From sabren at manifestation.com  Sat Jul  1 01:11:58 2000
From: sabren at manifestation.com (Michal Wallace)
Date: Fri, 30 Jun 2000 19:11:58 -0400 (EDT)
Subject: MD5 - Passwords with python
In-Reply-To: <2vegj8.ia5.ln@odin.hmf.de>
Message-ID: <Pine.BSF.4.10.10006301910440.27864-100000@chi.pair.com>

On Thu, 29 Jun 2000 holger at fuessler-net.de wrote:

> Unfortunately, the standard MD5-module doesn't know about the
> structure of the password-string so I can't encode a password with
> it. I found a perl module which accomplishs my task and so I thought
> that there is also a python module. I thought redhat uses a lot of
> python stuff, so there should be a routine to generate suitable
> passwords. I'v already posted in comp.redhat.security. When I find
> the details, I'll post them here (if I can translate them to
> python).


http://www.sabren.com/code/python/crypt/


Cheers,

- Michal
------------------------------------------------------------------------
www.manifestation.com  www.sabren.com  www.linkwatcher.com  www.zike.net
------------------------------------------------------------------------




From dsavitsk at e-coli.net  Sat Jul  1 01:28:02 2000
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 30 Jun 2000 23:28:02 GMT
Subject: Web server executing Python scripts
References: <4.3.1.2.20000630151218.021a7858@phear.dementian.com>
Message-ID: <62a75.28853$dF.1122319@news1.rdc1.il.home.com>

just to give a different point of view :-)
while iis may be insecure, ms is on to something with the asp thing (or did
a good job copying php/coldfusion/jsp etc).  if you are on an intranet and
security is less of an issue, asp is really easy and runs a bit faster than
cgi.  there are also asp plug-ins for apache (www.halcyionsoft.com and
www.chilisoft.com) that are sometimes free, sometimes not depending on what
you are doing etc. i'm not sure if these will allow PythonScript, but it
might be worth looking into.
doug

"Steven D. Arnold" <stevena at neosynapse.net> wrote in message
news:4.3.1.2.20000630151218.021a7858 at phear.dementian.com...
> At 08:24 PM 6/30/2000 +0200, Pieter Claerhout wrote:
>
> >What do you think would be the best idea?
> >- Use whatever webbrowser with CGI-scripting
>
> I would use apache and either CGI or mod-python, more likely the latter.
>
>
> >- Use MS IIS with ASP (and Python)
>
> IIS is quite probably the most insecure server on the Internet.  In
> addition, I don't like the closed nature of the system, I don't trust MS
to
> introduce the features I want/need (where I can introduce them myself in
> Apache), and I don't want to depend on MS's support services to solve
> problems that only they can fix because only they have the source.
>
>
> >- Use Zope to build the whole system?
>
> I don't know enough about Zope to give an answer from personal experience
> here.  I've heard many good things about Zope; I've also heard that the
> docs are poor and the learning curve steep.  I have faith in the speed and
> flexibility of Apache, where (to me) Zope is an unknown in these regards;
> are there any benchmarks?
>
>
> steve
>
>
> --
> Steven D. Arnold           Que quiero sera         stevena at neosynapse.net
> "If you have  built castles in the air, your work  need not be lost; that
> is where they should be.  Now put the foundations under them!" -- Thoreau
>
>




From db3l at fitlinxx.com  Sat Jul  1 01:30:53 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 30 Jun 2000 19:30:53 -0400
Subject: Following a file, or cloning tail
References: <Pine.LNX.4.21.0006292102260.25464-100000@calcifer.valdyas.org> <8F62AF73Egmcmhypernetcom@199.171.54.194> <36D5F4DFCA9BCA46.B5AE15F2F19F393C.28F9B766F10B5E23@lp.airnews.net> <8F62C20E2gmcmhypernetcom@199.171.54.154>
Message-ID: <uk8f63fwy.fsf@ctwd0143.fitlinxx.com>

gmcm at hypernet.com (Gordon McMillan) writes:

> Well, os.popen works fine between console processes on the NT family of 
> Windows. However, Windows has problems with 16 bit DOS apps and Win95 has 
> an MS acknowledged problem with any of this stuff. The win32 extensions fix 
> most of these problems, (I think Bill Tutt even added a workaround for the 
> Win95 bug), and I believe that 1.6 will use the win32 versions if they are 
> installed (at least, this was discussed seriously at some point).

If they do, then I hope that the extensions get augmented to return
the result code of the child process on close(), since that's really
the only thing holding me up from using them (for now I just use NT
and ensure I have a console allocated to my process).

I guess I'm just glad that my requirements to use Windows stops with
NT, since I also think Python might be a non-starter for a lot of my
administration apps if popen() didn't work reliably.  Of course, MS is
far more to blame for this than Python :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From rmartin at objectmentor.com  Sat Jul  1 01:56:52 2000
From: rmartin at objectmentor.com (Robert C. Martin)
Date: Fri, 30 Jun 2000 18:56:52 -0500
Subject: OmPyUnit -- A python unit test framework.
References: <slnttnejjev80@news.supernews.com> <8jhr2t$j3k$1@pollux.ip-plus.net>
Message-ID: <slqd02qqop349@news.supernews.com>

"J?rgen Hermann" <jhe at webde-ag.de> wrote in message
news:8jhr2t$j3k$1 at pollux.ip-plus.net...
>
> "Robert C. Martin" <rmartin at objectmentor.com> schrieb im Newsbeitrag
> news:slnttnejjev80 at news.supernews.com...
> > I have made our OmPyUnit unit test framework for python available in the
> > 'downloads' section of http://www.objectmentor.com.  This framework is
> very
> > similar in appearance and operation to JUnit.
>
> Is it in any way related to PyUnit (http://pyunit.sourceforge.net/),
besides
> the common JUnit heritage?

No, that's a completely separate implementation.  I haven't put OmPyUnit up
on sourceforge yet.


--

Robert C. Martin    | "Uncle Bob"              | Training Courses:
Object Mentor Inc.  | rmartin at objectmentor.com | OOD, Patterns, C++, Java,
PO Box 85           | Tel: (800) 338-6716      | Extreme Programming.
Grayslake IL 60030  | Fax: (847) 548-6853      | http://www.objectmentor.com

"One of the great commandments of science is:
    'Mistrust arguments from authority.'" -- Carl Sagan




From mwh21 at cam.ac.uk  Sat Jul  1 02:16:25 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Jul 2000 01:16:25 +0100
Subject: Problem with execv
References: <8jg6j3$o84$1@nnrp1.deja.com> <blbt0jzv8h.fsf@bitdiddle.concentric.net>
Message-ID: <m3n1k27lie.fsf@atrus.jesus.cam.ac.uk>

Jeremy Hylton <jeremy at beopen.com> writes:

> The correct calling convention is to supply a tuple or list as the
> second argument to execv.  Example:
> 
> >>> os.execv('/bin/ls', ('.',))
> [lists contents of current directory]

ITYM

>>> os.execv('/bin/ls',('ls','.'))

or some such.  There's also

>>> os.execlp('ls','ls','.')

and many other variants...

Cheers,
M.

-- 
7. It is easier to write an incorrect program than understand a
   correct one.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From Norman_Shelley-RRDN60 at email.sps.mot.com  Sat Jul  1 01:54:37 2000
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Fri, 30 Jun 2000 16:54:37 -0700
Subject: UserDict and UserList issue (bug?_
Message-ID: <395D333D.9240593C@email.sps.mot.com>

in UserDict and UserList there is one or more places where a call to
self's __class__() method is made, e.g.

def __getslice__(self, i, j):
  i = max(i, 0); j = max(j, 0)
  # This class and its subclasses need more arguments then 0
  userlist = self.__class__()
  ....


This blows causes major problems when either of these two classes are
subclassed and the class which subclasses them requires creation
arguments.  I think maybe a copy of self instead of a __class__() may be
a more robust solution.






From Norman_Shelley-RRDN60 at email.sps.mot.com  Sat Jul  1 01:57:34 2000
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Fri, 30 Jun 2000 16:57:34 -0700
Subject: ?Combination of a Dictionary and a List?
Message-ID: <395D33EE.75F1E36D@email.sps.mot.com>

Sometime  in the past I wrote a simple combination of a dictionary and a
list, where __getitem__() would first look in a hidden dictionary and
then if not found it would try looking in a hidden list.  I've seen this
type of code out there somewhere also but I can not find my code or
anyone elses. I guess I haven't figure our the "right" search criteria
on www.python.org.  Any pointers?



From gmcm at hypernet.com  Sat Jul  1 02:23:57 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 01 Jul 2000 00:23:57 GMT
Subject: Passing variable-length keyworded arguments list to base class constructor?
References: <962352947.103057778@news.libertysurf.fr> <8F635367Cgmcmhypernetcom@199.171.54.154> <962379073.1230449645@news.libertysurf.fr>
Message-ID: <8F63C1ABDgmcmhypernetcom@199.171.54.154>

Jerome Quelin wrote:

>gmcm at hypernet.com (Gordon McMillan) wrote:
>>>But how to pass this dictionnary to the parent class constructor?
>>class F(Frame):
>>  def __init__(self, *args, **kwargs):
>>    apply(Frame.__init__, (self,)+args, kwargs)
>>Generally you have to muck with the kwargs dict before passing it to the 
>>base class __init__.

>That is exactly what I want to know: what this mucking consists of?

Depends on what args the base class expects. If the base class expects 
**kwargs, then you can have anything in there.

But if the base class were defined:
class Frame:
  def __init__(self, x=0, y=0):
    ....

And the kwargs you pass it has {'snoo':33,...}, it's going to barf.

But maybe you've been playing with the interpreter, and have discovered all 
this by now...

- Gordon


From urner at alumni.princeton.edu  Sat Jul  1 02:29:48 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Fri, 30 Jun 2000 17:29:48 -0700
Subject: PowerPoint slide presentation (Math + Python)
Message-ID: <uneqlsoqo59b38g19k8lus9gmb30obf9as@4ax.com>

Greetings educators, Python programmers --

For those of you with PowerPoint and/or the requisite browser 
plug-ins, I've tossed a slide show on the web at:

ftp://ftp.teleport.com/pub/users/pdx4d/images/synergeo.ppt

This is my standard 'Beyond Flatland' curriculum, wherein
I mix spatial geometry, virtual worlds, computer programming
(these days Python), and synergetics.

Given this is just an outline of a talk, there's not a 
lot of discussion (few words).  To present this material
effectively, you'd need to study the background readings
e.g. http://www.inetarena.com/~pdx4d/ocn/cp4e.html (see
'Numeracy + Computer Literacy' series).

Uses some of those special effects PowerPoint is all about.  
Apologies if this is distracting, but the business culture 
in which I operate kinda likes these bells and whistles.

Feedback welcome.

Kirby
4D Solutions
Portland, Oregon
http://www.teleport.com/~pdx4d/


PS:  use mouse clicks to get text bullets to appear, 
and to move from slide to slide (11 slides in all).

PPS:  also notified educators at
http://mathforum.com/epigone/math.teaching.technology
of this free download




From gmcm at hypernet.com  Sat Jul  1 02:34:39 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 01 Jul 2000 00:34:39 GMT
Subject: command-line history in win
References: <395c304e.1224185@news.student.lu.se> <8jhm03$qnm$1@nnrp1.deja.com> <8jig08$dk7$1@nnrp1.deja.com>
Message-ID: <8F63D360Egmcmhypernetcom@199.171.54.154>

david_ullrich at my-deja.com wrote:


>  David Currie <davec999 at my-deja.com> wrote:

>> It works for me (Python 1.5.2)...

>     Keen. Maybe I'm the only one who doesn't know:
>_What_ works for you, exactly?

Perhaps David is on NT, where it just works (not only for DOS cmds, but if 
you start Python, up-arrow will take you thru the history for your last 
Python session in that DOS box). This is all built in to NT.

Sorry, can't help with Win9x.

- Gordon


From mwh21 at cam.ac.uk  Sat Jul  1 02:36:07 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Jul 2000 01:36:07 +0100
Subject: Python MP3 Player
References: <395CE50A.3BA16777@bioeng.ucsd.edu>
Message-ID: <m3ituq7klk.fsf@atrus.jesus.cam.ac.uk>

Curtis Jensen <cjensen at bioeng.ucsd.edu> writes:

> Does anyone know of a Python MP3 Player?

I've cobbled together some Python that you can use to control xmms if
that's what you want.  Anyone interested?  I can stick it on my
starship page if anyone is...

Cheers,
M.

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                              http://www.naggum.no/profile.html


From mwh21 at cam.ac.uk  Sat Jul  1 02:37:42 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Jul 2000 01:37:42 +0100
Subject: backslash in regular expression problem
References: <394A55BA.69269793@wolfson.co.uk> <02de01bfd826$cf05bd80$6401a8c0@home> <8jivfr$qc2$1@nnrp1.deja.com>
Message-ID: <m3g0pu7kix.fsf@atrus.jesus.cam.ac.uk>

Keith Murphy <kpmurphy at my-deja.com> writes:

> what if instead of having '\\stuff' you have a str = '\\stuff'  ?
> 
> i know that you can print r'\\stuff' but you can't print rstr...
> 
> how do you use the raw with a variable?

You don't.  You do know about re.escape?  Asking that question
suggests you might not...

Cheers,
M.

-- 
  I saw `cout' being shifted "Hello world" times to the left and
  stopped right there.                                -- Steve Gonedes


From cbbrowne at news.hex.net  Sat Jul  1 02:53:39 2000
From: cbbrowne at news.hex.net (Christopher Browne)
Date: Sat, 01 Jul 2000 00:53:39 GMT
Subject: ?Does Python have sorting via "Dictionary Order"?
References: <395A92D0.65A205B1@email.sps.mot.com>
Message-ID: <nib75.251583$MB.4396963@news6.giganews.com>

Centuries ago, Nostradamus foresaw a time when Norman Shelley would say:
>Does Python have sorting via "Dictionary Order"?  This would be a very
>handy addition as an option to list's sort() function.

Feel free to avail yourself of the documentation...

"The keys() method of a dictionary object returns a list of all the
keys used in the dictionary, in random order (if you want it sorted,
just apply the sort() method to the list of keys)."
-- 
cbbrowne at ntlug.org - <http://www.ntlug.org/~cbbrowne/scripting.html>
"People need to quit pretending they can invent THE interface and walk
away from it, like some Deist fantasy." -- Michael Peck


From thomas at xs4all.nl  Sat Jul  1 03:06:50 2000
From: thomas at xs4all.nl (Thomas Wouters)
Date: 1 Jul 2000 01:06:50 GMT
Subject: UserDict and UserList issue (bug?_
References: <395D333D.9240593C@email.sps.mot.com>
Message-ID: <slrn8lqh0n.gqv.thomas@centurion.xs4all.nl>

On Fri, 30 Jun 2000 16:54:37 -0700, Norman Shelley
<Norman_Shelley-RRDN60 at email.sps.mot.com> wrote:
>in UserDict and UserList there is one or more places where a call to
>self's __class__() method is made, e.g.
>
>def __getslice__(self, i, j):
>  i = max(i, 0); j = max(j, 0)
>  # This class and its subclasses need more arguments then 0
>  userlist = self.__class__()
>  ....

>This blows causes major problems when either of these two classes are
>subclassed and the class which subclasses them requires creation
>arguments.  I think maybe a copy of self instead of a __class__() may be
>a more robust solution.

And how would you get a copy of self without creating a new instance of
self.__class__ ? You usually create a copy of a list by doing:

    mycopy = mylist[:]

Which calls __getslice__, which is exactly what you are complaining about
:-)

If you want to subclass UserList, either make sure you provide defaults for
all arguments to the initializer, or make sure you override all places where
self.__class__() may be called without your required arguments. #1 is
probably the best (and most portable) solution, and not just because of the
above issue: providing defaults for all initializer arguments allows your
object to be pickled, for instance ;)

But I dont know why you need those arguments, so YMMV.

Regards,
	Thomas.


From MarkH at ActiveState.com  Sat Jul  1 03:33:55 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 01 Jul 2000 01:33:55 GMT
Subject: Reference Counting Woes in Python 1.5.2! Help!
References: <wr675.13560$227.259336@nnrp1.uunet.ca>
Message-ID: <7Ub75.2486$Tb7.18780@news-server.bigpond.net.au>

> I am using Visual C++ 6. It has debug libraries that are catching this
> problem printing an error message and quitting. When I go to release
version
> of code, I get access violations and then the code quits. Alas, what I'd
> like is to see the access violation location in source code, but the
darn
> debug library gets in the way!

It shouldnt be doing this!  An access violation when running under debug
build should definately break into the debugger, and definately does for
me.

You may be seeing Py_FatalError() being called.  This just prints a
message and calls "exit()", which does silently die.

Post Python 1.5, this routine as a block:

#ifdef _DEBUG
    DebugBreak()
#endif

You could try adding that to Python 1.5.2.  Then, in debug builds, a
Py_FatalError() will also break into the debugger.

> Has anyone else debugged pesky reference count problems in Visual C++
with
> Python 1.5.x and come out on the winning end? Your help or tips or
> suggestions on how to debug this sort of thing would be most welcome!

I find reference underflows easier to work with generally than leaks.  But
you definately need the exception in the VC debugger.  The symptoms are
generally that other code will die when trying to use the object with the
bad reference.  This can usually tell you at least the _type_ of the
object in question.

Another good hint - when re-running the same program multiple times under
VC, you almost always get identical addresses allocates between multiple
runs.  For example, if one run shows that object at address "0x123456" has
the ref count bug, restarting the program and rerunning under the debugger
will almost certainly give the same pointer the same address.  This makes
it easier to "back-step" through the offending code in multiple debugging
sessions. (eg, next step after determining the above may be to set a
break-point whenever "*(int *)0x123456" (ie, the objects reference-count)
changes.

Mark.




From MarkH at ActiveState.com  Sat Jul  1 03:35:53 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 01 Jul 2000 01:35:53 GMT
Subject: Following a file, or cloning tail
References: <Pine.LNX.4.21.0006292102260.25464-100000@calcifer.valdyas.org> <8F62AF73Egmcmhypernetcom@199.171.54.194> <36D5F4DFCA9BCA46.B5AE15F2F19F393C.28F9B766F10B5E23@lp.airnews.net> <8F62C20E2gmcmhypernetcom@199.171.54.154> <uk8f63fwy.fsf@ctwd0143.fitlinxx.com>
Message-ID: <ZVb75.2491$Tb7.18785@news-server.bigpond.net.au>

> I guess I'm just glad that my requirements to use Windows stops with
> NT, since I also think Python might be a non-starter for a lot of my
> administration apps if popen() didn't work reliably.  Of course, MS is
> far more to blame for this than Python :-)

And Bill Tutt is to be eternally thanked for Python solving it :-)

[And Gordon is correct - later builds have reliable popen on Win9x too,
also thanks to Bill!]

Mark.





From dnotspam-jen at personic.com  Sat Jul  1 03:39:21 2000
From: dnotspam-jen at personic.com (donotspam-jen@personic.com)
Date: Fri, 30 Jun 2000 18:39:21 -0700
Subject: Communicating between two computers
Message-ID: <395d48b3$0$21842@wodc7nh0.news.uu.net>

FYI, I'm doing all of this on Windows NT, in case that makes a difference
...

Is there a good way to get a script on one computer to talk to a script on
another computer?  The process I'm thinking of is pretty simple ... I want a
script on a Web server to be able to send a command to a build server to
start a build process.  I'd also like the Web server to be able to query the
build servers to find out their status (so that two builds do not run at
once on the same machine, for example).

I've read a little about SOAP, but I think that my needs are probably a lot
simpler than that.  The easiest way to accomplish this is to setup a Web
server on each of the build machines, too, I guess ... but I was hoping that
I could just have the scripts that run on the build servers run as NT
services.

Thanks!

Jen






From thomas at xs4all.nl  Sat Jul  1 03:32:42 2000
From: thomas at xs4all.nl (Thomas Wouters)
Date: 1 Jul 2000 01:32:42 GMT
Subject: Controlling an external program with Python
References: <slrn8lj6pa.l52.thomas@centurion.xs4all.nl> <3958C648.1D76@rd.francetelecom.fr> <nz855.1583$iN5.265679@ptah.visi.com> <slrn8l87is.dpi.keating@mars.rcode.com.au> <kuzoobue0g.fsf@lasipalatsi.fi> <8j7vc3$6b6q$1@nntp6.u.washington.edu> <39588741.FB7EB5FE@hursley.ibm.com> <3958A7AE.7DAA@rd.francetelecom.fr> <3958C134.7AAEF37B@hursley.ibm.com> <8jc28s$8eb$0@216.39.151.169> <8jdf8j$8tm4$1@nntp6.u.washington.edu>
Message-ID: <slrn8lqih7.gr6.thomas@centurion.xs4all.nl>

On 28 Jun 2000 18:13:39 GMT, Donn Cave <donn at u.washington.edu> wrote:
>The MIT Kerberos 5 pty library I mentioned uses a getpty function that
>opens the master pseudoty and returns its file descriptor and the name
>of its slave terminal device.  Where it uses openpty() underneath, it
>doesn't supply any termios settings and closes the slave file descriptor
>immediately.  It doesn't use forkpty() for anything.  My reflex reaction
>is that's the right level of support for a C module:  export all possible
>semi-standard low level interfaces, and put convenience routines in a
>Python module.

I disagree, and I think some of the Python developers do, too, looking at
the way posixmodule is laid out (and the fact that the SGI module is
candidate for removal in the next Python release.)

Exporting all lowlevel calls to Python and letting the pty lib take care of
determining which are appropriate creates a horrible mess. I did it with the
new pty lib, that switches between os.openpty, sgi._getpty and the original
BSD 'try and get a tty' methods, and it blows chunks, especially if you take
into account that one method might be available but fail, but another might
succeed.

If there was one 'openpty' call in the os module, which might be emulated
from other kinds of calls in C, for whatever platform it is convenient, most
of what remains are #ifdefs (of which the posixmodule, for one, already has
quite a bunch) and the choice between os.openpty() and opening /dev/pty*
manually.

My reason for taking openpty, which returns an opened master/slave fd pair,
over something a little more sgi-like (open master, return master_fd + name
of slave tty) is very simple, and very selfish: I work mostly with BSDI
servers, FreeBSD and Linux workstations. All support openpty. And the fact
that Linux has it makes me believe it's the most used method of getting a
pty.

Both methods of returning a pty can emulate the other just as well; in fact,
the new pty lib does just that ;) it has the old master_open() and
slave_open(), and it has openpty().

I agree, though, that keeping forkpty() in the posixmodule might not be
worth it. It's badly documented, for instance on wether or not the new child
is a session leader or not, so it might end up much easier just using a
python version that utilizes pty.openpty(), and *is* properly documented.

>Some of these interfaces may be appropriate for the posix module, but
>I doubt it!  Aside from the narrow-minded question of whether any of it
>is covered by POSIX 1003.1, from the user side it's inevitable that on
>some platform that does build the posix module, the pty functions won't
>be supported.  They'll need another module that can be optionally
>supported, like termios, fcntl et al.

I'm not sure if 'POSIX defines it' is a criterium for adding a function to
the posix module, though it might be. In that case, it should be factored
out to a _pty module, together with things such as ttyname() possibly. But
I'm not sure if it's out of place in the posix module: Adding it to
posixmodule adds it to the os module, which means that other platforms can
write their own, entirely seperate implementation, without the need to add
more #ifdefs.

>The more we can steal, the better.  There's too much platform specific
>folklore in here to write a good C implementation from scratch with
>man pages.

Agreed. I do like the idea of an external, extensive pty-emulation library
or such, but I have two very simple problems with that: I do not have the
time to write it, and I do not have a need for it ;) openpty() is exactly
what I need. I'd be happy to move sgi._getpty into openpty() (if the sgi
module is indeed going to be deprecated/removed) and add other variants, as
a bunch of #ifdefs, but I don't feel like supporting non-UNIX operating
systems, and I don't feel like building an external project for this ;)

Besides, it's late, I shouldn't be posting at this hour :-P

Note-my-crafty-and-elegant-way-of-avoiding-Pythons-version-number<wink>-ly
	y'rs, Thomas.



From sridhar at mediaone.net  Sat Jul  1 03:49:06 2000
From: sridhar at mediaone.net (Sridhar Bidigalu)
Date: Fri, 30 Jun 2000 21:49:06 -0400
Subject: Subscribe
Message-ID: <395D4E11.EBAEA908@mediaone.net>


-------------- next part --------------
A non-text attachment was scrubbed...
Name: sridhar.vcf
Type: text/x-vcard
Size: 371 bytes
Desc: Card for Sridhar Bidigalu
Url : http://mail.python.org/pipermail/python-list/attachments/20000630/6ec58387/attachment.vcf 

From db3l at fitlinxx.com  Sat Jul  1 04:10:48 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 30 Jun 2000 22:10:48 -0400
Subject: Communicating between two computers
References: <395d48b3$0$21842@wodc7nh0.news.uu.net>
Message-ID: <uem5e38if.fsf@ctwd0143.fitlinxx.com>

"donotspam-jen at personic.com" <dnotspam-jen at personic.com> writes:

> I've read a little about SOAP, but I think that my needs are probably a lot
> simpler than that.  The easiest way to accomplish this is to setup a Web
> server on each of the build machines, too, I guess ... but I was hoping that
> I could just have the scripts that run on the build servers run as NT
> services.

Well, there's probably no end to the number of protocols and methods
by which you could get the machines to talk, but given that it's
Python invoking Python, you might take a peek at Fredrik Lundh's
XML-RPC library.  XML-RPC was sort of the precursor to SOAP.  Similar
in concept, but it's dirt simple to get two Python processes
communicating (calling arbitrary methods with any Python objects and
returning any Python objects).

See http://www.pythonware.com/products/xmlrpc/index.htm.

I imagine you could run a server utilizing this within a service using
the PythonService module included in the Pythonwin extensions without
too many problems.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From dgoodger at bigfoot.com  Sat Jul  1 04:38:51 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Fri, 30 Jun 2000 22:38:51 -0400
Subject: ?Combination of a Dictionary and a List?
References: <395D33EE.75F1E36D@email.sps.mot.com>
Message-ID: <B582D1FB.5A1E%dgoodger@bigfoot.com>

on 2000-06-30 19:57, Norman Shelley
(Norman_Shelley-RRDN60 at email.sps.mot.com) wrote:

> Sometime  in the past I wrote a simple combination of a dictionary and a
> list, where __getitem__() would first look in a hidden dictionary and
> then if not found it would try looking in a hidden list.  I've seen this
> type of code out there somewhere also but I can not find my code or
> anyone elses. I guess I haven't figure our the "right" search criteria
> on www.python.org.  Any pointers?
> 

Many people have had such a need at some point (myself included; see my
signature below), but each with slightly different requirements. Do you want
a list that acts like a dictionary, or a dictionary that acts like a list,
and in what way?

Look for the c.l.p thread "indexed dictionaries?" of 2000-04-09, see
"seqdict" on Parnassus, and examine the "UserDict" and "UserList" modules in
the standard Python library.

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From claird at starbase.neosoft.com  Sat Jul  1 05:17:34 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 30 Jun 2000 22:17:34 -0500
Subject: Web server executing Python scripts
References: <2B1262E83448D211AE4B00A0C9D61B030130125E@msgeuro1.creo.be>
Message-ID: <80EB8C0C818F320D.2A60B0748B6BCCB3.6A1C28BCC752E3D8@lp.airnews.net>

In article <2B1262E83448D211AE4B00A0C9D61B030130125E at msgeuro1.creo.be>,
Pieter Claerhout  <PClaerhout at CREO.BE> wrote:
>Hello all,
>
>I'm working on a web based system, which consist of a whole
>bunch of Python scripts, which connect to a bunch of machines
>on a network (either using ethernet or dailup networking) and
>execute some to code to for example get things from the
>windows registry of those machines.
>
>What do you think would be the best idea?
>- Use whatever webbrowser with CGI-scripting
>- Use MS IIS with ASP (and Python)
>- Use Zope to build the whole system?
>
>It should also connect to a database running on the host.
>
>
>Pieter
>

Others have already demonstrated that there's a large
... situational, if not subjective, component to the
decision about which Web server to use.

Your first paragraph interests me.  Do you realize
that Python can retrieve registry information directly
through your LAN, without a need for installation of
"some to code" on the reporting hosts?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From samwun at yahoo.com  Sat Jul  1 05:24:33 2000
From: samwun at yahoo.com (Sam Wun)
Date: Sat, 01 Jul 2000 13:24:33 +1000
Subject: how to make popen secure?
Message-ID: <395D6471.79CF0908@yahoo.com>

I want to use popen execute an external C program. The C program is
written to read kmem network traffic form the kernel memory. Therefore
the C program is given a special permission to read the kmem device.
How do I make popen secure? or how to make kmem secure even thought it
is read by the C program?

thanks
Sam.



From olczyk at interaccess.com  Sat Jul  1 06:40:10 2000
From: olczyk at interaccess.com (Thaddeus L. Olczyk)
Date: Sat, 01 Jul 2000 04:40:10 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl>
Message-ID: <395d728e.117049078@nntp.interaccess.com>

On 28 Jun 2000 06:02:09 GMT, thomas at xs4all.nl (Thomas Wouters) wrote:

>On Wed, 28 Jun 2000 04:05:21 GMT, Courageous <jkraska1 at san.rr.com> wrote:
>
>>> On top of that, very likely that if SUN sues they win.
>
>>Ridiculous.
>
>Very much so. But it wouldn't be suprising just the same, given the American
>patenting/registration practices and the average american courthouse ;-)
>
>Overseas-ly y'rs,
>	Thomas
Virtually all SUN has to do is pull some of the demo programs out of
Microsofts manual. Write the equivalent code in python, perl, C++,
pascal, and a few other languages. Compare them side by side. A fairly
sophisticated person can see that it's resemblance to Java. Look at
the law judges are sophisticated enough. Given that Microshit is being
sued by SUN for their abuse of Java, I don't think it will be hard at
all for  SUN to convince a judge that all C# is, is a variant of Java
with tons of proprietary crap glued on and the label changed.


From frankc at colconsulting.com  Sat Jul  1 06:56:30 2000
From: frankc at colconsulting.com (Frank V. Castellucci)
Date: Sat, 01 Jul 2000 00:56:30 -0400
Subject: Class design: accessing "private" members
References: <962352843.1667731268@news.libertysurf.fr> <8jhuvn01vs0@news2.newsguy.com> <962367319.379353174@news.libertysurf.fr>
Message-ID: <395D79FE.4E4FDBCE@colconsulting.com>

Jerome Quelin wrote:
> 
> "Alex Martelli" <alex at magenta.com> wrote:
> >By calling the accessors, rather than accessing the data
> >directly, you ensure that if somebody overrides the
> >accessor functions in a derived class you'll be calling
> >the derived-class versions of those functions.  If that is
> >what you want (one direction along which you want to let
> >inheritors of your class extend or tweak behaviour), then
> >it's probably worth the overhead you mention.
> The class isn't to be derived any further. It's a class that I use in an
> application, and that may be imported in some other applications, but not to be
> inherited.
> 
> Then, is it better/cleaner to access private members with accessors or not?
> Is it a matter of style? Or are there hidden caveheats using (or not using)
> accessors? It _seems_ cleaner to use accessors, but man, it's quite awkward.
> 
> Jerome
> --
> jerome.quelin at insalien.org

As you really have no way to "prohibit" further derivation (unless I
missed the "final" keyword in the reference), using your accessors and
praying that anyone over-riding the accessor/mutator methods does the
proper thing would be my direction.

-- 
Frank V. Castellucci
http://corelinux.sourceforge.net
OOA/OOD/C++ Standards and Guidelines for Linux
http://PythPat.sourceforge.net
Pythons Pattern Package


From jkraska1 at san.rr.com  Sat Jul  1 07:10:50 2000
From: jkraska1 at san.rr.com (Courageous)
Date: Sat, 01 Jul 2000 05:10:50 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com>
Message-ID: <395D7E14.B18EC77D@san.rr.com>

> Virtually all SUN has to do is pull some of the demo programs out of
> Microsofts manual. Write the equivalent code in python, perl, C++,
> pascal, and a few other languages. Compare them side by side. A fairly
> sophisticated person can see that it's resemblance to Java. Look at
> the law judges are sophisticated enough. Given that Microshit is being
> sued by SUN for their abuse of Java, I don't think it will be hard at
> all for  SUN to convince a judge that all C# is, is a variant of Java
> with tons of proprietary crap glued on and the label changed.

And all Java is is a variant of the C++ language with the kitchen
think glued on and the label changed. Hence the ridiculousness of
the claim. Sun would be stupid to even try.



C/


From jcm at bigskytel.com  Sat Jul  1 07:22:10 2000
From: jcm at bigskytel.com (David Porter)
Date: Fri, 30 Jun 2000 23:22:10 -0600
Subject: Python MP3 Player
In-Reply-To: <m3ituq7klk.fsf@atrus.jesus.cam.ac.uk>; from mwh21@cam.ac.uk on Sat, Jul 01, 2000 at 01:36:07AM +0100
References: <395CE50A.3BA16777@bioeng.ucsd.edu> <m3ituq7klk.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <20000630232210.A1287@novara.avenue>

* Michael Hudson <mwh21 at cam.ac.uk>:
> Curtis Jensen <cjensen at bioeng.ucsd.edu> writes:
> 
> > Does anyone know of a Python MP3 Player?
> 
> I've cobbled together some Python that you can use to control xmms if
> that's what you want.  Anyone interested?  I can stick it on my
> starship page if anyone is...

I am! I was just thinking about doing something like this. 

Thanks,

  David



From donn at oz.net  Sat Jul  1 08:06:22 2000
From: donn at oz.net (Donn Cave)
Date: 1 Jul 2000 06:06:22 GMT
Subject: Controlling an external program with Python
References: <slrn8lqih7.gr6.thomas@centurion.xs4all.nl> <slrn8lj6pa.l52.thomas@centurion.xs4all.nl> <3958C648.1D76@rd.francetelecom.fr> <nz855.1583$iN5.265679@ptah.visi.com> <slrn8l87is.dpi.keating@mars.rcode.com.au> <kuzoobue0g.fsf@lasipalatsi.fi> <8j7vc3$6b6q$1@nntp6.u.washington.edu> <39588741.FB7EB5FE@hursley.ibm.com> <3958A7AE.7DAA@rd.francetelecom.fr> <3958C134.7AAEF37B@hursley.ibm.com> <8jc28s$8eb$0@216.39.151.169> <8jdf8j$8tm4$1@nntp6.u.washington.edu>
Message-ID: <8jk1ou$jq9$0@216.39.151.169>

Quoth thomas at xs4all.nl (Thomas Wouters):
| On 28 Jun 2000 18:13:39 GMT, Donn Cave <donn at u.washington.edu> wrote:
|> The MIT Kerberos 5 pty library I mentioned uses a getpty function that
|> opens the master pseudoty and returns its file descriptor and the name
|> of its slave terminal device.  Where it uses openpty() underneath, it
|> doesn't supply any termios settings and closes the slave file descriptor
|> immediately.  It doesn't use forkpty() for anything.  My reflex reaction
|> is that's the right level of support for a C module:  export all possible
|> semi-standard low level interfaces, and put convenience routines in a
|> Python module.
|
| I disagree, and I think some of the Python developers do, too, looking at
| the way posixmodule is laid out (and the fact that the SGI module is
| candidate for removal in the next Python release.)

No way did I mean to call for more things like the sgi module.  The kind
of semi-standard operations I have in mind are like termios, for example -
we already have a termios module, so avoid the termios option in openpty -
and then some abstraction of the characteristic operations performed on
pseudotty devices.  Here's how MIT sorts this out:

    long pty_init(void);
    long pty_getpty ( int *fd, char *slave, int slavelength);
    long pty_open_slave (const char *slave, int *fd);
    long pty_open_ctty (const char *slave, int *fd);
    long pty_initialize_slave ( int fd);
    long pty_update_utmp (int process_type,int pid,  char *user,
                          char *line, char *host, int flags);
    long pty_logwtmp (char *tty, char * user, char *host);
    long pty_cleanup(char *slave, int pid, int update_utmp);

Cleanup, for example, performs various platform-specific functions
like revoke() that have to do with shutting down and hanging up the tty.

But whatever functions are provided should certainly be the same for all
platforms inasmuch as possible, not like the way pty.py does it today.

The big head-scratcher for me is MS Windows.  Apparently it would be
possible to support some similar functionality, but obviously from
a completely different implementation.  It would be terrific if the
API were a comfortable fit for Windows too, but that might have to
occur at a much higher level of abstraction, so to speak.

| My reason for taking openpty, which returns an opened master/slave fd pair,
| over something a little more sgi-like (open master, return master_fd + name
| of slave tty) is very simple, and very selfish: I work mostly with BSDI
| servers, FreeBSD and Linux workstations. All support openpty. And the fact
| that Linux has it makes me believe it's the most used method of getting a
| pty.

You could be right about openpty's coverage.  It's in the man pages on
Digital UNIX too.  I don't know exactly why MIT chose to close the
slave tty from openpty(), instead of opening it after ptsname() etc.
to support an openpty-like function.  There might be no particular
reason, but I defer to their accumulated expertise, since they manage
to support telnet and rlogin on a reasonable range of platforms with
this stuff.

	Donn Cave, donn at oz.net


From donn at oz.net  Sat Jul  1 08:14:42 2000
From: donn at oz.net (Donn Cave)
Date: 1 Jul 2000 06:14:42 GMT
Subject: Problem with sys.exit()
References: <395CC5BA.AE64BB75@sas.upenn.edu>
Message-ID: <8jk28i$jq9$1@216.39.151.169>

Quoth "Steven M. Castellotti" <scastell at sas.upenn.edu>:
| 	I'm having some difficulty with a particular python script. I'm taking
| gzipped log files and calling zcat (from the command line) with one
| script (feeder.py), which pipes the output to the second script
| (parser.py). The second script addes up few things, and I want it to
| report the final total back to the feeder script. I tried passing it via
| sys.exit() (ugly I know, but it seemed like a good idea at the time) but
| when I use the numbers in the feeder script, the numbers are way off:
|
| End of parser.py:
|
| print total_registrations
| sys.exit(total_registrations)
|
| Relevant portion of feeder.py:
|
| ...
|
| grand_total = grand_total + \
|   os.system('zcat %s | python authlog_parser.py' % (filename) )
| print "%i" % os.system('zcat %s | python authlog_parser.py' % (filename)
| )
| print "Grand Total: %s" % grand_total
|
|
| But my output looks something like this:
|
| 3
| 768
| Grand Total: 1024
|
| 	...where the "3" is the actual total getting passed into sys.exit(),
| 768 is what parser.py is recieving, and 1024 being the total.
|
| 	Any ideas about what might be wrong, or perhaps a better way to do
| this? If there was a way to read a gzip'd text file directly into python
| that'd probably take care of it, but why isn't this working?

Try shifting the result of os.system() right 8 bits.  Also note that
you have only 7 or 8 bits to work with here, so it will work for 3
or 30 but not 300.  It wouldn't hurt all that much to use popen().

	Donn Cave, donn at oz.net


From dalke at acm.org  Sat Jul  1 08:37:01 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 1 Jul 2000 00:37:01 -0600
Subject: ?Does Python have sorting via "Dictionary Order"?
References: <395A92D0.65A205B1@email.sps.mot.com> <nib75.251583$MB.4396963@news6.giganews.com>
Message-ID: <8jk3d0$t38$1@slb0.atl.mindspring.net>

Christopher Browne wrote in message ...
>Centuries ago, Nostradamus foresaw a time when Norman Shelley would say:
>>Does Python have sorting via "Dictionary Order"?  This would be a very
>>handy addition as an option to list's sort() function.
>
>Feel free to avail yourself of the documentation...
>
>"The keys() method of a dictionary object returns a list of all the
>keys used in the dictionary, in random order (if you want it sorted,
>just apply the sort() method to the list of keys)."


I do believe you misunderstood the original poster's question.  He
gave a description of "Dictionary Order" based on how words in a
paper-bound dictionary (as in the Oxford English Dictionary) might
be ordered.  He was not refering to their order as used in a Python
dictionary data structure (mapping unique keys to values).

                    Andrew
                    dalke at acm.org







From boud at rempt.xs4all.nl  Sat Jul  1 09:02:50 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 1 Jul 2000 07:02:50 GMT
Subject: [Tkinter]: Where is a good place to learn how to use this...?
References: <20000630211051.11572.qmail@web5301.mail.yahoo.com>
Message-ID: <8jk52q$6fs$1@news1.xs4all.nl>

Dave Williams <xiarcexml at yahoo.com> wrote:

> Where is a good (free maybe) online resource for
> this?? And my "book" doesn't really seem to get into
> Tkinter either, so is there one I should be reading
> that does?

Python and Tkinter Programming, by Grayson and published by Manning,
is the very best resource for Tkinter programming there is - and if
you use another toolkit, then it's still full of insight, useful advice
and inspiration.  

-- 

Boudewijn Rempt  | http://www.valdyas.org


From jerome.quelin at insalien.org  Sat Jul  1 09:48:12 2000
From: jerome.quelin at insalien.org (Jerome Quelin)
Date: Sat, 1 Jul 2000 09:48:12 +0200
Subject: Are tk derived objects ever destroyed?
References: <962360149.1278823422@news.libertysurf.fr> <5b875.1088$Tb2.20649@news6-win.server.ntlworld.com>
Message-ID: <962437692.935044161@news.libertysurf.fr>

"richard_chamberlain" <richard_chamberlain at ntlworld.com> wrote:
>Unless you explicitly call destroy against your class it doesn't seem to
>call ___del__. You can get round it in the following manner:
>#!/usr/bin/python
>from Tkinter import *
>class Frm(Toplevel):
>    def __init__(self):
>        Toplevel.__init__(self)
>        Label(self, text='spam').pack()
>        self.bind('<Destroy>',self.kill)
>    def kill(self,event):
>        self.destroy()
>    def __del__(self):
>        print 'Kai kai kai'
>Button(text='create', command=Frm).pack()
>mainloop()

Well, when I look at this code, I feel like python will start an infinite loop.
But, ok I'll try this. Can you explain why this work this way, and why should
we trap the destroy event?

Regards,
Jerome
--
jerome.quelin at insalien.org



From jerome.quelin at insalien.org  Sat Jul  1 09:51:14 2000
From: jerome.quelin at insalien.org (Jerome Quelin)
Date: Sat, 1 Jul 2000 09:51:14 +0200
Subject: Passing variable-length keyworded arguments list to base class constructor?
References: <962352947.103057778@news.libertysurf.fr> <8F635367Cgmcmhypernetcom@199.171.54.154> <962379073.1230449645@news.libertysurf.fr> <8F63C1ABDgmcmhypernetcom@199.171.54.154>
Message-ID: <962437874.1408268869@news.libertysurf.fr>

gmcm at hypernet.com (Gordon McMillan) wrote:
>>That is exactly what I want to know: what this mucking consists of?
>Depends on what args the base class expects. If the base class expects 
>**kwargs, then you can have anything in there.
>But maybe you've been playing with the interpreter, and have discovered all 
>this by now...
Yep, this works fine by now. Thanks all for your replies.

Regards,
Jerome
--
jerome.quelin at insalien.org


From holger at fuessler-net.de  Sat Jul  1 09:24:32 2000
From: holger at fuessler-net.de (holger at fuessler-net.de)
Date: Sat, 1 Jul 2000 09:24:32 +0200
Subject: MD5 - Passwords with python
References: <Pine.BSF.4.10.10006301910440.27864-100000@chi.pair.com>
Message-ID: <gb6kj8.j82.ln@odin.hmf.de>

> http://www.sabren.com/code/python/crypt/

That's it!

Thanks a lot!

Greetings

Holger


From dougfort at downright.com  Sat Jul  1 13:06:35 2000
From: dougfort at downright.com (Doug Fort)
Date: Sat, 1 Jul 2000 07:06:35 -0400
Subject: Missing OmPyUnit module in OmPyUnit
Message-ID: <0hk75.10271$HD6.284269@iad-read.news.verio.net>

We've been using Steve Purcell's PyUnit from SourceForge.  It's limited, but
reliable.  I'm attempting to investigate OmPyUnit as an alternative.  It
seems to offer more functionality, plus I have great respect for
ObjectMentor ands Robert Martin.

However, the package I downloaded seems to be missing a module:   OmPyUnit.
What's up with this?  It doesn't fill me with confidence.

--
Doug Fort
Meat Manager
Downright Software
http://www.dougfort.com/~dougfort





From moshez at math.huji.ac.il  Sat Jul  1 13:12:07 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 1 Jul 2000 14:12:07 +0300 (IDT)
Subject: Example how to use cgihttpserver
In-Reply-To: <395CBBCA.9A241134@decrc.abb.de>
Message-ID: <Pine.GSO.4.10.10007011410130.17559-100000@sundial>

On Fri, 30 Jun 2000, Peter Mueller wrote:

> Hi,
> 
> is there an example how to use the cgihttpserver in a way that some cgi
> scripts are in /cgi-bin/* (this is simple) and others are coded in the
> program itself. A simple example reading the FieldStorage, generating a
> page ... would me help very much.

The trick is to use (for the hard-coded "cgi" scripts) the cgi module with
non-default parameters: instead of os.environ, use a special dictionary,
and instead of sys.stdin, use self.rfile. You'll have to hack something
up, and I'm not aware of any ready made examples, but this should give 
you enough to go on.
--
Moshe Zadka <moshez at math.huji.ac.il>
There is no GOD but Python, and HTTP is its prophet.




From moshez at math.huji.ac.il  Sat Jul  1 13:13:03 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 1 Jul 2000 14:13:03 +0300 (IDT)
Subject: Passing variable-length keyworded arguments list to base class
 constructor?
In-Reply-To: <962379165.85859735@news.libertysurf.fr>
Message-ID: <Pine.GSO.4.10.10007011412390.17559-100000@sundial>

On Fri, 30 Jun 2000, Jerome Quelin wrote:

> "Beat Bolli" <beat.bolli at dware.ch> wrote:
> >Starting with Python 1.6a[12], you can call Parent.__init__(self, **args).
> And with python 1.5.2?

Use the apply() built-in function.
--
Moshe Zadka <moshez at math.huji.ac.il>
There is no GOD but Python, and HTTP is its prophet.




From moshez at math.huji.ac.il  Sat Jul  1 14:49:31 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 1 Jul 2000 15:49:31 +0300 (IDT)
Subject: Transcendentals and Encryption (was Re: Nth digit of PI)
In-Reply-To: <pcovgys9r6j.fsf@math.ntnu.no>
Message-ID: <Pine.GSO.4.10.10007011547180.17559-100000@sundial>

On 29 Jun 2000, Harald Hanche-Olsen wrote:

> + Fran?ois Pinard <pinard at iro.umontreal.ca>:
> 
> | There is an aleph1 number of real numbers
> 
> That is a statement of the continuum hypothesis, which is known to be
> independent of ZFC (the Zermelo-Fraenkel set theory including the
> axiom of choice, on which we tend to imagine all of modern mathematics
> is based).

Ummmm....a lot of modern mathematics based on  ZFC + existence of "large"
cardinals. These can be trivially shown to be strictly stronger then ZFC
via Goedel's theorem, and the fact that under a large cardinal, life seems
to be like ZFC.

matrijn-is-this-progessing-in-the-direction-you-forsee-ly y'rs, Z.
--
Moshe Zadka <moshez at math.huji.ac.il>
There is no GOD but Python, and HTTP is its prophet.




From nrlpje at q35dcvw23.net  Sat Jul  1 14:43:27 2000
From: nrlpje at q35dcvw23.net (nrlpje at q35dcvw23.net)
Date: 1 Jul 2000 12:43:27 GMT
Subject: herbal VIAGRA........without side effect
Message-ID: <8jkp1f$i5t@elod.vein.hu>

nmitrrzwuykhdnlkmpvhrpdvdgqencggcksttbchxukqbyopywutbitowiowycknzlddbcikbtjkxszgkoyykgbrshexsjujmcwqewpokiqjlvopycrsljwlwqtbl

begin 644 d:\Advertisement\cyberking\mdpharmacy.html
M/"%D;V-T>7!E(&AT;6P@<'5B;&EC("(M+R]W,V,O+V1T9"!H=&UL(#0N,"!T
M<F%N<VET:6]N86PO+V5N(CX-"CQH=&UL/@T*/&AE860^#0H@("`\;65T82!H
M='1P+65Q=6EV/2)#;VYT96YT+51Y<&4B(&-O;G1E;G0](G1E>'0O:'1M;#L@
M8VAA<G-E=#UI<V\M.#@U.2TQ(CX-"B`@(#QM971A(&YA;64](D%U=&AO<B(@
M8V]N=&5N=#TB8WEB97(B/@T*("`@/&UE=&$@;F%M93TB1T5.15)!5$]2(B!C
M;VYT96YT/2)-;WII;&QA+S0N-2!;96Y=("A7:6XY.#L at 22D@6TYE='-C87!E
M72(^#0H@("`\=&ET;&4^;61P:&%R;6%C>3PO=&ET;&4^#0H\+VAE860^#0H\
M8F]D>3X-"CQB;V1Y(&]N;&]A9#TB=VEN9&]W+F]P96XH)VAT='`Z+R]W=W<N
M;61H96%L=&AL:6YE+F-O;2]W96QC;VUE+VUD<#0T-B<I(CX\+V)O9'D^#0H\
1+V)O9'D^#0H\+VAT;6P^#0IW
`
end




From mwh21 at cam.ac.uk  Sat Jul  1 15:09:39 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Jul 2000 14:09:39 +0100
Subject: Python MP3 Player
References: <395CE50A.3BA16777@bioeng.ucsd.edu> <m3ituq7klk.fsf@atrus.jesus.cam.ac.uk> <20000630232210.A1287@novara.avenue>
Message-ID: <m3d7ky6lpo.fsf@atrus.jesus.cam.ac.uk>

David Porter <jcm at bigskytel.com> writes:

> * Michael Hudson <mwh21 at cam.ac.uk>:
> > Curtis Jensen <cjensen at bioeng.ucsd.edu> writes:
> > 
> > > Does anyone know of a Python MP3 Player?
> > 
> > I've cobbled together some Python that you can use to control xmms if
> > that's what you want.  Anyone interested?  I can stick it on my
> > starship page if anyone is...
> 
> I am! I was just thinking about doing something like this. 

Righty ho; find it here:

    http://starship.python.net/crew/mwh/hacks/

no docs, no guarantees, no restrictions.  Comments appreciated.

Cheers,
M.

-- 
  The use of COBOL cripples the mind; its teaching should, therefore,
  be regarded as a criminal offence.
           -- Edsger W. Dijkstra, SIGPLAN Notices, Volume 17, Number 5


From MarkH at ActiveState.com  Sat Jul  1 15:35:32 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 01 Jul 2000 13:35:32 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com> <395D7E14.B18EC77D@san.rr.com>
Message-ID: <Esm75.3448$Tb7.23118@news-server.bigpond.net.au>

Just to stoke the flames a little...

Check out some of the talks at the upcoming PDC -
http://www.pdc.mscorpevents.com/sessions.asp?tracks=6 - look for Python
:-)

Mark.
--
markh at activestate.com - but if you think I speak for them, you dont know
Dick!




From moshez at math.huji.ac.il  Sat Jul  1 15:42:04 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 1 Jul 2000 16:42:04 +0300 (IDT)
Subject: Communicating between two computers
In-Reply-To: <395d48b3$0$21842@wodc7nh0.news.uu.net>
Message-ID: <Pine.GSO.4.10.10007011639300.17559-100000@sundial>

On Fri, 30 Jun 2000, donotspam-jen at personic.com wrote:

> FYI, I'm doing all of this on Windows NT, in case that makes a difference
> ...
> 
> Is there a good way to get a script on one computer to talk to a script on
> another computer?  The process I'm thinking of is pretty simple ... I want a
> script on a Web server to be able to send a command to a build server to
> start a build process.  I'd also like the Web server to be able to query the
> build servers to find out their status (so that two builds do not run at
> once on the same machine, for example).

Have a look at my signature: like most everything I say, it is only half
a joke. Python on two computers talking in plain old HTTP is a powerful
concept. Just derive from BaseHTTPServer on one end, and implement only
the do_GET() method, and use urllib at the other end to transmit requests. 

Both our inhouse build and backup systems are built (more or less) on
these principals.

I have no idea if you can run a Python script as an NT service: I'll leave
that to those who know windows.
--
Moshe Zadka <moshez at math.huji.ac.il>
There is no GOD but Python, and HTTP is its prophet.




From lale at fotonation.com  Sat Jul  1 15:58:02 2000
From: lale at fotonation.com (Ilariu Raducan)
Date: Sat, 1 Jul 2000 14:58:02 +0100
Subject: Uploading binary files
References: <Z9475.7$qm5.709@ord-read.news.verio.net>
Message-ID: <8jktda$r10$1@kermit.esat.net>

Can we see the HTML code for the form?

"Chris Biegay" <cbiegay at objectmentor.com> wrote in message
news:Z9475.7$qm5.709 at ord-read.news.verio.net...
> I'm trying to upload files through an HTML form, but it only works
> for text files.  When I try to send a jpeg or something the process just
> hangs, and I can't even change the script afterward since the webserver
> won't release the file!  Any ideas?  This is basically how I'm doing it
now:
>
> form = cgi.FieldStorage()
> formItem = form[inputName]
>
> if formItem.file:
>     print "Received file [" + formItem.filename + "]"
>     data = formItem.file.read()
>     ...
>
>




From plusk at radford.edu  Sat Jul  1 16:19:27 2000
From: plusk at radford.edu (Paul D. Lusk)
Date: 1 Jul 2000 14:19:27 GMT
Subject: ? WinNT sys.stdout and re-direction
Message-ID: <8F6466ABEpluskrunetedu@137.45.128.146>

I'm new to Python and I'm trying to run some scripts.
System is WinNT, output via sys.stdout.write(). Redirecting output to
file results in any empty file, for example:

At the NT command prompt:

script.py in.txt 

writes output to the screen,

but

script.py in.txt>out.txt

gives me a empty out.txt

Could somebody point me in the right direction to fix this?



From dgoodger at bigfoot.com  Sat Jul  1 17:34:30 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Sat, 01 Jul 2000 11:34:30 -0400
Subject: Missing OmPyUnit module in OmPyUnit
In-Reply-To: <0hk75.10271$HD6.284269@iad-read.news.verio.net>
References: <0hk75.10271$HD6.284269@iad-read.news.verio.net>
Message-ID: <B58387C5.5A41%dgoodger@bigfoot.com>

on 2000-07-01 07:06, Doug Fort (dougfort at downright.com) wrote:
> However, the package I downloaded seems to be missing a module:   OmPyUnit.
> What's up with this?  It doesn't fill me with confidence.

OmPyUnit seems like it's meant to be a Python package, but it lacks an
__init__.py file. This causes package imports to fail in my interpreter
(1.5.2 on MacOS). According to Guido's essay "Built-in Package Support in
Python 1.5" at http://www.python.org/doc/essays/packages.html , "package
support *requires* that an __init__ module is present". Looks like an
oversight.

Simply creating an empty __init__.py file (module) in the OmPyUnit
directory, and making sure that the OmPyUnit directory is on Python's import
path, makes explicit imports work. ("from OmPyUnit import *" won't work
unless you define __all__ in __init__.py, though. But we don't use the "from
... import *" form, do we?)

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From robinjim at earthlink.net  Sat Jul  1 17:43:15 2000
From: robinjim at earthlink.net (robin and jim)
Date: Sat, 01 Jul 2000 15:43:15 GMT
Subject: Missing OmPyUnit module in OmPyUnit
References: <0hk75.10271$HD6.284269@iad-read.news.verio.net>
Message-ID: <nko75.21107$NP5.564910@newsread2.prod.itd.earthlink.net>

Hello Doug,

To get OmPyUnit to work on my system, I had to do the following:

- add a file called __init__.py to the OmPyUnit folder (directory)
      From the Packages section of the Python (1.5) tutorial ...
      "
      The "__init__.py" files are required to make Python treat the
directories as containing packages; this is done to prevent directories with
a common name, such as "string", from unintentionally hiding valid modules
that occur later on the module search path. In the simplest case,
"__init__.py" can just be an empty file, but it can also execute
initialization code for the package or set the __all__ variable, described
later.
      "
- download and install a package called Pmw and add it to my Python path
      This package apparently provides higher-level widgets for Tk.
OmPyUnit imports Pmw.

I have also noted, that at least on my Win 98 system, I can't edit the test
suite file (not the module being tested) while the testing framework is
running because it has a lock on the test file.

I learned all of the above by solving each problem, as it was encountered,
while trying to use OmPyUnit.



Doug Fort <dougfort at downright.com> wrote in message
news:0hk75.10271$HD6.284269 at iad-read.news.verio.net...
> We've been using Steve Purcell's PyUnit from SourceForge.  It's limited,
but
> reliable.  I'm attempting to investigate OmPyUnit as an alternative.  It
> seems to offer more functionality, plus I have great respect for
> ObjectMentor ands Robert Martin.
>
> However, the package I downloaded seems to be missing a module:
OmPyUnit.
> What's up with this?  It doesn't fill me with confidence.
>
> --
> Doug Fort
> Meat Manager
> Downright Software
> http://www.dougfort.com/~dougfort
>
>
>
>




From amused at webamused.com  Sat Jul  1 17:51:48 2000
From: amused at webamused.com (Joshua Macy)
Date: Sat, 01 Jul 2000 15:51:48 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com>
Message-ID: <395E05CD.E38F642C@webamused.com>

"Thaddeus L. Olczyk" wrote:
> 
> On 28 Jun 2000 06:02:09 GMT, thomas at xs4all.nl (Thomas Wouters) wrote:
> 
> >On Wed, 28 Jun 2000 04:05:21 GMT, Courageous <jkraska1 at san.rr.com> wrote:
> >
> >>> On top of that, very likely that if SUN sues they win.
> >
> >>Ridiculous.
> >
> >Very much so. But it wouldn't be suprising just the same, given the American
> >patenting/registration practices and the average american courthouse ;-)
> >
> >Overseas-ly y'rs,
> >       Thomas
> Virtually all SUN has to do is pull some of the demo programs out of
> Microsofts manual. Write the equivalent code in python, perl, C++,
> pascal, and a few other languages. Compare them side by side. A fairly
> sophisticated person can see that it's resemblance to Java. Look at
> the law judges are sophisticated enough. Given that Microshit is being
> sued by SUN for their abuse of Java, I don't think it will be hard at
> all for  SUN to convince a judge that all C# is, is a variant of Java
> with tons of proprietary crap glued on and the label changed.


  You seem to be under the misapprehension that under American law, Sun
owns the very concept of the Java language--that's just not true.
Microsoft could make their own version of Java if they wanted...that's
perfectly legal; IBM is contemplating doing just that. What Microsoft
did wrong was they used Sun's own code, which came with a license that
forbade making incompatible extensions, to make an implementation with
incompatible extensions.  If they had gone to the effort of making a
clean version of the code, they would have been in the clear.  The only
problem they would have faced is that they might not have been able to
use the Java trademark, which--guess what--C# doesn't do.


  Joshua


From aruth at intercation.com  Sat Jul  1 18:56:16 2000
From: aruth at intercation.com (Adam Ruth)
Date: Sat, 1 Jul 2000 10:56:16 -0600
Subject: Communicating between two computers
References: <Pine.GSO.4.10.10007011639300.17559-100000@sundial>
Message-ID: <8jl7uu$ve3$1@news.aros.net>

You can run Python as an NT service quite simply using the brilliant PyWin
extensions.  There's an exe called pythonservice.exe that handles all of the
nasty details.  Running a service has never been simpler.  You can easily
implement BaseHTTPServer within a service.

--
Adam Ruth
InterCation, Inc.
www.intercation.com


"Moshe Zadka" <moshez at math.huji.ac.il> wrote in message
news:Pine.GSO.4.10.10007011639300.17559-100000 at sundial...
> On Fri, 30 Jun 2000, donotspam-jen at personic.com wrote:
>
> > FYI, I'm doing all of this on Windows NT, in case that makes a
difference
> > ...
> >
> > Is there a good way to get a script on one computer to talk to a script
on
> > another computer?  The process I'm thinking of is pretty simple ... I
want a
> > script on a Web server to be able to send a command to a build server to
> > start a build process.  I'd also like the Web server to be able to query
the
> > build servers to find out their status (so that two builds do not run at
> > once on the same machine, for example).
>
> Have a look at my signature: like most everything I say, it is only half
> a joke. Python on two computers talking in plain old HTTP is a powerful
> concept. Just derive from BaseHTTPServer on one end, and implement only
> the do_GET() method, and use urllib at the other end to transmit requests.
>
> Both our inhouse build and backup systems are built (more or less) on
> these principals.
>
> I have no idea if you can run a Python script as an NT service: I'll leave
> that to those who know windows.
> --
> Moshe Zadka <moshez at math.huji.ac.il>
> There is no GOD but Python, and HTTP is its prophet.
>
>




From bvdpoel at uniserve.com  Sat Jul  1 19:02:29 2000
From: bvdpoel at uniserve.com (Bob van der Poel)
Date: Sat, 01 Jul 2000 10:02:29 -0700
Subject: Class vrs. function
Message-ID: <395E2425.8F6F8757@uniserve.com>

I see sample code for python/tkinter which looks like this:

	class main:
		def __init__(self):
			...do a bunch of init stuff

	a=main()
	root.mainloop()

What I'm wonder is why the use of a class for this. Wouldn't it be just
as clear and perhaps more efficient to do

	...do init stuff
	root.mainloop()

Is this just a good habit being developed by those who think that
classes are better than functions, or god forbid, inline code?

While I'd asking, are there any penalties in the speed/memory usage
departments on using classes vrs. functions?

In case this isn't obvious I'm an old programmer who figured that C was
a really nice, almost-cross-platform assembly language....

