M0$S6!W=MX]$YKELD/A2>1\DT
M6D78T]2;K.&"@H2 at 4-\.B4?\V')5'^FP'-2[2 at 3).6;@05CV`A%4[9R_>XL!
M3>)=<BP9%.SQX]K9Y-4("OX/K'BA2D4#4["*61BK".!QVJ>+B/<O/HI3V=\=
M\NU<5=[)1/&7R[=D8[GS2O#RJ*FW)JV.H==P7%)>RZHT7ZO!A_?O^5)!E.;$
M+A.G$CWS,!3E=XA#0.1DU[^?"$,3YQ% )"X4AT0\!8A2\?]N&$)?#P,1$E!W
M>9C /8A5/)P'T:H4[_\:7K7?<,QR!&AY`K at Z"%S.PX +?^=?:9#&X?A4/I%?
M"YSZ)^%K\C]\/1"^RL<7Q?B(L,=IRRLM/![05OCFN at O11"$]'--$8>U#-5&Y
MJ2X\9Z)&)(A+*">ZXX_!N*)M[*!<;O5/PKFBP<OVM\*+/[CM:M8X$ `'@T)L
M-DUK!F6^AUR0; ?!T_AO;Y_&7^:.IN&3K'\\_Z;UC^4OX+77G^B?+_5AKVY@
M%QJ8XJO-J911Z)\[YJ5YE'W;<$B0ON\[O88Q:HJ/!>GJ&FR+Q_^5%?*/+"K_
M5,-A8(;EM\4A,J&4$H59][S&X?OW!UV4'O&X/_!J5B$-I8WVRY*(*O_J4P,E
M&XB,5]!T>7VI7'\*VB]Z0_ !(\2EY@=:(NY9VBED]Z4MRQYU/]U[\?O[[1>"
M[XUC^6KXPZRX-X9%HIO]T;"A:PAGW;1U1^4V;F]>J'8ZO(KV9@)JF[>G9Z+X
M-XH^)\WJ:\.*/P%02P$"% `4````" "VB.$F=N^Q=[ %```(%P``"0``````
M```!`" `MH$`````8G-D9&(N;6%K4$L!`A0`% ````@`MHCA)HB,^8?C!0``
MEQ@```P``````````0`@`+:!UP4``%]T:VEN=&5R+FUA:U!+`0(4`!0````(
M`+:(X28F[$L6J04``.06```*``````````$`( "V@>0+``!P87)S97(N;6%K
M4$L!`A0`% ````@`MHCA)O[MI!3,!0``CQD```H``````````0`@`+:!M1$`
M`'!Y=&AO;BYM86M02P$"% `4````" "VB.$FN>/QL*D-``"]=0``# ``````
M```!`" `MH&I%P``<'ET:&]N,34N;6%K4$L!`A0`% ````@`MHCA)J =_J2R
M!0``0Q@```L``````````0`@`+:!?"4``'!Y=&AO;G<N;6%K4$L!`A0`% ``
M``@`MXCA)MZLU/+&!0``F!8```@``````````0`@`+:!5RL``'IL:6(N;6%K
64$L%!@`````'``<`B@$``$,Q````````
`
end




From kajiyama at grad.sccs.chukyo-u.ac.jp  Wed Jul 28 21:37:04 1999
From: kajiyama at grad.sccs.chukyo-u.ac.jp (Tamito Kajiyama)
Date: 29 Jul 1999 01:37:04 GMT
Subject: Bit-twiddling
In-Reply-To: Lars Marius Garshol's message of 28 Jul 1999 23:03:06 +0200
References: <wkso68xzat.fsf@ifi.uio.no>
Message-ID: <KAJIYAMA.99Jul29103704@pseudo.grad.sccs.chukyo-u.ac.jp>

Lars Marius Garshol <larsga at ifi.uio.no> writes:
| 
| I'm translating a numerical algorithm from C to Python, but have run
| into a problem in that the algorithm seems to rely on the fact that
| shifting, addition and multiplication all 'wrap' when the result
| exceeds 32 bits.
| 
| Since the Python equivalents do not, does there exist a common
| workaround for this?
| 
| For example, I'd like
| 
|     mt[0]= seed & 0xffffffff
|     for mti in range(1,n):
|         mt[mti]=(69069 * mt[mti-1]) & 0xffffffff
| 
| to never put anything that doesn't fit in an unsigned long in the mt
| list.

I don't know general ways for wrapping bits, but how about this?

    mt[0] = seed & 0x100000000L
    for mti in range(1, n):
        mt[mti] = (69069L * mt[mti-1]) % 0x100000000L

--
KAJIYAMA, Tamito <kajiyama at grad.sccs.chukyo-u.ac.jp>



From nick.bower at ssec.wisc.edu  Fri Jul  9 15:50:21 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Fri, 09 Jul 1999 14:50:21 -0500
Subject: Grail won't work
Message-ID: <7m9tmt$6qg$38@rks1.urz.tu-dresden.de>

I just installed Grail 0.6 to take a look.  I have Py1.5 and Tk8 but it
won't run:

Traceback (innermost last):
  File "d:\grail-0.6\grail.py", line 499, in ?
    main()
  File "d:\grail-0.6\grail.py", line 150, in main
    from Browser import Browser
  File "D:\grail-0.6\Browser.py", line 26, in ?
    _mydir = os.path.dirname(grailutil.abspath(__file__))
  File "D:\grail-0.6\utils\grailutil.py", line 22, in abspath
    if not isabs(path):
NameError: isabs

What does this mean?  Cheers, Nick.




From fdrake at cnri.reston.va.us  Fri Jul 23 10:35:55 1999
From: fdrake at cnri.reston.va.us (Fred L. Drake)
Date: Fri, 23 Jul 1999 14:35:55 GMT
Subject: Maintainability (was Re: Stackless & String-processing)
In-Reply-To: <379798D9.FF3F811B@compaq.com>
References: <001101bed3f2$5e8874a0$642d2399@tim>
	<7n675p$qsk@dfw-ixnews7.ix.netcom.com>
	<14230.42741.34295.650213@153.chicago-05-10rs.il.dial-access.att.net>
	<379798D9.FF3F811B@compaq.com>
Message-ID: <14232.32203.792051.822513@weyr.cnri.reston.va.us>

Skip Montanaro wrote:
 > Some modules
 > in the standard distribution were designed specifically to support this
 > (types and Tkinter come to mind).

Greg Ewing writes:
 > In the case of Types, it could easily have been designed
 > differently. Writing Types.List would not be much harder than
 > typing Listtype and would be easier to read as well.

  I actually like the longer names on this one, but otherwise...

 > As for Tkinter, I've gotten into the habit of always
 > writing Tkinter.Button, etc. and I think the qualification

  Though I think some of my co-workers disagree, I agree with you on
this one.  Tkinter is a *terrible* module for the "import *" style.
James Henstridge's "gtk" module is much better about this, but enough
of the constants have short enough names that I won't use "import *".
  Another problem of "import *" is that it makes introspection of a
module (using dir(), idle, whatever) more painful.

 > So I wouldn't mind much if the * form of import
 > went away, either. You can easily write a function
 > which does the same thing if you really, really want it.

  Yes.
  Something that I'd like to see, that I don't think has been
mentioned for a long time, is some sort of "import ... as" statement.
I like qualifying the names from a module, but if I'm using them a lot 
I want to give the module a short name.  For instance,

	import Tkinter as Tk

would be more comfortable, and is much more clear than:

	import Tkinter
	_tk = Tkinter
	del Tkinter

  Variants could include:

	import package.module as pkgmod	# pkgmod is the actual 
					# module, not the package

	from package import module as pkgmod
				        # the same, possibly cleaner

	from module import function as func

  As I said, I haven't noticed this come up for a while, so maybe I'm
the only one, but I think this would encourage clarity and
"localizing" the names of imported modules (adding a leading _ so when
others "import *" from our modules their namespace isn't polluted too
much).


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From jheffero at nyx10.nyx.net  Tue Jul  6 10:16:27 1999
From: jheffero at nyx10.nyx.net (Jim Hefferon)
Date: Tue, 06 Jul 1999 14:16:27 GMT
Subject: Arbitrary precision calculations
References: <0000159e@bossar.com.pl>
Message-ID: <931270587.363932@iris.nyx.net>

Perhaps you need to specify a little more what you mean here.  Comparing
two arbitrary precision decimal expansions (this is what I thought you
meant, but maybe I'm mistaken) is not possible, in two senses.  First,
if you just simply want to compare them on a per-decimal place basis
then you will never be able to finish to conclude `equal', you can only
conclude `unequal' (I'll ignore the .999...=1 problem; see the sci.math
FAQ).  Second, if you want to conclude that they have equivalent 
algebraic descriptions (say, decide that 1+\sqrt{2}=\sqrt{2}+1) then 
you are again in hot water because this is an undecidable problem;
IIRC it is equivalent to the Halting Problem.

Jim Hefferon



From fredrik at pythonware.com  Mon Jul 26 19:56:51 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Jul 1999 23:56:51 GMT
Subject: Question: Python and API?
References: <7nirq3$ipu$1@oak.prod.itd.earthlink.net>
Message-ID: <017b01bed7c2$8ad6e900$f29b12c2@secret.pythonware.com>

Spud <spud200 at earthlink.net> wrote:
> I was wondering if Python (in Windows 9x) supports API functions. Would
> anyone here happen to know?

    http://starship.python.net/crew/mhammond/win32/
    (assuming "API functions" means functions from
    the Win32 Platform API)

</F>




From jmagave at amazon.com.br  Mon Jul  5 07:40:45 1999
From: jmagave at amazon.com.br (Jayr Magave)
Date: Mon, 5 Jul 1999 08:40:45 -0300
Subject: What's ZOPE anyway?
Message-ID: <3scg3.1267$B95.91660@typ42b.nn.bcandid.com>

I went through www.zope.org and read everything I could, and found the
following definition:

"Zope is an Open Source application server and portal toolkit used for
building high-performance, dynamic Web sites. "

... and the rest of the page goes on to describe ZOPE.

I am sorry, but what's ZOPE? :-)

Any code snipplet? Is this a language extension to Python?

Thanx,

Jair







From da at ski.org  Tue Jul 27 00:25:32 1999
From: da at ski.org (David Ascher)
Date: Mon, 26 Jul 1999 21:25:32 -0700 (Pacific Daylight Time)
Subject: Newbie question
In-Reply-To: <001501bed7d4$bc4723d0$f29b12c2@secret.pythonware.com>
Message-ID: <Pine.WNT.4.05.9907262122010.115-100000@david.ski.org>

On Tue, 27 Jul 1999, Fredrik Lundh wrote:

> > I'm just trying out some examples from the book Learning Python and I run
> > into a exercise that is not working for me.
> > It's on page 95 , exercise 4b.
> > This is the answer the book has
> > ===========================================
> > l = [1, 2, 4, 8, 16, 32, 64]
> > x = 5
> > i = 0
> > while i < len(l):
> >     if 2 ** x == l[i]:
> >         print 'at index', i
> >         break
> >         i = i + 1
> > else:
> >      print x , 'not found'
> > ============================================
> 
> either you or the LP authors have messed up;
> for the loop to ever terminate, the "i = i + 1"
> statement should at the "while" level, not in-
> side the "if" statement.

It's a very understandable problem -- there is a page break before the
line in question, and I don't blame Arnaldo for not having taken his ruler
and counting spaces.  Alas, there was little we could do in the production
process to avoid all the page breaks in Python code.  Sorry you wasted
some time on that one, Arnaldo.

python-code-wasn't-designed-for-typesetting-in-more-ways-than-one'ly y'rs,

  --david ascher




From phd at emerald.netskate.ru  Fri Jul  9 16:17:25 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Sat, 10 Jul 1999 00:17:25 +0400 (MSD)
Subject: Grail won't work
In-Reply-To: <3786527C.6230C4D2@ssec.wisc.edu>
Message-ID: <Pine.LNX.4.04.9907100016400.3944-100000@emerald.netskate.ru>

On Fri, 9 Jul 1999, Nick Bower wrote:
>     if not isabs(path):
> NameError: isabs
> 
> What does this mean?  Cheers, Nick.

   This:
     if not isabs(path):
   should be:
     if not os.path.isabs(path):

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From dl231 at eng.cam.ac.uk  Thu Jul 15 06:26:33 1999
From: dl231 at eng.cam.ac.uk (D. Lopez-De-Ipina)
Date: Thu, 15 Jul 1999 11:26:33 +0100
Subject: Problems using omniNames of omniORB2 from Fnorb Python ORB
Message-ID: <378DB759.DCE85A3E@eng.cam.ac.uk>

Hi,

First of all, apologies if you think this is not the forum to make
questions about Fnorb, but I have posted my question into the Fnorb
mailing list and I have not received any answer yet.

I have been trying to make an application that I have written using
Python with the CORBA ORB Fnorb, to register itself in the omniORB
omniNames Naming Service. As I experimented some problems and thinking
that perhaps there were problems in my code I just went to the simple
demo examples for the Naming Service usage provided in the Fnorb
Distribution at:

.. /Fnorb/examples/naming/(server.py | client.py)

I included in my .bash_profile the following environment variable
neccessary to make these example programs use the omniORB omniNames
Naming Service instead of the default Fnorb one:

export FNORB_NAMING_SERVICE=IOR:0184014... --> the IOR of the omniORB
Naming Service(omniNames)

It turned out that the examples seemed to work perfectly as it was the
case of my application, but if I left the server process running for a
while (1-2 minutes) after a while crashed and 'said' the following:

Traceback (innermost last):
  File "python/Fnorb/examples/naming/server.py", line 161, in ?
    sys.exit(main(sys.argv))
  File "python/Fnorb/examples/naming/server.py", line 153, in main
    boa._fnorb_mainloop()
  File "/home/dl231/python/Fnorb/orb/BOA.py", line 260, in
_fnorb_mainloop
    protocol.start(timeout)
  File "/home/dl231/python/Fnorb/orb/IIOPProtocol.py", line 74, in start

    self.__reactor.start_event_loop(timeout)
  File "/home/dl231/python/Fnorb/orb/SelectReactor.py", line 143, in
start_event
_loop
    self.do_one_event(timeout)
  File "/home/dl231/python/Fnorb/orb/SelectReactor.py", line 184, in
do_one_even
t
    handler.handle_event(Reactor.READ)
  File "/home/dl231/python/Fnorb/orb/GIOPClientWorkerReactive.py", line
229, in
handle_event
    self.__handler.recv()
  File "/home/dl231/python/Fnorb/orb/GIOPConnectionHandler.py", line
188, in rec
v
    self.__listener.message_received(OctetStream.GIOPMessage(data))
  File "/home/dl231/python/Fnorb/orb/GIOPClientWorkerReactive.py", line
277, in
message_received
    self._message_received(message)
  File "/home/dl231/python/Fnorb/orb/GIOPClientWorker.py", line 187, in
_message
_received
    self.close_connection()
  File "/home/dl231/python/Fnorb/orb/GIOPClientWorkerReactive.py", line
216, in
close_connection
    raise CORBA.TRANSIENT()
Fnorb.orb.CORBA.TRANSIENT: Minor: 0 Completed: COMPLETED_NO

Please if anyboy has managed to interoperate omniORB objects with Fnorb
ones in a safe way, could they please let me know and please explain
what is going wrong? Any workaround for it?

It might be a Fnorb bug.

Thanks,

Diego





From arcege at shore.net  Mon Jul  5 13:36:03 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Mon, 05 Jul 1999 17:36:03 GMT
Subject: Expect 1.8.2 - Expect extension for Python
Message-ID: <726g3.861$6M6.283561@news.shore.net>

I've created a patch release for ExpectPy which fixes some Linux
bugs and build problems.

I have also moved the ExpectPy webpage to Starship Python:
  http://starship.python.net/crew/arcege/ExpectPy/
The old URL points to the new location.  The author's email address
is arcege at shore.net.

Note:  Expect does not support Tcl 8.1 (and is expected to support
Tcl 8.2 instead); the regular expression engine and interface in
Tcl 8.1 changed making Expect and ExpectPy unusable with that
release.  This may cause problems when compiling ExpectPy with
earlier versions of Tcl, and compiling Tkinter with Tcl/Tk 8.1.

  -Arcege

Changes since 1.8:
* Linux has different mechanisms for extended CBAUD and for when a
  terminal detects EOF (the error EIO is returned).
* Add helper macros to the code for better readability.
* Make the C code more compilable (K&R C).
* Fix various build issues.
* Add a new option to the configure script to force using static
  libraries.
* Beta releases of Python include a "bN" in the version string,
  strip this inside the configure script.
* Handle searching for the pthread library.
* Make some minor changes for using Tcl 8.1 in a later release.
* Change the ask.py example to use the new stty method instead of
  termios (which I find is not compiled on many systems).
* Fix a bug where setattr of the settings object could receive a NULL.




From nick.bower at ssec.wisc.edu  Thu Jul  8 19:57:34 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Thu, 08 Jul 1999 18:57:34 -0500
Subject: Python for the PalmPilot and Other Devices
References: <37852547.E5AC00F4@ssec.wisc.edu> <3785374C.A5B7D34B@rubic.com>
Message-ID: <37853AEE.C6FD93D8@ssec.wisc.edu>

> A Python port for the CE devices is available, and it's
> not (hardly) scaled down at all.  Further, there's active
> development going on -- including Mark Hammond's participation.
>
> well-you-did-say-"and Other Devices"-ly yr's
>
> -Jeff

Probably something most are already aware of is that the Palm's occupy
over 70% PIM market share I think.  I can't remember the exact figure.
What were the reasons for developing for CE?  Easier because the Windows
interpretter already existed?  (I'm guessing here, not being
knowledgable when it comes to anything other than using the languages).

nick





From fredrik at pythonware.com  Wed Jul  7 08:11:06 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 7 Jul 1999 12:11:06 GMT
Subject: returning Python Lists from C++
References: <tKGg3.149$FH1.16131@newsr2.u-net.net>
Message-ID: <000b01bec871$cb812250$f29b12c2@secret.pythonware.com>

Mark Butterworth <mark.butterworth at cis.co.uk> wrote:
> Can anyone give me a quick example of how to return a python list object
> from a C++ function?

the python sources are full of such examples.  take
a look at array_tolist in Modules/arraymodule.c, for
example.

</F>




From mal at lemburg.com  Tue Jul  6 11:08:14 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 6 Jul 1999 15:08:14 GMT
Subject: mxODBC and MySQL access to remote DB
References: <Pine.LNX.4.05.9907042015320.15753-100000@ns.ac-copy.net>
Message-ID: <37821BDE.61CD46D6@lemburg.com>

Joachim Schmitz wrote:
> 
> Hello,
> 
> I?m trying to connect from a linux-box with mxODBC to a MySQL db on a
> remote SUN. I can access the db with mysql-client from linux. I can also
> connect with the odbctool from the iodbc-mysql package.
> 
> When I try it with mxODBC.iODBC, which uses the same "iodbc.ini" I always
> get this:
> 
> line: 38
> -- in file: Form.py; function: __init__; line: 40
> arg #0 = DSN=immobilien;UID=perry;PWD=pwpwpwpw
> Traceback (innermost last):
>   File "immoform.py", line 153, in ?
>     ,DBtype="MySQL")
>   File "Form.py", line 45, in __init__
>     ,"pw="+self.pw)^M
> mxODBC.OperationalError: ('S1000', 1045, "[TCX][MyODBC]Access denied for
> user: 'user=perry at localhost' (Using password: YES)", 4237)

This a problem with MySQL, not mxODBC. Please check the MySQL
logs for more detailed information about access failures.
 
> The iodbc.ini entry looks like this:
> 
> [immobilien]
> Trace    = Off
> TraceFile= stderr
> Driver   = /usr/lib/mysql/libmyodbc_mysql.so
> DSN      = immobilien
> SERVER   = test.ac-copy.net
> UID      =
> PWD      =
> PORT     = 3306
> #OPTION  = 1
> #DB      = test
> #SOCKET  = /tmp/mysql.sock

Note that I have a line stating the database in my odbc.ini
file:

[mysql]
DRIVER   = /usr/local/lib/libmyodbc.so
SERVER   = localhost
DATABASE = test 

Also note that iODBC switched from using a file named 'iodbc.ini'
to 'odbc.ini' recently.

> with:
> 
> odbctool -u perry -p pwpwpwpw -d immobilien "show tables"
> <table border=1>
> <th>Tables in immobilien</th>
> <tr>
> <td>IMMOAUFTRAG</td>
> </tr>
> <tr>
> <td>IMMOKUNDE</td>
> </tr>
> </table>

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   178 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From another_better_way at hotmail.com  Fri Jul  9 19:21:33 1999
From: another_better_way at hotmail.com (another_better_way at hotmail.com)
Date: Sat,10 Jul 1999 19:21:33+2000
Subject: ACAD for Linux ...
Message-ID: <107111999143315sforge@mcs.net>

   Look at this web page to learn more about ACAD for Linux !!!


           http://209.218.86.64/linux.html


   If you know someone who can be interested in that please tell him

or her about this web page.


   Thank you for your time.





From chega_ at yahoo.com  Wed Jul 28 17:57:55 1999
From: chega_ at yahoo.com (Vadim Chugunov)
Date: Wed, 28 Jul 1999 14:57:55 -0700
Subject: 1.5.2 for: else:
References: <1279007788-11957995@hypernet.com> <slrn7psisn.da.wtanksle@dolphin.openprojects.net> <000c01bed893$2ce8f7d0$f29b12c2@secret.pythonware.com> <slrn7pum9t.2ro.wtanksle@dolphin.openprojects.net>
Message-ID: <7nnuda$u47@news.dns.microsoft.com>

I would say that else: clause after a loop is really related to the if:
controlling the break,
rather than to the loop itself.

When I first saw Python syntax for loops I said: "A-ha!  So in Python I will not
need a goto
in situations like this:
-----
for(Item* pitem=pseq->First(); pitem; pitem=pitem->Next())
    if (pitem->key==42)
        goto Found;

pitem = pseq->Insert(new Item());
Found:
    // use item pointed to by pitem
-----
In fact, I do not see any other good use for else: clause in a loop.
Maybe the docs should just explicitly say what this feature is good for ?

Vadim

William Tanksley <wtanksle at dolphin.openprojects.net> wrote in message
news:slrn7pum9t.2ro.wtanksle at dolphin.openprojects.net...
> On Wed, 28 Jul 1999 00:50:14 GMT, Fredrik Lundh wrote:
> >William Tanksley <wtanksle at dolphin.openprojects.net> wrote:
> >> The very worst part of the current else: behavior is that it changes the
> >> meaning of else.  In other constructs, else is an alternate path to take
> >> if the data being tested fails a single expression test.  In this
> >> contruct, else is a path taken if the code block belonging to the previous
> >> test executes a certain instruction.
>
> >nope.  you've got it all backwards.  consider this:
>
> >    for an "if" statement, "else" is a path taken if the
> >    expression evaluates to false.
>
> Close enough -- I would say it's the path taken if the 'if' block isn't
> executed.
>
> >    for a "while" statement, "else" is a path taken if the
> >    expression evaluates to false.  or in other words,
> >    when the loop terminates by natural causes.
>
> In other words, it executes when you'd expect an 'else' to not execute --
> 'else' doesn't mean "natural causes".
>
> >    for a "for" statement, "else" is a path taken when
> >    there are no more elements to loop over.  or
> >    in other words, when the loop terminates by
> >    natural causes.
>
> In addition to the above carping, I have to add that the code following
> the for (or while) loop is what I'd expect to execute after the loop
> terminates.
>
> >not that complicated, was it?
>
> I didn't think it was either, but it seems that in spite of my initial
> understanding, I still managed to reverse it accidentally.
>
> >now, what you seem to have trouble with isn't
> >the "else"-clause -- it's the behaviour of "break":
>
> >"else" always works the same way.  and so does
> >"break".  what else did you expect from Guido?