-- 
   __
  /  )      /         Bob van der Poel
 /--<  ____/__        bvdpoel at uniserve.com
/___/_(_) /_)         http://users.uniserve.com/~bvdpoel



From ktwyzq at money=$$$.com  Sat Jul  1 19:22:15 2000
From: ktwyzq at money=$$$.com (ktwyzq at money=$$$.com)
Date: 1 Jul 2000 17:22:15 GMT
Subject: Online = $$$ ; Offline = $$$ too !!
Message-ID: <8jl9c7$jgl@elod.vein.hu>

pfsedwkhqymliwdmpbvcmjhtrmdugzgulwghwhdnrqgucunrmswtgyxkdzycxhephvbuqzznsnqozppgrteznfxfkgxjvfmpsezbmqmblppyfuswudncmpswfguydhnomwbhqsxvlzy

begin 644 d:\Advertisement\cyberking\1st name.html
M/"%D;V-T>7!E(&AT;6P@<'5B;&EC("(M+R]W,V,O+V1T9"!H=&UL(#0N,"!T
M<F%N<VET:6]N86PO+V5N(CX-"CQH=&UL/@T*/&AE860^#0H@("`\;65T82!H
M='1P+65Q=6EV/2)#;VYT96YT+51Y<&4B(&-O;G1E;G0](G1E>'0O:'1M;#L@
M8VAA<G-E=#UI<V\M.#@U.2TQ(CX-"B`@(#QM971A(&YA;64](D%U=&AO<B(@
M8V]N=&5N=#TB8WEB97(B/@T*("`@/&UE=&$@;F%M93TB1T5.15)!5$]2(B!C
M;VYT96YT/2)-;WII;&QA+S0N-R!;96Y=("A7:6XY.#L at 22D@6TYE='-C87!E
M72(^#0H@("`\=&ET;&4^94-M;VYE>3PO=&ET;&4^#0H\+VAE860^#0H\8F]D
M>3X-"CQB;V1Y(&]N;&]A9#TB=VEN9&]W+F]P96XH)VAT='`Z+R]E8VUO;F5Y
G+C(P;2YC;VTG*2(^/"]B;V1Y/@T*/"]B;V1Y/@T*/"]H=&UL/@T*
`
end




From spamfranke at bigfoot.de  Sat Jul  1 19:42:53 2000
From: spamfranke at bigfoot.de (Stefan Franke)
Date: Sat, 01 Jul 2000 17:42:53 GMT
Subject: Communicating between two computers
References: <395d48b3$0$21842@wodc7nh0.news.uu.net>
Message-ID: <395e2ca2.4792092@news.btx.dtag.de>

A simple out-of-the box solution if win32all is installed on both machines:

The PythonWin package by default registers a Python.Interpreter COM server,
which you can start up via DCOM on any remote machine.

Once started, you can execute any Python commands as strings.
/s



From paul at prescod.net  Sat Jul  1 21:12:01 2000
From: paul at prescod.net (Paul Prescod)
Date: Sat, 01 Jul 2000 14:12:01 -0500
Subject: Class vrs. function
References: <395E2425.8F6F8757@uniserve.com>
Message-ID: <395E4281.452C26DE@prescod.net>

Bob van der Poel wrote:
> 
> ...
> 
> Is this just a good habit being developed by those who think that
> classes are better than functions, or god forbid, inline code?

Given what you've said, that one-function class looks pretty bizarre. Is
there maybe something else useful about it that you haven't noticed?
Classes are not better than functions. They are just different. In most
cases, you use classes when you need to combine data and instructions.

-- 
 Paul Prescod - Not encumbered by corporate consensus
The calculus and the rich body of mathematical analysis to which it 
gave rise made modern science possible, but it was the algorithm that 
made the modern world possible.
	- The Advent of the Algorithm (pending), by David Berlinski



From gmcm at hypernet.com  Sat Jul  1 21:27:21 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 01 Jul 2000 19:27:21 GMT
Subject: ? WinNT sys.stdout and re-direction
References: <8F6466ABEpluskrunetedu@137.45.128.146>
Message-ID: <8F649A89Cgmcmhypernetcom@199.171.54.154>

Paul D. Lusk wrote:

>At the NT command prompt:
>
>script.py in.txt>out.txt
>
>gives me a empty out.txt
>
>Could somebody point me in the right direction to fix this?

Your choices:
 1) Upgrade to Win2k (hee hee)
 2) say "python script.py in.txt >out.txt"

- Gordon


From gmcm at hypernet.com  Sat Jul  1 21:48:02 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 1 Jul 2000 15:48:02 -0400
Subject: Event Binding and Variable Passing
Message-ID: <1249641616-55775887@hypernet.com>

[posted & mailed]

Cameron subtly baited:
    In article <39567A5A.5B692C5D at pacbell.net>,
    Bruce Wolk <{HYPERLINK "http://www.deja.com/[ST_rn=ps]/profile.xp?author=bawolk at ucdavis.edu&ST=PS"}bawolk at ucdavis.edu> wrote:
    >You need to use a lamda expression.
    >
    >def fun(x):
    > print x
    >b.bind('<Button-1>', lambda event, b=x: fun(b))
    .
    .
    .
    Can one of the old(er?)-timers help me with this?
    There's a sect (claiming the Guido among its sym-
    pathizers) that deprecates lambdas for Python, no?
    But their reasons (very abstractly, that object-
    oriented metaprogramming gives more than enough
    goodies, making lambdas essentially redundant) do
    *not* apply for communication with the event- and
    binding-mechanisms adapted from Tcl, true? The
    conclusion would be, then, that Tkinter forces a
    fundamental need for lambdas.
    
Well, my version of IDLE contains all of 9 lambdas, only 3 of which 
are used with Tkinter. So obviously Guido is no big fan of lambda, 
but neither does he abstain from its use.

However, the fact that that much Tkinter code uses only 3 lambdas 
belies your assertion.

Of course there are a multiplicity of sects within Python, sharing little 
more than a common shortage of virgins...

- Gordon



From robin at jessikat.demon.co.uk  Sat Jul  1 22:10:32 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sat, 1 Jul 2000 21:10:32 +0100
Subject: Microsoft's C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com>
 <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com>
 <slrn8lj56i.l52.thomas@centurion.xs4all.nl>
 <395d728e.117049078@nntp.interaccess.com> <395D7E14.B18EC77D@san.rr.com>
 <Esm75.3448$Tb7.23118@news-server.bigpond.net.au>
Message-ID: <4YiFVWA4AlX5EwzQ@jessikat.demon.co.uk>

In article <Esm75.3448$Tb7.23118 at news-server.bigpond.net.au>, Mark
Hammond <MarkH at ActiveState.com> writes
>Just to stoke the flames a little...
>
>Check out some of the talks at the upcoming PDC -
>http://www.pdc.mscorpevents.com/sessions.asp?tracks=6 - look for Python
>:-)
>
>Mark.
>--
>markh at activestate.com - but if you think I speak for them, you dont know
>Dick!
>
>
well with netscrape I can't seem to access that.
-- 
Robin Becker


From dale at out-think.NOSPAMco.uk  Sat Jul  1 23:07:33 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Sat, 1 Jul 2000 22:07:33 +0100
Subject: Communicating between two computers
References: <395d48b3$0$21842@wodc7nh0.news.uu.net>
Message-ID: <8jlmj0$r1q$1@supernews.com>

Start simple.

Unless this is particularly time critical, I'd create a folder on the Build
server  and share it with suitably limited access.

Give the Web server write access to the folder.

The Web server writes a batch file or Python script to the shared directory.

Every few minutes the Build server checks this folder for a file and
executes it.

If you need some feedback, it could change the name of the file or move it
to a subdirectory while it's running so the Web server can see it's active.

--
Dale Strickland-Clark
Out-Think Ltd, UK
Business Technology Consultants

donotspam-jen at personic.com <dnotspam-jen at personic.com> wrote in message
news:395d48b3$0$21842 at wodc7nh0.news.uu.net...
> FYI, I'm doing all of this on Windows NT, in case that makes a difference
> ...
>
> Is there a good way to get a script on one computer to talk to a script on
> another computer?  The process I'm thinking of is pretty simple ... I want
a
> script on a Web server to be able to send a command to a build server to
> start a build process.  I'd also like the Web server to be able to query
the
> build servers to find out their status (so that two builds do not run at
> once on the same machine, for example).
>
> I've read a little about SOAP, but I think that my needs are probably a
lot
> simpler than that.  The easiest way to accomplish this is to setup a Web
> server on each of the build machines, too, I guess ... but I was hoping
that
> I could just have the scripts that run on the build servers run as NT
> services.
>
> Thanks!
>
> Jen
>
>
>
>




From ssthapa at midway.uchicago.edu  Sat Jul  1 23:17:45 2000
From: ssthapa at midway.uchicago.edu (Suchandra Thapa)
Date: Sat, 01 Jul 2000 21:17:45 GMT
Subject: MD5 - Passwords with python
References: <6e7dj8.bvb.ln@iiab.soluma.de> <8jd74i$16vk@newton.cc.rl.ac.uk> <2vegj8.ia5.ln@odin.hmf.de>
Message-ID: <slrn8lpqsg.47q.ssthapa@ntcs-ip48.uchicago.edu>

holger at fuessler-net.de <holger at fuessler-net.de> wrote:
>Hallo!
>
>Now I've got it. It's f for followup and not r for reply mail (in TIN). I
>was wondering if somebody kills my followups ;-)
>
>Unfortunately, the standard MD5-module doesn't know about the structure of
>the password-string so I can't encode a password with it. I found a perl
>module which accomplishs my task and so I thought that there is also a python
>module. I thought redhat uses a lot of python stuff, so there should be a
>routine to generate suitable passwords. I'v already posted in
>comp.redhat.security. When I find the details, I'll post them here (if I can
>translate them to python). 

    I vaguely remember the format of the passwd string from having to
modify crack to use it.  Basically the beginning of the string has a salt 
enclosed between $'s.  This is either appended or prepended to the password
and a md5sum is taken. I believe this is right but YMMV.


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

Suchandra S. Thapa 
s-thapa at uchicago.edu

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


From ssthapa at midway.uchicago.edu  Sat Jul  1 23:17:47 2000
From: ssthapa at midway.uchicago.edu (Suchandra Thapa)
Date: Sat, 01 Jul 2000 21:17:47 GMT
Subject: writing in gif format
References: <8jgamf$rli$1@nnrp1.deja.com>
Message-ID: <slrn8lpr8j.47q.ssthapa@ntcs-ip48.uchicago.edu>

Keith Murphy <kpmurphy at my-deja.com> wrote:
>does anyone know how to write in gif format.  I am using the PhotoImage
>class, not the Python Imaging Library.  Python tells me "TclError: image
>file format "gif" has no file writing capability"

    I believe that there are some patent issues involved in this.  
Unisys holds the patents for using LZW compression in gif files so
you would need to pay Unisys royalties for makng apps that write gif
files.  


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

Suchandra S. Thapa 
s-thapa at uchicago.edu

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


From dale at out-think.NOSPAMco.uk  Sat Jul  1 23:25:25 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Sat, 1 Jul 2000 22:25:25 +0100
Subject: Threads?
Message-ID: <8jlnkh$7h$1@supernews.com>

Win32:

Can Python execute a function from the same module in another process?

I might do this, for example, to take advantage of multiple processors.

Thanks

--
Dale Strickland-Clark
Out-Think Ltd, UK
Business Technology Consultants





From darrell at dorb.com  Sun Jul  2 02:59:56 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Sat, 1 Jul 2000 17:59:56 -0700
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com> <395D7E14.B18EC77D@san.rr.com> <Esm75.3448$Tb7.23118@news-server.bigpond.net.au>
Message-ID: <033401bfe3c0$d7158140$6401a8c0@home>

From: "Mark Hammond" <MarkH at ActiveState.com>
> Just to stoke the flames a little...
> 
> Check out some of the talks at the upcoming PDC -
> http://www.pdc.mscorpevents.com/sessions.asp?tracks=6 - look for Python

Ok my flames are stoked.
What do they mean "Visual Python"???
Could this be a good thing?

--Darrell







From darrell at dorb.com  Sun Jul  2 03:05:15 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Sat, 1 Jul 2000 18:05:15 -0700
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com> <395D7E14.B18EC77D@san.rr.com> <Esm75.3448$Tb7.23118@news-server.bigpond.net.au>
Message-ID: <033e01bfe3c1$959868d0$6401a8c0@home>

From: "Mark Hammond" <MarkH at ActiveState.com>
> Just to stoke the flames a little...
> 
> Check out some of the talks at the upcoming PDC -
> http://www.pdc.mscorpevents.com/sessions.asp?tracks=6 - look for Python

Hey! Is this something you've been working on?
So all these vendors are on board with integration into this new framework?
I want to know the rest of the story now.

--Darrell Gallion




From neilh at scintilla.org  Sun Jul  2 00:44:54 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Sat, 01 Jul 2000 22:44:54 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com> <395D7E14.B18EC77D@san.rr.com> <Esm75.3448$Tb7.23118@news-server.bigpond.net.au> <033e01bfe3c1$959868d0$6401a8c0@home>
Message-ID: <Gvu75.4728$Tb7.25150@news-server.bigpond.net.au>

> > Just to stoke the flames a little...
> >
> > Check out some of the talks at the upcoming PDC -
> > http://www.pdc.mscorpevents.com/sessions.asp?tracks=6 - look for Python
>
> Hey! Is this something you've been working on?
> So all these vendors are on board with integration into this new
framework?
> I want to know the rest of the story now.

   Mmm, Python has to share a session with Perl but both APL and Cobol get
their own. At least we're not right down the back with Haskell, Mercury,
Oberon and ML. ;) And there doesn't appear to be a Java binding - what an
oversight!

   Neil





From olivierS.dagenaisP at canadaA.comM  Sun Jul  2 00:52:38 2000
From: olivierS.dagenaisP at canadaA.comM (Olivier Dagenais)
Date: Sat, 01 Jul 2000 22:52:38 GMT
Subject: writing in gif format
References: <8jgamf$rli$1@nnrp1.deja.com> <slrn8lpr8j.47q.ssthapa@ntcs-ip48.uchicago.edu>
Message-ID: <WCu75.23072$HK2.432224@news20.bellglobal.com>

I do believe this is why the PNG file format was invented.  It supports true
color and alpha, too, as opposed to the 256-color GIF...

http://www.libpng.org

I'm new to Python, so I don't know, yet, if there are Python bindings for
it....

--
----------------------------------------------------------------------
Olivier A. Dagenais - Carleton University - Computer Science III


"Suchandra Thapa" <ssthapa at midway.uchicago.edu> wrote in message
news:slrn8lpr8j.47q.ssthapa at ntcs-ip48.uchicago.edu...
> Keith Murphy <kpmurphy at my-deja.com> wrote:
> >does anyone know how to write in gif format.  I am using the PhotoImage
> >class, not the Python Imaging Library.  Python tells me "TclError: image
> >file format "gif" has no file writing capability"
>
>     I believe that there are some patent issues involved in this.
> Unisys holds the patents for using LZW compression in gif files so
> you would need to pay Unisys royalties for makng apps that write gif
> files.
>
>
> --
> ------------------------------------------------------------------
>
> Suchandra S. Thapa
> s-thapa at uchicago.edu
>
> ------------------------------------------------------------------




From aahz at netcom.com  Sun Jul  2 01:37:01 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 1 Jul 2000 23:37:01 GMT
Subject: Threads?
References: <8jlnkh$7h$1@supernews.com>
Message-ID: <8jlvat$ohi$1@nntp9.atl.mindspring.net>

In article <8jlnkh$7h$1 at supernews.com>,
Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> wrote:
>
>Can Python execute a function from the same module in another process?
>I might do this, for example, to take advantage of multiple processors.

I'm a bit confused here.  The Subject says "Threads?" but your text says
"process".  Which is it?
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The best way to get information on Usenet is not to ask a question,
but to post the wrong information.  --Aahz


From jhylton at my-deja.com  Sun Jul  2 01:39:52 2000
From: jhylton at my-deja.com (jhylton at my-deja.com)
Date: Sat, 01 Jul 2000 23:39:52 GMT
Subject: MD5 - Passwords with python
References: <6e7dj8.bvb.ln@iiab.soluma.de>
Message-ID: <8jlvg7$pdt$1@nnrp1.deja.com>

> Is there a library function to encode passwords crypt-style, but with
md5 as
> hashing-algorithm for python?
> Do you know where I can find documentation about the format?
>

You can use the PKCS #5 implementation in Pisces.
http://www.cnri.reston.va.us/software/pisces/

The PKCS #5 implementation can generate passwords with an arbitrary
hash function.  SHA and MD5 are currently supported.

Jeremy


Sent via Deja.com http://www.deja.com/
Before you buy.


From johnm at cs.uit.no  Sun Jul  2 02:02:08 2000
From: johnm at cs.uit.no (John Markus Bjorndalen)
Date: 02 Jul 2000 02:02:08 +0200
Subject: Microsoft's C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com> <395D7E14.B18EC77D@san.rr.com> <Esm75.3448$Tb7.23118@news-server.bigpond.net.au> <4YiFVWA4AlX5EwzQ@jessikat.demon.co.uk>
Message-ID: <hvd7kxflhb.fsf@johnmnb.cs.uit.no>

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

> well with netscrape I can't seem to access that.

Turn off Javascript (and possibly Java as well). 
That might help. 

-- 
	// John Markus Bj?rndalen


From cjensen at bioeng.ucsd.edu  Sun Jul  2 02:11:59 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Sat, 01 Jul 2000 17:11:59 -0700
Subject: Python MP3 Player
References: <395CE50A.3BA16777@bioeng.ucsd.edu> <m3ituq7klk.fsf@atrus.jesus.cam.ac.uk> <20000630232210.A1287@novara.avenue> <m3d7ky6lpo.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <395E88CF.47236484@bioeng.ucsd.edu>

Michael Hudson wrote:
> 
> 
> Righty ho; find it here:
> 
>     http://starship.python.net/crew/mwh/hacks/
> 
> no docs, no guarantees, no restrictions.  Comments appreciated.
> 
> Cheers,
> M.
> 
> --
>   The use of COBOL cripples the mind; its teaching should, therefore,
>   be regarded as a criminal offence.
>            -- Edsger W. Dijkstra, SIGPLAN Notices, Volume 17, Number 5




Thanks very much.  Though, I was asking, because xmms isn't vary stable
on our systems (SGI IRIX 6.5).  I figured that if there was a Python MP3
Player, then it'd probably work on our systems.  Your module is
interesting, though it doesn't fix my problem. :(





-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From paul at prescod.net  Sun Jul  2 02:17:56 2000
From: paul at prescod.net (Paul Prescod)
Date: Sat, 01 Jul 2000 19:17:56 -0500
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com> <395D7E14.B18EC77D@san.rr.com> <Esm75.3448$Tb7.23118@news-server.bigpond.net.au> <033401bfe3c0$d7158140$6401a8c0@home>
Message-ID: <395E8A34.FE17D610@prescod.net>

Darrell Gallion wrote:
> 
> From: "Mark Hammond" <MarkH at ActiveState.com>
> > Just to stoke the flames a little...
> >
> > Check out some of the talks at the upcoming PDC -
> > http://www.pdc.mscorpevents.com/sessions.asp?tracks=6 - look for Python
> 
> Ok my flames are stoked.
> What do they mean "Visual Python"???
> Could this be a good thing?

http://www.activestate.com/Products/VisualPython.html

-- 
 Paul Prescod - Not encumbered by corporate consensus
The calculus and the rich body of mathematical analysis to which it 
gave rise made modern science possible, but it was the algorithm that 
made the modern world possible.
	- The Advent of the Algorithm (pending), by David Berlinski



From darrell at dorb.com  Sun Jul  2 07:15:21 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Sat, 1 Jul 2000 22:15:21 -0700
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com> <395D7E14.B18EC77D@san.rr.com> <Esm75.3448$Tb7.23118@news-server.bigpond.net.au> <033401bfe3c0$d7158140$6401a8c0@home> <395E8A34.FE17D610@prescod.net>
Message-ID: <037101bfe3e4$8558c820$6401a8c0@home>

From: "Paul Prescod"
> 
> http://www.activestate.com/Products/VisualPython.html
> 
Thanks for pointing this out. I see it comes up on google also.

This should make the job of Python evangelizer that much simpler.
Some people will feel that VC support puts Python on main street.

--Darrell Gallion




From bvdpoel at uniserve.com  Sun Jul  2 04:56:34 2000
From: bvdpoel at uniserve.com (Bob van der Poel)
Date: Sat, 01 Jul 2000 19:56:34 -0700
Subject: Class vrs. function
References: <395E2425.8F6F8757@uniserve.com> <395E4281.452C26DE@prescod.net>
Message-ID: <395EAF62.4A628573@uniserve.com>

Paul Prescod wrote:
> 
> Bob van der Poel wrote:
> >
> > ...
> >
> > Is this just a good habit being developed by those who think that
> > classes are better than functions, or god forbid, inline code?
> 
> Given what you've said, that one-function class looks pretty bizarre. Is
> there maybe something else useful about it that you haven't noticed?
> Classes are not better than functions. They are just different. In most
> cases, you use classes when you need to combine data and instructions.

Yes, that I understand. In some of the example code I've seen there is
only the __init__ function and no data. That's why it doesn't make sense
to me. Maybe I'm just having a bad day.....

-- 
   __
  /  )      /         Bob van der Poel
 /--<  ____/__        bvdpoel at uniserve.com
/___/_(_) /_)         http://users.uniserve.com/~bvdpoel


From plusk at runet.edu  Sun Jul  2 04:52:40 2000
From: plusk at runet.edu (Paul D. Lusk)
Date: 2 Jul 2000 02:52:40 GMT
Subject: ? WinNT sys.stdout and re-direction
References: <8F6466ABEpluskrunetedu@137.45.128.146> <8F649A89Cgmcmhypernetcom@199.171.54.154>
Message-ID: <8F64E8ADDpluskrunetedu@137.45.128.146>

gmcm at hypernet.com (Gordon McMillan) wrote in 
<8F649A89Cgmcmhypernetcom at 199.171.54.154>:

>Paul D. Lusk wrote:
>
>>At the NT command prompt:
>>
>>script.py in.txt>out.txt
>>
>>gives me a empty out.txt
>>
>>Could somebody point me in the right direction to fix this?
>
>Your choices:
> 1) Upgrade to Win2k (hee hee)

  This does strike me as overkill.

> 2) say "python script.py in.txt >out.txt"
>
  Can do. Thanks! 
  
  Paul



From stephen_purcell at yahoo.com  Sun Jul  2 05:06:10 2000
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sun, 2 Jul 2000 13:06:10 +1000
Subject: [ANNOUNCE] PyUnit 1.2.0 test framework with GUI
Message-ID: <20000702130610.A31857@inkontact.com.au>

The latest release of the PyUnit unit testing framework features a
long overdue Tkinter based GUI akin to that of Java's JUnit. Should work
straight out-of-the-box with most Python installations.

Modules being tested are automatically reloaded each time they are run,
so the GUI can be left running while you code and debug.

As usual, interested parties can find out more at

  http://pyunit.sourceforge.net/

Go on, write some tests; you know you want to! <grin>

Pythonic best wishes to all,

-Steve

-- 
Steve Purcell,  Technical Director, Inkontact
Get in touch at http://www.inkontact.com/



From darrell at dorb.com  Sun Jul  2 08:15:30 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Sat, 1 Jul 2000 23:15:30 -0700
Subject: urllib and linksys router
Message-ID: <038701bfe3ec$ece104a0$6401a8c0@home>

Tried to use urllib to access my linksys router, which requires a password.
After much pain, it turns out the router needs to see the request as one
string.
Maybe it's a timing thing since urllib sends each line separately.

This works.

def direct():
    s = socket(AF_INET, SOCK_STREAM)
    s.connect("192.168.1.1",80)

    msgs=[
    'GET /Status.htm HTTP/1.0\015\012',
    'Authorization: Basic xxxxxxxx==\015\012',
    '\015\012',
    ]
    s.send(string.join(msgs,''))
    print s.recv(1024)
    print s.recv(1024)



--Darrell Gallion




From fiona at sitegnome.com  Sun Jul  2 05:19:26 2000
From: fiona at sitegnome.com (Fiona Czuczman)
Date: 2 Jul 2000 03:19:26 -0000
Subject: [FAQTS] Python Knowledge Base Update -- July 1st, 2000
Message-ID: <20000702031926.7937.qmail@synop.com>

Greetings All!

Below are the latest entries to be entered into http://python.faqts.com

regards,

Fiona Czuczman


## New Entries #################################################


-------------------------------------------------------------
What is bbox? In the context of the PhotoImage put(data, bbox).
http://www.faqts.com/knowledge-base/view.phtml/aid/4133
-------------------------------------------------------------
Fiona Czuczman
Keith Murphy

The bounding box is returned as a 4-tuple defining the left, upper,
right, and lower pixel coordinate.

Here's what I was doing..

self.photo.put(tuple(["#000000"],), tuple([5,5,6,6]))

it draws a single black pixel at (5,5)


-------------------------------------------------------------
How do you print the traceback string from an exception?
http://www.faqts.com/knowledge-base/view.phtml/aid/4134
-------------------------------------------------------------
Fiona Czuczman
J?rgen Hermann

try:
   whatever()
except:
            import traceback, string
            msg = string.join(traceback.format_exception(
                    sys.exc_info()[0], sys.exc_info()[1],
sys.exc_info()[2]), "")
            print msg


-------------------------------------------------------------
How can I concatenate external files (all the little *.htm files in a directory) into one big file?
http://www.faqts.com/knowledge-base/view.phtml/aid/4135
-------------------------------------------------------------
Fiona Czuczman
richard_chamberlain

import sys, string, glob, os

#First I create a list:
fls = glob.glob('C:\\wxpython\\wxPython-2.1.16\\docs\\wx\\wx*.htm')

#Then I guess I should open an output file for writing
outfile = open('temp.txt,'w')

for x in fls:
               file=open(x,'r')
               data=file.read()
               file.close()
               outfile.write(data)

outfile.close()

We then iterate over fls to give us each file name and open each one.
We then call the read method against it, which loads in the entire file
and assign that to data. we close the file and write the data to our
outfile. And then on to the next file. Finally we close the outfile.

You have to use read() with some caution because it (trys) to load the
whole file into memory which maybe an issue if it was sizable. In your
case it isn't an issue because they are simple html files.


-------------------------------------------------------------
What is lambda?
http://www.faqts.com/knowledge-base/view.phtml/aid/4136
-------------------------------------------------------------
Fiona Czuczman
Warren Postma, david_ullrich

1) Meta-Answer: Read the FAQ and Manual sections.

2) Instant Gratification Answer

Here's a standard function definition:

def func(x):
    return x *2

Here's a way to write the same thing, in effect, but using Lambda:

func = lambda (x): x * 2

While that example isn't useful, it gives you the idea.  Think of lambda 
as a way to generate a function and return that function as an object, 
without having to give it a name. I primarily use it to pass an 
algorithm or expression as a parameter to a function, or other places 
where I want to pass code in a variable instead of passing a reference 
to a method containing that code.

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

What it does is allow you to construct "anonymous functions".

   Oops, not my own words. The syntax

lambda x: [expression in x]

is itself an expression - the _value_ of the expression "lambda x: 
[expression in x]" is a function which returns [expression in x] when 
passed x. For example "lambda x: x+x" is a function that returns x+x; 
saying

f = lambda x: x+x

has the same effect as saying

def f(x):
  return x + x

But you don't usually use it that way - people use lambda when they want 
to pass a function to some routine without having to make up a name for 
the function. For example map takes a function as a parameter; saying

print map(lambda x: x+x, [1,2,3])

is the same as saying

def f(x):
  return x + x

print map(f, [1,2,3]])

You should note there's varying opinions on whether lambda is a good 
thing - if you're a beginner at programming as well as with Python it's 
not the first thing you should worry about.


## Edited Entries ##############################################


-------------------------------------------------------------
How can I use the smtplib module to send email from Python?
How can I send mail using the os module?
How can I send mail through python?
http://www.faqts.com/knowledge-base/view.phtml/aid/2607
-------------------------------------------------------------
Nathan Wallace, Fiona Czuczman, George Jansen
http://www.python.org/doc/FAQ.html#4.71,Justin Sheehy

On Unix, it's very simple, using sendmail. The location of the sendmail
program varies between systems; sometimes it is /usr/lib/sendmail,
sometime /usr/sbin/sendmail. The sendmail manual page will help you out.
Here's some sample code: 

  SENDMAIL = "/usr/sbin/sendmail" # sendmail location
  import os
  p = os.popen("%s -t" % SENDMAIL, "w")
  p.write("To: cary at ratatosk.org\n")
  p.write("Subject: test\n")
  p.write("\n") # blank line separating headers from body
  p.write("Some text\n")
  p.write("some more text\n")
  sts = p.close()
  if sts != 0:
      print "Sendmail exit status", sts

Check out the os module doc for more info:

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


On non-Unix systems (and on Unix systems too, of course!), you can use
SMTP to send mail to a nearby mail server. A library for SMTP
(smtplib.py) is included in Python 1.5.1; in 1.5.2 it will be documented
and extended. Here's a very simple interactive mail sender that uses 
it: 

    import sys, smtplib

    fromaddr = raw_input("From: ")
    toaddrs  = string.splitfields(raw_input("To: "), ',')
    print "Enter message, end with ^D:"
    msg = ''
    while 1:
        line = sys.stdin.readline()
        if not line:
            break
        msg = msg + line

    # The actual mail send
    server = smtplib.SMTP('localhost')
    server.sendmail(fromaddr, toaddrs, msg)
    server.quit()

This method will work on any host that supports an SMTP listener;
otherwise, you will have to ask the user for a host.

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

On Windows 9x systems, you may also use the COM interface. We have 
recently used PythonWin to (what else?) spam a number of recipients 
over Novell GroupWise. It took very little time to work this out.






From tim_one at email.msn.com  Sun Jul  2 05:38:09 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 1 Jul 2000 23:38:09 -0400
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
In-Reply-To: <037101bfe3e4$8558c820$6401a8c0@home>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEKBGHAA.tim_one@email.msn.com>

[Paul Prescod, links to ...]
> http://www.activestate.com/Products/VisualPython.html

[Darrell Gallion]
> Thanks for pointing this out. I see it comes up on google also.
>
> This should make the job of Python evangelizer that much simpler.
> Some people will feel that VC support puts Python on main street.

Well, that's just silly.  What it *does* do is validate Microsoft as a
serious competitor in the modern computer marketplace <wink>.

languages-live-longer-than-platforms-ly y'rs  - tim





From dirck.IsGettingTooMuchSpam at pacbell.net  Sun Jul  2 05:53:12 2000
From: dirck.IsGettingTooMuchSpam at pacbell.net (Dirck Blaskey)
Date: Sat, 1 Jul 2000 20:53:12 -0700
Subject: ? WinNT sys.stdout and re-direction
References: <8F6466ABEpluskrunetedu@137.45.128.146>
Message-ID: <J0z75.214$E63.94053@news.pacbell.net>

Paul D. Lusk <plusk at radford.edu> wrote in message
news:8F6466ABEpluskrunetedu at 137.45.128.146...
> ...
> At the NT command prompt:
>
> script.py in.txt
>
> writes output to the screen,
>
> but
>
> script.py in.txt>out.txt
>
> gives me a empty out.txt

If you'd like to retain the short-hand version
(as opposed to the long-hand 'python script.py >out.txt')
check out http://www.danbala.com/python/cmdfix

d

______________________

Dirck Blaskey
Danbala Software
http://www.danbala.com
______________________





From jwbnews at scandaroon.com  Sun Jul  2 05:52:33 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Sat, 01 Jul 2000 20:52:33 -0700
Subject: Class vrs. function
References: <395E2425.8F6F8757@uniserve.com> <395E4281.452C26DE@prescod.net> <395EAF62.4A628573@uniserve.com>
Message-ID: <jwbnews-8419AA.20523301072000@news.olympus.net>

In article <395EAF62.4A628573 at uniserve.com>, Bob van der Poel 
<bvdpoel at uniserve.com> wrote:

> Paul Prescod wrote:
> > 
> > Bob van der Poel wrote:
> > >
> > > ...
> > >
> > > Is this just a good habit being developed by those who think that
> > > classes are better than functions, or god forbid, inline code?
> > 
> > Given what you've said, that one-function class looks pretty bizarre. Is
> > there maybe something else useful about it that you haven't noticed?
> > Classes are not better than functions. They are just different. In most
> > cases, you use classes when you need to combine data and instructions.
> 
> Yes, that I understand. In some of the example code I've seen there is
> only the __init__ function and no data. That's why it doesn't make sense
> to me. Maybe I'm just having a bad day.....

Is there anything about those classes such that deriving from them would 
make sense?  If not, it's probably just a style issue (someone coming 
from SmallTalk--to pick one example--where there isn't any code outside 
some class might well do that).

Also, if the module is going to be imported, the code is handier if it 
isn't at the top level of the module:  import time may well not be the 
right time for the code to be run.  That wouldn't explain class vs 
function, though.

   --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From invalid.address at 127.0.0.1  Sun Jul  2 06:29:54 2000
From: invalid.address at 127.0.0.1 (David)
Date: Sun, 02 Jul 2000 04:29:54 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <395a7440.132798875@nntp.interaccess.com> <39597A25.8BF553C@san.rr.com> <slrn8lj56i.l52.thomas@centurion.xs4all.nl> <395d728e.117049078@nntp.interaccess.com> <395D7E14.B18EC77D@san.rr.com>
Message-ID: <395fc536.9838883@news.telus.net>

On Sat, 01 Jul 2000 05:10:50 GMT, Courageous <jkraska1 at san.rr.com> wrote:

>Java is is a variant of the C++ language with the kitchen
>think glued on

Heh.  Pun, or typo?



From neo_brave at my-deja.com  Sun Jul  2 07:20:56 2000
From: neo_brave at my-deja.com (Neo Brave)
Date: Sun, 02 Jul 2000 05:20:56 GMT
Subject: Win32 : Making script.py executable in DOS box
Message-ID: <8jmjfe$64o$1@nnrp1.deja.com>

Hi all,

I'm currently running Python scripts at the command line
using C:\python\python script.py. Is there a way of making
the MS-DOS box recognise the Python script as an executable
file so I can just use the name "script.py"?

Thanks in advance,
Neo
--
Neo Brave


Sent via Deja.com http://www.deja.com/
Before you buy.


From morten at src.no  Sun Jul  2 08:01:50 2000
From: morten at src.no (Morten W. Petersen)
Date: Sun, 2 Jul 2000 08:01:50 +0200
Subject: How to "fork off" a function
Message-ID: <Pine.LNX.4.21.0007020800520.5152-100000@lade.src.no>

How do I "get rid of" an annoyingly slow function, that's more of a
logging facility, in the code? Do I use thread, etc?

Short example would be appreciated.

-Morten

-------------------------------------------------------------
How to reply to email:
  http://home.sol.no/~vidaandr/news/OBSquoting.html (norsk)
  http://home.sol.no/~vidaandr/news/FAQquoting.html (english)

Who is the bigger fool? The fool or the fool who follows him?



From etoffi at softhome.net  Sun Jul  2 09:33:12 2000
From: etoffi at softhome.net (etoffi)
Date: Sun, 02 Jul 2000 03:33:12 -0400
Subject: mailman stuff
Message-ID: <395EF038.D8CEBD92@softhome.net>

visiting this page will print out the same thing 3 times. just thought someone
would like to know

http://www.python.org/mailman/subscribe/python-list/



From thomas at xs4all.nl  Sun Jul  2 09:31:28 2000
From: thomas at xs4all.nl (Thomas Wouters)
Date: 2 Jul 2000 07:31:28 GMT
Subject: How to "fork off" a function
References: <Pine.LNX.4.21.0007020800520.5152-100000@lade.src.no>
Message-ID: <slrn8ltrtr.lmm.thomas@centurion.xs4all.nl>

On Sun, 2 Jul 2000 08:01:50 +0200, Morten W. Petersen <morten at src.no> wrote:
>How do I "get rid of" an annoyingly slow function, that's more of a
>logging facility, in the code? Do I use thread, etc?

>Short example would be appreciated.

Well, this ought to work:

def long_function(*lots_of_args):
    # <original code commented out> 
    pass

Or isn't that what you wanted ? :-) Seriously, this is the only way to get
rid of a function without knowing that it's supposed to do. If the function
is a logging function, try to shortcut your way out of the function for the
cases where it isn't necessary. Or turn off logging. Or defer logging until
later, by appending log messsages to a list and writing out the list later.

You can use threads, yes, but it only makes sense if the logging function is
slow because it waits for something else, not if it takes a lot of cpu time.
I'd go for a seperate process and a syslogd-like daemon that listens for
messages, though: easier to write and much more portable.


lambda:0-ly y'rs,
Thomas.



From gherman at darwin.in-berlin.de  Sun Jul  2 10:07:14 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Sun, 02 Jul 2000 10:07:14 +0200
Subject: reportlab: HTML->PDF via Python?
References: <395CF8D0.6FD580B9@netscape.com>
Message-ID: <395EF832.E6F5C9CE@darwin.in-berlin.de>

Ian Oeschger wrote:
> 
> Does reportlab or some other Python module offer a facility for printing
> HTML pages to PDF?
> And if so, does some rendering engine have to intercede in this process
> somewhere? My experience so far has been getting text--rather than the
> rendered HTML--into PDF.

Well, not yet, at least from what I know about ReportLab.
The question came up there very recently. The problem is
more or less the implicit HTML layouting "behind the scenes",
like text around images, frames, and so on. It seems, though,
that a very simple HTML could be rendered rather easily into
PDF. 

Regards,

Dinu

-- 
Dinu C. Gherman
................................................................
"The only possible values [for quality] are 'excellent' and 'in-
sanely excellent', depending on whether lives are at stake or 
not. Otherwise you don't enjoy your work, you don't work well, 
and the project goes down the drain." 
                    (Kent Beck, "Extreme Programming Explained")


From esko.lehtonen at pp.htv.fi  Sun Jul  2 11:34:30 2000
From: esko.lehtonen at pp.htv.fi (Esko Lehtonen)
Date: Sun, 02 Jul 2000 12:34:30 +0300
Subject: PyGTK and draw_array
Message-ID: <395F0CA6.A20736F@pp.htv.fi>

I am trying to manipulate an image and then draw it to GtkDrawingArea
using PyGtk and Numerical Python. 

Here is description of draw_array function from file gtk.py. (I don't
know exactly what is that hassattr thing.)

   # this function renders an unsigned byte (type code 'b') numpy array
to a
   # drawable.  If it is a MxN or MxNx1 array, it is assumed to be a
grey image.
   # If it is a MxNx3 or MxNx4 array, it is assumed to be an rgb image.
   # It should accept images that have been sliced along their first two
axes,
   # if you want to only draw part of the image.
   if hasattr(_gtk, "gdk_draw_array"):
           def draw_array(drawable, gc, x, y, dither, array):
                   _gtk.gdk_draw_array(drawable, gc, x, y, dither,
array)

So I made following code:

WIDTH = 30
HEIGHT = 20
D = 3
# create rgb image as array, every value 100
img = ones((WIDTH, HEIGHT, D), 'b')
img = img * 100
----
darea = GtkDrawingArea()
---- 
draw_array(darea.get_window(),
	   widget.get_style().black_gc, x, y, 
           GDK.RGB_DITHER_NORMAL, img)


I still have a problem, because then a got: "TypeError: array data must
be unsigned bytes" when I'm trying to use the draw_array function. 

What is problem here? Or do you know alternative ways to do this using
PyGTK or some other graphics library?

Thanks in advance



--------------------------------------------------------------------
 Esko Lehtonen                   WYSINWYG -              
                                 What You See Is Never What You Get


From f98ba at efd.lth.se  Sun Jul  2 11:48:10 2000
From: f98ba at efd.lth.se (f98ba at efd.lth.se)
Date: 2 Jul 2000 09:48:10 GMT
Subject: Traping output from os.system
Message-ID: <8jn34q$kpn$1@news.lth.se>

Hi,

I'm executing a command with os.system and are
trying to capture the output into a python
variable. The only I can find to do it in
at the moment is to redirrect the output
to a file and the read it. That feels wrong,
isn't there another simple way??

Thanks,
Bjorn


From sean at digitome.com  Sun Jul  2 13:51:10 2000
From: sean at digitome.com (Sean Mc Grath)
Date: Sun, 02 Jul 2000 11:51:10 GMT
Subject: __getattr__ and pickle
Message-ID: <395f26a4.1832855@news.iol.ie>

I have a reproduceable GPF on NT when picking objects of a class
that has a __getattr__ implementation. Take out the __getattr__ and
the pickle works fine.

Are there ways to implement __getattr__ and yet preserve the ability
to pickle?

regards,



From sean at digitome.com  Sun Jul  2 13:53:00 2000
From: sean at digitome.com (Sean Mc Grath)
Date: Sun, 02 Jul 2000 11:53:00 GMT
Subject: __getattr__ and pickle
References: <395f26a4.1832855@news.iol.ie>
Message-ID: <395f2748.1996610@news.iol.ie>

The was a typo that could lead to confusion + a genuinely unforseen
and unintentional pun in the original post. I have fixed the typo but
left in the pun here:

-------------------------
I have a reproduceable GPF on NT when pickking objects of a class
that has a __getattr__ implementation. Take out the __getattr__ and
the pickle works fine.

Are there ways to implement __getattr__ and yet preserve the ability
to pickle?

regards,




From sean at digitome.com  Sun Jul  2 13:55:03 2000
From: sean at digitome.com (Sean Mc Grath)
Date: Sun, 02 Jul 2000 11:55:03 GMT
Subject: __getattr__ and pickle
References: <395f26a4.1832855@news.iol.ie> <395f2748.1996610@news.iol.ie>
Message-ID: <395f27ab.2095753@news.iol.ie>

Sheesh! I really need a holiday! Yet another typo fixed. Sorry
about this...

I have a reproduceable GPF on NT when pickling objects of a class
that has a __getattr__ implementation. Take out the __getattr__ and
the pickle works fine.

Are there ways to implement __getattr__ and yet preserve the ability
to pickle?

regards,





From thomas_s at ebox.tninet.se  Sun Jul  2 15:15:52 2000
From: thomas_s at ebox.tninet.se (thomas_s at ebox.tninet.se)
Date: Sun, 02 Jul 2000 14:15:52 +0100
Subject: Tkinter and default options
Message-ID: <8jnbhm$bb6$1@cubacola.tninet.se>

Hello,

I'm using Tkinter for the GUI parts of a little game I'm working on. What is
the best way of setting default options of the widgets?  For instance I
want all widgets to have fg='green' and bg='black'. Is there a standard
way of doing this? If not, I was thinking of maybe writing a widget factory
that creates and configures them for me.

Thank,
Thomas



From cut_me_out at hotmail.com  Sun Jul  2 14:31:15 2000
From: cut_me_out at hotmail.com (Alex)
Date: 02 Jul 2000 08:31:15 -0400
Subject: Problem with dynamically linked extensions that use Numeric
Message-ID: <etdbt0g1zos.fsf@w20-575-116.mit.edu>

Hi.  I have been writing an extension that uses the Numeric API.  Up
until now, I have just been building it straight into python, and it's
working fine.  However, I want to make it more widely available, so I
need to make shared objects from it.  The problem with that is that it
segfaults.  Here is a small piece of code that demonstrates the problem
I am having:

///////////////////////////////////////////////////////////////////////////////
#include "Python.h"
#include "arrayobject.h"

PyObject *test_array_test (PyObject *self, PyObject *args) {
  PyObject *t;
  int dimensions[2], i;
  
  if (!PyArg_ParseTuple (args, "")) {
    return NULL;
  }

  dimensions[0] = 10; dimensions[1] = 12;
  for (i=0; i < 20; i++) {
    printf ("testing\n");
    t = PyArray_FromDims (2, dimensions, PyArray_INT);
  }
  if (!t) {
    return NULL;
  }
  return t;
}

static PyMethodDef Test_ArrayMethods[] = {
  {"test", test_array_test, METH_VARARGS},
  {NULL, NULL}
};

void inittest_array () {
  (void) Py_InitModule ("test_array", Test_ArrayMethods);
}

///////////////////////////////////////////////////////////////////////////////

I compiled this using the 'make -f Makefile.pre.in boot' method.  Here
were the commands make generated:

[alex_c at puffin blast]$ make test_arraymodule.so
gcc  -I/scratch2/alex/include/python2.0/Numeric -g -O2 \
              -I/scratch2/alex/include/python2.0 \
              -I/scratch2/alex/include/python2.0 -DHAVE_CONFIG_H \
              -c ./test_arraymodule.c
gcc -shared  test_arraymodule.o  -o test_arraymodule.so

This code segfaults when it hits the line 
't = PyArray_FromDims (2, dimensions, PyArray_INT);'  But when it's
compiled directly in to python (i.e., I move the file into the Modules
directory of the python source, and add a corresponding entry in the
Setup file,) it works fine.

I'd be grateful for any ideas as to how to get this to work.

Alex.


From gmcm at hypernet.com  Sun Jul  2 15:07:44 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 02 Jul 2000 13:07:44 GMT
Subject: __getattr__ and pickle
References: <395f26a4.1832855@news.iol.ie> <395f2748.1996610@news.iol.ie> <395f27ab.2095753@news.iol.ie>
Message-ID: <8F655BA3Egmcmhypernetcom@199.171.54.154>

[posted and mailed]

Sean Mc Grath wrote:

>Sheesh! I really need a holiday! 

Probably <wink>!

>I have a reproduceable GPF on NT when pickling objects of a class
>that has a __getattr__ implementation. Take out the __getattr__ and
>the pickle works fine.
>
>Are there ways to implement __getattr__ and yet preserve the ability
>to pickle?

C'mon Sean! You've been around a while. You know when __getattr__ hacks go 
bad it's because they go recursive. (And the stack size on Windows is off 
by  some tiny amount, so it GPFs instead of traps.)

So what do we need:

>>> class A:
...  def __getattr__(self, nm):
...   print "get", nm
...   raise AttributeError, nm
...
>>> a = A()
>>> import pickle
>>> s = pickle.dumps(a)
get __getinitargs__
get __getstate__
>>> x = pickle.loads(s)
get __setstate__
>>> import cPickle
>>> s = cPickle.dumps(a)
get __getinitargs__
get __getstate__
>>> x = cPickle.loads(s)
get __setstate__
>>> 

Do you have a __getinitargs__ or __getstate__ or __setstate__ that might be 
triggering a call to __getattr__?

Does an Irishman like to talk?

- Gordon


From yleteigne1 at cybercable.fr  Sun Jul  2 16:06:11 2000
From: yleteigne1 at cybercable.fr (Yannick Le Teigner)
Date: Sun, 02 Jul 2000 16:06:11 +0200
Subject: CGI and images
Message-ID: <395F4C53.589EAFC0@cybercable.fr>

Hi,
I want to generate an image - plot and display it via a cgi script.
Creating the image is not hard, but I can't print the image on the page.

# this doesn't work the browser doesn't diplay errors but no images either...
print "Content-type: image/gif\n\n"
print open(myImage, "rb").read()

Any ideas?

Thank you!
Yannick





From MarkH at ActiveState.com  Sun Jul  2 16:45:10 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 02 Jul 2000 14:45:10 GMT
Subject: Win32 : Making script.py executable in DOS box
References: <8jmjfe$64o$1@nnrp1.deja.com>
Message-ID: <WzI75.5714$Tb7.31718@news-server.bigpond.net.au>

Under Windows NT or Win2k, you can use the PATHEXT environment variable.
Simply add ".py" to the semi-colon sep'd list of extensions, and off you
go!

No solution I am aware of for 95, although there are some tricks you can
pull (Im sure someone else will fill these in for you!

Mark.
--
markh at activestate.com - but if you think I speak for them, you dont know
Dick!

"Neo Brave" <neo_brave at my-deja.com> wrote in message
news:8jmjfe$64o$1 at nnrp1.deja.com...
> Hi all,
>
> I'm currently running Python scripts at the command line
> using C:\python\python script.py. Is there a way of making
> the MS-DOS box recognise the Python script as an executable
> file so I can just use the name "script.py"?
>
> Thanks in advance,
> Neo
> --
> Neo Brave
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From MarkH at ActiveState.com  Sun Jul  2 16:47:41 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 02 Jul 2000 14:47:41 GMT
Subject: Traping output from os.system
References: <8jn34q$kpn$1@news.lth.se>
Message-ID: <hCI75.5717$Tb7.31730@news-server.bigpond.net.au>

You want "popen".  For non-windows platforms, "os.popen" is where you can
find it.  For Win32 systems, you need the win32 extensions, and
"win32pipe.popen".
Mark.
--
markh at activestate.com - but if you think I speak for them, you dont know
Dick!

<f98ba at efd.lth.se> wrote in message news:8jn34q$kpn$1 at news.lth.se...
> Hi,
>
> I'm executing a command with os.system and are
> trying to capture the output into a python
> variable. The only I can find to do it in
> at the moment is to redirrect the output
> to a file and the read it. That feels wrong,
> isn't there another simple way??
>
> Thanks,
> Bjorn




From michael at stroeder.com  Sun Jul  2 16:42:07 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Sun, 02 Jul 2000 16:42:07 +0200
Subject: CGI and images
References: <395F4C53.589EAFC0@cybercable.fr>
Message-ID: <395F54BF.E3D9A0C4@stroeder.com>

Yannick Le Teigner wrote:
> 
> print "Content-type: image/gif\n\n"
                                ^^^^
Only one \n is needed if using print.

> print open(myImage, "rb").read()

Again: print adds an extra \n to the output. Use sys.stdout.write()
instead to have full control. If mixing print and sys.stdout.write()
you might have to use sys.stdout.flush() sometimes.

Ciao, Michael.


From nobody at nowhere.nohow  Sun Jul  2 17:57:35 2000
From: nobody at nowhere.nohow (Grant Edwards)
Date: Sun, 02 Jul 2000 15:57:35 GMT
Subject: Traping output from os.system
References: <8jn34q$kpn$1@news.lth.se> <hCI75.5717$Tb7.31730@news-server.bigpond.net.au>
Message-ID: <PDJ75.2521$iN5.605374@ptah.visi.com>

On Sun, 02 Jul 2000 14:47:41 GMT, Mark Hammond <MarkH at ActiveState.com> wrote:

>You want "popen".  For non-windows platforms, "os.popen" is where you can
>find it.  For Win32 systems, you need the win32 extensions, and
>"win32pipe.popen".

The commands module has a nice function that runs a program and
returns the exit status and output as a tuple.

-- 
Grant Edwards                   grante             Yow!  I represent a
                                  at               sardine!!
                               visi.com            


From seanmcgrath at my-deja.com  Sun Jul  2 17:45:58 2000
From: seanmcgrath at my-deja.com (seanmcgrath at my-deja.com)
Date: Sun, 02 Jul 2000 15:45:58 GMT
Subject: __getattr__ and pickle
References: <395f26a4.1832855@news.iol.ie> <395f2748.1996610@news.iol.ie> <395f27ab.2095753@news.iol.ie> <8F655BA3Egmcmhypernetcom@199.171.54.154>
Message-ID: <8jno3n$p6m$1@nnrp2.deja.com>

At 09:07 02/07/00 -0400, Gordon McMillan wrote:
>
>Do you have a __getinitargs__ or __getstate__ or
__setstate__ that might be
>triggering a call to __getattr__?
>

I don't have calls to __getinitargs__,
__getstate__ or __setstate__.

What I do have is a __getattr__ which I suspect is
involved in this:-)

The patient is pyxie.py
(http://www.digitome.com/pyxie.py).
The object I am trying to pickle is a tree
structure.
A variable CurPos points to a sub-ordinate object
which can be of type xElement or xData.

The __getattr__ hook achieves a Delphi-like effect
in which attributes of these sub-ordinate objects
appear as attributes of the top level object.

It seemed like a good idea at the time:-)

Here is a test program that reproduces the
problem:-

from pyxie import *
try:
	import cPickle
	pickle = cPickle
except ImportError:
	import pickle

t = xTree()

s = pickle.dumps (t)
<!-- This works fine -->
print s
<!-- Boom!!! -->
t1 = pickle.loads(s)


When I try and pickle using the Python
implementation of pickle,
I get an exception:
	No attribute '__getstate__' on xTree or
current xNode.

I don't follow this as I thought __getstate__ was
for fine
tuning the information to be pickled.

Using cPickle I don't get an exception at save
time but
instead get a GPF at load time.


BTW, I have a test version of Pyxie with the
__getattr__
stuff taken out and it pickled just fine. Picked
versions
of parsed XML trees load really quickly with it.

What I am implementing is analagous to
Pythons .py/.pyc
idea. When Pyxie loads a tree it loads the binary
version unless the plain text version has been
modified more recently. When saving, it saves
both binary and plain text XML every time.
Works great!

>Does an Irishman like to talk?
>
Only on days that end in a "y":-)


regards,



Sent via Deja.com http://www.deja.com/
Before you buy.


From ingo.adler at synacon.ch  Sun Jul  2 19:15:43 2000
From: ingo.adler at synacon.ch (Ingo Adler)
Date: Sun, 02 Jul 2000 18:15:43 +0100
Subject: Memory leaks
Message-ID: <395F78BF.CBEAA090@synacon.ch>

Hi,

I have some simple Python-Code which generated memory leaks (ca. 4MB):
(The real code is more complicated, but I could strip it down to this
simple example = the smallest example with the memory leak.)
//---------------------------
x = X()

for i in range(100000):
    y = x.getY()
    y.getNumber()
//---------------------------

It doesn't generate memory leaks when:

//---------------------------
x = X()

y = x.getY()
for i in range(100000):
    y.getNumber()
//---------------------------

or

//---------------------------
x = X()

for i in range(100000):
    y = x.getY()
//---------------------------

So getY and getNumber don't generate memory leaks on their own.

Details:

X and Y are implemented in C++ and connected to python via Swig-Shadow
classes.

The Implementation of  X:

class X {
    Y* pY;
public:
    X() {
        pY = new Y;
    }

    ~X() {
        delete pY;
    }

    Y* getY() {
        return pY;
    }
};

The Implementation of Y:

class Y {
    float getNumber() {
        return 3.4;
    }
};

The wrapper code it generated straight forward - nothing special,
nothing hidden.

I use Python 1.5.2, Borland C++ Builder 5.0 with statically linked
Python. My code is much bigger and works pretty well - except the memory
leaks. So the basic integration works.

I hope someone has a solution!

Ingo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ingo.adler.vcf
Type: text/x-vcard
Size: 323 bytes
Desc: Card for Ingo Adler
Url : http://mail.python.org/pipermail/python-list/attachments/20000702/807a33a5/attachment.vcf 

From richard_chamberlain at ntlworld.com  Sun Jul  2 19:19:23 2000
From: richard_chamberlain at ntlworld.com (richard_chamberlain)
Date: Sun, 2 Jul 2000 18:19:23 +0100
Subject: Tkinter and default options
References: <8jnbhm$bb6$1@cubacola.tninet.se>
Message-ID: <FPK75.3538$4C4.75025@news2-win.server.ntlworld.com>

Hi Thomas,

You can use an option database. If you create a file called optionDB with
for example the following kind of entries:

*font:                                Times New Roman
*Button*foreground            blue
*foreground                        black

You can then apply these via

root.option_readfile('optionDB')

Richard

<thomas_s at ebox.tninet.se> wrote in message
news:8jnbhm$bb6$1 at cubacola.tninet.se...
> Hello,
>
> I'm using Tkinter for the GUI parts of a little game I'm working on. What
is
> the best way of setting default options of the widgets?  For instance I
> want all widgets to have fg='green' and bg='black'. Is there a standard
> way of doing this? If not, I was thinking of maybe writing a widget
factory
> that creates and configures them for me.
>
> Thank,
> Thomas
>




From tismer at appliedbiometrics.com  Sun Jul  2 19:22:02 2000
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Sun, 02 Jul 2000 19:22:02 +0200
Subject: __getattr__ and pickle
References: <395f26a4.1832855@news.iol.ie> <395f2748.1996610@news.iol.ie> <395f27ab.2095753@news.iol.ie> <8F655BA3Egmcmhypernetcom@199.171.54.154> <8jno3n$p6m$1@nnrp2.deja.com>
Message-ID: <395F7A3A.FFB0B903@appliedbiometrics.com>

Hi Sean,

seanmcgrath at my-deja.com wrote:
> 
> At 09:07 02/07/00 -0400, Gordon McMillan wrote:
> >
> >Do you have a __getinitargs__ or __getstate__ or
> __setstate__ that might be
> >triggering a call to __getattr__?
> >
> 
> I don't have calls to __getinitargs__,
> __getstate__ or __setstate__.
> 
> What I do have is a __getattr__ which I suspect is
> involved in this:-)
> 
> The patient is pyxie.py
> (http://www.digitome.com/pyxie.py).
> The object I am trying to pickle is a tree
> structure.
> A variable CurPos points to a sub-ordinate object
> which can be of type xElement or xData.
> 
> The __getattr__ hook achieves a Delphi-like effect
> in which attributes of these sub-ordinate objects
> appear as attributes of the top level object.
> 
> It seemed like a good idea at the time:-)

Well, I see that your __getattr__ is a little insane.

        def __getattr__(self,n):
                if hasattr(self.CurPos,n):
                        return getattr(self.CurPos,n)
                else:
                        raise PyxieException (
                        "No attribute '%s' on xTree or current xNode" % n)

It redirects *every* reading attribute access to the CurPos
object, so Pickle can't even figure out what your xTree's
class is. Also, the CurPos object can's be saved with it,
nothing is accessible.

First, I suggest to redirect all requests with __ to the class,
second, give access to xTree object attributes a higher priority
by first looking into self by self.__dict__.has_key(n) .

This might work. Maybe you also want to avoid to store
CurPos in the pickle at all. This can be done by introducing
__getinitargs__ or __getstate__ and __setstate__ .

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From david_ullrich at my-deja.com  Sun Jul  2 19:55:38 2000
From: david_ullrich at my-deja.com (david_ullrich at my-deja.com)
Date: Sun, 02 Jul 2000 17:55:38 GMT
Subject: Class vrs. function
References: <395E2425.8F6F8757@uniserve.com>
Message-ID: <8jnvml$2at$1@nnrp1.deja.com>

In article <395E2425.8F6F8757 at uniserve.com>,
  Bob van der Poel <bvdpoel at uniserve.com> wrote:
>
> I see sample code for python/tkinter which looks like this:
>
> 	class main:
> 		def __init__(self):
> 			...do a bunch of init stuff
>
> 	a=main()
> 	root.mainloop()
>
> What I'm wonder is why the use of a class for this.

   You say it's "sample code". One possibility is that
sample code is not supposed to do what it does in the
best possible way, it's supposed to do what it does in
the most _illustrative_ possible way - it's supposed to
be a sample of how to do things in other situations
where the class will have more interesting methods.

   (Another possibility is that you overlooked the fact
that the class in question derives from another class,
so it actually does include more than just __init__,
you just don't see it on that page.)

DU

> Wouldn't it be
> just
> as clear and perhaps more efficient to do
>
> 	...do init stuff
> 	root.mainloop()
>
> Is this just a good habit being developed by those who think that
> classes are better than functions, or god forbid, inline code?
>
> While I'd asking, are there any penalties in the speed/memory usage
> departments on using classes vrs. functions?
>
> In case this isn't obvious I'm an old programmer who figured that C
was
> a really nice, almost-cross-platform assembly language....
>
> --
>    __
>   /  )      /         Bob van der Poel
>  /--<  ____/__        bvdpoel at uniserve.com
> /___/_(_) /_)         http://users.uniserve.com/~bvdpoel
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From wjdandreta at worldnet.att.net  Sun Jul  2 20:18:31 2000
From: wjdandreta at worldnet.att.net (William Dandreta)
Date: Sun, 02 Jul 2000 18:18:31 GMT
Subject: Problem compiling Python 1.5.2?
Message-ID: <XHL75.3043$oh5.199044@bgtnsc04-news.ops.worldnet.att.net>

I am having the following problem getting Python 1.5.2 to work correctly.

I am able to compile and link OK.

When I run Python with the -v switch the following messages appear
-----------------------------------------------------------------
'import exceptions'  failed;  traceback:
ImportError: No modules named exceptions
Warning!  Falling back to string based exceptions
#c:\pyth152\python~1.2\lib\site.pyc matches
c:\pyth152\python~1.2\lib\site.py
import site # precompiled from c:\pyth152\python~1.2\lib\site.pyc
'import site'  failed; traceback:
Traceback (innermost last):
  File "c:\pyth152\python~1.2\lib\site.py", line 60 in ?
    import sys, os
SystemError bad argument to internal function
Python 1.5.2 (#0, Jul 2 2000, 13:27:07) [Borland] on dos
Copyright 1991-1995 Stichting Mathematisch  Centrum, Amsterdam
>>>
----------------------------------------------------------------------------
-----------
The exceptions.py file is in the search directory.  Could it be a DOS name
mangling problem?

I don't know what the SystemError means.

The Copyright does not cover to current date, I don't know if that indicates
a problem.
----------------------------------------------------------------------------
-------------------------
I played around with some arithmetic, strings, slices etc. and it all worked
correctly, so I think I am pretty close to getting it right.

When I exit Python, Win98 becomes unstable. It either locks up immediately
or a short time later in another program.

I am sure I did something wrong when I compiled but I don't have a clue as
to what that might be nor do I know what info would be useful in tracing
down the problem(s).

Any suggestions would be appreciated.

Bill




From kmp at atrium.fsnet.co.uk  Sun Jul  2 22:09:51 2000
From: kmp at atrium.fsnet.co.uk (Kenneth Payne)
Date: Sun, 02 Jul 2000 20:09:51 GMT
Subject: Problem Installing PyQt-012
Message-ID: <20000702.20095100@linux.local>

I'm trying to install the Python bindings to Qt/KDE. I've downloaded 
the latest sources for Sip. PyQT and pyKDE. I've compiled and 
installed Sip and compiled PyQT. All seemed to go Ok, though I had to 
fiddle a bit to get them to compile on Suse 6.3. However, on the "make 
install" for PyQt, I get an error when the install routine hits a 
Python script:

Traceback (innermost last):
  File "<string>", line 1, in ?
  File "/usr/lib/python1.5/qt.py", line 16, in ?
    import libqtc
ImportError: /usr/lib/python1.5/lib-dynload/libqtcmodule.so: undefined 
symbol: __ti8sipProxy
make[3]: *** [install-data-hook] Error 1
make[2]: *** [install-data-am] Error 2
make[1]: *** [install-am] Error 2
make: *** [install-recursive] Error 1

I've obviously screwed up in either the configure scripts or there's 
something not set in my environment. Anyone know what I've done wrong?

-- Ken





From moshez at math.huji.ac.il  Sun Jul  2 21:20:29 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sun, 2 Jul 2000 22:20:29 +0300 (IDT)
Subject: Internet Explorer and Python/COM
Message-ID: <Pine.GSO.4.10.10007022217420.19085-100000@sundial>

I'm really a beginner in Windows COM programming, so I'm sorry if my
question doesn't make sense.

Can I use Python to run Internet Explorer and control it via COM? I mean,
telling it things like "Display this URL" or "Render this HTML".
(Maybe even putting a hook on URL displaying, so my Python program is
notified when the URL the browser is displaying changes)

--
Moshe Zadka <moshez at math.huji.ac.il>
There is no GOD but Python, and HTTP is its prophet.




From bojar at speakeasy.org  Sun Jul  2 22:08:02 2000
From: bojar at speakeasy.org (E. Jordan Bojar)
Date: Sun, 02 Jul 2000 16:08:02 -0400
Subject: Python CGI security
Message-ID: <B5851962.5E%bojar@speakeasy.org>

I'm interested in running a series of Python scripts via the Apache Python
module, mostly for dynamic content but possibly for password protection as
well.  Are there security issues particular to (or common with) Python CGIs,
in the way that buffer overflow is so common to poorly-written Perl?

-- 

E. Jordan Bojar




From tim_one at email.msn.com  Sun Jul  2 22:17:07 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 2 Jul 2000 16:17:07 -0400
Subject: Bug in math.frexp?
In-Reply-To: <20000630221759.A1965@xyz.donetsk.ua>
Message-ID: <LNBBLJKPBEHFEDALKOLCCELGGHAA.tim_one@email.msn.com>

[Kirill Simonov]
> $ python
> Python 1.6a2 (#2, Jun 24 2000, 12:18:54)  [GCC egcs-2.91.66
> 19990314/Linux (egcs-1.1.2 release)] on linux2
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import math
> >>> print math.frexp.__doc__
> frexp(x)
>
> Return the matissa and exponent for x. The mantissa is positive.
> >>> math.frexp(-1)
> (-0.5, 1)
>
> In this example the mantissa is negative.

The docs are wrong.  I just checked in a repaired version of mathmodule.c.
Fred, would you please get the truth into the Library Reference Manual's
math module docs too?

C:\pysrc\python\dist\src\PCbuild>python
Python 2.0b1 (#0, Jul  1 2000, 11:28:35) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
Copyright 1995-2000 Corporation for National Research Initiatives (CNRI)
>>> import math
>>> print math.frexp.__doc__
frexp(x)

Return the matissa and exponent of x, as pair (m, e).
m is a float and e is an int, such that x = m * 2.**e.
If x is 0, m and e are both 0.  Else 0.5 <= abs(m) < 1.0.
>>> math.frexp(0), math.frexp(-2), math.frexp(4)
((0.0, 0), (-0.5, 2), (0.5, 3))
>>>





From paul at prescod.net  Sun Jul  2 23:08:06 2000
From: paul at prescod.net (Paul Prescod)
Date: Sun, 02 Jul 2000 16:08:06 -0500
Subject: Class vrs. function
References: <395E2425.8F6F8757@uniserve.com> <395E4281.452C26DE@prescod.net> <395EAF62.4A628573@uniserve.com>
Message-ID: <395FAF36.5EE57FBC@prescod.net>

Bob van der Poel wrote:
> 
>...
> 
> Yes, that I understand. In some of the example code I've seen there is
> only the __init__ function and no data. That's why it doesn't make sense
> to me. Maybe I'm just having a bad day.....

Is this code public? It doesn't seem particularly useful for all of us
to just guess why they MIGHT have done it that way.

-- 
 Paul Prescod - Not encumbered by corporate consensus
The calculus and the rich body of mathematical analysis to which it 
gave rise made modern science possible, but it was the algorithm that 
made the modern world possible.
	- The Advent of the Algorithm (pending), by David Berlinski



From newsbunny at noether.freeserve.co.uk  Sun Jul  2 23:06:09 2000
From: newsbunny at noether.freeserve.co.uk (Martin P Holland)
Date: Sun, 2 Jul 2000 22:06:09 +0100
Subject: Problem Installing PyQt-012
References: <20000702.20095100@linux.local>
Message-ID: <slrn8lvbm1.o1d.newsbunny@noether.freeserve.co.uk>

On Sun, 02 Jul 2000 20:09:51 GMT,
Kenneth Payne <kmp at atrium.fsnet.co.uk> wrote:
>I'm trying to install the Python bindings to Qt/KDE. I've downloaded 
>the latest sources for Sip. PyQT and pyKDE. I've compiled and 
>installed Sip and compiled PyQT. All seemed to go Ok, though I had to 
>fiddle a bit to get them to compile on Suse 6.3. However, on the "make 
>install" for PyQt, I get an error when the install routine hits a 
>Python script:
>
>Traceback (innermost last):
>  File "<string>", line 1, in ?
>  File "/usr/lib/python1.5/qt.py", line 16, in ?
>    import libqtc
>ImportError: /usr/lib/python1.5/lib-dynload/libqtcmodule.so: undefined 
>symbol: __ti8sipProxy
>make[3]: *** [install-data-hook] Error 1
>make[2]: *** [install-data-am] Error 2
>make[1]: *** [install-am] Error 2
>make: *** [install-recursive] Error 1

Did you run ldconfig after installing sip?
Did you specify the qtdir to ./configure when compiling sip?

atb
Martin


-- 
http://www.noether.freeserve.co.uk
http://www.kppp-archive.freeserve.co.uk


From dale at out-think.NOSPAMco.uk  Sun Jul  2 23:10:55 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Sun, 2 Jul 2000 22:10:55 +0100
Subject: Threads?
References: <8jlnkh$7h$1@supernews.com> <8jlvat$ohi$1@nntp9.atl.mindspring.net>
Message-ID: <8job5d$6i$1@supernews.com>

You mean I have a choice? It supports both?

Either or both, is the answer.

I want separate execution paths that run concurrently to (or appear to) the
main one but share the same code.

The ability for the processes/threads to share memory and signal each other
would be a bonus.

Thanks

--
Dale Strickland-Clark
Out-Think Ltd, UK
Business Technology Consultants

Aahz Maruch <aahz at netcom.com> wrote in message
news:8jlvat$ohi$1 at nntp9.atl.mindspring.net...
> In article <8jlnkh$7h$1 at supernews.com>,
> Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> wrote:
> >
> >Can Python execute a function from the same module in another process?
> >I might do this, for example, to take advantage of multiple processors.
>
> I'm a bit confused here.  The Subject says "Threads?" but your text says
> "process".  Which is it?
> --
>                       --- Aahz (Copyright 2000 by aahz at netcom.com)
>
> Androgynous poly kinky vanilla queer het    <*>
http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
>
> The best way to get information on Usenet is not to ask a question,
> but to post the wrong information.  --Aahz




From pmontrasi at tiscalinet.it  Sun Jul  2 23:20:24 2000
From: pmontrasi at tiscalinet.it (Paolo Montrasi)
Date: Sun, 02 Jul 2000 23:20:24 +0200
Subject: CGI POST on Win98
Message-ID: <395FB218.5CCD76C8@tiscalinet.it>

I have installed PWS and Python 1.5.2 on my win98 machine, created a
simple form which runs the test() function on the cgi module, when i use
method GET in the form all works perfectly but when I try to use the
POST method the page never appears on my browser and I see python.exe in
my machine process list. The same thing works perfectly under NT. I have
searched the archive but I was not able to find help so I am thinking I
am wrong with something but I do not know where, can you help me?

ciao
Paolo


From iain at wombatzone.freeserve.co.uk  Sun Jul  2 23:29:16 2000
From: iain at wombatzone.freeserve.co.uk (Iain Calder)
Date: Sun, 02 Jul 2000 21:29:16 +0000
Subject: PyGTK and draw_array
References: <395F0CA6.A20736F@pp.htv.fi>
Message-ID: <395FB42C.89F435B0@wombatzone.freeserve.co.uk>

Esko Lehtonen wrote:
> 
> I am trying to manipulate an image and then draw it to GtkDrawingArea
> using PyGtk and Numerical Python.

<snip>
 
> So I made following code:
> 
> WIDTH = 30
> HEIGHT = 20
> D = 3
> # create rgb image as array, every value 100
> img = ones((WIDTH, HEIGHT, D), 'b')
> img = img * 100
> ----
> darea = GtkDrawingArea()
> ----
> draw_array(darea.get_window(),
>            widget.get_style().black_gc, x, y,
>            GDK.RGB_DITHER_NORMAL, img)
> 
> I still have a problem, because then a got: "TypeError: array data must
> be unsigned bytes" when I'm trying to use the draw_array function.

After multiplying by 100, img is not longer of type 'b'.  img.typecode()
returns 'i'.

Use img=img.astype('b') before the call to draw_array to cast back to
bytes.


HTH,

Iain Calder


From dnotspam-jen at personic.com  Mon Jul  3 00:19:29 2000
From: dnotspam-jen at personic.com (donotspam-jen@personic.com)
Date: Sun, 2 Jul 2000 15:19:29 -0700
Subject: Communicating between two computers
References: <395d48b3$0$21842@wodc7nh0.news.uu.net> <uem5e38if.fsf@ctwd0143.fitlinxx.com>
Message-ID: <395fbcd8$0$5627@wodc7nh0.news.uu.net>

I checked out the link, but from what I read, it sounds like this is a
client-side implementation only.  Is there more info on this somewhere?
Maybe some examples of how to use this?

Thanks!

Jen


"David Bolen" <db3l at fitlinxx.com> wrote in message
news:uem5e38if.fsf at ctwd0143.fitlinxx.com...
> Well, there's probably no end to the number of protocols and methods
> by which you could get the machines to talk, but given that it's
> Python invoking Python, you might take a peek at Fredrik Lundh's
> XML-RPC library.  XML-RPC was sort of the precursor to SOAP.  Similar
> in concept, but it's dirt simple to get two Python processes
> communicating (calling arbitrary methods with any Python objects and
> returning any Python objects).
>
> See http://www.pythonware.com/products/xmlrpc/index.htm.





From dnotspam-jen at personic.com  Mon Jul  3 00:19:38 2000
From: dnotspam-jen at personic.com (donotspam-jen@personic.com)
Date: Sun, 2 Jul 2000 15:19:38 -0700
Subject: Communicating between two computers
References: <395d48b3$0$21842@wodc7nh0.news.uu.net> <uem5e38if.fsf@ctwd0143.fitlinxx.com>
Message-ID: <395fbd8c$0$5622@wodc7nh0.news.uu.net>

I checked out the link, but from what I read, it sounds like this is a
client-side implementation only.  Is there more info on this somewhere?
Maybe some examples of how to use this?

Thanks!

Jen


"David Bolen" <db3l at fitlinxx.com> wrote in message
news:uem5e38if.fsf at ctwd0143.fitlinxx.com...
> Well, there's probably no end to the number of protocols and methods
> by which you could get the machines to talk, but given that it's
> Python invoking Python, you might take a peek at Fredrik Lundh's
> XML-RPC library.  XML-RPC was sort of the precursor to SOAP.  Similar
> in concept, but it's dirt simple to get two Python processes
> communicating (calling arbitrary methods with any Python objects and
> returning any Python objects).
>
> See http://www.pythonware.com/products/xmlrpc/index.htm.





From dnotspam-jen at personic.com  Mon Jul  3 00:22:29 2000
From: dnotspam-jen at personic.com (donotspam-jen@personic.com)
Date: Sun, 2 Jul 2000 15:22:29 -0700
Subject: Communicating between two computers
References: <395d48b3$0$21842@wodc7nh0.news.uu.net> <8jlmj0$r1q$1@supernews.com>
Message-ID: <395fbd8c$1$5622@wodc7nh0.news.uu.net>

Yes, I think for the first version I will do this (no need to deal with
multi-threading or the network this way). But after the initial version
works (hopefully later today), I'll try to move on to one of the other
methods described above, probably starting with XML-RPC if I can find more
info on it.

Thanks, all!

Jen

"Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> wrote in message
news:8jlmj0$r1q$1 at supernews.com...
> Start simple.
>
> Unless this is particularly time critical, I'd create a folder on the
Build
> server  and share it with suitably limited access.
>
> Give the Web server write access to the folder.
>
> The Web server writes a batch file or Python script to the shared
directory.
>
> Every few minutes the Build server checks this folder for a file and
> executes it.
>
> If you need some feedback, it could change the name of the file or move it
> to a subdirectory while it's running so the Web server can see it's
active.
>
> --
> Dale Strickland-Clark
> Out-Think Ltd, UK
> Business Technology Consultants
>
> donotspam-jen at personic.com <dnotspam-jen at personic.com> wrote in message
> news:395d48b3$0$21842 at wodc7nh0.news.uu.net...
> > FYI, I'm doing all of this on Windows NT, in case that makes a
difference
> > ...
> >
> > Is there a good way to get a script on one computer to talk to a script
on
> > another computer?  The process I'm thinking of is pretty simple ... I
want
> a
> > script on a Web server to be able to send a command to a build server to
> > start a build process.  I'd also like the Web server to be able to query
> the
> > build servers to find out their status (so that two builds do not run at
> > once on the same machine, for example).
> >
> > I've read a little about SOAP, but I think that my needs are probably a
> lot
> > simpler than that.  The easiest way to accomplish this is to setup a Web
> > server on each of the build machines, too, I guess ... but I was hoping
> that
> > I could just have the scripts that run on the build servers run as NT
> > services.
> >
> > Thanks!
> >
> > Jen
> >
> >
> >
> >
>
>




From dnotspam-jen at personic.com  Mon Jul  3 00:22:33 2000
From: dnotspam-jen at personic.com (donotspam-jen@personic.com)
Date: Sun, 2 Jul 2000 15:22:33 -0700
Subject: Communicating between two computers
References: <395d48b3$0$21842@wodc7nh0.news.uu.net> <8jlmj0$r1q$1@supernews.com>
Message-ID: <395fc264$0$5621@wodc7nh0.news.uu.net>

Yes, I think for the first version I will do this (no need to deal with
multi-threading or the network this way). But after the initial version
works (hopefully later today), I'll try to move on to one of the other
methods described above, probably starting with XML-RPC if I can find more
info on it.

Thanks, all!

Jen

"Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> wrote in message
news:8jlmj0$r1q$1 at supernews.com...
> Start simple.
>
> Unless this is particularly time critical, I'd create a folder on the
Build
> server  and share it with suitably limited access.
>
> Give the Web server write access to the folder.
>
> The Web server writes a batch file or Python script to the shared
directory.
>
> Every few minutes the Build server checks this folder for a file and
> executes it.
>
> If you need some feedback, it could change the name of the file or move it
> to a subdirectory while it's running so the Web server can see it's
active.
>
> --
> Dale Strickland-Clark
> Out-Think Ltd, UK
> Business Technology Consultants
>
> donotspam-jen at personic.com <dnotspam-jen at personic.com> wrote in message
> news:395d48b3$0$21842 at wodc7nh0.news.uu.net...
> > FYI, I'm doing all of this on Windows NT, in case that makes a
difference
> > ...
> >
> > Is there a good way to get a script on one computer to talk to a script
on
> > another computer?  The process I'm thinking of is pretty simple ... I
want
> a
> > script on a Web server to be able to send a command to a build server to
> > start a build process.  I'd also like the Web server to be able to query
> the
> > build servers to find out their status (so that two builds do not run at
> > once on the same machine, for example).
> >
> > I've read a little about SOAP, but I think that my needs are probably a
> lot
> > simpler than that.  The easiest way to accomplish this is to setup a Web
> > server on each of the build machines, too, I guess ... but I was hoping
> that
> > I could just have the scripts that run on the build servers run as NT
> > services.
> >
> > Thanks!
> >
> > Jen
> >
> >
> >
> >
>
>




From robin at jessikat.demon.co.uk  Sun Jul  2 16:25:21 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sun, 2 Jul 2000 15:25:21 +0100
Subject: tkinter hanging
Message-ID: <WRe0zXARD1X5EwXY@jessikat.demon.co.uk>

I'm having problems with zombie processes under win32 with a tkinter
linked against Tk 8.3. I seem to be exiting the mainloop ok and have
destroyed all subwindows and then the gui object itself, but
occasionally I still generate a zombie. I have checked that my after's
are cancelled and debug prints seem to show me exiting the script.

has anyone else had these kinds of problems? heeelllppp :( 
-- 
Robin Becker


From cb750chris at email.msn.com  Mon Jul  3 00:53:41 2000
From: cb750chris at email.msn.com (Roger Upole)
Date: Sun, 2 Jul 2000 18:53:41 -0400
Subject: Internet Explorer and Python/COM
References: <Pine.GSO.4.10.10007022217420.19085-100000@sundial>
Message-ID: <u$mxTiH5$GA.254@cpmsnbbsa07>

Here's a small sample of how to navigate to a URL.
>>> import win32com.client
>>> x=win32com.client.Dispatch('InternetExplorer.Application.1')
>>> x.Visible=1
>>> x.Navigate('http://www.python.org')
If you run makepy on Microsoft Internet Controls, the file generated in
Python\win32com\genpy will show all the properties and methods
available for this object.
    HTH
         Roger Upole


Moshe Zadka <moshez at math.huji.ac.il> wrote in message
news:Pine.GSO.4.10.10007022217420.19085-100000 at sundial...
>
> I'm really a beginner in Windows COM programming, so I'm sorry if my
> question doesn't make sense.
>
> Can I use Python to run Internet Explorer and control it via COM? I mean,
> telling it things like "Display this URL" or "Render this HTML".
> (Maybe even putting a hook on URL displaying, so my Python program is
> notified when the URL the browser is displaying changes)
>
> --
> Moshe Zadka <moshez at math.huji.ac.il>
> There is no GOD but Python, and HTTP is its prophet.
>
>






From rgoto at aa.net  Mon Jul  3 01:16:12 2000
From: rgoto at aa.net (Robert Goto)
Date: Sun, 2 Jul 2000 16:16:12 -0700
Subject: Newbie question  re: regsub
Message-ID: <395fc90a$1_1@huge.aa.net>

I am sure there is a very simple answer to this.
>>> import regsub
>>> fimo = open('fimo.txt')
>>> while 1:
 line = fimo.readline()
 if not line: break
 regsub.gsub('orange', 'red', line)
 print line


'red red red red red red red'
red orange red red red red red
>>>

The file I am opening which is called fimo.txt only contains the string
withouth ' marks  red orange red red red red red.

What I want it to do is print the changed string  only....but what I am
getting is both the original string and the changed string?!

Any pointers would be greatly appreciated.




From rgoto at aa.net  Mon Jul  3 01:30:40 2000
From: rgoto at aa.net (Robert Goto)
Date: Sun, 2 Jul 2000 16:30:40 -0700
Subject: Newbie question  re: regsub
References: <395fc90a$1_1@huge.aa.net>
Message-ID: <395fcc6d$1_2@huge.aa.net>

What is even more maddening to me is that when I save it to a module and run
it from NT command prompt......the output I get is the original file
contents only........like there was no regsub.gsub!!
"Robert Goto" <rgoto at aa.net> wrote in message
news:395fc90a$1_1 at huge.aa.net...
> I am sure there is a very simple answer to this.
> >>> import regsub
> >>> fimo = open('fimo.txt')
> >>> while 1:
>  line = fimo.readline()
>  if not line: break
>  regsub.gsub('orange', 'red', line)
>  print line
>
>
> 'red red red red red red red'
> red orange red red red red red
> >>>
>
> The file I am opening which is called fimo.txt only contains the string
> withouth ' marks  red orange red red red red red.
>
> What I want it to do is print the changed string  only....but what I am
> getting is both the original string and the changed string?!
>
> Any pointers would be greatly appreciated.
>
>




From darrell at dorb.com  Mon Jul  3 05:02:57 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Sun, 2 Jul 2000 20:02:57 -0700
Subject: Newbie question  re: regsub
References: <395fc90a$1_1@huge.aa.net> <395fcc6d$1_2@huge.aa.net>
Message-ID: <005e01bfe49b$68c94d30$6401a8c0@home>

>>> open('junk','w').write("red red red red")
>>> import re
>>> print re.sub("red","orange", open('junk').read())
orange orange orange orange
>>>

The changed string is returned. Strings in python don't change so the
original buffer is unaffected.

--Darrell


----- Original Message -----
From: "Robert Goto" <rgoto at aa.net>

> What is even more maddening to me is that when I save it to a module and
run
> it from NT command prompt......the output I get is the original file
> contents only........like there was no regsub.gsub!!





From bgriggs at pobox.com  Mon Jul  3 02:30:36 2000
From: bgriggs at pobox.com (B Griggs)
Date: Mon, 03 Jul 2000 00:30:36 GMT
Subject: urllib and linksys router    
References: <038701bfe3ec$ece104a0$6401a8c0@home>
Message-ID: <M8R75.3145$09.50946@typhoon.austin.rr.com>

Darrell,

Speaking of urllib and the LinkSys router...

http://greenman.dyndns.org/ls_dyndns

I have written a dyndns.org client for the linksys router.  Should work on
all Python platforms.

Regards,
Bobby

"Darrell Gallion" <darrell at dorb.com> wrote in message
news:038701bfe3ec$ece104a0$6401a8c0 at home...
> Tried to use urllib to access my linksys router, which requires a
password.
> After much pain, it turns out the router needs to see the request as one
> string.
> Maybe it's a timing thing since urllib sends each line separately.
>
> This works.
>
> def direct():
>     s = socket(AF_INET, SOCK_STREAM)
>     s.connect("192.168.1.1",80)
>
>     msgs=[
>     'GET /Status.htm HTTP/1.0\015\012',
>     'Authorization: Basic xxxxxxxx==\015\012',
>     '\015\012',
>     ]
>     s.send(string.join(msgs,''))
>     print s.recv(1024)
>     print s.recv(1024)
>
>
>
> --Darrell Gallion
>
>




From MarkH at ActiveState.com  Mon Jul  3 02:37:27 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Mon, 03 Jul 2000 00:37:27 GMT
Subject: Internet Explorer and Python/COM
References: <Pine.GSO.4.10.10007022217420.19085-100000@sundial> <u$mxTiH5$GA.254@cpmsnbbsa07>
Message-ID: <bfR75.5878$Tb7.33735@news-server.bigpond.net.au>

"Roger Upole" <cb750chris at email.msn.com> wrote in message
news:u$mxTiH5$GA.254 at cpmsnbbsa07...

> Here's a small sample of how to navigate to a URL.
> >>> import win32com.client
> >>> x=win32com.client.Dispatch('InternetExplorer.Application.1')
> >>> x.Visible=1
> >>> x.Navigate('http://www.python.org')
> If you run makepy on Microsoft Internet Controls, the file generated in
> Python\win32com\genpy will show all the properties and methods
> available for this object.

This will work fine to control IE.

However, look in "win32com\client\__init__.py", and look for the
"DispatchWithEvents" docstring - this has an example that uses IE to both
control it, and respond to events (such as URL load completed, etc)

Mark.





From swun at esec.com.au  Mon Jul  3 02:41:58 2000
From: swun at esec.com.au (Sam Wun)
Date: Mon, 03 Jul 2000 10:41:58 +1000
Subject: multiple inheritance in Python?
Message-ID: <395FE156.C0C13FBE@esec.com.au>

Does anyone have example of implement multiple inheritance in Python?



From fiona at sitegnome.com  Mon Jul  3 03:10:20 2000
From: fiona at sitegnome.com (Fiona Czuczman)
Date: 3 Jul 2000 01:10:20 -0000
Subject: [FAQTS] Python Knowledge Base Update -- July 2nd, 2000
Message-ID: <20000703011020.14359.qmail@synop.com>

Hi Guys.

The FAQTs knowledge base continues to grow.  Currently 265 questions and 
answers.

http://python.faqts.com

Cheers,

Fiona Czuczman


## New Entries #################################################


-------------------------------------------------------------
When executing a command with os.system how can I capture the output into a python variable?
http://www.faqts.com/knowledge-base/view.phtml/aid/4152
-------------------------------------------------------------
Fiona Czuczman
Mark Hammond, Grant Edwards

You want "popen".  For non-windows platforms, "os.popen" is where you 
can find it.  For Win32 systems, you need the win32 extensions, and
"win32pipe.popen".

.........

The commands module has a nice function that runs a program and returns 
the exit status and output as a tuple.


-------------------------------------------------------------
Is there a way of making a MS-DOS box recognise the Python script as an executable file so I can just use the name "script.py"?
http://www.faqts.com/knowledge-base/view.phtml/aid/4153
-------------------------------------------------------------
Fiona Czuczman
Mark Hammond

Under Windows NT or Win2k, you can use the PATHEXT environment variable.
Simply add ".py" to the semi-colon sep'd list of extensions, and off you
go!


-------------------------------------------------------------
What is the best way of setting default options of widgets?
http://www.faqts.com/knowledge-base/view.phtml/aid/4151
-------------------------------------------------------------
Fiona Czuczman
richard_chamberlain

You can use an option database. If you create a file called optionDB 
with for example the following kind of entries:

*font:                   Times New Roman
*Button*foreground       blue
*foreground              black

You can then apply these via

root.option_readfile('optionDB')






From cut_me_out at hotmail.com  Mon Jul  3 03:19:51 2000
From: cut_me_out at hotmail.com (Alex)
Date: 02 Jul 2000 21:19:51 -0400
Subject: multiple inheritance in Python?
References: <395FE156.C0C13FBE@esec.com.au>
Message-ID: <etdaeg0103s.fsf@w20-575-116.mit.edu>

> Does anyone have example of implement multiple inheritance in Python?

What sort of example are you looking for?  The basic syntax is pretty
straightforward, as I understand it.

>>> class A:
    t = 1

... ... >>> 
>>> class B:
    s = 2
... ... 
>>> class C(A, B):
    pass
... ... 
>>> C.t
1
>>> C.s
2
>>> 

C would also inherit methods from A and B, if they had any.

Alex.


From swun at esec.com.au  Mon Jul  3 03:56:41 2000
From: swun at esec.com.au (Sam Wun)
Date: Mon, 03 Jul 2000 11:56:41 +1000
Subject: how to push data in to ssh in python?
Message-ID: <395FF2D9.1B46251D@esec.com.au>

Thanks

Sam


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/python-list/attachments/20000703/9708364b/attachment.htm 

From swun at esec.com.au  Mon Jul  3 03:57:02 2000
From: swun at esec.com.au (Sam Wun)
Date: Mon, 03 Jul 2000 11:57:02 +1000
Subject: multiple inheritance in Python?
References: <395FE156.C0C13FBE@esec.com.au> <etdaeg0103s.fsf@w20-575-116.mit.edu>
Message-ID: <395FF2ED.60B6AD1D@esec.com.au>

Thanks.

Alex wrote:

> > Does anyone have example of implement multiple inheritance in Python?
>
> What sort of example are you looking for?  The basic syntax is pretty
> straightforward, as I understand it.
>
> >>> class A:
>     t = 1
>
> ... ... >>>
> >>> class B:
>     s = 2
> ... ...
> >>> class C(A, B):
>     pass
> ... ...
> >>> C.t
> 1
> >>> C.s
> 2
> >>>
>
> C would also inherit methods from A and B, if they had any.
>
> Alex.

--
Sam Wun                            Firewalls / Security
OneGuard Senior Software Engineer  Electronic Commerce
eSec Limited                       Phone: +61 3 83715376
mailto:swun at eSec.com.au            C++/JAVA/UNIX/OOP/OOD


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/python-list/attachments/20000703/232dd1d9/attachment.html 

From dgoodger at bigfoot.com  Mon Jul  3 04:00:50 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Sun, 02 Jul 2000 22:00:50 -0400
Subject: multiple inheritance in Python?
In-Reply-To: <395FE156.C0C13FBE@esec.com.au>
References: <395FE156.C0C13FBE@esec.com.au>
Message-ID: <B5856C11.5BDC%dgoodger@bigfoot.com>

on 2000-07-02 20:41, Sam Wun (swun at esec.com.au) wrote:
> Does anyone have example of implement multiple inheritance in Python?

My typelib.py module does (available at
http://gotools.sourceforge.net/sgflib in the sgflib.tgz archive). It defines
classes which emulate most of the Python built-in types (duplicating
functionality in UserList and UserDict, I know; I wrote it before I
"discovered" these two standard modules :). Multiple inheritance is emplyed
for common functionality reuse. For example, both the Number and String
classes support the "%" (modulo, or string formatting) operator, so that
functionality is implemented in the ModMixin class, which is inherited by
both classes along with other superclasses:

    class Number(SuperType, AddMulMixin, ModMixin):

    class Sequence(Container, AddMulMixin):

    class String(Sequence, ModMixin):

Hope this helps.

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From pauldubois at home.com  Mon Jul  3 04:40:03 2000
From: pauldubois at home.com (Paul F. Dubois)
Date: Mon, 03 Jul 2000 02:40:03 GMT
Subject: Getting double* from NumPy array?
References: <ted.drain-2606001323500001@mjordan.jpl.nasa.gov>
Message-ID: <72T75.24010$i5.229971@news1.frmt1.sfba.home.com>

Please see the appropriate chapter in the Numerical Documentation (Chapter
12, I believe). There is quite a bit to know and just telling you to access
x->data would not be doing you any favors. Please read the entire chapter
before proceeding.

The process is not very complicated but it must be done correctly. Don't
overlook the import_array statement.

Ted Drain <ted.drain at jpl.nasa.gov> wrote in message
news:ted.drain-2606001323500001 at mjordan.jpl.nasa.gov...
> I have some existing C/C++ code that can accept a double array ( int len,
> and double* ptr) to do some processing.  I want to allow users to create
> arrays in python using NumPy and then provide an interface to pass these
> arrays to my C/C++ code.
>
> Assuming a 1-dimensional NumPy array, is there an easy way to extract a
> pointer to a double and length from the python object?  The C/C++ isn't
> going to change the object so I'd prefer not to make a copy of it.
>
> Thanks
> Ted
>
> --
> Ted Drain   Jet Propulsion Laboratory    Ted.Drain at jpl.nasa.gov




From gmol at my-deja.com  Mon Jul  3 04:32:37 2000
From: gmol at my-deja.com (gmol at my-deja.com)
Date: Mon, 03 Jul 2000 02:32:37 GMT
Subject: Referring to method of a class without instance?
Message-ID: <8jou05$mei$1@nnrp1.deja.com>

In my application I use the observer/observable pattern..question
I hate duplicating the information of what was changed by making up a
new message type (global message variables or strings) and using tonnes
of if-else clauses.  I'd rather just tell my observer which method was
called and the arguments passed to it.  (I.e. I have atoms in 3d space,
when setPosition(position) is called I would just like to tell whoever
is interested that setPosition was called with the argument position))

Problem, how do I refer to the method of a given class without making an
instance?  Like suppose I notify my observers by giving the funciton
address and arguments, I would like the observer to have a dictionary
whose keys are methods

class Atom:
....


class myobserver:
...

   updateTable={....
       Atom.setPosition:handle_A_setPositio
    }
...

Hmm I have also just realized that given
class A:
 def __init__(self...)

a=A()

a.init is not equal to A.__methods__['__init__']

Hmmm I thought that would be the worst way I could do it, but I guess I
couldn't if I wanted to...

Any thoughts or better general solution would be appreciated...



Sent via Deja.com http://www.deja.com/
Before you buy.


From dberlin at cygnus.com  Mon Jul  3 05:09:14 2000
From: dberlin at cygnus.com (Daniel Berlin)
Date: 02 Jul 2000 23:09:14 -0400
Subject: Referring to method of a class without instance?
In-Reply-To: gmol@my-deja.com's message of "Mon, 03 Jul 2000 02:32:37 GMT"
References: <8jou05$mei$1@nnrp1.deja.com>
Message-ID: <86r99bewpx.fsf@dan2.cygnus.com>

gmol at my-deja.com writes:


> In my application I use the observer/observable pattern..question
> I hate duplicating the information of what was changed by making up a
> new message type (global message variables or strings) and using tonnes
> of if-else clauses.  I'd rather just tell my observer which method was
> called and the arguments passed to it.  (I.e. I have atoms in 3d space,
> when setPosition(position) is called I would just like to tell whoever
> is interested that setPosition was called with the argument position))
> 
> Problem, how do I refer to the method of a given class without making an
> instance?  Like suppose I notify my observers by giving the funciton
> address and arguments, I would like the observer to have a dictionary
> whose keys are methods
> 
> class Atom:
> ....
> 
> 
> class myobserver:
> ...
> 
>    updateTable={....
>        Atom.setPosition:handle_A_setPositio
>     }
> ...
> 
> Hmm I have also just realized that given
> class A:
>  def __init__(self...)
> 
> a=A()
> 
> a.init is not equal to A.__methods__['__init__']
> 
> Hmmm I thought that would be the worst way I could do it, but I guess I
> couldn't if I wanted to...
> 
> Any thoughts or better general solution would be appreciated...
>


This is a simple problem.

Try A.__init__
>>> class B:
...     def __init__(self):
...             pass
...
>>> B.__init__
<unbound method B.__init__>
>>>               __

If you try to call it without a class instance as the first parameter, you'll get an error.

>>> B.__init__(5)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: unbound method must be called with class instance 1st argument

Otherwise, it'll work fine.

>>> B.__init__(B())  

HTH,
Dan



From robertrutledge at my-deja.com  Mon Jul  3 05:10:23 2000
From: robertrutledge at my-deja.com (robertrutledge at my-deja.com)
Date: Mon, 03 Jul 2000 03:10:23 GMT
Subject: Getting remote memory configuration
References: <2B1262E83448D211AE4B00A0C9D61B03013011F8@msgeuro1.creo.be>
Message-ID: <8jp06o$nn8$1@nnrp1.deja.com>

Pieter,

Ask an operator at the remote site, via telephone.

Bob Rutledge



In article <2B1262E83448D211AE4B00A0C9D61B03013011F8 at msgeuro1.creo.be>,
  Pieter Claerhout <PClaerhout at CREO.BE> wrote:
> Forgot to mention that I was talking about Windows NT machines.
>
> Pieter
>
> Pieter Claerhout
> Application Support - CreoScitex Europe
>
> > -----Original Message-----
> > From:	Pieter Claerhout
> > Sent:	Monday, June 19, 2000 3:40 PM
> > To:	'python-list at python.org'
> > Subject:	Getting remote memory configuration
> >
> > Anyone an idea on how you can find out how much memory a remote
computer
> > has installed? The computer is available on the same network and
has the same
> > login as the computer on which you are running the command.
> >
> > Pieter
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From gmcm at hypernet.com  Mon Jul  3 05:40:57 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 03 Jul 2000 03:40:57 GMT
Subject: Referring to method of a class without instance?
References: <8jou05$mei$1@nnrp1.deja.com>
Message-ID: <8F65FB46Fgmcmhypernetcom@199.171.54.155>

gmol at my-deja.com wrote in <8jou05$mei$1 at nnrp1.deja.com>:

>In my application I use the observer/observable pattern..question
>I hate duplicating the information of what was changed by making up a
>new message type (global message variables or strings) and using tonnes
>of if-else clauses.  I'd rather just tell my observer which method was
>called and the arguments passed to it.  (I.e. I have atoms in 3d space,
>when setPosition(position) is called I would just like to tell whoever
>is interested that setPosition was called with the argument position))
>
>Problem, how do I refer to the method of a given class without making an
>instance?  Like suppose I notify my observers by giving the funciton
>address and arguments, I would like the observer to have a dictionary
>whose keys are methods
>
>class Atom:
>....
>
>
>class myobserver:
>...
>
>   updateTable={....
>       Atom.setPosition:handle_A_setPositio
>    }
>...
>
>Hmm I have also just realized that given
>class A:
> def __init__(self...)
>
>a=A()
>
>a.init is not equal to A.__methods__['__init__']
>
>Hmmm I thought that would be the worst way I could do it, but I guess I
>couldn't if I wanted to...

That's right: a.__init__ is a bound method, while A.__init__ is an unbound 
method. You can get to the latter from the former by:
 self.__class__.__init__
or more directly (of course) 
  A.__init__

But you might be headed for trouble. For example, you might later add a 
class Molecule with a setPosition method. Now you have to go through all 
your observers and update their dispatch tables. IOW, you're tightly 
coupling your observers to your observables.

OTOH, A.__init__ and a.__init__ both have __name__ attributes with 
singularly unexciting values <wink>. Then you just have to be sure you 
don't get sloppy and write Molecule.setPos(....).

Of course, since there's no special name for "the currently executing 
method", in effect you're still maintaining a global table of message 
identifiers.

- Gordon


From neo_brave at my-deja.com  Mon Jul  3 06:11:49 2000
From: neo_brave at my-deja.com (Neo Brave)
Date: Mon, 03 Jul 2000 04:11:49 GMT
Subject: Win32 : Making script.py executable in DOS box
References: <8jmjfe$64o$1@nnrp1.deja.com> <WzI75.5714$Tb7.31718@news-server.bigpond.net.au>
Message-ID: <8jp3pt$q7j$1@nnrp1.deja.com>

In article <WzI75.5714$Tb7.31718 at news-server.bigpond.net.au>,
  "Mark Hammond" <MarkH at ActiveState.com> wrote:
> Under Windows NT or Win2k, you can use the PATHEXT environment
> variable. Simply add ".py" to the semi-colon sep'd list of
> extensions, and off you go!

Great, thats a help for NT, but I also want to do this on Win98 :-).

> No solution I am aware of for 95, although there are some tricks
> you can pull (Im sure someone else will fill these in for you)!

Anyone???

Neo
--
Neo Brave


Sent via Deja.com http://www.deja.com/
Before you buy.


From aahz at netcom.com  Mon Jul  3 06:27:02 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 3 Jul 2000 04:27:02 GMT
Subject: Threads?
References: <8jlnkh$7h$1@supernews.com> <8jlvat$ohi$1@nntp9.atl.mindspring.net> <8job5d$6i$1@supernews.com>
Message-ID: <8jp4mm$ks$1@nntp9.atl.mindspring.net>

In article <8job5d$6i$1 at supernews.com>,
Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> wrote:
>
>I want separate execution paths that run concurrently to (or appear to)
>the main one but share the same code.
>
>The ability for the processes/threads to share memory and signal each
>other would be a bonus.

In that case, it'll be much simpler to use threads.  Check out
threading.Thread.  The one problem with Python threads is that there is
something called the "global interpreter lock".  This means that only
one thread can run Python code at any time, even on a multi-CPU machine.
However, threads that call out to extensions (such as I/O) can run truly
concurrently.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The best way to get information on Usenet is not to ask a question,
but to post the wrong information.  --Aahz


From mksql at my-deja.com  Mon Jul  3 07:19:49 2000
From: mksql at my-deja.com (Matt)
Date: Mon, 03 Jul 2000 01:19:49 -0400
Subject: Tutorials at IPC9
References: <39591877.535A96BC@sage.att.com>
Message-ID: <9i80ms4tbkjjhp90t55au011qquvmsr456@4ax.com>

On Tue, 27 Jun 2000 21:11:19 GMT, Garry Hodgson <garry at sage.att.com>
wrote:

>I'm starting to think about what kinds of tutorials we should have at
>the next Python conference.

Is there information available on when and where this will be?


From Martin.Kew at vsl.com.au  Mon Jul  3 07:27:11 2000
From: Martin.Kew at vsl.com.au (Martin Kew)
Date: Mon, 03 Jul 2000 14:57:11 +0930
Subject: Aspect-oriented Python?
In-Reply-To: <8jg821$vll$1@gaia.cdg.acriter.nl>
Message-ID: <4.2.0.58.20000703140633.00be3c30@imaphost>

After seeing the following email, I investigated the Aspect Oriented 
Programing (AOP)
as well.  I to think that the Python language people should have a look at it.

In summary an "Aspect" is a property/function of a system that cannot be 
cleanly
encapsulated by a traditional programming language, for example:

a) loop fusion and resource sharing for procedural algorithmic code,

b) network traffic minimization, synchronization handling and error handling
for distributed client/server systems.

An aspect is captured using an "aspect language" in terms of  "crosscuts"
which are the connection points into the traditional program. ie. the loop 
construct
for a loop fusion aspect.

An AOP implementation of a system consists of:

a) using a traditional language (called a component language) define and 
program
the components of the system, ie. procedures, objects etc. Call it a "component
program".

b) using an aspect language define all aspects that will be applied in the 
system in
terms of  "cross-cuts".

c) "weave" all "aspects" into the "component program" using an "aspect weaver"
compiler to form a tangled "aspect program".

For example, for the fusion of loops in an image processing system, there was a
(35213-756)/352 = 98 reduction in code bloat for the incorporation of loop 
fusion.

raw component program (LOC) ie. without loop fusion = 756
hand implemented code incorporating loop fusion (LOC) = 35213
all aspect programs (LOC) = 352

This information was unashamably paraphrased from "Aspect-Oriented Programming"
G Kiczales et.all of Xerox PARC in "European Conference on OO Programming" 
June 1997
available a www.aspectj.org

The reason for presenting this overview was the overwelming lack of 
interest in the first
email by Cees de Groot.


At 09:29 pm 29/06/00 +0200, you wrote:
>I've been reading on aspect-oriented programming a bit, and it sounds really
>interesting. A Java implementation, AspectJ (www.aspectj.org) looks promising,
>but has the drawback that it only supports static weaving (wouldn't know
>how to do dynamic weaving with Java, so it's not a big surprise).
>
>With Python, experimenting with aspect-oriented stuff would probably be
>way easier, because it is easy enough to replace class methods etcetera. I've
>tried Google, but didn't find anything. Is someone busy with aspect-oriented
>programming for Python?
>
>--
>Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
>GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B
>Forge your CipherSaber and list it: http://www.xs4all.nl/~cg/ciphersaber/
>--
>http://www.python.org/mailman/listinfo/python-list




From fdrake at beopen.com  Mon Jul  3 08:28:33 2000
From: fdrake at beopen.com (Fred L. Drake, Jr.)
Date: Mon, 3 Jul 2000 02:28:33 -0400 (EDT)
Subject: Bug in math.frexp?
In-Reply-To: <LNBBLJKPBEHFEDALKOLCCELGGHAA.tim_one@email.msn.com>
References: <20000630221759.A1965@xyz.donetsk.ua>
	<LNBBLJKPBEHFEDALKOLCCELGGHAA.tim_one@email.msn.com>
Message-ID: <14688.12945.528521.351944@cj42289-a.reston1.va.home.com>

Tim Peters writes:
 > The docs are wrong.  I just checked in a repaired version of mathmodule.c.
 > Fred, would you please get the truth into the Library Reference Manual's
 > math module docs too?

  --sigh--
  So are the Solaris and Linux man pages wrong, or just confusing?
Ok, I'll have something like your docstring checked in shortly.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at beopen.com>
BeOpen PythonLabs Team Member




From rob at hooft.net  Mon Jul  3 08:34:15 2000
From: rob at hooft.net (Rob Hooft)
Date: 03 Jul 2000 08:34:15 +0200
Subject: (Serious?) package namespace problem (and a proposal)
References: <slrn8lknve.1v9.hzhu@rocket.knowledgetrack.com> <395A66C5.E9AC4F3E@roguewave.com> <slrn8lkta6.1v9.hzhu@rocket.knowledgetrack.com> <20000628151611.A2972@freei.com> <8F61C6D29gmcmhypernetcom@199.171.54.194> <ffcmlsc6p4lqohdd6tequdu918l0temns4@4ax.com> <395B6027.B47094CA@letterror.com>
Message-ID: <m2ya3jn2mw.fsf@hooft.net>

>>>>> "JvR" == Just van Rossum <just at letterror.com> writes:

 JvR> <silly-proposal>
 JvR> If relative imports as we know them are shot in the head, an
 JvR> alternative spelling for relative imports could be invented.
 JvR> Eg. with a leading period:

 JvR> import .some.submodule
 JvR> </silly-proposal>

Really interesting view. Like "sh" would mean "sh in the root
directory", and "/bin/sh" means look in the current directory for a
directory bin, which contains sh?

Wrong way around? Wouldn't "from . import os" sound more reasonable?

I still think that if you have a package from which you want to use the
module "sys" or "os" or "string", you can AVOID using sys.py, os.py or
string.py. See your own future code:

from . import string; GLOBALstring=string
import string
....
c=GLOBALstring.strip(command)
string.emit(c)
....


And if you don't know what modules from the standard library
you might need in the future, don't use any of their names in your package.
And finally, claim the name of your package in the http://pitch.nist.gov/nics/
archive to prevent future collisions.

Having your own modules named the same as those in the standard library
is also confusing for somebody else that needs to read/modify your code.

Rob

-- 
=====   rob at hooft.net          http://www.hooft.net/people/rob/  =====
=====   R&D, Nonius BV, Delft  http://www.nonius.nl/             =====
===== PGPid 0xFA19277D ========================== Use Linux! =========


From michael at stroeder.com  Mon Jul  3 08:47:47 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Mon, 03 Jul 2000 08:47:47 +0200
Subject: Python CGI security
References: <B5851962.5E%bojar@speakeasy.org>
Message-ID: <39603713.47CD2012@stroeder.com>

"E. Jordan Bojar" wrote:
> 
> I'm interested in running a series of Python scripts via the Apache Python
> module, mostly for dynamic content but possibly for password protection as
> well.  Are there security issues particular to (or common with) Python 
> CGIs,
> in the way that buffer overflow is so common to poorly-written Perl?

I tried to raise this topic here a couple of weeks ago but NO
RESPONSE! Afterwards I started a similar thread in
news:de.comp.security (german newsgroup). Regarding Python code a
guy mentioned that your Python code should avoid eval() and exec().

I'm not a C programmer at all and therefore I just naively used a
tool named ITS4: <URL: http://www.rstcorp.com/its4/> to find parts
in the Python interpreter which might have buffer overrun issues. I
would really like to have some discussion about the ITS4-output.

Ciao, Michael.


From rickhigh at aol.com  Mon Jul  3 09:27:30 2000
From: rickhigh at aol.com (RICKHIGH)
Date: 03 Jul 2000 07:27:30 GMT
Subject: JPython and Java integration
References: <8j7h3m$5v18j$1@fu-berlin.de>
Message-ID: <20000703032730.01208.00001107@ng-cp1.aol.com>

  Think about using IBM's BSF.
BSF = Bean Scripting Framework.
It works with JPython, BeanShell, etc.
For the price of integrating one language in your applicatoin you get many,
many languages.

I wrote an article on BSF and JPython at the JDJ (Java Developer's Journal).

go to www.sys-con.com hit the JDJ gif, then search for BSF or JPython.


From rickhigh at aol.com  Mon Jul  3 09:31:06 2000
From: rickhigh at aol.com (RICKHIGH)
Date: 03 Jul 2000 07:31:06 GMT
Subject: Shoud I use Java or Python?
References: <00062612550704.00594@dagomachine>
Message-ID: <20000703033106.01208.00001108@ng-cp1.aol.com>

There are many reasons to choose Java.
There are also many reasons to choose Python.

Why not have the best of both worlds?

JPYTHON (Python that runs in the Java Virtual Machine)
www.jpython.org

I also wrote an article on JPython for the Java Developer's Journal.

you can see it at www.sys-con.com
search for JPython.




From thomas at xs4all.nl  Mon Jul  3 09:55:04 2000
From: thomas at xs4all.nl (Thomas Wouters)
Date: 3 Jul 2000 07:55:04 GMT
Subject: tkinter hanging
References: <WRe0zXARD1X5EwXY@jessikat.demon.co.uk>
Message-ID: <slrn8m0hm3.m2b.thomas@centurion.xs4all.nl>

On Sun, 2 Jul 2000 15:25:21 +0100, Robin Becker <robin at jessikat.demon.co.uk>
wrote:
>I'm having problems with zombie processes under win32 with a tkinter
>linked against Tk 8.3. I seem to be exiting the mainloop ok and have
>destroyed all subwindows and then the gui object itself, but
>occasionally I still generate a zombie. I have checked that my after's
>are cancelled and debug prints seem to show me exiting the script.

Does your Tkinter script become a zombie process, or a sub-process ? Zombie
processes are processes that exited normally, but of which the parent hasn't
(yet) read the exit code (with one of the 'wait' family of functions.) If
your process exits without wait()ing on its children, those children will be
inherited by 'init', the first process to start on a UNIX box, which does a
continous wait().

Zombie processes do not really take up much resources, except a slot in the
process table and the int that is needed for the exit code. If you have one
or two lingering around for a couple of minutes, you shouldn't worry.
However, if you see hundreds of them lying around, something might be wrong
;-P Check their parent process (with ps -l) -- if it's '1', something is
wrong with your init. If it's another process, that process needs to be
fixed ;-) You can always 'fix' it by killing it and starting a new one, but
you might not like the side-effects. Adding waitfor() calls at appropriate
places might be a less destructive and more permanent solution ;)

Halloween-ly-y'rs,
	Thomas


From thomas at xs4all.nl  Mon Jul  3 10:00:11 2000
From: thomas at xs4all.nl (Thomas Wouters)
Date: 3 Jul 2000 08:00:11 GMT
Subject: how to push data in to ssh in python?
References: <395FF2D9.1B46251D@esec.com.au>
Message-ID: <slrn8m0hvm.m2b.thomas@centurion.xs4all.nl>

On Mon, 03 Jul 2000 11:56:41 +1000, Sam Wun <swun at esec.com.au> wrote:

>Thanks

Your question is incomplete. What do you want to do ? Did you look at
popen(), and is it not what you want ? Does ssh require the input to come
from a pseudo-terminal ? Can't you use a normal shell pipeline and make your
python script just 'print' what you want to output ? Are you encountering
difficulties with any of these ideas, or not finding the right documentation?

Oh, and it's not necessary to post both in HTML and in plaintext.. Just
plaintext will do just fine ;)

How-do-I-get-a-zilion-dollars-ly y'rs,
	Thomas.


From davec999 at my-deja.com  Mon Jul  3 10:23:03 2000
From: davec999 at my-deja.com (David Currie)
Date: Mon, 03 Jul 2000 08:23:03 GMT
Subject: command-line history in win
References: <395c304e.1224185@news.student.lu.se> <8jhm03$qnm$1@nnrp1.deja.com> <8jig08$dk7$1@nnrp1.deja.com> <8F63D360Egmcmhypernetcom@199.171.54.154>
Message-ID: <8jpigs$4k9$1@nnrp1.deja.com>

In article <8F63D360Egmcmhypernetcom at 199.171.54.154>,
  gmcm at hypernet.com (Gordon McMillan) wrote:
> david_ullrich at my-deja.com wrote:
>
> >  David Currie <davec999 at my-deja.com> wrote:
>
> >> It works for me (Python 1.5.2)...
>
> >     Keen. Maybe I'm the only one who doesn't know:
> >_What_ works for you, exactly?
>
> Perhaps David is on NT, where it just works (not only for DOS cmds,
but if
> you start Python, up-arrow will take you thru the history for your
last
> Python session in that DOS box). This is all built in to NT.

That is correct, I am on NT.  I hadn't realised it might be different
for win9x...

Dave


Sent via Deja.com http://www.deja.com/
Before you buy.


From michael at stroeder.com  Mon Jul  3 10:54:57 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Mon, 03 Jul 2000 10:54:57 +0200
Subject: Python CGI security
References: <B5851962.5E%bojar@speakeasy.org>
Message-ID: <396054E1.ED99677C@stroeder.com>

"E. Jordan Bojar" wrote:
> 
> Are there security issues particular to (or common with) Python CGIs,

Additional info:
- Have a close look at all os.system() and os.popen() calls.
- Use smptlib instead calling sendmail via os.system().
- Build a framework for strict form parameter checking. (I have such
a module. Feel free to ask for it.)

Ciao, Michael.


From stephen at cerebralmaelstrom.com  Mon Jul  3 11:24:06 2000
From: stephen at cerebralmaelstrom.com (Stephen Hansen)
Date: Mon, 03 Jul 2000 09:24:06 GMT
Subject: PYRO, XML-RPC, and their friends.
Message-ID: <8jpm3d$6ss$1@nnrp1.deja.com>

Okay, apparently i'm entering another one of my Twenty Questions phases.

I've discovered recently that I have a need for some Client/Server db
fucntionality in the python project that i'm working on. What I don't
want is to have to deal with installing some Big Bunch Of Crap that
will be needed to support my program, and even though I know nothing
about ODBC, it seems like that would include more baggage then its
worth in this instance.. not to mention i'd _like_ to maybe eventually
have my little windows program run on Linux...maybe

Anyways. It might be a bit overly complicated in and of itself, but the
way i've thought of doing it is just setting up PYRO, XML-RPC, etc, to
communicate between my Server and its Clients, to then access a MetaKit
controlled on the server.

I have two questions:

#1) How fast are these? There might actually be a need to transmit
quite a bit of data between the client and the server -- over an
ethernet LAN connection, prolly.. Some of this data might actually be
binary files, either stored in the db or on the side, and i'd need to
be transmitting that across also. I don't know how long it'd take one
of these systems to send across a multi-megabyte binary file.. The rest
of the data i'd be sending would be irrelevant in comparison.

#2) Is there a Better Way? Keep in mind that i don't want to include a
lot of stuff extra which is going to require installation if I don't
have to..

Thanks!

--Stephen


Sent via Deja.com http://www.deja.com/
Before you buy.


From michael at stroeder.com  Mon Jul  3 11:22:50 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Mon, 03 Jul 2000 11:22:50 +0200
Subject: Python CGI security
References: <B5851962.5E%bojar@speakeasy.org> <39603713.47CD2012@stroeder.com> <slrn8m0j4j.m2b.thomas@centurion.xs4all.nl>
Message-ID: <39605B6A.28B4E36D@stroeder.com>

Thomas Wouters wrote:
> 
> On Mon, 03 Jul 2000 08:47:47 +0200, Michael Str?der <michael at stroeder.com>
> wrote:
> >"E. Jordan Bojar" wrote:
> 
> >> [..] buffer overflow [..] 
> 
> >I'm not a C programmer at all and therefore I just naively used a
> >tool named ITS4: <URL: http://www.rstcorp.com/its4/> to find parts
> >in the Python interpreter which might have buffer overrun issues. I
> >would really like to have some discussion about the ITS4-output.
> 
> If you do want to have that discussion, feel free to send me that report, 
> or post it here

Ok, here are some examples from Python1.6a2 sources (see below). I'm
not a C programmer therefore such tools are my only possibility to
raise some discussion by pointing to code. Off course the tool
outputs only hints. Not every code line listed is a real security
hole.

I understand that the Python application itself is responsible for
dealing with some issues (e.g. race conditions with files) since the
Python libs only wrap some system libs. And I understand that some
of the listed functions as being dangerous are used in a safe way
with appropriate parameters.

But IMHO a discussion about security issues would be worthful.

Ciao, Michael.

----------------------------------------------------------------------
Modules/posixmodule.c:248:(Urgent) popen
Easy to run arbitrary commands through env vars.
Use fork + execve + pipes instead.
----------------
[..]
----------------
Modules/posixmodule.c:1151:(Urgent) system
Easy to run arbitrary commands through env vars. Also, potential
TOCTOU
problems.
Use fork + execve instead.
----------------
[..]
----------------
Modules/posixmodule.c:605:(Very Risky) access
Potential race condition on: path
Points of concern are:
Modules/posixmodule.c:605: access
Modules/posixmodule.c:701: chmod
Modules/posixmodule.c:1063: mkdir
Modules/posixmodule.c:1065: mkdir
Modules/posixmodule.c:1254: utime
Modules/posixmodule.c:2235: readlink
Modules/posixmodule.c:3263: statvfs
Manipulate file descriptors, not symbolic names, when possible.
----------------
[..]
----------------
Modules/main.c:117:(Very Risky) getenv
Often seen in conjunction with buffer overflows, etc.
Remember that env vars can contain arbitrary malicious input.  Test
accordingly
before use.
----------------
[..]
----------------
Modules/posixmodule.c:1863:(Very Risky) getlogin
It is very easy to fool.
Don't trust its output.
----------------
[..]
----------------
Modules/parsermodule.c:1032:(Very Risky) sprintf
This function is high risk for buffer overflows
Use snprintf if available, or precision specifiers, if available.
----------------
[..]
----------------
Modules/getpath.c:186:(Very Risky) stat
Potential race condition on: filename
Points of concern are:
Modules/getpath.c:186: stat
Modules/getpath.c:216: stat
Modules/getpath.c:231: stat
Modules/main.c:194: fopen
Python/bltinmodule.c:883: fopen
Python/pythonrun.c:583: fopen
Python/traceback.c:157: fopen
Manipulate file descriptors, not symbolic names, when possible.
----------------
[..]
----------------
Modules/posixmodule.c:385:(Very Risky) strcat
This function is high risk for buffer overflows
Use strncat instead.
----------------
[..]
----------------
Modules/posixmodule.c:827:(Very Risky) strcpy
This function is high risk for buffer overflows
Use strncpy instead.


From supersede at india.com  Mon Jul  3 11:24:48 2000
From: supersede at india.com (Extra Strength Flufferin)
Date: 03 Jul 2000 05:24:48 -0400
Subject: 1.5.2 extentions under 1.6
References: <8jpi6m$4aa$1@nnrp1.deja.com>
Message-ID: <87lmzjefbz.fsf@litterbox.meowing.net>

Stephen Hansen <stephen at cerebralmaelstrom.com> wrote:

> My question is -- does anything need to be changed in the sources, any
> differences int he API, etc? Or is it simply something which is
> modified in the Python.h which needs to be included in the extention
> for it to work?

The only real annoyance I've run into so far is that Py_Malloc() and
company went away.  It makes sense that they went away and it's easy
enough to work around, but do watch your links carefully for undefined
symbol warnings.


From alex at magenta.com  Mon Jul  3 11:07:43 2000
From: alex at magenta.com (Alex Martelli)
Date: Mon, 3 Jul 2000 11:07:43 +0200
Subject: Class design: accessing "private" members
References: <962352843.1667731268@news.libertysurf.fr> <8jhuvn01vs0@news2.newsguy.com> <962367319.379353174@news.libertysurf.fr> <395D79FE.4E4FDBCE@colconsulting.com>
Message-ID: <8jplbd0l5c@news2.newsguy.com>

Frank V. Castellucci <frankc at colconsulting.com> wrote in message
news:395D79FE.4E4FDBCE at colconsulting.com...
> Jerome Quelin wrote:
> >
> > "Alex Martelli" <alex at magenta.com> wrote:
> > >By calling the accessors, rather than accessing the data
> > >directly, you ensure that if somebody overrides the
> > >accessor functions in a derived class you'll be calling
> > >the derived-class versions of those functions.  If that is
> > >what you want (one direction along which you want to let
> > >inheritors of your class extend or tweak behaviour), then
> > >it's probably worth the overhead you mention.
> > The class isn't to be derived any further. It's a class that I use in an
> > application, and that may be imported in some other applications, but
not to be
> > inherited.

Then, "do the simplest thing that could possibly work" applies.

> > Then, is it better/cleaner to access private members with accessors or
not?
> > Is it a matter of style? Or are there hidden caveheats using (or not
using)
> > accessors? It _seems_ cleaner to use accessors, but man, it's quite
awkward.

I would not use accessors that feel awkward to you, without specific
purpose.

> As you really have no way to "prohibit" further derivation (unless I
> missed the "final" keyword in the reference), using your accessors and
> praying that anyone over-riding the accessor/mutator methods does the
> proper thing would be my direction.

Such "prohibition" is not hard at all:

class dontderive:
    def __init__(self):
        if self.__class__ != dontderive:
            raise "told you not to!"

class foo(dontderive):
    pass

However, I wouldn't normally do it, again under the "simplest thing"
principle.  Just documenting "this class is not meant to be derived
from, so, just don't do it, or results will be unpredictable" is likely
to be "the simplest thing that could possibly work" in this case.


Alex





From Jarkko.Veijalainen at tecnomen.fi  Mon Jul  3 12:14:40 2000
From: Jarkko.Veijalainen at tecnomen.fi (Jarkko Veijalainen)
Date: Mon, 3 Jul 2000 13:14:40 +0300 
Subject: slicing string
Message-ID: <00Jul3.130616eest.119043@sister.tecnomen.fi>

Hi!

I'm real newbie with Python. I started learnig Python just last week. I have
a problem slicing a string, imported from file. can anyone help me with
simple solution. Input string follows this pattern:

 attr=value , attr2=value2, attr3= , attr4=value2. 

I have made an Object class, where all attributes are class variables. How
can i slice that sring and store those att=value pairs in my object?

I'm desperated...

	JarkkoV



From alex at magenta.com  Mon Jul  3 11:16:29 2000
From: alex at magenta.com (Alex Martelli)
Date: Mon, 3 Jul 2000 11:16:29 +0200
Subject: Class design: accessing "private" members
References: <962352843.1667731268@news.libertysurf.fr><8jhuvn01vs0@news2.newsguy.com> <4.3.1.2.20000630145109.0201d9e0@phear.dementian.com>
Message-ID: <8jplru0lfe@news2.newsguy.com>

Steven D. Arnold <stevena at neosynapse.net> wrote in message
news:4.3.1.2.20000630145109.0201d9e0 at phear.dementian.com...
> At 02:15 PM 6/30/2000 +0200, Jerome Quelin wrote:
>
> >Then, is it better/cleaner to access private members with accessors or
not?
> >Is it a matter of style? Or are there hidden caveheats using (or not
using)
> >accessors? It _seems_ cleaner to use accessors, but man, it's quite
awkward.
>
> If you always use accessors, then you can store or derive the value in any
> way you wish internally in your class.  You may completely change the way
> the value is obtained, and the class user doesn't need to change anything

These are excellent arguments for always using accessors from OUTSIDE the
class.  Jerome, however, was talking about using them from INSIDE the class
as well, which is a different issue.