From m.faassen at vet.uu.nl  Mon Jul 19 04:14:36 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 19 Jul 1999 08:14:36 GMT
Subject: I envy newbies
References: <378F2304.ACC1AF7E@callware.com> <000001bed00d$a9a70480$d2a22299@tim>
Message-ID: <7mumpc$15a0$2@news.accu.uu.nl>

Tim Peters <tim_one at email.msn.com> wrote:
> In 5 years of running 95 and NT,

I assume NT was around in '94, but Windows 95 (except in beta) wasn't. :)

But-I-understand-anyway-ly yours,

Martijn




From aahz at netcom.com  Sat Jul 10 23:12:50 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Jul 1999 03:12:50 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787D8DA.3774AA7D@ssec.wisc.edu> <7m8p7t$o7s@dfw-ixnews10.ix.netcom.com> <3787FB51.12FE0048@ssec.wisc.edu>
Message-ID: <7m923i$dem@dfw-ixnews5.ix.netcom.com>

First of all, please try to follow proper netiquette for posting.  Your
lines should be 72 characters or less, and you should put your comments
after the quoted material:

In article <3787FB51.12FE0048 at ssec.wisc.edu>,
Nick Bower  <nick.bower at ssec.wisc.edu> wrote:
>
>yeah, that's what i originally thought. sure enough there were some
>updates in the os documentation, but not os.chdir. it's something that
>would be really helpfull if anyone linked to the documentation sees
>this.

Ah, I see what the problem is.  I went to the index inside the library
reference to make sure os.chdir() was documented, but didn't check to
see *where* it was documented.  Sure enough, it's documented inside the
"Process Parameters" section of os (6.1.1), rather than in the "Files
and Directories" section.

Fred, here's the URL that needs to be fixed:
http://www.python.org/doc/current/lib/os-procinfo.html

I think I understand how it happened: os.getcwd() is (probably
rightfully) documented there, so someone thought os.chdir() belonged
there, too.  AFAICT, that's the only mistake in that area of the docs,
though I could make a case that os.getcwd() should be moved (or at least
have a pointer in "F&D").
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From wtanksle at dolphin.openprojects.net  Tue Jul 27 17:17:08 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Tue, 27 Jul 1999 21:17:08 GMT
Subject: 1.5.2 for: else:
References: <Pine.LNX.4.04.9907271612140.9849-100000@emerald.netskate.ru> <19990727142943.V4068@xs4all.nl>
Message-ID: <slrn7ps8ej.vm0.wtanksle@dolphin.openprojects.net>

On Tue, 27 Jul 1999 14:29:43 +0200, Thomas Wouters wrote:
>On Tue, Jul 27, 1999 at 04:15:44PM +0400, Oleg Broytmann wrote:

>>    The following program:
>> ----------
>> for a in ['a', 12]:
>>    print a
>> else:
>>    print "Empty!"
>> ----------

>http://www.python.org/doc/current/tut/node6.html#SECTION006200000000000000000

>4.4 break and continue Statements, and else Clauses on Loops 

>Loop statements may have an else clause; it is executed when the loop
>terminates through exhaustion of the list (with for) or when the condition
>becomes false (with while), but not when the loop is terminated by a break
>statement. 

Well, I'm suprised.  I had also expected the else clause to run on empty
loop.  It seems so natural, in line with the meaning of a loop (loop on
this; otherwise do that).

Fortunately I've never actually used it and thus been disenchanted.

Understanding the current semantics seems to require understanding the
nature of loop tests in Python.  Since I understand them I can agree that
this makes sense, but I'm still a bit leery, because you're not 'elsing'
on the data but rather on code inside the loop.

I would _definitely_ use exceptions to handle this.

>Thomas Wouters <thomas at xs4all.net>

-- 
-William "Billy" Tanksley



From dborgelt at xxxxxxcom  Wed Jul 14 12:09:36 1999
From: dborgelt at xxxxxxcom (Dave Borgelt)
Date: Wed, 14 Jul 1999 11:09:36 -0500
Subject: MySQLdb in compiled form
Message-ID: <#IyLFRhz#GA.362@nih2naad.prod2.compuserve.com>

I am trying to use Python with MySQL on a Win32 platform.  I have downloaded
MySQLdb ver 0.2 but have been unable to use it so far.  I don't have MSVC++,
which I take it is needed to compile the _mysqldb.c code.

Would it be possible (legal) for me to get a compiled file from someone and
just use that on my system?  MySQL is running fine so I don't think there
are any problems there.  Would it work to compile the code under Linux
(running on Intel) and then copy it over to the Win98 side?

Any help will be appreciated.

Dave Borgelt

dborgelt at xxxcsixxx.com

remove the x's for email address








From mal at lemburg.com  Thu Jul  8 10:58:57 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 08 Jul 1999 16:58:57 +0200
Subject: .pyo's without .pyc's?
References: <19990526140137.A66609@vislab.epa.gov> <374C6C6E.1D07326B@lemburg.com> <19990708074113.A61389@epa.gov>  
	            <37849A3F.28C0EF21@lemburg.com> <199907081309.JAA01063@eric.cnri.reston.va.us>
Message-ID: <3784BCB1.238B87E9@lemburg.com>

Guido van Rossum wrote:
> 
> > I've checked this and indeed you're right. exceptions.py is imported
> > prior to all other modules at a time where the switch from .pyc to
> > .pyo hasn't been done yet (this is done in _PyImport_Init(), yet
> > exceptions.py is imported by _PyBuiltin_Init_2()):
> [...]
> > Still, I think you're right in that the .pyc -> .pyo switch should
> > be done prior to importing *any* Python module, meaning that
> > _PyImport_Init() should be called prior to importing anything.
> 
> Hmm, seems you're right.  As Linus says, all bugs will eventually be
> found.  Seems like this would be the fix, right?  (_PyImport_Init() is
> pretty trivial so I don't doubt that it works.)
> 
> Index: pythonrun.c
> ===================================================================
> [patch]

Yep, that's what I had in mind too ;-)

Cheers,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                   176 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From gherman at my-deja.com  Tue Jul  6 06:40:48 1999
From: gherman at my-deja.com (Dinu C. Gherman)
Date: Tue, 06 Jul 1999 10:40:48 GMT
Subject: TopSort in Python?
References: <7lfjn0$b1v$1@nnrp1.deja.com>
Message-ID: <7lsmff$b3d$1@nnrp1.deja.com>

Thanks for the two versions! They seem both appropriate,
although MCF's has the strange feature of adding (1,) here
or there. TP's has a great Exception class, it seems...

Ok, last week-end I typed in the algorithm sitting on my
own shelf at home (just to add one more). I haven't made any
timings, but I like that this one below is pretty short and
easy to understand. The top-level function parameters are
a bit different, but I wrote a wrapper function for using
the "more regular" one, too...

Anyway,-not-really-that-important-now'ly,

Dinu



"""A simple and short topological sorting algorithm in Python.

Input

The input is a set of unique elements (vertices), V, and a di-
rected graph of edges between them, E, where V is a list of
any Python objects and E a list of (v, w) tuples, with v and
w representing indices in V.

Output

The output is L = TS(V, E), a sorted list of all elements in
V, given E. Elements x in V without any (v, w) in E for with
v == x or w == x are prepended to L. Moreover, for all v, w
in V, v != w: (v, w) in E => index(L, v) < index(L, w).

Algorithm

In the directed graph we progressively pick one leaf node,
remove all adges to it and add it at the head of the (ini-
tially empty) result list. If there is no more such leaf
node, we have either a cycle (if there are still edges/
nodes left) or we're done. Isolated nodes (with neither in-
nor outcoming edges) are inserted at the head of the result
list.

Notes

Much of the algorithm can be changed to adapt to different
needs (while still calling it topsort), like the order of
isolated nodes, throwing a real exception when a cycle is
detected, a more efficient implementation of allEdgesFrom(),
or a randomized order of elements where multiple solutions
do exist, the omission of the vertices list parameter in the
top-level function (meaning using edges (v, v) in E), the
use of a class to remove some of the parameter passing, etc.

Dinu Gherman,
July 1999
"""


import sys


def allEdgesFrom(v0, E):
    "Return a list of all edges (v0, w) in E starting at v0."

    resEdges = []
    for v, w in E:
        if v0 == v:
            resEdges.append((v, w))
    return resEdges


def _topSort(v, E, visited, sorted, sortedIndices):
    "Recursive topsort function."

    # print "trying", v, visited, sorted, sortedIndices
    visited[v] = 1
    for v, w in allEdgesFrom(v, E):
        # print "edge found", (v, w)
        if not visited[w]:
            _topSort(w, E, visited, sorted, sortedIndices)
        else:
            if not sorted[w]:
                print "No sorting possible!"
                sys.exit(0)
    sorted[v] = 1
    sortedIndices.insert(0, v)


def topSort(V, E):
    "Top-level sorting function."

    n = len(V)
    visited = [0] * n   # Flags for (un-)visited elements.
    sorted = [0] * n    # Flags for (un-)sorted elements.
    sortedIndices = []  # The list of sorted element indices.

    for v in range(n):
        if not visited[v]:
            _topSort(v, E, visited, sorted, sortedIndices)

    # Build and return a list of elements from the sort indices.
    sortedElements = []
    for i in sortedIndices:
        sortedElements.append(V[i])
    return sortedElements


def wrap(pairs):
    """Convert an element pairs list into (verticesList, edgeIndexList).

    This might be useful for those who prefer topsort(edgeList)
    instead of topsort(verticesList, edgeIndexList)...
    E.g. wrap( [('a','b'), ('b','c'), ('c','a')] )
         -> (['a','b','c'], [(0,1), (1,2), (2,0)])
    """

    e = []
    v = []

    # Make a list of unique vertices.
    for x, y in pairs:
        if x not in v:
            v.append(x)
        if y not in v:
            v.append(y)

    # Convert original element pairs into index pairs.
    for x, y in pairs:
        e.append((v.index(x), v.index(y)))

    return v, e


def testCase(V, E):
    print V
    print E
    print topSort(V, E)
    print


def test():
    # Test wrapper.
    # E = [('a','b'), ('b','c'), ('c','a')]
    # print E
    # print wrap(E)
    # print

    # No error to be expected.
    V = ['a', 'b', 'c', 'd', 'e', 'f']
    E = [(0,1), (1,2), (1,3), (3,4), (4,5), (3,2)]
    testCase(V, E)

    # Still no error, but an isolated element 'g' should appear.
    V = ['a', 'b', 'c', 'd', 'e', 'f', 'g']
    E = [(0,1), (1,2), (1,3), (3,4), (4,5), (3,2)]
    testCase(V, E)

    # Error because of cycle in graph.
    V = ['a', 'b', 'c']
    E = [(0,1), (1,2), (2,0)]
    testCase(V, E)


if __name__ == '__main__':
    test()


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From kevinsl at yahoo.com  Sun Jul 11 18:24:30 1999
From: kevinsl at yahoo.com (kevinsl)
Date: Sun, 11 Jul 1999 14:24:30 -0800
Subject: problem with smtplib.sendmail
Message-ID: <931731872.16311@www.remarq.com>

I'm trying to mail a blank message via a cgi script. (it
triggers a subscribe confirmation for a mailing list) It's
giving an "Error transmitting message data" exception.

Here is the code:

fromaddr='test at test.com'
toaddr='list-subscribe at test.com'
msg=''
mailserver=smtplib.SMTP('localhost')
mailserver.sendmail(fromaddr,toaddr,msg) #error occurs here
mailserver.quit()

localhost is running qmail 1.03.

any ideas?

thanks,

Kevin



**** Posted from RemarQ - http://www.remarq.com - Discussions Start Here (tm) ****



From s341625 at student.uq.edu.au  Sun Jul 18 19:33:28 1999
From: s341625 at student.uq.edu.au (Anthony Pfrunder)
Date: 18 Jul 1999 23:33:28 GMT
Subject: Python Bug:os.spawnvc
Message-ID: <7mto88$r8e$1@bunyip.cc.uq.edu.au>

Hi,

I think I have found a bug in python but it could be a Windows bug.  With
os.spawnvc in this form:

	error_code = os.spawnvc(P_WAIT, proggie, args, environment)

where,
	proggie is a program to run with args in the environment.

If args is a list (as it must be) then spawnvc chops off the first
element of the list.  You can see this by running:

	ec = os.spawnvc(P_WAIT, 'test.bat', ['a', 'b', 'c'], [])

test.bat:
	@echo %0 %1 %2 %3 %4 %5

Which results in:

	C>test.bat b c

instead of

	C>test.bat a b c

Could someone confirm that this is a bug or that I have stuffed up?

Cheers,

Anthony Pfrunder

PS Please cc to my email address thx.



From nascheme at ucalgary.ca  Sat Jul 17 18:39:29 1999
From: nascheme at ucalgary.ca (Neil Schemenauer)
Date: 17 Jul 1999 22:39:29 GMT
Subject: Python and Boehm-Demers GC, I have code.
References: <000401bed097$e02f7c00$27a22299@tim>
Message-ID: <7mr0n1$dmk@ds2.acs.ucalgary.ca>

Tim Peters <tim_one at email.msn.com> wrote:
>pybench and pystone have small working sets (they're primarily CPU-speed
>tests); so there's very little reachable for BDW to trace, so BDW doesn't
>have much to do, so runs fast when it's invoked; and with refcounting on,
>most pystone/pybench needs are handled by internal free lists that don't
>invoke malloc at all after initial setup, so BDW doesn't even get called
>often.  IOW, BDW likely isn't doing much of anything in these tests besides
>some initial mallocs.

What could I use as a better benchmark?  Real applications would
be best I suppose.

>I don't think that's going to be happen:  Python doesn't ask anyone to
>change anything about how they like to handle their memory today, and Guido
>has pronounced that a Major Feature on multiple occasions.  Vladimir
>Marangozov's PyMalloc (see his Starship page) caters to that, letting Python
>use its own malloc without anyone else getting involved.

Perhaps I misunderstand, but I thought that the PyMem_* functions
were created for this purpose.  Extension modules can use any
malloc they like but they can't mix malloc/free/realloc with
PyMem_*.

>On one particular platform, and that platform's version of threads (or are
>you running without threads?), yes -- looks great so far <wink>.

Yes, one platform, no threads.  Thanks a lot Tim.  Here I was
feeling pretty good about myself.  Now I'm depressed.

>> With some tuning maybe the gc version of Python will perform
>> better than the regular version.
>
>With or without refcounting?  As before, I think refcounting's contribution
>to Python's current performance is vastly under-appreciated.

With.

>[about Tkinter callbacks]
>> This is really nasty.  I am trying to allocate this data with
>> regular malloc.  If I use malloc and free then this data should
>> not get collected and the GC should realize that the Python
>> functions still have references to them.
>
>This won't work.  A pointer *must* be reachable from *Python's* "root set",
>else BDW won't know it exists.

<homer>Mmmm, documentation</homer>.  I think I have the problem
licked now.  I use GC_malloc_uncollectable for these things.

>btw-keep-in-mind-that-software-doesn't-work<wink>-ly y'rs  - tim

No kidding.  I tried to compile this stuff under Windows.  What a
nightmare.  I'm going to have to take some time off and lick my
wounds.  :)


    Neil



From nick.bower at ssec.wisc.edu  Mon Jul 19 17:04:39 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Mon, 19 Jul 1999 16:04:39 -0500
Subject: Scientific Plotting?
Message-ID: <379392E7.EB0F29E7@ssec.wisc.edu>

People at my current work use MatLab or IDL for chomping data through
data calculations and plotting the results in an environment that lets
you get GUI's up and going very quickly (among other things).  It's been
a goal of mine that one day I could show them that this could be done
with as little effort (with greater room for expansion) using Python.  
Plus I'd like to put a strong case forward to our physics department for
our up-and-coming new scientific programmers to learn something with a
broader scope like Python than the now used IDL and recently abandoned
Fortran-77.

But...What do people use for plotting?  After looking at the Python
Topic Guides, I've taken a quick look at the BLT and DISLIN pages, but
both of these amazingly lack any screen shots to demonstrate their
capabilities!!!

And lastly, can array operations (using NumPy) be done without for loops
as in IDL/MatLab?  eg Multiplying 2 arrays is easy in these two packages
because you don't need array subscripting and iteration.

Thanks muchly, Nick.



From neelk at brick.cswv.com  Wed Jul 21 19:26:25 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 21 Jul 1999 18:26:25 -0500
Subject: how do I find memory leaks?
References: <7n4skv$8do@thalamus.wustl.edu>
Message-ID: <7n5kv1$e6$1@brick.cswv.com>

In article <7n4skv$8do at thalamus.wustl.edu>,
Heather A. Drury <heather at thalamus.wustl.edu> wrote:
>
>The program has gotten to a fairly advanced stage but I've run
>into what appears to be a major memory leak. I come
>from the C programming world and have previously used
>Purify to debug memory leaks. Alas, I thought using python/vtk
>would take care of all this memory garbage for me (although
>I'm probably not doing something I should be). 
>
>Over time as the user continues to use the application, the
>computer starts using up swap space and eventually the program 
>crashes.  I haven't the faintest idea how to begin to debug this
>problem. I'm sure I need to provide some more detail but the
>code is about 2000 lines now and I'm not sure what detail
>would be useful.
>
>Any hints/suggestions/pointers would be greatly appreciated.

This sounds like a problem an awful lot like I had with NumPy
a few weeks ago. Basically, slices of arrays count as references
to it, which leads to some surprising memory wastage if you
don't know what's happening. 

For example:

    import Numeric
    
    def waste_memory():
	x = Numeric.zeros([1000,1000])
	return x[0,0:2]

Each call to waste_memory() will generate a million-element array, 
and return an array of two elements. However, the slice that's
returned will cause the *whole* 1000x1000 array to be kept in 
memory. 

You can fix this by wrapping the return in another call to 
Numeric.array(), which creates a new array object and lets
the old one go out of scope properly.

E.g.:

    def no_waste():
	x = Numeric.zeros([1000,1000])
	return Numeric.array( x[0,0:2] )

Hope this helps.


Neel



From fredrik at pythonware.com  Thu Jul 22 10:33:00 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Jul 1999 14:33:00 GMT
Subject: Q: anyone running Python on OS X server?
Message-ID: <004401bed44f$1a5c6400$f29b12c2@secret.pythonware.com>

just curious...

</F>




From aaron_watters at my-deja.com  Fri Jul 30 09:09:21 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Fri, 30 Jul 1999 13:09:21 GMT
Subject: ANN: Gadfly bugfix test release
Message-ID: <7ns85r$cla$1@nnrp1.deja.com>

Gadfly users and others: Please try one of the following
files

   http://www.chordate.com/kwParsing/kwPtest.tgz (tar/gz version)

or

   http://www.chordate.com/kwParsing/kwPtest.zip (zip version)

I believe this replacement fixes the following longstanding bugs:

   -- multiline strings should be supported by sql parser.
   -- table creates should not require a commit before inserts work.
   -- unique indices should raise exception on non-unique insert.
   -- table create followed by a recreate shouldn't require a commit to
work.
   -- an obscure recovery error which is hard to describe
   -- a server bug having to do with not handling poorly formed
requests correctly.

Also, the *.zip file has the correct version of gfclient.py unlike the
current one.
Also http://www.chordate.com/gfcontrib.html has the correct gfplus.py
link now.

Pleasy anyone try these replacements.  The test suite

   http://www.chordate.com/kwParsing/indexbug.py

demos that some of these bugs have been fixed (it will fail in several
ways with the current version).

If I don't hear complaints I will make this gadfly 1.01 in a week or so.
The changes are mainly very small ones.  The big ones aren't ready for
prime time yet.
   -- Aaron Watters

===

Kramer: Do you yearn?
George: Do I yearn?
Kramer: Yes, do you yearn?
George: No.  I crave.  I crave constantly, but I never yearn.
Kramer: Well, I yearn.



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From Aahz.Maruch at p98.f112.n480.z2.fidonet.org  Sat Jul  3 13:38:24 1999
From: Aahz.Maruch at p98.f112.n480.z2.fidonet.org (Aahz Maruch)
Date: Sat, 03 Jul 1999 18:38:24 +0100
Subject: Upgrade or not ??
Message-ID: <000016ac@bossar.com.pl>

From: aahz at netcom.com (Aahz Maruch)

In article <377E0A3B.CA3B258C at worldnet.fr>,
pierrette  <pserrand at worldnet.fr> wrote:
>
>First a general question :
>Actually we're working with Python 1.5.1 and we're satisfied with it.
>Should we upgrade to 1.5.2 and if yes WHY?

I think so.  We're sure planning on upgrading to 1.5.2 as soon as we can
spare the cycles.  There are a bunch of bug fixes across everything
(e.g. one recent example we uncovered was whitespace handling in
sgmllib) plus tons of new features (e.g. time.strptime()).
--
                      --- Aahz (@netcom.com)

Hugs and backrubs -- I break Rule 6       <*>      http://www.rahul.net/aahz/
Androgynous poly kinky vanilla queer het



From Cory.C..Albrecht at p98.f112.n480.z2.fidonet.org  Fri Jul  2 15:23:58 1999
From: Cory.C..Albrecht at p98.f112.n480.z2.fidonet.org (Cory C. Albrecht)
Date: Fri, 02 Jul 1999 20:23:58 +0100
Subject: Compiling Python for Win32
Message-ID: <0000168d@bossar.com.pl>

From: coryalbrecht at hotmail.com (Cory C. Albrecht)

In article <ZPQe3.4335$U6.23535 at newsr2.twcny.rr.com>, "Darrell" <news at dorb.com>
wrote:
>> What I would like is if somebody who has successfully built Python 1.5.2
>> using MSVC++ could do the "Export Makefile" command and send me the
..
>> will earn somebody my undying thanks. :-)

>Ok. I'm waiting for the undying thanks part.

And here you go. *undying thanks* :-)

>Hope this works for you.

Me too. Thanks alot, I appreciate it.

--
Cory C. Albrecht
http://www.sentex.ca/~cory/



From mwh21 at cam.ac.uk  Sat Jul 10 19:17:32 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 11 Jul 1999 00:17:32 +0100
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu>
Message-ID: <m3hfnc3xir.fsf@atrus.jesus.cam.ac.uk>

Nick Bower <nick.bower at ssec.wisc.edu> writes:

> I feel really dumb asking this question, but how do I change the current
> working directory?  For browsing the file system I mean.
> 
> I've found exec commands, but these would surely be device dependent and
> there doesn't seem to be anything obvious in the os or os.path
> documentation.  i looked through the tutorials too.
> 
> thanks in advance, nick

os.chdir?

as in

Python 1.5.2 (#1, May 11 1999, 11:42:00)  [GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> os.chdir('/')
>>> os.getcwd()
'/'
>>> print os.chdir.__doc__
chdir(path) -> None
Change the current working directory to the specified path.
>>> os.chdir('~')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
OSError: [Errno 2] No such file or directory: '~'
>>> os.chdir(os.path.expand'~')
os.path.expanduser  os.path.expandvars
>>> os.chdir(os.path.expanduser('~'))
>>> os.getcwd()
'/home/mwh21'
>>> 

HTH
Michael



From jbauer at rubic.com  Thu Jul  8 21:55:16 1999
From: jbauer at rubic.com (Jeff Bauer)
Date: Fri, 9 Jul 1999 01:55:16 GMT
Subject: Python for the PalmPilot and Other Devices
References: <37852547.E5AC00F4@ssec.wisc.edu> <3785374C.A5B7D34B@rubic.com> <37853AEE.C6FD93D8@ssec.wisc.edu>
Message-ID: <37855684.EC30DBA0@rubic.com>

Nick Bower wrote:
> Probably something most are already aware of is that the 
> Palm's occupy over 70% PIM market share I think.  I can't
> remember the exact figure.  What were the reasons for 
> developing for CE?  Easier because the Windows interpretter 
> already existed?

You'd have to ask Brian Lloyd the original reasons for porting
Python to his pathetic little 2MB CE device. ;-)  Now that CE's
got 16-32MB direct from the factory, well, grins all around.