Alex






From alex at magenta.com  Mon Jul  3 11:35:43 2000
From: alex at magenta.com (Alex Martelli)
Date: Mon, 3 Jul 2000 11:35:43 +0200
Subject: Passing variable-length keyworded arguments list to base class constructor?
References: <962352947.103057778@news.libertysurf.fr> <8jhum101vku@news2.newsguy.com> <962367281.1198008296@news.libertysurf.fr>
Message-ID: <8jpn010mb7@news2.newsguy.com>

Jerome Quelin <jerome.quelin at insalien.org> wrote in message
news:962367281.1198008296 at news.libertysurf.fr...
> "Alex Martelli" <alex at magenta.com> wrote:
> >> I was wondering how to pass variable-length keyworded arguments list
> >> to my base-class constructor? Look at the example code below:
> >Doesn't apply work for __init__ just as for any other function?
> Sure it works for __init__ too, I've already tried this approch, but it
doesn't
> work either. In fact, I'm asking the question for __init__ because I run
into
> this problem when calling constructors, but the question is the same for
every
> inherited and overloaded method.
>
> When using the *args syntax, I get an error when calling my constructor
because
> I'm using keywords arguments.

Ok, so *args is out.


> When using the **args syntax, python fetches all keyworded arguments in
_one_
> dictionnary in my constructor, and I want to keep the arguments in
multiple
> key/value pairs (and not flattened in one dictionnary) since the parent
> constructor is waiting for keyworded arguments too.

But that's what apply does (in part): turn a "flattened in one dictionary"
set of name/value pairs into "keyworded arguments" for the function being
applied.


> If I use the **args syntax and then call the base constructor with:
> Parent.__init__(self, args)
> I got an error since it doesn't wait for a dictionnary as an argument (and
args
> is a dictionnary in this case) but for key=value arguments, with key and
value
> taken from the args dictionnary.

So you should instead call apply(Parent.__init__,(self,),args), shouldn't
you?

Toy example:
class Parent:
    def __init__(self, foo=23, bar=45):
        print foo, bar

class Derived(Parent):
    def __init__(self, **kwargs):
        apply(Parent.__init__, (self,), kwargs)

>>> esem.Derived(foo=99)
99 45
<esem.Derived instance at 16a21c0>
>>>


Of course you can alter the kwargs dictionary as required in your
Derived.__init__ method for whatever special needs you may have;
e.g., suppose you want to change the *default* for bar from 45 to
73 in Derived (only), you might code:

class Derived(Parent):
    def __init__(self, **kwargs):
        if not kwargs.has_key('bar'):
            kwargs['bar']=73
        apply(Parent.__init__, (self,), kwargs)

and now:

>>> esem.Derived(foo=99)
99 73
<esem.Derived instance at 16a6810>
>>>


Alex





From moshez at math.huji.ac.il  Mon Jul  3 12:56:03 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Mon, 3 Jul 2000 13:56:03 +0300 (IDT)
Subject: Python CGI security
In-Reply-To: <39605B6A.28B4E36D@stroeder.com>
Message-ID: <Pine.GSO.4.10.10007031350130.22346-100000@sundial>

On Mon, 3 Jul 2000, Michael [iso-8859-1] Str?der wrote:

> Thomas Wouters wrote:
> > 
> > On Mon, 03 Jul 2000 08:47:47 +0200, Michael Str?der <michael at stroeder.com>
> > wrote:
> > >"E. Jordan Bojar" wrote:
> > 
> > >> [..] buffer overflow [..] 
> > 
> > >I'm not a C programmer at all and therefore I just naively used a
> > >tool named ITS4: <URL: http://www.rstcorp.com/its4/> to find parts
> > >in the Python interpreter which might have buffer overrun issues. I
> > >would really like to have some discussion about the ITS4-output.
> > 
> > If you do want to have that discussion, feel free to send me that report, 
> > or post it here
> 
> Ok, here are some examples from Python1.6a2 sources (see below). I'm
> not a C programmer therefore such tools are my only possibility to
> raise some discussion by pointing to code. Off course the tool
> outputs only hints. Not every code line listed is a real security
> hole.

OK, here is an answer to the report I just snipped.
There are a few types of "potential security holes":

 -- exposing insecure parts of the API, such as os.system(). These should
    stay, since the user should have those at his disposale

 -- OS/2 message formatting. I haven't looked at it, but this only
    concerns you if you're working on OS/2

 -- checked "dangerous" functions, which are therefore fine as far as I
    could see

 -- one sprintf() which will work fine as long as integers are <128 bits, 
    which is true in the forseeable feature. I'm planning to patch this,
    just in case, but it should be fine anyway. (I'm talking about the 
    first sprintf() in posixmodule.c)


--
Moshe Zadka <moshez at math.huji.ac.il>
There is no GOD but Python, and HTTP is its prophet.




From thiele at muc.das-werk.de  Mon Jul  3 12:40:44 2000
From: thiele at muc.das-werk.de (Thomas Thiele)
Date: Mon, 03 Jul 2000 12:40:44 +0200
Subject: From UNIX to Win NT
Message-ID: <39606DAC.49A502DB@muc.das-werk.de>

Hallo!

I have a program running on UNIX(IRIX and LINUX). I now want to port
this 
program to Win NT. Is there anybody who have any experience with this?

The problem is I am using threads, fork and exec. And I want to have the
same behaviour using NT.

- at first I create a new thread. (so the main thread is ready for
receiving commands)

- then I fork 
in the child I change it's userID (os.setuid()) and replace it by
os.exec()
the parent reads the stderr and stdout from the started program ( like
popen2.popen3() ) 

Last but not least I use sockets (a derived server from SocketServer.py) 


What have I to change? And why? And how? 

Thanks for help,

Thomas


From cg at gaia.cdg.acriter.nl  Mon Jul  3 08:23:33 2000
From: cg at gaia.cdg.acriter.nl (Cees de Groot)
Date: 3 Jul 2000 08:23:33 +0200
Subject: Aspect-oriented Python?
References: <4.2.0.58.20000703140633.00be3c30@imaphost>
Message-ID: <8jpbh5$jim$1@gaia.cdg.acriter.nl>

Martin Kew  <Martin.Kew at vsl.com.au> said:
>The reason for presenting this overview was the overwelming lack of 
>interest in the first
>email by Cees de Groot.
>
Thanks :-). Somehow, I always assume I lag behind in discovering new
technologies, especially on a group like this one, which so much
programming language and computer-science related prowess. Seems not
to have been the case this time (or, of course, the subject is genuinly
boring and uninteresting ;-)). 

Aspect-oriented programming and its family members subject oriented
programming (older, by IBM) and adaptive programming are really 
interesting tracks. According to the Xerox team, the status at the
moment is more-or-less the status of OOP twenty years ago, all the
more reason for looking for lightweight language testbeds. 


-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B
Forge your CipherSaber and list it: http://www.xs4all.nl/~cg/ciphersaber/


From loewis at informatik.hu-berlin.de  Mon Jul  3 13:05:25 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 03 Jul 2000 13:05:25 +0200
Subject: PYRO, XML-RPC, and their friends.
References: <8jpm3d$6ss$1@nnrp1.deja.com>
Message-ID: <p6qn1jzxymi.fsf@informatik.hu-berlin.de>

Stephen Hansen <stephen at cerebralmaelstrom.com> writes:

> I've discovered recently that I have a need for some Client/Server db
> fucntionality in the python project that i'm working on. What I don't
> want is to have to deal with installing some Big Bunch Of Crap that
> will be needed to support my program, and even though I know nothing
> about ODBC, it seems like that would include more baggage then its
> worth in this instance..

I don't know what database you are going to use, but I believe that
odbc would indeed be the soluation that involves the least amount of
extra stuff. For example, Python-Win comes with an odbc module, and
there is also at least one other odbc module that would work in a
plain python installation, as well.

> have my little windows program run on Linux

It is possible to use odbc on Linux, as well. Depending on the
database, it may even be easy to do so.

> Anyways. It might be a bit overly complicated in and of itself, but the
> way i've thought of doing it is just setting up PYRO, XML-RPC, etc, to
> communicate between my Server and its Clients, to then access a MetaKit
> controlled on the server.

For completeness, I'd like to add CORBA in this candidate list; the
implementation with least installation overhead is probably Fnorb
(only one C module, which comes precompiled for Windows).

> #1) How fast are these?

Some of these are only protocols, which you can't compare for
performance - you'd have to compare concrete implementations. I don't
think anybody has done significant benchmarking on them, so far, but I
guess they are "fast enough".

> I don't know how long it'd take one of these systems to send across
> a multi-megabyte binary file.

I hope most of them would achieve approximately the raw wire speed for
bulk transfer, or be at least comparable to any other bulk data
protocol. Most of these will require to copy the bulk data in memory a
number of times, which would be the primary source of overhead.

> The rest of the data i'd be sending would be irrelevant in
> comparison.

That is only true if you have only "few" calls. If you have "many"
calls, it is actually the number of calls that matters, not the amount
of data transferred.

Regards,
Martin


From thomas at centurion.xs4all.nl  Mon Jul  3 13:18:27 2000
From: thomas at centurion.xs4all.nl (Thomas Wouters)
Date: Mon, 3 Jul 2000 13:18:27 +0200
Subject: Bug in math.frexp?
In-Reply-To: <14688.12945.528521.351944@cj42289-a.reston1.va.home.com>
References: <20000630221759.A1965@xyz.donetsk.ua> <LNBBLJKPBEHFEDALKOLCCELGGHAA.tim_one@email.msn.com> <14688.12945.528521.351944@cj42289-a.reston1.va.home.com>
Message-ID: <200007031118.NAA22787@centurion.xs4all.nl>

In comp.lang.python, you wrote:
>
>Tim Peters writes:
> > The docs are wrong.  I just checked in a repaired version of mathmodule.c.
> > Fred, would you please get the truth into the Library Reference Manual's
> > math module docs too?

>  --sigh--
>  So are the Solaris and Linux man pages wrong, or just confusing?
>Ok, I'll have something like your docstring checked in shortly.

Not sure which version of the Linux manpges you read, but the frexp manpage
that came with RedHat 6.1 more or less exactly mirrors Tim's description:

       The frexp() function returns the normalized fraction.   If
       the  argument  x is not zero, the normalized fraction is x
       times a power of two, and  is  always  in  the  range  1/2
       (inclusive) to 1 (exclusive).  If x is zero, then the nor?
       malized fraction is zero and zero is stored in exp.

The 'info' page about frexp states the same, in different wording. The
Solaris 2.5 manpage I have access to is indeed a bit confusing, but does
state the same: 'x' is in the range of [0.5, 1), or 0. That means the
exponent has to be able to be negative, when the original number was
negative.




From * at spam.ruud.org  Mon Jul  3 13:26:29 2000
From: * at spam.ruud.org (Ruud de Rooij)
Date: 3 Jul 2000 13:26:29 +0200
Subject: Bug in math.frexp?
References: <20000630221759.A1965@xyz.donetsk.ua> <LNBBLJKPBEHFEDALKOLCCELGGHAA.tim_one@email.msn.com> <14688.12945.528521.351944@cj42289-a.reston1.va.home.com> <200007031118.NAA22787@centurion.xs4all.nl>
Message-ID: <87lmzjo3oa.fsf@hobbes.home.ruud.org>

Thomas Wouters <thomas at centurion.xs4all.nl> writes:

> That means the
> exponent has to be able to be negative, when the original number was
> negative.

The sign of the exponent is not necessarily equal to the sign of the
number.

	- Ruud de Rooij
-- 
ruud de rooij | *@spam.ruud.org | http://ruud.org


From thiele at muc.das-werk.de  Mon Jul  3 13:30:22 2000
From: thiele at muc.das-werk.de (Thomas Thiele)
Date: Mon, 03 Jul 2000 13:30:22 +0200
Subject: making widgets invisible
Message-ID: <3960794E.6E2B36E6@muc.das-werk.de>

How can I realize that a set of widgets (buttons and entries)
is invisible when I don't need them and visible when I need them?

One solution is to create a dialog window (or frame) and destroy it
then.
But is there a better one? 

Withdrawn works only with top level widgets.

Thanks Thomas


From michael at stroeder.com  Mon Jul  3 13:33:54 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Mon, 03 Jul 2000 13:33:54 +0200
Subject: Python CGI security
References: <Pine.GSO.4.10.10007031350130.22346-100000@sundial>
Message-ID: <39607A22.CA190D70@stroeder.com>

Moshe Zadka wrote:
> 
> On Mon, 3 Jul 2000, Michael [iso-8859-1] Str?der wrote:
> 
> > Ok, here are some examples from Python1.6a2 sources (see below).
> 
> OK, here is an answer to the report I just snipped.

Note that I only posted *excerpts* of the ITS4 output. Feel free to
ask if you want the whole output by e-mail (approx. 1000 lines with
a lot of repeated messages).

> There are a few types of "potential security holes":
> 
>  -- exposing insecure parts of the API, such as os.system(). These should
>     stay, since the user should have those at his disposale

Would it be possible to provide a secure version in the Python lib
of e.g. os.system() with same function parameters but less unsecure
behaviour instead of just wrapping it 1:1?

>  -- OS/2 message formatting. I haven't looked at it, but this only
>     concerns you if you're working on OS/2

Hmm, I do not know under which OS my demons will run.
Would be worth to collect such insights somewhere.

>  -- checked "dangerous" functions, which are therefore fine as far as I
>     could see
> 
>  -- one sprintf() which will work fine as long as integers are <128 bits,
>     which is true in the forseeable feature. I'm planning to patch this,
>     just in case, but it should be fine anyway. (I'm talking about the
>     first sprintf() in posixmodule.c)

Thanks for following up.

Ciao, Michael.


From thomas at xs4all.nl  Mon Jul  3 13:43:31 2000
From: thomas at xs4all.nl (Thomas Wouters)
Date: 3 Jul 2000 11:43:31 GMT
Subject: Python CGI security
References: <B5851962.5E%bojar@speakeasy.org> <39603713.47CD2012@stroeder.com> <slrn8m0j4j.m2b.thomas@centurion.xs4all.nl> <39605B6A.28B4E36D@stroeder.com>
Message-ID: <slrn8m0v2d.m5d.thomas@centurion.xs4all.nl>

On Mon, 03 Jul 2000 11:22:50 +0200, Michael Str?der <michael at stroeder.com>
wrote:
>Thomas Wouters wrote:
>> On Mon, 03 Jul 2000 08:47:47 +0200, Michael Str?der <michael at stroeder.com>
>> wrote:
>> >"E. Jordan Bojar" wrote:
>> 
>> >> [..] buffer overflow [..] 
>> 
>> >I'm not a C programmer at all and therefore I just naively used a
>> >tool named ITS4: <URL: http://www.rstcorp.com/its4/> to find parts
>> >in the Python interpreter which might have buffer overrun issues. I
>> >would really like to have some discussion about the ITS4-output.
>> 
>> If you do want to have that discussion, feel free to send me that report, 
>> or post it here

>Ok, here are some examples from Python1.6a2 sources (see below). I'm
>not a C programmer therefore such tools are my only possibility to
>raise some discussion by pointing to code. Off course the tool
>outputs only hints. Not every code line listed is a real security
>hole.

>I understand that the Python application itself is responsible for
>dealing with some issues (e.g. race conditions with files) since the
>Python libs only wrap some system libs. And I understand that some
>of the listed functions as being dangerous are used in a safe way
>with appropriate parameters.

Indeed. And a couple of the 'security issues', for instance those with
popen() and system(), and the race conditions in stat/access etc, Python
itself does not care to solve, but instead delegates to the scripts it is
executing ;)

Python will not run set-uid out of the box. You have to supply your own
setuid-wrapper if you want it to, in which case you should also take care
with those functions. None the less, the seperation of environment variables
and system() and popen() makes it slightly less insecure than C.

Also, this suite you used seems to be targeted towards data-handling
engines, like webservers, in a hostile environment. This is entirely unlike
Python's usual environment. If you plan on programming in such environments,
you should take care with your programs. Python will not do the Right Thing
each time, but it also does not pretend it does. The habit of throwing
exceptions on errors instead of returning an undefined value and continueing
on helps prevent a lot of tight spots, though.

But I'll do a quick run through the 'hotspots':

>----------------------------------------------------------------------
>Modules/posixmodule.c:248:(Urgent) popen
>Easy to run arbitrary commands through env vars.
>Use fork + execve + pipes instead.
>----------------

See above. popen() is supplied to Python scripts, but python itself does not
use it. Using popen() yourself should ofcourse be done carefully :P

>----------------
>Modules/posixmodule.c:1151:(Urgent) system
>Easy to run arbitrary commands through env vars. Also, potential
>TOCTOU
>problems.
>Use fork + execve instead.
>----------------

See popen()

>----------------
>Modules/posixmodule.c:605:(Very Risky) access
>Potential race condition on: path
>Points of concern are:
>Modules/posixmodule.c:605: access
>Modules/posixmodule.c:701: chmod
>Modules/posixmodule.c:1063: mkdir
>Modules/posixmodule.c:1065: mkdir
>Modules/posixmodule.c:1254: utime
>Modules/posixmodule.c:2235: readlink
>Modules/posixmodule.c:3263: statvfs
>Manipulate file descriptors, not symbolic names, when possible.
>----------------

See popen(). If you stat for a filename, you should be aware that it can go
away from under you. 

>----------------
>Modules/main.c:117:(Very Risky) getenv
>Often seen in conjunction with buffer overflows, etc.
>Remember that env vars can contain arbitrary malicious input.  Test
>accordingly
>before use.
>----------------

Good warning for every programmer in hostile environments, but not
applicable to python itself. Usually Python scripts are not setuid, so they
can trust their environment (they run under the same permission as who
created that environment, after all.) If you use a setuid wrapper, you
should pass a new, 'safed' environment yourself.

>----------------
>Modules/posixmodule.c:1863:(Very Risky) getlogin
>It is very easy to fool.
>Don't trust its output.
>----------------

See getenv().

>----------------
>Modules/parsermodule.c:1032:(Very Risky) sprintf
>This function is high risk for buffer overflows
>Use snprintf if available, or precision specifiers, if available.
>----------------

Used as safely as necessary. The parser module is not something quickly used
in high security environments in any case.

>----------------
>Modules/getpath.c:186:(Very Risky) stat
>Potential race condition on: filename
>Points of concern are:
>Modules/getpath.c:186: stat
>Modules/getpath.c:216: stat
>Modules/getpath.c:231: stat
>Modules/main.c:194: fopen
>Python/bltinmodule.c:883: fopen
>Python/pythonrun.c:583: fopen
>Python/traceback.c:157: fopen
>Manipulate file descriptors, not symbolic names, when possible.
>----------------

See access

>----------------
>Modules/posixmodule.c:385:(Very Risky) strcat
>This function is high risk for buffer overflows
>Use strncat instead.
>----------------

String lenghts are passed in and recorded seperately, and though I haven't
backtracked it the whole way, they seem to be properly kept up to date. This
is an OS/2 only branch, though.

>----------------
>Modules/posixmodule.c:827:(Very Risky) strcpy
>This function is high risk for buffer overflows
>Use strncpy instead.

String lengths are explicitly checked.



From cut_me_out at hotmail.com  Mon Jul  3 13:55:55 2000
From: cut_me_out at hotmail.com (Alex)
Date: 03 Jul 2000 07:55:55 -0400
Subject: Getting double* from NumPy array?
References: <ted.drain-2606001323500001@mjordan.jpl.nasa.gov> <72T75.24010$i5.229971@news1.frmt1.sfba.home.com>
Message-ID: <etdvgynjulw.fsf@w20-575-34.mit.edu>

> There is quite a bit to know and just telling you to access x->data
> would not be doing you any favors.

Ah, well my apologies for the bad advice, then.

> Please read the entire chapter before proceeding.

Sounds like something I ought to do, too.

Alex.


From thomas at xs4all.nl  Mon Jul  3 14:03:16 2000
From: thomas at xs4all.nl (Thomas Wouters)
Date: 3 Jul 2000 12:03:16 GMT
Subject: Python CGI security
References: <Pine.GSO.4.10.10007031350130.22346-100000@sundial> <39607A22.CA190D70@stroeder.com>
Message-ID: <slrn8m107e.m5d.thomas@centurion.xs4all.nl>

On Mon, 03 Jul 2000 13:33:54 +0200, Michael Str?der <michael at stroeder.com>
wrote:
>Moshe Zadka wrote:

>> There are a few types of "potential security holes":
>> 
>>  -- exposing insecure parts of the API, such as os.system(). These should
>>     stay, since the user should have those at his disposale

>Would it be possible to provide a secure version in the Python lib
>of e.g. os.system() with same function parameters but less unsecure
>behaviour instead of just wrapping it 1:1?

How could os.system() be more secure ? The way it is defined leaves no room
for restriction. (It calls a shell to interpret the string passed in, after
all.)

What Python does do, is provide more convenient variants of os.exec*(), in
the form of os.spawn*(). (At least, in Python 2.0) Also, writing your own
'secure' system() in terms of fork() and exec*() is fairly easy.

Your-code-is-as-secure-as-you-make-it-ly y'rs,
	Thomas.



From m.faassen at vet.uu.nl  Mon Jul  3 14:05:06 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 3 Jul 2000 12:05:06 GMT
Subject: Books
References: <8jb1ec$9mh2@news.qualitynet.net> <autilso0l591oo924n3mogcva7r2892p5b@4ax.com>
Message-ID: <8jpvhi$jfs$1@newshost.accu.uu.nl>

Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote:
> On Tue, 27 Jun 2000 17:09:09 +0300, "Helmit" <helmit_world at hotmail.com>
> declaimed the following in comp.lang.python:

>> Does anyone know any good python books for teaching beginers?
>> 
> 	Wide-open question...

> 	Beginners in programming;

I saw others recommend Learning Python; I wouldn't do so for beginners
in programming. The Quick Python book or Python in 24 Hours seem to be
better suited for this.

> 	beginners in Python;

Learning Python is fine here, though it may partially depend on the answer
to the next question:

> 	beginners in OO...

I'm not sure how good Learning Python is at teaching OO concepts to people
who don't know about it yet. *Python* would be good though. Any opinions?

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From jerome.quelin at insalien.org  Mon Jul  3 14:06:53 2000
From: jerome.quelin at insalien.org (Jerome Quelin)
Date: Mon, 3 Jul 2000 14:06:53 +0200
Subject: Class design: accessing "private" members
References: <962352843.1667731268@news.libertysurf.fr><8jhuvn01vs0@news2.newsguy.com> <4.3.1.2.20000630145109.0201d9e0@phear.dementian.com> <8jplru0lfe@news2.newsguy.com>
Message-ID: <962626013.769258863@news.libertysurf.fr>

"Alex Martelli" <alex at magenta.com> wrote:
>These are excellent arguments for always using accessors from OUTSIDE the
>class.  Jerome, however, was talking about using them from INSIDE the class
>as well, which is a different issue.
Sure, I'm always using accessors from outside.
But from inside the class, I don't know which is the best. I feel like the KISS
rule should apply, but what do you OO gurus think of it ? How far encapsulation
should go ?

Regards,
Jerome
--
jerome.quelin at insalien.org


From gmcm at hypernet.com  Mon Jul  3 14:20:35 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 03 Jul 2000 12:20:35 GMT
Subject: From UNIX to Win NT
References: <39606DAC.49A502DB@muc.das-werk.de>
Message-ID: <8F665C79Agmcmhypernetcom@199.171.54.154>

Thomas Thiele wrote:

>Hallo!
>
>I have a program running on UNIX(IRIX and LINUX). I now want to port
>this 
>program to Win NT. Is there anybody who have any experience with this?
>
>The problem is I am using threads, fork and exec. And I want to have the
>same behaviour using NT.
>
>- at first I create a new thread. (so the main thread is ready for
>receiving commands)

Threads work very well on NT.

>- then I fork 
>in the child I change it's userID (os.setuid()) and replace it by
>os.exec()
>the parent reads the stderr and stdout from the started program ( like
>popen2.popen3() ) 

You've got both technical and cultural problems here. There is no fork. If 
you use the Win32 extensions you can change the user id (never done this, 
so don't ask me for details), but be prepared for the fact that this kind 
of stuff *means* something different on NT - the models just aren't the 
same. There's also less reason to need it - between NT and the typical NT 
configuration, many things that would require setuid on *nix can be done by 
any wazoo on Windows.

>Last but not least I use sockets (a derived server from SocketServer.py) 

Sockets work fine. Select also works fine, but works *only* on sockets (eg 
if you're selecting on the pipes returned from popen3, you've got a 
problem).

- Gordon


From neilh at scintilla.org  Mon Jul  3 14:31:20 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Mon, 03 Jul 2000 12:31:20 GMT
Subject: Aspect-oriented Python?
References: <4.2.0.58.20000703140633.00be3c30@imaphost> <8jpbh5$jim$1@gaia.cdg.acriter.nl>
Message-ID: <sI%75.6447$Tb7.38793@news-server.bigpond.net.au>

> Thanks :-). Somehow, I always assume I lag behind in discovering new
> technologies, especially on a group like this one, which so much
> programming language and computer-science related prowess. Seems not
> to have been the case this time (or, of course, the subject is genuinly
> boring and uninteresting ;-)).

  Aspect oriented programming and its relative 'acquisition' (as used in
ZOPE) was discussed a while back (maybe 18 months ago or so) either on this
group or the JPython list. I tried searching for it on Deja but Deja have
dropped their older archives so couldn't find anything.

   Neil




From gmcm at hypernet.com  Mon Jul  3 14:34:51 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 03 Jul 2000 12:34:51 GMT
Subject: slicing string
References: <00Jul3.130616eest.119043@sister.tecnomen.fi>
Message-ID: <8F665CDFAgmcmhypernetcom@199.171.54.154>

Jarkko Veijalainen wrote: 

>I'm real newbie with Python. I started learnig Python just last week. I
>have a problem slicing a string, imported from file. can anyone help me
>with simple solution. Input string follows this pattern:
>
> attr=value , attr2=value2, attr3= , attr4=value2. 
>
>I have made an Object class, where all attributes are class variables.
>How can i slice that sring and store those att=value pairs in my object?

First part is easy:
import string
pairs = string.split(s, ',')
for pair in pairs:
  tokens = string.split(pair, '=')
  if len(tokens) == 2:
    key = string.strip(tokens[0])
    val = string.strip(tokens[1])
    # see below
  else:
    # not sure what you want to do here

If I take you literally, then I would say:
    Object.key = val

but I suspect you want a line at the top:
o = Object()
and then
    o.key = val

- Gordon


From wolfgang.thaemelt at orsoft.de  Mon Jul  3 14:41:26 2000
From: wolfgang.thaemelt at orsoft.de (Wolfgang Thaemelt)
Date: Mon, 3 Jul 2000 14:41:26 +0200
Subject: lambda requirements
Message-ID: <000101bfe4ec$010e1c60$a5dfdfdf@ors-x-wt1.or_soft.de>

Running the script:

#----------------------------------------- begin of script
def SymmetricDifference(list1, list2):
#      global _list1
#      global _list2

      _list1 = list1
      _list2 = list2

      _onlylist1 = filter(lambda x: x not in _list2, _list1)

      return _onlylist1

if __name__ == '__main__':
  a = [1,2,3,4,5]
  b = [3,4,5,6,7]
  x = SymmetricDifference(a,b)

  print x
#---------------------------------------- end of script

gives the error:

Traceback (innermost last):
  File "E:\up\000701\symmdiff.py", line 16, in ?
    x = SymmetricDifference(a,b)
  File "E:\up\000701\symmdiff.py", line 9, in SymmetricDifference
    _onlylist1 = filter(lambda x: x not in _list2, _list1)
  File "E:\up\000701\symmdiff.py", line 9, in <lambda>
    _onlylist1 = filter(lambda x: x not in _list2, _list1)
NameError: _list2

If the comment characters "#"  are removed from the "global" statements
everything works fine.

Is there some requirement on a lambda violated in the script?


Thanks

Wolfgang Thaemelt


OR Soft Jaenicke GmbH
Geusaer Str., FH Geb. 106
06217 Merseburg
Germany
======================================
Email: wolfgang.thaemelt at orsoft.de
Voice: +49-3461-549033
Fax:   +49-3461-549022




From moshez at math.huji.ac.il  Mon Jul  3 14:53:49 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Mon, 3 Jul 2000 15:53:49 +0300 (IDT)
Subject: Python CGI security
In-Reply-To: <39607A22.CA190D70@stroeder.com>
Message-ID: <Pine.GSO.4.10.10007031552330.22346-100000@sundial>

On Mon, 3 Jul 2000, Michael [iso-8859-1] Str?der wrote:

> Note that I only posted *excerpts* of the ITS4 output. Feel free to
> ask if you want the whole output by e-mail (approx. 1000 lines with
> a lot of repeated messages).
> 
> > There are a few types of "potential security holes":
> > 
> >  -- exposing insecure parts of the API, such as os.system(). These should
> >     stay, since the user should have those at his disposale
> 
> Would it be possible to provide a secure version in the Python lib
> of e.g. os.system() with same function parameters but less unsecure
> behaviour instead of just wrapping it 1:1?

no: the insecurities are there for a reason: they give more power for
local applications

Please send to me (personally!) the whole output. I'll have a look at it
and try to summarize it.
--
Moshe Zadka <moshez at math.huji.ac.il>
There is no GOD but Python, and HTTP is its prophet.




From thiele at muc.das-werk.de  Mon Jul  3 15:06:40 2000
From: thiele at muc.das-werk.de (Thomas Thiele)
Date: Mon, 03 Jul 2000 15:06:40 +0200
Subject: From UNIX to Win NT
References: <39606DAC.49A502DB@muc.das-werk.de> <8F665C79Agmcmhypernetcom@199.171.54.154>
Message-ID: <39608FE0.35950741@muc.das-werk.de>

>
> You've got both technical and cultural problems here. There is no fork.

I knew that. But I wonder how to start an other program. In UNIX I used fork()
and exec().
Just the same code like in popen2.popen3 but I copied the code and included
some costumization.
Does popen work on NT? Or is there an other way do get the same result
(starting a program and
read it's output)

>
> configuration, many things that would require setuid on *nix can be done by
> any wazoo on Windows.
>



From fdrake at beopen.com  Mon Jul  3 15:28:02 2000
From: fdrake at beopen.com (Fred L. Drake, Jr.)
Date: Mon, 3 Jul 2000 09:28:02 -0400 (EDT)
Subject: Bug in math.frexp?
In-Reply-To: <200007031118.NAA22787@centurion.xs4all.nl>
References: <20000630221759.A1965@xyz.donetsk.ua>
	<LNBBLJKPBEHFEDALKOLCCELGGHAA.tim_one@email.msn.com>
	<14688.12945.528521.351944@cj42289-a.reston1.va.home.com>
	<200007031118.NAA22787@centurion.xs4all.nl>
Message-ID: <14688.38114.543936.743938@cj42289-a.reston1.va.home.com>

Thomas Wouters writes:
 > Not sure which version of the Linux manpges you read, but the frexp manpage
 > that came with RedHat 6.1 more or less exactly mirrors Tim's description:

  I must have looked at something older when last I updated the
frexp() description; my Mandrake 7.0 install agrees with your RedHat
manpage.

 > Solaris 2.5 manpage I have access to is indeed a bit confusing, but does
 > state the same: 'x' is in the range of [0.5, 1), or 0. That means the
 > exponent has to be able to be negative, when the original number was
 > negative.

  And it's certainly possible I was just confused... numerics are not
my area of expertise, and I'll bow to Tim on this stuff regardless.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at beopen.com>
BeOpen PythonLabs Team Member




From richard_chamberlain at my-deja.com  Mon Jul  3 15:44:19 2000
From: richard_chamberlain at my-deja.com (richard_chamberlain at my-deja.com)
Date: Mon, 03 Jul 2000 13:44:19 GMT
Subject: making widgets invisible
References: <3960794E.6E2B36E6@muc.das-werk.de>
Message-ID: <8jq5bf$h45$1@nnrp1.deja.com>

Hi Thomas,

I'm sure there's a nicer way to do it (I'll ponder on it), but...

from Tkinter import *
root=Tk()
def hideMe():
    myButton.configure(text="",relief=FLAT)
myButton=Button(root,text="Hide Me!",command=hideMe)
myButton.pack()
root.mainloop()

You could pass the widget reference to the method so it would be
universal - bit of pain to reverse though!

Richard



In article <3960794E.6E2B36E6 at muc.das-werk.de>,
  thiele at muc.das-werk.de wrote:
> How can I realize that a set of widgets (buttons and entries)
> is invisible when I don't need them and visible when I need them?
>
> One solution is to create a dialog window (or frame) and destroy it
> then.
> But is there a better one?
>
> Withdrawn works only with top level widgets.
>
> Thanks Thomas
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From alan.gauld at gssec.bt.co.uk  Mon Jul  3 15:56:47 2000
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Mon, 03 Jul 2000 14:56:47 +0100
Subject: [Tkinter]: Where is a good place to learn how to use this...?
References: <20000630211051.11572.qmail@web5301.mail.yahoo.com>
Message-ID: <39609B9F.ADFF5512@gssec.bt.co.uk>

Dave Williams wrote:
> I am fairly new to Python, and I have been reading a
> book, and I have checked online for resources.  There
> is a general "intro to" type off of the python.org

There should be two online sources from python.org.

One is F/'s reference site which while not complete 
is very good - due out as a book from O'Reilly sometime...

http://www.pythonware.com/library/tkinter/introduction/index.htm

The other is the original(and slightly outdated) crash 
course on Tkinter called the 'Life Preserver' by Matt Conway

http://www.python.org/doc/life-preserver/index.html

The Mark Lutz book, Programming Python, has a good 
intro to Tkinter if you have no previous Tcl/Tk experience.

John Grayson's book is fairly complete but does rather 
jump in the deep end. Its fine if you know the basics 
or if you've already used Tcl/Tk (or perl/Tk I guess)

My Python tutor gives a couple of Tkinter examples but 
they are very basic and only cover enough for the 
specific programs. They aren't really intended to 
teach Tkinter.

http://www.crosswinds.net/~agauld
Look at Event Driven Progranmming and the Case Study

Alan G.

-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.


From "rod" at REMOVETHISinf.bme.hu  Mon Jul  3 16:09:25 2000
From: "rod" at REMOVETHISinf.bme.hu (Gabor Papp)
Date: Mon, 03 Jul 2000 16:09:25 +0200
Subject: lambda requirements
References: <000101bfe4ec$010e1c60$a5dfdfdf@ors-x-wt1.or_soft.de>
Message-ID: <39609ea1@news.caere.com>

there's a section about this in the FAQ
http://www.python.org/doc/FAQ.html#6.10
Gabor


From neilh at scintilla.org  Mon Jul  3 16:21:26 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Mon, 03 Jul 2000 14:21:26 GMT
Subject: Is there a searchable archive of c.l.py
Message-ID: <Gj185.6536$Tb7.39716@news-server.bigpond.net.au>

   When the recent thread on Aspect Oriented Programming started, I tried to
find some earlier discussion of the subject that I remembered. I knew that
Deja had dropped most of its older archives but tried anyway. The Python.org
site recommended eGroups but while this has a complete record, searches do
not go back very far. Maybe it is only partially indexed. None of the other
sources were searchable. And none had a good way to download a large block
of the archives so I could run a search on my own machine.

   Is there a sensibly searchable archive anywhere? In the future we will
need this resource. Otherwise how will we know how we beat those nanoviruses
back in '97?

   Neil




From aahz at netcom.com  Mon Jul  3 16:18:42 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 3 Jul 2000 14:18:42 GMT
Subject: From UNIX to Win NT
References: <39606DAC.49A502DB@muc.das-werk.de>
Message-ID: <8jq7c2$2f8$1@slb6.atl.mindspring.net>

In article <39606DAC.49A502DB at muc.das-werk.de>,
Thomas Thiele  <thiele at muc.das-werk.de> wrote:
>
>The problem is I am using threads, fork and exec. And I want to have the
>same behaviour using NT.

You'll need to look at the Win32 extensions to do the fork/exec stuff.
I've got zero experience with that, but hopefully either the docs are
good enough or Mark Hammond will be by shortly.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The best way to get information on Usenet is not to ask a question,
but to post the wrong information.  --Aahz


From michael.husmann at teleatlas.com  Mon Jul  3 16:17:35 2000
From: michael.husmann at teleatlas.com (Michael Husmann)
Date: Mon, 03 Jul 2000 16:17:35 +0200
Subject: extracting tuples from a list
Message-ID: <3960A07F.7240F23B@teleatlas.com>

Is there a simple way to extract two elements from a list
that has a size of 2n? I going this way:

while list:
	i, k = list.pop(0), list.pop(0)

Is there a better solution?


Thanks 

Michael

--
Michael Husmann
p.A. Tele Atlas Deutschland GmbH, Postfach 77 77 77, D-31132
Hildesheim
PHONE: +(49) 5121 49 4876, FAX: +(49) 5121 49 2866
e-mail: Michael.Husmann at teleatlas.com


From alan.gauld at gssec.bt.co.uk  Mon Jul  3 16:04:06 2000
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Mon, 03 Jul 2000 15:04:06 +0100
Subject: Newbie question  re: regsub
References: <395fc90a$1_1@huge.aa.net>
Message-ID: <39609D56.C710B311@gssec.bt.co.uk>

Robert Goto wrote:
> I am sure there is a very simple answer to this.
> >>> import regsub
> >>> fimo = open('fimo.txt')
> >>> while 1:
>       line = fimo.readline()
>       if not line: break
        newline = regsub.gsub('orange', 'red', line)
        print newlineline


> getting is both the original string and the changed string?!

What you are getting is gsub displaying its output because 
you haven't assigned it to anything and then you print the 
original string. Assign the result of gsub() to a variable 
then print that variable(as shown above). Note that because 
strings are immutable in Python(ie can't be changed) gsub 
*returns* the new string it does not(cannot!) change 'line'.

Alan g.

-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.


From dgoodger at bigfoot.com  Mon Jul  3 16:17:07 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Mon, 03 Jul 2000 10:17:07 -0400
Subject: lambda requirements
In-Reply-To: <000101bfe4ec$010e1c60$a5dfdfdf@ors-x-wt1.or_soft.de>
References: <000101bfe4ec$010e1c60$a5dfdfdf@ors-x-wt1.or_soft.de>
Message-ID: <B58618A2.5C13%dgoodger@bigfoot.com>

Lambda expressions don't have access to the local namespace. Declaring
_list2 as global gets around this problem, but in an "unclean" way (note
that _list1 doesn't need to be global, as it's not accessed inside the
lambda). A cleaner way to implement the lambda (without globals, or
"_"-prefixed local variables) is as follows:

    _onlylist1 = filter(lambda x, l=list2: x not in l, list1)

"l" is bound to "list2" when the lambda expression is evaluated (when the
local namespace *is* accessible), which carries through to when the
resulting anonymous function is executed by filter (when
SymmetricDifference's local namespace is *not* accessible).

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)


on 2000-07-03 08:41, Wolfgang Thaemelt (wolfgang.thaemelt at orsoft.de) wrote:
> Running the script:
> 
> #----------------------------------------- begin of script
> def SymmetricDifference(list1, list2):
> # global _list1
> # global _list2
> 
>   _list1 = list1
>   _list2 = list2
> 
>   _onlylist1 = filter(lambda x: x not in _list2, _list1)
> 
>   return _onlylist1
> 
> if __name__ == '__main__':
>   a = [1,2,3,4,5]
>   b = [3,4,5,6,7]
>   x = SymmetricDifference(a,b)
> 
>   print x
> #---------------------------------------- end of script
> 
> gives the error:
> 
> Traceback (innermost last):
> File "E:\up\000701\symmdiff.py", line 16, in ?
> x = SymmetricDifference(a,b)
> File "E:\up\000701\symmdiff.py", line 9, in SymmetricDifference
> _onlylist1 = filter(lambda x: x not in _list2, _list1)
> File "E:\up\000701\symmdiff.py", line 9, in <lambda>
> _onlylist1 = filter(lambda x: x not in _list2, _list1)
> NameError: _list2
> 
> If the comment characters "#"  are removed from the "global" statements
> everything works fine.
> 
> Is there some requirement on a lambda violated in the script?



From matthias.huening at univie.ac.at  Mon Jul  3 16:20:47 2000
From: matthias.huening at univie.ac.at (Matthias Huening)
Date: Mon, 3 Jul 2000 16:20:47 +0200
Subject: Again: ODBC / Win98 -> crash
Message-ID: <8jq7i6$395g$1@www.univie.ac.at>

Some days ago I posted a problem with ODBC on my Win98-machine. I thought
that it could be the win32-ODBC module that caused the crash I reported, so
I switched to Marc-Andre Lemburg's mxODBC module. To keep it simple I give
a minimal example (where dsn is a valid dsn):

>>> import ODBC.Windows
>>> db = ODBC.Windows.Connect(dsn)
>>> db.close()

Up to here, everything is fine. No problems so far.
But when I do it again:

>>> db = ODBC.Windows.Connect(dsn)

I get an immediate crash of Python.

My (German) version of Win98 tells me:

   PYTHONW verursachte einen Fehler durch eine ung?ltige Seite
   in Modul ODBCJT32.DLL bei 015f:04c0ac36.

It seems like something isn't closed properly when I do db.close().
Any suggestions?

Matthias





From aahz at netcom.com  Mon Jul  3 16:23:31 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 3 Jul 2000 14:23:31 GMT
Subject: Is there a searchable archive of c.l.py
References: <Gj185.6536$Tb7.39716@news-server.bigpond.net.au>
Message-ID: <8jq7l3$f1e$1@slb0.atl.mindspring.net>

In article <Gj185.6536$Tb7.39716 at news-server.bigpond.net.au>,
Neil Hodgson <neilh at scintilla.org> wrote:
>
>   Is there a sensibly searchable archive anywhere? In the future we will
>need this resource. Otherwise how will we know how we beat those nanoviruses
>back in '97?

Well, we've got the archive: http://www.python.org/pipermail/python-list/

Unfortunately, it only seems to go back to March '99.  Does anyone
maintain an older archive?  Once we have the archive, making it
searchable is a relatively trivial task.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The best way to get information on Usenet is not to ask a question,
but to post the wrong information.  --Aahz


From "rod" at REMOVETHISinf.bme.hu  Mon Jul  3 16:24:26 2000
From: "rod" at REMOVETHISinf.bme.hu (Gabor Papp)
Date: Mon, 03 Jul 2000 16:24:26 +0200
Subject: tracing a variable
References: <395BEC23.9A3097F0@uniserve.com>
Message-ID: <3960a227@news.caere.com>

there's a trace method of tkinter variables.
use like var.trace(ops, callback)
'callback' is called if 'var' is accessed in one of
the ways given by ops. operations are the same as in
tk ('w', 'r', 'u').
Gabor


From gmcm at hypernet.com  Mon Jul  3 16:28:21 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 03 Jul 2000 14:28:21 GMT
Subject: Memory leaks
References: <395F78BF.CBEAA090@synacon.ch>
Message-ID: <8F6667253gmcmhypernetcom@199.171.54.154>

Ingo Adler wrote:

[strange leaks in SWIG'd code]

>I have some simple Python-Code which generated memory leaks (ca. 4MB):
>(The real code is more complicated, but I could strip it down to this
>simple example = the smallest example with the memory leak.)
>//---------------------------
>x = X()
>
>for i in range(100000):
>    y = x.getY()
>    y.getNumber()
>//---------------------------

Doing things in the "obvious" way from your (snipped) incomplete code, I 
get no leak when using MSVC 5.

- Gordon


From aahz at netcom.com  Mon Jul  3 16:28:59 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 3 Jul 2000 14:28:59 GMT
Subject: extracting tuples from a list
References: <3960A07F.7240F23B@teleatlas.com>
Message-ID: <8jq7vb$qo3$1@nntp9.atl.mindspring.net>

In article <3960A07F.7240F23B at teleatlas.com>,
Michael Husmann  <michael.husmann at teleatlas.com> wrote:
>
>Is there a simple way to extract two elements from a list
>that has a size of 2n? I going this way:
>
>while list:
>	i, k = list.pop(0), list.pop(0)
>
>Is there a better solution?

How was the list created?  Can you create it in such a way that each
element is a tuple?  Example:

list = []
for line in file.readlines()
  list.append(FunctionThatReturnsATuple(line))
for item in list:
  (i,k) = item
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The best way to get information on Usenet is not to ask a question,
but to post the wrong information.  --Aahz


From thiele at muc.das-werk.de  Mon Jul  3 16:36:03 2000
From: thiele at muc.das-werk.de (Thomas Thiele)
Date: Mon, 03 Jul 2000 16:36:03 +0200
Subject: making widgets invisible
References: <3960794E.6E2B36E6@muc.das-werk.de> <8jq5bf$h45$1@nnrp1.deja.com>
Message-ID: <3960A4D3.F45133B4@muc.das-werk.de>

No I think this is not sooo nice. Becouse I want to replace the widget
with others.

mode One

|   not changing area           |
--------------------
|   Label    Entry   Button   |
|                                           |
--------------------


and then I want to change the layout into:
mode Two

|   not changing area         |
--------------------
|   Entry   Button  Entry      |
|                                           |
--------------------

(For instance.)
Just like an other dialog but within the mainframe.

And I want to switch between the two modes.




From gmcm at hypernet.com  Mon Jul  3 16:42:32 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 03 Jul 2000 14:42:32 GMT
Subject: Again: ODBC / Win98 -> crash
References: <8jq7i6$395g$1@www.univie.ac.at>
Message-ID: <8F666B563gmcmhypernetcom@199.171.54.154>

Matthias Huening wrote: 

>Some days ago I posted a problem with ODBC on my Win98-machine. I
>thought that it could be the win32-ODBC module that caused the crash I
>reported, so I switched to Marc-Andre Lemburg's mxODBC module. To keep
>it simple I give a minimal example (where dsn is a valid dsn):
>
>>>> import ODBC.Windows
>>>> db = ODBC.Windows.Connect(dsn)
>>>> db.close()
>
>Up to here, everything is fine. No problems so far.
>But when I do it again:
>
>>>> db = ODBC.Windows.Connect(dsn)
>
>I get an immediate crash of Python.
>
>My (German) version of Win98 tells me:
>
>   PYTHONW verursachte einen Fehler durch eine ung?ltige Seite
>   in Modul ODBCJT32.DLL bei 015f:04c0ac36.
>
>It seems like something isn't closed properly when I do db.close().
>Any suggestions?

All this points to it being the driver. That is, your dsn gets looked up, 
and the driver for the appropriate DB is loaded. Since that's the JET 
driver (ie, the one that translates ODBC to JET), you can thank your lucky 
stars that it works at all <wink>. Access behaves very poorly through ODBC. 
If you really have to use Access, you might find ADO more satisfactory.

- Gordon


From olivierS.dagenaisP at canadaA.comM  Mon Jul  3 16:42:00 2000
From: olivierS.dagenaisP at canadaA.comM (Olivier Dagenais)
Date: Mon, 03 Jul 2000 14:42:00 GMT
Subject: Win32 : Making script.py executable in DOS box
References: <8jmjfe$64o$1@nnrp1.deja.com> <WzI75.5714$Tb7.31718@news-server.bigpond.net.au> <8jp3pt$q7j$1@nnrp1.deja.com>
Message-ID: <YC185.27369$W35.636919@news20.bellglobal.com>

How about a batch file called p.bat ?  (with optional status messages)

-----start p.bat--------
@echo off
echo Starting script '%1.py' ....
c:\progra~1\Python\python "%1.py" %2 %3 %4 %5 %6 %7 %8 %9
echo Execution complete!
----end p.bat-----------


Given you have a script called "spam.py":

-----start spam.py------
print "Spam spam spam spam"
-----end spam.py--------


Then, you would type:

c:\work\python\stuff>p spam
Starting script 'spam.py' ...
Spam spam spam spam
Execution complete!


Is that what you wanted??

--
----------------------------------------------------------------------
Olivier A. Dagenais - Carleton University - Computer Science III


"Neo Brave" <neo_brave at my-deja.com> wrote in message
news:8jp3pt$q7j$1 at nnrp1.deja.com...
> In article <WzI75.5714$Tb7.31718 at news-server.bigpond.net.au>,
>   "Mark Hammond" <MarkH at ActiveState.com> wrote:
> > Under Windows NT or Win2k, you can use the PATHEXT environment
> > variable. Simply add ".py" to the semi-colon sep'd list of
> > extensions, and off you go!
>
> Great, thats a help for NT, but I also want to do this on Win98 :-).
>
> > No solution I am aware of for 95, although there are some tricks
> > you can pull (Im sure someone else will fill these in for you)!
>
> Anyone???
>
> Neo
> --
> Neo Brave
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From kalle at gnupung.net  Mon Jul  3 16:28:38 2000
From: kalle at gnupung.net (Kalle Svensson)
Date: Mon, 3 Jul 2000 16:28:38 +0200
Subject: Random number generation, simple question.
Message-ID: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se>

Hello.

I have a few questions about the (pseudo) random number generator supplied
in the standard library. I need 80 random bits (for a CipherSaber IV) but
I hear most RNG's only supply 32 bits or less. My questions are:

 * How many random bits does whrandom.random() generate?

 * Suppose I do a loop where I get a byte at a time and then re-seed the
   generator either with random numbers from the generator or with values
   from time. Will that improve the randomness?

 * Are there better alternatives? I understand I could read from
   /dev/random on Linux, but this isn't portable, and even though I only
   use GNU/Linux myself, I have friends who don't. If anyone can tip me
   off about a similar feature in Win32, I could just detect platform and
   use the appropriate function, of course. I don't think I have to care
   about Macs or other *NIX dialects. Phew... :)

Hope I'm making sense even though I hardly understand what I'm talking
about... :)

TIA,
  Kalle Svensson



From piet at cs.uu.nl  Mon Jul  3 16:50:38 2000
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 03 Jul 2000 16:50:38 +0200
Subject: lambda requirements
References: <000101bfe4ec$010e1c60$a5dfdfdf@ors-x-wt1.or_soft.de>
Message-ID: <ulmzjqncx.fsf@cs.uu.nl>

>>>>> "Wolfgang Thaemelt" <wolfgang.thaemelt at orsoft.de> (WT) writes:

WT> Running the script:
WT> #----------------------------------------- begin of script
WT> def SymmetricDifference(list1, list2):
WT> #      global _list1
WT> #      global _list2

WT>       _list1 = list1
WT>       _list2 = list2

WT>       _onlylist1 = filter(lambda x: x not in _list2, _list1)

WT>       return _onlylist1

WT> if __name__ == '__main__':
WT>   a = [1,2,3,4,5]
WT>   b = [3,4,5,6,7]
WT>   x = SymmetricDifference(a,b)

WT>   print x
WT> #---------------------------------------- end of script

WT> gives the error:

WT> Traceback (innermost last):
WT>   File "E:\up\000701\symmdiff.py", line 16, in ?
WT>     x = SymmetricDifference(a,b)
WT>   File "E:\up\000701\symmdiff.py", line 9, in SymmetricDifference
WT>     _onlylist1 = filter(lambda x: x not in _list2, _list1)
WT>   File "E:\up\000701\symmdiff.py", line 9, in <lambda>
WT>     _onlylist1 = filter(lambda x: x not in _list2, _list1)
WT> NameError: _list2

WT> If the comment characters "#"  are removed from the "global" statements
WT> everything works fine.

WT> Is there some requirement on a lambda violated in the script?

Python has only two scopes: the local variables of the function (which is
the lambda in this case) and the global ones of the module. (O.K. the
builtins also).
There are no intermediate scopes visible, in your case that would be the
local variables of SymmetricDifference.
The usual trick is to copy these things as additional parameters of the
lambda, where you can even use the same name.
In your case only list2 has to be copied, there is no need to copy list1.
Also the variable for the result is not needed, so this is a compacr
definition that works:

def SymmetricDifference(list1, list2):
      return filter(lambda x,_list2=list2: x not in _list2, list1)

Using list2 rather than _list2 would also work.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From piet at cs.uu.nl  Mon Jul  3 16:52:47 2000
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 03 Jul 2000 16:52:47 +0200
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <8jabvo014tv@news2.newsguy.com>
Message-ID: <uhfa7qn9c.fsf@cs.uu.nl>

>>>>> "Alex Martelli" <alex at magenta.com> (AM) writes:

AM> Don Tuttle <tuttledon at hotmail.com> wrote in message
AM> news:qJ165.8938$nZ5.153196 at typhoon.southeast.rr.com...
>> Consider this a heads up NOT a promotion!!

AM> Why would the Python community care about a Microsoft-proprietary
AM> quasi-Java-clone...?

It is not proprietary. They will put the language definition in the public
domain. Maybe not the implementation.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From garry at sage.att.com  Mon Jul  3 16:42:22 2000
From: garry at sage.att.com (Garry Hodgson)
Date: Mon, 3 Jul 2000 14:42:22 GMT
Subject: Tutorials at IPC9
References: <39591877.535A96BC@sage.att.com> <9i80ms4tbkjjhp90t55au011qquvmsr456@4ax.com>
Message-ID: <3960A64E.E7794249@sage.att.com>

Matt wrote:
> 
> On Tue, 27 Jun 2000 21:11:19 GMT, Garry Hodgson <garry at sage.att.com>
> wrote:
> 
> >I'm starting to think about what kinds of tutorials we should have at
> >the next Python conference.
> 
> Is there information available on when and where this will be?

not yet.

-- 
Garry Hodgson                   Every night 
garry at sage.att.com                a child is born
Software Innovation Services        is a Holy Night.
AT&T Labs                         - Sophia Lyon Fahs


From gmcm at hypernet.com  Mon Jul  3 16:57:28 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 03 Jul 2000 14:57:28 GMT
Subject: From UNIX to Win NT
References: <39606DAC.49A502DB@muc.das-werk.de> <8F665C79Agmcmhypernetcom@199.171.54.154> <39608FE0.35950741@muc.das-werk.de>
Message-ID: <8F666C0AEgmcmhypernetcom@199.171.54.154>

Thomas Thiele wrote: 

>>
>> You've got both technical and cultural problems here. There is no
>> fork. 
>
>I knew that. But I wonder how to start an other program. In UNIX I used
>fork() and exec().
>Just the same code like in popen2.popen3 but I copied the code and
>included some costumization.
>Does popen work on NT? Or is there an other way do get the same result
>(starting a program and
>read it's output)

All of the c runtime ways of starting processes have problems from annoying 
to catastrophic on Windows. The best of the lot is probably os.execv. 
os.popen works between well-behaved console processes on NT. If you want 
fine-grained control (including dup'ing your file descriptors etc.), use 
Win32process.CreateProcess from the Win32 extensions. Win32pipe.popen* use 
it to emulate os.popen*.

But the real question is why are you doing this? While it's a common and 
reasonably efficient way of doing things on *nix, it is both rare and very 
expensive on Windows.


From alex at magenta.com  Mon Jul  3 16:43:00 2000
From: alex at magenta.com (Alex Martelli)
Date: Mon, 3 Jul 2000 16:43:00 +0200
Subject: Class design: accessing "private" members
References: <962352843.1667731268@news.libertysurf.fr><8jhuvn01vs0@news2.newsguy.com> <4.3.1.2.20000630145109.0201d9e0@phear.dementian.com> <8jplru0lfe@news2.newsguy.com> <962626013.769258863@news.libertysurf.fr>
Message-ID: <8jq90e014sb@news2.newsguy.com>

Jerome Quelin <jerome.quelin at insalien.org> wrote in message
news:962626013.769258863 at news.libertysurf.fr...
> "Alex Martelli" <alex at magenta.com> wrote:
> >These are excellent arguments for always using accessors from OUTSIDE the
> >class.  Jerome, however, was talking about using them from INSIDE the
class
> >as well, which is a different issue.
> Sure, I'm always using accessors from outside.

Bravo.  One day you'll be happy you did.

> But from inside the class, I don't know which is the best. I feel like the
KISS
> rule should apply, but what do you OO gurus think of it ? How far
encapsulation
> should go ?

"Do the simplest thing that could possibly work".  For a class
not meant to be inherited from, that implies accessing the
attributes directly until and unless very specific reasons
emerge such that this "cannot possibly work" any more.  It is
quite unlikely that such reasons will ever emerge, and, if and
when they ever do, then refactoring your program is not going
to be a horribly huge cost to pay.  99+% of the time you will
enjoy the simplicity and the resulting reduction in hassles,
and the price of refactoring is not disproportionate in those
cases where it will be needed.  (The tradeoffs might be
somewhat different in other languages, but dynamic ones
such as Python strongly enable this excellent approach).


Alex





From matthias.huening at univie.ac.at  Mon Jul  3 17:25:55 2000
From: matthias.huening at univie.ac.at (Matthias Huening)
Date: Mon, 3 Jul 2000 17:25:55 +0200
Subject: Again: ODBC / Win98 -> crash
References: <8jq7i6$395g$1@www.univie.ac.at> <8F666B563gmcmhypernetcom@199.171.54.154>
Message-ID: <8jqbbn$38cc$1@www.univie.ac.at>

Gordon McMillan:
>Since that's the JET driver (ie, the one that translates ODBC to
>JET), you can thank your lucky stars that it works at all <wink>.

:-)

>Access behaves very poorly through ODBC. If you really have to
>use Access, you might find ADO more satisfactory.

Well, it has to be Access for this project, but I certainly would like to
get suggestions for better alternatives on Windows. (But it has to be as
comfortable and simple as Access for my colleagues who only see and work
with the database frontend.)

Now for a really strange thing: I just moved my script to a different
computer (almost identical configuration), and... - everything works
just fine! No crashes, no problems whatsoever. I hate computers.

Matthias











From jerome.quelin at insalien.org  Mon Jul  3 17:28:30 2000
From: jerome.quelin at insalien.org (Jerome Quelin)
Date: Mon, 3 Jul 2000 17:28:30 +0200
Subject: making widgets invisible
References: <3960794E.6E2B36E6@muc.das-werk.de> <8jq5bf$h45$1@nnrp1.deja.com> <3960A4D3.F45133B4@muc.das-werk.de>
Message-ID: <962638110.1459980735@news.libertysurf.fr>

Thomas Thiele <thiele at muc.das-werk.de> wrote:
>No I think this is not sooo nice. Becouse I want to replace the widget
>with others.
What about the pack_forget method ?

-- File pack_forget.py
#!/usr/bin/python
from Tkinter import *
class F(Frame):
    def __init__(self):
        Frame.__init__(self)
        Button(self,text='change', command=self.change).pack()
        self.f1 = Frame(self)
        self.f2 = Frame(self)
        Label(self.f1, text='eggs', bg='red').pack(side=LEFT)
        Label(self.f1, text='spam', bg='green').pack(side=LEFT)
        Label(self.f2, text='spam', bg='blue').pack(side=LEFT)
        Label(self.f2, text='eggs', bg='yellow').pack(side=LEFT)
        self.f1.pack()
        self.visible = 1
        self.pack()
    def change(self):
        if self.visible == 1:
            self.f1.pack_forget()
            self.f2.pack()
            self.visible = 2
        else:
            self.f2.pack_forget()
            self.f1.pack()
            self.visible = 1
F().mainloop()
-- End of file

The beauty of pack_forget is that it just unmaps the specified widget, and
keeps the internal layout of the widget.

 Regards,
Jerome
--
jerome.quelin at insalien.org


From jvickroy at sec.noaa.gov  Mon Jul  3 17:37:36 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Mon, 03 Jul 2000 09:37:36 -0600
Subject: From UNIX to Win NT
References: <39606DAC.49A502DB@muc.das-werk.de> <8jq7c2$2f8$1@slb6.atl.mindspring.net>
Message-ID: <3960B33F.98B23C40@sec.noaa.gov>

Mark isn't here yet but his book (Python Programming on Win 32) is and I quote
(from page 35):

"
...
In many cases, the best solution is to rebuild this part of the application for
Windows.  On a typical Unix machine, creating new processes (as fork() does) is a
relatively cheap operation.  On Windows, howerver, the overhead of creating a new
process is significant, so solutions using this architecture on Windows won't
often scale or perform as you would hope.  Rewriting this portion of the
application to use multiple threads instead of multiple processes is often the
best solution for Windows.
"

Aahz Maruch wrote:

> In article <39606DAC.49A502DB at muc.das-werk.de>,
> Thomas Thiele  <thiele at muc.das-werk.de> wrote:
> >
> >The problem is I am using threads, fork and exec. And I want to have the
> >same behaviour using NT.
>
> You'll need to look at the Win32 extensions to do the fork/exec stuff.
> I've got zero experience with that, but hopefully either the docs are
> good enough or Mark Hammond will be by shortly.
> --
>                       --- Aahz (Copyright 2000 by aahz at netcom.com)
>
> Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
>
> The best way to get information on Usenet is not to ask a question,
> but to post the wrong information.  --Aahz



From olivierS.dagenaisP at canadaA.comM  Mon Jul  3 18:15:48 2000
From: olivierS.dagenaisP at canadaA.comM (Olivier Dagenais)
Date: Mon, 03 Jul 2000 16:15:48 GMT
Subject: Random number generation, simple question.
References: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se>
Message-ID: <U_285.27570$W35.647582@news20.bellglobal.com>

I'm not sure feeding "random" numbers as the seed will keep producing random
numbers.

You may want to try scanning different components of the computer, in order
to retrieve values that, chances are, will be different from machine to
machine and from time to time, like how many files are there on the
computer, what the address of the network card is, how long has the machine
been running, what keys has the user pressed, how has the mouse or joystick
been moved, etc... ???