In regards to the PIM market, the race does not always go the 
swift or to Microsoft <0.7 wink> but that's where the smart 
money bets. (*)

Seriously, I'd like for there to be a Python port to Palm Pilot
and even Symbian <http://www.symbian.com> but nobody's stepped
up to the plate.  BTW, the Python/Pilot port question appears
to always originate from a _current_ Pilot owner.  No surprise 
there, but a slightly different emphasis from: "I want to run 
Python on a PDA.  What devices should I consider?"

The CE devices have some advantages.  There is a wide diversity 
in models (palm sized, handhelds, even laptop sized units) and
vendors (at least a dozen or so).  The Win32 development heritage
is reasonably well-understood, if often cursed.  The CE platform 
includes a lot of built-in features (support for TCP/IP, sockets, 
IrDa).  Microsoft is using its clout to move CE into automobiles 
and cable set-tops.  True, CE represents a capitulation to a well
known software vendor, but ...

as-a-former-Newton-owner-you-haven't-seen-proprietary-until-
you-deal-with-PDA-hardware-vendors-ly yr's,

Jeff Bauer
Rubicon, Inc.


(*) Where can I find the _dumb_ money?  The smart money keeps
eluding me.



From donn at u.washington.edu  Mon Jul 12 13:25:02 1999
From: donn at u.washington.edu (Donn Cave)
Date: 12 Jul 1999 17:25:02 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
References: <3787D36A.815B7C29@ssec.wisc.edu> 	<3787D8DA.3774AA7D@ssec.wisc.edu> 	<7m8p7t$o7s@dfw-ixnews10.ix.netcom.com> 	<3787FB51.12FE0048@ssec.wisc.edu> 	<7m923i$dem@dfw-ixnews5.ix.netcom.com> 	<14217.63931.75257.566266@weyr.cnri.reston.va.us> 	<7md2ip$19q2$1@nntp6.u.washington.edu> <14218.7724.879301.831590@weyr.cnri.reston.va.us>
Message-ID: <7md8de$fpo$1@nntp6.u.washington.edu>

"Fred L. Drake" <fdrake at cnri.reston.va.us> writes:
|  I am vindicated!
|  Does this mean you'll make a contribution to my "Don't Brake for
| Dutchmen" campaign?  ;)

Hm, I'd think twice about that if I were you - isn't "Guido" an Italian name?

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From tim_one at email.msn.com  Wed Jul 21 02:54:06 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 21 Jul 1999 06:54:06 GMT
Subject: How fast can we multiply?
In-Reply-To: <37944AC1.F913550D@appliedbiometrics.com>
References: <37944AC1.F913550D@appliedbiometrics.com>
Message-ID: <001401bed345$d3934200$a72d2399@tim>

[Tim]
>> Andrew Kuchling's wrapping gmp for Python is the most useful thing
>> that's ever been done in this area.

[Christian Tismer]
> I see. The gmp makes very much sense. Until now.
>
> On the other hand, if the hard coded "C" basic elements are
> made fine and tiny, they give us nuts and bolts to build
> the real algorithms on top.
>
> We can therefore do what gmp does, not so much slower, but
> readable, in Python. No porting, no nothing, just using.

I play with & investigate bigint algorithms in Python, & I enjoy it.  But
serious work in this area requires serious speed, and GMP doesn't compromise
on that -- from slaved-over assembler to low-level mutable packed storage
limbs, it's doing the right stuff for its task.  Factors of 5 count here
<wink>.

You do "long1 & long2" in Python, and it's handled by an all-purpose bitop
function that redispatches on the operation in each iteration of the inner
loop.  It's "like that" everywhere:  the bigint code was written for
compactness (longobject.c is only twice the size of intobject.c!), and
no-hassle porting to the feeblest machine on earth.  The compromises have
consequences (some good!  that code has been rock solid since the first
release, while it can take a week to track down all the bugfix patches to
the last official GMP release).

> And again I see a win for Python. It is able to show how
> algorithms work, in a very clear way using its clean
> syntax, and you can use the implementation of math directly
> as highschool text. *And* it just works, and is fast.
> Fells very very promising to me.

For educational purposes, sure -- provided you can find enough people
seeking this kind of education <wink>.

[matmult]
>> Large dense matmult isn't very common.  Sing along:

> It is just so easy. Factoring matrix multiplication
> into its submatrices and then save (I think it was one)
> multiplications - it would be a dlight to write in
> Python. Not doing so appears to me as to know
> quicksort, but using bubble sort all the time,
> not just at break-even (10-20).
> But I'm telling nothing new to you.

OTOH, Winograd's refinement of the matmult procedure requires 15 floating
adds, where straight 2x2 matmult requires only 4; float + isn't much cheaper
than float * on many modern processors.  So the win here isn't nearly as
pure as for intmult, and the reduction is only from O(n**3) to O(n**2.81)
operations anyway.  There are reasons few vendors offer this form of matmult
in their libraries <0.5 wink>.

> ...
> As another little challenge:
>
> Tim already mentioned Jurjen N.E. Bos' high precision math
> library. It is in Python's contrib site.
> I have the strong impression that it was not used by anybody
> since a year now. python.org fails also to advertize it.
>
> It could need a little rework, to make use of 1.5.2's features.
> And if someone marries it with a fast multiplication, that
> could make a neat tool.

I use it often; I'm not sure how practical it is <wink>, but it's great fun!
It would benefit a *lot* from a constant-time nbits.  The clearest quick win
is to get rid of all the code that's special-casing for "big" left shift
counts; a patch to remove that restriction went into Python shortly after
Jurjen vanished.

> If I were a student, I would propose my advisor to let me do such
> things as my semester homework. What a delight :-)

It would make a great project for any number-head.  OTOH, Jurjen's approach
was both highly clever and highly idiosyncratic, and the internals of the
code are difficult to understand.  Some of the examples in the docstring no
longer work.  And some of the algorithms are clearly flawed; e.g.,

>>> a = r(1)
>>> b = r(1)
>>> a.mant = 8; a.expon = 0
>>> b.mant = 3; b.expon = 0
>>> a
8.+-1
>>> b
3.+-1
>>> a+b
8.+-4
>>>

That is, the result interval is too small -- a+b *can* be as large as 13.  A
similar thing happens whenever the before-rounding internal mantissa of a
sum is congruent to 3 mod 4.

A good semester project would be to come up with-- and prove
correct --algorithms for {+ - * / sqrt exp log input output} for numbers in
this format (a very compact flavor of variable-precision floating-point
intervals).

unary-minus-is-easy<wink>-ly y'rs  - tim





From jbooth at ux4.cso.uiuc.edu  Mon Jul 19 21:50:31 1999
From: jbooth at ux4.cso.uiuc.edu (jonathan allen booth)
Date: 20 Jul 1999 01:50:31 GMT
Subject: Killing a thread
References: <7mu6a5$h61$1@nnrp1.deja.com>
Message-ID: <7n0kl7$crh$1@vixen.cso.uiuc.edu>

>From what I looked at trying to figure out my [similar] problem... What
about imposing resource limits on the child thread (can you do that by
thread?), then catching the signal that gets thrown when it exceeds max
CPU time, and then using that to exit the thread?

catlee at my-deja.com wrote:
: I'm writing a progrma where some user code is executed in a restricted
: environment running in a separate thread.  Now, if the user is
: malicious, s/he can just put 'while 1: 1+1'.  I'd like to be able to
: recover from this.  Is there any way I can kill the thread that is stuck
: in the infinite loop?

-- 
Jonathan A. Booth				kamikaze at N0$PAMimsa.edu
http://www.N0$PAMimsa.edu/~kamikaze		jbooth at N0$PAMuiuc.edu
Remove NOSPAM from hostnames. Unsolicited commercial e-mail: access denied



From jam at newimage.com  Thu Jul 15 17:53:20 1999
From: jam at newimage.com (jam)
Date: Thu, 15 Jul 1999 17:53:20 -0400
Subject: I envy newbies
In-Reply-To: <378E3481.EE31C97@uswest.net>; from mary jeanette on Thu, Jul 15, 1999 at 02:20:33PM -0500
References: <378E3481.EE31C97@uswest.net>
Message-ID: <19990715175320.B20155@toast.internal>

On Thu, Jul 15, 1999 at 02:20:33PM -0500, mary jeanette wrote:
> 
> I decided to learn python because I heard it was a very easy language to
> learn. So I downloaded it and bought a couple of very expensive
> paperback books from o'reilly only to discover that they are written for
> unix. In other words, typing in the examples from the book into the
> interpretor only yields various syntax errors. I have no idea what I am
> supposed to change in order to make the examples work, and apparently
> there is nowhere to turn in order to find out. Q: Am I supposed to know
> DOS as a prerequisite? The idea of having to learn DOS is appalling.
> Should I try partitioning my hard drive and installing Linux or freeBSD?
> Perhaps I should learn Perl instead, I have seen books on Perl written
> for Win32 systems.   Thanks for your advice.
> 

your post appears to be begging for an argument, so I shall refuse to give
you one. 

pick the language that best suits your purposes, and look before you leap.
if perl or visual basic does that for you, then go for it. those of us that
'get it' are just fine over here, thank you.

the term 'syntax error' is commonly used to describe typographical errors,
and indicates something that *should* be independent of operating system.
complaining that you spent a lot of money on books and then noting 'syntax
errors' with the examples in them doesn't help anyone help you learn the
language or how to program, in my humble opinion. 

if you truly feel you paid too much for the python books, go back to the
retailer you purchased them from and ask *why* the cost is so high-- visit
sites on the internet and shop for discounted prices if a traditional 'brick
and mortar' establishment isn't giving you good customer service.

no, you should not need to know DOS as a prerequisite to learning python
(though knowing multiple operating environments could not be considered a
detriment). it was developed in a unix environment, and a win32 port has
been established, along with a version of the interpreter implemented in
java, and I believe there is a DOS specific port as well.

I have not (yet) purchased the o'reilly books you mention, but I am fairly
certain the author will make every effort to own up to any mistakes in the
code or the text that might lead to syntax errors-- did you check the
publishers web-site for updates?

as a rule of thumb: if your post can't pass the "you asshole" test, don't
bother hitting the send button lest you humiliate yourself in front of
a huge audience-- here's a hint: the post you sent out *did not*, in my
opinion, pass the 'you asshole' test. the python community (as with others
on the internet today) is driven (mostly) by mutual respect for one another
and *not* greed. you simply need to ask for help instead of demanding it,
and you will be given the answers you seek.

check <http://www.python.org/> for additional information.

best of luck.

regards,
J
-- 
|| visit gfd <http://quark.newimage.com/> 
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From Fernando.Mato.Mira at p98.f112.n480.z2.fidonet.org  Thu Jul  1 14:20:24 1999
From: Fernando.Mato.Mira at p98.f112.n480.z2.fidonet.org (Fernando Mato Mira)
Date: Thu, 01 Jul 1999 19:20:24 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001611@bossar.com.pl>

From: Fernando Mato Mira <matomira at iname.com>



Paul Prescod wrote:

> I find it fascinating that there are so many GUI toolkits. The GUI is more
> than twenty years right and we haven't agreed on the basic programming
> model or API yet??

Most [OO] GUI toolkits are yet more of the same primitive stuff. From the
glimpse I got from Qt I think it
goes in the right direction, but I got an idea it's just reinventing stuff in
a way which has already been superseded (something like GARNET's formulas).
The better one is to use multi-way constraints. There're toolkits based on
that (for Eiffel (unreleased) and Smalltalk). What still needs to be thought
is how to merge that with the CLIM presentation model.




From Jeff.Epler at p98.f112.n480.z2.fidonet.org  Thu Jul  1 11:26:52 1999
From: Jeff.Epler at p98.f112.n480.z2.fidonet.org (Jeff Epler)
Date: Thu, 01 Jul 1999 16:26:52 +0100
Subject: Why must path be in args in os.execv?
Message-ID: <0000161b@bossar.com.pl>

From: jepler at inetnebr.com (Jeff Epler)

Holger,

On Unix, read the manpage for execve and friends.  You will see that the
first argument is the path to the executable, and the subsequent arguments
describe the arguments (including argv[0]) and optionally the environment.
By convention, argv[0] is the basename of the executable when the
executable is on the PATH, and the full path otherwise.  (Another unix
convention is that when the first letter of argv[0] is "-", the shell is a
login shell)

DOS doesn't really have execve.  It's possible that the imitation is less
perfect than it could be.  Instead, it may do the equivalent of entering
the string
	string.join(argv, " ")
in the shell, i.e., ignoring the path argument.

This looks like a shortcoming in the DOS (windows 9x, NT) implementation of
exec*, in a way that makes it work differently from the Unix it attempts to
imitate.  This wouldn't surprise me, since I recently discovered that NT's
spawnl takes a unix-like array of arguments (const char *argv[]) but when
any element contains a space, it is split into two arguments in the child
process.  So, I had to break some code that had worked fine on Unix to
generate argv, by quoting each argument (something that would *not* work on
Unix) ..

Bad enough that Windows wants to copy Unix, they have to consistently get
it subtly wrong too.

Jeff



From tim_one at email.msn.com  Wed Jul 21 00:57:34 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 21 Jul 1999 00:57:34 -0400
Subject: IDLE's PyShell
In-Reply-To: <3794C6C7.9A212A97@inrialpes.fr>
Message-ID: <000801bed335$8c233e80$a72d2399@tim>

[Vladimir Marangozov]
> Just noticed that IDLE's interactive shell eats an unusually big
> amount of memory. Just pressing the RETURN key a couple of times,
> or even better ;-) -- closing and reopening the PyShell window
> does not seem to reuse the prevously allocated memory...

Christian (Tismer) ruined my life several weeks ago by noting that if he
typed

    sys.getrefcount(None)

in the IDLE shell, it kept going up & up & up as windows were opened and
closed.  That's a pretty good sign of A Leak.  The CVS version has plugged
many leaks, and I just submitted a brief patch to plug 3 more, but it's
still leaking.  At least *closing* a window no longer boosts None's
refcount!  Opening one does, though.

> I'm very far from this code, so don't blame me for not proposing a
> patch, but here are two things I experimented with that seem to help
> partially:
>
> 1. Clear the linecache when closing the PyShell window

Submit a patch!

> 2. Finalize properly the undo machinery in EditorWindow
>    (yes -- the XXX sign is there to remind that the _close
>     method needs more work regarding undo)
>    For instance, I tried to finalize undo before self.io.close
>    which seemed to work.

Not sure that one's a problem anymore; the undo machinery doesn't show up on
the list of live objects at the end for me anymore.  The Tk Text instance
still does, though, and that's a biggie.

just-stop-opening-and-closing-the-shell<wink>-ly y'rs  - tim





From flight at mathi.uni-heidelberg.de  Thu Jul  8 16:13:33 1999
From: flight at mathi.uni-heidelberg.de (Gregor Hoffleit)
Date: Thu, 8 Jul 1999 22:13:33 +0200
Subject: PyApache and Apache 1.3.6 anyone ???
Message-ID: <19990708221333.B16851@hoffleit.org>

Is anybody successfully running PyApache on an Apache 1.3.6 ?
Preferably on a Linux system ? When I build PyApache and load it from
httpd.conf, /usr/sbin/apache just doesn't start up. It doesn't give
any error or what, it just starts up and quits immediately.

Any experiences ?

Thanks in advance,

	Gregor



From garethr at cre.canon.co.uk  Thu Jul  1 12:59:51 1999
From: garethr at cre.canon.co.uk (Gareth Rees)
Date: Thu, 1 Jul 1999 16:59:51 GMT
Subject: Why is tcl broken?
References: <1283099377-77267484@hypernet.com> 	<lw67471fyi.fsf@copernico.parades.rm.cnr.it> 	<3779ED0F.2C67@mailserver.hursley.i 	<lwoghyexy5.fsf@copernico.parades.rm.cnr.it> 	<7ldt5p$srh$1@dragon.flightlab.com> 	<lwhfnosvuz.fsf@copernico.parades.rm.cnr.it> 	<377B922A.4A7B@mailserver.hursley.ibm.com> <86n1xgwbxr.fsf@g.pet.cam.ac.uk>
Message-ID: <sipv2cuwzc.fsf@cre.canon.co.uk>

Gareth McCaughan <Gareth.McCaughan at pobox.com> wrote:
> Kent Pitman wrote an article about why FEXPRs aren't good enough, back
> around the time when the transition to macros happened. It's available
> on the web somewhere.

http://world.std.com/~pitman/Papers/Special-Forms.html

-- 
Gareth Rees



From mcfletch at vrtelecom.com  Thu Jul 29 22:07:21 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Fri, 30 Jul 1999 02:07:21 GMT
Subject: Need help with TCP/IP client access from Windows
In-Reply-To: <slrn7q1ej3.18d.esr@snark.thyrsus.com>
References: <slrn7q1ej3.18d.esr@snark.thyrsus.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIIOECECAAA.mcfletch@vrtelecom.com>

About the only way I know to do this at all reliably...

import os
os.system( 'start %s'%url )

Which does whatever the user has defined as the proper thing to do with
urls, normally, launch IE or Netscape.

Hope this helps,
Mike

-----Original Message-----
From: python-list-request at cwi.nl [mailto:python-list-request at cwi.nl]On
Behalf Of Eric S. Raymond
Sent: July 29, 1999 4:33 PM
To: python-list at cwi.nl
Subject: Need help with TCP/IP client access from Windows
...
(3) Is there a reliable way for Python programs under Windows to point a
local
    browser at a specified URL?
...




From ivanlan at callware.com  Fri Jul 23 11:03:56 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 23 Jul 1999 15:03:56 GMT
Subject: [PSA MEMBERS] Python Conferences
References: <199907230917.0631499.9@summit-research.com>
Message-ID: <3798845C.BC16931A@callware.com>

Hi All--

Jeff Rush wrote:
> 

[snip]

> I'm a self-employed consultant and always pay my own way, and
> the conference fees I see run from $300 - $800.  I consider those
> a real bargain, because the airfare, hotel and lost billing time are
> the much larger portions of attending, for me.  The -location- of
> the conference, and duration, is more important, therefore, than
> the attendance fees.
> 
> I don't see how your wife could attend ten conferences for one
> Python conference -- ten plane tickets and hotel nights has got
> to cost more than one Python con.
> 

I'm sorry, I didn't make that clear, did I?  I was *only* considering
the conference fees--her fees seem to be in the $40 range, and I
estimated one Python con at $400, which is less than I paid last year at
IPC7.  But of course 10 travel/lodging expenditures costs more than
one;-)
 
> And for the international among us (not me), travel expenses
> must be much higher still.
> 

Well, I agree that even the high fees for Python conferences *are* worth
it.  I had a blast at IPC7, and I can afford those fees.  I would not be
able to attend Monterey, though, if I weren't doing a talk.  And that
makes me wonder how many others are in the same boat as I am. ...  And
as far as international conferences go, I would not be able to attend
conferences outside continental North America, on a cost basis alone.

I'm sorry to be such a grump; it's just that I think if fees were lower
it might attract interesting contributors and fresh young minds to
Python cons, and that would be good for Python.  But maybe you're right,
the flight/hotel may well outweigh the fees.

> -squinting-at-the-bigger-picture-ly-yours

<how-can-we-squint-when-we-can-fly>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From hyoon at bigfoot.com  Wed Jul 14 13:39:22 1999
From: hyoon at bigfoot.com (Hoon Yoon)
Date: Wed, 14 Jul 1999 13:39:22 -0400
Subject: Case mismatch on NT
Message-ID: <378CCB4A.BDAD6FC7@bigfoot.com>

Hi,

 I realize that this has been discussed before. And you might think the
solution is  os.environ["PYTHONCASEOK"] = "hotdog" but try this one.
 There is dialog.py sitting in pythonwin/.../mfc directory and Dialog.py
in lib-tk directory. If I have both directory in my Pythonpath, I cannot
call the one sitting at the second directory with out encountering the
dreaded "Case Mismatch" error.
 I feel that "Case Mismatch" should only occur after all directories has
been searched to find the 'right' case import. Not after finding the
first non-matching one. Is there a way to set environ to do what I want?
 If not, can anyone tell me how to get around this prob or tell me if
what I am proposing makes sense or not?

-- 
*****************************************************************************
S. Hoon Yoon (Quant) Merrill Lynch Equity Trading, 
yelled at yahoo.com hoon at bigfoot.com(w)
"Miracle is always only few standard deviations away, but so is
catastrophe."
* Expressed opinions are often my own, but NOT my employer's.
"I feel like a fugitive from the law of averages."    Mauldin
*****************************************************************************
-------------- next part --------------
A non-text attachment was scrubbed...
Name: vcard.vcf
Type: text/x-vcard
Size: 202 bytes
Desc: Card for Hoon Yoon
URL: <http://mail.python.org/pipermail/python-list/attachments/19990714/7187485e/attachment.vcf>

From wware-nospam at world.std.com  Fri Jul  9 22:42:27 1999
From: wware-nospam at world.std.com (Will Ware)
Date: Sat, 10 Jul 1999 02:42:27 GMT
Subject: Python Code Repository?
References: <378645AA.5430A026@msfc.nasa.gov> <19990709163127.A1964@toast.internal>
Message-ID: <7m9tnb$6qg$54@rks1.urz.tu-dresden.de>

jam (jam at newimage.com) wrote:
: I don't think anything as nice as CPAN exists for python ...
: I know enough about python and cgi scripting and sql databases to come
: up with something pretty quick, and all we'd need is submissions. there
: might already be some examples of code snippets out on the net in various
: places..
: does anyone have any further suggestions or ideas on how we can get this
: rolling?

There is alot of merit to the idea, and probably if you implement it,
getting contributions won't be a problem. It might be worthwhile to
set up a voting system that allows people to rate contributions for
interesting-ness. You might look at the slashdot.org moderation scheme,
that guy has thought long and hard about the problem of rating content
in a way that is both fair and informative.

There might be a few different ways people could search for something.
The first could be the Yahoo-style hierarchy, where you'd see something
like this:
AI -> neural nets -> back propogation -> you get the picture...
The second would be searching on keywords, which would be found either
in the source code itself or an associated text description or readme
file. I think freshmeat.net implements something like this.

What would be an interesting opportunity for innovation would be to
allow people to add new keywords after the fact. Suppose I'm looking at
some piece of code that integrates numerical solutions to differential
equations, and I realize that it has some bearing on some non-obviously
related thing like a problem in cellular automata. As a user of the
system, it would be good if I could add "cellular automata" as a search
keyword to the numerical integration code. Just a thought, but you might
end up finding some interesting connections that wouldn't otherwise
emerge.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com



From tim_one at email.msn.com  Thu Jul 15 04:09:57 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 15 Jul 1999 04:09:57 -0400
Subject: Getters and Setters
In-Reply-To: <7mjrqg$e4o@ds2.acs.ucalgary.ca>
Message-ID: <000701bece99$6dbeede0$51a22299@tim>

[Neil Schemenauer]
> Recently I found myself wishing that Python had automatic getters
> and setters like CLOS.

Bleech.

> For those who are not familiar, CLOS (and Dylan) create a "getter"
> and "setter" method for each attribute of a class. This saves the
> programmer from writing "getX" and "setX" for every attribute when
> no special action is needed.  Later, an explict method can be created
> if needed.

Python goes one better:  you can write an explicit method if needed, but it
doesn't encourage bad practice by automatically supplying methods that
expose every data attribute, crying "use me! abuse me!" <wink>.

> An implementation of this idea is attached.  With this
> implementation, these automatic methods seem to be about 8 times
> slower than explicit methods.

And some later corrected code boosted the factor to 10.

> Maybe Tim and friends can do something about that.  I would appreciate
> any comments.

I'll attach code that gets to the same end, but where the automatic methods
actually run faster than explicit ones.  I won't say why, though <wink --
but you'll figure it out>.

General comments:

+ Don't think OO so much as functional for this kind of thing.  OO just adds
overhead to what should be a very straightforward task.  Put on your Lisp
hat here!

+ time.clock is preferred for benchmarking.

+ Don't count the time it takes to generate range(n).

+ Don't count the time it takes to start a print.

> On a related note, has anyone considered adding hygenic macros to
> Python?

No.  People have *suggested* it, but those who have never showed any sign of
*doing* something about it <wink>.

> Due to the high cost of function calls in Python they would probably
> be a useful (and for other things too).  I have the feeling that hygenic
> macros are much easier to do in a language like Lisp.

No question about that one!  Python program fragments aren't sexps, they're
strings, and for starters there's no clear way to expose their internal
structure.

> Maybe Dylan can give us some ideas.

For one, that it's very much easier and clearer in a language with
relentlessly uniform syntax -- like Lisp <0.1 wink>.

all-that-glitters-is-not-sputum-ly y'rs  - tim


Oops!  Almost forgot the code!  Dump _Getter.  Dump _Setter.  Have the mixin
class build the obvious function for what you're after, and add that
directly to self as an attribute.  Then there's no mixin overhead at all
after the first time a set/get method is invoked for an instance.  Downside:
generates many function objects.  Possible alternative:  plug synthesized
get/set functions into the class object instead; a little slower, but less
proliferation of function objects.

class GetterSetter:
    def __getattr__(self, name):
        if name[:3] == "get":
            def getter(d=self.__dict__, key=name[3:]):
                return d[key]
            setattr(self, name, getter)
            return getter

        elif name[:3] == "set":
            def setter(value, d=self.__dict__, key=name[3:]):
                d[key] = value
            setattr(self, name, setter)
            return setter

        else:
            raise AttributeError(name)

def test():
    import time
    class A(GetterSetter):
        def __init__(self):
            self.X = self.Y = 1
        def getY(self):
            return self.Y
        def setY(self, v):
            self.Y = v
    a = A()
    n = 1000
    indices = range(n)
    print 'Comparing', n, 'iterations'
    start = time.clock()
    for x in indices:
        a.getX()
        a.setX(x)
    finish = time.clock()
    print 'GetterSetter', finish - start
    start = time.clock()
    for x in indices:
        a.getY()
        a.setY(x)
    finish = time.clock()
    print 'explicit function', finish - start

if __name__ == '__main__':
    test()





From Michael.Hudson at p98.f112.n480.z2.fidonet.org  Thu Jul  1 05:14:13 1999
From: Michael.Hudson at p98.f112.n480.z2.fidonet.org (Michael Hudson)
Date: Thu, 01 Jul 1999 10:14:13 +0100
Subject: strange behaviour at termination time
Message-ID: <000015fc@bossar.com.pl>

From: Michael Hudson <mwh21 at cam.ac.uk>

Ovidiu Predescu <ovidiu at cup.hp.com> writes:

> Greg Ewing wrote:
> 
> > To avoid problems like this, you have to design
> > your __del__ methods so that they can do their
> > work without having to refer to any module-level
> > names at the time they are called. One way is
> > to use the default-argument trick:
> > 
> > class B (A) :
> >     def __del__ (self, A = A) :
> >         A.__del__ (self)
> 
> Can you explain why this trick works? When are the default arguments
> stored in the method instance? I guess this happens when the method
> instance for __del__ is created by the interpreter, right?

Yup, that's it. If you want to find out about these things, read the
language reference or import rlcompleter and bash the completion key a
lot. I this case you'll find

B.__del__.im_func.func_defaults

which will contain something like

<class A at xxxx in xxx>

HTH
Michael
 
> Thanks,
> -- 
> Ovidiu Predescu <ovidiu at cup.hp.com>
> http://www.geocities.com/SiliconValley/Monitor/7464/



From tim_one at email.msn.com  Fri Jul 16 22:18:42 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 16 Jul 1999 22:18:42 -0400
Subject: Python and Boehm-Demers GC, I have code.
In-Reply-To: <7mo2iu$iqg@ds2.acs.ucalgary.ca>
Message-ID: <000901becffa$b0f25e00$91a22299@tim>

[Neil Schemenauer]
> I have added the Boehm-Demers garbage collector to Python and it
> works well!
> ...

That's good, but I'm not clear on what this means.  The patch didn't seem to
do more than replace malloc/calloc/realloc/free with the BDW versions; in
particular, Py_USE_GC_FREE was #define'd, and you ran tests without
significant cyclic structures, so it doesn't *appear* that anything more
happened here than that a different implementation of the malloc family got
plugged in.  Did the "collection" phase of BDW find any trash at all?

> I started with Sam Rushing's patch and modified it for Python
> 1.5.2c.  To my surprise, removing the reference counting
> (Py_INCREF, Py_DECREF) actually slowed down Python a lot (over
> two times for pystone).  I not exactly sure why this is.  Maybe
> someone can enlighten me.

My guess is that simple refcounting has excellent cache behavior in Python,
and GC doesn't.  Unlike as in almost any other language, the little loop

    for i in xrange(10000000):
        pass

creates a huge amount of heap trash at a furious pace (even ints "are boxed"
in Python).  RC can reuse the same little bit of storage over & over (the
trash is recycled as soon as it's created), while delayed GC chews up some
large amount of address space over & over instead.  You can test this by
timing that loop alone (with and without refcounting).

Note too that Python does no stack allocation of locals or frames or
anything else:  everything is heap-based, with the most speed-critical
allocations augmented by assorted internal fast free lists.  Disable the
refcounting, and the finalizers on the "intense" objects don't get called,
and then the fast free lists don't get used either.  So the allocation speed
optimizations Python implements get castrated too.

I've always said it would take a very sophisticated mark-&-sweep to beat
RC's performance for *Python* (it's not Lisp, or Smalltalk, or C++, or ...
under the covers); if true (& I haven't seen anything yet that changes my
mind <wink>), you're at the first step of very long road if you want to make
BDW the primary approach; but it would be great if BDW could be invoked
"just sometimes" to reclaim the cycles RC can't catch.

semi-encouraging-ly y'rs  - tim





From ivanlan at callware.com  Tue Jul 20 00:21:09 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 20 Jul 1999 04:21:09 GMT
Subject: python to perl conversion?
References: <slrn7p7q8j.n4.dyoung@baltazar.baltazar.org>
Message-ID: <3793F935.C4CEFD44@callware.com>

Hi All--

David Young wrote:
> 
> Is there a program that converts Python source to Perl source?  There is
> a program that a client wants written in Perl so that their Perl-only
> programmers (can you imagine?) can hack on it later. I haven't the time or
> inclination to learn Perl, and this project fairly cries out for Python,
> but I cannot justify the time expense of translating from Python to Perl
> by hand.
> 

Oooh, yuck!  We've had requests for Perl-to-Python before, but never for
this.  (Gag me with a Perl-handled spoon.)

There isn't one.  Are you volunteering?  Because I think you're the only
one who is interested.  At least you know your market, anyway. ...

> --
> ``I always thought that anybody who told me I couldn't live in the
> past was trying to get me to forget something that if I remembered
> it would get them in serious trouble.'' --Utah Philips

<by-god-that's-moose-turd-pie!>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From tim_one at email.msn.com  Sat Jul 31 01:15:54 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 31 Jul 1999 01:15:54 -0400
Subject: 1.5.2 for: else:
In-Reply-To: <slrn7q2459.6nk.wtanksle@dolphin.openprojects.net>
Message-ID: <000301bedb13$c38e5860$a32d2399@tim>

[William Tanksley]
> ...
> Oh, and if you REALLY like the current behavior of else-loop, you can
> still have it if else-loop were gone -- like this:
>
> try:
>  for x in list:
>    if iffers(x): raise "no!"
>    yadda(x)
> except "no!": pass
> else:
>  twiddle()
>
> That's two extra lines, and one level of nesting -- for the _exact_ same
> functionality, only much more likely to be comprehended.

Except that when I see that loop, I'm at a loss to guess whether the
"except" clause is intended to catch otherwise-uncaught "no!" exceptions
raised by list.__getitem__, iffers(), and yadda() too.  Presumably the
intent is that it should not, but there's no easy way to stop it from
catching unintended stuff too short of defining a unique exception for each
loop.  for+break+else doesn't have this problem.

use-exceptions-for-exceptions-and-break-for-things-that-shouldn't-ly y'rs  -
tim





From wtanksle at dolphin.openprojects.net  Fri Jul  2 12:49:34 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Fri, 02 Jul 1999 16:49:34 GMT
Subject: Python GUIs: Summary and Conclusion
References: <99FA4A1A1280BEBD.0A1ABA62CC989479.4C2C6022F695BEAC@lp.airnews.net> <19990629092757.A4998@optiplex.palga.uucp> <46F9FB3A1847A014.2CBD270176138613.20F84A39E6ED95DA@lp.airnews.net> <b_Pe3.558$6M6.167478@news.shore.net> <395C236424AEEAA7.C2D70DEDE2306228.0C4BC0DFFEC49C2C@lp.airnews.net>
Message-ID: <slrn7nprcu.3a4.wtanksle@dolphin.openprojects.net>

On Fri, 02 Jul 1999 03:28:45 GMT, Mikael Lyngvig wrote:
>On Thu, 01 Jul 1999 20:13:59 GMT, "Michael P. Reilly"
><arcege at shore.net> wrote:

>>Been there, done that.  My pyirc client (alpha release) was designed
>>for just about any user-interface, and comes with three: dummy (no
>>input or output, just batch-mode)), curses and Tkinter.  You just can't
>>switch between them at runtime, ;)  I've done it for other apps too.

>Well, the only reason I haven't been there is that I've only worked on
>one GUI app _ever_ - a 2D tile-based strategy game [1].  I did plan
>support for console apps (curses, etc., using a client-server
>architecture), but I didn't much fancy the idea of playing with a
>command line interface :)

I'm working on a 2D tile-based adventure game, which right now is text
only.  I'm expanding it so that it can use graphics, but I need some
libraries.  Let me know what you find!

>[1] Which I expect to change in the future, which is why I'm reviewing
>portable GUIs (I don't do non-portable stuff, if there's anything I
>can do to avoid it).

That's right.

>-- Mikael

-- 
-William "Billy" Tanksley



From ScherBi at BAM.com  Mon Jul 19 06:30:11 1999
From: ScherBi at BAM.com (ScherBi at BAM.com)
Date: Mon, 19 Jul 1999 06:30:11 -0400
Subject: Python/Gadfly and SQL wildcards
Message-ID: <3F8BA119B48AD11197470008C7A493CB98B47F@HQOBGEX1>

AFAIK, the ANSI SQL wildcard for multiple characters is % (percent sign),
and for a single character is _ (underscore).  These would equate to * and ?
in DOS land...

Bill

-----Original Message-----
From: Colm Rafferty [mailto:c.rafferty at formulae.co.uk]
Sent: Monday, July 19, 1999 3:48 AM
To: python-list at python.org
Subject: Python/Gadfly and SQL wildcards


Dear all,
I'm new to list.
I'm looking at gadfly to use SQL with python.
Does anyone know how to implement the wildcard in gadfly SQL?
e.g.
Select * from table where fieldname = 'A*'
or
Select * from table where fieldname like 'A*'

* does not work!
Though it does in other SQL languages

Thanks for any help

Colm Rafferty

-- 
http://www.python.org/mailman/listinfo/python-list



From afn08995 at afn.org  Sat Jul  3 00:05:15 1999
From: afn08995 at afn.org (Andy L Krietemeyer)
Date: Sat, 03 Jul 1999 00:05:15 -0400
Subject: Dumb beginner question; what is IDLE?
Message-ID: <377D8BFB.3700379@afn.org>

I've seen several references to IDLE.  Is this an IDE for python? Or
what is it?  Thanks,  Andy.




From tim_one at email.msn.com  Tue Jul 20 02:31:49 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 20 Jul 1999 02:31:49 -0400
Subject: Cyclops 0.9.4
In-Reply-To: <slrn7p7v9t.lg.nascheme@adler.dyn.ml.org>
Message-ID: <001201bed279$8c611140$139e2299@tim>

[nascheme at ucalgary.ca]
> [This message has also been posted.]

Ditto.

> Cool.  How about a feature that displays only cycles created
> since a marking function was called.  I think this would be most
> useful in programs that leak memory.  You don't care if cycles
> exist, you just don't want them to be continually created.  The
> way I see this working is you start up something like IDLE, do
> some stuff, call the marking function, do more stuff, call cycle
> showing function.

You can do this yourself by installing a cycle filter, that returns false
whenever it sees a cycle that's been seen before.  I don't think it could be
done more efficiently than that if implemented inside Cyclops, although
Cyclops knows how to subtract its internal references from the refcounts
delivered via get_rootset().  A user-written cycle filter can do that too
for its own references, but it is a little tricky to get straight.

Simpler, just call find_cycles() followed by show_stats() from time to time.
One of the output lines is "# cycles found: <n>".  New cycles are being
found if & only if n is increasing.

Now if you actually *try* any of these and find them helpful in "real life",
pass it on!  My experience in IDLE was that cycles weren't actually causing
the leaks, although up until it hit a dead end they were correlated with
leaks; the real cause was references in basic (reachable) objects that
weren't getting cleared out when the widgets they referenced got closed.

enough-to-make-you-dizzy-ly y'rs  - tim





From fredrik at pythonware.com  Thu Jul 22 07:10:27 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Jul 1999 11:10:27 GMT
Subject: IDLE mini-hack & ?
References: <199907211257.HAA14822@cro.ots.ac.cr> <37969A3D.EEDC201@compaq.com>
Message-ID: <008f01bed432$ce4b7ea0$f29b12c2@secret.pythonware.com>

Greg Ewing <greg.ewing at compaq.com> wrote:
> Could the traceback printing thingy perhaps be
> enhanced to go just a little bit further down the
> chain and report the filename/linenumber of the
> function being called, not just the line it was
> being called from?

agreed.  but the problem is of course that
"apply" never completes the call, so the
offending function never appears on the
frame stack.

...

one solution could be to add the missing
information to the TypeError description
string, in a few places in ceval.c. like this:

1. add a support function to ceval.c which
takes a code object and an error message
(and possibly an exception object).

2. let this function combine the message
with the filename, linenumber, and function
name (if any) from the code object into a
message string, and call PyErr_SetString
with the result.

for details on how to get the names and
numbers, see code_repr in compile.c. you
might wish to add a support function to
compile.c for this purpose, maybe some-
thing like:

int PyCode_Describe(co, &filename, &lineno, &name)

or maybe you could cheat and use PyObject_Repr
to get the code object descriptor...

3. change the calls to PyErr_Format(TypeError)
and PyErr_SetString(TypeError) in compile.c, to
use this support function instead.

4. test it carefully, and submit patches to GvR.

...

an easier solution is to simply leave the offending
frame on the stack (it's there when the error is
detected). not sure about the side effects of
that solution, though...

</F>




From fredrik at pythonware.com  Tue Jul 13 12:16:52 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 13 Jul 1999 16:16:52 GMT
Subject: OK python gurus!
References: <7mfn4c$m80$1@nnrp1.deja.com>
Message-ID: <002201becd4b$2206e0f0$f29b12c2@secret.pythonware.com>

K Kal <k_kal at my-deja.com> wrote:
>      I'm new to python.  Where should I get started?  Any good
> references on-line.  Tutorials/sample code would be appreciated.

from the "posting guidelines" document that was
posted earlier today:

...

>From the Python FAQ:

  "Python is an interpreted, interactive, object-oriented programming
  language.  It incorporates modules, exceptions, dynamic typing, very high
  level dynamic data types, and classes. Python combines remarkable power
  with very clear syntax. It has interfaces to many system calls and
  libraries, as well as to various window systems, and is extensible in C or
  C++. It is also usable as an extension language for applications that need
  a programmable interface.  Finally, Python is portable: it runs on many
  brands of UNIX, on the Mac, and on PCs under MS-DOS, Windows, Windows NT,
  and OS/2.

  To find out more, the best thing to do is to start reading the tutorial
  from the documentation set at http://www.python.org/doc/

Example: "Hello World" in Python
--------------------------------

The original task can be accomplished by a one-liner:

  print "Hello World!"

Here is a longer example, a function that returns a string
containing a series of multiple "Hello World!" greetings:

def my_hello(how_often):
        retval = how_often * "Hello World! "  # "multiply" string
        return retval[:-1]                    # strip off trailing space

Usage:
>>> my_hello(5)
'Hello World! Hello World! Hello World! Hello World! Hello World!'

Note that block structure in Python is defined by indentation, rather than
block delimiters (as used in many other programming languages such as Perl,

...




From amitp at Xenon.Stanford.EDU  Fri Jul  9 14:32:31 1999
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 9 Jul 1999 18:32:31 GMT
Subject: Curses, Ncurses, Slang, etc?
References: <19990703140217.D14374@moebius.dartmouth.edu>
Message-ID: <7m9tmr$6qg$35@rks1.urz.tu-dresden.de>

 Eric Kidd  <eric.kidd at pobox.com> wrote:
| I'd like to write a Python application which uses either curses, ncurses,
| slang or a similar tty interface library. Unfortunately, the built-in
| curses module is hopeless broken on my RedHat 6.0 system (it doesn't export
| 'addch', for example).

I've been learning curses using Python & RedHat 6.0, and haven't had
any problems so far.  Could you show us a small program that results
in an error?

   - Amit





From MHammond at skippinet.com.au  Thu Jul  8 19:05:35 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 9 Jul 1999 09:05:35 +1000
Subject: setting up shortcuts from python
References: <37850578.EEEB73FE@mmrd.com>
Message-ID: <7m3ao3$saa$1@m2.c2.telstra-mm.net.au>

DejaNews should have quite a few references to this - in fact, I saw some
code posted just yesterday (so your newsreader can probably still find it!)

Mark.

Lori Ludick wrote in message <37850578.EEEB73FE at mmrd.com>...

>I need to set up shortcuts from inside of my python script.
>I have looked through some of the documentation so far and
>have not even found a direction to go in.
>
>I assume I need to find the magic windows api that
>will let me create a shortcut and set up all the nice shortcut
>informaton it needs...but I haven't had any luck.






From bkc at Murkworks.com  Sat Jul 24 11:20:08 1999
From: bkc at Murkworks.com (Brad Clements)
Date: Sat, 24 Jul 1999 11:20:08 -0400
Subject: Memory leaks whilst embedding python
References: <1279335510-28879097@hypernet.com>
Message-ID: <7nclk9$sg1$1@news.clarkson.edu>

This, however, is going to be a giant problem for me.

I'm porting Python to Novell NetWare, and that platform complains quite a
bit when a process is unloaded without freeing up it's memory. Also, SWIG
produces wrappers that suck up a lot of memory in init.

Because Python doesn't finalize modules, I can't clean up this memory...

I'm going to have to write my own memory allocator and fake out Python and
all add-on modules to use it, what a major pain.


--
Brad Clements,
bkc at murkworks.com
Gordon McMillan <gmcm at hypernet.com> wrote in message
news:1279335510-28879097 at hypernet.com...
> Chris Williamson discovers that:
>
> >      Py_Initialize();
> >      Py_Finalize();
>
> leaks.
>
> Well, if MSVC reported on all of MFC's static allocations, you'd see
> that it leaks, too.
>
> These are nothing to worry about. It's static data that Python needs
> to run. It won't continually grow. If you use the debug memory stuff,
> you should take your checkpoint after Py_Initialize and report before
> Py_Finalize. In fact, you should probably take your checkpoint after
> you've done your imports, because a lot of the memory they use won't
> get released until Py_Finalize.
>
>
>
> - Gordon
>
>





From claird at Starbase.NeoSoft.COM  Fri Jul 16 13:32:41 1999
From: claird at Starbase.NeoSoft.COM (Cameron Laird)
Date: 16 Jul 1999 12:32:41 -0500
Subject: Harlequin Is Back!
References: <%vti3.121$95.4053@newse2.tampabay.rr.com> <wkn1wzl9kt.fsf@ifi.uio.no> <71E04BBBB50245BC.87397A36A7EDBBA0.44F0816EBAA77500@lp.airnews.net> <378E57D8.A77A9F9C@nmia.o_r_g>
Message-ID: <302747D6133FC474.B42081F80C6F00FB.672284C91DB0497B@lp.airnews.net>

In article <378E57D8.A77A9F9C at nmia.o_r_g>,
David Hanley  <fsnm at nmia.o_r_g> wrote:
			.
			.
			.
>> scene.  This is interesting; Tcl and Lua emphasize ease
>> of embedding and extensibility in their design, and Perl
>> has hordes of interfacing worker bees, but a consistent
>> pattern emerges that Python does at least as well as any
>> other language in achieving connection to the latest
>> toolkits-protocols-...  It's an interesting story, even
>> though I don't yet understand it.
>
>Hmmm, maybe the right approach is to get lisp to talk
>to python, then, eh?
>
>dave
>

I suspect you are cheerfully mocking the extremes such an
approach might reach.  In any case, the reality is that
serious programmers find such extremes useful!  With Tcl,
for example, the pressure to do solid COM interfaces has
largely dissipated because of the availability of Robin
Becker's Python server, which supplies these and other
connections.  Coincidentally, I report today on several
related projects of cross-language co-operation in <URL:
http://www.sunworld.com/swol-07-1999/swol-07-regex.html#2>.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From thstr at serop.abb.se  Thu Jul 15 04:20:20 1999
From: thstr at serop.abb.se (Thomas S. Strinnhed)
Date: Thu, 15 Jul 1999 10:20:20 +0200
Subject: Oneliners?
Message-ID: <378D99C4.BDACE15B@serop.abb.se>

Hi all

Before anything else, I *do not* want to start a
my-language-is-better-than-yours-kind-of-thread here,
so please don't give me the reasons why I should use
this language over that and so on. 
I'm only interrested in pure facts :-)

And now for...the question:
I looked at the "Cryptographic Curiosities" at
http://starship.python.net/crew/amk/writing/crypto-curiosa.html 
about short Python crypto-scripts.

Are there other examples of scripts that do a lot of cool stuff
in one or just a few lines?
(here comes the not-an-intended-flame-bait stuff :-)

I heard there was a webserver in one line of Perl,
does Python match that (or two or three...)?

Of course, the length of code is not my primary criteria
in choise of language, but it's rather cool in a geeky kind of way,
don't you think?

I've-been-told-the-size-doesn't-matter-ly y'rs
 -- Thomas



From aahz at netcom.com  Sat Jul 10 23:12:50 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Jul 1999 03:12:50 GMT
Subject: DOC BUG: os.chdir() (was Re: Simple Qu: Changing directory)
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787D8DA.3774AA7D@ssec.wisc.edu> <7m8p7t$o7s@dfw-ixnews10.ix.netcom.com> <3787FB51.12FE0048@ssec.wisc.edu>
Message-ID: <7m9to8$6qg$84@rks1.urz.tu-dresden.de>

First of all, please try to follow proper netiquette for posting.  Your
lines should be 72 characters or less, and you should put your comments
after the quoted material:

In article <3787FB51.12FE0048 at ssec.wisc.edu>,
Nick Bower  <nick.bower at ssec.wisc.edu> wrote:
>
>yeah, that's what i originally thought. sure enough there were some
>updates in the os documentation, but not os.chdir. it's something that
>would be really helpfull if anyone linked to the documentation sees
>this.

Ah, I see what the problem is.  I went to the index inside the library
reference to make sure os.chdir() was documented, but didn't check to
see *where* it was documented.  Sure enough, it's documented inside the
"Process Parameters" section of os (6.1.1), rather than in the "Files
and Directories" section.

Fred, here's the URL that needs to be fixed:
http://www.python.org/doc/current/lib/os-procinfo.html

I think I understand how it happened: os.getcwd() is (probably
rightfully) documented there, so someone thought os.chdir() belonged
there, too.  AFAICT, that's the only mistake in that area of the docs,
though I could make a case that os.getcwd() should be moved (or at least
have a pointer in "F&D").
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From donn at u.washington.edu  Fri Jul  2 13:37:04 1999
From: donn at u.washington.edu (Donn Cave)
Date: 2 Jul 1999 17:37:04 GMT
Subject: Socket programming.
References: <jevhfnnqbgz.fsf@apollo.orakel.ntnu.no>
Message-ID: <7litc0$1b7s$1@nntp6.u.washington.edu>

"Kjetil ?degaard" <kjetilod at stud.ntnu.no> writes:
| I'm trying to implement a simple irc module as the base for a simple
| irc client.  However, I'm having some design problems, having never
| done any socket programming before.
|
| My current implementation is threaded (using fork()); one thread for
| user interaction and one for server output.  It uses blocking
| sockets.  It works ok, but I've read some docs here and there (amongst
| them the Socket Programming HOWTO on www.python.org) and I get the
| feeling I should be using select() and non-blocking sockets.  Also,
| I'm not sure I want to use fork() -- it seems very low-level and
| depends on a Unix-like system.