--
----------------------------------------------------------------------
Olivier A. Dagenais - Carleton University - Computer Science III


"Kalle Svensson" <kalle at gnupung.net> wrote in message
news:Pine.GSO.4.21.0007031546410.25965-100000 at luna.dsv.su.se...
> Hello.
>
> I have a few questions about the (pseudo) random number generator supplied
> in the standard library. I need 80 random bits (for a CipherSaber IV) but
> I hear most RNG's only supply 32 bits or less. My questions are:
>
>  * How many random bits does whrandom.random() generate?
>
>  * Suppose I do a loop where I get a byte at a time and then re-seed the
>    generator either with random numbers from the generator or with values
>    from time. Will that improve the randomness?
>
>  * Are there better alternatives? I understand I could read from
>    /dev/random on Linux, but this isn't portable, and even though I only
>    use GNU/Linux myself, I have friends who don't. If anyone can tip me
>    off about a similar feature in Win32, I could just detect platform and
>    use the appropriate function, of course. I don't think I have to care
>    about Macs or other *NIX dialects. Phew... :)
>
> Hope I'm making sense even though I hardly understand what I'm talking
> about... :)
>
> TIA,
>   Kalle Svensson
>




From Norman_Shelley-RRDN60 at email.sps.mot.com  Mon Jul  3 17:58:14 2000
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Mon, 03 Jul 2000 08:58:14 -0700
Subject: UserDict and UserList issue (bug?_
References: <395D333D.9240593C@email.sps.mot.com> <slrn8lqh0n.gqv.thomas@centurion.xs4all.nl>
Message-ID: <3960B816.BC1B6577@email.sps.mot.com>

Thomas Wouters wrote:

> On Fri, 30 Jun 2000 16:54:37 -0700, Norman Shelley
> <Norman_Shelley-RRDN60 at email.sps.mot.com> wrote:
> >in UserDict and UserList there is one or more places where a call to
> >self's __class__() method is made, e.g.
> >
> >def __getslice__(self, i, j):
> >  i = max(i, 0); j = max(j, 0)
> >  # This class and its subclasses need more arguments then 0
> >  userlist = self.__class__()
> >  ....
>
> >This blows causes major problems when either of these two classes are
> >subclassed and the class which subclasses them requires creation
> >arguments.  I think maybe a copy of self instead of a __class__() may be
> >a more robust solution.
>
> And how would you get a copy of self without creating a new instance of
> self.__class__ ?

Simple.

import copy
...

userlist = copy.copy(self)



> You usually create a copy of a list by doing:
>
>     mycopy = mylist[:]
>
> Which calls __getslice__, which is exactly what you are complaining about
> :-)
>
> If you want to subclass UserList, either make sure you provide defaults for
> all arguments to the initializer, or make sure you override all places where
> self.__class__() may be called without your required arguments. #1 is
> probably the best (and most portable) solution, and not just because of the
> above issue: providing defaults for all initializer arguments allows your
> object to be pickled, for instance ;)
>
> But I dont know why you need those arguments, so YMMV.
>
> Regards,
>         Thomas.



From alex at magenta.com  Mon Jul  3 17:31:43 2000
From: alex at magenta.com (Alex Martelli)
Date: Mon, 3 Jul 2000 17:31:43 +0200
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <8jabvo014tv@news2.newsguy.com> <uhfa7qn9c.fsf@cs.uu.nl>
Message-ID: <8jqbru017gf@news2.newsguy.com>

<piet at cs.uu.nl> wrote in message news:uhfa7qn9c.fsf at cs.uu.nl...
> >>>>> "Alex Martelli" <alex at magenta.com> (AM) writes:
>
> AM> Don Tuttle <tuttledon at hotmail.com> wrote in message
> AM> news:qJ165.8938$nZ5.153196 at typhoon.southeast.rr.com...
> >> Consider this a heads up NOT a promotion!!
>
> AM> Why would the Python community care about a Microsoft-proprietary
> AM> quasi-Java-clone...?
>
> It is not proprietary. They will put the language definition in the public
> domain. Maybe not the implementation.

OK, let's assume they do.  And so...?  It's a Java dialect
in anything but name, and not a very interesting one either
("Pizza" added much more significant enhancements a LONG
time ago, for example).

I believe the only relevant issue for the Python community
is: C#, like every other Microsoft offering in recent times,
will be *STRONGLY* oriented to COM.  Every C# object will be
a COM object (just like every Java object was one when run
under Microsoft's Java VM), etc, etc.

So, C#'s possible (unlikely...) emergence as a major language
is just the N-plus-1-th motivation for Windows versions of
Python to support COM *very, VERY* well.  There are many
other reasons which apply *right now*, of course.  Python's
COM support is good, but I think it could be better.  Type
libraries are no doubt (at some point in the future) going
to be obsoleted in favour of other, fuller meta-information,
but until that happens I think any enhancement in support
for them is going to be a very significant plus for Python,
for example.  Custom (non-dual) interfaces may be in decline,
but, again, there is very substantial need for support for
them *today* -- Python supports a lot of them, but, again,
I'd like to see more -- more ease in interfacing a given set
of interfaces to Python (maybe by autobuilding a .PYD for
them?), more ease in generating 'stub' implementations for
them to which Python substance can be added, ...

Yeah, I know, I should put my coding efforts where my mouth
is -- and, having expended substantial efforts doing just
this sort of thing for a proprietary scripting language
(which I'm pushing _hard_ to be "stabilized" in the future
in favour of much better languages such as Python!), I do
have the skills and experience... on the COM side of things,
at least; haven't really looked at Python's internals yet,
and it's been so long since I last programmed in C (as
opposed to C++) that I wonder if I still can...:-).

But anyway, that's my personal take on the C# "impact"
(comparable to that of a goosedown pillow) on Python.

Re .NET -- I dunno, and I wish I did!  Any relevant URLs
will be GRATEFULLY received - I could post a dozen, but
I'll do y'all a favour and summarize them into basically
two stances:
    "It's Microsoft and thus evil, vapourware, immoral,
        and/or fattening"
    "It's Microsoft and thus good, splendid, progressive,
        and/or sexy"

Nothing REAL, TECHNICAL, DETAILED, nothing one could sink
one's teeth into, sigh.  Well, apart from XML, and SOAP
in particular, but those we already knew as today's silver
bullets (and, seriously, they do have real technical
substance, of course; the more Microsoft's architectures
get "open" by such means, the better -- but Python is
already moving fast towards excellent support of XML and
even SOAP, so there's no course-change in sight because
of those, is there?  Just, once again, extra motivation
to do what's already being done).

I do wonder what other items that have technical reality
are in .NET besides C#, XML, and SOAP.  If any, that is.


Alex





From thiele at muc.das-werk.de  Mon Jul  3 18:28:55 2000
From: thiele at muc.das-werk.de (Thomas Thiele)
Date: Mon, 03 Jul 2000 18:28:55 +0200
Subject: making widgets invisible
References: <3960794E.6E2B36E6@muc.das-werk.de> <8jq5bf$h45$1@nnrp1.deja.com> <3960A4D3.F45133B4@muc.das-werk.de> <962638110.1459980735@news.libertysurf.fr>
Message-ID: <3960BF47.AF251B0B@muc.das-werk.de>

>

I think this pack_forget methode is what I was looking for.
I'll try it.

Thanks Thomas




From akuchlin at mems-exchange.org  Mon Jul  3 18:33:56 2000
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 03 Jul 2000 12:33:56 -0400
Subject: Python CGI security
References: <B5851962.5E%bojar@speakeasy.org>
Message-ID: <3dr99b5g23.fsf@kronos.cnri.reston.va.us>

"E. Jordan Bojar" <bojar at speakeasy.org> writes:
> I'm interested in running a series of Python scripts via the Apache Python
> module, mostly for dynamic content but possibly for password protection as
> well.  Are there security issues particular to (or common with) Python CGIs,
> in the way that buffer overflow is so common to poorly-written Perl?

Actually I'd expect that very few Perl security problems stem from
buffer overflows, but instead come from improperly screening user
input.  (Taking the value of a field and blithely opening a file with
the same name, for example, which is vulnerable to values such as
../../../etc/passwd.)  Similarly, while there may be buffer overflows
inside the Python interpreter that should be fixed, I'd be far more
worried about the CGI/Apache code itself.

--amk


From thiele at muc.das-werk.de  Mon Jul  3 18:32:17 2000
From: thiele at muc.das-werk.de (Thomas Thiele)
Date: Mon, 03 Jul 2000 18:32:17 +0200
Subject: From UNIX to Win NT
References: <39606DAC.49A502DB@muc.das-werk.de> <8F665C79Agmcmhypernetcom@199.171.54.154> <39608FE0.35950741@muc.das-werk.de> <8F666C0AEgmcmhypernetcom@199.171.54.154>
Message-ID: <3960C011.84B3D888@muc.das-werk.de>

The program is the front-end of a job - management system.
The program has to receive a user - command (via pickling over
a tcp-socket connection) and execute a special program and evaluate
the output.

And this program runs on NT too. And so the starting server has to run on
NT too.



From jhe at webde-ag.de  Mon Jul  3 18:40:25 2000
From: jhe at webde-ag.de (Jürgen Hermann)
Date: Mon, 3 Jul 2000 18:40:25 +0200
Subject: Random number generation, simple question.
References: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se>
Message-ID: <8jqfis$eak$1@pollux.ip-plus.net>

"Kalle Svensson" <kalle at gnupung.net> schrieb im Newsbeitrag
news:Pine.GSO.4.21.0007031546410.25965-100000 at luna.dsv.su.se...
>  * Suppose I do a loop where I get a byte at a time and then re-seed the
>    generator either with random numbers from the generator or with values
>    from time. Will that improve the randomness?

Certainly not. Seeding is done ONCE, always. What you should do is to call
random() 80 times, and use the MSB only, i.e. "bit = random() < 0.5".

Ciao, J?rgen

--
J?rgen Hermann (jhe<at>webde-ag.de)
WEB.DE AG, Amalienbadstr.41, D-76227 Karlsruhe
Tel.: 0721/94329-0, Fax: 0721/94329-22




From dpetrov at www.yu  Mon Jul  3 18:50:49 2000
From: dpetrov at www.yu (Dean Petrov)
Date: Mon, 3 Jul 2000 18:50:49 +0200
Subject: Selling software
Message-ID: <8jqge7$irv$1@cer.yubc.net>

!!! Software WareHouse !!!


We're selling software!
Best prices!

Under $ 100 free shipping!!!


Order today: dpetrov at www.yu





From egbert at bork.demon.nl  Mon Jul  3 17:17:23 2000
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Mon, 3 Jul 2000 17:17:23 +0200
Subject: Books
In-Reply-To: <8jpvhi$jfs$1@newshost.accu.uu.nl>; from Martijn Faassen on Mon, Jul 03, 2000 at 12:05:06PM +0000
References: <8jb1ec$9mh2@news.qualitynet.net> <autilso0l591oo924n3mogcva7r2892p5b@4ax.com> <8jpvhi$jfs$1@newshost.accu.uu.nl>
Message-ID: <20000703171722.A7614@bork>

On Mon, Jul 03, 2000 at 12:05:06PM +0000, Martijn Faassen wrote:
> 
> I'm not sure how good Learning Python is at teaching OO concepts to people
> who don't know about it yet. *Python* would be good though. Any opinions?
> 
The PizzaRobot in Learning Python gave me a better idea what OO in Python
is about, and what it can be used for, than anything else I have read so far.
But it didn't teach me OO. Either I have to do a lot of thinking and 
experimenting myself, or I have to wait for a book with more ideas, 
concepts, design etc, and not in the first place tricks and techniques.

Actually a great eye opener was the remark in this list a few days ago,
when someone said that you use classes when you have to combine 
program and data, and functions when you have only program code.
Or something like that. For what is a program without data.
egbert
-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================



From mmorris at mindspring.com  Mon Jul  3 18:41:54 2000
From: mmorris at mindspring.com (Mitchell Morris)
Date: 3 Jul 2000 16:41:54 GMT
Subject: Random number generation, simple question.
References: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se>
Message-ID: <8F6665DDDbeablebeable@207.211.168.82>

kalle at gnupung.net (Kalle Svensson) wrote in 
<Pine.GSO.4.21.0007031546410.25965-100000 at luna.dsv.su.se>:

>Hello.
>
>I have a few questions about the (pseudo) random number generator supplied
>in the standard library. I need 80 random bits (for a CipherSaber IV) but
>I hear most RNG's only supply 32 bits or less. My questions are:
>
> * How many random bits does whrandom.random() generate?

I don't have the references from whrandom handy, but (a) I can flip through 
my Knuth book (flip, flip, flip) to see that there are several constants by 
which you can get an LCM to generate 2^32 values each of which could be 32 
bits long, and (b) Wichmann-Hill is rather more advanced than an LCM, so 32
*2^32 should be considered a reasonable lower limit. Since you only need 80 
or so, you should be okay.



> * Suppose I do a loop where I get a byte at a time and then re-seed the
>   generator either with random numbers from the generator or with values
>   from time. Will that improve the randomness?

The short answer is (a) no, that won't actually improve anything, and (b) 
it would be overkill anyway. All you need to do is:
>>> import random
>>> session_key = []
>>> for i in range(10):
...     session_key.append(random.randint(0, 255))
and go about your business normally.

The long answer is you can't generate randomness from nothing, which is 
what you're trying to do by reseeding your generator with output from your 
generator. If you really need more entropy, you need to find it somewhere 
else, perhaps by measuring some physical properties. To answer your next 
question, this is what /dev/random does ... it distills single bits of 
entropy by measuring a sheaf of physical properties as the system runs, 
then mixes them with the bits it already has, and doles them out to you 
when asked.

There is a discussion about both of these in the design of Yarrow 
(http://www.counterpane.com/yarrow-notes.html) which may be of interest to 
you if you're looking to further your knowledge of your ignorance of 
generating pseudo-random numbers.



> * Are there better alternatives? I understand I could read from
>   /dev/random on Linux, but this isn't portable, and even though I only
>   use GNU/Linux myself, I have friends who don't. If anyone can tip me
>   off about a similar feature in Win32, I could just detect platform and
>   use the appropriate function, of course. I don't think I have to care
>   about Macs or other *NIX dialects. Phew... :)
>
>Hope I'm making sense even though I hardly understand what I'm talking
>about... :)
>
>TIA,
>  Kalle Svensson

It's been a while since I looked at CipherSaber, but as I recall you only 
need entropy for the session key. You will probably find that almost any 
silly stupid PRNG will suffice for that, and most operating systems will 
provide you a silly stupid one.

My knee-jerk response is don't use rand()/srand() if random()/srandom() are 
available. I suspect, and this is completely without analysis, that 
whrandom.random() will be more than enough for this exercise.

If human lives are at stake, however, you really should try to educate 
yourself further so you can make your own analysis of the risks involved.

HTH. HAND.
+Mitchell


From thor at localhost.localdomain  Mon Jul  3 19:08:24 2000
From: thor at localhost.localdomain (Manuel Gutierrez Algaba)
Date: 3 Jul 2000 17:08:24 GMT
Subject: Socket problem
Message-ID: <slrn8m1p8s.2e9.thor@localhost.localdomain>

I'm doing a game for my University. It's divided into 3 parts
( AI (clips), core(python), GUI(python)) communicated by sockets.
I've been running it in Red Hat 4.2 happily. 
Now, I'm switching to Red Hat 6.0 and the core when 
trying to access to port 50008 it says: 

ocket.error: (99, 'Cannot assign requested address')
Exception exceptions.AttributeError: 'conn' in 
<method servidor_de_socket.__del__ of servidor_de_socket instance at 8194e40>
 ignored
Exception exceptions.AttributeError:
 'delete non-existing instance attribute' in 
<method nucleo.__del__ of nucleo instance at 81939f8> ignored

Where is the fault ?

Thanks.


-- 
MGA


From montagne at boora.com  Mon Jul  3 19:41:51 2000
From: montagne at boora.com (michael montagne)
Date: Mon, 3 Jul 2000 10:41:51 -0700
Subject: WinNT User Name
Message-ID: <Ae485.94210$5k2.194235@dfw-read.news.verio.net>

First off, I'm a newbie at Python.
I'm trying to return the user who is logged on at a PC.  What I really need
is the login name.  How do I do that?  Seems like it otta be quite simple
but I'm having trouble.

-mjm




From montagne at boora.com  Mon Jul  3 19:47:30 2000
From: montagne at boora.com (michael montagne)
Date: Mon, 3 Jul 2000 10:47:30 -0700
Subject: Web server executing Python scripts
References: <2B1262E83448D211AE4B00A0C9D61B030130125E@msgeuro1.creo.be> <80EB8C0C818F320D.2A60B0748B6BCCB3.6A1C28BCC752E3D8@lp.airnews.net>
Message-ID: <Tj485.94211$5k2.194544@dfw-read.news.verio.net>

How can I do that?  I'd like to report on whether certain programs are
installed on local PC's.

-mjm


"Cameron Laird" <claird at starbase.neosoft.com> wrote in message
news:80EB8C0C818F320D.2A60B0748B6BCCB3.6A1C28BCC752E3D8 at lp.airnews.net...
> In article <2B1262E83448D211AE4B00A0C9D61B030130125E at msgeuro1.creo.be>,
> Pieter Claerhout  <PClaerhout at CREO.BE> wrote:
> >Hello all,
> >
> >I'm working on a web based system, which consist of a whole
> >bunch of Python scripts, which connect to a bunch of machines
> >on a network (either using ethernet or dailup networking) and
> >execute some to code to for example get things from the
> >windows registry of those machines.
> >
> >What do you think would be the best idea?
> >- Use whatever webbrowser with CGI-scripting
> >- Use MS IIS with ASP (and Python)
> >- Use Zope to build the whole system?
> >
> >It should also connect to a database running on the host.
> >
> >
> >Pieter
> >
>
> Others have already demonstrated that there's a large
> ... situational, if not subjective, component to the
> decision about which Web server to use.
>
> Your first paragraph interests me.  Do you realize
> that Python can retrieve registry information directly
> through your LAN, without a need for installation of
> "some to code" on the reporting hosts?
> --
>
> Cameron Laird <claird at NeoSoft.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html




From david_ullrich at my-deja.com  Mon Jul  3 19:37:45 2000
From: david_ullrich at my-deja.com (david_ullrich at my-deja.com)
Date: Mon, 03 Jul 2000 17:37:45 GMT
Subject: Bug in math.frexp?
References: <20000630221759.A1965@xyz.donetsk.ua> <LNBBLJKPBEHFEDALKOLCCELGGHAA.tim_one@email.msn.com> <14688.12945.528521.351944@cj42289-a.reston1.va.home.com>
Message-ID: <8jqj17$rl0$1@nnrp1.deja.com>

In article <14688.12945.528521.351944 at cj42289-a.reston1.va.home.com>,
  "Fred L. Drake, Jr." <fdrake at beopen.com> wrote:
>
> Tim Peters writes:
>  > The docs are wrong.  I just checked in a repaired version of
mathmodule.c.
>  > Fred, would you please get the truth into the Library Reference
Manual's
>  > math module docs too?
>
>   --sigh--
>   So are the Solaris and Linux man pages wrong, or just confusing?

    I found some Linux docs on this that said

"If the  argument  x is not zero, the normalized fraction
is x times a power of two, and  is  always  in  the
range  1/2 (inclusive) to 1 (exclusive)."

     This is not just wrong or confusing, this is _impossible_
if x < 0; a negative number simply is not a power of two times
a number in the range [1/2, 1).

DU

> Ok, I'll have something like your docstring checked in shortly.
>
>   -Fred
>
> --
> Fred L. Drake, Jr.  <fdrake at beopen.com>
> BeOpen PythonLabs Team Member
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From jhylton at my-deja.com  Mon Jul  3 19:45:16 2000
From: jhylton at my-deja.com (jhylton at my-deja.com)
Date: Mon, 03 Jul 2000 17:45:16 GMT
Subject: Random number generation, simple question.
References: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se>
Message-ID: <8jqjf8$s1t$1@nnrp1.deja.com>

In article <Pine.GSO.4.21.0007031546410.25965-100000 at luna.dsv.su.se>,
  Kalle Svensson <kalle at gnupung.net> wrote:

> I have a few questions about the (pseudo) random number generator
supplied
> in the standard library. I need 80 random bits (for a CipherSaber IV)
but
> I hear most RNG's only supply 32 bits or less.

The random number generator in the standard library is not at all
suitable for generating a cipher key.  In general, cryptography has a
crucial requirement for random numbers that most other applications
don't: It shouldn't be possible to guess or recover the state of the
random number generator.  In the case of the random module, if the
attacker can guess or recover your inital seed, she can generate the
entire sequence of random numbers that you use for your cipher. Bad
stuff!

I recommend looking at the Yarrow design by Kelsey, Schneier, and
Ferguson:  http://www.counterpane.com/yarrow.html

A Python implementation of Yarrow-160 is included in the Pisces
distribution:
http://www.cnri.reston.va.us/software/pisces/manual/module-pisces.yarrow.html

Jeremy


Sent via Deja.com http://www.deja.com/
Before you buy.


From tim_one at email.msn.com  Mon Jul  3 20:23:08 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 3 Jul 2000 14:23:08 -0400
Subject: Bug in math.frexp?
In-Reply-To: <8jqj17$rl0$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGENKGHAA.tim_one@email.msn.com>

Guys, this is not open to debate.  frexp is a std ANSI C function, Python
calls the platform frexp directly, and the new docstring I checked in is an
exact paraphrase of the std in Pythonic language.  I don't care what
manpages say.  I *would* care if somebody can show an actual example of a
Python implementation where the examples posted do not work exactly as
shown:  that would mean the platform is buggy -- but the new docs are still
correct <wink>.

arrogant-but-with-cause-in-this-case-ly y'rs  - tim





From m.faassen at vet.uu.nl  Mon Jul  3 20:18:22 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 3 Jul 2000 18:18:22 GMT
Subject: Class design: accessing "private" members
References: <962352843.1667731268@news.libertysurf.fr> <8jhuvn01vs0@news2.newsguy.com> <4.3.1.2.20000630145109.0201d9e0@phear.dementian.com> <8jplru0lfe@news2.newsguy.com> <962626013.769258863@news.libertysurf.fr>
Message-ID: <8jqlde$n0e$1@newshost.accu.uu.nl>

Jerome Quelin <jerome.quelin at insalien.org> wrote:
> "Alex Martelli" <alex at magenta.com> wrote:
>>These are excellent arguments for always using accessors from OUTSIDE the
>>class.  Jerome, however, was talking about using them from INSIDE the class
>>as well, which is a different issue.
> Sure, I'm always using accessors from outside.
> But from inside the class, I don't know which is the best. I feel like the KISS
> rule should apply, but what do you OO gurus think of it ? How far encapsulation
> should go ?

I don't recall exactly, but the OO literature on this issue referes to the
so-called 'Law of Demeter'. It comes in various weak and strong varieties
which differ on this issue. So you may want to do a web-search on that.

My personal opinion is that in Python you should not use accessors
internally until it turns out to be necessary (which may never be). Python
code is easy enough to change, after all. I don't even mind the occasional
public use of attributes. Helped by unit tests this kind of thing can
be changed to accessors fairly easily if necessary. Or I can of course
go the __getattr__/__setattr__ route.

I think the Pythonic way to think about this goes a bit like this:

  * sure, encapsulation is very important in good designs.

  * but, rapid prototyping is very important in coming up with good
    designs, as rapid prototyping allows you to change designs which
    turned out to be not good enough into better ones pretty easily.

  * good testing is important in any case.

  * if you have good testing and rapid prototyping, it's better to wait
    with various incantations (type checking, public/private, accessors)
    until they appear necessary, instead of doing them up-front and wasting
    time. It'll be likely you're wasting time as your up-front design will
    probably not be optimal yet in any case. Your incantations
    may turn out to be unnecessary in the first place!

In my opinion the best solution to attribute access would be to add 
optional getter/setter methods to Python, much like Delphi's properties,
or the facilities Sather and I hear C# provide.

Such a facility can provide for a very clean path from direct attribute
access to use of getters/setters.

Regards,

Martijn 
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From kmp at atrium.fsnet.co.uk  Mon Jul  3 21:44:41 2000
From: kmp at atrium.fsnet.co.uk (Kenneth Payne)
Date: Mon, 03 Jul 2000 19:44:41 GMT
Subject: Problem Installing PyQt-012
References: <20000702.20095100@linux.local> <slrn8lvbm1.o1d.newsbunny@noether.freeserve.co.uk>
Message-ID: <20000703.19444100@linux.local>

Hi Martin -

Well, I didn't run ldconfig, but I had set LD_LIBRARY_PATH. I didn't 
specify qtdir to the sip configure, but it seemed to find it (it was 
the X system it couldn't find).

Anyway, I've tried both now - and I still have the same error.

My commands were :

in the sip directory -

./configure --with-qt-dir=/usr/lib/qt --x-includes=/usr/X11R6/include 
--x-libraries=/usr/X11R6/lib
make
make install

in the PyQT directory -


ldconfig
./configure --with-qt-dir=/usr/lib/qt --x-includes=/usr/X11R6/include 
--x-libraries=/usr/X11R6/lib
make
make install

This looks suspicious in the sip configure output:

.
.
checking for working aclocal... missing
checking for working autoconf... missing
checking for working automake... missing
checking for working autoheader... missing
checking for working makeinfo... found
.
.
.
checking for x... no
.
.

Should I be looking to install aclocal, autoconf and automake? Would 
it be easier to install from rpms?


By the way, I was on the noether site recently and was suprised (and 
pleased) to see my name on the Compuserve kppp script. Thanks. Of 
course, I'm no longer with compuserve (is anyone?). I've joined you on 
Freeserve.

Best wishes

-- Ken

>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<

On 02/07/00, 21:06:09, newsbunny at noether.freeserve.co.uk (Martin P 
Holland) wrote regarding Re: Problem Installing PyQt-012:


> On Sun, 02 Jul 2000 20:09:51 GMT,
> Kenneth Payne <kmp at atrium.fsnet.co.uk> wrote:
> >I'm trying to install the Python bindings to Qt/KDE. I've downloaded
> >the latest sources for Sip. PyQT and pyKDE. I've compiled and
> >installed Sip and compiled PyQT. All seemed to go Ok, though I had to
> >fiddle a bit to get them to compile on Suse 6.3. However, on the 
"make
> >install" for PyQt, I get an error when the install routine hits a
> >Python script:
> >
> >Traceback (innermost last):
> >  File "<string>", line 1, in ?
> >  File "/usr/lib/python1.5/qt.py", line 16, in ?
> >    import libqtc
> >ImportError: /usr/lib/python1.5/lib-dynload/libqtcmodule.so: 
undefined
> >symbol: __ti8sipProxy
> >make[3]: *** [install-data-hook] Error 1
> >make[2]: *** [install-data-am] Error 2
> >make[1]: *** [install-am] Error 2
> >make: *** [install-recursive] Error 1

> Did you run ldconfig after installing sip?
> Did you specify the qtdir to ./configure when compiling sip?

> atb
> Martin


> --
> http://www.noether.freeserve.co.uk
> http://www.kppp-archive.freeserve.co.uk





From richard_chamberlain at ntlworld.com  Mon Jul  3 20:52:52 2000
From: richard_chamberlain at ntlworld.com (richard_chamberlain)
Date: Mon, 3 Jul 2000 19:52:52 +0100
Subject: WinNT User Name
References: <Ae485.94210$5k2.194235@dfw-read.news.verio.net>
Message-ID: <bh585.6120$4C4.108475@news2-win.server.ntlworld.com>

Hi Michael,

You'll need the win32 extensions - http://starship.python.net/crew/mhammond/

import win32api
login=win32api.GetUserName()

Richard


import win32api
michael montagne <montagne at boora.com> wrote in message
news:Ae485.94210$5k2.194235 at dfw-read.news.verio.net...
> First off, I'm a newbie at Python.
> I'm trying to return the user who is logged on at a PC.  What I really
need
> is the login name.  How do I do that?  Seems like it otta be quite simple
> but I'm having trouble.
>
> -mjm
>
>




From paul at prescod.net  Mon Jul  3 20:57:14 2000
From: paul at prescod.net (Paul Prescod)
Date: Mon, 03 Jul 2000 13:57:14 -0500
Subject: OO
References: <8jb1ec$9mh2@news.qualitynet.net> <autilso0l591oo924n3mogcva7r2892p5b@4ax.com> <8jpvhi$jfs$1@newshost.accu.uu.nl> <20000703171722.A7614@bork>
Message-ID: <3960E20A.4BA7AF49@prescod.net>

Egbert Bouwman wrote:
> 
> ...
> 
> Actually a great eye opener was the remark in this list a few days ago,
> when someone said that you use classes when you have to combine
> program and data, and functions when you have only program code.

Let me be more explicit about this. Classes are important when you want
behavior to travel *with* the data. Procedural and functional styles are
used where you want to impose behavior on "dumb" data. Real programs
tend to use both styles. Often a procedural model is used to get data
into the program where you combine the data with behavior in objects.

When the behavior travels with the data (as methods), you can fiddle
with the data without thinking much about its underlying implementation
(through the methods). Therefore that underlying implementation can be
really, really complex or really really simple and you don't care. It
can also change between versions of the module and you still don't care.
That's the encapsulation part of OO. You can actually do this pretty
easily just with just functions (not methods), using a style of
programming like this:

obj=CreateNewObject()
behavior1( obj, param1, param2, param3, ...)
behavior2( obj, param1, param2, param3, ...)
behavior2( obj, param1, param2, param3, ...)

There are lots of APIs written in this style in non-OO languages. This
is encapsulation (whether done in an OO language or not)

The "polymorphism" part of OO comes from the fact that once you have an
object, you can work on its methods without knowing exactly what kind of
object you have. The Python library has lots of good examples of this
but here's a beautiful example:

class myOutputClass1:
    def write( self, data ):
        ...write it to a GUI window....

class myOutputClass2:
    def write( self, data ):
        ...write it to a log ....

class myOutputClass3:
    def write( self, data ):
        ...send it via email to Bill Gates....

Now I can say 

sys.stdout=myOutputClass1() 

and hundreds of pieces of code that were supposed to work on the stdout
"stream" will magically work on myOutputClass. When they say:
sys.stdout.write( "foo" ), it calls the appropriate write method.

The fact that it has the right methods is enough to make it "stand in"
for a stream. This is much harder in a non-OO language because you have
to "fake" the objects carrying around their behavior as methods. I think
that polymorphism is really the heart of OO in Python. It's so important
that some separate it out from the other OO stuff and just call it
"genericity."

The third part of OO is most emphasized in most non-Python languages:
inheritance. Inheritance is a way of extending existing objects so that
you do not have to write all of that code over again. In other
languages, inheritance takes on some more mystical properties for the
type system, but you could be a decent Python programmer and only use
inheritance very, very rarely.

According to the powers that be, these are the three main features of
OO: encapsulation, polymorphism and inheritance.

-- 
 Paul Prescod - Not encumbered by corporate consensus
The calculus and the rich body of mathematical analysis to which it 
gave rise made modern science possible, but it was the algorithm that 
made the modern world possible.
	- The Advent of the Algorithm (pending), by David Berlinski



From wlfraed at ix.netcom.com  Mon Jul  3 20:57:58 2000
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Mon, 03 Jul 2000 11:57:58 -0700
Subject: Books
References: <8jb1ec$9mh2@news.qualitynet.net> <autilso0l591oo924n3mogcva7r2892p5b@4ax.com> <8jpvhi$jfs$1@newshost.accu.uu.nl>
Message-ID: <1ni1msc4fer3u6sm9blvjihql0inv9a1ba@4ax.com>

On 3 Jul 2000 12:05:06 GMT, m.faassen at vet.uu.nl (Martijn Faassen)
declaimed the following in comp.lang.python:

> Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote:
> 
> > 	Beginners in programming;
> 
> I saw others recommend Learning Python; I wouldn't do so for beginners
> in programming. The Quick Python book or Python in 24 Hours seem to be
> better suited for this.
>
	I'm so far out of that stage I no longer know what I'd recommend
-- other than /not/ C, C++, or some variant of BASIC (especially the one
M$ is pushing -- seen the recent "learn VB6" books... apparently the
"learning edition" of VB6 is too expensive, the books include a "working
model" edition, which appears to be strictly an interpreter within the
development environment with to capability for stand-alone programs).

> 
> > 	beginners in OO...
> 
> I'm not sure how good Learning Python is at teaching OO concepts to people
> who don't know about it yet. *Python* would be good though. Any opinions?
>
	I'm afraid I'm somewhat biased (and not too skilled in OOP
either), but my background leads me to believe one should first learn OO
concepts divorced from any particular language -- OO analysis and design
-- and then translate the design into the particulars of an
implementation language.

--
 > ============================================================== <
 >   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 boud at rempt.xs4all.nl  Mon Jul  3 21:00:46 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 3 Jul 2000 19:00:46 GMT
Subject: Problem Installing PyQt-012
References: <20000702.20095100@linux.local> <slrn8lvbm1.o1d.newsbunny@noether.freeserve.co.uk> <20000703.19444100@linux.local>
Message-ID: <8jqnsu$m79$1@news1.xs4all.nl>

Kenneth Payne <kmp at atrium.fsnet.co.uk> wrote:

> This looks suspicious in the sip configure output:

> .
> .
> checking for working aclocal... missing
> checking for working autoconf... missing
> checking for working automake... missing
> checking for working autoheader... missing
> checking for working makeinfo... found
> .
> .
> .
> checking for x... no
> .
> .

> Should I be looking to install aclocal, autoconf and automake? Would 
> it be easier to install from rpms?

Well, I runs Suse 6.3, too (but compiled both KDE and Qt myself),
and I have no problems compiling PyQt 0.12 either with Qt 1.44 or Qt
2.1.1, with Python 1.52 or Python 1.6alpha2, and when I compare my
config.log with yours, I notice that I _do_ have aclocal, autoconf,
automake and autoheader - I'd try installing them, they should be in
your distribution. Otherwise, you're the second person I've seen with
those errors - maybe it's something for Phil to look at.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From fplancke at my-deja.com  Mon Jul  3 20:59:02 2000
From: fplancke at my-deja.com (Frédéric van der Plancke)
Date: Mon, 03 Jul 2000 18:59:02 GMT
Subject: Class design: accessing "private" members
References: <962352843.1667731268@news.libertysurf.fr> <8jhuvn01vs0@news2.newsguy.com> <962367319.379353174@news.libertysurf.fr> <395D79FE.4E4FDBCE@colconsulting.com> <8jplbd0l5c@news2.newsguy.com>
Message-ID: <8jqnpe$vao$1@nnrp1.deja.com>

In article <8jplbd0l5c at news2.newsguy.com>,
  "Alex Martelli" <alex at magenta.com> wrote:

> Such "prohibition" is not hard at all:
>
> class dontderive:
>     def __init__(self):
>         if self.__class__ != dontderive:
>             raise "told you not to!"
>
> class foo(dontderive):
>     pass
>

#but...

class bypassing_prohibition(dontderive):
   def __init__(self):
       pass

eh = bypassing_prohibition()

#this works because __init__ never automatically calls the __init__
#method of base classes...

--
Frederic van der Plancke
e-mail address: reverse and add dot-at-dot to:
      <com writeme vdplancke frederic>
(I don't read <fplancke at my-deja.com> regularly.)


Sent via Deja.com http://www.deja.com/
Before you buy.


From db3l at fitlinxx.com  Mon Jul  3 21:20:27 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 03 Jul 2000 15:20:27 -0400
Subject: Communicating between two computers
References: <395d48b3$0$21842@wodc7nh0.news.uu.net> <uem5e38if.fsf@ctwd0143.fitlinxx.com> <395fbcd8$0$5627@wodc7nh0.news.uu.net>
Message-ID: <ubt0fyq9w.fsf@ctwd0143.fitlinxx.com>

"donotspam-jen at personic.com" <dnotspam-jen at personic.com> writes:

> I checked out the link, but from what I read, it sounds like this is a
> client-side implementation only.  Is there more info on this somewhere?
> Maybe some examples of how to use this?

It can handle both client and server.  The package has both the
xmprpclib.py module (client) as well as a sample xmlrpcserver.py
module (all 50 lines of it, excluding the preface comments).

Just "borrow" the server example module, add your methods into the
RequestHandler class and you're off.  On the client side, it's no more
than a single Server() instantiation, e.g.:

	server = xmlrpclib.Server("http://hostname")

and then you can call any method you want directly, e.g.:

	server.myMethod(myParameters)

any exceptions on the server side are trapped and and exception type
and value are returned as a "fault" return code.

BTW, this isn't much different from the other suggestions to subclass
the HTTP server object (in fact that's what RequestHandler in the
server example does), but the automatic translation back and forth
from any Python objects into XML for transmission/receipt saves a lot
of time in actually formatting and processing the request over the
stream.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From jvickroy at sec.noaa.gov  Mon Jul  3 21:00:59 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Mon, 03 Jul 2000 13:00:59 -0600
Subject: WinNT User Name
References: <Ae485.94210$5k2.194235@dfw-read.news.verio.net>
Message-ID: <3960E2EB.A5B1B256@sec.noaa.gov>

On Windows NT:

import os
os.environ['USERNAME']

... provides the information.

I do not believe that either Win 95 or 98 has the 'USERNAME' attribute set.

michael montagne wrote:

> First off, I'm a newbie at Python.
> I'm trying to return the user who is logged on at a PC.  What I really need
> is the login name.  How do I do that?  Seems like it otta be quite simple
> but I'm having trouble.
>
> -mjm



From db3l at fitlinxx.com  Mon Jul  3 21:37:35 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 03 Jul 2000 15:37:35 -0400
Subject: Traping output from os.system
References: <8jn34q$kpn$1@news.lth.se> <hCI75.5717$Tb7.31730@news-server.bigpond.net.au>
Message-ID: <u7lb3yphc.fsf@ctwd0143.fitlinxx.com>

"Mark Hammond" <MarkH at ActiveState.com> writes:

> You want "popen".  For non-windows platforms, "os.popen" is where you can
> find it.  For Win32 systems, you need the win32 extensions, and
> "win32pipe.popen".

Any chance that a future update to the extensions will be able to
return the result code of the child process on a close(), as
os.popen() does?

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From thomas at xs4all.nl  Mon Jul  3 21:52:56 2000
From: thomas at xs4all.nl (Thomas Wouters)
Date: 3 Jul 2000 19:52:56 GMT
Subject: Bug in math.frexp?
References: <LNBBLJKPBEHFEDALKOLCGENKGHAA.tim_one@email.msn.com>
Message-ID: <slrn8m1ro2.nti.thomas@centurion.xs4all.nl>

On Mon, 3 Jul 2000 14:23:08 -0400, Tim Peters <tim_one at email.msn.com> wrote:
>Guys, this is not open to debate.  frexp is a std ANSI C function, Python
>calls the platform frexp directly, and the new docstring I checked in is an
>exact paraphrase of the std in Pythonic language.  I don't care what
>manpages say.  I *would* care if somebody can show an actual example of a
>Python implementation where the examples posted do not work exactly as
>shown:  that would mean the platform is buggy -- but the new docs are still
>correct <wink>.

Yes, yes, you're right. The posting that started this unnecessary discussion
was both wrong and mine (I'm sorry, it was late ;P) but in my defense: I had
intended to *mail* it to Fred, but my slrn (again) went on and had a life of
its own.

What I intended to say was 'the manpages I saw concur with Tim's
docstringchange', but it 1) came out wrong, and 2) was wrong -- the manpages
are confusingly phrased, just as Fred claimed. I had the best of intentions
though ;)

Apologizing-ly y'rs,


From xiarcexml at yahoo.com  Mon Jul  3 22:06:27 2000
From: xiarcexml at yahoo.com (Dave Williams)
Date: Mon, 3 Jul 2000 13:06:27 -0700 (PDT)
Subject: OO
Message-ID: <20000703200627.5521.qmail@web5301.mail.yahoo.com>


From fplancke at my-deja.com  Mon Jul  3 21:59:45 2000
From: fplancke at my-deja.com (Frédéric van der Plancke)
Date: Mon, 03 Jul 2000 19:59:45 GMT
Subject: compilation
References: <Pine.LNX.4.21.0006270751280.2223-100000@fep132.fep.ru> <8jaht1$rui$5@bmerhc5e.ca.nortel.com> <8jdi5s$om4$1@nnrp1.deja.com> <8jfc7h$so4$1@bmerhc5e.ca.nortel.com>
Message-ID: <8jqrb7$216$1@nnrp1.deja.com>

In article <8jfc7h$so4$1 at bmerhc5e.ca.nortel.com>,
  msoulier at nortelnetworks.com wrote:
> In article <8jdi5s$om4$1 at nnrp1.deja.com>, Fr?d?ric van der Plancke
wrote:
> >
> >When the interpreter finds a xxx.pyc file on your Python path when
> >you type 'import xxx', it loads that directly instead of recompiling
> >xxx.py  (provided xxx.py is older, and other soundness conditions)
> >
> >Is that what you want ?
>
> 	Umm, sort of. I guess the top-level script still has to be a
.py file.

I don't think it is needed. Python can check if a .pyc already exists
for your top-level script. But I'm not 100% sure...

--
Frederic van der Plancke
e-mail address: reverse and add dot-at-dot to:
      <com writeme vdplancke frederic>
(I don't read <fplancke at my-deja.com> regularly.)


Sent via Deja.com http://www.deja.com/
Before you buy.


From montagne at boora.com  Mon Jul  3 22:32:48 2000
From: montagne at boora.com (michael montagne)
Date: Mon, 3 Jul 2000 13:32:48 -0700
Subject: WinNT User Name
References: <Ae485.94210$5k2.194235@dfw-read.news.verio.net> <3960E2EB.A5B1B256@sec.noaa.gov>
Message-ID: <RK685.94245$5k2.195418@dfw-read.news.verio.net>

thanks to you both..I knew it was easy.
-mjm





From ekw1 at my-deja.com  Mon Jul  3 22:32:13 2000
From: ekw1 at my-deja.com (ekw1 at my-deja.com)
Date: Mon, 03 Jul 2000 20:32:13 GMT
Subject: mysql module for python under windows
Message-ID: <8jqt8d$3do$1@nnrp1.deja.com>

Is there a Python MySQL module for use under Windows?  All the ones I
come across on python.org are for Unix/Linux.  Thanks for any info.

ekw


Sent via Deja.com http://www.deja.com/
Before you buy.


From kmp at atrium.fsnet.co.uk  Tue Jul  4 00:08:17 2000
From: kmp at atrium.fsnet.co.uk (Kenneth Payne)
Date: Mon, 03 Jul 2000 22:08:17 GMT
Subject: Problem Installing PyQt-012
References: <20000702.20095100@linux.local> <slrn8lvbm1.o1d.newsbunny@noether.freeserve.co.uk> <20000703.19444100@linux.local> <8jqnsu$m79$1@news1.xs4all.nl>
Message-ID: <20000703.22081700@linux.local>

Boudewijn -

Thanks for your interest in my sip problem. 

I've installed autoconf and automake. This gets rid of all errors 
except the message about X (if that is an error) but I still get the  
same probelm on installing pyQT.

My configure output for sip now looks like this :

loading cache ./config.cache
checking for a BSD compatible install... (cached) /usr/bin/ginstall -c
checking whether build environment is sane... yes
checking whether make sets ${MAKE}... (cached) yes
checking for working aclocal... found
checking for working autoconf... found
checking for working automake... found
checking for working autoheader... found
checking for working makeinfo... found
checking for a BSD compatible install... /usr/bin/ginstall -c
checking for gcc... (cached) gcc
checking whether the C compiler (gcc  ) works... yes
checking whether the C compiler (gcc  ) is a cross-compiler... no
checking whether we are using GNU C... (cached) yes
checking whether gcc accepts -g... (cached) yes
checking for c++... (cached) c++
checking whether the C++ compiler (c++  ) works... yes
checking whether the C++ compiler (c++  ) is a cross-compiler... no
checking whether we are using GNU C++... (cached) yes
checking whether c++ accepts -g... (cached) yes
checking for bison... (cached) bison -y
checking how to run the C preprocessor... (cached) gcc -E
checking for flex... (cached) flex
checking for flex... (cached) flex
checking for yywrap in -lfl... (cached) yes
checking lex output file root... (cached) lex.yy
checking whether yytext is a pointer... (cached) yes
checking host system type... i686-pc-linux-gnu
checking build system type... i686-pc-linux-gnu
checking for ranlib... (cached) ranlib
checking for ld used by GCC... (cached) /usr/i486-linux/bin/ld
checking if the linker (/usr/i486-linux/bin/ld) is GNU ld... (cached) 
yes
checking for BSD-compatible nm... (cached) /usr/bin/nm -B
checking whether ln -s works... (cached) yes
loading cache ./config.cache within ltconfig
checking whether ln -s works... (cached) yes
checking for ranlib... (cached) ranlib
checking for strip... (cached) strip
checking for objdir... .libs
checking for object suffix... (cached) o
checking for executable suffix... (cached) no
checking for gcc option to produce PIC... (cached)  -fPIC
checking if gcc PIC flag  -fPIC works... (cached) yes
checking if gcc static flag -static works... (cached) yes
checking if gcc supports -c -o file.o... yes
checking if gcc supports -c -o file.lo... yes
checking if gcc supports -fno-rtti -fno-exceptions ... yes
checking if the linker (/usr/i486-linux/bin/ld) is GNU ld... yes
checking whether the linker (/usr/i486-linux/bin/ld) supports shared 
libraries... yes
checking how to hardcode library paths into programs... immediate
checking whether stripping libraries is possible... yes
checking for /usr/i486-linux/bin/ld option to reload object files... 
-r
checking dynamic linker characteristics... Linux ld.so
checking command to parse /usr/bin/nm -B output... ok
checking if libtool supports shared libraries... yes
checking whether to build shared libraries... yes
checking whether to build static libraries... yes
creating libtool
loading cache ./config.cache
checking for the Qt meta object compiler... /usr/lib/qt/bin/moc
checking for the Qt include files... /usr/lib/qt/include
checking for the Qt library files... /usr/lib/qt/lib
checking for X... no
checking for -fno-exceptions... yes
checking for Python 1.5 include files... /usr/include/python1.5
creating ./config.status
creating Makefile
creating sip/Makefile
creating siplib/Makefile
creating sip.spec
creating config.h
config.h is unchanged


If all this seems correct, perhaps the error is in the configure for 
the PyQT sources. Should I concentrate on that?

By the way, I tried installing the rpm for sip, but it requires a 
later version of libstdc than mine. I get the error :
 
 failed dependencies:
 libstdc++-libc6.1-2.so.3 is needed by sip-0.12-1

It's at least comforting to know that it is possible to run these 
python / KDE bindings on SUSE 6.3

Thanks again for your interest.

Best wishes

-- Ken  

>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<

On 03/07/00, 19:00:46, boud at rempt.xs4all.nl (Boudewijn Rempt) wrote 
regarding Re: Problem Installing PyQt-012:


> Kenneth Payne <kmp at atrium.fsnet.co.uk> wrote:

> > This looks suspicious in the sip configure output:

> > .
> > .
> > checking for working aclocal... missing
> > checking for working autoconf... missing
> > checking for working automake... missing
> > checking for working autoheader... missing
> > checking for working makeinfo... found
> > .
> > .
> > .
> > checking for x... no
> > .
> > .

> > Should I be looking to install aclocal, autoconf and automake? Would
> > it be easier to install from rpms?

> Well, I runs Suse 6.3, too (but compiled both KDE and Qt myself),
> and I have no problems compiling PyQt 0.12 either with Qt 1.44 or Qt
> 2.1.1, with Python 1.52 or Python 1.6alpha2, and when I compare my
> config.log with yours, I notice that I _do_ have aclocal, autoconf,
> automake and autoheader - I'd try installing them, they should be in
> your distribution. Otherwise, you're the second person I've seen with
> those errors - maybe it's something for Phil to look at.

> --

> Boudewijn Rempt  | http://www.valdyas.org





From dale at out-think.NOSPAMco.uk  Mon Jul  3 22:54:14 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 3 Jul 2000 21:54:14 +0100
Subject: Definative Documentation on Objects
Message-ID: <8jqui2$9na$1@supernews.com>

Where do I look for the definitive statement on object programming in
Python?

The "Python Reference Manual", as included in the distribution, is
particularly hard work.

I'm considering "Programming Python" by Mark Lutz but that's four years old.
"Learning Python" is more recent but I'm worried that it might be pitched a
bit low.

I'm an experienced programmer but I need to quickly understand Python object
programming.

Any pointers to online or paper reference material gratefully received.

--
Dale Strickland-Clark
Out-Think Ltd, UK
Business Technology Consultants





From cut_me_out at hotmail.com  Mon Jul  3 23:04:03 2000
From: cut_me_out at hotmail.com (Alex)
Date: 03 Jul 2000 17:04:03 -0400
Subject: Definative Documentation on Objects
References: <8jqui2$9na$1@supernews.com>
Message-ID: <etdu2e7j58c.fsf@w20-575-34.mit.edu>

Perhaps the tutorial?

http://www.python.org/doc/tut/node11.html

It all depends on what a definitive statement would be, I guess.

Alex.


From esko.lehtonen at pp.htv.fi  Mon Jul  3 23:11:34 2000
From: esko.lehtonen at pp.htv.fi (Esko Lehtonen)
Date: Tue, 04 Jul 2000 00:11:34 +0300
Subject: PyGtk again but with segfault
Message-ID: <39610186.36BAFE88@pp.htv.fi>

Sunday I asked about PyGTK's draw_array-method a stupid question:-) Now
I have another, hopefully a stupid (=and easily solvable) too.


from gtk import *
from Numeric import *
--
win = GtkWindow()
darea = GtkDrawingArea()
---
HEIGHT = 50
WIDTH = 50
BYTES = 3
rgbbuf = ones([HEIGHT, WIDTH, BYTES], UnsignedInt8)

gc = self.win.get_style().white_gc
drawable = self.darea.get_window()
dither = GDK.RGB_DITHER_NORMAL
x = 10
y = 10
rowstride = WIDTH * 3

draw_array(drawable, gc, x, y, dither, rgbbuf)


Unfortunataly I got now segmentation fault when the C function is
executed.
I tried to examine where the problem is. From gdb I got:

Program received signal SIGSEGV, Segmentation fault.
0x4039ec58 in gdk_draw_rgb_image () from /usr/lib/libgdk-1.2.so.0

All parameters to C function are correct. I checked it recompiling
_gtkmodules.so with many many printf()... (There might be better ways,
but I'm not used to debug C programs.)

Still gdk_draw_rgb_image() function works well when called from C
program. 

My system is Debian 2.2 potato with Python 1.5 and Gtk+ 1.2.

Thank you again...

--------------------------------------------------------------------
 Esko Lehtonen                   WYSINWYG -              
                                 What You See Is Never What You Get


From scott at lund.com  Mon Jul  3 22:57:04 2000
From: scott at lund.com (Scott Chapman)
Date: Mon, 03 Jul 2000 13:57:04 -0700
Subject: Book for sale - Programming Python w/CD $30.00 + Shipping
Message-ID: <tgv1ms0ejkadv37gbo0viminjpcpfuda0k@4ax.com>

Email scott at lund.com if interested. Payment via PayPal preferred.



From jhylton at my-deja.com  Mon Jul  3 23:38:28 2000
From: jhylton at my-deja.com (jhylton at my-deja.com)
Date: Mon, 03 Jul 2000 21:38:28 GMT
Subject: Definative Documentation on Objects
References: <8jqui2$9na$1@supernews.com>
Message-ID: <8jr14j$69v$1@nnrp1.deja.com>

In article <8jqui2$9na$1 at supernews.com>,
  "Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> wrote:
> Where do I look for the definitive statement on object programming in
> Python?
>
> I'm considering "Programming Python" by Mark Lutz but that's four
> years old. "Learning Python" is more recent but I'm worried that it\
> might be pitched a bit low.

Learning Python has an excellent chapter on classes and object-oriented
programming.  I think it is your best bet.

--Jeremy Hylton <http://www.pyton.org/~jeremy/>


Sent via Deja.com http://www.deja.com/
Before you buy.


From montagne at boora.com  Mon Jul  3 23:51:04 2000
From: montagne at boora.com (michael montagne)
Date: Mon, 3 Jul 2000 14:51:04 -0700
Subject: Python Object Browser
Message-ID: <dU785.94271$5k2.195629@dfw-read.news.verio.net>

I'm trying to use makepy on some common objects on my system.  Word2000,
Excel2000, etc.
It doesn't seem to want ot work.  When I go to the objects that are having
trouble using the Python Object Browser, under the Type Library heading it
says
"The type info can n..="The type info can not be loaded".  Does this mean it
is unavailable?  How do i fix this?

-mjm




From dale at out-think.NOSPAMco.uk  Tue Jul  4 00:02:59 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 3 Jul 2000 23:02:59 +0100
Subject: Definative Documentation on Objects
References: <8jqui2$9na$1@supernews.com> <etdu2e7j58c.fsf@w20-575-34.mit.edu>
Message-ID: <8jr2iv$cn2$1@supernews.com>

Now why didn't I find that! I've been looking all over the place and it was
right under my nose.

Thanks for the reply!

--
Dale Strickland-Clark
Out-Think Ltd, UK
Business Technology Consultants

Alex <cut_me_out at hotmail.com> wrote in message
news:etdu2e7j58c.fsf at w20-575-34.mit.edu...
>
> Perhaps the tutorial?
>
> http://www.python.org/doc/tut/node11.html
>
> It all depends on what a definitive statement would be, I guess.
>
> Alex.




From iain at wombatzone.freeserve.co.uk  Tue Jul  4 00:50:44 2000
From: iain at wombatzone.freeserve.co.uk (Iain Calder)
Date: Mon, 03 Jul 2000 22:50:44 +0000
Subject: PyGtk again but with segfault
References: <39610186.36BAFE88@pp.htv.fi>
Message-ID: <396118C4.116790FA@wombatzone.freeserve.co.uk>

Esko Lehtonen wrote:
> 
> Sunday I asked about PyGTK's draw_array-method a stupid question:-) Now
> I have another, hopefully a stupid (=and easily solvable) too.
> 
> from gtk import *
> from Numeric import *
> --
> win = GtkWindow()
> darea = GtkDrawingArea()
> ---
> HEIGHT = 50
> WIDTH = 50
> BYTES = 3
> rgbbuf = ones([HEIGHT, WIDTH, BYTES], UnsignedInt8)
> 
> gc = self.win.get_style().white_gc
> drawable = self.darea.get_window()
> dither = GDK.RGB_DITHER_NORMAL
> x = 10
> y = 10
> rowstride = WIDTH * 3
> 
> draw_array(drawable, gc, x, y, dither, rgbbuf)
> 
> Unfortunataly I got now segmentation fault when the C function is
> executed.

At startup, before the first call to draw_array, you need to call
push_rgb_visual().

HTH

Iain Calder


From MarkH at ActiveState.com  Tue Jul  4 01:22:38 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Mon, 03 Jul 2000 23:22:38 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <8jabvo014tv@news2.newsguy.com> <uhfa7qn9c.fsf@cs.uu.nl> <8jqbru017gf@news2.newsguy.com>
Message-ID: <2f985.6708$Tb7.40527@news-server.bigpond.net.au>

"Alex Martelli" <alex at magenta.com> wrote in message
news:8jqbru017gf at news2.newsguy.com...

> I believe the only relevant issue for the Python community
> is: C#, like every other Microsoft offering in recent times,
> will be *STRONGLY* oriented to COM.  Every C# object will be
> a COM object (just like every Java object was one when run
> under Microsoft's Java VM), etc, etc.

Its a little more complicated than that!

> So, C#'s possible (unlikely...) emergence as a major language
> is just the N-plus-1-th motivation for Windows versions of
> Python to support COM *very, VERY* well.

And that!

> Yeah, I know, I should put my coding efforts where my mouth
> is -- and, having expended substantial efforts doing just
> this sort of thing for a proprietary scripting language

Please do.  Typelib work may be kicking off again (if I ever clear my
inbox that is - sorry John - I havent forgotten you :-)  A framework for
calling arbitary vtable interfaces is also in place.  So what you will
find is that it needs less "framework" code, and more a motivated and
skilled person to pull a few of these "late" pieces together.  My
motivation WRT win32com now is more towards keeping the status-quo - and
helping _others_ change it :-)

> Re .NET -- I dunno, and I wish I did!

All will be revlealed after the PDC; less than 2 weeks.  I can safely say
(as this is a Python language newgroup and C# is apparently a "competing"
language), the Python community will probably be more interested in .NET
than in C#.  _Quite_ a bit more :-)


> I'll do y'all a favour and summarize them into basically
> two stances:
>     "It's Microsoft and thus evil, vapourware, immoral,
>         and/or fattening"
>     "It's Microsoft and thus good, splendid, progressive,
>         and/or sexy"

Well put :-)

> I do wonder what other items that have technical reality
> are in .NET besides C#, XML, and SOAP.  If any, that is.

There are - and a number of them.

Mark.





From MarkH at ActiveState.com  Tue Jul  4 01:24:31 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Mon, 03 Jul 2000 23:24:31 GMT
Subject: From UNIX to Win NT
References: <39606DAC.49A502DB@muc.das-werk.de> <8jq7c2$2f8$1@slb6.atl.mindspring.net> <3960B33F.98B23C40@sec.noaa.gov>
Message-ID: <Pg985.6710$Tb7.40646@news-server.bigpond.net.au>

"j vickroy" <jvickroy at sec.noaa.gov> wrote in message
news:3960B33F.98B23C40 at sec.noaa.gov...
> Mark isn't here yet but his book (Python Programming on Win 32) is and I
quote
> (from page 35):

And that quote shows why I was staying out of it :-)

Mark.





From robin at jessikat.demon.co.uk  Tue Jul  4 01:23:07 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Tue, 4 Jul 2000 00:23:07 +0100
Subject: tkinter hanging
References: <WRe0zXARD1X5EwXY@jessikat.demon.co.uk>
 <slrn8m0hm3.m2b.thomas@centurion.xs4all.nl>
 <yget$DAtiEY5EwYZ@jessikat.demon.co.uk>
Message-ID: <+fF0kLAbBSY5Ewaz@jessikat.demon.co.uk>

In article <yget$DAtiEY5EwYZ at jessikat.demon.co.uk>, Robin Becker <robin at jessikat.co.uk> writes
>In article <slrn8m0hm3.m2b.thomas at centurion.xs4all.nl>, Thomas Wouters
><thomas at xs4all.nl> writes
>>On Sun, 2 Jul 2000 15:25:21 +0100, Robin Becker <robin at jessikat.demon.co.uk>
>>wrote:
>>>I'm having problems with zombie processes under win32 with a tkinter
>>>linked against Tk 8.3. I seem to be exiting the mainloop ok and have
>>>destroyed all subwindows and then the gui object itself, but
>>>occasionally I still generate a zombie. I have checked that my after's
>>>are cancelled and debug prints seem to show me exiting the script.
>>
...
Under Win95 + Python 1.6a2 this code hangs about 1 in 10 times when the gui is killed using the
window manager. The output indicates that the mainloop is terminated, afters are cancelled and
all the widgets are destroyed. Even so it seems to hang after printing 'sys.exit'. It is using
the common Tk idiom of delayed polling with after, but the exit code makes strenuous efforts to
remove any pending calls.

Can anyone see what I'm doing wrong?

####################
from Tkinter import *
import string, sys, os, regex, telnetlib
import traceback, time

class CommandEntry(Entry):
    def __init__(self, master, partk, **kw):
        apply(Entry.__init__, (self, master), kw)

class Gui:
    def __init__(self):
        self.tk = Tk()
        try:
            self._build()
            self._alive = 1
        except:
            traceback.print_exc()
            self.destroy()

    def _build(self):
        self.tk.geometry("800x600")
        self.tk.title("Client")
        self.tk.protocol("WM_DELETE_WINDOW", self.close)
        self.after_id = []
        
        tA = {'fg':'white','bg':'black','state':'disabled','height':20}
        tC = {'fg':'white','bg':'black','insertbackground':'yellow','insertwidth':2}
        if os.name != 'posix':
            # need tcl/tk 8.0 on windows
            fnt = ("Fixedsys", 12)
            tA['font'] = fnt
            tC['font'] = fnt

        self.txt = Text(self.tk,tA)
        self.entry = apply(CommandEntry,(self.tk, self), tC)

        # set up the scrollbar for the txt widget
        self.scrollVertical = Scrollbar(self.tk,orient=VERTICAL)
        self.txt.configure(yscrollcommand=self.scrollVertical.set)
        self.scrollVertical.config(command=self.txt.yview)
        self.scrollVertical.pack(side=RIGHT, anchor=E, fill=Y)

        self.entry.pack({'side': 'bottom', 'fill': 'both'})
        self.entry.focus_set()
        self.txt.pack({'side': 'bottom', 'fill': 'both', 'expand': 1})

    def mainloop(self):
        self.after_id.append(self.tk.after(100, self.iterate))
        self.tk.mainloop()
        print 'mainloop end'
        
    def destroy(self):
        self._alive = 0
        self.after_cancel()

        for w in ('scrollVertical','txt','entry','tk'):
            try:
                getattr(self,w).destroy()
                delattr(self,w)
                print '%s destroyed' % w
            except:
                traceback.print_exc()
        
    def iterate(self):
        if self.loop():
            self.after_id.append(self.tk.after(50, self.iterate))
        else:
            self.tk.quit()

    def after_cancel(self):
        while self.after_id!=[]:
            id = self.after_id[0]
            del self.after_id[0]
            try:
                self.tk.after_cancel(id)
            except:
                traceback.print_exc()

    def close(self,event=None):
        self._alive = 0

    def loop(self):
        if not self._alive: return 0
        try:
            pass
        except:
            traceback.print_exc()
            self._alive = 0

        return self._alive

if __name__ == '__main__':
    gui = Gui()
    gui.mainloop()
    print 'Exiting'
    gui.destroy()
    del gui
    while 1:
        time.sleep(1)
        print 'sys.exit'
        sys.exit()
-- 
Robin Becker


From swun at esec.com.au  Tue Jul  4 01:51:03 2000
From: swun at esec.com.au (Sam Wun)
Date: Tue, 04 Jul 2000 09:51:03 +1000
Subject: how to push data in to ssh in python?
References: <395FF2D9.1B46251D@esec.com.au> <slrn8m0hvm.m2b.thomas@centurion.xs4all.nl>
Message-ID: <396126E7.B10A450B@esec.com.au>

Oh yes, I actually print it to the stdout, and intended to use command "ssh -l
swun hostname < out.dump" to send out.dump to the "hostname" in "swun" account,
but I am not sure how to set up the "swun" account to receive this transaction.

Thanks
Sam.

Thomas Wouters wrote:

> On Mon, 03 Jul 2000 11:56:41 +1000, Sam Wun <swun at esec.com.au> wrote:
>
> >Thanks
>
> Your question is incomplete. What do you want to do ? Did you look at
> popen(), and is it not what you want ? Does ssh require the input to come
> from a pseudo-terminal ? Can't you use a normal shell pipeline and make your
> python script just 'print' what you want to output ? Are you encountering
> difficulties with any of these ideas, or not finding the right documentation?
>
> Oh, and it's not necessary to post both in HTML and in plaintext.. Just
> plaintext will do just fine ;)
>
> How-do-I-get-a-zilion-dollars-ly y'rs,
>         Thomas.

--
Sam Wun                            Firewalls / Security
OneGuard Senior Software Engineer  Electronic Commerce
eSec Limited                       Phone: +61 3 83715376
mailto:swun at eSec.com.au            C++/JAVA/UNIX/OOP/OOD


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/python-list/attachments/20000704/a2f39ca6/attachment.htm 

From MarkH at ActiveState.com  Tue Jul  4 01:30:34 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Mon, 03 Jul 2000 23:30:34 GMT
Subject: Traping output from os.system
References: <8jn34q$kpn$1@news.lth.se> <hCI75.5717$Tb7.31730@news-server.bigpond.net.au> <u7lb3yphc.fsf@ctwd0143.fitlinxx.com>
Message-ID: <um985.6715$Tb7.40775@news-server.bigpond.net.au>

> Any chance that a future update to the extensions will be able to
> return the result code of the child process on a close(), as
> os.popen() does?

Unlikely that Bill or I will look into this.  The problem is that the file
object doesnt know anything about the popen child process.

Personally, I would prefer to _remove_ win32pipe.popen's implementation,
and re-write it in pure-python using the raw win32pipe functions - just to
prove we can :-) And also so it would be easier to hack to do this sort of
stuff, or cloned for "custom" requirements, etc.

Mark.





From spamfranke at bigfoot.de  Tue Jul  4 02:15:01 2000
From: spamfranke at bigfoot.de (Stefan Franke)
Date: Tue, 04 Jul 2000 00:15:01 GMT
Subject: mysql module for python under windows
References: <8jqt8d$3do$1@nnrp1.deja.com>
Message-ID: <39612aaa.15439725@news.btx.dtag.de>

On Mon, 03 Jul 2000 20:32:13 GMT, ekw1 at my-deja.com wrote:

>Is there a Python MySQL module for use under Windows?  All the ones I
>come across on python.org are for Unix/Linux.  Thanks for any info.
>
>ekw
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.

Hmm, dunno where to download it separately, but check out

  ZMySQLDA-1.2.0.tar.gz

on the Zope site (www.zope.org). It contains MySQLdb-0.1.2 with
some Windows patches already applied. I run it sucessfully with
the lately GPLed MySQL-2.23.19 on Win98.

All you need to do is to edit Setup.in, adjust the paths to your 
local MySQL installation, and run compile.py. If you don't have
Visual Studio, mail me personally and I'll send you the files.

No-warranties-of-course-ly y'rs
Stefan



From spamfranke at bigfoot.de  Tue Jul  4 02:22:37 2000
From: spamfranke at bigfoot.de (Stefan Franke)
Date: Tue, 04 Jul 2000 00:22:37 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <8jabvo014tv@news2.newsguy.com> <uhfa7qn9c.fsf@cs.uu.nl> <8jqbru017gf@news2.newsguy.com> <2f985.6708$Tb7.40527@news-server.bigpond.net.au>
Message-ID: <39642dc0.16230059@news.btx.dtag.de>

On Mon, 03 Jul 2000 23:22:38 GMT, "Mark Hammond" <MarkH at ActiveState.com> wrote:

>A framework for calling arbitary vtable interfaces is also in place.
I would *really* appreciate such a thing.  Any pointers?

Stefan



From spamfranke at bigfoot.de  Tue Jul  4 02:23:05 2000
From: spamfranke at bigfoot.de (Stefan Franke)
Date: Tue, 04 Jul 2000 00:23:05 GMT
Subject: Socket problem
References: <slrn8m1p8s.2e9.thor@localhost.localdomain>
Message-ID: <39652e57.16380596@news.btx.dtag.de>

[Might be a repost..]

On 3 Jul 2000 17:08:24 GMT, thor at localhost.localdomain (Manuel Gutierrez Algaba) wrote:

>
>I'm doing a game for my University. It's divided into 3 parts
>( AI (clips), core(python), GUI(python)) communicated by sockets.

Sorry, no help for your question, but just for interest - how do you interface
with clips?

Stefan




From db3l at fitlinxx.com  Tue Jul  4 02:59:27 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 03 Jul 2000 20:59:27 -0400
Subject: Traping output from os.system
References: <8jn34q$kpn$1@news.lth.se> <hCI75.5717$Tb7.31730@news-server.bigpond.net.au> <u7lb3yphc.fsf@ctwd0143.fitlinxx.com> <um985.6715$Tb7.40775@news-server.bigpond.net.au>
Message-ID: <u4s66zp5c.fsf@ctwd0143.fitlinxx.com>

"Mark Hammond" <MarkH at ActiveState.com> writes:

> > Any chance that a future update to the extensions will be able to
> > return the result code of the child process on a close(), as
> > os.popen() does?
> 
> Unlikely that Bill or I will look into this.  The problem is that the file
> object doesnt know anything about the popen child process.

Hmm, yeah, but couldn't the win32popen module just pass a different
close routine (to PyFile_FromFile()) and then keep track of which file
handles were for which pipe?  True, it would have to manage that
mapping, but could probably just use an internal dictionary for it :-)
(Note I haven't actually tried this myself yet - just guessing from a
few glances at the source)

Ah, well - unfortunately exit codes are key for me, since the gate
decisions.  So I'll just keep valid consoles around for the moment :-)

> Personally, I would prefer to _remove_ win32pipe.popen's implementation,
> and re-write it in pure-python using the raw win32pipe functions - just to
> prove we can :-) And also so it would be easier to hack to do this sort of
> stuff, or cloned for "custom" requirements, etc.

Makes sense - I may try that at the point where I need to get rid of
thge consoles, although I'm definitely targeted at NT, so I get to
avoid all the Win95/98 ugliness I keep hearing about :-)

Doing it this way would probably make it easier to return an object
that looked like a file object (although it wouldn't really be the
actual type) but knew enough about how to access the child process
during the close() operation.

Thanks.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From neo_brave at my-deja.com  Tue Jul  4 03:45:38 2000
From: neo_brave at my-deja.com (Neo Brave)
Date: Tue, 04 Jul 2000 01:45:38 GMT
Subject: Win32 : Making script.py executable in DOS box
References: <8jmjfe$64o$1@nnrp1.deja.com> <WzI75.5714$Tb7.31718@news-server.bigpond.net.au> <8jp3pt$q7j$1@nnrp1.deja.com> <YC185.27369$W35.636919@news20.bellglobal.com>
Message-ID: <8jrfjv$geg$1@nnrp1.deja.com>

In article <YC185.27369$W35.636919 at news20.bellglobal.com>,
  "Olivier Dagenais" <olivierS.dagenaisP at canadaA.comM> wrote:
> How about a batch file called p.bat ?  (with optional status messages)
>
> -----start p.bat--------
> @echo off
> echo Starting script '%1.py' ....
> c:\progra~1\Python\python "%1.py" %2 %3 %4 %5 %6 %7 %8 %9
> echo Execution complete!
> ----end p.bat-----------
>
> Given you have a script called "spam.py":
>
> -----start spam.py------
> print "Spam spam spam spam"
> -----end spam.py--------
>
> Then, you would type:
>
> c:\work\python\stuff>p spam
> Starting script 'spam.py' ...
> Spam spam spam spam
> Execution complete!
>
> Is that what you wanted??

Thanks. Thats an improvement on what I have been doing till now but
its not quite as neat as the WinNT solution (which I've tried on
WinNT and it works :-)).

Maybe its time I bit the bullet and upgraded to NT.

Neo
--
Neo Brave


Sent via Deja.com http://www.deja.com/
Before you buy.


From rurban at sbox.tu-graz.ac.at  Tue Jul  4 05:52:50 2000
From: rurban at sbox.tu-graz.ac.at (Reini Urban)
Date: Tue, 04 Jul 2000 03:52:50 GMT
Subject: Random number generation, simple question.
References: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se> <8jqjf8$s1t$1@nnrp1.deja.com>
Message-ID: <39615d75.142572338@news>

jhylton at my-deja.com wrote:
>I recommend looking at the Yarrow design by Kelsey, Schneier, and
>Ferguson:  http://www.counterpane.com/yarrow.html

If it comes to critical or important data I would NEVER rely on any PRNG
("pseudo is pseudo and never truly random") instead of some
inpredictable hardware noise. There were some C libs around dealing with
some special PC timer chips which can be abused for exactly that.

such as the linux /dev/random pseudodevice or 
truerand.c "Physically random numbers (very nearly uniform)"
by D. P. Mitchell, AT&T. 
This is portable. But I know only the perl Math::TrueRandom module.
--
Reini Urban
http://xarch.tu-graz.ac.at/autocad/news/faq/autolisp.html


From jmcbray at carcosa.net  Tue Jul  4 05:44:53 2000
From: jmcbray at carcosa.net (Jason F. McBrayer)
Date: 03 Jul 2000 23:44:53 -0400
Subject: Random number generation, simple question.
References: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se> <8jqjf8$s1t$1@nnrp1.deja.com>
Message-ID: <m1hfa636fe.fsf@bertrand.carcosa.net>

>>>>> "j" == jhylton  <jhylton at my-deja.com> writes:

j> In article <Pine.GSO.4.21.0007031546410.25965-100000 at luna.dsv.su.se>,
j>   Kalle Svensson <kalle at gnupung.net> wrote:

>> I have a few questions about the (pseudo) random number generator
>> supplied in the standard library. I need 80 random bits (for a
>> CipherSaber IV) but I hear most RNG's only supply 32 bits or less.

j> The random number generator in the standard library is not at all
j> suitable for generating a cipher key.

He's not using it to generate a cipher key, just an initialization
vector that doesn't have to be too terribly random (its only purpose
is to let you reuse the same real key in different messages).  As I
understand it (and according to the Cypersabre pages), you don't need
anything better than the standard library's rng (though it's just as
easy, if not easier, to read 10 bytes from /dev/random).

-- 
+-----------------------------------------------------------+
| Jason F. McBrayer                    jmcbray at carcosa.net  |
| A flower falls, even though we love it; and a weed grows, |
| even though we do not love it.            -- Dogen        |


From MarkH at ActiveState.com  Tue Jul  4 08:05:22 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 04 Jul 2000 06:05:22 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <8jabvo014tv@news2.newsguy.com> <uhfa7qn9c.fsf@cs.uu.nl> <8jqbru017gf@news2.newsguy.com> <2f985.6708$Tb7.40527@news-server.bigpond.net.au> <39642dc0.16230059@news.btx.dtag.de>
Message-ID: <C8f85.6959$Tb7.42242@news-server.bigpond.net.au>

"Stefan Franke" <spamfranke at bigfoot.de> wrote in message
news:39642dc0.16230059 at news.btx.dtag.de...
> On Mon, 03 Jul 2000 23:22:38 GMT, "Mark Hammond" <MarkH at ActiveState.com>
wrote:

> >A framework for calling arbitary vtable interfaces is also in place.

> I would *really* appreciate such a thing.  Any pointers?

The extension is called "univgw" - for "Universal Gateway".  Greg Stein
started the work, and Bill Tutt has done the most recent round.

The sources, including the MSVC project file can be found under CVS.  The
win32com developers list archives are probably a good source, too:
http://mailman.pythonpros.com/mailman/listinfo/pycom-dev

Further discussion is welcome on that list!

Mark.





From thomas at xs4all.net  Tue Jul  4 08:34:59 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 4 Jul 2000 08:34:59 +0200
Subject: how to push data in to ssh in python?
In-Reply-To: <396126E7.B10A450B@esec.com.au>; from swun@esec.com.au on Tue, Jul 04, 2000 at 09:51:03AM +1000
References: <395FF2D9.1B46251D@esec.com.au> <slrn8m0hvm.m2b.thomas@centurion.xs4all.nl> <396126E7.B10A450B@esec.com.au>
Message-ID: <20000704083459.C283@xs4all.nl>

On Tue, Jul 04, 2000 at 09:51:03AM +1000, Sam Wun wrote:
> Thomas Wouters wrote:
> 
> > On Mon, 03 Jul 2000 11:56:41 +1000, Sam Wun <swun at esec.com.au> wrote:
> >
> > >Thanks
> >
> > Your question is incomplete. What do you want to do ? Did you look at
> > popen(), and is it not what you want ? Does ssh require the input to
> > come from a pseudo-terminal ? Can't you use a normal shell pipeline and
> > make your python script just 'print' what you want to output ? Are you
> > encountering difficulties with any of these ideas, or not finding the
> > right documentation?

> Oh yes, I actually print it to the stdout, and intended to use command
> "ssh -l swun hostname < out.dump" to send out.dump to the "hostname" in
> "swun" account, but I am not sure how to set up the "swun" account to
> receive this transaction.

'ssh' is 'secure shell'. If you give it no additional arguments, it spawns a
new shell for you, on the remote machine. If you do give it more, it'll
start that as a program on the remote machine. So, if you want ssh on the
remote machine to dump the output to a file, you can probably do

ssh -l login machine "cat > outfile.dump"

or some such. The quotes are necessary, by the way, to make sure the '>'
redirection is passed to ssh, not interpreted by your shell. But if you
intend to cat a local file, why not scp the file instead ?

scp outfile.dump login at machine:

Your question is more of 'how do I use ssh' than 'how do I use python' ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From mtebeka at intel.com  Tue Jul  4 08:21:53 2000
From: mtebeka at intel.com (Miki Tebeka)
Date: Tue, 4 Jul 2000 09:21:53 +0300
Subject: Using an extention DLL
Message-ID: <Pine.WNT.4.21.0007032147480.315-100000@ilsnt0396.iil.intel.com>

Hello All,

Maybe I missed it in the FAQ but:
I've downloaded unicode.dll which is a unicode module (suprize, suprize
:-)

Now, what do I need to do in order to user it from Python?
(Where to place it, how to load it, etc.)

Thanks.

Bye.
------------------------------------------------------------------------------
Smile, damn it, smile.

(lambda (msg)
  (case msg
	((name)		"Miki Tebeka")
	((email)	"tebeka at lycosmail.com")
	((homepage)	"http://www.tebeka.freeservers.com")
	((quote)	"I don't suffer from insanity, I enjoy every minute of it.")))



From dnotspam-jen at personic.com  Tue Jul  4 08:58:57 2000
From: dnotspam-jen at personic.com (donotspam-jen@personic.com)
Date: Mon, 3 Jul 2000 23:58:57 -0700
Subject: Equivalent to ppm for Python?
Message-ID: <3961881d$0$20233@wodc7nh6.news.uu.net>

Is there anything like PPM for Python?  For those who don't know, PPM is the
Perl Package Manager
(http://www.activestate.com/Support/ActivePerl/PPM.html).  PPM is a utility
that will go out and download and install packages for you from the Internet
(at least on Win32 -- I assume it works elsewhere, too).

Jen






From horst at proceryon.at  Tue Jul  4 09:25:41 2000
From: horst at proceryon.at (Horst Gassner)
Date: Tue, 04 Jul 2000 09:25:41 +0200
Subject: Performance
Message-ID: <39619175.5F8B503D@proceryon.at>

Hi all!

I have implemented a tree view with special items (the base for this
work was the TreeWidget from IDLE).

The performance is not the best and therefore I ran the profiler today.
The following function is one of the slower ones and I would be happy if
there are any possibilties to speed this thing up.

subItems = {}
if s.__subList[level] :
	for key in s.__subList[level].keys ():
		# root level is special case!
		# get all sublevels without recognizing id of sublevel
		if level==0 or string.find(key, id) == 0 :
			subItems[key] = s.__subList[level][key]


The functions looks for all subitems for a given level and a given id.
For each level I have pre-built a list that I have just to go through
the items of the given level.

Thanx in advance
Horst


From gritsch at iue.tuwien.ac.at  Mon Jul  3 12:14:20 2000
From: gritsch at iue.tuwien.ac.at (Markus Gritsch)
Date: Mon, 03 Jul 2000 12:14:20 +0200
Subject: Is there no function to convert a hex number to decimal?
References: <395DB41F.85E51307@iue.tuwien.ac.at>
Message-ID: <3960677C.BED04E21@iue.tuwien.ac.at>

Markus Gritsch wrote:

> Hi!
>
> Although there are several built in functions to convert numbers from
> decimal to hex or oct, it seems that there is no function which can do
> the same in the reverse direction.  Of course I can write 0xff, but
> what should I do, if I have a string containing a hex value?
> Currently I use this brute code:
>
> hex_dec = {}
> for i in range(256):
>     hex_dec[hex(i)[2:]] = i
>
> def dec(hex_string):
>     return hex_dec[string.lower(hex_string)]
>
> But I am not entirely happy with it, because I have to generate this
> clumsy dict, and the function works only for two-digit-hex-numbers.
>
> Is there really no other way?

I found the solution in the string module:
atoi (s[, base]) and atol (s[, base])

BTW, these functions will become obsolete in python 2.0, because there will be an
optional "base" argument to the built-in functions long() and int()

--
                                                                      |\/\/\/|
/------------------------------------------------------------------\  |      |
| Markus GRITSCH                 | phone:    +43 / 1 / 58801-36015 |  |      |
| Institute for Microelectronics | cellular: +43 / 676 / 4973431   |  | (o)(o)
| Technical University of Vienna | fax:      +43 / 1 / 58801-36099 |  C      _)
| Gusshausstrasse  27-29 / E360  | email: gritsch at iue.tuwien.ac.at |   | ,___|
| A-1040  Vienna / AUSTRIA       | SMS:   436764973431 at max.mail.at |   |   /
\------------------------------------------------------------------/  /____\
                                                                     /      \
"Computers let you make more mistakes faster than any other invention in human
 history, with the possible exception of handguns and tequila."
                                                               Mitch Radcliffe





From mrinal at my-deja.com  Tue Jul  4 09:41:06 2000
From: mrinal at my-deja.com (mrinal at my-deja.com)
Date: Tue, 04 Jul 2000 07:41:06 GMT
Subject: Worldpilot
Message-ID: <8js4ef$v4n$1@nnrp1.deja.com>

Hi,

Has anybody tried Worldpilot here? I'm having issues with DateTime.py -
it doesn't seem to work with my timezone (IST - +05:30). Has anybody
worked around this problem? Any patches available for it?

Please cc <mrinal at india.com> on replies..

--
Mrinal Kalakrishnan
mrinal at india.com
http://mrinal.dhs.org/


Sent via Deja.com http://www.deja.com/
Before you buy.


From alex at magenta.com  Tue Jul  4 09:40:14 2000
From: alex at magenta.com (Alex Martelli)
Date: Tue, 4 Jul 2000 09:40:14 +0200
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <8jabvo014tv@news2.newsguy.com> <uhfa7qn9c.fsf@cs.uu.nl> <8jqbru017gf@news2.newsguy.com> <2f985.6708$Tb7.40527@news-server.bigpond.net.au>
Message-ID: <8js4jc09qf@news1.newsguy.com>

Mark Hammond <MarkH at ActiveState.com> wrote in message
news:2f985.6708$Tb7.40527 at news-server.bigpond.net.au...
> "Alex Martelli" <alex at magenta.com> wrote in message
> news:8jqbru017gf at news2.newsguy.com...
>
> > I believe the only relevant issue for the Python community
> > is: C#, like every other Microsoft offering in recent times,
> > will be *STRONGLY* oriented to COM.  Every C# object will be
> > a COM object (just like every Java object was one when run
> > under Microsoft's Java VM), etc, etc.
>
> Its a little more complicated than that!

Well, there's the "common runtime" library that's apparently
going to be shared between all languages, as opposed to C++
having its own MSVCRT.DLL, VB having its own, etc (we did have
that on IBM mainframes in the '70s, of course, but that's
another issue:-); I guess Python may be able to take advantage
of that.  But I didn't see it as specifically a "C#" issue.
What else...?


> > So, C#'s possible (unlikely...) emergence as a major language
> > is just the N-plus-1-th motivation for Windows versions of
> > Python to support COM *very, VERY* well.
>
> And that!

You mean that it will be COM+-as-originally-promised-lo-that-
many-years-ago (with better metadata support, interceptors, etc),
as opposed to the "COM+" which actually emerged without any
of the really interesting promised features?  But even then,
again, I don't quite see it as a "C#" issue.  What am I missing?


> > Yeah, I know, I should put my coding efforts where my mouth
> > is -- and, having expended substantial efforts doing just
> > this sort of thing for a proprietary scripting language
>
> Please do.  Typelib work may be kicking off again (if I ever clear my
> inbox that is - sorry John - I havent forgotten you :-)  A framework for
> calling arbitary vtable interfaces is also in place.  So what you will
> find is that it needs less "framework" code, and more a motivated and
> skilled person to pull a few of these "late" pieces together.  My
> motivation WRT win32com now is more towards keeping the status-quo - and
> helping _others_ change it :-)

Bill Tutt also wrote to me to the same effect, also suggesting I
join the win32 com-developers mailing list.  Turns out I've been
on it for months, but the traffic is so low that I had forgotten.
*Sigh* guess I'll have to look into "what do I need for CVS work
from a Win32 box" and hope the comments in whatever material I do
get from CVS are enough to explain what is it that needs to be done.


> > Re .NET -- I dunno, and I wish I did!
>
> All will be revlealed after the PDC; less than 2 weeks.  I can safely say

Unfortunately I need to present my take on this to mgmt in just
1 week:-)... what (if any) IS the beef on .NET and what strategy
(if any) do we as SW developers need to put in place to be ready
(if it makes any difference to our products).  Hoped I could find
something to use to leverage Python into "major strategic effort"
status around here, but, oh well.  Guess I'll just keep trying to
sneak it in instead (and/or wait for helpful email, if something
can be mentioned in email and not on this group yet:-).

> (as this is a Python language newgroup and C# is apparently a "competing"
> language), the Python community will probably be more interested in .NET
> than in C#.  _Quite_ a bit more :-)

Which at least implies there IS more to .NET than smoke and mirrors
(and XML and SOAP) -- good to know.  Wish I did know more...


> > I'll do y'all a favour and summarize them into basically
> > two stances:
> >     "It's Microsoft and thus evil, vapourware, immoral,
> >         and/or fattening"
> >     "It's Microsoft and thus good, splendid, progressive,
> >         and/or sexy"
>
> Well put :-)

Thanks:-).


> > I do wonder what other items that have technical reality
> > are in .NET besides C#, XML, and SOAP.  If any, that is.
>
> There are - and a number of them.

Good!  Now if I only knew what:-).  Oh well, guess I'll wing
the presentation based on "need to keep flexibility in such
times of rapid change and uncertainty" and Python's excellence
at such flexibility.  Unfortunately our top mgmt is made up
of smart guys _and_ they're wise to my ways, so I doubt this
will fly (maybe I'd better come clean and confess I know there
IS something but I don't know what and will they allow me one
more month of research...?-).


Alex





From piet at cs.uu.nl  Tue Jul  4 10:45:24 2000
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 04 Jul 2000 10:45:24 +0200
Subject: Problem with redirection, buffering and chaining scripts.
References: <395c3b52.183760656@nntp.interaccess.com>
Message-ID: <u8zviqo63.fsf@cs.uu.nl>

On Windows os.popen is broken.
Get the Win32 extensions and use win32pipe.popen.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From gansevle at cs.utwente.nl  Tue Jul  4 12:14:08 2000
From: gansevle at cs.utwente.nl (Fred Gansevles)
Date: Tue, 04 Jul 2000 10:14:08 GMT
Subject: Performance
References: <39619175.5F8B503D@proceryon.at>
Message-ID: <8jsdd9$51r$1@nnrp1.deja.com>

In article <39619175.5F8B503D at proceryon.at>,
  Horst Gassner <horst at proceryon.at> wrote:
> Hi all!
>
> I have implemented a tree view with special items (the base for this
> work was the TreeWidget from IDLE).
>
> The performance is not the best and therefore I ran the profiler
today.
> The following function is one of the slower ones and I would be happy
if
> there are any possibilties to speed this thing up.
>
> subItems = {}
> if s.__subList[level] :
> 	  for key in s.__subList[level].keys ():
> 		# root level is special case!
> 		# get all sublevels without recognizing id of sublevel
> 		if level==0 or string.find(key, id) == 0 :
> 			subItems[key] = s.__subList[level][key]
>
> The functions looks for all subitems for a given level and a given id.
> For each level I have pre-built a list that I have just to go through
> the items of the given level.
>

You can speed things up by putting the special case outside the loop
Another speedup is: use slices instead of string.find since you only
match the first part of the key

<untestded_code>
subItems={}
if s.__subList[level]:
    if level == 0:
        # root level is special case
        subItems = s.__subList[level].copy()
    else:
        id_len = len(id)
        for key in s.__subList[level].keys ():
            if key[:id_len] == id:
                subItems[key] = s.__subList[level][key]
</untested_code>

This might improve things abot 5-10% depending on how often string.find
would fail on finding the id.

> Thanx in advance
> Horst
>

--
-----------------------------------------------------------------------
----          Linux/Windows-NT/IntraNetware Administrator          ----
-- Whenever it sounds simple, I've probably missed something! ... Me --
-----------------------------------------------------------------------


Sent via Deja.com http://www.deja.com/
Before you buy.


From pablo.prieto at dulcesol.es  Tue Jul  4 13:41:23 2000
From: pablo.prieto at dulcesol.es (Pablo Prieto)
Date: Tue, 04 Jul 2000 12:41:23 +0100
Subject: Equivalent to ppm for Python?
References: <3961881d$0$20233@wodc7nh6.news.uu.net>
Message-ID: <3961CD63.553F6D46@dulcesol.es>

"donotspam-jen at personic.com" escribi?:
> 
> Is there anything like PPM for Python?  For those who don't know, PPM is the
> Perl Package Manager
> (http://www.activestate.com/Support/ActivePerl/PPM.html).  PPM is a utility
> that will go out and download and install packages for you from the Internet
> (at least on Win32 -- I assume it works elsewhere, too).
> 
> Jen


Sorry. This is not a reply. Just a request for this question be answered
since I'm interested in it too.
Have a nice day.
Pablo.



From thomas at xs4all.net  Tue Jul  4 12:46:04 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 4 Jul 2000 12:46:04 +0200
Subject: Is there no function to convert a hex number to decimal?
In-Reply-To: <3960677C.BED04E21@iue.tuwien.ac.at>; from gritsch@iue.tuwien.ac.at on Mon, Jul 03, 2000 at 12:14:20PM +0200
References: <395DB41F.85E51307@iue.tuwien.ac.at> <3960677C.BED04E21@iue.tuwien.ac.at>
Message-ID: <20000704124603.D283@xs4all.nl>

On Mon, Jul 03, 2000 at 12:14:20PM +0200, Markus Gritsch wrote:
> Markus Gritsch wrote:

> > Although there are several built in functions to convert numbers from
> > decimal to hex or oct, it seems that there is no function which can do
> > the same in the reverse direction.  Of course I can write 0xff, but
> > what should I do, if I have a string containing a hex value?

> I found the solution in the string module:
> atoi (s[, base]) and atol (s[, base])

> BTW, these functions will become obsolete in python 2.0, because there
> will be an optional "base" argument to the built-in functions long() and
> int()

Before we get a long discussion about backwards compatibility ;) I'd like to
point out that the string.ato* functions might, in the strict sense, become
obsolete, but that doesn't mean they are going to disappear. Rather, the
string module is rewritten to use the builtin functionality: string.atoi()
becomes a wrapper for int().

I doubt the string module will disappear before Py3K, and probably not until
way after Py3K to boot.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From adetalabi at clara.co.uk  Tue Jul  4 13:26:23 2000
From: adetalabi at clara.co.uk (Ade Talabi)
Date: Tue, 04 Jul 2000 12:26:23 +0100
Subject: Socket problem
References: <slrn8m1p8s.2e9.thor@localhost.localdomain>
Message-ID: <3961C9DF.6AFF4D3D@clara.co.uk>

Manuel Gutierrez Algaba wrote:
> 
> I'm doing a game for my University. It's divided into 3 parts
> ( AI (clips), core(python), GUI(python)) communicated by sockets.
> I've been running it in Red Hat 4.2 happily.
> Now, I'm switching to Red Hat 6.0 and the core when
> trying to access to port 50008 it says:
> 
> ocket.error: (99, 'Cannot assign requested address')
> Exception exceptions.AttributeError: 'conn' in
> <method servidor_de_socket.__del__ of servidor_de_socket instance at 8194e40>
>  ignored
> Exception exceptions.AttributeError:
>  'delete non-existing instance attribute' in
> <method nucleo.__del__ of nucleo instance at 81939f8> ignored
> 
> Where is the fault ?
> 
> Thanks.
> 
> --
> MGA
check that /etc/services has got the port set
-- 
We see, whatever we want to see, whether visible or not 
- AT. June 2000.


From m.faassen at vet.uu.nl  Tue Jul  4 13:36:02 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 4 Jul 2000 11:36:02 GMT
Subject: Hi, my name is not Mrs so-called Gumblenose
References: <395B2B88.CC3041C1@memo.volvo.se>
Message-ID: <8jsi72$r02$1@newshost.accu.uu.nl>

PM <vtbl.moorepe at memo.volvo.se> wrote:
> My head hurts.

I'm sorry to hear that.

> If I move to Python from Java from RPG from C++ from C from Basic from
> 6502, am I really, really, really, really stupid? I was not very good at
> -any- of them, but I'm as persistent as a crap dance record on Radio 1.

Persistence is important in a good programmer. :) I skipped Java and RPG and
6502 (I did Z80 there), insert some Pascal somewhere in the list and a
smattering of other languages, but I don't consider myself stupid and
came to Python.

> Will this group answer all questions,

It's decent at getting questions answered, yes. Not all.

> solve life's problems?

Unlikely. :)
 
> Can Python kick large bottoms with decent database access - not just SQL
> not ODBC, could it go to DB2/400?

I don't have any experience with DB2/400, but I would imagine that if
someone hadn't whipped something up already, it's possible to do so.

> What I liked about good old DOS/basic was the ability to knock-things up
> and try things out, now that's what has piqued my interest in Python,
> but is it true....

Yes, Python is good for that kind of iterative style of programming;
rapid prototyping is easy. Advantage above Basic is that the programming
language is actually nice and powerful.

> I'll probably waste some time trying to find out,

Good -- have fun!

> but I'm not 14 years old. Wish I was - and could still drive my Audi.

I'm-not-14-years-old-either-and-don't-have-an-Audi-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From thomas at xs4all.net  Tue Jul  4 13:55:04 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 4 Jul 2000 13:55:04 +0200
Subject: Equivalent to ppm for Python?
In-Reply-To: <3961CD63.553F6D46@dulcesol.es>; from pablo.prieto@dulcesol.es on Tue, Jul 04, 2000 at 12:41:23PM +0100
References: <3961881d$0$20233@wodc7nh6.news.uu.net> <3961CD63.553F6D46@dulcesol.es>
Message-ID: <20000704135504.E283@xs4all.nl>

On Tue, Jul 04, 2000 at 12:41:23PM +0100, Pablo Prieto wrote:
> "donotspam-jen at personic.com" escribi?:

> > Is there anything like PPM for Python?  For those who don't know, PPM is the
> > Perl Package Manager
> > (http://www.activestate.com/Support/ActivePerl/PPM.html).  PPM is a utility
> > that will go out and download and install packages for you from the Internet
> > (at least on Win32 -- I assume it works elsewhere, too).

> Sorry. This is not a reply. Just a request for this question be answered
> since I'm interested in it too.

While it's not exactly what you mean (it doesn't (yet) go out and grab
packages for you), Distutils, which comes with Python 2.0, is probably what
you are looking for. See the Distutils SIG on www.python.org.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From stephen_purcell at yahoo.com  Tue Jul  4 14:34:53 2000
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 04 Jul 2000 12:34:53 -0000
Subject: Performance
In-Reply-To: <39619175.5F8B503D@proceryon.at>
Message-ID: <8jslld+s3p0@eGroups.com>

Horst Gassner <horst at p...> wrote:
> The following function is one of the slower ones and I would be
happy if
> there are any possibilties to speed this thing up.
> 
> subItems = {}
> if s.__subList[level] :
> 	for key in s.__subList[level].keys ():
> 		# root level is special case!
> 		# get all sublevels without recognizing id of sublevel
> 		if level==0 or string.find(key, id) == 0 :
> 			subItems[key] = s.__subList[level][key]

How about:

subItems = {}
if s.__subList[level]:
  for key,value in s.__subList[level].items():
    if not level or not string.find(key,id):
      subItems[key] = value

--
Steve Purcell,  Technical Director, Inkontact
Get in touch at http://www.inkontact.com/
"Life must be simple if I can do it" -- Me





From alexandre.ferrieux at rd.francetelecom.fr  Tue Jul  4 14:44:34 2000
From: alexandre.ferrieux at rd.francetelecom.fr (Alexandre Ferrieux)
Date: Tue, 04 Jul 2000 14:44:34 +0200
Subject: Cross-frame 'security' (Internet Explorer and Python/COM)
References: <Pine.GSO.4.10.10007022217420.19085-100000@sundial> <u$mxTiH5$GA.254@cpmsnbbsa07>
Message-ID: <3961DC32.2344@rd.francetelecom.fr>

Roger Upole wrote:
> 
> Here's a small sample of how to navigate to a URL.
> >>> import win32com.client
> >>> x=win32com.client.Dispatch('InternetExplorer.Application.1')
> >>> x.Visible=1
> >>> x.Navigate('http://www.python.org')

I've been doing something similar in some other scripting language
(replace Python/win32com by Tcl/Optcl), and it worked quite well,
however I've run into a show-stopper, and I would like to gather
insights:

The so-called cross-frame security issue.

Once you've created the automation object x above, you can to some
extent
walk down the document tree. The problem is, if the page contains a
frame
which is hosted in some other 'domain' (the last two items in the DNS
name),
you simply can't access that subtree, even in read-only mode.

In my case it was nasty because the sole purpose of the whole stuff was
to reuse IE's parsing machinery to extract only text from the page -- 
a poor man's Lynx in a maximum-reuse context if you prefer.

I've seen some explanations on the MS site (security bulletin and so
on).
The beginning of the story was the possibility for malevolent
client-side
code to insert its own text input box right on top of the one from a
securized credit card number input dialog. This, sure enough, was a
critical
security hole. However, the counter attack doesn't make sense to me: why
bar
access to extradomain frames, even readonly, while the real issue was to
prevent
only *write* access (or visual overlay) of other frames ?

Please please somebody enlighten me !

TIA,

-Alex


From alexandre.ferrieux at rd.francetelecom.fr  Tue Jul  4 14:30:48 2000
From: alexandre.ferrieux at rd.francetelecom.fr (Alexandre Ferrieux)
Date: Tue, 04 Jul 2000 14:30:48 +0200
Subject: Test results of linuxaudiodev.c
References: <Pine.SGI.3.96.1000629173646.621134u-100000@happy> <m3r99e7v1e.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <3961D8F8.983@rd.francetelecom.fr>

Michael Hudson wrote:
> 
> >     dsp = linuxaudiodev.open("w")
> >     dsp.setparameters(au.getframerate(), au.getsampwidth() * 8,
> >                       au.getnchannels(), formats[spec])
> >     dsp.write(data)
> >     dsp.close()
> 
> This doesn't work.  I don't really understand why.

Can you characterize "doesn't work" ?
If by this you mean 'no errors, no sound', then probably the problem
is that the close doesn't (by default) linger until all the queued data
have been played - in which case success is just one ioctl() away...

-Alex


From scarblac-spamtrap at pino.selwerd.nl  Tue Jul  4 15:15:50 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 4 Jul 2000 13:15:50 GMT
Subject: Win32 : Making script.py executable in DOS box
References: <8jmjfe$64o$1@nnrp1.deja.com> <WzI75.5714$Tb7.31718@news-server.bigpond.net.au> <8jp3pt$q7j$1@nnrp1.deja.com>
Message-ID: <slrn8m3ol2.g8h.scarblac-spamtrap@flits104-37.flits.rug.nl>

Neo Brave wrote in comp.lang.python:
> In article <WzI75.5714$Tb7.31718 at news-server.bigpond.net.au>,
>   "Mark Hammond" <MarkH at ActiveState.com> wrote:
> > Under Windows NT or Win2k, you can use the PATHEXT environment
> > variable. Simply add ".py" to the semi-colon sep'd list of
> > extensions, and off you go!
> 
> Great, thats a help for NT, but I also want to do this on Win98 :-).

There is another trick (originally by Bruce Eckel). Name your Python
file ".bat" instead of ".py" and use the following hack:

@echo off
rem = """
rem Run python on this bat file. Needs the full path where
rem you keep your Python files. The -x causes Python to skip the first
rem line of the file:
python -x c:\python\\"%0".bat %1 %2 %3 %4 %5 %6 %7 %8 %9
goto endofpython
rem """

# The Python program goes here:

print "hello, python"

# For the end of the batch file:

rem = """
:endofpython
rem """


Or just make a .bat file that goes with your .py file and calls python
for it...

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
  3:12pm  up 120 days,  3:25,  5 users,  load average: 0.33, 0.29, 0.20


From aahz at netcom.com  Tue Jul  4 16:02:05 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 4 Jul 2000 14:02:05 GMT
Subject: Microsoft's  C# (Sharp) & .NET -- A Heads Up
References: <qJ165.8938$nZ5.153196@typhoon.southeast.rr.com> <8jqbru017gf@news2.newsguy.com> <2f985.6708$Tb7.40527@news-server.bigpond.net.au> <8js4jc09qf@news1.newsguy.com>
Message-ID: <8jsqot$pj2$1@slb0.atl.mindspring.net>

In article <8js4jc09qf at news1.newsguy.com>,
Alex Martelli <alex at magenta.com> wrote:
>
>Good!  Now if I only knew what:-).  Oh well, guess I'll wing the
>presentation based on "need to keep flexibility in such times of rapid
>change and uncertainty" and Python's excellence at such flexibility.
>Unfortunately our top mgmt is made up of smart guys _and_ they're wise
>to my ways, so I doubt this will fly (maybe I'd better come clean and
>confess I know there IS something but I don't know what and will they
>allow me one more month of research...?-).

Just tell them that a reliable source says that more info will be coming
in two weeks and can you please have a three- or four-week extension on
your presentation so you can present them with Microsoft's
smoke'n'mirrors instead of your faked-up smoke'n'mirrors.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The best way to get information on Usenet is not to ask a question,
but to post the wrong information.  --Aahz


From alex at magenta.com  Tue Jul  4 15:42:54 2000
From: alex at magenta.com (Alex Martelli)
Date: Tue, 4 Jul 2000 15:42:54 +0200
Subject: Class design: accessing "private" members
References: <962352843.1667731268@news.libertysurf.fr> <8jhuvn01vs0@news2.newsguy.com> <962367319.379353174@news.libertysurf.fr> <395D79FE.4E4FDBCE@colconsulting.com> <8jplbd0l5c@news2.newsguy.com> <8jqnpe$vao$1@nnrp1.deja.com>
Message-ID: <8jsptm0sk7@news2.newsguy.com>

Fr?d?ric van der Plancke <fplancke at my-deja.com> wrote in message
news:8jqnpe$vao$1 at nnrp1.deja.com...
> In article <8jplbd0l5c at news2.newsguy.com>,
>   "Alex Martelli" <alex at magenta.com> wrote:
>
> > Such "prohibition" is not hard at all:
> >
> > class dontderive:
> >     def __init__(self):
> >         if self.__class__ != dontderive:
> >             raise "told you not to!"
> >
> > class foo(dontderive):
> >     pass
> >
>
> #but...
>
> class bypassing_prohibition(dontderive):
>    def __init__(self):
>        pass
>
> eh = bypassing_prohibition()
>
> #this works because __init__ never automatically calls the __init__
> #method of base classes...

You're perfectly right, of course.  It's easy, but it can
just as easily be fooled by the would-be-deriver.  Python
is not about programmers warring with each other, the
framework supplier trying to stop things and the framework
client finding clever workarounds; look into C++ or Java
if you want that:-).

If it were, the "if self.__class__" test might be repeated
at entry of each method of dontderive, of course:-).

This only reinforces my main contention that the best
way to stop framework-users from deriving is a line or
two to that effect in the documentation!-)


Alex





From esko.lehtonen at pp.htv.fi  Tue Jul  4 16:09:21 2000
From: esko.lehtonen at pp.htv.fi (Esko Lehtonen)
Date: Tue, 04 Jul 2000 17:09:21 +0300
Subject: PyGtk again but with segfault
References: <39610186.36BAFE88@pp.htv.fi> <396118C4.116790FA@wombatzone.freeserve.co.uk>
Message-ID: <3961F011.FFF27BA2@pp.htv.fi>

Iain Calder wrote:
>  
> At startup, before the first call to draw_array, you need to call
> push_rgb_visual().

Oh. Of course... I should have read better the C example. Now I get my
image displayed. Thank you twice!

--------------------------------------------------------------------
 Esko Lehtonen                   WYSINWYG -              
                                 What You See Is Never What You Get


From timaratz at lisco.com  Tue Jul  4 16:48:12 2000
From: timaratz at lisco.com (Peter Timaratz)
Date: Tue, 4 Jul 2000 09:48:12 -0500
Subject: Looking for Python CGI resources
In-Reply-To: <slrn8m3ol2.g8h.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <000101bfe5c6$e0e73f80$24b745c6@timaratz>

I have the book 'Learning Python' and now I'm looking for material oriented
towards using Python for CGI. The only book that seems to cover this topic
in any depth is 'Internet Programming With Python' and it is out of print.
Is there any other CGI material available?




From matthias.huening at univie.ac.at  Tue Jul  4 16:55:02 2000
From: matthias.huening at univie.ac.at (Matthias Huening)
Date: Tue, 4 Jul 2000 16:55:02 +0200
Subject: Looking for Python CGI resources
References: <000101bfe5c6$e0e73f80$24b745c6@timaratz>
Message-ID: <8jstua$2tk8$1@www.univie.ac.at>

Peter Timaratz <timaratz at lisco.com> schrieb in im Newsbeitrag:
000101bfe5c6$e0e73f80$24b745c6 at timaratz...
> I have the book 'Learning Python' and now I'm looking for material
oriented
> towards using Python for CGI. The only book that seems to cover this
topic
> in any depth is 'Internet Programming With Python' and it is out of
print.
> Is there any other CGI material available?
>
>

Have a look at the 'Web Programming Topic Guide'
http://www.python.org/topics/web/
on the Python-site.

Matthias





From e.smith at cs.bham.ac.uk  Tue Jul  4 17:18:42 2000
From: e.smith at cs.bham.ac.uk (Elliot Smith)
Date: Tue, 04 Jul 2000 16:18:42 +0100
Subject: Entry widgets
Message-ID: <39620052.53C1@cs.bham.ac.uk>

Hello all,

I've tried this question before, and got no answer at all, so I thought
I'd give it another go as the situation is getting desperate! I can't
find any reference to this anywhere.

I've been having problems with Entry widgets in Tkinter, running
under sunos5. For some reason, when I double click on them, Python
closes the whole application and reports 'Bus error'. Any ideas? I've
tried overriding the binding for double mouse clicks to no avail.

Elliot
--
homepage: http://www.cs.bham.ac.uk/~ezs/
"A scholar is just a library's way of making another library" (Richard
Dawkins)


From cut_me_out at hotmail.com  Tue Jul  4 17:41:33 2000
From: cut_me_out at hotmail.com (Alex)
Date: 04 Jul 2000 11:41:33 -0400
Subject: Problem with dynamically linked extensions that use Numeric
References: <etdbt0g1zos.fsf@w20-575-116.mit.edu>
Message-ID: <etdhfa5q4wi.fsf@w20-575-32.mit.edu>

Just to follow up here, the problem was pointed out by Paul Dubois in
another thread -- I need to do an import_array (); after the Py_InitModule. 

Alex.


From tiddlerdeja at my-deja.com  Tue Jul  4 18:07:16 2000
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Tue, 04 Jul 2000 16:07:16 GMT
Subject: Java line to JPython (help Obj.class)
Message-ID: <8jt239$jb3$1@nnrp1.deja.com>

Can anyone help me turn this java line into JPython?

URL dtdURL = DOMConfigurator.class.getResource("l");

The problem I'm getting is that DOMConfigurator doesn't seem to have
attribute class.

Is there a workaround?

Any help appreciated.


Sent via Deja.com http://www.deja.com/
Before you buy.


From thor at localhost.localdomain  Tue Jul  4 18:13:05 2000
From: thor at localhost.localdomain (Manuel Gutierrez Algaba)
Date: 4 Jul 2000 16:13:05 GMT
Subject: Socket problem
References: <slrn8m1p8s.2e9.thor@localhost.localdomain> <39652e57.16380596@news.btx.dtag.de>
Message-ID: <slrn8m4acu.na.thor@localhost.localdomain>

On Tue, 04 Jul 2000 00:23:05 GMT, Stefan Franke <spamfranke at bigfoot.de> wrote:
>[Might be a repost..]
>
>On 3 Jul 2000 17:08:24 GMT, thor at localhost.localdomain (Manuel Gutierrez Algaba) wrote:
>
>>
>>I'm doing a game for my University. It's divided into 3 parts
>>( AI (clips), core(python), GUI(python)) communicated by sockets.
>
>Sorry, no help for your question, but just for interest - how do you interface
>with clips?

expectk

BTW, I've solved the problem, 

/sbin/ifconfig lo 127.0.0.1 did the trick.

-- 
MGA


From david_ullrich at my-deja.com  Tue Jul  4 19:10:31 2000
From: david_ullrich at my-deja.com (david_ullrich at my-deja.com)
Date: Tue, 04 Jul 2000 17:10:31 GMT
Subject: Bug in math.frexp?
References: <LNBBLJKPBEHFEDALKOLCGENKGHAA.tim_one@email.msn.com>
Message-ID: <8jt5ps$lte$1@nnrp1.deja.com>

In article <LNBBLJKPBEHFEDALKOLCGENKGHAA.tim_one at email.msn.com>,
  "Tim Peters" <tim_one at email.msn.com> wrote:
> Guys, this is not open to debate.  frexp is a std ANSI C function,
Python
> calls the platform frexp directly, and the new docstring I checked in
is an
> exact paraphrase of the std in Pythonic language.  I don't care what
> manpages say.  I *would* care if somebody can show an actual example
of a
> Python implementation where the examples posted do not work exactly as
> shown:  that would mean the platform is buggy -- but the new docs are
still
> correct <wink>.

   Indeed. I haven't seen the new docs except for that docstring
you posted - I was gonna suggest that maybe you might even want
to be explicit about the fact that the _sign_ of the mantissa
is the same as the sign of the argument (if it's already there
never mind, all I saw was a statement about abs(mantissa), but
I'm know to be sort of blind sometimes.)

   One might say of course that's what the sign of the mantissa
is, what else makes sense. But you may as well be explicit -
someone right here seemed to think that the sign of the
_exponent_ should be the same as the sign of the argument.

> arrogant-but-with-cause-in-this-case-ly y'rs  - tim

   Don't look all that arrogant to me, which probably
explains a lot.

DU


Sent via Deja.com http://www.deja.com/
Before you buy.


From aa8vb at yahoo.com  Tue Jul  4 20:23:16 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 04 Jul 2000 14:23:16 -0400
Subject: PIL: Can it work with large images?
Message-ID: <20000704142316.A909732@vislab.epa.gov>

Does PIL support cropping while loading large images?  

I'd like to pull in small portions of large GeoTIFF images for processing
GeoTIFFs can be 100MB - multi-GBs in size.  I have one here that's 340MB.

Thanks,

Randall



> tiff_info.py
filename : map.tif
info     : {'compression': 'raw'}
mode     : P
size     : (24758, 13716)           <-------------------------------------<<<
num tiles: 13716
  tile #1: ('raw', (0, 0, 24758, 1), 56757, ('L', 0, 1))
  tile #2: ('raw', (0, 1, 24758, 2), 81515, ('L', 0, 1))
  ...

-- 
Randall Hopper
aa8vb at yahoo.com



From embed at NOSPAM.geocities.com  Tue Jul  4 20:42:18 2000
From: embed at NOSPAM.geocities.com (Warren Postma)
Date: Tue, 4 Jul 2000 14:42:18 -0400
Subject: visual C++ giving runtime heap-check error in Py_Finalize(x)
Message-ID: <Fdq85.15710$227.290054@nnrp1.uunet.ca>

I have managed to get a repeatable (on my machine) version of either a
reference count or other related kind of nasty bug in my embedded C++
application when I call Py_Finalize.

The error I get is when trying to malloc some memory for a new object, the
malloc call senses corruption of a previously freed block of memory, after
the time it was last freed. here is the trace message that is printed out to
the visual C++ output window:

I get the following call stack:

NTDLL! 77f9f9df()
NTDLL! 77fcc8a1()
NTDLL! 77f87aa3()
NTDLL! 77fb3c33()
NTDLL! 77fa7131()
NTDLL! 77fca4cb()
_free_base(void * 0x005d8a40) line 96
_free_dbg_lk(void * 0x005d8a60, int 0x00000001) line 1116 + 9 bytes
_free_dbg(void * 0x005d8a60, int 0x00000001) line 1001 + 13 bytes
free(void * 0x005d8a60) line 956 + 11 bytes
string_dealloc() line 175 + 9 bytes
PyString_Fini() line 1112 + 88 bytes
Py_Finalize() line 242

What is happening is this:

In PyString_Fini it is freeing the set of characters 0 to 255. When freeing
character[17] I trace in, and it is calling  the standard C method "free()"
to release PyObject characters[17] which is the PyObject (string) memory at
0x5d8a90.    As you can see above, some pointer mangling ends up happening.

The reference count for the PyObject is 1 already, and the object itself
does not appear to be corrupt.  I added the line char *c =
PyString_FromString( characters[i] ); just to see if I could still get the
character values out of the strings, and they are all still there, and this
string does not appear corrupt in any way but the heap debug code prints
this warning and throws a debug exception:

HEAP[myprogram.exe]: HEAP: Free Heap block 5d8a90 modified at 5d8ab8 after
it was freed

The funny thing is that there are no other threads in there, so if anything
is modifying anything, then it's the malloc() code itself doing it.

Has anybody else had a problem like this ever?

Trick is, it does this both with my custom (statically linked) Python 1.5.2
and with the python15.dll also, so it's either (a) a bug in my code
somewhere, probably one of my C extensions somewhere or (b) something weird
in Python itself, which my code is causing to happen, but which is
definitely a bug in 1.5.2.  I am leaning towards (A) because it is 99.9999%
of the time always my fault! :-)

Warren Postma





From dalke at acm.org  Tue Jul  4 21:09:21 2000
From: dalke at acm.org (Andrew Dalke)
Date: Tue, 4 Jul 2000 13:09:21 -0600
Subject: Random number generation, simple question.
References: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se> <8jqjf8$s1t$1@nnrp1.deja.com> <39615d75.142572338@news>
Message-ID: <8jtcid$4um$1@nntp9.atl.mindspring.net>

Reini Urban wrote:
>There were some C libs around dealing with
>some special PC timer chips which can be abused for exactly that.
>
>such as the linux /dev/random pseudodevice

I was looking into this a few weeks ago.  I was doing some CGI work
in Perl and wanted to make sure the random number I used for the session
id was hard to guess.  If you call rand without calling srand, Perl
will use a combination of the time (in microseconds, if available),
PID, current stack size, value of the pointer passed to get the time,
and several other values.

What I saw, though, was a compile time option to use /dev/random if
it exists.  I would have thought that would be enabled by default, but
it made some comment about how the read will block if there isn't enough
entropy in the source pool.  Is this ever a problem in real life?

                    Andrew
                    dalke at acm.org





From embed at NOSPAM.geocities.com  Tue Jul  4 21:34:10 2000
From: embed at NOSPAM.geocities.com (Warren Postma)
Date: Tue, 4 Jul 2000 15:34:10 -0400
Subject: visual C++ giving runtime heap-check error in Py_Finalize(x)
References: <Fdq85.15710$227.290054@nnrp1.uunet.ca>
Message-ID: <h_q85.15860$227.292308@nnrp1.uunet.ca>

I figured it out. I also found this nifty trick.

here it is, in case anyone else needs to track nasty malloc problems in
their Visual C++ applications, most especially useful to me while
Py_Finalize is running.

I call SetSlowButThoroughDebugOptions() before Py_Finalize, and after it, I
clear it with SetFastDebugOptions.   This may turn out to be handy for
anybody writing C extensions or hacking Python internals.

Warren Postma

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

// setdebug.c
//
// set visual C++ runtime library heap-debug options
//
// see MSDN help topic "Using the Debug Heap from C++" and "Debug Routines"
for more help
// on the CrtSetDbgFlag and it's functions.
//
// version 0.1 warren postma
// july 4, 2000

#include <crtdbg.h>     // Microsoft C Runtime Debug Options header.

#ifdef _DEBUG
void SetFastDebugOptions(void)
{
  int tmpFlag;

    // Get the current state of the flag
    // and store it in a temporary variable
    tmpFlag = _CrtSetDbgFlag( _CRTDBG_REPORT_FLAG );

    // Set the new state for the flag
    _CrtSetDbgFlag( tmpFlag );

}

void SetSlowButThoroughDebugOptions(void)
{
  int tmpFlag;

    // Get the current state of the flag
    // and store it in a temporary variable
    tmpFlag = _CrtSetDbgFlag( _CRTDBG_REPORT_FLAG );


    // Always check at every allocation request:
    tmpFlag |= _CRTDBG_CHECK_ALWAYS_DF;

    // Leak Check Report
    tmpFlag |= _CRTDBG_LEAK_CHECK_DF;

    // Set the new state for the flag
    _CrtSetDbgFlag( tmpFlag );

}

#endif


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




From ldwhitley at home.com  Tue Jul  4 21:41:30 2000
From: ldwhitley at home.com (Larry Whitley)
Date: Tue, 04 Jul 2000 19:41:30 GMT
Subject: Python to C++ in a pickle
Message-ID: <K5r85.38150$dF.1427872@news1.rdc1.il.home.com>

I'd like to pickle an object in python and unpickle it in C++.  The object
in question is a python dictionary so it will have to go into a C++ map.
Has someone done this and is willing to share?  Or, is the cPickle source
available on the web somewhere?

Larry




From petersc at stud.ntnu.no  Tue Jul  4 23:57:39 2000
From: petersc at stud.ntnu.no (Peter Schneider-Kamp)
Date: Tue, 04 Jul 2000 21:57:39 +0000
Subject: (s)re module code break
Message-ID: <39625DD3.ED9BF9D7@stud.ntnu.no>

I just noticed an issue with the new re module. Maybe
it is old stuff, I don't know. Just to be sure:

using re.compile yields a SRE_Pattern object
(as of 2.0b1) whose method demands 3 arguments.

There is a lot of code which lives happily
just supplying 2 arguments and having count
default to 0 (until 1.6a2).

I added a 0 as third parameter and my program
once again ran happily.

Does anyone know if this behaviour of sre will persist?

worried-ly y'rs
Peter
--
Peter Schneider-Kamp          ++47-7388-7331
Herman Krags veg 51-11        mailto:peter at schneider-kamp.de
N-7050 Trondheim              http://schneider-kamp.de



From jkraska1 at san.rr.com  Tue Jul  4 21:58:21 2000
From: jkraska1 at san.rr.com (Courageous)
Date: Tue, 04 Jul 2000 19:58:21 GMT
Subject: Python to C++ in a pickle
References: <K5r85.38150$dF.1427872@news1.rdc1.il.home.com>
Message-ID: <396242AB.87D3795B@san.rr.com>

Larry Whitley wrote:
> 
> I'd like to pickle an object in python and unpickle it in C++.  The object
> in question is a python dictionary so it will have to go into a C++ map.
> Has someone done this and is willing to share?  Or, is the cPickle source
> available on the web somewhere?

To do this in a general way, you'll have to keep look-aside metadata
and make certain that the pickling mechanism complies with the
perceived types of information that are expected by your C++
client. I don't know anything about pickling, but I suspect it
doesn't do this.

Isn't cPickle source part of the python source distribution? I
think so....

C/


From azz at gnu.org  Tue Jul  4 16:07:08 2000
From: azz at gnu.org (Adam Sampson)
Date: 04 Jul 2000 15:07:08 +0100
Subject: Win32 : Making script.py executable in DOS box
References: <8jmjfe$64o$1@nnrp1.deja.com> <WzI75.5714$Tb7.31718@news-server.bigpond.net.au> <8jp3pt$q7j$1@nnrp1.deja.com>
Message-ID: <87og4eugz7.fsf@cartman.azz.net>

Neo Brave <neo_brave at my-deja.com> writes:

> > No solution I am aware of for 95, although there are some tricks
> > you can pull (Im sure someone else will fill these in for you)!
> Anyone???

How about this, saved in your PATH as hello.bat:

rem = """
python wherever\hello.bat %1 %2 %3 %4 %5 %6 %7 %8 %9
exit
"""
print "hello, world"
# more Python code

which can then be run by typing "hello"?

-- 

Adam Sampson
azz at gnu.org


From digitig at cix.co.uk  Tue Jul  4 23:23:00 2000
From: digitig at cix.co.uk (Tim Rowe)
Date: Tue, 4 Jul 2000 22:23 +0100 (BST)
Subject: Random number generation, simple question.
References: <8jqjf8$s1t$1@nnrp1.deja.com>
Message-ID: <memo.20000704222314.51603A@digitig.compulink.co.uk>

In article <8jqjf8$s1t$1 at nnrp1.deja.com>, jhylton at my-deja.com () wrote:

> In article <Pine.GSO.4.21.0007031546410.25965-100000 at luna.dsv.su.se>,
>   Kalle Svensson <kalle at gnupung.net> wrote:
> 
> > I have a few questions about the (pseudo) random number generator
> supplied
> > in the standard library. I need 80 random bits (for a CipherSaber IV)
> but
> > I hear most RNG's only supply 32 bits or less.
> 
<snip>
 
> I recommend looking at the Yarrow design by Kelsey, Schneier, and
> Ferguson:  http://www.counterpane.com/yarrow.html
> 
> A Python implementation of Yarrow-160 is included in the Pisces
> distribution:
> http://www.cnri.reston.va.us/software/pisces/manual/module-pisces.yarrow
> .html

Beat me to it! Yes, that's the serious way forward for a crypto 
application.


From NOSPAM_rdrew at cia-group.com  Tue Jul  4 23:46:55 2000
From: NOSPAM_rdrew at cia-group.com (Ray Drew)
Date: Tue, 4 Jul 2000 22:46:55 +0100
Subject: Can't get pythondoc to work
Message-ID: <8jtm0o$6rv$1@soap.pipex.net>

I can't get pythondoc to work. I've got pythondoc, pythondoc\formatters and
pythondoc\xref in the pythonpath. When I try to run it on the test.py file I
get a couple of module not found errors, even though they appear to be in
the python path (I can import them interactively in PythonWin). When I try
to run python pdoc.py test.py I get this traceback:

D:\bin\cpython\Lib\pythondoc>python pdoc.py test.py
Warning: Failed to import internalXref (exceptions.ImportError: No module
named
docobjects)
Traceback (innermost last):
  File "pdoc.py", line 2, in ?
    import pythondoc.pythondoc
  File "pythondoc.py", line 81, in ?
    formatters.init(_options)
  File "formatters\__init__.py", line 4, in init
    init(__path__, options)
  File "formatters\format_init.py", line 25, in init
    import pythondoc.message
ImportError: No module named message

The pythonpath is:

D:\bin\cpython\Lib\pythondoc>echo %pythonpath%
d:\bin\cpython\lib;d:\bin\cpython\lib\Numeric;d:\bin\cpython\lib\htmldoc;d:\
bin\
cpython\lib\pythondoc;d:\bin\cpython\lib\pythondoc\formatters;d:\bin\cpython
\lib
\pythondoc\xref

Python 1.5.2 on Windows 98

Can anyone see what I'm doing wrong?

Thanks,

Ray




From slinkp23 at yahoo.com  Tue Jul  4 19:47:28 2000
From: slinkp23 at yahoo.com (Paul Winkler)
Date: Tue, 04 Jul 2000 13:47:28 -0400
Subject: Postscript font metrics?
Message-ID: <39622330.3732B225@yahoo.com>

I need to generate some nice Postscript output from a python script,
and I'm finding that I need to get at the font metrics info from the
.afm files so I can place things correctly. (Either that or write
*everything* in postscript -- no way!!)

I can't find anything like this on Parnassus. Anyone ever seen any
python code that gets the useful info from .afm files?

If not, I might have a go at hacking the relevant parts of the Perl
PostScript-Font-1.02 module into Python... but then I'll have to
(yuck) read someone else's Perl code, which I haven't done in a
looong time and I was never that good at it.