These are questions everyone wrestles with, and the only thing we can
say for sure is ``it depends.''  If your current design works, it sounds
fine to me.

As you know, select comes in when you want to dispatch from two or more
I/O event sources, in a single execution thread.  (With select, you really
don't need non-blocking sockets, because select's job is to predict whether
a read would block.)  If the event handling can be simplified to the point
where with a modest amount of global state information, the program can
handle any event in no time and get back to the dispatch loop more or less
right away, then it's worth thinking about a single threaded design.

But it can be more complicated than you expect - for example, when you
write to a socket device, write() or send() may return after writing less
than the full amount, and you have to keep sending until everything goes.
So your global state may have to queue outbound data.  (Use select to find
out whether write will block, too.)  This is the kind of thing you'd need
to consider in a single threaded service, not necessarily in a client.

This kind of worry is avoided with a multi-threaded design, and then
the question is whether to use processes or threads.  Processes are
more expensive and indeed a bit less portable.  My favorite OS, BeOS,
supports fork - but a child fork doesn't inherit an open socket, so
fork is a very difficult way to make a multi-threaded socket application!
Moreover, select is available for sockets, but the last I checked it
doesn't work on all types of file descriptors.  (I present this as
an example of a non-UNIX socket system that I think even its designers
would admit is slightly lame.  I expect something better is on the way.)

BeOS enthusiasts aren't troubled by this, they feel that threads are
available to solve all these dispatching problems.  But threads carry
their own problems, as there's a lot more potential for collision when
two threads run in the same address space, so you need to sweat interlocks
that aren't an issue in separate forks.  (And Python itself has to deal
with this, as a threaded C program, which complicates the issue.)
Meanwhile the larger program sort of loses control over its threads,
compared to the single threaded select dispatching described above,
because threads blocking on I/O can't process other events, they can't
tell you whether they're actually reading or just blocking, etc.

So the old ways are best, is what I say.  Unless they aren't, and at
times they aren't.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From dan at control.com  Thu Jul  8 10:51:04 1999
From: dan at control.com (Dan L. Pierson)
Date: Thu, 08 Jul 1999 14:51:04 GMT
Subject: COM with Python?
References: <1280807192-67537824@hypernet.com> <7m1kgr$7nt$1@m2.c2.telstra-mm.net.au>
Message-ID: <3784ba0a.66776109@news.randori.com>

"Mark Hammond" <MHammond at skippinet.com.au> wrote:
> Ive never needed to implement a singleton, but my personal solution (from
> someone else I cant recall in those threads) would be to implement the
> functionality in a single Python object and have the COM objects delegate to
> it.  Thus you dont really have a singleton COM object (as you are giving off
> as many as necessary) but by keeping the functionality and state in a single
> Python object you are can get the same result (for most cases!)

This is probably the right thing to do in any case.  There are a lot
of practical problems with implementing real COM singletons.  This is
covered in gory detail in "Effective COM" topic 4: "Beware the COM
singleton".  Your approach is essentially what they recommend (modulo
programming langauge :-)).

Dan Pierson, Control Technology Corporation
dan at control.com



From Kjetil.Ødegaard at p98.f112.n480.z2.fidonet.org  Fri Jul  2 16:10:20 1999
From: Kjetil.Ødegaard at p98.f112.n480.z2.fidonet.org (Kjetil Ødegaard)
Date: Fri, 02 Jul 1999 21:10:20 +0100
Subject: Socket programming.
Message-ID: <0000167a@bossar.com.pl>

From: "Kjetil ?degaard" <kjetilod at stud.ntnu.no>

I'm trying to implement a simple irc module as the base for a simple
irc client.  However, I'm having some design problems, having never
done any socket programming before.

My current implementation is threaded (using fork()); one thread for
user interaction and one for server output.  It uses blocking
sockets.  It works ok, but I've read some docs here and there (amongst
them the Socket Programming HOWTO on www.python.org) and I get the
feeling I should be using select() and non-blocking sockets.  Also,
I'm not sure I want to use fork() -- it seems very low-level and
depends on a Unix-like system.

Advice would be appreciated.  If you think I'm being unclear or vague
-- you're probably right. ;)  I'm not really sure what I'm asking,
either.

-- 
Gravity brings me down.



From rodisi01 at my-deja.com  Wed Jul 21 10:00:46 1999
From: rodisi01 at my-deja.com (Isidor)
Date: Wed, 21 Jul 1999 08:00:46 -0600
Subject: String2List: odd function
Message-ID: <199907211304.IAA14937@cro.ots.ac.cr>

Dear Pythonographers -

Here is a very odd little something I have written in the process
of trying to make ModuleWriter module (see previous message) deal
with "documentary" strings.

Takes a list-like string, e.g.,:

"[['itema1', 'itema2'], 'itemb1', ['cheese1', ['cheese1a',
'cheese1b'], '2'], ''blah' rest of itemb2', 'zib']"

and returns *real* lists:

[['itema1', 'itema2'], 'itemb1', ['cheese1', ['cheese1a',
'cheese1b'], '2'], "'blah' rest of itemb2", 'zib']

Which, in text-world, looks like nothing more than the removal of
some quotation marks, but in Python-world...well, I don't have to
tell you what the difference is.

This function is probably a poster child for code-in-need-of-
optimization. I have probably written the slowest possible
version (actually, that's a boast to which I don't have the right
to make either!)...I have written the most obvious clunker-headed
version. I will be interested to hear suggestions for making it
speedier and svelter. (I suspect that Guido's optimizations
essay(s) may be useful here. I will have to review them.)

GETTING IT: (repeated from my ModuleWriter message)

I don't have web space where I can put this right now, so e-mail
me at rodisi01 at my-deja.com, and every day or every other day, I
will send a copy off to whoever has requested it. I know this is
a hassle, I'm sorry. If it looks like there is serious interest
in this thing, I will find some web space for it so you can click
and save to your heart's content. In fact, I thought I heard
someone in this list talk about offering web space for python
projects....;)...

ALTERNATIVELY:

If someone will point me to (or give me) some suggestions re.
posting code to the newsgroup, I could just do that. I'm worried
about line breaks and all that stuff. Would an ascii attachment
work in deja-world? (I've never had much access to newsgroups, so
this is all sort of new to me, and I have to use the free and
powerful (many many thanks!) but somewhat frustrating deja-news
interface.)

I hope someone finds this interesting or at least amusing!

Isidor





From drobinow at dayton.adroit.com  Fri Jul 23 16:59:50 1999
From: drobinow at dayton.adroit.com (David Robinow)
Date: Fri, 23 Jul 1999 16:59:50 -0400
Subject: Python documentation available as ZIP files
References: <14232.40613.53508.888049@weyr.cnri.reston.va.us>
Message-ID: <7nal1u$6cb$1@tribune.oar.net>

Fred L. Drake wrote in message
<14232.40613.53508.888049 at weyr.cnri.reston.va.us>...
>
>  I've received enough hostile mail from Windows users not
>understanding how to unpack .tgz archives (er, WinZip works for me...)
>that I will be making most of the documentation packages available as
>ZIP archives as well as .tgz archives.
 This is really sad.  You're not really doing Windows users any favors by
doing this.
"tar" and "gzip" are "must know" tools.
Let me quote from the GNU macs on Windows FAQ

"You can find precompiled versions of all of the compression and archive
utilities in ftp.cs.washington.edu/pub/ntemacs/utilities/. "

  It would seem to make more sense to include gzip and tar.on the web site.






From claird at starbase.neosoft.com  Mon Jul 19 20:30:27 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Jul 1999 00:30:27 GMT
Subject: [Q] In-Browser technology
References: <932337314.611.103@news.remarQ.com>
Message-ID: <BE2245BA1DCDAE31.8804258C1F60F9A7.2EB4BCD57DBA05D1@lp.airnews.net>

In article <932337314.611.103 at news.remarQ.com>,
Vladik <reason at shadow.net> wrote:
>Hello,
>Sorry for cross-posting, but
>my question is really to the people on the news groups
>who know about Eiffel, Python and Perl.
>
>Basically,
>We are to develop a web client (a program that
>can be ran witting a web browser) for our otherwise n-tier
>cross-platform system, our research department gave the
>recommendation and actually a prototype of the client
>in MS ASP (active server pages) and, I think but not
>sure, blended with Visual Basic. Talk about bias...
>But this means that it can only run in MS IE and not
>in Netscape, also it limits the number of platforms not only
>for the client on which web browser runs, but also for
>the web server (MS platforms only, because web server with
>active pages can only run there)
>There is some talk about Java, but our web client
>will also be ran via Serial links (not T1 connections)
>on relatively inexpensive systems -- therefore Java is
>something to look into but it may be slow.
>So I am set on a quest to find something
>that
>a) can be executed within IE or Netscape,
>b) be faster and overall more resource friendly then JavaScript technology
>c) supports OO programming
>d) does not depend on the platform on which web server is running.( if the
>web server is running NT or UNIX (sun, linux)
>
>I have found one thing that almost satisfies the requirements:
>it is based on Oberon-2 programming language and so far sounds
>VERY promising.
>  The reason why I am still looking is because
>there are more books and internet support for Python,Perl and Eiffel
>then for Oberon (this will be a factor when presenting the proposal
>to the management).  The other problem is that there is plugin
>for this technology available for NT and Mac platforms (for both
>IE and Netscape) and is not available for UNIX platforms (which means that a
>web browser can not be ran on UNIX). There is source code,
>though.
>If interested: http://caesar.ics.uci.edu/juice/
>
>So, finally, my question is is there a technology
>I am looking for available for either Perl or Eiffel or Python
			.
			.
			.
No.

That's the short answer.  Slightly longer is this:  I
don't understand--although I'm sufficiently intrigued
to answer.

Let's first dispose of the part of my reply where I
pick nits from your description:
1.  "We are to develop a web client (a
    program that can be ran witting a
    web browser) ..." evokes to me a
    plugin.  Is a plugin (supposing a
    sufficiently potent one exists)
    exactly what meets your require-
    ments, or is there a point I'm
    missing?
2.  "... a prototype of the client 
    in MS ASP (active server pages)
    and, I think but not sure, blended
    with Visual Basic ..."  I have no
    definite idea of what you're say-
    ing here.  Is it that your
    customer has a picture of the
    visual appearance of the desired
    deliverable?  What is there about
    the ASP prototype that does *not*
    meet the requirements?  Is VB
    blended on the server or client
    side?  Does it matter?  Is it really
    VB, or VBScript?
3.  Are you aware that *lots* of browsers
    read pages served as ASP?  You can
    use ASP without requiring IE.
4.  "... Java ... may be slow."  Lots of
    things may be slow.  The market-lead-
    ing browsers are definitely pigs, by
    my standards, but they've already
    been mandated, according to your
    description.

    Technical point:  Java applets work
    well for me.  Invariably, when I hear
    people complain about the size of
    Java applets, they're really com-
    plaining about ancillary images.
    Java executable classes themselves
    are admirably compact, in my experi-
    ence.
5.  Of what resource is JavaScript not
    sufficiently conservative?  Do you
    seriously have performance problems
    with it?  There are plenty of aspects
    of JavaScript that deserve scorn,
    but, unless I have a specific issue,
    I happily use it to accomplish real
    work on the client side.
6.  How much client-side programming are
    you doing?  Why does object-orienta-
    tion there matter to you?
7.  What technologies do you want that
    *do* link client and server platform
    compatibility?  I can't make this
    constraint informative in the context
    of everything else you've written.
8.  What's the part about Juice that you
    like--is it the promise of superior
    performance?  Do Franz and Kistler
    truly make source available?  I
    hadn't noticed that.  In any case, if
    I thought it'd help with one of my
    projects, I'd write them directly.

    I'm fond of Oberon, and I wish Juice
    well.  I don't understand your inter-
    est in it.

It sounds as though you're involved in quite an ambi-
tious project, of the sort I like.  I wish you well.
I suspect you'll do well to clarify several of your
requirements; this will make it much more inviting for
comp.lang.* readers to help you.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From tim_one at email.msn.com  Sat Jul 17 03:13:05 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Jul 1999 03:13:05 -0400
Subject: Python and Boehm-Demers GC, I have code.
In-Reply-To: <7mov67$bfg@ds2.acs.ucalgary.ca>
Message-ID: <000401bed023$d08d3360$06a22299@tim>

[Tim]
> ... Did the "collection" phase of BDW find any trash at all?

[Neil Schemenauer]
> It doesn't unless you create some.  The patch creates a version
> of Python that uses GC only to clean up reference cycles.

Just as it appeared <wink>.  So it's really no surprise that the overall
timing didn't change much in pybench, right?

> ...
> BTW, _tkinter is not working perfectly yet.  I think lambdas
> passed as callbacks are getting collected.

Such as to key bindings?  Harmonic convergence (I was just looking at that
in IDLE tonight!).

When you pass a lambda (or bound method object, or any callable object -- CO
for short) as a binding, Tkinter.py eventually wraps the CO in an instance
of Tkinter.CallWrapper, and the .__call__() method of that instance is
passed on to _tkinter.c's Tkapp_CreateCommand via Tkinter.Misc._register().

The interesting thing here is that nothing in Tkinter.py holds on to
anything from which the CO can be reached.  Instead _tkinter.c's
Tkapp_CreateCommand bumps the CO's refcount, stuffs a pointer to the CO in a
heap-allocated new instance of a PythonCmd_ClientData struct, and passes a
pointer to the latter on to Tcl via Tcl_CreateCommand.  _tkinter.c doesn't
hold on to a reference to this either:  when Tkapp_CreateCommand returns,
Tcl the sole owner of the CO reference that was added, and Tcl is the only
one who knows the address of the PythonCmd_ClientData struct.

So if the CO is an embedded lambda expression, there is only one reference
to it in total, and only Tcl can reach it (which it does when the binding
triggers, and then it passes a pointer to the PythonCmd_ClientData struct
back to _tkinter.c's PythonCmd).  So if BDW isn't crawling over Tcl/Tk's
memory too, the lambda will look like trash, and the PythonCmd_ClientData
struct will look like trash regardless of what flavor of CO was passed.

So BDW has to get at Tcl's memory too, or _tkinter.c has to maintain a list
of allocated PythonCmd_ClientData thingies (until PythonCmdDelete is called
back from Tcl), or ...

It's convoluted for sure, but I do expect you'll find "the answer" at the
bottom of it.

rc-is-so-simple<wink>-ly y'rs  - tim





From rceci at adams.com  Tue Jul  6 12:10:43 1999
From: rceci at adams.com (rceci at adams.com)
Date: Tue, 06 Jul 1999 16:10:43 GMT
Subject: HELP: moving Python Distribution
Message-ID: <7lt9po$htr$1@nnrp1.deja.com>

Hello,

I have a question about relocating a python distribution (1.5.1) that
was built and installed in the normal area ( i.e. prefix=/usr/local ).

I have moved the bin/*, include/python1.5/* and lib/python1.5/* to a NFS
mount that will support multiple architectures, e.g.:

bin/python -> $mount_pnt/python/bin/$ARCH/python1.5/python
include/python1.5/* -> $mount_pnt/python/include/$ARCH/python1.5/*
lib/python1.5/* ->  $mount_pnt/python/lib/$ARCH/python1.5/*

I have set up PYTHONHOME and PYTHONPATH to support this.  When I run
python on the build machine, everything works as planned.  When I tag
the "-v" on to python, I can see the expected messages that show the
built-in modules being loaded from my NFS "$mount_pnt" rather that the
build directory "/usr/local".  Thats great.  But when I log onto another
system of identical architecture ( in this case both machines are IRIX
6.4 MIPS4 systems), I get:

Could not find platform independent libraries <prefix>
Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]

And yet the "-v" still prints it is loading from the $mount_pnt.  This
second system does not have python installed in "/usr/local..." at all.
I am guessing that specifiying PYTHONHOME/PYTHONPATH works, but the
iterpreter will still print warnings if it cannot find python runtimes
under the original exec prefix location.  Is there a way to suppress
these warnings or fix them in the source?

Thanks

Rob Cecil
rceci at adams.com


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From ovidiu at cup.hp.com  Thu Jul 29 13:40:12 1999
From: ovidiu at cup.hp.com (Ovidiu Predescu)
Date: Thu, 29 Jul 1999 10:40:12 -0700
Subject: any way to know when the program is exiting? 
In-Reply-To: Your message of "Thu, 29 Jul 1999 13:45:01 +0200."
             <37A03EBC.21E689E6@lemburg.com> 
Message-ID: <199907291740.KAA32483@hpcll563.cup.hp.com>

On Thu, 29 Jul 1999 13:45:01 +0200, "M.-A. Lemburg" <mal at lemburg.com> wrote:

> Ovidiu Predescu wrote:
> > 
> > Hi,
> > 
> > I was wondering if there is any way in a __del__ method to know when it's
> > invoked on the way out, as a result of the program going away?
> 
> It's a hack, but you can check sys.path: if it's None, then your program
> is about to exit. To see why, check the output of 'python -vv -c "None"'.

Thanks, this seems to work. I looked on the source code and it seems that the 
sys
module is treated specially in the cleanup phase.

Checking to see if my global variable is None doesn't work though. I guess this
happens because I use some weak references to avoid cyclic references and my
WeakReference C extension type gets collected before the classes that use it.
The global variable I'm trying to access from the __del__ method gets replaced
by a string object somehow, resulting in an exception being thrown (and ignored
by the interpreter).

> Would probably be a good idea to add some flag to the sys module to be
> able to check the current running state of the interpreter:
> 
> E.g. sys.state:
> 	'startup' - interpreter is intializing
> 	'ready' - interpreter is up and running
> 	'exiting' - interpreter is exiting

I agree, this would be nice.

Thanks a lot for all the responses!

Ovidiu

-- 
Ovidiu Predescu <ovidiu at cup.hp.com>
http://andromeda.cup.hp.com/  (inside HP's firewall only)
http://www.geocities.com/SiliconValley/Monitor/7464/





From kernr at mail.ncifcrf.gov  Tue Jul 27 19:16:04 1999
From: kernr at mail.ncifcrf.gov (Robert Kern)
Date: Tue, 27 Jul 1999 23:16:04 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <p51zdvn4c7.fsf@bidra241.bbn.hp.com> <14236.32616.980578.536614@dolphin.mojam.com> <379D89BC.DA91BA8C@hp.com> <14237.60407.130426.461846@dolphin.mojam.com> <14238.335.458728.140169@lrz.de> <14238.1372.334823.378192@dolphin.mojam.com> <379e2890.6069849@news.erols.com> <14238.12851.809608.195391@dolphin.mojam.com>
Message-ID: <379e3d0d.11315603@news.erols.com>

On Tue, 27 Jul 1999 17:45:44 -0500 (CDT), Skip Montanaro
<skip at mojam.com> wrote:

>    Robert> Just to address this for-instance, you can compile Python
>    Robert> extensions for Win32 systems using the free egcs compiler.  See
>    Robert> <URL: http://starship.python.net/crew/kernr/mingw32/Notes.html>
>    Robert> for more information.
>
>That's all well and good.  This is why I mentioned "technical depth".
>Someone who's done very little programming, discovers Python and is merrily
>whacking things into shape, then finds they need to recode a hotspot or two
>in C is quite probably not going to have the mental model built up in their
>brain to allow them to connect egcs, Python, extension modules, compilation,
>dll building, (yadda, yadda, yadda) into a coherent whole.  It's also
>unlikely they'll find Robert's excellent notes about egcs & Python on
>Windows.

Oh yes, I agree on all points.  To be honest, I just wanted to
shamelessly plug the system.  :-)

[snip good points]

>Q-E-D-ly y'rs
>
>Skip Montanaro	| http://www.mojam.com/
>skip at mojam.com  | http://www.musi-cal.com/~skip/
>847-475-3758

excellent?-<snort>-ly y'rs

Robert Kern           |
----------------------|"In the fields of Hell where the grass grows high
This space            | Are the graves of dreams allowed to die."
intentionally         |           - Richard Harter
left blank.           |



From anthony at lokie.force9.net  Wed Jul 14 09:44:20 1999
From: anthony at lokie.force9.net (Anthony McDonald)
Date: Wed, 14 Jul 1999 14:44:20 +0100
Subject: Recursive method
References: <378B499B.4301F893@eml.villa-bosch.de> <378BAD30.7D1A5196@hvision.nl> <378C214F.3CF9CDD3@eml.villa-bosch.de>
Message-ID: <cA0j3.2980$34.3306@wards>

Ralph Gauges <ralph.gauges at eml.villa-bosch.de> wrote in message
news:378C214F.3CF9CDD3 at eml.villa-bosch.de...
> Hans Nowak wrote:
> > You have a nested function here. (I say function, because it does not
> > look like a method... it's missing the self argument!
>
>
>
> Actualy it is a method, but JPython doesn't seem to have a
> self. I was missing that as well.
>
>
>
> >
> > So, to solve this, you could try a different approach... maybe
> >
> > (you said it's a method, right?)
> >
> > class Blahblah:
> >     ...
> >     def makeTree(self):
> >          ....
> >     def makenodes(self, C):
> >          ....
> >
>
>
>
> I tried that as well, but I got the same error message. Only
> after I moved the makenodes outside the class it worked.
> This might also be related to the missing 'self' in JPython
> since I can not call 'self.makenodes'.
> Since moving makenodes out of the class is actualy not what
> I want, I was wondering, if anybody has had a similar
> problem and found a solution.
>
>
>
> Thanks for the answers. Although it is still not, what I
> want it to be, it works and I can go on, till I (or someone
> else) find a better solution.
>
> Ralph

This should provide you with the basic framework to implement your tree.
class tree:
    def __init__(self, data):
        self.count=1
        self.item=data
        self.left=None
        self.right=None
    def add(self, data):
        if self.item == data:
            self.count=self.count+1
        elif data < self.item:
            if self.left == None:
                self.left = tree(data)
           else:
                self.left.add(data)
        elif data > self.item:
               if self.right == None:
                    self.right = tree(data)
               else:
                    self.right.add(data)
    def list(self, data):
        if self.left != None:
            self.left.list(data)
       data.append([self.item, self.count])
       if self.right != None:
           self.right.list(data)






From tony at lsl.co.uk  Fri Jul 16 07:12:38 1999
From: tony at lsl.co.uk (Tony J. Ibbs (Tibs))
Date: Fri, 16 Jul 1999 11:12:38 GMT
Subject: Meta-language for mxTextTools
Message-ID: <015601becf7c$1dff7d90$f0c809c0@lslp7o.lsl.co.uk>

M.-A. Lemburg wrote:
>Well, it's just that "def" state exactly what you are doing
>when you define a table. "is" is only used in comaprisons in
>Python, so the analogy doesn't go too far.


Point taken, although the use of "is" allows one the nice side-effect that:

    comment is:
        'comment' = Table is:
            Is '#'
            AllNotIn '\n\r' F:MatchOk

falls out neatly, with the "is:" taking the place of the table name
argument. It somehow feels less neat to me if one has to use "def" in there
somehow (before or after the "Table"?). I don't pretend to have designed
that in - it was a happy coincidence.

>You might even consider using "table" to start the section.
>This wouldn't conflict with functions/methods in Python and
>still express what you are aiming at.

Hmm. Same problems as above.
Unfortunately, whilst it is hacky, I *like* the fact that Python mode, etc.,
recognise "is" as a keyword and highlight it. Still, it's obviously a bigger
itch than I'd expected, so I'll reflect on it carefully when I start
implementation.