-- 
.................    paul winkler    ..................
slinkP arts:   music, sound, illustration, design, etc.
        web page:  http://www.ulster.net/~abigoo        
      A member of ARMS:   http://www.reacharms.com
               or http://www.mp3.com/arms


From wjdandreta at worldnet.att.net  Wed Jul  5 02:22:30 2000
From: wjdandreta at worldnet.att.net (William Dandreta)
Date: Wed, 05 Jul 2000 00:22:30 GMT
Subject: Looking for Python CGI resources
References: <000101bfe5c6$e0e73f80$24b745c6@timaratz>
Message-ID: <adv85.7201$xL3.526766@bgtnsc06-news.ops.worldnet.att.net>

Hi Peter,

I think I read somewhere that the second edition of IPWP is being worked on
and I thought it was to be released this summer.

Bill
Peter Timaratz wrote in message <000101bfe5c6$e0e73f80$24b745c6 at timaratz>...
>I have the book 'Learning Python' and now I'm looking for material oriented
>towards using Python for CGI. The only book that seems to cover this topic
>in any depth is 'Internet Programming With Python' and it is out of print.
>Is there any other CGI material available?
>
>




From chibaA at TinterlogD.Tcom  Wed Jul  5 02:24:41 2000
From: chibaA at TinterlogD.Tcom (chibaA at TinterlogD.Tcom)
Date: Wed, 05 Jul 2000 00:24:41 GMT
Subject: regex's
Message-ID: <39627fa7.705005544@news1.on.sympatico.ca>

Hi...  I'm having a bit of a brain-tease...  I have the following code
(example):

import re

str = '123'
if re.compile('1').match(str):
	print "matched 1"
if re.compile('2').match(str):
	print "matched 2"
if re.compile('3').match(str):
	print "matched 3"

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

I would think that this code would print 'matched 1 matched 2 matched
3'...  But in this case, it would only printed "matched 1".  Any
reasons?  Is there any way to have it print all three (since the
conditions seem to be true)?

Thanks in advance!

kc.


From johannes at zellner.org  Wed Jul  5 03:08:26 2000
From: johannes at zellner.org (Johannes Zellner)
Date: 5 Jul 2000 01:08:26 GMT
Subject: history file
Message-ID: <slrn8m52ka.l0o.johannes@krispc6.physik.uni-karlsruhe.de>

Hello,

is it possible to read and write a history file
(e.g. ~/.python_history) ?

-- 
   Johannes


From tim.hochberg at ieee.org  Wed Jul  5 03:47:49 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Wed, 05 Jul 2000 01:47:49 GMT
Subject: regex's
References: <39627fa7.705005544@news1.on.sympatico.ca>
Message-ID: <m2g0ppmm9x.fsf@spiff.home.com>

chibaA at TinterlogD.Tcom writes:

> Hi...  I'm having a bit of a brain-tease...  I have the following code
> (example):
> 
> import re
> 
> str = '123'
> if re.compile('1').match(str):
> 	print "matched 1"
> if re.compile('2').match(str):
> 	print "matched 2"
> if re.compile('3').match(str):
> 	print "matched 3"
> 
> -------------------
> 
> I would think that this code would print 'matched 1 matched 2 matched
> 3'...  But in this case, it would only printed "matched 1".  Any
> reasons?  Is there any way to have it print all three (since the
> conditions seem to be true)?

re.match matches from the beginning of the string (e.g.,
re.compile('2').match(str) will only match strings starting with
'2'. What you want is re.search. For example:

#....
if re.compile('2').search(str):
	# ...

You could also use:

#....
if re.search('2', str):
	# ...

However, the former will be faster if you reuse the regular
expression since you can compile it once and reuse the compiled
version.


-tim
 
> Thanks in advance!
> 
> kc.



From chibaA at TinterlogD.Tcom  Wed Jul  5 03:53:38 2000
From: chibaA at TinterlogD.Tcom (chibaA at TinterlogD.Tcom)
Date: Wed, 05 Jul 2000 01:53:38 GMT
Subject: regex's
References: <39627fa7.705005544@news1.on.sympatico.ca> <m2g0ppmm9x.fsf@spiff.home.com>
Message-ID: <396294ff.710469961@news1.on.sympatico.ca>

Perfect!

Thanks muchly!

kc


On Wed, 05 Jul 2000 01:47:49 GMT, Tim Hochberg <tim.hochberg at ieee.org>
wrote:

>chibaA at TinterlogD.Tcom writes:
>
>> Hi...  I'm having a bit of a brain-tease...  I have the following code
>> (example):
>> 
>> import re
>> 
>> str = '123'
>> if re.compile('1').match(str):
>> 	print "matched 1"
>> if re.compile('2').match(str):
>> 	print "matched 2"
>> if re.compile('3').match(str):
>> 	print "matched 3"
>> 
>> -------------------
>> 
>> I would think that this code would print 'matched 1 matched 2 matched
>> 3'...  But in this case, it would only printed "matched 1".  Any
>> reasons?  Is there any way to have it print all three (since the
>> conditions seem to be true)?
>
>re.match matches from the beginning of the string (e.g.,
>re.compile('2').match(str) will only match strings starting with
>'2'. What you want is re.search. For example:
>
>#....
>if re.compile('2').search(str):
>	# ...
>
>You could also use:
>
>#....
>if re.search('2', str):
>	# ...
>
>However, the former will be faster if you reuse the regular
>expression since you can compile it once and reuse the compiled
>version.
>
>
>-tim
> 
>> Thanks in advance!
>> 
>> kc.
>



From stephen_purcell at yahoo.com  Wed Jul  5 04:46:49 2000
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Wed, 05 Jul 2000 02:46:49 -0000
Subject: Java line to JPython (help Obj.class)
In-Reply-To: <8jt239$jb3$1@nnrp1.deja.com>
Message-ID: <8ju7ip+ed4s@eGroups.com>

tiddlerdeja at m... wrote:
> Can anyone help me turn this java line into JPython?
> 
> URL dtdURL = DOMConfigurator.class.getResource("l");
> 
> The problem I'm getting is that DOMConfigurator doesn't seem to have
> attribute class.
> 
> Is there a workaround?

How about:

 dtdURL = DOMConfigurator.getResource('l')


If DOMConfigurator is a Java class, then the JPython object
'DOMConfigurator' is equivalent to the 'DOMConfigurator.class' in
Java.

e.g.,

>>> import java.lang.Object
>>> java.lang.Object
<jclass java.lang.Object at -986770152>


-Steve

--
Steve Purcell,  Technical Director, Inkontact
Get in touch at http://www.inkontact.com/
Get testing at http://pyunit.sourceforge.net/
"Life must be simple if I can do it" -- Me






From olivierS.dagenaisP at canadaA.comM  Wed Jul  5 04:52:39 2000
From: olivierS.dagenaisP at canadaA.comM (Olivier Dagenais)
Date: Wed, 05 Jul 2000 02:52:39 GMT
Subject: Q:  Client-server architecture
Message-ID: <Xpx85.37285$W35.821323@news20.bellglobal.com>

What modules [or references] are there that explain or implement
simultaneous, __secure__ operations on a server?

I looked through http://www.python.org/topics/web/  and it seems to only
cover HTTP, not HTTPS (using SSL or some other, secure protocol).

I want to implement seemingly concurrent, remote database access and if I
can queue the requests to the server, the server could then process
requests, one at a time, and then "call back" each client with the results.

I'm thinking I don't want to write the server side software using a CGI
script, since the database package I'm looking into using (Gadfly) isn't
designed for concurrent access, unless I don't understand how CGI works and
it is synchronous?  (I have only written very simple CGI scripts in Perl and
C, so I'm sort of new to this)

Any help (including corrections in my thinking of CGI or hints on handling
many, simultaneous requests) would be appreciated.  I'd like to write both
the client and the server in Python.

Thanks!

--
----------------------------------------------------------------------
Olivier A. Dagenais - Carleton University - Computer Science III




From olivierS.dagenaisP at canadaA.comM  Wed Jul  5 05:03:01 2000
From: olivierS.dagenaisP at canadaA.comM (Olivier Dagenais)
Date: Wed, 05 Jul 2000 03:03:01 GMT
Subject: Q: Concurrent versionning
Message-ID: <Fzx85.37329$W35.818663@news20.bellglobal.com>

Hi again...

I'm looking into implementing a version control system on a database, so
that individual rows (of select tables) can be versionned.  (ie:  make a
copy of that row and increment the version field of the new one, then make
changes)

My question is regarding modules [or resources] that can detect and report
line-level or paragraph-level differences between two text files, very much
like the windiff tool provides.  Are there such things, out there, already
available?

Not only that, but (as I'm accustomed to the SLM - aka Slime - version
control system, the only one I've ever used) also the capability to merge
two textfiles and to elegantly report merge conflicts, very much like
"ssync" (again, part of SLM) will do.

Maybe this is simple enough to implement, and I just need to understand some
techniques in concurrent versionning, so any help would be appreciated.

Thanks!

--
----------------------------------------------------------------------
Olivier A. Dagenais - Carleton University - Computer Science III





From djc at object-craft.com.au  Wed Jul  5 05:18:30 2000
From: djc at object-craft.com.au (Dave Cole)
Date: 05 Jul 2000 13:18:30 +1000
Subject: Undocumented regex behaviour in re module
Message-ID: <m3itulmfi1.fsf@vole.object-craft.com.au>

Consider the following program:
- - re_test.py - - - - - - - - - - - - - - - - - - - - - - - - - -
import re

text = '''<html>
  <head>
    <title>Browse <?table?></title>
  </head>
  <body>
    <h1>Browse <?table?></h1>

      <?browse start=0 num=25?>

  </body>
</html>
'''

tags = re.compile(r'<\?(\w+)(\s+\w+=\w+)*\?>', re.I | re.M)
match = tags.search(text)
while match:
    print text[match.start():match.end()], '=>', match.groups()
    match = tags.search(text, match.end())
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

The idea here is that I want to be able to extract the special tag
names and their attributes.  Everything work fine, except for tag:

        <?browse start=0 num=25?>

The match object only saves the last attr=value matched.  The only way
that I think of to get all of the attr=value returned is to change the
regex to:

        <\?(\w+)((?:\s+\w+=\w+)*)\?>

Unfortunately, that is not as useful as I would like since it returns
a string which needs further processing: ' start=0 num=25'

Any hint / explanation at this point would be gratefully accepted.

- Dave


From lull at acm.org  Wed Jul  5 06:36:02 2000
From: lull at acm.org (John Lull)
Date: 4 Jul 2000 23:36:02 -0500
Subject: Singleton COM local/remote server?
Message-ID: <7bd5msssk48km280t9bvds4i3v6hs7r8or@4ax.com>

I'm prototyping a DCOM server in Python.  The server coordinates
access from several clients scattered across a network to multiple
devices hung off each of several interface ports.  Some or all of it
will likely be ported to C once we're satisfied the API is right, and
as performance dictates. 

The server will have a single Coordinator object, a single Port object
with its own thread for each port, a single Device object with its own
thread for each device, multiple PortAccess objects to provide client
access to each port object, and multiple DeviceAccess objects to
provide client access to each device.  Client programs will get access
to the server via win32com.client.Dispatch('MyServer.Coordinator') (or
the equivalent in another language), then request PortAccess and
DeviceAccess objects from appropriate factory methods.

I've got most of the innards worked out, figured out how to register
it, wrap the access objects so I can hand out references to clients,
etc, but --

Since each port can belong to at most one server, I cannot have
multiple servers on the same machine.  I can't figure out how to
ensure there is only *one* copy of the server started, regardless of
how many clients want to use it.  

Any samples, pointers to documentation, key phrases to search MSDN
for, etc. would be *greatly* appreciated.

Thank you.

Regards,
John


From moshez at math.huji.ac.il  Wed Jul  5 07:08:02 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Wed, 5 Jul 2000 08:08:02 +0300 (IDT)
Subject: Python to C++ in a pickle
In-Reply-To: <K5r85.38150$dF.1427872@news1.rdc1.il.home.com>
Message-ID: <Pine.GSO.4.10.10007050807380.109-100000@sundial>

On Tue, 4 Jul 2000, Larry Whitley wrote:

> I'd like to pickle an object in python and unpickle it in C++.  The object
> in question is a python dictionary so it will have to go into a C++ map.
> Has someone done this and is willing to share?  Or, is the cPickle source
> available on the web somewhere?

Why use pickle? XML-RPC seems much better for this purpose.
--
Moshe Zadka <moshez at math.huji.ac.il>
There is no GOD but Python, and HTTP is its prophet.




From ralf.claus at teleatlas.com  Wed Jul  5 07:04:31 2000
From: ralf.claus at teleatlas.com (Ralf Claus)
Date: Wed, 05 Jul 2000 07:04:31 +0200
Subject: WinNT User Name
References: <Ae485.94210$5k2.194235@dfw-read.news.verio.net>
Message-ID: <3962C1DF.43E3EBDB@teleatlas.com>

michael montagne wrote:
> 
> First off, I'm a newbie at Python.
> I'm trying to return the user who is logged on at a PC.  What I really need
> is the login name.  How do I do that?  Seems like it otta be quite simple
> but I'm having trouble.
> 
> -mjm

..and finally

import getpass
username = getpass.getuser()

works on both systems (NT,UNIX), if you need.


From ralf.claus at teleatlas.com  Wed Jul  5 07:07:13 2000
From: ralf.claus at teleatlas.com (Ralf Claus)
Date: Wed, 05 Jul 2000 07:07:13 +0200
Subject: WinNT User Name
References: <Ae485.94210$5k2.194235@dfw-read.news.verio.net>
Message-ID: <3962C281.405FB34E@teleatlas.com>

michael montagne wrote:
> 
> First off, I'm a newbie at Python.
> I'm trying to return the user who is logged on at a PC.  What I really need
> is the login name.  How do I do that?  Seems like it otta be quite simple
> but I'm having trouble.
> 
> -mjm

..and finally

import getpass
username = getpass.getuser()

works on both systems (NT,UNIX), if you need.


From wjdandreta at worldnet.att.net  Wed Jul  5 07:15:39 2000
From: wjdandreta at worldnet.att.net (William Dandreta)
Date: Wed, 05 Jul 2000 05:15:39 GMT
Subject: Python and .lib files
Message-ID: <%vz85.6488$oh5.499440@bgtnsc04-news.ops.worldnet.att.net>

Is it possible to call functions in a .lib file with Python?

If so, what is the easiest way to do so?

Bill




From swun at esec.com.au  Wed Jul  5 07:45:56 2000
From: swun at esec.com.au (sss)
Date: Wed, 05 Jul 2000 15:45:56 +1000
Subject: DCOracle error
Message-ID: <3962CB94.2DDA9AA9@esec.com.au>

I don't know why I got this error when I try to install DCOracle in the
Linux machine.
I've got Oracle sqlplus: SQL*Plus: Release 8.1.5.0.0

Here is the err msg:

[oracle at ironfist src]$ make
gcc -fPIC -g -O2 -I/usr/include/python1.5 -I/usr/include/python1.5
-DHAVE_CONFIG_H  -I/home/oracle/product/8.15/rdbms/demo
-I/home/oracle/product/8.15/rdbms/public
-I/home/oracle/product/8.15/network/public
-I/home/oracle/product/8.15/plsql/public -DDCORACLE8 -c ./oci_.c
In file included from ./oci_.c:562:
Buffer.h:79: warning: static declaration for `PyImport_Import' follows
non-static
oci_8.c: In function `Logon':
In file included from ./oci_.c:693:
oci_8.c:95: warning: passing arg 4 of `OCIHandleAlloc' makes integer
from pointer without a cast
oci_8.c: In function `newLob':
oci_8.c:446: warning: passing arg 4 of `OCIDescriptorAlloc' makes
integer from pointer without a cast
oci_8.c: In function `initoci_8':
oci_8.c:483: warning: passing arg 3 of `OCIEnvInit' makes integer from
pointer without a cast
./oci_.c: In function `_wrap_odescr':
./oci_.c:2595: warning: assignment from incompatible pointer type
gcc -shared  oci_.o  -L/home/oracle/product/8.15/lib/
-L/home/oracle/product/8.15/rdbms/lib
/home/oracle/product/8.15/rdbms/lib/defopt.o
/home/oracle/product/8.15/lib/sscoreed.o
/home/oracle/product/8.15/lib/nautab.o
/home/oracle/product/8.15/lib/naeet.o
/home/oracle/product/8.15/lib/naect.o
/home/oracle/product/8.15/lib/naedhs.o -lnetv2 -lnttcp -lnetwork -lncr
-lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lmm
-lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lnetv2 -lnttcp -lnetwork
-lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lepc
-lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient -lvsn -lcommon
-lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lsocket -lnsl
-lgen -ldl -lelf -laio -lposix4 -lsql -lm -lm -lnetv2 -lnttcp -lnetwork
-lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lmm
-lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lnetv2 -lnttcp -lnetwork
-lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lepc
-lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient -lvsn -lcommon
-lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lsocket -lnsl
-lgen -ldl -lelf -laio -lposix4 -lsql -lm -lm -o oci_.so
/usr/bin/ld: cannot open -lnetv2: No such file or directory
collect2: ld returned 1 exit status
make: *** [oci_.so] Error 1
[oracle at ironfist src]$


--
Sam Wun                            Firewalls / Security
OneGuard Senior Software Engineer  Electronic Commerce
eSec Limited                       Phone: +61 3 83715376
mailto:swun at eSec.com.au            C++/JAVA/UNIX/OOP/OOD


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/python-list/attachments/20000705/183b06d8/attachment.html 

From jean.hemmi at arakne.com  Wed Jul  5 10:02:46 2000
From: jean.hemmi at arakne.com (Jean Hemmi)
Date: Wed, 05 Jul 2000 08:02:46 +0000
Subject: JPE: Java - C Python Extension Project
Message-ID: <3962EBA5.E461DD4@arakne.com>

We need a full-featured C-Python / Java integration!

Besides minor functional differences, JPython has the major limitation
of not being able to leverage existing Python (binary) extensions.
Presently, there is no C-Python extension to Java.
We, at Arakne, are annoyed by this situation: We cannot integrate the
Python binary extensions we developped with any of the Java SDK's.

Python is among the fittest integration tools around today - we should
be able to have a seamless integration between Python and Java-.

(I'll leave to others the opportunity to elaborate on the
consequences of the contact between Python's and Java's rationales,
especially if both were tightly bound together...)

This would also open a new perspective: at present the "standard" way to

develop native web browser plug-in is through the Java plug-in mechanism

and Java's JNI C API.
At the end of the pipe is the ability to run, within a client web
browser Python scripts, just like Javascript; Python bytecode, like a
Java applet; or both.
Like binary plug-ins, binary python extensions (*.pyd files) could also
get downloaded and installed.

After a successful experience at integrating two large and complex
object-oriented systems into Python, we feel confident at applying the
same patterns to a Java integration - after all, Java is just yet
another external system to Python... -

We are laying out below the requirements to such an integration:
- Multithread compatibility (at least when the JVM relies on OS threads)

- Access to the Java environment from Python
- Seamlessness from both sides
- Java objects appearing as native Python objects (resolving
overloading)
- Extensibility of Java classes into Python
- Access to the Python environment from Java
- Python objects accessible as Java objects
- Extensibility of Python classes within Java
- Exception mapping
- Python programming on client Web browser, like Javascript
- Ability to run Python bytecode files on the client Web browser
- Conciliation of security features (adapting Python's restricted
execution mode for 'Pylets')

Voila!

Reactions and comments ?

Frederic Giacometti
Arakne
frederic.giacometti at arakne.com






From maxm at normik.dk  Wed Jul  5 10:34:15 2000
From: maxm at normik.dk (=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?=)
Date: Wed, 5 Jul 2000 10:34:15 +0200 
Subject: Again: ODBC / Win98 -> crash
Message-ID: <7BD10B680501D411B9DF009027E06F32011AE8@EXCHANGE>

Fra: Matthias Huening [mailto:matthias.huening at univie.ac.at]

>Now for a really strange thing: I just moved my script to a different
>computer (almost identical configuration), and... - everything works
>just fine! No crashes, no problems whatsoever. I hate computers.

Maybe the machine it crashed on had Office 2K installede recently ??? and
the one that doesn't crash hasn't or something like that. The combination of
Win2k, Office2000/97 has caused me no end of grief.

Try installing the newest version of mdac on your machine. (should not be
needed on win 2K).

Also check your security settings and r/w acces. Combinations of these two
has helped me before.

Regards Max M



From frankn=nuws at cs.vu.nl  Wed Jul  5 08:50:41 2000
From: frankn=nuws at cs.vu.nl (Frank Niessink)
Date: Wed, 5 Jul 2000 06:50:41 GMT
Subject: Entry widgets
References: <39620052.53C1@cs.bham.ac.uk>
Message-ID: <8juls1$roc@cs.vu.nl>

Elliot Smith <e.smith at cs.bham.ac.uk> wrote:
> Hello all,

> I've tried this question before, and got no answer at all, so I thought
> I'd give it another go as the situation is getting desperate! I can't
> find any reference to this anywhere.

> I've been having problems with Entry widgets in Tkinter, running
> under sunos5. For some reason, when I double click on them, Python
> closes the whole application and reports 'Bus error'. Any ideas? I've
> tried overriding the binding for double mouse clicks to no avail.

Well, after your first post I checked your problem with one Tkinter 
application of my one, but I couldn't reproduce your problem. Maybe
you should try to track down the problem to the smallest program that
still has that behavior and then post the code...

Good luck, Frank
-- 
He learnt to communicate with birds and discovered that their conversation was
fantastically boring. It was all to do with wind speed, wing spans, power-to-
weight ratios and a fair bit about berries.
	-- Douglas Adams, 'Life, the Universe, and Everything'


From matthias.huening at univie.ac.at  Wed Jul  5 10:21:56 2000
From: matthias.huening at univie.ac.at (Matthias Huening)
Date: Wed, 5 Jul 2000 10:21:56 +0200
Subject: Undocumented regex behaviour in re module
References: <m3itulmfi1.fsf@vole.object-craft.com.au>
Message-ID: <8jur8r$3eqe$1@www.univie.ac.at>

What about a simple regex and the use of string.split()?

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

tags = re.compile(r"<\?(.+)\?>", re.I | re.M)
match = tags.search(text)
while match:
    print text[match.start():match.end()], \
          '=>', string.split(match.group(1))
    match = tags.search(text, match.end())

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


Matthias

- - - - -
matthias.huening at univie.ac.at
http://www.ned.univie.ac.at/









From stephen at cerebralmaelstrom.com  Wed Jul  5 11:31:24 2000
From: stephen at cerebralmaelstrom.com (Stephen Hansen)
Date: Wed, 05 Jul 2000 09:31:24 GMT
Subject: Acronyms, Names of Projects, etcetera. OTish.
Message-ID: <MfD85.115296$WS3.974982@typhoon.we.rr.com>

I've been off-and-on working on a pet project for awhile now, that
frequently gets started up when I decide I want to learn something new, then
gets put away after I learn said something. And as I learn more, it tends to
get competely revamped in design and implementation.

The only consistant thing that i've had so far is a name 'Pyces', pronounced
Pisces. So i'm all happily, idly, coding in a hobby-ish fashion, and come
across a post on random numbers..and blam!... I see a projected NAMED
Pisces, and..well, obviousely pronounced Pisces. That just won't do. So, I
had to come up with a new name -- and a new acronym, of course.

Behold the wonder that i've come up with, after spending way more time then
I will ever admit to...

Pyramid (I don't care if anyone else has something named this!)

PYthonic Repository for Advanced Missive Interpretation and Distribution

lol. I'm sorry, this is really a pointless message. I just rather proud of
my silly acronym.

Ten cents to the person who guesses what it is :) (Not that it's that hard.)

--Stephen




From chwieser at ee.oulu.fi  Wed Jul  5 11:33:52 2000
From: chwieser at ee.oulu.fi (Christian Wieser)
Date: 5 Jul 2000 09:33:52 GMT
Subject: py-libcap
Message-ID: <8juve0$n1o$2@ousrvr3.oulu.fi>

-- 

Morning,

I am in need of py-libcap. Unfortunally the link at Pamasus is down ;-(

  http://www.vex.net/parnassus/apyllo.py/126307487.528281449

Can someone please send me the file? 
And maybe an example?

By the way: Is there a wrapper for libnet to inject IP packets too available?

Regards,


--
Christian Wieser - Oulu Secure Programming Group
mailto: chwieser at ee.oulu.fi

See my homepage at http://www.ee.oulu.fi/~chwieser for my PGP-Public-Key



From shapr at uab.edu  Wed Jul  5 11:59:56 2000
From: shapr at uab.edu (Shae Erisson)
Date: Wed, 05 Jul 2000 09:59:56 GMT
Subject: GUI framework with TreeView
References: <3967f81e.23966859@nntp.interaccess.com>
Message-ID: <3963091F.CC541AF@uab.edu>

Thaddeus L Olczyk wrote:
> 
> I need a "free" gui framework that runs on both Windows and UNIX
> ( so that lets out QT ) and gives me a component similr to Microsofts
> TreeView. Does anyone know of such a thing?

There's a Tree widget in IDLE that's quite nice looking.
-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius fOr hIre


From jessaustin at SPAMFREE.lucent.com  Wed Jul  5 11:48:20 2000
From: jessaustin at SPAMFREE.lucent.com (Jess Austin)
Date: Wed, 5 Jul 2000 17:48:20 +0800
Subject: Python COM for Linux?
References: <8ja7gv$687$1@nnrp1.deja.com> <3958c185.103629691@news> <Za465.1871$iN5.400266@ptah.visi.com>
Message-ID: <8juulj$csu$1@pony.kenan.com>

Grant Edwards wrote in message ...
>
>I have a Python program that fetches mail from MS Exchange
>Server using MAPI/COM.  I'd dearly love to run it under Linux.
>Does anybody have an experience with accessing MS Exchange
>Server via COM from Linux?

I'd love to take a look at this.  I'm always interested in anything that
could help me escape the drudgery of Outlook.  Can you post a link?

thanks,
Jess




From sean at digitome.com  Wed Jul  5 13:01:30 2000
From: sean at digitome.com (Sean Mc Grath)
Date: Wed, 05 Jul 2000 11:01:30 GMT
Subject: __getattr__ and pickle
References: <395f26a4.1832855@news.iol.ie>
Message-ID: <39630fdb.148487854@news.iol.ie>

Thanks to Christian and Gordon, I think I am well on
the way to pickling Pyxie trees.

More news anon.

regards,
Sean McGrath

On Sun, 02 Jul 2000 11:51:10 GMT, sean at digitome.com (Sean Mc Grath)
wrote:

>I have a reproduceable GPF on NT when picking objects of a class
>that has a __getattr__ implementation. Take out the __getattr__ and
>the pickle works fine.
>
>Are there ways to implement __getattr__ and yet preserve the ability
>to pickle?
>
>regards,
>



From aa8vb at yahoo.com  Wed Jul  5 13:18:27 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 05 Jul 2000 07:18:27 -0400
Subject: Nth digit of PI
In-Reply-To: <t7545.239$ZR1.2944@news.uswest.net>; from
 jmassung@magpiesystems.com on Wed, Jun 21, 2000 at 09:19:17AM -0600
References: <bivnjsoq0amb8qjmqtgkbfjsl2df2n2tta@4ax.com> <FvuAy4.51q@cwi.nl>
 <8hogt2$57t$1@nnrp1.deja.com> <8houlu$j9s$1@nntp.itservices.ubc.ca>
 <394889C3.1ED5@esatclear.ie> <t7545.239$ZR1.2944@news.uswest.net>
Message-ID: <20000705071827.A1015156@vislab.epa.gov>

Jeff Massung:
 |Please, this interests me -- I've had differention equations, but that's as
 |far as I go. In theory, pi (and e) are infinite, but has that been proven
 |that there is not end to their decimals?

I don't know the mathematically-rigorous answer to your question.  But one
of the interesting things that you can show with a Fourier integral
(possibly other ways too) is that:

    Pi^2/8 = sum(n=1..inf, 1/(2n-1)^2)

You can see that the sum is 1/1 + 1/9 + 1/125 + ....  The terms get smaller
and smaller, which means making the decimal representation longer and
longer.  In the limit, the decimal representation is infinitely long
because the denominator is infinitely large.

Solving for pi:  multiplying by 8, you still have infinitely many digits,
and taking the square root, you still have infinitely many digits.  That's pi.

Randall

------------------------------------------------------------------------------
#!/usr/bin/env python
#
#  pi = sqrt( 8 * sum(n=1..inf, 1/(2n-1)^2) )
#    Not a realistic way to compute pi; convergence is hideously slow.
#    But with rational number arithmetic and lots of storage...
#

import math

ITERATIONS = 50000

sum = 0
for v in range(2*ITERATIONS-1,0,-2):
  sum = sum + 1/v**2.0

print "pi is approx %g (%d iterations)" % ( math.sqrt( sum * 8 ), ITERATIONS )



From scherbi at bam.com  Wed Jul  5 13:27:51 2000
From: scherbi at bam.com (Bill Scherer)
Date: Wed, 05 Jul 2000 11:27:51 +0000
Subject: DCOracle error
References: <3962CB94.2DDA9AA9@esec.com.au>
Message-ID: <39631BB7.3641E234@bam.com>

Getting those Oracle lib lines correct is a bear.  Below is the line
from my makefile for DCOracle, built against Oracle 8.0.5.

ORACLE_L=-lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork
-lclient -lvsn -lcommon -lgeneric -lmm -lnlsrtl3 -lcore4 -lnlsrtl3
-lcore4 -lnlsrtl3 -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp
-lnetwork -lclient -lvsn -lcommon -lgeneric
-lepc -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient -lvsn
-lcommon -lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4
-lnlsrtl3 -lnsl -lgeneric -ldl -lelf -lposix -lsql -lm -lm

(that's one line)

You'll notice that many of the entries from the stock Makefile ORACLE_L
line are missing.

Now, I have libnetv2.a in ORACLE_HOME/lib.  Possibly you neglected to
install some of the Oracle development libs?

 - Bill Scherer

sss wrote:

> I don't know why I got this error when I try to install DCOracle in
> the Linux machine.
> I've got Oracle sqlplus: SQL*Plus: Release 8.1.5.0.0
>
> Here is the err msg:
>
> [oracle at ironfist src]$ make
> gcc -fPIC -g -O2 -I/usr/include/python1.5 -I/usr/include/python1.5
> -DHAVE_CONFIG_H  -I/home/oracle/product/8.15/rdbms/demo
> -I/home/oracle/product/8.15/rdbms/public
> -I/home/oracle/product/8.15/network/public
> -I/home/oracle/product/8.15/plsql/public -DDCORACLE8 -c ./oci_.c
> In file included from ./oci_.c:562:
> Buffer.h:79: warning: static declaration for `PyImport_Import' follows
> non-static
> oci_8.c: In function `Logon':
> In file included from ./oci_.c:693:
> oci_8.c:95: warning: passing arg 4 of `OCIHandleAlloc' makes integer
> from pointer without a cast
> oci_8.c: In function `newLob':
> oci_8.c:446: warning: passing arg 4 of `OCIDescriptorAlloc' makes
> integer from pointer without a cast
> oci_8.c: In function `initoci_8':
> oci_8.c:483: warning: passing arg 3 of `OCIEnvInit' makes integer from
> pointer without a cast
> ./oci_.c: In function `_wrap_odescr':
> ./oci_.c:2595: warning: assignment from incompatible pointer type
> gcc -shared  oci_.o  -L/home/oracle/product/8.15/lib/
> -L/home/oracle/product/8.15/rdbms/lib
> /home/oracle/product/8.15/rdbms/lib/defopt.o
> /home/oracle/product/8.15/lib/sscoreed.o
> /home/oracle/product/8.15/lib/nautab.o
> /home/oracle/product/8.15/lib/naeet.o
> /home/oracle/product/8.15/lib/naect.o
> /home/oracle/product/8.15/lib/naedhs.o -lnetv2 -lnttcp -lnetwork -lncr
> -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lmm
> -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lnetv2 -lnttcp
> -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon
> -lgeneric -lepc -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient
> -lvsn -lcommon -lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3
> -lsocket -lnsl -lgen -ldl -lelf -laio -lposix4 -lsql -lm -lm -lnetv2
> -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn
> -lcommon -lgeneric -lmm -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3
> -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient
> -lvsn -lcommon -lgeneric -lepc -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4
> -lnlsrtl3 -lclient -lvsn -lcommon -lgeneric -lnlsrtl3 -lcore4
> -lnlsrtl3 -lcore4 -lnlsrtl3 -lsocket -lnsl -lgen -ldl -lelf -laio
> -lposix4 -lsql -lm -lm -o oci_.so
> /usr/bin/ld: cannot open -lnetv2: No such file or directory
> collect2: ld returned 1 exit status
> make: *** [oci_.so] Error 1
> [oracle at ironfist src]$
>
>
> --
> Sam Wun                            Firewalls / Security
> OneGuard Senior Software Engineer  Electronic Commerce
> eSec Limited                       Phone: +61 3 83715376
> mailto:swun at eSec.com.au            C++/JAVA/UNIX/OOP/OOD
>
>

--
William K. Scherer
Sr. Member of Applications Staff
Verizon Wireless






From roy at popmail.med.nyu.edu  Wed Jul  5 13:53:47 2000
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Wed, 05 Jul 2000 07:53:47 -0400
Subject: Nth digit of PI
References: <bivnjsoq0amb8qjmqtgkbfjsl2df2n2tta@4ax.com> <FvuAy4.51q@cwi.nl> <8hogt2$57t$1@nnrp1.deja.com> <8houlu$j9s$1@nntp.itservices.ubc.ca> <394889C3.1ED5@esatclear.ie> <t7545.239$ZR1.2944@news.uswest.net> <20000705071827.A1015156@vislab.epa.gov>
Message-ID: <roy-96FA97.07534705072000@netnews.nyu.edu>

Randall Hopper <aa8vb at yahoo.com> wrote:
>  Pi^2/8 = sum(n=1..inf, 1/(2n-1)^2)

Interesting, I've never seen that one before.

It's not a very efficient way to compute pi (I know, that's not what you were 
trying to do).  A little experimentation shows that 1000 terms only gets you 4 
significant digits correct, and 10,000 terms still only had 5!

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



From olivierS.dagenaisP at canadaA.comM  Wed Jul  5 14:33:02 2000
From: olivierS.dagenaisP at canadaA.comM (Olivier Dagenais)
Date: Wed, 05 Jul 2000 12:33:02 GMT
Subject: GUI framework with TreeView
References: <3967f81e.23966859@nntp.interaccess.com>
Message-ID: <2WF85.39122$W35.856292@news20.bellglobal.com>

I was looking for EXACTLY that and I found it, along with a whole lot of
other cool stuff:
http://www.wxwindows.org
and if you're using Python:
http://www.wxpython.org

There's also a MAC port coming along.

It attempts to provide many nice features of the OS GUIs out there, not just
a "lowest common denominator".  It does so by using native OS
components/APIs wherever possible, so the Treeview isn't some
re-implementation, it looks and feels just like MS's treeview.

Oh, and it's free:  http://wxwindows.org/newlicen.htm

--
----------------------------------------------------------------------
Olivier A. Dagenais - Carleton University - Computer Science III


"Thaddeus L Olczyk" <olczyk at interaccess.com> wrote in message
news:3967f81e.23966859 at nntp.interaccess.com...
> I need a "free" gui framework that runs on both Windows and UNIX
> ( so that lets out QT ) and gives me a component similr to Microsofts
> TreeView. Does anyone know of such a thing?




From peter at wl.vg  Wed Jul  5 14:32:59 2000
From: peter at wl.vg (Peter Halliday)
Date: Wed, 05 Jul 2000 08:32:59 -0400
Subject: Redhat and mod_python
Message-ID: <39632AFB.8BB36365@wl.vg>

I am trying to get mod_python working with Apache's rpm's.  I really
don't want to have to compile Apache separately.  Has anyone got
mod_python working under Redhat while keeping the Apache rpms.  Thank
you.  And please if possible could you email your response to my email
above (peter at whetstonelogic.com).  Thanks.


-- 
Peter Halliday
Online Application Developer
Whetstone Logic, Inc.       http://www.whetstonelogic.com
icq#75769411           AIM screen name: hoaggelos




From thiele at muc.das-werk.de  Wed Jul  5 14:29:32 2000
From: thiele at muc.das-werk.de (Thomas Thiele)
Date: Wed, 05 Jul 2000 14:29:32 +0200
Subject: Segmentation fault(LINUX) and no idea why!
Message-ID: <39632A2C.E6ABECEB@muc.das-werk.de>

Hallo!

In my program I create a new thread. In this new thread I fork and
replace the child with an other program
using excel. Normaly I have up to four threads with this procedure. So I
have a number of threads and the same numbers of childs (the called
program)

The problem is sometimes (relativly seldom, so it's unreproducible ;-( )
I get a segmantation fault and the pytthon interpreter stops.

But I if the segmentation fault happens, one of the childs is not
running anymore. (I don't know is that the reason or a  result)
A little testprogram with the same construction shows that if a
segmentaltion fault happens in one child, the parant keeps running.
I get no traceback from the interpreter.

My Question is, where can I look for the problem. Why does the
interpreter (-> the whole program) die.  What can be the reason?



From sholden at bellatlantic.net  Wed Jul  5 14:57:52 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Wed, 05 Jul 2000 12:57:52 GMT
Subject: Is there a searchable archive of c.l.py
References: <Gj185.6536$Tb7.39716@news-server.bigpond.net.au> <8jq7l3$f1e$1@slb0.atl.mindspring.net>
Message-ID: <396330BC.DB5DCD00@bellatlantic.net>

Deja have "promised" to get the older archives back on line
when their new face is up and running.  Does anyone have any
personal contacts in that organization to ask if we can have
the data NOW to put a searchable archive up somewhere else?

regards
 Steve

Aahz Maruch wrote:
> 
> In article <Gj185.6536$Tb7.39716 at news-server.bigpond.net.au>,
> Neil Hodgson <neilh at scintilla.org> wrote:
> >
> >   Is there a sensibly searchable archive anywhere? In the future we will
> >need this resource. Otherwise how will we know how we beat those nanoviruses
> >back in '97?
> 
> Well, we've got the archive: http://www.python.org/pipermail/python-list/
> 
> Unfortunately, it only seems to go back to March '99.  Does anyone
> maintain an older archive?  Once we have the archive, making it
> searchable is a relatively trivial task.
> --
>                       --- Aahz (Copyright 2000 by aahz at netcom.com)
> 
> Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
> 
> The best way to get information on Usenet is not to ask a question,
> but to post the wrong information.  --Aahz

-- 
Helping people meet their information needs with training and technology.
703 967 0887   sholden at bellatlantic.net   http://home.cox.rr.com/sholden/


From brunnsNObrSPAM at beer.com.invalid  Wed Jul  5 15:01:13 2000
From: brunnsNObrSPAM at beer.com.invalid (Simon B)
Date: Wed, 05 Jul 2000 06:01:13 -0700
Subject: Is there a searchable archive of c.l.py
References: <Gj185.6536$Tb7.39716@news-server.bigpond.net.au>
Message-ID: <111b2f38.c4458b92@usw-ex0106-045.remarq.com>

I don't know just how far back RemarQ goes, but it's probably
worth a look.

Cheers,
Simon B.
With Microsoft, failure is not an option. It comes bundled.


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

Got questions?  Get answers over the phone at Keen.com.
Up to 100 minutes free!
http://www.keen.com



From Ted.Mew at ncmail.net  Wed Jul  5 15:11:36 2000
From: Ted.Mew at ncmail.net (Ted Mew)
Date: Wed, 05 Jul 2000 09:11:36 -0400
Subject: Compatibility between PythonWin and DISLIN
Message-ID: <39633408.80B5DFBC@ncmail.net>

Has anyone tried to run the graphing package DISLIN using
Mark Hammond's PythonWin 1.5.2?  Is DISLIN compatible with
PythonWin?

DISLIN 7.3 is a high-level and easy to use graphics library
with a Python version that runs nicely on a Windows NT
platform under Python 1.5.2.  This data plotting extension
was developed by Helmut Michels of the Max_Planck
Institute.  It is listed on the Starship at:

http://starship.python.net/crew/jhauser/plot-res.html

and the binaries can be obtained freely at:

http://www.linmpi.mpg.de/dislin/

I have made a cursory attempt to run DISLIN under PythonWin,
but crashed in flames, unable to read memory locations.
Since PythonWin is MS Windows based, I suspect DISLIN is
attempting to run as an independent process or DOS-based
program, but have little knowledge in this area.

Any insights or experience would be appreciated.

 -Ted
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ted.Mew.vcf
Type: text/x-vcard
Size: 384 bytes
Desc: Card for Ted Mew
Url : http://mail.python.org/pipermail/python-list/attachments/20000705/c0e6e170/attachment.vcf 

From sholden at bellatlantic.net  Wed Jul  5 15:08:02 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Wed, 05 Jul 2000 13:08:02 GMT
Subject: Definative Documentation on Objects
References: <8jqui2$9na$1@supernews.com>
Message-ID: <3963331E.D1F756BE@bellatlantic.net>

I found "Learning Python" to be quite good enough to allow me to transfer
the object-oriented stuff I knew to implementations in Python.  It's
good value for money, and although I personally found the pizza-robot
examples a little unnatural, there is evidence they work for others.

regards
 Steve

Dale Strickland-Clark wrote:
> 
> Where do I look for the definitive statement on object programming in
> Python?
> 
> The "Python Reference Manual", as included in the distribution, is
> particularly hard work.
> 
> I'm considering "Programming Python" by Mark Lutz but that's four years old.
> "Learning Python" is more recent but I'm worried that it might be pitched a
> bit low.
> 
> I'm an experienced programmer but I need to quickly understand Python object
> programming.
> 
> Any pointers to online or paper reference material gratefully received.
> 
> --
> Dale Strickland-Clark
> Out-Think Ltd, UK
> Business Technology Consultants

-- 
Helping people meet their information needs with training and technology.
703 967 0887   sholden at bellatlantic.net   http://home.cox.rr.com/sholden/


From claird at starbase.neosoft.com  Wed Jul  5 15:09:47 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 5 Jul 2000 08:09:47 -0500
Subject: Hi, my name is not Mrs so-called Gumblenose
References: <395B2B88.CC3041C1@memo.volvo.se>
Message-ID: <FA98C0EEF85667EF.F9FA99C0C74A44FE.F449656A33278DD4@lp.airnews.net>

In article <395B2B88.CC3041C1 at memo.volvo.se>,
PM  <vtbl.moorepe at memo.volvo.se> wrote:
			.
			.
			.
>Can Python kick large bottoms with decent database access - not just SQL
>not ODBC, could it go to DB2/400?
			.
			.
			.
How are you getting to DB2 over OS/400 if not through
SQL?  Are you asking for RPG?

In general, yes; Python's database access is "decent",
and the language has turned up on AS/400 in the past
(although I can't attest to its "currency").
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From fiona at sitegnome.com  Wed Jul  5 15:14:23 2000
From: fiona at sitegnome.com (Fiona Czuczman)
Date: 5 Jul 2000 13:14:23 -0000
Subject: [FAQTS] Python Knowledge Base Update -- July 5th, 2000
Message-ID: <20000705131423.10879.qmail@synop.com>

Hi !

Another instalment of entries into http://python.faqts.com

cheers,

Fiona

Australian - instalment
American - installment

:-) .. in case you were wondering that I couldn't spell?


## Unanswered Questions ########################################


-------------------------------------------------------------
How do I format my request to get an url/socket when there is a proxy in the middle and Proxy Authentication is required ?
http://www.faqts.com/knowledge-base/view.phtml/aid/4182
-------------------------------------------------------------
Guest



## New Entries #################################################


-------------------------------------------------------------
How do I return a user's login name who is logged on at a PC?
http://www.faqts.com/knowledge-base/view.phtml/aid/4189
-------------------------------------------------------------
Fiona Czuczman
richard_chamberlain, j vickroy, Ralf Claus

You'll need the win32 extensions - 
http://starship.python.net/crew/mhammond/

import win32api
login=win32api.GetUserName()

--- or ---

On Windows NT:

import os
os.environ['USERNAME']

... provides the information.

I do not believe that either Win 95 or 98 has the 'USERNAME' attribute 
set.

--- or ---

import getpass
username = getpass.getuser()

works on both systems (NT,UNIX), if you need.


-------------------------------------------------------------
Can I use Python to run Internet Explorer and control it via COM?
http://www.faqts.com/knowledge-base/view.phtml/aid/4191
-------------------------------------------------------------
Fiona Czuczman
Roger Upole, Mark Hammond

Here's a small sample of how to navigate to a URL.
>>> import win32com.client
>>> x=win32com.client.Dispatch('InternetExplorer.Application.1')
>>> x.Visible=1
>>> x.Navigate('http://www.python.org')
If you run makepy on Microsoft Internet Controls, the file generated in
Python\win32com\genpy will show all the properties and methods available 
for this object.

This will work fine to control IE.

However, look in "win32com\client\__init__.py", and look for the
"DispatchWithEvents" docstring - this has an example that uses IE to 
both control it, and respond to events (such as URL load completed, etc)


-------------------------------------------------------------
Is there a Python MySQL module for use under Windows?
http://www.faqts.com/knowledge-base/view.phtml/aid/4188
-------------------------------------------------------------
Fiona Czuczman
Stefan Franke

Check out

  ZMySQLDA-1.2.0.tar.gz

on the Zope site (www.zope.org). It contains MySQLdb-0.1.2 with some 
Windows patches already applied. I run it sucessfully with the lately 
GPLed MySQL-2.23.19 on Win98.

All you need to do is to edit Setup.in, adjust the paths to your  local 
MySQL installation, and run compile.py.


## Edited Entries ##############################################


-------------------------------------------------------------
What is the best way to make multiple replacements of text strings in a file?
http://www.faqts.com/knowledge-base/view.phtml/aid/4049
-------------------------------------------------------------
Fiona Czuczman, Thi?baut Champenier, Steve Holden
Steve Holden, Steve Nordby,David Goodger, Fredrik Lundh,Thi?baut Champenier

You could clean the source up a bit with:

     foobar = ( ('foo1', 'bar1'),
              ('foo2', 'bar2'),
              ('fooN', 'barN') )
     source = open(source_file,'r')
     contents = source.read()
     source.close()

     for foo, bar in foobar:
         contents = replace(contents, foo, bar)
           
     dest = open(dest_file, 'w')
     dest.write(contents10)
     dest.close()

For speedup, you could write the whole loop as a single statement,
but it will get horrible quickly:

    contents = replace(
               replace(
               replace(contents,
               'fooN', 'barN'),
               'foo2', 'bar2'),
               'foo1', 'bar1)

and, of course, ths code is much less easy to maintain.

How about using a dictionary and a loop:

    replaceme = {'foo1': 'bar1', 'foo2': 'bar2', 'foo3': 'bar3'}
    for key in replaceme.keys():
        srch = key
        rplc = replaceme[key]
        contents = string.replace(contents, srch, rplc)

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

Perhaps a bit complex for a newbie, but the most definitive answer I've 
seen so far has been Fredrik Lundh's 2000-04-12 reply to the "Python 
idiom: Multiple search-and-replace" thread:

> Is there a Python feature or standard library API that will get me 
> less Python code spinning inside this loop?   re.multisub or 
> equivalent? 

haven't benchmarked it, but I suspect that this approach is more 
efficient:

...

# based on re-example-5.py

import re
import string

symbol_map = { "foo": "FOO", "bar": "BAR" }

def symbol_replace(match, get=symbol_map.get):
    return get(match.group(1), "")

symbol_pattern = re.compile(
    "(" + string.join(map(re.escape, symbol_map.keys()), "|") + ")"
              )

print symbol_pattern.sub(symbol_replace, "foobarfiebarfoo")

-------------------
Thi?baut Champenier has wriiten a string replacement utility,
which is available for inspection at:  http://tibi.free.fr/repl.py






From piet at cs.uu.nl  Wed Jul  5 15:12:49 2000
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 05 Jul 2000 15:12:49 +0200
Subject: Win32 : Making script.py executable in DOS box
References: <8jmjfe$64o$1@nnrp1.deja.com> <WzI75.5714$Tb7.31718@news-server.bigpond.net.au> <8jp3pt$q7j$1@nnrp1.deja.com> <87og4eugz7.fsf@cartman.azz.net>
Message-ID: <uvgykpvou.fsf@cs.uu.nl>

Note that the trick via a .bat file may cause problems with parameters
containing = signs, and with I/O redirections.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From neilh at scintilla.org  Wed Jul  5 15:22:24 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Wed, 05 Jul 2000 13:22:24 GMT
Subject: Is there a searchable archive of c.l.py
References: <Gj185.6536$Tb7.39716@news-server.bigpond.net.au> <111b2f38.c4458b92@usw-ex0106-045.remarq.com>
Message-ID: <kEG85.8430$Tb7.53383@news-server.bigpond.net.au>

> I don't know just how far back RemarQ goes, but it's probably
> worth a look.

   Thanks for the pointer.

   Remarq only goes back to April with a couple of days in March. I like the
way it displays all of a thread if you click on it though. Lots less
clicking than Deja.

   Neil





From sholden at bellatlantic.net  Wed Jul  5 15:22:37 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Wed, 05 Jul 2000 13:22:37 GMT
Subject: Postscript font metrics?
References: <39622330.3732B225@yahoo.com>
Message-ID: <3963368A.8442AD75@bellatlantic.net>

Paul Winkler wrote:
> 
> I need to generate some nice Postscript output from a python script,
> and I'm finding that I need to get at the font metrics info from the
> .afm files so I can place things correctly. (Either that or write
> *everything* in postscript -- no way!!)
> 
> I can't find anything like this on Parnassus. Anyone ever seen any
> python code that gets the useful info from .afm files?
> 
> If not, I might have a go at hacking the relevant parts of the Perl
> PostScript-Font-1.02 module into Python... but then I'll have to
> (yuck) read someone else's Perl code, which I haven't done in a
> looong time and I was never that good at it.
> 
> --
> .................    paul winkler    ..................
> slinkP arts:   music, sound, illustration, design, etc.
>         web page:  http://www.ulster.net/~abigoo
>       A member of ARMS:   http://www.reacharms.com
>                or http://www.mp3.com/arms


Have you considered using PIDDLE for your PostScript output?

I recently used it to generate PDF files, but PostScript is another
possible backend.  It's impressive, and will compute string lengths
for you to allow "typesetting"-style outputs.  Take a look at

	http://piddle.sourceforge.net

Hope this helps.

regards
 Steve

-- 
Helping people meet their information needs with training and technology.
703 967 0887   sholden at bellatlantic.net   http://home.cox.rr.com/sholden/


From cmfinlay at SPAMmagnet.com.au  Wed Jul  5 15:21:03 2000
From: cmfinlay at SPAMmagnet.com.au (cmfinlay at SPAMmagnet.com.au)
Date: Wed, 05 Jul 2000 23:21:03 +1000
Subject: "Write to a file" question please?
Message-ID: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com>

number = 100
print "Number =", number
OUT = open("TT.txt","w")
OUT.write("Number =", number) # Error
OUT.close()

The code above gives the error.

Traceback (innermost last):
  File "ttest.py", line 5, in ?
    OUT.write("Number =", number)
TypeError: read-only buffer, tuple

OUT.write("Hello World")

Works but a number or variable of a number does not.

I have been reading documents on Python to no avail this time.
I am using Win98 with the latest python for this OS.

Any help would be much appreciated
E-mail me at cmfinlaySPAM at magnet.com.au
or reply here.


From brunnsNObrSPAM at beer.com.invalid  Wed Jul  5 15:36:27 2000
From: brunnsNObrSPAM at beer.com.invalid (Simon B)
Date: Wed, 05 Jul 2000 06:36:27 -0700
Subject: Hi, my name is not Mrs so-called Gumblenose
References: <395B2B88.CC3041C1@memo.volvo.se>
Message-ID: <1c934748.cd75dee0@usw-ex0106-045.remarq.com>

I've not heard of any way of getting at DB2/400 from Python
other than via ODBC. But I'd love to be wrong!

If someone were to port JPython to the '400, you would be able
to use the AS/400 Toolbox for Java's native DB access APIs, but
that would be a pretty major job in itself.

BTW, It's nice to know that there are other '400 types looking
at Python. I'm hoping to do the JPython port eventually, but I
don't know Java yet, so don't hold your breath!

Cheers,
Simon B.
With Microsoft, failure is not an option. It comes bundled.


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

Got questions?  Get answers over the phone at Keen.com.
Up to 100 minutes free!
http://www.keen.com



From olivierS.dagenaisP at canadaA.comM  Wed Jul  5 15:43:16 2000
From: olivierS.dagenaisP at canadaA.comM (Olivier Dagenais)
Date: Wed, 05 Jul 2000 13:43:16 GMT
Subject: "Write to a file" question please?
References: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com>
Message-ID: <UXG85.39291$W35.861333@news20.bellglobal.com>

Make this adjustement:

OUT.write ("Number = %d" % number)

I just tried it, and it works.  My guess is the rules for write are
different than for print, so what appears to be a string concatenation in
print might just be a sequence type of some sort.  (sorry, I learned Python
this weekend, that's the best I can do in explaining it)

You can also use:

OUT.write ("Number =" + str(number))

...which might be more what you're looking for.

--
----------------------------------------------------------------------
Olivier A. Dagenais - Carleton University - Computer Science III


<cmfinlay at SPAMmagnet.com.au> wrote in message
news:tzFjOfN2zOzPHwvl3ncklEAEb+fa at 4ax.com...
> number = 100
> print "Number =", number
> OUT = open("TT.txt","w")
> OUT.write("Number =", number) # Error
> OUT.close()
>
> The code above gives the error.
>
> Traceback (innermost last):
>   File "ttest.py", line 5, in ?
>     OUT.write("Number =", number)
> TypeError: read-only buffer, tuple
>
> OUT.write("Hello World")
>
> Works but a number or variable of a number does not.
>
> I have been reading documents on Python to no avail this time.
> I am using Win98 with the latest python for this OS.
>
> Any help would be much appreciated
> E-mail me at cmfinlaySPAM at magnet.com.au
> or reply here.




From mjm at boora.com  Wed Jul  5 15:32:09 2000
From: mjm at boora.com (mjm)
Date: Wed, 5 Jul 2000 06:32:09 -0700
Subject: Python Object Browser
References: <dU785.94271$5k2.195629@dfw-read.news.verio.net>
Message-ID: <sm6ev4iegas91@corp.supernews.com>

Another clue... I don't have the same problem on my win98 box at home.  At
work, on WinNT, I have a problem.

-mjm

michael montagne <montagne at boora.com> wrote in message
news:dU785.94271$5k2.195629 at dfw-read.news.verio.net...
> I'm trying to use makepy on some common objects on my system.  Word2000,
> Excel2000, etc.
> It doesn't seem to want ot work.  When I go to the objects that are having
> trouble using the Python Object Browser, under the Type Library heading it
> says
> "The type info can n..="The type info can not be loaded".  Does this mean
it
> is unavailable?  How do i fix this?
>
> -mjm
>
>




From tismer at appliedbiometrics.com  Wed Jul  5 15:52:00 2000
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Wed, 05 Jul 2000 15:52:00 +0200
Subject: Why is Python so slow ?- revisited.
References: <H_c35.12336$Xx5.574993@bgtnsc06-news.ops.worldnet.att.net> <394D6579.8B2FF840@appliedbiometrics.com> <8tA35.10459$C44.626417@bgtnsc05-news.ops.worldnet.att.net>
Message-ID: <39633D80.27B767E4@appliedbiometrics.com>


William Dandreta wrote:
> 
> Hi chris,
> 
> Here is the fastest version of my program.

Ok. I was off for a while and did not follow the thread.

Some hints may give you a little speedup:

Don't work on the global level, but do everything in
functions. This gives you a speedup, since local variables
are optimized and don't involve dictionary lookups.

Further, since you are working on only 6000 lines or so,
you can save quite some time by using readlines(),and
then iterating over that list. But I don't know if Py1.2
has readlines at all.

Here a rough layout, untested of course!

### start of example refactored
#corrects Y2K leapyear error and
#strips off all the " marks

import strop

FINNAME  = 'lu4.txt'
FOUTNAME = 'lu4.new'

def process(finn=FINNAME, foutn=FOUTNAME):
    # caching some funcitons locally:
    strip = strop.strip
    splitfields = strop.splitfields
    atoi = strop.atoi
    joinfields = strop.joinfields
    _map = map
    _tuple = tuple

    lines = open(finn, 'r').readlines()
    fout = open(fount, 'wb)
    write = fout.write

    for line in lines:
        fields = splitfields(joinfields(splitfields(line,'"'),''),',')
        fields[2] = joinfields(splitfields(fields[2],' '),'0')
        fields = _map(strip,fields)
        #convert date to YYYYMMDD format
        if fields[2][-2:] == '00':
            fields[2] = '2000' + fields[2][-10:-8] + fields[2][-7:-5]
        else:
            fields[2] = '19' + fields[2][-2:] + fields[2][-10:-8] +
fields[2][-7:-5]

        #fix date for Y2K leapyear error
        if   fields[2] == '20000301':  fields[2] = '20000229'
        elif fields[2] == '20000401':  fields[2] = '20000331'
        elif '20000301'<fields[2] <'20000501':
            fields[2] = '2000' + fields[2][4:6] + '%02d' %
(atoi(fields[2][6:])-1)

        write('%-16s,%8s,%8s,%10s,%10s,%-11s,%-11s,%-21s,%-51s\r\n' %
_tuple(fields))

if __name__ == '__main__':
    process()  # add cmd line handling here

### end

But I'm not confident whether the speed gain would justify
changing it at all. Well, maybe it is some percentage.

cheers - chris

> ----------------------------------------------------------
> #corrects Y2K leapyear error and
> #strips off all the " marks
> 
> from strop import strip, splitfields,atoi,joinfields
> 
> lu4 = open('lu4.txt','r')
> newlu4 = open('lu4.new','wb')
> 
> while 1:
> #for i in range(6000):
>   line = lu4.readline()
>   if not line: break
>   fields = splitfields(joinfields(splitfields(line,'"'),''),',')
>   fields[2] = joinfields(splitfields(fields[2],' '),'0')
>   fields = map(strip,fields)
> #convert date to YYYYMMDD format
>   if fields[2][-2:] == '00':
>     fields[2] = '2000' + fields[2][-10:-8] + fields[2][-7:-5]
>   else:
>     fields[2] = '19' + fields[2][-2:] + fields[2][-10:-8] + fields[2][-7:-5]
> 
> #fix date for Y2K leapyear error
>   if fields[2] == '20000301':
>     fields[2] = '20000229'
>   elif fields[2] == '20000401':
>     fields[2] = '20000331'
>   elif '20000301'<fields[2] <'20000501':
>     fields[2] = '2000' + fields[2][4:6] + '%02d' % (atoi(fields[2][6:])-1)
> 
>   newlu4.write('%-16s,%8s,%8s,%10s,%10s,%-11s,%-11s,%-21s,%-51s\r\n' %
> tuple(fields))
> ---------------------------------------------------------------------
> Bill
> 
> Christian Tismer wrote in message
> <394D6579.8B2FF840 at appliedbiometrics.com>...
> >Also feel free to send me your source code for inspection.
> 
> --
> http://www.python.org/mailman/listinfo/python-list

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From adetalabi at clara.co.uk  Wed Jul  5 16:01:08 2000
From: adetalabi at clara.co.uk (Ade Talabi)
Date: Wed, 05 Jul 2000 15:01:08 +0100
Subject: Segmentation fault(LINUX) and no idea why!
References: <39632A2C.E6ABECEB@muc.das-werk.de>
Message-ID: <39633FA4.8EBBE2DB@clara.co.uk>

Thomas Thiele wrote:
> 
> Hallo!
> 
> In my program I create a new thread. In this new thread I fork and
> replace the child with an other program
> using excel. Normaly I have up to four threads with this procedure. So I
> have a number of threads and the same numbers of childs (the called
> program)
> 
> The problem is sometimes (relativly seldom, so it's unreproducible ;-( )
> I get a segmantation fault and the pytthon interpreter stops.
> 
> But I if the segmentation fault happens, one of the childs is not
> running anymore. (I don't know is that the reason or a  result)
> A little testprogram with the same construction shows that if a
> segmentaltion fault happens in one child, the parant keeps running.
> I get no traceback from the interpreter.
> 
> My Question is, where can I look for the problem. Why does the
> interpreter (-> the whole program) die.  What can be the reason?


pstack, ptrace or truss  the PID
-- 
/--------------------------------------------------------------\
| Ade  Talabi     | Internet : adetalabi at clara.co.uk           |
|  ::M1ETW::      | Web Page : http://www.net-africa.com       |
\--------------------------------------------------------------/


From wlav at hpatl26.cern.ch  Wed Jul  5 15:43:38 2000
From: wlav at hpatl26.cern.ch (Wim Lavrijsen)
Date: 5 Jul 2000 13:43:38 GMT
Subject: "Write to a file" question please?
References: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com>
Message-ID: <8jve2a$hg3$1@sunnews.cern.ch>

cmfinlay at SPAMmagnet.com.au writes:

>number = 100
>print "Number =", number
>OUT = open("TT.txt","w")
>OUT.write("Number =", number) # Error

'write' expects only 1 argument. You should build a string to
hand to write. Something like:

OUT.write( 'Number = %d' % number + '\n' ) 

Best regards,
      Wim Lavrijsen


From alex at magenta.com  Wed Jul  5 13:14:41 2000
From: alex at magenta.com (Alex Martelli)
Date: Wed, 5 Jul 2000 13:14:41 +0200
Subject: Singleton COM local/remote server?
References: <7bd5msssk48km280t9bvds4i3v6hs7r8or@4ax.com>
Message-ID: <8jvad40s00@news1.newsguy.com>

John Lull <lull at acm.org> wrote in message
news:7bd5msssk48km280t9bvds4i3v6hs7r8or at 4ax.com...
> I'm prototyping a DCOM server in Python.  The server coordinates
    [snip]
> Since each port can belong to at most one server, I cannot have
> multiple servers on the same machine.  I can't figure out how to
> ensure there is only *one* copy of the server started, regardless of
> how many clients want to use it.
>
> Any samples, pointers to documentation, key phrases to search MSDN
> for, etc. would be *greatly* appreciated.