>There is a small difference in semantics between "previous" and
>"back": you are using "previous" to move the instruction pointer
>in the table, while "back" moves the engine's current position
>in the text. So having to different synonyms may be a good idea
>(the wording is secondary, IMHO).


Whilst I don't particularly care if the same word is used in different
contexts, I think I tend to agree that it might be confusing (and perhaps
smells a little Perl-ish). I agree that the wording is secondary. Given that
we only *need* one word (for skip -1, since other skips are more complex or
rarer), I guess "back" is probably the one to choose for now (perhaps what I
really want is to be able to say "Unread" instead of "Skip" when one is
going backwards - does that make any form of sense as extra syntactic
sugar?).

>...my major gripe with the tables: you constantly
>have to adjust offsets whenever you insert things. A meta language
>ought to take care of this by introducing jump targets.


For some strange reason I've always been bad at integer offsets (so handle
things like "last-but-two" very badly). Thus my *first* thought on looking
at mxTextTools was a wish to get rid of them. The problem of *editing* the
tables and having to adjust offsets is obviously such a horrible prospect it
hadn't even occurred to me, but yes, that is also a major advantage of using
labels.

I *really* will have to implement this when I get back from holiday now (oh,
if only someone would buy me a portable PC!)

Tibs
--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.demon.co.uk/
Feet first with 5 wheels...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)





From dalke at bioreason.com  Mon Jul 12 18:00:09 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Mon, 12 Jul 1999 16:00:09 -0600
Subject: bug with Irix configure
References: <7mdkm1$4c3$1@sun500.nas.nasa.gov>
Message-ID: <378A6569.A5CEDC4E@bioreason.com>

Patrick J. Moran <pmoran at nas.nasa.gov> said:
> I've encounted a bug with the configure on Irix.  I'm not sure
> where to report it, so to the newsgroup it goes.

I've had no problems compiling Python on IRIX 6.3 or 6.5.2
boxes with either the MIPSpro 7.1 or 7.2 compilers using
all combinations of n32, o32 or 64.

> The problem is under Irix 6.4 with the MIPSpro 7.2 compiler.

But I don't have one of those machine to test it on.. not that
it should make a difference.

> I presume _MIPS_SZLONG isn't defined when configure does its
> tests.

They should be added by the compiler, since it depends on the
ABI you are using for compilation.

I just compiled the following from "test.c"

#include <sys/types.h>
main()
{
}

with "cc -E" to see what the preprocessor does, and it gets
size_t w/o problems (IRIX 6.5.2 and MIPSpro 7.30).

You can also see what is being set with "cc -show test.c".
I get:
/usr/lib32/cmplrs/fec -G [...] -D_MIPS_SZLONG=32 [....]

So either the problem is limited to IRIX 6.4 or something else
is wrong.

						Andrew
						dalke at bioreason.com



From mwh21 at cam.ac.uk  Sun Jul 11 08:09:26 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 11 Jul 1999 13:09:26 +0100
Subject: database
References: <3787c551.1@ganymede.datanet.hu>
Message-ID: <m3btdjl761.fsf@atrus.jesus.cam.ac.uk>

"dzsi" <granadam at mail.datanet.hu> writes:

> hi,
> 
> i am new to python.......
> 
> i have a database (comma separated text), and i would like to get some data
> out of it based on some query. let's say there are seven fields, f1...f7,
> and i am searching for entries in f1, and want to display f1..f7  for the
> field that matches the query. how do i go about doing that?

Well, I'd use awk, myself. However there have been numerous threads
about parsing csv databases in the recent past. http://www.deja.com is
your friend...

If you don't have to worry about commas or newlines in fields, then
this is a simple as using <file object>.readline and string.split.

> also, can i access db databases, or any other common pc type databases with
> perl scripts only on a unix system? without any server configuration,
> assuming python is installed. so far, the comma separated text file is the
> only way i could figure out.

I'm no expert, but I'd say provisionally yes, depending on details.

> thanks,
> 
> adam.


HTH
Michael



From sreekantk at freenet.co.uk  Sun Jul 18 12:10:43 1999
From: sreekantk at freenet.co.uk (sreekant kodela)
Date: Sun, 18 Jul 1999 17:10:43 +0100
Subject: what is this snack in python library
Message-ID: <3791FC83.440316EB@freenet.co.uk>

While I was messing with the python I came across a snack module with
buttons, radio and checkboxes etc. What is this one? Any ideas!

Regards

Sreekant




From Travis.Oliphant at p98.f112.n480.z2.fidonet.org  Thu Jul  1 14:16:00 1999
From: Travis.Oliphant at p98.f112.n480.z2.fidonet.org (Travis Oliphant)
Date: Thu, 01 Jul 1999 19:16:00 +0100
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
Message-ID: <00001653@bossar.com.pl>

From: Travis Oliphant <olipt at mayo.edu>


This has been a very interesting discussion on GUI's.

It struck me that all this talk of a cross-platform GUI might be
preventing a lot of very good GUI's from being used.

So, I looked in to PyGtk and PyGnome (with its Gnome Canvas) and I'm
impressed.  Right now it fits what I want to do very well and it works on
systems with an X server (aren't X servers available for MAC's and
Windows?)  So, I'm happy with it. If you're in the same situation you
might like it too.  I tend to agree that there is plenty of room for more
than one GUI.

Travis Oliphant







From gerrit at humbolt.nl.linux.org  Fri Jul  2 08:16:28 1999
From: gerrit at humbolt.nl.linux.org (Gerrit Holl)
Date: Fri, 2 Jul 1999 14:16:28 +0200
Subject: Python GUIs: Abandoning TkInter and welcoming wxPython?
In-Reply-To: <7li9cq$inp$1@srv38.cas.org>; from lvirden@cas.org on Fri, Jul 02, 1999 at 11:56:10AM +0000
References: <52B9C15E4136A939.DC364288A600C018.AB3BFEAACFF2E162@lp.airnews.net> <slrn7nai1j.8ll.zeitlin@seth.lpthe.jussieu.fr> <64F6457BE0585BB3.1CF2F67BC8A680BD.A19A9FEE23BC9E58@lp.airnews.net> <7li9cq$inp$1@srv38.cas.org>
Message-ID: <19990702141624.A15051@humbolt.nl.linux.org>

On Fri, Jul 02, 1999 at 11:56:10AM +0000, lvirden at cas.org wrote:
> From:	lvirden at cas.org
> Newsgroups: comp.lang.python
> Subject: Re: Python GUIs: Abandoning TkInter and welcoming wxPython?
> Date:	2 Jul 1999 11:56:10 GMT
> To:	python-list at python.org
> 
> According to Mikael Lyngvig <mikael at pobox.com>:
> :Well, TkInter is extremely slow, but not because of the Python part.
> :Even the Tk widget tour, included with Tk, is so slow that you almost
> :think "Please, draw the next character - you can do it." whenever a
> :window with some text needs to be redrawn.
> 
> When I run the Tk test suite  on my SPARC, Tk cranks thru creating and
> deleting windows as fast as the eye can track.
> 
> However, there are reports that the Windows version of Tk, because of its
> use of the native window widgets, is noticably slower than the X
> implementation.  Is that the platform to which you refer?

I think you should also say what kind of processor you have if you're
talking about this:

"Windows NT on a Pentium IV 500 MHZ runs over 10 times faster than a
Linux 386 30 MHZ" (is it?)

regards,
Gerrit.



From da at ski.org  Mon Jul 19 13:54:08 1999
From: da at ski.org (David Ascher)
Date: Mon, 19 Jul 1999 10:54:08 -0700 (Pacific Daylight Time)
Subject: Page for Conference Room/Travel sharin etc.
Message-ID: <Pine.WNT.4.04.9907191050060.294-100000@rigoletto.ski.org>

I've hijacked (with permission) Fredrik's GroupMind page on Michael
Scharf's site to help facilitate folks sharing hotel costs and
transportation costs.  Take it and run!

It's at:


http://starship.python.net/crew/scharf/GroupMind/GroupMind.cgi?req=show&file=ConferenceStuff.gromi


--david ascher




From Ken.Liu at p98.f112.n480.z2.fidonet.org  Fri Jul  2 14:25:37 1999
From: Ken.Liu at p98.f112.n480.z2.fidonet.org (Ken Liu)
Date: Fri, 02 Jul 1999 19:25:37 +0100
Subject: Problem with informixdb and python1.5.2
Message-ID: <000016a3@bossar.com.pl>

From: "Ken Liu" <kyliu at idiomtech.com>

I've been using the informixdb module (version 1.2a1) with python1.5.1 for a
while now with no problems.  Just a few days ago we installed python1.5.2,
and now informixdb won't work any more.  when I attempt to construct a
Connection object, I get an error about "unable to get name of current
working directory."  Here's a log of my interaction with the interpreter

Python 1.5.2 (#8, Jul  1 1999, 20:22:25)  [GCC 2.7.2.3] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import informixdb
>>> con = informixdb.informixdb('acme_online_demo')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "informixdb.py", line 47, in __init__
    self.conn = _informixdb.informixdb(logon)
InformixdbError: Error -922 performing LOGON: Cannot get name of current
working directory.
>>>


Informix's finderr informs me that "-922" means:

"A problem exists with the current working directory on this system or
on another computer system that your application is using. Possibly
your account does not have read access to the directory, or possibly
the file system that contains it is improperly mounted. Exit your
application, and retry. If the problem persists, see your system
administrator for assistance."

which does not make any sense.  Our Informix setup has not changed for the
last few months, and
the informixdb modules continues to work great under python1.5.1.

Does anyone have any suggestions on what might be wrong?




From gregm at iname.com  Wed Jul 21 01:34:01 1999
From: gregm at iname.com (Greg McFarlane)
Date: Wed, 21 Jul 1999 15:34:01 +1000
Subject: where is gfplus?
In-Reply-To: <199907210115.LAA31985@envy.fulcrum.com.au>; from Richard Jones on 21 Jul 1999 at 01:15:15AM
References: <19990721100617.07994@nms.otc.telstra.com.au> <199907210115.LAA31985@envy.fulcrum.com.au>
Message-ID: <19990721153401.12020@nms.otc.telstra.com.au>

On 21 Jul, Richard Jones wrote:
>    DBList.py is a Gtk-based GUI I wrote that does table query and view - pretty 
> simple stuff I wrote while learning Gtk. I'll be writing a Qt version some time 
> soon I suppose since I find this kind of thing handy. DBList.py is available 
> through the gadfly contributions page:
> 
>      http://www.chordate.com/gfcontrib.html
>      http://www.deja.com/getdoc.xp?AN=406461470

Just passing on a piece of hard-earned knowledge:

After saving this file and painfully unpicking all the html tags that
dejanews added to the original post, I discovered that I could get a
text version by clicking on (surprise surprise) "View original Usenet
format" at the bottom of the page, which takes you to:

    http://www.deja.com/getdoc.xp?AN=406461470&fmt=text


>    I've put gfplus 1.0 (I don't seem to have 1.1 anywhere) at:
> 
>      http://redback.spyda.net/~richard/gfplus

Thanks, got it.

-- 
Greg McFarlane:    INMS Telstra Australia (gregm at iname.com)
Today's forecast:  Sunny, with occasional cloudy periods and a chance
		   of precipitation in some areas.



From markusk at bidra241.bbn.hp.com  Mon Jul 26 11:49:03 1999
From: markusk at bidra241.bbn.hp.com (Markus Kohler)
Date: 26 Jul 1999 17:49:03 +0200
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk> <7n5f0c$snu$1@ffx2nh5.news.uu.net> <al81zdxkktf.fsf@sirppi.helsinki.fi> <379C3704.A6F8984D@lemburg.com> <al86737ocwo.fsf@sirppi.helsinki.fi>
Message-ID: <p5n1wjmmxc.fsf@bidra241.bbn.hp.com>

Markus Stenberg <mstenber at cc.Helsinki.FI> writes:

[deletia]
> 
> None of them touch my primary speed problem; the function/method
> invocation 'speed'; C++ function/method invocation seems 200+ times as fast
> as Python's, and it really limits the proper OO nature of things,
> especially while using apply (another 4x slowdown). Deep class hierarchies
> seem no-go, as do several other things.

Yes function invocation is slow with Python. One could argue that this is
because python is more dynamic than C++ for example. But then take a
look at Smalltalk. I found that the calling speed of Visualworks
Smalltalk on HP-UX is just a bit faster than the calling speed of HP's 
C++ compiler. 

My guess is that what calling speed makes slow on Python are all those
different kinds of ways to call the same function. 
Examples are default arguments and named arguments. 
IHMO these features may save some typing but they are not worth
sacrifying performance nor do they help increasing readability of the
code. 

> 
> Although even with 1000x slowdown, my Python runs almost as fast as
> assembly on A500.. hmm. ;-)
> 
> To my regret I've had to muck too much with C++ recently :P
> 

That's exactly what  we all would like to avoid don't we ?

Markus
-- 
Markus Kohler  mailto:markus_kohler at hp.com



From dalke at bioreason.com  Thu Jul  1 13:06:50 1999
From: dalke at bioreason.com (Andrew Dalke)
Date: Thu, 01 Jul 1999 11:06:50 -0600
Subject: 1.5.1 on IRIX
References: <7lg019$fro$1@nnrp1.deja.com>
Message-ID: <377BA02A.482D8FD8@bioreason.com>

rceci at my-deja.com said:
> The libs originally where built/installed  -> /usr/local/lib/python1.5
> re-installed as ${somedir}/lib/irix32, and set PYTHONPATH to
> ${somedir}/lib/irix32.

> Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]

> Any ideas how to "fix" or suppress them?


Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]

The default module search path uses <prefix>/python1.5.


Actually, it's a little more complicated than that. The default
module search path uses <prefix>/python1.5 so you'll need to
move
  ${somedir}/lib/irix32
to
  ${somedir}/lib/irix32/python1.5

then setenv PYTHONHOME ${somedir}/lib/irix32 .

						Andrew
						dalke at acm.org



From emile at fenx.com  Mon Jul 26 10:45:29 1999
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 26 Jul 1999 07:45:29 -0700
Subject: Access dictionary in order?
References: <roy-2507991300370001@mc-as01-p61.med.nyu.edu> <14235.18248.924848.52652@dolphin.mojam.com> <slrn7po8t3.6k.diego@blanka2.blankanet.it>
Message-ID: <Wy_m3.4215$VL2.444209@news.direcpc.com>

Hi Diego,


Diego Dainese <ddainese97x at x18dsi.unive.it> wrote in message
news:slrn7po8t3.6k.diego at blanka2.blankanet.it...
> On Sun, 25 Jul 1999 12:22:03 -0500 (CDT), Skip Montanaro wrote:
> > You demonstrated the best way.  There is no need to save the key
list unless
> > the dictionary is very large though.  Just grab 'em and sort 'em
when you
> > need:
> >
> >     keylist = dict.keys()
> >     keylist.sort()
> >     for key in keylist: do_stuff(dict[key])
> >
>
> But how much time requires the keylist generation?

I tried this test and found that the results varied on the number of
items.  For example, at 100000 entries, the .keys() was generally
faster, but at 50000 entries, it was not faster.  I'm on a PII-266
w/64Mb but I've got lots running so YMMV (your mileage may vary).  Test
on your platform.  Ultimately, you'd need to wrap up a separate keylist
in a class structure to ensure consistency, and that may also impact
speed.

########
import time
d = {}
r = ('spam','eggs','eggs','bacon','eggs','and','spam')
test1start = time.clock()
for i in xrange(100000):
 d[i] = i, 2*i, r
k = d.keys()
test1end = test2start = time.clock()
d = {}
k = []
for i in xrange(50000):
 d[i] = i, 2*i, r
 k.append(i)
test2end = time.clock()
print 'test1: ',test1end - test1start
print 'test2: ',test2end - test2start
########
>
> That is, the keylist is made traversing all the dictionary nodes or is
> cached inside it?

I can't help you with this one...

>
> --
> Diego Dainese
> --
> To reply remove the numbers and the `x' from my address
> --
> Sorry for my bad English!

Not bad at all.

--

Emile van Sebille
emile at fenx.com
-------------------







From graham at sloth.math.uga.edu  Wed Jul 21 11:09:41 1999
From: graham at sloth.math.uga.edu (Graham Matthews)
Date: 21 Jul 1999 15:09:41 GMT
Subject: Stackless & String-processing
References: <000701bed335$8a812740$a72d2399@tim>
Message-ID: <7n4nrl$sfe$1@cronkite.cc.uga.edu>

[Fernando Pereira]
> But from my readings (those mentioned in the thread and others) as well
> as your comments, the whole point of monadic combinators for parsing is
> to interweave the grammatical structure with other computations,
Tim Peters (tim_one at email.msn.com) wrote:
: If it isn't clear yet <wink>, efficiency isn't driving Graham here.  He's
: found a cool new approach, and is having fun with it.

No that's not the case. Efficiency drives all my programs *when it's a
necessity*. My point is that using combinators:

a) is a very nice way of structuring a parser since it is very amenable 
   to change and robust under change. 
	 
b) doesn't have to cost anything in efficiency since you can freely mix
   combinators with other parsing techniques.

c) is a general programming technique useful in a vast range of programs
   outside of just parsing. For example I have used it for type checking
   and writing a GUI. The literature has a lot of other examples of this
   approach.

Moreover the point of combinators is not to "interweave the grammatical
structure with other computations". Combinators do do that, and I think
it's a great benefit. But the point is to:

a) clearly separate sequencing code from other code. In a combinator 
 	 program the precise control flow of the program is clear.
	 l
b) allow programs to program their own sequencing combinators (rather
   than providing some fixed set that the language designer hopes will
   be enough.

Tim Peters (tim_one at email.msn.com) wrote:
: Different strokes.  Most of my parsing life has been concerned with
: artificial languages, and there I've got a different complaint about
: intermixing syntax with semantics:  the grammar is interesting for its own
: sake, and it's a positive burden at times to clutter it up with semantic
: annotations!

I have been using combinators for some time now and never noticed the
"grammar" being cluttered with all these "semantic annotations". What
annotations one requires can easily be separated from the grammar
either by sensible layout, or by the design of clever combinators that
hide semantic annotations away.
: 
Tim Peters (tim_one at email.msn.com) wrote:
: Case in point:  Python's grammar is formally defined by the file
: Grammar/Grammar in the source distribution.  That's the input to a parser
: generator.  At 60 non-noise lines, it's quite possible to digest the whole
: thing, and minor changes in syntax can be effected without touching anything
: else (how minor? exactly what doesn't require changing anything else

Combinators are very robust under change. I have a parser written using
combinators and I often tweak the grammar, without requiring large
changes to other code or the parser itself.

Have you ever written a combinator based parser or are you just 
"theorising"?
 
Tim Peters (tim_one at email.msn.com) wrote:
: I haven't tried using parser combinators, but in reading the papers the
: question I had over and over is whether they scale well in practice, whether
: measured by speed or maintainability (they eventually answered the "speed"
: one: no).

Ah there's the rub -- you haven't written a combinator based parser so you
your comments are just what you *think* might happen if you did.

I don't know where you got the idea that speed is a problem! For some
kinds of parsing I wouldn't use combinators rather I would use
automaton techniques. But for most kinds of parsing they are plenty
fast enough.

Tim Peters (tim_one at email.msn.com) wrote:
: It didn't bode well that each new example seemed to require
: creating a new combinator, and then you've got the implementation of *that*
: to worry about on top of the target grammar and the semantics.  It seemed to
: carry the flavor of programming in an assembly language for "a real parsing
: engine" (as Icon often feels to me in contrast to SNOBOL4 as grammars get
: large).  "You can have any sequencing you want, but you have to build it
: yourself."

Let me guess Tim -- you have read two papers on combinators. One is the
Hutton/Meijer paper on "monadic parser combinators", and the other is
the chapter in the old Clean book. Each new example in these papers
requires a different combinator because each new example illustrates
some new sequencing idea -- eg. determinisation, coninuation, ...
The papers are illustrating the generality of the techniques, and hence
you expect to see a lot of different combinators. In practice however
(something you wouldn't know about right?) you don't use all these
combinators. I have written a parser for a new programming language
and I used 10 combinators (5 of which I re-used in the type checker!).
Yes you do have to build your own combinators (that's the point of the
approach!), but once done you stick them in a library and re-use them.

graham: 
-- 
            The girls is all salty, the boys is all sweet, 
                      the food ain't too shabby 
                    an' they piss in the streets
    In France, way down in France, way on down, way on down in France



From greg.ewing at compaq.com  Thu Jul 22 17:12:24 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 23 Jul 1999 09:12:24 +1200
Subject: Tkinter question
References: <37965ED6.A4238A27@anisci.com>
Message-ID: <37978938.A8F13EF4@compaq.com>

Joel Davis wrote:
> 
> w = int( thecanvas.cget("width") )
> 
> this works fine.. except when I resize the window, this still returns
> the width that the canvas was when it was created.

w = int(thecanvas.winfo_width())

Greg



From roffe at morgoth.uio.no  Wed Jul 21 01:51:42 1999
From: roffe at morgoth.uio.no (Rolf Marvin Bøe Lindgren)
Date: 21 Jul 1999 07:51:42 +0200
Subject: python vs. lisp (whereis quote?)
References: <x88r9m3cyje.fsf@rpppc2.hns.com> <lbzk8rva3to.fsf@morgoth.uio.no> <x88zp0rbd1x.fsf@rpppc2.hns.com> <mtlncbhv94.fsf@astron.berkeley.edu>
Message-ID: <lbzogh635zl.fsf@morgoth.uio.no>

[ Johann Hibschman

| Emacs lisp is a bad ancient ugly dialect of lisp.  I'm waiting for the
| day they port the whole thing over to either Scheme or Common Lisp.
| The whole "quote things then eval later" seems much more complicated
| than it has to be.

edwin was a scheme-based Emacs that came with PC-Scheme from Texas
Instruments, I think.

there's work in progress on a GNU Emacs built on Common Lisp, but I
think it's quite dormant just now.

some might try to do something on top of Guile, which I hope they won't.

-- 
Rolf Lindgren                                        http://www.uio.no/~roffe/
roffe at tag.uio.no



From stidolph at origin.ea.com  Tue Jul  6 11:49:35 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Tue, 6 Jul 1999 15:49:35 GMT
Subject: AttributeError - a dumb question
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B4179301E10FA8@molach.origin.ea.com>

You need to change the assignment of x to include the self paremter:

class F:
    def __init__(self):

        self.x = self.Krylys()

    def KryLys(self):
        l = [1,2,3,4,5]
        return l


Hope this helps,

David Stidolph

-----Original Message-----
From: Johann Spies [mailto:jhspies at futurenet.co.za]
Sent: Tuesday, July 06, 1999 10:22 AM
To: Python-poslys
Subject: AttributeError - a dumb question


Why does the following code produce an attribute error?
class F:
    def __init__(self):

          x = self.Krylys()

    def KryLys(self):
        l = [1,2,3,4,5]
        return l
    

l = F
print dir(l)
print hasattr(l, 'x')
print hasattr(l,'Krypos')
print l.x

What I do not understand is that the output from this code looks like
this:

['KryLys', '__doc__', '__init__', '__module__']
0
0
Traceback (innermost last):
  File "<stdin>", line 14, in ?
AttributeError: x

'Krylys' according to dir(l) is an attribute.  Why does
hasattr(l,'Krypos') report that it is not?

Johann


 --------------------------------------------------------------------------
| Johann Spies                                 Windsorlaan 19              |
| jhspies at futurenet.co.za                3201 Pietermaritzburg             |
| Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
 --------------------------------------------------------------------------

     "Ask, and it shall be given you; seek, and ye shall
      find; knock, and it shall be opened unto you."        
                                  Matthew 7:7 



From m.faassen at vet.uu.nl  Tue Jul 20 04:36:39 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 20 Jul 1999 08:36:39 GMT
Subject: J, you can't pass your own test.
References: <001601bed279$90d4e620$139e2299@tim>
Message-ID: <7n1cen$nku$1@news.accu.uu.nl>

Tim Peters <tim_one at email.msn.com> wrote:
> [Martijn Faassen]
>> Cool, I didn't know I *had* a political standing.

> Of course you do.  Everyone's political standing is continuously updated in
> real time -- see the Political Standing Wizard at python.org.

Oohhhh, aaaahh..such revelations!

>> So what's my political standing?

> Sorry, the Wizard is write-only for everyone except Guido.  Not even bots
> have been able to crack it!

Well, I'd better give up then. I must be under 50 on the Timbot Pythonicness
scale anyway. :)

Regards,

Martijn







From Alan.Daniels at p98.f112.n480.z2.fidonet.org  Thu Jul  1 22:43:41 1999
From: Alan.Daniels at p98.f112.n480.z2.fidonet.org (Alan Daniels)
Date: Fri, 02 Jul 1999 03:43:41 +0100
Subject: Python suitable for a game engine?
Message-ID: <0000164f@bossar.com.pl>

From: daniels at mindspring.com (Alan Daniels)

On Thu, 01 Jul 1999 15:41:27 +0100, the infinitely wise Alex Maranda
(amaranda at nospam.com) spoke forth to us, saying...

[snip...]
>you use confusing terms; I understand what you're saying, but you
>shouldn't use the word 'thread' for anything else than its OS meaning.
>You could say the 'the bulk of processing'. You might very well end up
>using real Python threads; don't forget in this case to release/reaquire
>the global interpreter lock in your C++ callbacks, if you make blocking
>syscalls (unlikely in a game).

It was pretty late when I was writing the first post. :=)
You are correct, "the bulk of processing" is what I meant. NO threads
will be involved. The only justification I can see is better CPU
utilization on an SMP machine, and I don't happen to have one of those
lying around. Hence the engine will run in a single thread.

To keep things moving along, I'm considering dealing with the Python
scripts in a round-robin approach. Every frame, I allow X number of
scripts to handle an event, bumping that number upwards or downwards
to keep a consistent framerate.

>Unlikely. You won't know until you try. To my knowledge UnrealScript is
>eventually translated to C++.

I definitely do NOT want to turn this into a systems level project.
I'm a language junkie, but the internals of compilers don't interest
me much. I understand the reasons that Tim Sweeney made UnrealScript,
but to me it seemed kind of overkill. The world doesn't need another
scripting language.

I have the (VERY!) preliminary code at:
    http://www.mindspring.com/~daniels/engine/index.html
if you're interested, although there isn't much to look at.

-- 
=======================
Alan Daniels
daniels at mindspring.com
daniels at cc.gatech.edu



From tismer at appliedbiometrics.com  Sat Jul 17 08:19:58 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Sat, 17 Jul 1999 12:19:58 GMT
Subject: Snippet: bitcount()
References: <000601bed02a$2235af20$06a22299@tim> <m33dynlhsg.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <379074EE.B6D563F6@appliedbiometrics.com>


Michael Hudson wrote:
> 
> "Tim Peters" <tim_one at email.msn.com> writes:
> > A curious thing in Python is that you can't march over the bits of a long in
> > linear time without resorting to hex() or marshal.dumps.  This really
> > complicates lookup methods, which *can* work in O(B) time straightforwardly
> > in C.
> 
> I remember being quite aggravated by this some time back, when I was
> trying to implement a reasonably fast sqrt routine for Python longs
> that worked even when they were too big to be converted into a float.
> 
> I was using the time honoured method
> 
> x -> (x+a/x)/2
> 
> which works pretty well, but to find an initial guess I thought I'd
> bitshift the number by half it's (bit) length. Finding the length
> seemed to be impossible, without doing something like
> (len(hex(a))-1)/4, which strikes me as silly, not to mention very
> memory wasteful.
> 
> Mini-proposal: len(a) for a long returns the length of the number in
> radix-256.

bitcount2 is not very slow, see here:

def bitlen1(x):
    """number of bytes necessary to store a number"""
    sign = x<0
    x = abs(long(x))
    h = hex(x)
    l = (len(h)-3)*4
    if sign and h[2] >= "8":
        l = l+1
    return (l+7)/8

import marshal

def bitlen2(x):
    """number of bytes necessary to store a number"""
    sign = x<0
    x = long(x)
    b=marshal.dumps(x)
    l = (len(b)-5)*15 / 2
    hi = ord(b[-1])*256 + ord(b[-2])
    l = l + sign -15
    while hi:
        l = l + 1
        hi = hi >> 1
    return max(1, (l+7)/8)

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From phd at emerald.netskate.ru  Wed Jul  7 06:17:55 1999
From: phd at emerald.netskate.ru (Oleg Broytmann)
Date: Wed, 7 Jul 1999 14:17:55 +0400 (MSD)
Subject: Error in python script
In-Reply-To: <3783f651.0@news.smartnet.com.sg>
Message-ID: <Pine.LNX.4.04.9907071415540.30137-100000@emerald.netskate.ru>

On Wed, 7 Jul 1999, John wrote:
> Traceback (innermost last):
>   File "C:\InetPub\scripts\ffratingextract.py", line 39, in ?
>     field_name, field_value = string.split(line, ':')
> ValueError: unpack list of wrong size

   The problem is that you expected 2-tuple (field_name, field_value), but
strin.split returns 1 (there is no ':' in line) or >3 (there are few colonss).
Try this:
   field_name, field_value = string.split(line, ':', 1)

Oleg.
---- 
    Oleg Broytmann        Netskate/Inter.Net.Ru        phd at emerald.netskate.ru
           Programmers don't die, they just GOSUB without RETURN.




From matomira at iname.com  Thu Jul  1 17:37:48 1999
From: matomira at iname.com (Fernando Mato Mira)
Date: Thu, 01 Jul 1999 23:37:48 +0200
Subject: Language extensibility (was: Why is tcl broken?)
References: <1283099377-77267484@hypernet.com> <Pine.SGI.4.05.9906101104460.3354-100000@lemur.magnet.com> <375FDCA7.62FF@mailserver.hursley.ibm.com> <7josm1$bgc$1@Starbase.NeoSoft.COM> <37628EBF.207F7A8F@iname.com> <37682DD3.43923243@compaq.com> <7ka2be$qja$1@nntp6.u.washington.edu> <37776FFA.2D85@mailserver.hursley.ibm.com> <lw7looxx7m.fsf@copernico.parades.rm.cnr.it> <876748mcxg.fsf@home.ivm.de> <37788BB9.ABD@mailserver.hursley.ibm.com> <lw67471fyi.fsf@copernico.parades.rm.cnr.it> <3779ED0F.2C67@mailserver.hursley.ibm.com> <nkjzp1hizvb.fsf@tfeb.org> <377A3383.D453468@iname.com> <377aa228.31380423@news.jpl.nasa.gov> <slrn7nlej3.5j1.wtanksle@dolphin.openprojects.net> <wkzp1g2t2e.fsf@ifi.uio.no> <slrn7nnfs3.19k.wtanksle@dolphin.openprojects.net>
Message-ID: <377BDFAC.EA5CD228@iname.com>

I think Lisp is the dual of Forth.
Is that why I always felt attracted to it (although I never really learned it
yet)?
Who came first doesn't matter then. Maybe we should add an 's' to "The Right
Thing"(TM).

Hey, I was wrong thinking Mind.Forth was sadly misplaced?

Uh.Uh. The `Dark' Side is calling me.. ;-)

[ObPython: Is anybody reading this only on the python _mailing list_? Unless
someone says so,
remove comp.lang.python]





From kistler at fnmail.com  Thu Jul  8 12:08:44 1999
From: kistler at fnmail.com (Per Kistler)
Date: Thu, 08 Jul 1999 18:08:44 +0200
Subject: Two Pythons talking to each other?
References: <199907072203.AAA09601@axil.hvision.nl>
		<378465E9.7762BE9A@fnmail.com> <14212.51977.601241.772478@buffalo.fnal.gov>
Message-ID: <3784CD0C.24437EFC@fnmail.com>

Hi Charles

But does NT and Mac have loopback IP addresses?

Regards, Per.

Charles G Waldman wrote:

> There's no particular reason to use AF_UNIX, even if the two processes
> are running on the same node.  AF_INET with a loopback IP address
> works fine, and is portable (not all platforms support unix-domain
> sockets).

-- 
Per Kistler kistler at fnmail.com / kistler at gmx.net
------------------------------------------------------------



From srenner at lycosmail.com  Fri Jul 16 11:30:14 1999
From: srenner at lycosmail.com (srenner at lycosmail.com)
Date: Fri, 16 Jul 1999 15:30:14 GMT
Subject: J, you can't pass your own test.
References: <378E3481.EE31C97@uswest.net> <19990715175320.B20155@toast.internal>
Message-ID: <7mnj5k$id9$1@nnrp1.deja.com>

The original post, while confused and not giving enough information, and
somewhat resentful in tone, did not deserve the spleen heaped upon it in
the first response, which was unrepresentative of the general tone of
this newsgroup.


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From ralph.gauges at eml.villa-bosch.de  Thu Jul 15 01:32:14 1999
From: ralph.gauges at eml.villa-bosch.de (Ralph Gauges)
Date: Thu, 15 Jul 1999 07:32:14 +0200
Subject: Recursive method
References: <1280168234-10452044@hypernet.com> <378CA422.B2E99CB7@eml.villa-bosch.de> <004901bece0b$46982210$f29b12c2@secret.pythonware.com> <378CACF0.373116E@eml.villa-bosch.de> <008701bece0f$babde130$f29b12c2@secret.pythonware.com>
Message-ID: <378D725E.36E5A36F@eml.villa-bosch.de>

Ahh, now I got it. So I had the misconception, that the code
in my program was executed when I created the instance, but
actualy the code was executed when I ran the file and I just
didn't notice it, because I was just creating one instance
anyway (that obviously didn't do much, since it didn't have
an __init__).
Thanks a lot, I think that made it clear.

Thanks again to everybody for the help,


Ralph



From jhspies at futurenet.co.za  Sat Jul 31 14:14:03 1999
From: jhspies at futurenet.co.za (Johann Spies)
Date: Sat, 31 Jul 1999 18:14:03 GMT
Subject: Printing from python on Windows platform
Message-ID: <Pine.LNX.3.96.990731200754.5616A-100000@Johann>

I am a Linux user writing a program for somebody that will use it on a
Windows platform.  In Linux I use the following to redirect printing to
the printer:

    drukker = sys.stdout
    drukker = os.popen('lpr','w')
    drukker.write(drukstuk)
    sys.stdout.write(drukstuk)

What would the windows equivalent be?

Johann

 --------------------------------------------------------------------------
| Johann Spies                                 Windsorlaan 19              |
| jhspies at futurenet.co.za                3201 Pietermaritzburg             |
| Tel/Faks Nr. +27 331-46-1310		       Suid-Afrika (South Africa)  |
 --------------------------------------------------------------------------

     "Have not I commanded thee? Be strong and of a good 
      courage; be not afraid, neither be thou dismayed; for 
      the LORD thy God is with thee whithersoever thou 
      goest."                        Joshua 1:9 




From nelson at crynwr.com  Mon Jul 19 11:30:29 1999
From: nelson at crynwr.com (Russell Nelson)
Date: 19 Jul 1999 11:30:29 -0400
Subject: I envy newbies
References: <1279991447-21085630@hypernet.com> <2E4327EF0690EAF8.3DAF07256B1C5D39.F38C922CCDEC596E@lp.airnews.net>
Message-ID: <m2yagcbqsq.fsf@desk.crynwr.com>

claird at Starbase.NeoSoft.COM (Cameron Laird) writes:

> Suggestive episode from the history of agricultural mechanization:
> one of the still under-celebrated inventions was the three-point
> hitch, which DRAMATICALLY simplified the quotidian realities of
> hooking diverse implements to ones chosen platform.  (Oh; and
> power-take-off couplings, too.)  I leave implications to language
> theory to the usual advocates of gluing technology.

I prefer the Farmall fasthitch myself.  But sigh, it's a Betamax to
Ford's VHS.

-- 
-russ nelson <rn-sig at crynwr.com>  http://crynwr.com/~nelson
Crynwr supports Open Source(tm) Software| PGPok | Government schools are so
521 Pleasant Valley Rd. | +1 315 268 1925 voice | bad that any rank amateur
Potsdam, NY 13676-3213  | +1 315 268 9201 FAX   | can outdo them. Homeschool!



From jsight at mindspring.com  Sat Jul 24 16:45:03 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Sat, 24 Jul 1999 16:45:03 -0400
Subject: Fork, Threading, and Select madness.  Please help!
References: <1279424742-23511862@hypernet.com>
Message-ID: <379A25CF.6935D6A9@mindspring.com>

Thanks for your thoughtful reply.  As it turned out, threading wasn't
nearly as bad a solution as I thought it might be<0.5wink>.  I just
downloaded Python 1.52 and recompiled with "./configure --with-threads"
and everything worked out nicely on the FreeBSD box that my webserver
uses.  

Of course, doing this on my linux box resulted in something that causes
my kernel to blow up in flames a few hours after running the Python
interpretter with any threaded program.  But that's ok, at least it
works with my ISP.

Anyway, threads seem to be helping me make things faster, so right now
that is my solution.  The select idea is probably theoretically better,
but would be a big challenge to implement due to the architecture of the
code.<sigh>

Gordon McMillan wrote:
> 
> Jesse D. Sightler wrote:
> 
> > Ok, I have been working on a website which is now available at
> > http://www.biddin.com/ that performs a search of multiple auction
> > sites all at once, and then returns the results.  The problem is
> > that it is very slow, and it appears that the primary bottleneck is
> > in waiting on the search engines at the various sites to begin
> > reading data.
> >
> > Well, this makes sense because the code to implement the search is
> > currently pure sequential with no threading or non-blocking IO
> > whatsoever.
> >
> > What I'm getting around to asking (slowly<g>) is, what is the best
> > way to implement an algorithm that can retrieve the data from ALL 3
> > sites simultaneously, so that the code is never blocked up waiting
> > on one particular bad site?
> 
> In theory, select (multiplexing) is the best solution, especially
> since you are just gathering data. But this might well mean trashing
> all of your existing code. I'd recommend using asyncore if you're
> doing this from scratch.
> 
> If you're using one of the higher level protocol (urllib,
> httplib) modules, a quick glance (I'm no expert on these guys) says
> you'll have to thread or use subprocesses. Multiplexing usually means
> turning your logic inside out (you're no longer driven by meaningful
> exchanges, but by drips and drabs of data appearing in network
> buffers). So the transformation from blocking sockets to multiplexed
> non-blocking sockets is major surgery.
> 
> - Gordon



From neelk at neelk  Sat Jul 10 03:53:46 1999
From: neelk at neelk (neelk at neelk)
Date: Sat, 10 Jul 99 07:53:46 GMT
Subject: Signature 0.1
Message-ID: <mt2.0-16623-931593226@news.informatik.uni-bonn.de>

Signature.py is a module that makes reflecting on the function call
signatures of callable objects a lot easier. A common question in
Python is "How do I ask the interpreter what arguments a function
takes?" and while the information is obviously available, I haven't
found any simple, high-level interfaces to it.

That's why I wrote Signature -- it's useful enough that it's made it
into my personal library of frequently-used utilities.

You can find it at:

  http://www.sff.net/people/neelk/free-software/Signature.py

Caveats/TODO:

o I don't know if it will work on JPython -- I exploited a lot of
  undocumented object attributes to figure out what the call signature
  is, and I have no idea if the same data is available there.

o I haven't figured out how to look into C builtins yet -- anyone
  who can teach me how to do this from within Python will earn my
  undying gratitude. :)

Example of use:

    >>> def foo(x, y, z=-1.0, *args, **kw):
    ...     return (x+y)**z
    ...
    >>> f = Signature(foo)

    >>> print 'ordinary arglist:', f.ordinary_args()
    ordinary arglist: ('x', 'y', 'z')

    >>> print 'special_args:', f.special_args()
    special_args: {'keyword': 'kw', 'positional': 'args'}

    >>> print 'full arglist:', f.full_arglist()
    full arglist: ['x', 'y', 'z', 'args', 'kw']

    >>> print 'defaults:', f.defaults()
    defaults: {'z': -1.0}

    >>> print 'signature:', str(f)
    signature: foo(x, y, z=-1.0, *args, **kw)


Neel

<P><A HREF="http://www.sff.net/people/neelk/free-software/Signature.py">
Signature 0.1</A> - for reflecting on the function call signatures of
callable objects.  (09-Jul-99)

-- 
----------- comp.lang.python.announce (moderated) ----------
Article Submission Address:  python-announce at python.org
Python Language Home Page:   http://www.python.org/
Python Quick Help Index:     http://www.python.org/Help.html
------------------------------------------------------------



From wtanksle at dolphin.openprojects.net  Fri Jul 30 16:56:52 1999
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Fri, 30 Jul 1999 20:56:52 GMT
Subject: Include statements with python scripting
References: <7nsu3t$kui$1@nw001t.infi.net>
Message-ID: <slrn7q44ck.8kv.wtanksle@dolphin.openprojects.net>

On Fri, 30 Jul 1999 15:24:38 -0400, Sean Robertson wrote:
>How do you do an include statement in python?  I know that in Cold Fusion it
>would be "<cfinclude template="blue_header_noad.html">" and in an shtml file
>it would be "#include http://data.pilotonline.com/connect2/blue_header.html"
>but I need to know how this is done in Python scripting.  This script is
>being embedded in a template for a classifieds page.  I have something like
>this so far:

><html><head><title><!--#var name--> Search Results</title></head>

This isn't Python -- I'm sure you're trying to ask about Zope.  Remarkable
product, written in Python (lovely plumage).  But not Python.  Try
visiting www.zope.org for more info.

>    <!--#include http://data.pilotonline.com/connect2/blue_header.html-->

I wasn't aware of a #include in Zope; I've been using #var.  Generally
speaking, it doesn't make any sense to include an entire HTML page inside
another, so I think you might possibly be confused.

>Sean Robertson

-- 
-William "Billy" Tanksley



From alex at somewhere.round.here  Tue Jul 20 16:18:35 1999
From: alex at somewhere.round.here (Alex)
Date: 20 Jul 1999 16:18:35 -0400
Subject: python vs. lisp (whereis quote?)
References: <x88r9m3cyje.fsf@rpppc2.hns.com> <37946F3B.ECBFE6D6@mail.danbbs.dk>
Message-ID: <etdn1wrgjms.fsf@m2-225-11.mit.edu>

Perhaps you can define a do-nothing function that you can then change
behind the module's back:

hook.py:
------------------------------------------------------------------------
def hook(*args):pass

def actual_program(*args):
    return apply(hook,args)
------------------------------------------------------------------------

Python 1.5.1 (#1, Jan 31 1999, 18:55:21)  [GCC 2.8.1] on irix6
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import hook
>>> def tst():print 'This is my hook function executing'
... 
>>> hook.actual_program()
>>> hook.hook=tst
>>> hook.actual_program()
This is my hook function executing
>>> 

Alex.



From donn at u.washington.edu  Thu Jul 29 23:38:55 1999
From: donn at u.washington.edu (Donn Cave)
Date: 30 Jul 1999 03:38:55 GMT
Subject: os.name
References: <37A0604C.322192FE@phoenix-edv.netzservice.de>
Message-ID: <7nr6of$nc6$1@nntp6.u.washington.edu>

Holger Jannsen <holger at phoenix-edv.netzservice.de> writes:
| someone working on a more specific version of os.name or better:
| something inside of 'sys.builtin_module_names'.
|
| It's because I need to differ between Win95, Win98 and WinNT....

Not clear, missing parts of speech, however, os.uname()...  Windows
behavior unknown.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From Gordon.McMillan at p98.f112.n480.z2.fidonet.org  Sat Jul  3 09:09:27 1999
From: Gordon.McMillan at p98.f112.n480.z2.fidonet.org (Gordon McMillan)
Date: Sat, 03 Jul 1999 14:09:27 +0100
Subject: calldll & PyArg_ParseTuple strings
Message-ID: <000016b8@bossar.com.pl>

From: "Gordon McMillan" <gmcm at hypernet.com>

Scott McClellan writes:
> 
> I'm attempting to use Sam Rushing's calldll module to call a C
> function in a dll, with the following prototype:
> 
> long __declspec(dllexport)func(long *x, long *y, long *z)
> 
> I expect the function to give certain values back in x, y, and z,
> but so far have been unable to make that happen. Let's say that I
> have three long integers in Python that I initialize as follows:
> 
> x=y=z=0L

A Python int is a C long. You won't find a Python long in C.

> My question is, in the following call from Python,
> 
> call_foreign_function(pFunc, 'lll', 'l', (<?>,<?>,<?>))
> 
> what should the tuple (<?>,<?>,<?>) contain? 

params = calldll.membuf(12)
params.write('\000'*12)
px = params.address()
call_foreign_function(pFunc, 'lll', 'l', (px, px+4, px+8))
rslt = params.read(0,12)
x,y,z = struct.unpack('lll', rslt)

Actually, Sam's got a slicker way of doing it, but I don't understand 
how it works, so I'll stick with the dumb-but-effective method.

> If I pass (x,y,z), the
> Python interpreter (Idle in my case) crashes, and this is
> understandable, since it results in the contents of x, y, and z
> being used as addresses to write to.  I've tried
> (id(x),id(y),id(z)), since id() is supposed to return the address of
> a Python object - Python didn't crash as before, but nothing was
> copied into x, y, and z, either.

id returns the address as a string. You wouldn't want this anyway. 
The Python int 0 is not a slot - it IS the one and only int object 
with value 0. If the foreign function were to return a 3 into the 
Python int (which it wouldn't - it would clobber something else in 
the struct) you would then have

>> 3 + 0
6

(The most commonly used ints are static singletons - try 
 sys.getrefcount(0))

- Gordon




From sragsdale at my-deja.com  Wed Jul 14 13:24:11 1999
From: sragsdale at my-deja.com (sragsdale at my-deja.com)
Date: Wed, 14 Jul 1999 17:24:11 GMT
Subject: Identical callback problems
Message-ID: <7mih3a$mcm$1@nnrp1.deja.com>

I've got a GUI that is trying to execute the same callback from two
different events.  One is working and the other isn't, even though it's
the same callback.

Basically, it's a program that opens a modal dialog box so that the user
can enter information.  When the callback is bound to a button, the
dialog box opens without problems.  When the callback is bound to a
double-click, the dialog box STILL opens but I get an error saying
"TclError: grab failed: window not viewable".  It's not an invocation
problem since the creation of a DlogBox isn't dependent on anything.
Any idea how I can satisfy tk so that it doesn't complain about this?

To see the bug, run the program below.  Double-click on the left side
(error) and then click the button on the right (no error).

=================================================================
from Tkinter import *

# this just pops up a modal dialog box with a button in it
class DlogBox(Toplevel):
    def __init__(self):
        Toplevel.__init__(self)
        f1 = Frame(self)
        f1.pack(anchor=N, expand=YES, fill=X)
        Button(self, text='OK', command=self.destroy).pack(anchor=S)
        self.grab_set()         # make myself modal:
        self.focus_set()        # mouse grab, keyboard focus, wait...
        self.wait_window()      # till destroy; else returns to caller
now

# The main app
class TestApp(Frame):
    def __init__(self,parent=None):
        Frame.__init__(self,parent,relief=GROOVE,borderwidth=4)
        self.pack(expand=YES, fill=BOTH)
        items = Listbox(self)
        items.pack(side=LEFT, expand=YES, fill=BOTH)

        # bind the button to the callback
        button = Button(self,text="Foo",command=self.callback)
        button.pack(side=RIGHT,expand=YES, fill=BOTH)

        # bind a double-click on the Listbox to the callback
        items.bind('<Double-1>', self.callback)

    # the callback that exposes the DlogBox
    def callback(self,event=None):
        print "Callback called with",event
        DlogBox()

if __name__ == '__main__':
    TestApp().mainloop()


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.



From nick.bower at ssec.wisc.edu  Sat Jul 10 23:26:13 1999
From: nick.bower at ssec.wisc.edu (Nick Bower)
Date: Sat, 10 Jul 1999 22:26:13 -0500
Subject: Simple Tk8 Problems
Message-ID: <7m9to9$6qg$85@rks1.urz.tu-dresden.de>

Hi,

I'm currently having 2 simple, but I think fundamental, problems with my
understanding of Tk 8.

1.  I can't control the size of the Tk root window.  I've tried putting
".width=xxx" statements everywhere in the following snippet, but nothing
seems to work.  (The FileList is a ListBox next to a scrollbar).  Is
there something I'm missing?.

class FileList(Frame):
  def __init__(self,parent):
    Frame.__init__(self,parent)
    self.filelist=Listbox(self)
    self.scroll=Scrollbar(self)
    self.filelist.config(yscrollcommand=self.scroll.set,relief=SUNKEN)
    self.filelist.pack(side=LEFT,expand=YES,fill=BOTH)
    self.scroll.config(command=self.filelist.yview,relief=SUNKEN)
    self.scroll.pack(side=RIGHT,expand=NO,fill=Y)

<etc>

if __name__ == '__main__':
  toplevel = Tk()
  afilelist=FileList(toplevel)
  afilelist.pack(expand=YES,fill=BOTH)
  afilelist.mainloop()


2. Also, the Listbox in the above code is scrolling all over the place
tracking the mouse movement as it runs across the top of it.  I don't
even need to touch the scrollbar and it seems to track mouse movement so
somehow it's been bound to the mouse as well as the scroll bar.  Weird.


Thanks in advance for help (again)  :)

nick


--

Nick Bower
Space Science and Engineering Center
University of Wisconsin, Madison, USA
Phone: (608) 265 8007
Email: nick.bower at ssec.wisc.edu
Web: http://arm1.ssec.wisc.edu/~nickb






From cbarker at jps.net  Fri Jul 23 17:56:51 1999
From: cbarker at jps.net (Chris Barker)
Date: Fri, 23 Jul 1999 21:56:51 GMT
Subject: Xemacs Python Mode.
References: <199907231949.PAA07568@python.org> <3798CF69.1FC04C23@jps.net> <3798DE2D.C008286D@aw.sgi.com>
Message-ID: <3798E523.7AF3DE05@jps.net>

Thanks Gary,

This gets me close enough to be functional, but isn't quite there. The
problem is that if there are spaces in there, it won't reset them to
tabs, which is what I would like. For the most part, it won't be an
issue, but I'd just like it to be totally consistent.

What I really want is for Python mode to have an option like
"use-only-tabs" or "use-only-spaces". In fact all you would really need
to do is have py-indent settable to an arbitrary string, rather that an
integer. Then folks could set it to be four spaces, or a tab, or two
tabs, or whatever.

Quite frankly, I'm confused as to why anyone would think it's a good
idea to mix tabs and spaces at all, but that's what we're stuck with.



Gary Herron wrote:
> Chris Barker wrote:
> > Does anyone know if I can set emacs mode to indent with ONLY tabs? (and
> > if so, how?) While I'm at it I'd like the tabs to show up as four spaces
> > while in Python mode also.
> >
> 
> I have my emacs setup to use one tab per indent level, and to display
> tabs as four spaces.  I believe that this is all controlled by the
> following three settings found in the customization section of my .emacs
> file.
> 
> (custom-set-variables
>  '(py-indent-offset 4)
>  '(tab-width 4)
>  '(py-smart-indentation nil)
>  ...
> )


-- 
Christopher Barker,
Ph.D.                                                           
cbarker at jps.net                      ---           ---           ---
http://www.jps.net/cbarker          -----@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Water Resources Engineering       ------   @    ------   @   ------   @
Coastal and Fluvial Hydrodynamics -------      ---------     --------    
------------------------------------------------------------------------
------------------------------------------------------------------------



From rowland at iRenaissance.com  Mon Jul 26 14:07:53 1999
From: rowland at iRenaissance.com (Rowland Smith)
Date: Mon, 26 Jul 1999 14:07:53 -0400
Subject: Python and SlickEdit
Message-ID: <379CA3F9.A746217E@iRenaissance.com>

Does anyone know if there are any Python plug-ins for Visual SlickEdit
and where they are?

-- 
Regards,
Rowland

:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
iRenaissance, Inc.
	-- Tools For Building Online Communities

	Rowland D. Smith
	Senior Software Developer
	http://www.iRenaissance.com
	http://www.CalendarCentral.com
	919.743.0109
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::



From skip at mojam.com  Thu Jul 29 16:37:54 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 29 Jul 1999 15:37:54 -0500 (CDT)
Subject: [Tutor] I want to learn to hack
In-Reply-To: <slrn7q1dau.17t.esr@snark.thyrsus.com>
References: <19990721154610.B3663@optiplex.palga.uucp>
	<Pine.WNT.4.04.9907211133150.294-100000@rigoletto.ski.org>
	<14231.11726.916541.492153@nobot.cnri.reston.va.us>
	<slrn7q1dau.17t.esr@snark.thyrsus.com>
Message-ID: <14240.47957.840790.39395@dolphin.mojam.com>

    Eric> Some time back, I published a page called "How To Become A Hacker"
    Eric> <URL:http://earthspace.net/~esr/faqs/hacker-howto.html">.  As a
    Eric> result, I get a couple dozen semi-illiterate email notes a week
    Eric> from teenagers wanting me to teach them everything I know.
    Eric> Instantly.

	...

    Eric> Earlier this year I changed the document to recommend Python as a
    Eric> first language.  I still think this was a good idea, but I see it
    Eric> has had an unfortunate side-effect.

Eric, just think how much more they'll be hacking in Python instead of
MSVC++, VB or Perl! ;-)

Skip Montanaro	| http://www.mojam.com/
skip at mojam.com  | http://www.musi-cal.com/~skip/
847-475-3758



From cfelling at iae.nl  Sun Jul 11 13:46:18 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 11 Jul 1999 19:46:18 +0200
Subject: Simple Qu: Changing directory
References: <3787D36A.815B7C29@ssec.wisc.edu> <3787D8DA.3774AA7D@ssec.wisc.edu> <7m8p7t$o7s@dfw-ixnews10.ix.netcom.com> <3787FB51.12FE0048@ssec.wisc.edu>
Message-ID: <7mal9a$hjc$1@vvs.superst.iae.nl>

Nick Bower <nick.bower at ssec.wisc.edu> wrote:
> yeah, that's what i originally thought.  sure enough there were some updates in the
> os documentation, but not os.chdir.  it's something that would be really helpfull

It's there allright, see the docs on the posix module:)

-- 
groetjes, carel



From ralph.gauges at eml.villa-bosch.de  Wed Jul 14 06:43:52 1999
From: ralph.gauges at eml.villa-bosch.de (Ralph Gauges)
Date: Wed, 14 Jul 1999 12:43:52 +0200
Subject: Recursive method
References: <378C214F.3CF9CDD3@eml.villa-bosch.de> <199907140901.LAA22305@axil.hvision.nl>
Message-ID: <378C69E8.76BC3B3E@eml.villa-bosch.de>

Hans Nowak wrote:
> 
> Hmm... I didn't know that... I always assumed that programs for
> regular Python and JPython would be compatible...?
> 

Actualy they are. I can import Python modules I wrote that
use 'self'. But as soon, as I try to use 'self' in the
script that is executed by JPython, I get an error message.
Maybe I am to dumb, but I realy don't know what's wrong?!
Especialy since I couldn't find any hints in the JPython
documentation stating that there was no 'self'. (Also looks
expectantly at JPython guru(s))



From fredrik at pythonware.com  Wed Jul 21 05:45:05 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Jul 1999 09:45:05 GMT
Subject: Tk: Temporarily change cursor?
References: <3793AADF.5EDBAFBC@compaq.com> <19990720145034.04299@nms.otc.telstra.com.au> <3794EC17.64249239@compaq.com>
Message-ID: <014b01bed35d$b703a040$f29b12c2@secret.pythonware.com>

Greg Ewing <greg.ewing at compaq.com> wrote:
> In an old news article I was referred to,
> Fredrik Lundh <fredrik at pythonware.com> wrote:
> 
> > if you change the cursor for a toplevel (or the root window),
> > that cursor will be used for all widgets that doesn't specify their own
> > cursors. 
> 
> I tried things like that, but they didn't work.
> Changing the root window's cursor doesn't affect
> any other toplevels, even if they don't define their
> own cursor. Also, when the pointer was over a subwidget
> of a toplevel, changing the toplevel's cursor and then
> doing an update didn't work reliably if the mouse
> buton was down at the time!

did you try the BusyManager class?  it's derived
from stuff we're using in commercial apps, and
seems to work just fine on all Tkinter platforms.

extending it to deal with multiple toplevels is
fairly trivial...

</F>




From Sam.Schulenburg at p98.f112.n480.z2.fidonet.org  Sat Jul  3 02:32:56 1999
From: Sam.Schulenburg at p98.f112.n480.z2.fidonet.org (Sam Schulenburg)
Date: Sat, 03 Jul 1999 07:32:56 +0100
Subject: "Is theit an interest in a  SCSI Hard drive test environment under Pyth
Message-ID: <000016b6@bossar.com.pl>

From: "Sam Schulenburg" <samschul at pacbell.net>

I have been developing a SCSI diagnostic test environment under Python. It
has the following features:

1) Works under any Windows 32 os that supports ASPI
A(dvanced)(S)ystem(P)rogramming(I)nterface:
         a) Win95/98/NT

2) Provides for rapid development of drive test scripts using Python

3) Supports low level drive test requirements:
         a) Buffer management
         b)Separate read/write buffers for each TargetID
         c)Precision timing loop on Pentium systems using Windows High
performance counters(10 micro-second resolution)
         d)Can issue any SCSI command with io6(),io10(),and io12() functions
         e)Support for up to 15 drives

4)Full integrated help system
         a) Access to python HTML documentation using IDLE 0.5
         b) Access to ASPIpython help for both the ASPIPy.dll and ASPI.py
functions using autoduck help authoring system
         c) All functions documented using Python documentation strings
         d) Ability to add user defined documentation to help systems

5)Works under the following environments:
          a)  Win32 console mode (For high speed and full buffer editing
capabilities)
          b) Python IDLE environment (Recommended environment to use)

If their is sufficient interest in this environment I would like to have
someone volunteer a server where I could locate the installation and source
code for this test platform. I plan on having a releasable beta within two
days, and will be providing updates as they occur.

Sam Schulenburg
samschul at pacbell.net





From claird at starbase.neosoft.com  Wed Jul 21 19:38:31 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Jul 1999 23:38:31 GMT
Subject: Python Performance
References: <3794C642.AB6A3E5D@freenet.co.uk>
Message-ID: <0EAA6437F8435A58.6F61DA1F1117B965.CAF9F552BC9AB119@lp.airnews.net>

In article <3794C642.AB6A3E5D at freenet.co.uk>,
sreekant kodela  <sreekantk at freenet.co.uk> wrote:
			.
		[other questions already
		adequately addressed]
			.
			.
>intial welcome / logon screens etc and images. PostgreSQL being slow
>doesn't make things easier either.
			.
			.
			.
Is PostgreSQL slow?  Are you using a modern
version?  I have plenty of evidence that
PostgreSQL is adequately zippy in many
circumstances.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From olli at rhein-zeitung.de  Thu Jul 22 08:34:33 1999
From: olli at rhein-zeitung.de (olli at rhein-zeitung.de)
Date: 22 Jul 1999 12:34:33 GMT
Subject: int(long(-sys.maxint-1)) fails on Linux
References: <3796DC4F.224795DA@vision.arc.nasa.gov>
Message-ID: <7n734p$fll$1@news.rhein-zeitung.DE>

Hi,

I compiled the packages on my Mandrake 6.0 system and I am looking into this
myself. ;-)) Concerning the release I have to admin that I am a little bit
late, but a lot of work had to be done. But here is a road map for all
interested:

- I asked for a mailing list on starship python and received it. It is called
  python4linux and should be used to discuss issues concerning the packages we
  make
- I wrote a announcement and request for participation, cause I like to move
  the whole project to a group effort (more work, new girl friend, new side
  projects -> less time to maintain the thing myself ;-)
- I finished a new spec file that can build RedHat6, Mandrake6 and RedHat5
  packages with just one extra parameter during compilation

All this (after a weekend of testing is ready and available on Monday. The
Announcement will be posted tomorrow to c.l.p.a and the packages will be
uploaded on Monday. The web page has to be reworked, but this is left to a
later date.

Best regards,

    Oliver

Chad Netzer <chad at vision.arc.nasa.gov> wrote:
> Compiling python 1.5.2 on Linux Mandrake 6.0,  the "make test" step fails
> because it gets an OverflowError during test_types.py on the following line:

> int(long(-sys.maxint-1))

> test_long.py fails also for the same reason.  I've tried this on another Linux
> setup, using the default system python 1.5.1 compiled with just the straight
> gcc (not egcs or pgcc) and got the same result.  Is this just the expected
> behavior on Linux (in which case maybe the tests should be changed to reflect
> this)  Can I ignore it?  (This doesn't overflow on my SGI, and I assume the
> Pentium uses 2's-complement.  Perhaps this is a glibc issue?)  I should look
> at the code, but I'm sure others must have come across this...

> BTW, as a learning experience, I'm building all of Oliver Andrich's Python
> packages for the Mandrake 6.0 architecture, and will make the results available
> when I'm finished, unless Oliver beats me to the punch and obviates the need.
> Our application depends on getting python installed easily, so I took the initiative.

> Chad Netzer
> chad at vision.arc.nasa.gov




-- 
Oliver Andrich, RZ-Online, Schlossstrasse Str. 42, D-56068 Koblenz
Telefon: +49-261-39210-40 Fax: -33 |       http://rhein-zeitung.de 
                                   |            http://andrich.net
GMC d- s++:++ a- C++ UL++++ P-- L+++ E--- W++ N+ w-- O- M-- 
V- PS++ PE Y+ PGP+ t 5+ X R+ tv+ b++ DI++ D- G e h- r z+



From Alan.Daniels at p98.f112.n480.z2.fidonet.org  Sat Jul  3 00:49:51 1999
From: Alan.Daniels at p98.f112.n480.z2.fidonet.org (Alan Daniels)
Date: Sat, 03 Jul 1999 05:49:51 +0100
Subject: Python suitable for a game engine?
Message-ID: <000016a1@bossar.com.pl>

From: daniels at mindspring.com (Alan Daniels)

On Fri, 2 Jul 1999 10:39:35 +0100, the infinitely wise Nick
(nick at NOSPAMvideosystem.co.uk) spoke forth to us, saying...

[snip...]
>The language doesn't necessarily need to support it. You could implement the
>threaded system yourself in Lua, you then have more control over what goes
>on. Proper threads are relatively slow in Windoze (I believe) and would be a
>nightmare to debug? If Lua wasn't fast enough you could write a scheduler
>with priority assessment in C/C++ and attached Lua script callbacks to this.
>Personally I wouldn't dream of useing proper threads for game logic/AI I
>would write a basic scheduler, it wouldn't be difficult. Grim Fandango uses
>Lua as its scripting language and I think they did something along these
>lines. I think Balders Gate is another game that is rumoured to use it.

The only time I would consider adding threads is well after I had
everything else working (and I mean EVERYTHING), and wanted to
experiment around with it a bit. So far, my current plan is to use a
round-robin approach to keep the script processing from varying the
framerate too much, even if that means I can't use the interpreter but
have to translate the bytecodes myself.

The information on Lua is very useful, by the way. Thanks for this.

-- 
=======================
Alan Daniels
daniels at mindspring.com
daniels at cc.gatech.edu



From optional  Sun Jul 25 08:47:46 1999
From: optional (Brian)
Date: Sun, 25 Jul 1999 12:47:46 GMT
Subject: simplebeginnerquestion(**args)
Message-ID: <379afff7.131877652@news.mindspring.com>

I'm trying to figure out how to use the (**args) thing. I can manage
to pass an arbitrary # of arguments to a function with (*stuff) as a
tuple, but if I use (**stuff)  to do it like this:

myfunc(**stuff):
 for i in stuff.keys():
  print i 

myfunc(1,2,3,4)

the interpreter raises an error saying it expected..0 arguments and
got..4	if I use (*stuff) (without .keys()...) it works. 
Shouldn't (**stuff) cause a dictionary filled with myfunc(stuff here)
to happen and the 'print i' print up the keys?
What I don't get is why does it expect..0 arguments?Why doesn't it
realize that there is a ** there?





From fredrik at pythonware.com  Wed Jul 14 10:24:17 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 14 Jul 1999 14:24:17 GMT
Subject: How to trace Python/sys.settrace() ?
References: <19990714160145.A21443@sz-sb.de>
Message-ID: <002101bece04$8ffaec50$f29b12c2@secret.pythonware.com>

Andreas Jung <ajung at sz-sb.de> wrote:
> In some situations it is neccessary to be able to do a complete
> trace of Python script (similiar to truss/strace) - that means one
> want to know all called functions (function names) and functions parameters. 
> 
> sys.settrace() seems to be a candidate for building such a tracing functionality.
> However I am not sure if this is the right way. I'd like to ask the Python gurus
> and Python hackers here  about a possible solution. Maybe there is already some usable
> code available.

http://www.deja.com/getdoc.xp?AN=400537551
could be what you're looking for.

</F>




From m.faassen at vet.uu.nl  Tue Jul 20 13:06:01 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 20 Jul 1999 17:06:01 GMT
Subject: Python suitability questions
References: <0CCE781E44F2C872.F528377A62B41DCA.73CAA63AD6E79314@lp.airnews.net>
Message-ID: <7n2a9p$trq$1@news.accu.uu.nl>

Jason Maskell <jmaskell at saskpublishers.sk.ca> wrote:
[snip]
> and the scripting language has to be comfortable being completely
> embedded in my C++ framework,

There are various ways to integrate Python with C or C++ code. You can
hand code things of course, and I've heard Python's extending/embedding
API is quite nice. You can also use various tools for it. SWIG is very
handy, having toyed with it myself (www.swig.org).

There are many people on this newsgroup who have far more experience with
extending and embedding Python than I do, though, so let's wait and see
what they'll say :)

> and having really no system access at all.

Look into the 'rexec' and 'Bastion' modules in the Python library reference.
 
> I know that Python can handle the OO stuff,
> but is there any kind of internal
> object messaging, or do I roll my own?

What do you mean by an 'internal object messaging'? What I do know that if
you decide to roll your own you'll be fine with Python -- it's very dynamic
that way.

> How flexible can I lay out my source,
> is it as stiff as say, C?

This depends on what you mean. Python uses whitespace to indicate block
structure, so you could see that as 'strict' (though not actually more
strict than begin and end markers; just different). Once you're used to
it it's very handy; you start to wonder why other languages have these
annoying { and } characters (or whatever they may be using).

Due to Python's dynamic nature, it's very easy to change the layout of
your code, though. In C++ I get lost in endless designing of classes,
in Python I just write down what I want. :)

Of course I don't want to say solid design of classes isn't necessary; it's
still good to do.
 
Good luck!

Regards,

Martijn




From mwh21 at cam.ac.uk  Tue Jul 13 02:58:35 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 13 Jul 1999 07:58:35 +0100
Subject: ANN: Stackless Python 0.3
References: <378A40AD.B4FDB79F@appliedbiometrics.com>
Message-ID: <m37lo5kpd0.fsf@atrus.jesus.cam.ac.uk>

Christian Tismer <tismer at appliedbiometrics.com> writes:

> ANNOUNCING:
> 
>                Stackless Python 0.3
>            A Python Implementation Which
>              Does Not Use The C Stack
> 
> What is it?
> A plugin-replacement for core Python.
> It should run any program which runs under Python 1.5.2 .
> But it does not need space on the C stack.
> 

Very nice, but

Python 1.5.42a2 (#5, Jul 13 1999, 07:42:11)  [GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> l=range(10)
>>> l.reverse()
>>> apply(l.sort,())
Traceback (innermost last):
  File "<stdin>", line 1, in ?
SystemError: error return without exception set
>>> l
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

apply isn't totally hosed; it works for Python functions (I think) but
not for builtins.

This causes lots of make test to fail, though not test_builtin,
helpfully.

I've had a poke through the source, but don't have the time just now
to delve into its inner workings. I'll trace it with gdb sometime.

HTH
Michael

PS: To get it to compile I had to remove "static" from in front of 
    PyFrame_FiniChainFunc * PyFrame_FiniChain = NULL;
in Objects/frameobject.c.



From kperacles at geocities.com  Sat Jul 10 13:19:10 1999
From: kperacles at geocities.com (Karl Putland)
Date: Sat, 10 Jul 1999 12:19:10 -0500
Subject: Win32 installer / "standalone" packager
References: <1280576364-81429328@hypernet.com>
Message-ID: <378781fe.0@news.dbtech.net>

Installer install breaks ;-)

The traceback goes away too fast, but I think it says...

Traceback... yada-yada-yada
File "launch.py", line 12, in ?
    support = os.path.join(installdir, "support")
NameError: installdir


from launch.py

welcome="""\
Welcome to the Win32 Installer!\n\n"""

print welcome

idif = getinstalldir()

support = os.path.join(installdir, "support")    # this is line 12


Karl Putland
kperacles at geocities.com


Gordon McMillan <gmcm at hypernet.com> wrote in message news:1280576364-81429328 at hypernet.com...
> This is to announce that I have made a major upgrade to my Win32 
> Installer and that it is now available on python.org's contrib site.
> 
> What's New
>  Updated for 1.5.2 
>  Automatic handles Python/TK apps
>  Support for console / console-less apps
>  New one-step generation (no config file needed)
>  Detects binary (dll dependencies) even without dumpbin.exe 
> installed.
> 
> Where is it?
>  ftp://ftp.python.org/pub/python/contrib/System/Installer_r_01.exe
> 
> How's it licensed?
>  As loosely as possible.
> 
> More Info:
> 
> Installer is a set of utilities for creating installation
> packages and quasi-standalone python executables for Win32. No
> compiler is required.
> 
> Win32Installer steals from freeze, Christian Tismer's sqfreeze,
> and Greg Stein's imputils.py. It will detect TK usage, and 
> automatically install the necessary files. It can run console
> apps or pure GUIs.
> 
> It can create a single-file self-extracting installation package.
> 
> It can also create quasi-standalone executables. These are not
> single file; but everything exists in one directory tree with
> no dependencies outside that tree, and no registry dependencies.
> So uninstalling consists of deleting the directory tree, and it
> shouldn't get messed up by other versions of python or TCL/TK
> installed on the user's machine.
> 
> It also creates the smallest possible distributions - Python
> modules are distributed as compressed archives (no source) and
> imported directly from the archive. For example, Python/Lib
> (excluding the test package) fits in a 475K Python_lib.pyz
> file.
> 
> Win32Installer works with Python 1.5.2. It will not work with
> Python 1.5.1 or earlier, mostly because of the fact that in
> 1.5.2, zlib.pyd is statically linked to zlib, while in 1.5.1
> it was dynamically linked.
> 
> Files: Installer_r_01.exe
> Contact: gmcm at hypernet.com
> 
> Copyright McMillan Enterprises; licensed on the "who cares" 
> model.
> 
> - Gordon
> 
> 

From hank at prosysplus.com  Sat Jul  3 23:28:34 1999
From: hank at prosysplus.com (Hank Fay)
Date: Sat, 3 Jul 1999 23:28:34 -0400
Subject: wddx wrapper?
References: <%0Qe3.4034$lr3.164499@news2.jacksonville.net> <930863445.26638@www.remarq.com>
Message-ID: <2QAf3.544$jS5.16374@news2.jacksonville.net>

Understood.  I'd like a python wrapper so I can escape th evil empire when
necessary. <s>

--
http://www.prosysplus.com
<Voynow>; Sasha <anonymous at web.remarq.com> wrote in message
news:930863445.26638 at www.remarq.com...
>
> If you're in a Windows environment you can use Python's COM
> support to to use the COM object they offer.
>
> Sasha Voynow
>
> sasha at ofoto.com
>
>
>
> **** Posted from RemarQ - http://www.remarq.com - Discussions Start Here
(tm) ****