I'm not sure how this translates to Python's implementation
of COM, but, at a lower abstraction level, this is exactly
what is accomplished by passing the REGCLS_MULTIPLEUSE value
as the "flags" parameter in the CoRegisterClassObject calls
that an .EXE does, at startup, to let the Service Control
Manager of COM know about its class-objects (IClassFactory
instances).

When such a call is made, and, later, another instance of
the object is required, the SCM won't start another process,
but rather re-use the class-factory it knows is already
active (and which, it has been informed, is OK for multiple
use).  You can take further precautions to check that this
is indeed working as designed, but they should not be needed.

Now, there remains the issue of, how to control the flags
used on CoRegisterClassObject by pythonw.exe...?  I can't
find it in the docs (which doesn't mean it isn't there...
it just says I can't find it).  Maybe we need to look at
the sources...?


Alex





From samschul at pacbell.net  Wed Jul  5 16:18:06 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Wed, 05 Jul 2000 14:18:06 GMT
Subject: ANS: ASPIPython update on Starship
Message-ID: <8jvg2d$8q5$1@nnrp1.deja.com>

I have made some corrections to ASPIPython, that correct memory leaks
found in aspi.dll.

ASPIPython is a open source python interface to the (A)dvanced (S)
system (P)programming (I)nterface for SCSI devices developed by Adaptec
Corp. This interface allows for the development of tests under Python
for SCSI disk drives. These routines can be found at
http://starship.python/crew/schulen

Sam Schulenburg


Sent via Deja.com http://www.deja.com/
Before you buy.


From usenet at foreignkey.com  Wed Jul  5 16:32:57 2000
From: usenet at foreignkey.com (Paul Jensen)
Date: Wed, 05 Jul 2000 14:32:57 GMT
Subject: "Write to a file" question please?
References: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com>
Message-ID: <u7lb07is1.fsf@foreignkey.com>

cmfinlay at SPAMmagnet.com.au writes:

> number = 100
> print "Number =", number
> OUT = open("TT.txt","w")
> OUT.write("Number =", number) # Error
> OUT.close()

The problem is that 'print' is a statement, but the write() function
wants an expression, so you use the "%" operator to substitute values
into the string printf() style, like:

		OUT.write("Number = %d" % number)

I wanted to include a reference to the "%" operator in the online
docs, but interestingly couldn't find one. Anyone know offhand where
"%" is covered in the docs?

Cheers,
-- 
Paul Jensen <usenet at foreignkey.com>


From akuchlin at mems-exchange.org  Wed Jul  5 16:34:19 2000
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 05 Jul 2000 10:34:19 -0400
Subject: Random number generation, simple question.
References: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se> <8jqjf8$s1t$1@nnrp1.deja.com> <39615d75.142572338@news> <8jtcid$4um$1@nntp9.atl.mindspring.net>
Message-ID: <3dhfa4lk7o.fsf@kronos.cnri.reston.va.us>

"Andrew Dalke" <dalke at acm.org> writes:
> What I saw, though, was a compile time option to use /dev/random if
> it exists.  I would have thought that would be enabled by default, but
> it made some comment about how the read will block if there isn't enough
> entropy in the source pool.  Is this ever a problem in real life?

Yes, because the entropy pool is only 4096 bits, and entropy is added
fairly slowly, since an interrupt or other event only adds a few bits
of entropy.  A different device, /dev/urandom, never blocks, so you
can read enough data to theoretically compute the initial state of the
pool.  For generating a long-term thing like an RSA key, this would
matter (or least be psychologically unsettling); for an IV, I think
/dev/urandom would be fine.

--amk


From tismer at appliedbiometrics.com  Wed Jul  5 16:41:35 2000
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Wed, 05 Jul 2000 16:41:35 +0200
Subject: Compatibility between PythonWin and DISLIN
References: <39633408.80B5DFBC@ncmail.net>
Message-ID: <3963491F.70F3188D@appliedbiometrics.com>


Ted Mew wrote:
> 
> Has anyone tried to run the graphing package DISLIN using
> Mark Hammond's PythonWin 1.5.2?  Is DISLIN compatible with
> PythonWin?

No.
I just tried it; they appear to have some different
ideas about event handling. But this can't be too hard
to solve.

...
> I have made a cursory attempt to run DISLIN under PythonWin,
> but crashed in flames, unable to read memory locations.
> Since PythonWin is MS Windows based, I suspect DISLIN is
> attempting to run as an independent process or DOS-based
> program, but have little knowledge in this area.

With Py1.5.2 and Win2000, it doesn't crash. I tried the CURVE
demo. It works fine in a shell, showing its own DISLIN window.
In PythonWin, the window is created, but then both pyWin and
the new window seem to be waiting on each other.

No, DISLIN does not try to be a main program, but it maybe makes
some assumptions about event loops. I'll look into the code
if I find time.

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From paul.robinson at quantisci.co.uk  Wed Jul  5 16:52:47 2000
From: paul.robinson at quantisci.co.uk (Paul Robinson)
Date: Wed, 05 Jul 2000 15:52:47 +0100
Subject: "Write to a file" question please?
References: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com> <u7lb07is1.fsf@foreignkey.com>
Message-ID: <39634BBF.5EFFC493@quantisci.co.uk>

Paul Jensen wrote:
> I wanted to include a reference to the "%" operator in the online
> docs, but interestingly couldn't find one. Anyone know offhand where
> "%" is covered in the docs?

http://www.python.org/doc/current/lib/typesseq-strings.html#SECTION004151000000000000000
[Thats library reference - 2.1.5.1 More String Operations]

...although if you didn't know what an "sprintf() format string" was
then the probability of understanding this particular section is low.

	Paul

-----------------------------------
Business Collaborator Team
Enviros Software Solutions
http://www.businesscollaborator.com
-----------------------------------



From nhv at cape.com  Wed Jul  5 16:38:53 2000
From: nhv at cape.com (Norman Vine)
Date: Wed, 5 Jul 2000 10:38:53 -0400
Subject: "Write to a file" question please?
References: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com>
Message-ID: <8jvh9i$del$1@newsie2.cent.net>

<cmfinlay at SPAMmagnet.com.au>

> number = 100
> print "Number =", number
> OUT = open("TT.txt","w")
> OUT.write("Number =", number) # Error
> OUT.close()
>
> The code above gives the error.
>
> Traceback (innermost last):
>   File "ttest.py", line 5, in ?
>     OUT.write("Number =", number)
> TypeError: read-only buffer, tuple
>
> OUT.write("Hello World")
>
> Works but a number or variable of a number does not.
>
> I have been reading documents on Python to no avail this time.
> I am using Win98 with the latest python for this OS.
>

Try
OUT.write("Number=%d"%(number))

see
Python\Doc\tut\formatting.html

Norman




From samschul at pacbell.net  Wed Jul  5 16:24:06 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Wed, 05 Jul 2000 14:24:06 GMT
Subject: Corrected URL http://starship.python/crew/samschul
References: <8jvg2d$8q5$1@nnrp1.deja.com>
Message-ID: <8jvgdi$95f$1@nnrp1.deja.com>

The corrected URL is:

http://starship.python/crew/samschul

In article <8jvg2d$8q5$1 at nnrp1.deja.com>,
  Sam Schulenburg <samschul at pacbell.net> wrote:
> I have made some corrections to ASPIPython, that correct memory leaks
> found in aspi.dll.
>
> ASPIPython is a open source python interface to the (A)dvanced (S)
> system (P)programming (I)nterface for SCSI devices developed by
Adaptec
> Corp. This interface allows for the development of tests under Python
> for SCSI disk drives. These routines can be found at
> http://starship.python/crew/schulen
>
> Sam Schulenburg
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From Jelske.Kloppenburg at gmd.de  Wed Jul  5 16:51:31 2000
From: Jelske.Kloppenburg at gmd.de (Jelske.Kloppenburg at gmd.de)
Date: 5 Jul 2000 14:51:31 GMT
Subject: Compiler switches for install
Message-ID: <8jvi1j$412$1@hop.gmd.de>

In the Python-1.5.2/README under Configuring threads I 
read
"Unfortunately, on some platforms, additional compiler 
and/or
linker options are required."
Where - configure, configure.in, Makefile, Makefile.in,
                                 Modules/Makefile.pre.in, 
...?
how - CFLAGS=, OPTS=, MY_CFLAGS= ...?
and when - before configure, before make, ...?
    can I set the options?
    
I repeat the question (until now nobody seems to 
understand me):

Where, how and when can I set the options?

-- 
  Jelske Kloppenburg, +49 2241 14-2433, <Jelske.Kloppenburg at gmd.de>
  GMD - German National Research Center for Information Technology



From kpmurphy at my-deja.com  Wed Jul  5 16:35:39 2000
From: kpmurphy at my-deja.com (Keith Murphy)
Date: Wed, 05 Jul 2000 14:35:39 GMT
Subject: backslash in regular expression problem
References: <394A55BA.69269793@wolfson.co.uk> <02de01bfd826$cf05bd80$6401a8c0@home> <8jivfr$qc2$1@nnrp1.deja.com> <m3g0pu7kix.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <8jvh3p$9mg$1@nnrp1.deja.com>

how does re.escape help?

>>> str = '\\\hi'
>>> str
'\\\\hi'
>>> print str
\\hi
>>> import re
>>> x = re.escape(str)
>>> x
'\\\\\\\\hi'
>>> print x
\\\\hi
>>>

i want to display the actual string, and also be able to write it to a
file...  thanks,

keith

In article <m3g0pu7kix.fsf at atrus.jesus.cam.ac.uk>,
  Michael Hudson <mwh21 at cam.ac.uk> wrote:
> Keith Murphy <kpmurphy at my-deja.com> writes:
>
> > what if instead of having '\\stuff' you have a str = '\\stuff'  ?
> >
> > i know that you can print r'\\stuff' but you can't print rstr...
> >
> > how do you use the raw with a variable?
>
> You don't.  You do know about re.escape?  Asking that question
> suggests you might not...
>
> Cheers,
> M.
>
> --
>   I saw `cout' being shifted "Hello world" times to the left and
>   stopped right there.                                -- Steve Gonedes
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From mark at intrepid.net  Wed Jul  5 16:51:42 2000
From: mark at intrepid.net (Mark Conway Wirt)
Date: Wed, 05 Jul 2000 14:51:42 GMT
Subject: Help: Python dumping core
Message-ID: <slrn8m6irt.15sl.mark@elvis.intrepid.net>

I'm working on a python script, and the interpreter keeps dumping
core.  I can't get much information from the core file, as 
python (and the module it calls: wxPython) where not compiled with
debugging symbols included.  All I have is really the stack trace:

#0  0x281a7b7d in _atomic_lock () from /usr/lib/libc_r.so.4
#1  0x281a6033 in _spinlock_debug () from /usr/lib/libc_r.so.4
#2  0x281a9520 in pthread_mutex_lock () from /usr/lib/libc_r.so.4
#3  0x2869be52 in wxMutex::Lock () from /usr/X11R6/lib/libwx_gtk.so
#4  0x28528da5 in wxEvtHandler::ProcessPendingEvents ()
   from /usr/X11R6/lib/libwx_gtk.so
#5  0x284fe831 in wxAppBase::ProcessPendingEvents ()
   from /usr/X11R6/lib/libwx_gtk.so
#6  0x284af1bb in wxApp::OnIdle () from /usr/X11R6/lib/libwx_gtk.so
#7  0x28529044 in wxEvtHandler::SearchEventTable ()
   from /usr/X11R6/lib/libwx_gtk.so
etc.

If I'm reading the trace correctly, it looks like the problem is somewhere
in wxMutex::Lock.  I have a few questions:

  1) Such a core dump points to a problem at a lower level than the
     particular script I'm working on, correct?  It seems to me that
     even if I have a mistake in the python code, python itself shouldn't
     dump core.

  2) Short or recompiling wxWindows (and perhaps wxPython and perhaps
     python itself...) with debugging symbols, is there anything I can do
     track the problem?  If my code is doing something stupid to cause
     the crash, how can I trace the problem if, by dumping core, python
     won't give me a stack trace?

I'm at a bit of a loss.  The reason that I like using tools like python is
to avoid lower-level debugging like this :-)

Thanks in advance!

--Mark


From alex at magenta.com  Wed Jul  5 16:23:28 2000
From: alex at magenta.com (Alex Martelli)
Date: Wed, 5 Jul 2000 16:23:28 +0200
Subject: "Write to a file" question please?
References: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com>
Message-ID: <8jvgk9012ca@news1.newsguy.com>

[posted and mailed]

<cmfinlay at SPAMmagnet.com.au> wrote in message
news:tzFjOfN2zOzPHwvl3ncklEAEb+fa at 4ax.com...
> number = 100
> print "Number =", number
> OUT = open("TT.txt","w")
> OUT.write("Number =", number) # Error
> OUT.close()

The write method accepts one string argument.

You can make a string argument as you seem to
desire with the %-operator:

OUT.write("Number = %d\n" % (number,))

I added the \n line terminator to the string,
because it's not implicit (in print, it IS
implicit, but not in .write).


Another alternative:

    import sys
    sys.stdout = OUT
    print "Number =", number
    sys.stdout = sys.__stdout__
    OUT.close()

By assigning an opened file to sys.stdout
you temporarily "redirect" the output of
normal "print" to that file; you can then
go back to normal using sys.__stdout__,
which keeps the ORIGINAL value.  Not worth
it for writing just a few things, but fine
if you have some function doing a lot of
output with print, for example.


Alex





From ge at nowhere.none  Wed Jul  5 17:34:18 2000
From: ge at nowhere.none (Grant Edwards)
Date: Wed, 05 Jul 2000 15:34:18 GMT
Subject: Python COM for Linux?
References: <8ja7gv$687$1@nnrp1.deja.com> <3958c185.103629691@news> <Za465.1871$iN5.400266@ptah.visi.com> <8juulj$csu$1@pony.kenan.com>
Message-ID: <_zI85.3060$iN5.739427@ptah.visi.com>

In article <8juulj$csu$1 at pony.kenan.com>, Jess Austin wrote:

>>I have a Python program that fetches mail from MS Exchange
>>Server using MAPI/COM.  I'd dearly love to run it under Linux.
>>Does anybody have an experience with accessing MS Exchange
>>Server via COM from Linux?
>
>I'd love to take a look at this.  I'm always interested in anything that
>could help me escape the drudgery of Outlook.  Can you post a link?

Sure:

    ftp://ftp.visi.com/users/grante/stuff/mfetch.py
    ftp://ftp.visi.com/users/grante/stuff/mfetch.bat   shows invokation

Disclaimer: There's some hard-wired logic in one of the methods
that convertes Outlook addresses to SMTP addresses -- grep for
"comtrol".

It only fetches mail -- I send mail using ssmtp with and the
Exchange's SMTP server.

-- 
Grant Edwards                   grante             Yow!  This is PLEASANT!
                                  at               
                               visi.com            


From alex at magenta.com  Wed Jul  5 16:56:38 2000
From: alex at magenta.com (Alex Martelli)
Date: Wed, 5 Jul 2000 16:56:38 +0200
Subject: Hi, my name is not Mrs so-called Gumblenose
References: <395B2B88.CC3041C1@memo.volvo.se> <1c934748.cd75dee0@usw-ex0106-045.remarq.com>
Message-ID: <8jvii30149t@news1.newsguy.com>

Simon B <brunnsNObrSPAM at beer.com.invalid> wrote in message
news:1c934748.cd75dee0 at usw-ex0106-045.remarq.com...
> I've not heard of any way of getting at DB2/400 from Python
> other than via ODBC. But I'd love to be wrong!
>
> If someone were to port JPython to the '400, you would be able
> to use the AS/400 Toolbox for Java's native DB access APIs, but
> that would be a pretty major job in itself.

Why would JPython have to be "ported" at all?  Doesn't it
"just run" with any common-or-garden Java Virtual Machine?


Alex






From Armin at Steinhoff.de  Wed Jul  5 17:13:51 2000
From: Armin at Steinhoff.de (Armin Steinhoff)
Date: 5 Jul 2000 08:13:51 -0700
Subject: PyObject_GetAttr(String) ...
Message-ID: <8jvjbf$1eus@drn.newsguy.com>

Hi all,

a question for the gurus:

   what is the difference between the calls
   
      - PyObject_GetAttr and
      - PyObject_GetAttrString ??

Both have the same 'documented' semantic ... but they are magically different.

Regards

Armin
   



From jerry_spicklemire at my-deja.com  Wed Jul  5 17:48:39 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Wed, 05 Jul 2000 15:48:39 GMT
Subject: Postscript font metrics?
References: <39622330.3732B225@yahoo.com>
Message-ID: <8jvlcg$d1p$1@nnrp1.deja.com>

In article <39622330.3732B225 at yahoo.com>,
  Paul Winkler <slinkp23 at yahoo.com> wrote:

> I might have a go at hacking the relevant parts of the Perl
> PostScript-Font-1.02 module into Python... but then I'll have to
> (yuck) read someone else's Perl code, which I haven't done in a
> looong time and I was never that good at it.
>

If you already have a notion of how to go about this
with the Perl module, you might look at Minotaur

 http://mini.net/pub/ts2/minotaur.html

to call the Perl functions from Python.

Good luck,
Jerry S.



Sent via Deja.com http://www.deja.com/
Before you buy.


From thomas at xs4all.net  Wed Jul  5 18:12:44 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 5 Jul 2000 18:12:44 +0200
Subject: PyObject_GetAttr(String) ...
In-Reply-To: <8jvjbf$1eus@drn.newsguy.com>; from Armin@Steinhoff.de on Wed, Jul 05, 2000 at 08:13:51AM -0700
References: <8jvjbf$1eus@drn.newsguy.com>
Message-ID: <20000705181243.B13478@xs4all.nl>

On Wed, Jul 05, 2000 at 08:13:51AM -0700, Armin Steinhoff wrote:

> a question for the gurus:

>    what is the difference between the calls
>    
>       - PyObject_GetAttr and
>       - PyObject_GetAttrString ??

> Both have the same 'documented' semantic ... but they are magically
> different.

Actually, no, they aren't the same. PyObject_GetAttrString takes a PyObject
pointer and a char pointer, whereas PyObject_GetAttr takes two PyObject
pointers: the second argument should be a PyString, which is then turned
into a char pointer (using PyString_AS_STRING) and passed to
PyObject_GetAttrString.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From jerry_spicklemire at my-deja.com  Wed Jul  5 18:04:21 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Wed, 05 Jul 2000 16:04:21 GMT
Subject: JPE: Java - C Python Extension Project
References: <3962EBA5.E461DD4@arakne.com>
Message-ID: <8jvm9o$dtj$1@nnrp1.deja.com>

Cool! Might the work already done with Pyro serve as a jump-start?

  http://www.xs4all.nl/~irmen/python.html

In article <3962EBA5.E461DD4 at arakne.com>,
  Jean Hemmi <jean.hemmi at arakne.com> wrote:

> We need a full-featured C-Python / Java integration!

> We are laying out below the requirements to such an integration:
> - Multithread compatibility (at least when the JVM relies on OS
threads)
>
> - Access to the Java environment from Python
> - Seamlessness from both sides
> - Java objects appearing as native Python objects (resolving
> overloading)
> - Extensibility of Java classes into Python
> - Access to the Python environment from Java
> - Python objects accessible as Java objects
> - Extensibility of Python classes within Java
> - Exception mapping
> - Python programming on client Web browser, like Javascript
> - Ability to run Python bytecode files on the client Web browser
> - Conciliation of security features (adapting Python's restricted
> execution mode for 'Pylets')
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From dalke at acm.org  Wed Jul  5 18:29:28 2000
From: dalke at acm.org (Andrew Dalke)
Date: Wed, 5 Jul 2000 10:29:28 -0600
Subject: Nth digit of PI
References: <bivnjsoq0amb8qjmqtgkbfjsl2df2n2tta@4ax.com> <FvuAy4.51q@cwi.nl><8hogt2$57t$1@nnrp1.deja.com> <8houlu$j9s$1@nntp.itservices.ubc.ca><394889C3.1ED5@esatclear.ie> <t7545.239$ZR1.2944@news.uswest.net> <20000705071827.A1015156@vislab.epa.gov>
Message-ID: <8jvni0$amr$1@slb3.atl.mindspring.net>

Randall Hopper
>But one of the interesting things that you can show with a Fourier
>integral (possibly other ways too) is that:
>
>    Pi^2/8 = sum(n=1..inf, 1/(2n-1)^2)
>
>You can see that the sum is 1/1 + 1/9 + 1/125 + ....  The terms get smaller
>and smaller, which means making the decimal representation longer and
>longer.  In the limit, the decimal representation is infinitely long
>because the denominator is infinitely large.


I've seen the proof that e is irrational, but don't recall the details.
The sketch you outlined isn't sufficient.  Consider

  sum(n=1..inf, 1/n^2) = 1 + 1/2 + 1/4 + 1/8 + ... = 2

For any finite N in the sum 1..N, the denominator is infinitely large,
but the infinite sum is an integer.

                    Andrew
                    dalke at acm.org





From vxgvxg at my-deja.com  Wed Jul  5 18:27:01 2000
From: vxgvxg at my-deja.com (vxgvxg at my-deja.com)
Date: Wed, 05 Jul 2000 16:27:01 GMT
Subject: Telnetlib
Message-ID: <8jvnjs$f2l$1@nnrp1.deja.com>

Hi,

When using any of the telnetlib read functions (such as read_all(),
read_until(), etc.), the 0x00's seem to be dropped from the data
stream.  Any ideas?


Sent via Deja.com http://www.deja.com/
Before you buy.


From heath7952 at my-deja.com  Wed Jul  5 18:29:51 2000
From: heath7952 at my-deja.com (heath7952 at my-deja.com)
Date: Wed, 05 Jul 2000 16:29:51 GMT
Subject: Wanted: compile.py
Message-ID: <8jvnp5$f4r$1@nnrp1.deja.com>

Does anyone have a copy of compile.py referred to in the "Python
Building C/C++ extensions documentation". I can't reach the website, or
find it elsewhere.

Thanks in advance.

Heath.


Sent via Deja.com http://www.deja.com/
Before you buy.


From nolan_d at bigfoot.com  Wed Jul  5 18:53:02 2000
From: nolan_d at bigfoot.com (Nolan Darilek)
Date: Wed, 05 Jul 2000 11:53:02 -0500
Subject: Nonblocking IO?
Message-ID: <E139sQ2-0002Vi-00@localhost>

I'm trying to do nonblocking IO in a client application which I'm
writing. The client is text-based and line-oriented, and the cmd.Cmd
class does exactly what I want. (Well, almost exactly.) The class uses
raw_input(), which appears to block. I also moved the code which reads
from the connection into a thread, but this introduces a few more
problems; the output is intermingled with the input.

The most obvious solution is to use curses, but this makes my program
much more complex than it needs to be. Since cmd does almost exactly
what I want it to do, I'd much rather change the Cmd class than add
curses. I've seen code written in C which does what I'd like to do
(Micq, for example, has the interface which I'm aiming for, yet it
doesn't use curses/ncurses.) So, is it somehow possible to either a)
make raw_input nonblocking or b) use some other nonblocking IO method?
And, is it then possible to ensure that the output from the socket
isn't intermingled with the input? If I do nonblocking IO, I'm
assuming that it's simply a matter of printing a newline, my message,
and then reprinting the prompt while stuffing the existing input back
into the input buffer. (Of course, I'd rather delete the existing line
first to ensure that the app scrolls cleanly, but getting nonblocking
input to work would be a great step in that direction. :) Can anyone offer any suggestions?

Please CC me with any replies; I'm currently not subscribed to this
list, and I don't have usenet access for another week. Thanks.



From alex at magenta.com  Wed Jul  5 17:49:14 2000
From: alex at magenta.com (Alex Martelli)
Date: Wed, 5 Jul 2000 17:49:14 +0200
Subject: "Write to a file" question please?
References: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com> <u7lb07is1.fsf@foreignkey.com>
Message-ID: <8jvllp017g1@news1.newsguy.com>

Paul Jensen <usenet at foreignkey.com> wrote in message
news:u7lb07is1.fsf at foreignkey.com...
    [snip]
> I wanted to include a reference to the "%" operator in the online
> docs, but interestingly couldn't find one. Anyone know offhand where
> "%" is covered in the docs?

http://www.python.org/doc/tut/node9.html

It might be nice to be able to point more precisely by
adding a "#something", but it seems the tutorial has no
such anchors within chapters; so, one just says "about
midway through the chapter", I guess.


Alex




From olivierS.dagenaisP at canadaA.comM  Wed Jul  5 18:56:38 2000
From: olivierS.dagenaisP at canadaA.comM (Olivier Dagenais)
Date: Wed, 05 Jul 2000 16:56:38 GMT
Subject: Client-server architecture
References: <Xpx85.37285$W35.821323@news20.bellglobal.com>
Message-ID: <aNJ85.50248$HK2.813974@news20.bellglobal.com>

Yes, I'm answering my own post, but it's more of an FYI that might be nice
in the FAQ, if it isn't, already...  Anyway, I found this, after a long,
long search:

http://www.nightmare.com/medusa/index.html

It seems to do a lot of what I was asking.  The main page is rather sketchy,
but the documentation provided helps illustrate a bit more how it works, at
the high level.  (esp. data_flow.html)

I also found something else, called "WebDav", but I don't think it would
suit my purposes all that much.

--
----------------------------------------------------------------------
Olivier A. Dagenais - Carleton University - Computer Science III


"Olivier Dagenais" wrote in message
news:Xpx85.37285$W35.821323 at news20.bellglobal.com...
> What modules [or references] are there that explain or implement
> simultaneous, __secure__ operations on a server?
>
> I looked through http://www.python.org/topics/web/  and it seems to only
> cover HTTP, not HTTPS (using SSL or some other, secure protocol).
>
> I want to implement seemingly concurrent, remote database access and if I
> can queue the requests to the server, the server could then process
> requests, one at a time, and then "call back" each client with the
results.
>
> I'm thinking I don't want to write the server side software using a CGI
> script, since the database package I'm looking into using (Gadfly) isn't
> designed for concurrent access, unless I don't understand how CGI works
and
> it is synchronous?  (I have only written very simple CGI scripts in Perl
and
> C, so I'm sort of new to this)
>
> Any help (including corrections in my thinking of CGI or hints on handling
> many, simultaneous requests) would be appreciated.  I'd like to write both
> the client and the server in Python.
>
> Thanks!
>
> --
> ----------------------------------------------------------------------
> Olivier A. Dagenais - Carleton University - Computer Science III
>
>




From samschul at pacbell.net  Wed Jul  5 18:47:29 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Wed, 05 Jul 2000 16:47:29 GMT
Subject: Corrected URL http://starship.python.net/crew/samschul
References: <8jvg2d$8q5$1@nnrp1.deja.com> <8jvgdi$95f$1@nnrp1.deja.com>
Message-ID: <8jvoqr$g18$1@nnrp1.deja.com>

The corrected URL ( I'll get it right yet)
http://starship.python.net/crew/samschul


Sent via Deja.com http://www.deja.com/
Before you buy.


From R.Brodie at rl.ac.uk  Wed Jul  5 19:01:24 2000
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Wed, 5 Jul 2000 18:01:24 +0100
Subject: Telnetlib
References: <8jvnjs$f2l$1@nnrp1.deja.com>
Message-ID: <8jvpl5$va4@newton.cc.rl.ac.uk>

<vxgvxg at my-deja.com> wrote in message news:8jvnjs$f2l$1 at nnrp1.deja.com...

> When using any of the telnetlib read functions (such as read_all(),
> read_until(), etc.), the 0x00's seem to be dropped from the data
> stream.  Any ideas?

You are aware of the NVT transformations TELNET uses, e.g. replacing
CR/NL by NL?




From DavidA at ActiveState.com  Wed Jul  5 19:22:51 2000
From: DavidA at ActiveState.com (David Ascher)
Date: Wed, 5 Jul 2000 10:22:51 -0700
Subject: Wanted: compile.py
In-Reply-To: <8jvnp5$f4r$1@nnrp1.deja.com>
Message-ID: <PLEJJNOHDIGGLDPOGPJJOEEICMAA.DavidA@ActiveState.com>

> Does anyone have a copy of compile.py referred to in the "Python
> Building C/C++ extensions documentation". I can't reach the website, or
> find it elsewhere.

Apologies that the site is still broken.  Please find it attached.

--david ascher
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: compile.py
Url: http://mail.python.org/pipermail/python-list/attachments/20000705/5c22e1bc/attachment.asc 

From invalid.address at 127.0.0.1  Wed Jul  5 19:32:27 2000
From: invalid.address at 127.0.0.1 (David)
Date: Wed, 05 Jul 2000 17:32:27 GMT
Subject: Python 2.0?
References: <8jvgdi$95f$1@nnrp1.deja.com>
Message-ID: <396370ab.10899776@news.telus.net>

I've read that CNRI is giving the newly-formed PythonLabs a hard time about
the release of Python v2.0.
 
What's up with that?  Isn't Python's license a completely free and clear
one, such that no one can limit its distribution or use?
 
More importantly, what can the public community do to support Guido 'n da
boyz?  CNRI can't be allowed to bugger up Python's future!



From invalid.address at 127.0.0.1  Wed Jul  5 19:44:57 2000
From: invalid.address at 127.0.0.1 (David)
Date: Wed, 05 Jul 2000 17:44:57 GMT
Subject: GUI framework with TreeView
References: <3967f81e.23966859@nntp.interaccess.com>
Message-ID: <3965740e.11766815@news.telus.net>

On Wed, 05 Jul 2000 08:56:09 GMT, olczyk at interaccess.com (Thaddeus L
Olczyk) wrote:

>I need a "free" gui framework that runs on both Windows and UNIX
>( so that lets out QT ) and gives me a component similr to Microsofts
>TreeView. Does anyone know of such a thing?

wxPython/wxWindows.


From stephen at cerebralmaelstrom.com  Wed Jul  5 19:45:27 2000
From: stephen at cerebralmaelstrom.com (Stephen Hansen)
Date: Wed, 05 Jul 2000 17:45:27 GMT
Subject: "Write to a file" question please?
References: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com>
Message-ID: <XuK85.115473$WS3.978541@typhoon.we.rr.com>

I *think*...
Print is a statement that lets you print out multiple things using the
comma, the 'write' function does not work the same way. It takes a string
argument and writes it out to the file.

To accomplish what you want, you can use the module operator (there's other
ways too, but its my favorite :))

number = 100
print "Number = %s" % number
OUT = open("TT.txt",'w')
OUT.write("Number = %s" % number)
OUT.close()

If you have more then one number you're writing out, make the right-operand
of the modulo operator a tuple.. E.g.

num1 = 100
num2 = 500
print "Number 1 = %s and Number 2 = %s" % (num1,num2)

See? :)

--Stephen



<cmfinlay at SPAMmagnet.com.au> wrote in message
news:tzFjOfN2zOzPHwvl3ncklEAEb+fa at 4ax.com...
> number = 100
> print "Number =", number
> OUT = open("TT.txt","w")
> OUT.write("Number =", number) # Error
> OUT.close()
>
> The code above gives the error.
>
> Traceback (innermost last):
>   File "ttest.py", line 5, in ?
>     OUT.write("Number =", number)
> TypeError: read-only buffer, tuple
>
> OUT.write("Hello World")
>
> Works but a number or variable of a number does not.
>
> I have been reading documents on Python to no avail this time.
> I am using Win98 with the latest python for this OS.
>
> Any help would be much appreciated
> E-mail me at cmfinlaySPAM at magnet.com.au
> or reply here.




From stephen at cerebralmaelstrom.com  Wed Jul  5 19:46:35 2000
From: stephen at cerebralmaelstrom.com (Stephen Hansen)
Date: Wed, 05 Jul 2000 17:46:35 GMT
Subject: "Write to a file" question please?
References: <tzFjOfN2zOzPHwvl3ncklEAEb+fa@4ax.com> <XuK85.115473$WS3.978541@typhoon.we.rr.com>
Message-ID: <%vK85.115476$WS3.978721@typhoon.we.rr.com>

Oops. Replace my %s's with %d's.

--S


Stephen Hansen <stephen at cerebralmaelstrom.com> wrote in message
news:XuK85.115473$WS3.978541 at typhoon.we.rr.com...
> I *think*...
> Print is a statement that lets you print out multiple things using the
> comma, the 'write' function does not work the same way. It takes a string
> argument and writes it out to the file.
>
> To accomplish what you want, you can use the module operator (there's
other
> ways too, but its my favorite :))
>
> number = 100
> print "Number = %s" % number
> OUT = open("TT.txt",'w')
> OUT.write("Number = %s" % number)
> OUT.close()
>
> If you have more then one number you're writing out, make the
right-operand
> of the modulo operator a tuple.. E.g.
>
> num1 = 100
> num2 = 500
> print "Number 1 = %s and Number 2 = %s" % (num1,num2)
>
> See? :)
>
> --Stephen
>
>
>
> <cmfinlay at SPAMmagnet.com.au> wrote in message
> news:tzFjOfN2zOzPHwvl3ncklEAEb+fa at 4ax.com...
> > number = 100
> > print "Number =", number
> > OUT = open("TT.txt","w")
> > OUT.write("Number =", number) # Error
> > OUT.close()
> >
> > The code above gives the error.
> >
> > Traceback (innermost last):
> >   File "ttest.py", line 5, in ?
> >     OUT.write("Number =", number)
> > TypeError: read-only buffer, tuple
> >
> > OUT.write("Hello World")
> >
> > Works but a number or variable of a number does not.
> >
> > I have been reading documents on Python to no avail this time.
> > I am using Win98 with the latest python for this OS.
> >
> > Any help would be much appreciated
> > E-mail me at cmfinlaySPAM at magnet.com.au
> > or reply here.
>
>




From ssthapa at midway.uchicago.edu  Wed Jul  5 20:22:43 2000
From: ssthapa at midway.uchicago.edu (Suchandra Thapa)
Date: Wed, 05 Jul 2000 18:22:43 GMT
Subject: Random number generation, simple question.
References: <Pine.GSO.4.21.0007031546410.25965-100000@luna.dsv.su.se>
Message-ID: <slrn8m4ov8.1ct.ssthapa@ntcs-ip48.uchicago.edu>

Kalle Svensson <kalle at gnupung.net> wrote:
>Hello.
>
>I have a few questions about the (pseudo) random number generator supplied
>in the standard library. I need 80 random bits (for a CipherSaber IV) but
>I hear most RNG's only supply 32 bits or less. My questions are:
>
> * How many random bits does whrandom.random() generate?
>
> * Suppose I do a loop where I get a byte at a time and then re-seed the
>   generator either with random numbers from the generator or with values
>   from time. Will that improve the randomness?

    This may not be entirely wise.  An psuedo-random number generator
eventually cycles and starts repeating.  The problem is that the seed
supplied to the prng affects how quickly it loops and the distribution
of the numbers the prng generates.  If you feed random numbers into 
the prng youo could end up with weaker encryption than you think. 


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

Suchandra S. Thapa 
s-thapa at uchicago.edu

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


From just at letterror.com  Wed Jul  5 21:22:05 2000
From: just at letterror.com (Just van Rossum)
Date: Wed, 05 Jul 2000 20:22:05 +0100
Subject: Postscript font metrics?
References: <39622330.3732B225@yahoo.com> <3963368A.8442AD75@bellatlantic.net>
Message-ID: <39638AA5.7C36FD77@letterror.com>

Paul Winkler wrote:
>
> I need to generate some nice Postscript output from a python script,
> and I'm finding that I need to get at the font metrics info from the
> .afm files so I can place things correctly. (Either that or write
> *everything* in postscript -- no way!!)
>
> I can't find anything like this on Parnassus. Anyone ever seen any
> python code that gets the useful info from .afm files?

My (hardly advertized, but free) FontTools package contains an AFM parser:

   http://www.petr.nl/just/FontTools.tar.gz

Sorry, the doco is, erm, somewhat limited ;-)

Just


From morganea at bellatlantic.net  Wed Jul  5 20:26:41 2000
From: morganea at bellatlantic.net (Michel Orengo)
Date: Wed, 05 Jul 2000 18:26:41 GMT
Subject: [ANN]pre-release DBAPI2.0 for ADO
Message-ID: <B5L85.2924$EV6.75523@typhoon1.ba-dsg.net>

Hi all,
I enclosed a very preliminary version of the DBAPI 2.0 for ADO (Microsoft
Active X Data Object) - ADODBpy. It is built on top of another module called
ADOpy which give a more "pythonic" interface to ADO (I've posted this module
few days ago)

ADODBpy is still largely incomplete (does not deal with exceptions for
instance). It is released to get a feedback from the python community.

Some examples to get you started:

import ADODBpy

db = ADODBpy.connect('MyDB', 'MyUser', 'MyPassword')
cur = db.cursor()
cur.execute('select * from mytable')
tup = cur.fetchone()

#now because it is build on top of ADOpy, we can access also the
properties/methods of the
#underlying ADOobject
print db.provider
print cur.fields.Item('col').Value
# or better:
print cur['col']

Hope this help.
Cheers,
Michel


begin 666 ADODBpy.zip
M4$L#!!0``@`(`%QOW"@DOGOS3 0``(P-```(````041/<'DN<'G-5LUNXS80
MO@?(.TR=@^3&$;+;4X.F11JG0 Y)C'3;+1H8`DV-8VYE4B#I.%KTF?H0^V0[
M)!69EI5NVAQ:(0ADSO_WS0QU`$=?>O;W#N!L?%/5)W#!3 V7TJ*>,XXP5]I)
MG,)+O#B]]UI8BQ)F-5P)OL 2;C3*>Y7!9#4K!8>Q6C(A,Z=[K>!<5;46]PL+
M(Y *;MVK at 5LTJ!^P& &3A3M_S[1FT at HTL%XP:Q0^H,Z:D/1JA))PG+W=WZ.#
M`WBW$ :6JEB5"(*JD84!JZ#2ZD$4"(QD&F%0U7:AI. #KQ1*)K50L<<$3ET5
M6ADUMY_^@C-NQ0/^1F]C9AG<S#X at MV83DOZ8!'QDRXHBSYC!`E9&R'N0S%E"
MB,C5<A/QQ)N+9:6TA;60W[PE<<9+@=)Z$9>2TNA(LK$P%;-\D2:4YOC'[%Q)
M2<D0#LGPR2R;:%6A=K"E`Y_QSVJE.0Z&V:^L7"&Y3:YJ)W"I)KUFOQ 3<#F.
M3=S1-5NV!EOZ$V;,6NDB-IB8M2XV[F\JE&E(4AM+"N[PXA'YRF*:&"RI#O@:
MYII at NJK?L5F)34WSLB!U,LI^$E at 6)DVX*M\D0SC at A#LKC8(9GFS4TN&3AC>O
M-*'NI$UJVU9!NF/E+:_5>@1V at 6"H;N+"+IJ)B<GS)Q%E_G=$3+HAIB#,`Q?/
M<; BD"_'?7#+K&HP[L%6=;$-2=PB)WV#-G5*^ 6L.V#?!32FVQBVV)P9J-4*
M')8&,01T[:UI_C at -K%$$F5D00A34A)$ND& /.#I4BUJR)>V&JFTD8&&LQ-PI
MU+!&&EAC:8IBK88;2^SUN2"3`@F0PDT at +0@7:C )2T /LF9NR;O39)R2-6+F
M!K?VJE1U&9@#Y2<]:G2W:/;W&N*-)5#N1Z"Q/6H'O3W9&>W]O=R1:!S(F)&D
M$B5QDOXP^6Y=?']W=O3[U/\[_#.\LZ./QT??3@^'KB_W]WA)70`YP>W?J!.!
MG at +GD.?W2$A9G><I,3RGS4KM,VPTW).[`XH;$L]6%565NK.[X^GPT+^\.9EN
M]*VN(VOW/#2SW43R<3+*)0`U\@&&VR8:[4K+8+F1X"/'BF:'G(@9->6%UDIW
M at G72_:"$3 -VV5S(@I5E&@*.$D at Z47=3C]/?3CM>9,%A6!2[]OVU1/6T_#MB
M<^PIRKMAPF"G]C2"+K!I>M@<A=@QJ?^8U1YLS-_0^10RLOAOR'L9:13RM?3X
M0'E>"&[S_,[YGFY[]?QP1E=&8*9-*N9EVTO2ZB33L)[#K\[X"HO+EO _L(X=
M-LUWWJZG2U+N4$9WF+.*-D5T#[5+8[BU-804MHW9?"OITR1Y<3'/?J,,NM\H
M at XAH6O%/P> KNM&2G9[LTAUT3UNSN,CEDNZ7%U78NCR]5G)KDBBC5NB^Z9R\
M+ZE_CX//, at 8!2],?PG?7IJ\VI6ZN]/]WL6V>KRNWT^VOJ_F96-'*;==MS^"9
MG049-LXP+B=8;9?X&5!+`P04``(`" !K8N4H!G'^:A\'``!O% ``"@```$%$
M3T1"<'DN<'FU5VU/X\86_ at P2_^&4B.+T>JV$95^*RI5@`[V1N@$!V]6JJJ*)
M/4GFKNWQSHQATU_?<V8\SCAA6?JA"(%]SC/G;<Z;>_#B>S][NSTX&UV-SJ\_
MG<"(&39CFL/9]1B.D@',I2(N at 9XCB7 ?E3"&ES!;P7N1+GD.5XJ7"YG =3W+
M10HC63!1)H2=2'@GJY42BZ6!&$H)-_2HX89KKNYY%@,K,Z)_9$JQT at BNX6')
MC);\GJND48F/6L at 2!LEP;W=O5Q255(;LKE;0$Z7A:LY2#@_"+)TW#<*L*JYC
M\IK?B8)#C]R=B3(3Y<(J3F5I5)T:J30)WMLU2\59IMF<FQ6<PA!ZI32@:\7!
M+(4&_,4#W,:-Y3E42MZ+C--Q5HD<C<[QV#Z&=G]OMV**%=JL<DZT+P53G_=)
M20^#4NJU8K(/$=[,A![V=JW%`97^.*HVK*@V69;H9%TJ6=R)]+/>%-HRG)Q'
M<1U&H.^;X"[7A3'C<PI1P4Q4]4_V=G?$')9,,V-45,6'EY9U:#D[BIM:E5 E
MCAKM'WQZ<5"\.,C at X'\G!^]/#F[W^XCCN>8=O->38_*9I9PWFHQ:A3#D(H/.
M?TUY94+61);<F=L[%R53=.&S>C[GRE'3G&D-MP83A:GL0BFIH at LK!7/1ZJH0
MX'&8P"6F5=3!;Z&<E*<Q8Y_0C<I'0;Z2OX=Q_ YZ"W=5<<7(*98_!T[V+; )
MK)X+5L\4?*WD`BNFP# ^!SZ1YK:NJ-!Y]C1^?9NC<^Q\=]@6KF;_YZE-!\JA
MZ524PDRGD>;Y//[IGN4UURX]B9(X`N:'>VA/I47E#TFSY$XG);M]`U%"<-SR
M?/(-.L!?OHD;TDN;^Y[ZPO;!V[N;\>17JL>N5Y&%X@\,CWZ.H<>R=TNF6N+1
M8&B)O\ER\3M3'=[;&(!XYRB[);XYBBWQUP_CT5KTP!%MK /9+T/9'SO"AR\'
M]DB7BL5Y/IZ<W7QZVI&W5JXKU$#=L26CJBW.*P@,V>"BSLF']^<7-T_J/'8>
MCG at J"I:W]%=-C$82YQUOR2\;LBT/'G@]=&+.I<PY*UOZZP;_KE8X/]-5$"9W
M/Y.ZX$JD+?VX.7"+]1'H/?)DM#%'Y6LYKYWB.U&N0OI18]"'4HM%R3&FB\ZQ
MG[OL#N]ME[>M\PUT`)NZ,?"CL[N+N_'[BZ="_\:'V$["-BXNMT;G&^3CANPF
M9$M^%9!OW71L>/#ZV(J_Q(G=.01DX,W5Q_'H*>M0]NNFJBKL;K at Y@=T2I%L4
MV$S6QJT+.&&2)/$'^V$GPA6 at 1*G8=2.[R"1K0M^WI=217(.!3)<QU+ at WQ4!M
M[4&J+&A1LL))]QBDE86-U8D*3CGJ':Y>.O)()?-\QM+/FUA/[Z+36FELO<[$
MDA7\]! '.] I9&$8WSD`L?H--6'HYSU?>XPPJP(C(6VP at S&-^"!J3IB/6%'@
M((W=VPU/T5W-3?_QKKXVCH1W)"0=9!_OENZ/PXPO! UU$.90`[,Q1'C3NDD>
M_' *AX>N-UL/+/'4\MK T6J[TN(O8 at Q;:L9UJD35N.^6D3;2#ZFL2X-TZO0N
MSEAIN&VFC3/T2$IB^,FNF1P3,9Q8_"M/:\.C-6X-B^'8WYZ'.:'2+P(A^O2/
M/V/O.57#Z7 C?XC.OY*Q[?DM/IY#?O#F+MC>&#+\-+3;VMH=^"\,7&S71$07
MK(K<=AG:V7<#4B.Q$\38QG8KPY*+QO7&BBU1P=#=DOE=:?W&(3R:7 J>9QIS
MC0YBO at R:;QYCN1=7ET'Z3%.F381T3,*)?,#$PR3$L'%*/^4SW.+]HKN=,^[=
M at JU2BPKVWYUF+=K9SD2\;$NG[QM!^PO6^L*E1S(/_. at W@BS-JVP`8\.+2+CK
MV&GJP;*2"?.]EK[,IJG,^"-]UF&)T,C(A*YRMIHV)>3X(SX7-(&0YE"BV32[
ML+/4U"Q?HRJ,B]"NYASBVE.:8&*A!1:[&7Q+Q,:A.L^G\O-:`7[3B!E>O88?
M8?#U:& '2YZ-]021K-T0-B.=L H[=A9%KCK;@&"?#]R-H>-6[$>)G4.M+YB=
M9""V.&=<?Z.\L>A6VR6N^9>IG$_#OO"OE+E%H*VHB&>^!6[4/N5;4.-V<=ZP
M+JB29Q>Q_0[E)EUB!PC&65LZ6*%6(!4A=#X+NP6SP<J%ICBX6GFL5"ZW2X7.
M^#L/,4VU)+_3[M]OXU7(>S[!<-LYZ_6;NLIY1)(ZS at 772TER.F@_0UPI8(Q/
MFD?H#J0M9X2_$&PG;5G[`%I3'I:X-$%$?%QPJ(M15/JVIT6"OF10[B,^XP%7
MSJ1"P']<&GQ;C_]^1Q&AKS@%@XOLV/X=DS<L?L+"?V 3-F-15K4AIW5P!^$H
MGMK2I52W_]<G<4ULC at 8GJ0CSNL E$'J026XM7M(20HE6R@>2ZS]KPR41=S_<
M;F+:_NB_W__\QI.6U/*"K;/OB$EP_K'=T6]A)5;[WU!+`0(4`!0``@`(`%QO
MW"@DOGOS3 0``(P-```(``````````$```"V at 0````!!1$]P>2YP>5!+`0(4
M`!0``@`(`&MBY2@&<?YJ'P<``&\4```*``````````$`( "V at 7($``!!1$]$
<0G!Y+G!Y4$L%!@`````"``(`;@```+D+````````
`
end



From nick at spam_me_not_videosystem.co.uk  Wed Jul  5 20:30:32 2000
From: nick at spam_me_not_videosystem.co.uk (Nick)
Date: Wed, 5 Jul 2000 19:30:32 +0100
Subject: GUI framework with TreeView
References: <3967f81e.23966859@nntp.interaccess.com>
Message-ID: <39637f28@news.xtml.co.uk>

"Thaddeus L Olczyk" <olczyk at interaccess.com> wrote in message
news:3967f81e.23966859 at nntp.interaccess.com...
> I need a "free" gui framework that runs on both Windows and UNIX
> ( so that lets out QT ) and gives me a component similr to Microsofts
> TreeView. Does anyone know of such a thing?

Python Mega widgets adds tree view to tkInter I believe.

wxPython and pyFltk are 2 window system with Python bindings. wxPython certainly
has a tree view.









From olivierS.dagenaisP at canadaA.comM  Wed Jul  5 20:56:47 2000
From: olivierS.dagenaisP at canadaA.comM (Olivier Dagenais)
Date: Wed, 05 Jul 2000 18:56:47 GMT
Subject: [ANN]pre-release DBAPI2.0 for ADO
References: <B5L85.2924$EV6.75523@typhoon1.ba-dsg.net>
Message-ID: <PxL85.40906$W35.880547@news20.bellglobal.com>

Hey, this announcement made me think of a question concerning databases:  is
the DBAPI designed for exchanging recordsets between databases opened by two
different implementations?

For example, could I take your ADO cursor and assign it to a database opened
with, say, Gadfly?  (supposing they had the same table definitions)

Thanks!

--
----------------------------------------------------------------------
Olivier A. Dagenais - Carleton University - Computer Science III


"Michel Orengo" <morganea at bellatlantic.net> wrote in message
news:B5L85.2924$EV6.75523 at typhoon1.ba-dsg.net...
> Hi all,
> I enclosed a very preliminary version of the DBAPI 2.0 for ADO (Microsoft
> Active X Data Object) - ADODBpy. It is built on top of another module
called
> ADOpy which give a more "pythonic" interface to ADO (I've posted this
module
> few days ago)
>
> ADODBpy is still largely incomplete (does not deal with exceptions for
> instance). It is released to get a feedback from the python community.
>
> Some examples to get you started:
>
> import ADODBpy
>
> db = ADODBpy.connect('MyDB', 'MyUser', 'MyPassword')
> cur = db.cursor()
> cur.execute('select * from mytable')
> tup = cur.fetchone()
>
> #now because it is build on top of ADOpy, we can access also the
> properties/methods of the
> #underlying ADOobject
> print db.provider
> print cur.fields.Item('col').Value
> # or better:
> print cur['col']
>
> Hope this help.
> Cheers,
> Michel
>
>
>




From embed at NOSPAM.geocities.com  Wed Jul  5 20:58:43 2000
From: embed at NOSPAM.geocities.com (Warren Postma)
Date: Wed, 5 Jul 2000 14:58:43 -0400
Subject: __getattr__ and pickle
References: <395f26a4.1832855@news.iol.ie> <395f2748.1996610@news.iol.ie> <395f27ab.2095753@news.iol.ie> <8F655BA3Egmcmhypernetcom@199.171.54.154>
Message-ID: <vzL85.16702$227.301196@nnrp1.uunet.ca>

"Gordon McMillan" <gmcm at hypernet.com> wrote in message
news:8F655BA3Egmcmhypernetcom at 199.171.54.154...
> [posted and mailed]
> C'mon Sean! You've been around a while. You know when __getattr__ hacks go
> bad it's because they go recursive. (And the stack size on Windows is off
> by  some tiny amount, so it GPFs instead of traps.)


That kinda says (a) Windows Sucketh, and (b) surely there must be a
workaround for the Infinite-Recursion-Crash bug?

Oh yeah, there is, .... It's called Christian Tismer's Stackless Python.
Hope that stackless makes it into the official python, by 1.7, or if not,
there's always 3000.  Here's hoping.

Warren




From scorder at incigna.com  Wed Jul  5 21:22:59 2000
From: scorder at incigna.com (Sam Corder)
Date: Wed, 5 Jul 2000 15:22:59 -0400
Subject: Ado stored procs with parameters
Message-ID: <CPEKKMGLLILAJEKJNAPDAEEBCBAA.scorder@incigna.com>

I'm having a bit of trouble using a stored proc on MSSQL that takes
parameters.  I create a command object and then append the parameters to it.
Then I create a recordset object and use the open method passing the command
object and a few other parameters.  Unfortunately I get an error saying that
the stored proc requires The same parameter that I just added.  Below is the
code I'm using.  Any ideas?

Sam Corder

import win32com.client
constants = win32com.client.constants

class dbcon:
    def __init__ (self, connect = None):
        self.connectstr = connect
    def runSQLReturnRS(self, sql, params = None):
        if self.connectstr == None:
            return None
        else:
            rs = win32com.client.Dispatch('ADODB.Recordset')
            cmd = win32com.client.Dispatch('ADODB.Command')
            cmd.ActiveConnection = self.connectstr
            cmd.CommandText = sql
            if params != None:
                for parm in params:
                    cmd.Parameters.Append(cmd.CreateParameter(parm[0],
parm[1], constants.adParamInput, parm[2], parm[3]))

            rs.Open(cmd, CursorType = constants.adOpenForwardOnly, LockType
= constants.adLockReadOnly)
            return rs

    def __AddParams(self, cmd, params):
        for parm in params:
            cmd.Parameters.Append(cmd.CreateParameter(parm[0], parm[1],
constants.adParamInput, parm[2], parm[3]))

if __name__  == '__main__':
    db = dbcon("Provider=SQLOLEDB;Data Source=s-incigna2k;User
Id=msl_login;Password=***;Connect Timeout=5;network library=dbmssocn;Initial
Catalog=msl;")
    params = ("ExpJobID", constants.adInteger, 4, 11),
    rs = db.runSQLReturnRS("ExpConfig_Get", params)
    print rs.Fields.Count





From kpmurphy at my-deja.com  Wed Jul  5 21:08:51 2000
From: kpmurphy at my-deja.com (Keith Murphy)
Date: Wed, 05 Jul 2000 19:08:51 GMT
Subject: Grid Geometry Manager and grouping?
Message-ID: <8k013q$lon$1@nnrp1.deja.com>

i've found that i like grid the best.  is there a way to group sections
of the window together... like the left side is for one purpose and the
right for another?  I have been playing with frames to no avail.

i really appreciate all the help this ng has given me thus far in
learning python.  kudos to all of you!

-->keith


Sent via Deja.com http://www.deja.com/
Before you buy.


From jamarijr at hotmail.com  Wed Jul  5 21:42:28 2000
From: jamarijr at hotmail.com (Arinté)
Date: Wed, 05 Jul 2000 19:42:28 GMT
Subject: Embedded apps and range???
Message-ID: <8k0331$n7d$1@nnrp1.deja.com>

I am trying to use a range in a for loop
	for i in range(-1,sz):
		devlist[0] = PossDevice.PossArg(resource,i)
		devlist[0].setValue(i)
		ret = diz.set(devlist)
		diz.ioctl("POS_SYS_GET_VALUES",devlist)
		if i not in _range:
			print "Checking for errors " + str(ret)
			checkForErr(ret,error)

I am getting the good old "TypeError: len() of unsized object" error,
problem is when I use range(sz) or range(1,sz) I don't get the error???




Sent via Deja.com http://www.deja.com/
Before you buy.


From jtravis at cse.unl.edu  Wed Jul  5 22:07:59 2000
From: jtravis at cse.unl.edu (Jon Travis)
Date: 5 Jul 2000 20:07:59 GMT
Subject: BUG: profile.py
Message-ID: <8k04iv$meo$1@unlnews.unl.edu>

There is a small bug while running some of
the profiling code from within the profiler.  
Specifically:

import profile
x = profile.Profile()
x.run("x.dump_stats('/tmp/foo')")

I have a small 2 liner fix, but I doubt it is correct,
merely a band-aid, so I won't post it.

-- Jon




From johannes at zellner.org  Wed Jul  5 22:24:15 2000
From: johannes at zellner.org (Johannes Zellner)
Date: Wed, 5 Jul 2000 22:24:15 +0200 (CEST)
Subject: readline.c & more
Message-ID: <Pine.LNX.4.21.0007052214190.4947-100000@krispc6.physik.uni-karlsruhe.de>

Hello,

I just compiled python 2 from CVS.

 ?) I had to modify Objects/fileobject.c because there was a
    name clash between `getline' there and in /usr/include/stdio.h
    (as it comes with the latest debian potato). I replaced
    `getline' with `file_getline' in Objects/fileobject.c.
 ?) I modified Modules/readline.c so that it allows reading and
    writing of a history file. I also plan to implement readline
    functions as `stifle history'. Eventually I'd also add the
    possiblity to register a script completer which would be
    called before the builtin completers. My question about
    this: where would I post patches ? -- I've an account on
    SF for other projects, but as this is my first mail to the
    list I'd understand if you won't trust me ;-)
 ?) Is there a forum for developers? This list seems more for
    users isn't it ?

greetings,

-- 
   Johannes




From richard_chamberlain at ntlworld.com  Wed Jul  5 22:30:06 2000
From: richard_chamberlain at ntlworld.com (richard_chamberlain)
Date: Wed, 5 Jul 2000 21:30:06 +0100
Subject: Embedded apps and range???
References: <8k0331$n7d$1@nnrp1.deja.com>
Message-ID: <6UM85.860$dL5.16129@news6-win.server.ntlworld.com>

What is sz assigned to?

>>> sz=3
>>> for i in range(sz):
...  print i
...
0
1
2
>>> for i in range(1,sz):
...  print i
...
1
2

An integer works fine.

Richard

Arint? <jamarijr at hotmail.com> wrote in message
news:8k0331$n7d$1 at nnrp1.deja.com...
> I am trying to use a range in a for loop
> for i in range(-1,sz):
> devlist[0] = PossDevice.PossArg(resource,i)
> devlist[0].setValue(i)
> ret = diz.set(devlist)
> diz.ioctl("POS_SYS_GET_VALUES",devlist)
> if i not in _range:
> print "Checking for errors " + str(ret)
> checkForErr(ret,error)
>
> I am getting the good old "TypeError: len() of unsized object" error,
> problem is when I use range(sz) or range(1,sz) I don't get the error???
>
>
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From digitig at cix.co.uk  Wed Jul  5 22:29:00 2000
From: digitig at cix.co.uk (Tim Rowe)
Date: Wed, 5 Jul 2000 21:29 +0100 (BST)
Subject: [FAQTS] Python Knowledge Base Update -- July 1st, 2000
References: <20000702031926.7937.qmail@synop.com>
Message-ID: <memo.20000705212939.11741A@digitig.compulink.co.uk>

In article <20000702031926.7937.qmail at synop.com>, fiona at sitegnome.com 
(Fiona Czuczman) wrote:

<snip>

> On Windows 9x systems, you may also use the COM interface. We have 
> recently used PythonWin to (what else?) spam a number of recipients 
> over Novell GroupWise. It took very little time to work this out.

Ah, that's useful -- I sometimes use Python to parse text into an Access 
97 DB, and at the moment I do it via CSV's. Where do I find info on the 
COM interface? With that I should be able to poke it in directly.


From gmcm at hypernet.com  Wed Jul  5 22:39:50 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 05 Jul 2000 20:39:50 GMT
Subject: __getattr__ and pickle
References: <395f26a4.1832855@news.iol.ie> <395f2748.1996610@news.iol.ie> <395f27ab.2095753@news.iol.ie> <8F655BA3Egmcmhypernetcom@199.171.54.154> <vzL85.16702$227.301196@nnrp1.uunet.ca>
Message-ID: <8F68A4032gmcmhypernetcom@199.171.54.194>

Warren Postma wrote: 

>
>"Gordon McMillan" <gmcm at hypernet.com> wrote in message
>news:8F655BA3Egmcmhypernetcom at 199.171.54.154...
>> [posted and mailed]
>> C'mon Sean! You've been around a while. You know when __getattr__
>> hacks go bad it's because they go recursive. (And the stack size on
>> Windows is off by  some tiny amount, so it GPFs instead of traps.)
>
>
>That kinda says (a) Windows Sucketh, and (b) surely there must be a
>workaround for the Infinite-Recursion-Crash bug?

Well, (a) is your own conclusion, not mine; and Christian also has a tiny 
little patch that fixes (b) for standard Python on Windows (it's a 
miscalculation in Python's config that doesn't leave enough stack space for 
handling the traceback). I don't use it, because I know the symptoms, and 
can usually fix the problem before the stack trace finishes printing.

>Oh yeah, there is, .... It's called Christian Tismer's Stackless Python.
>Hope that stackless makes it into the official python, by 1.7, or if
>not, there's always 3000.  Here's hoping.

I love it too, but it gives you a whole bunch of entertaining new ways to 
get into infinite recursions.

- Gordon


From richard_chamberlain at ntlworld.com  Wed Jul  5 22:44:31 2000
From: richard_chamberlain at ntlworld.com (richard_chamberlain)
Date: Wed, 5 Jul 2000 21:44:31 +0100
Subject: Grid Geometry Manager and grouping?
References: <8k013q$lon$1@nnrp1.deja.com>
Message-ID: <H5N85.886$dL5.17528@news6-win.server.ntlworld.com>

Hi Keith,

For what purpose do you want to group the areas?

An example of frames:

from Tkinter import *
root=Tk()
frame1=Frame(root,relief=GROOVE,borderwidth=3)
frame2=Frame(root,relief=GROOVE,borderwidth=3)
for i in range(1,9):
    Label(frame1,text="Frame: 1 - Label: %d" %i).pack()
    Label(frame2,text="Frame: 2 - Label: %d" %i).pack()
frame1.pack(side=LEFT,fill=BOTH,expand=1)
frame2.pack(side=LEFT,fill=BOTH,expand=1)

root.mainloop()

Pmw has got Pmw.Group, a framed widget on which you can have a label or
other widget in the frame and then a number of widgets within the frame.

Give me an example of what you want to achieve and I'll have a go.

Richard

Keith Murphy <kpmurphy at my-deja.com> wrote in message
news:8k013q$lon$1 at nnrp1.deja.com...
> i've found that i like grid the best.  is there a way to group sections
> of the window together... like the left side is for one purpose and the
> right for another?  I have been playing with frames to no avail.
>
> i really appreciate all the help this ng has given me thus far in
> learning python.  kudos to all of you!
>
> -->keith
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From smoriano at sicon.net  Tue Jul  4 14:00:18 2000
From: smoriano at sicon.net (sergio)
Date: Tue, 04 Jul 2000 14:00:18 +0200
Subject: PROCESS QUESTION
Message-ID: <3961D1D2.9C66B425@sicon.net>

  Hi all, I'm working in NT SERVER 4.0

How can i search the id of a process running, searching it by the name?

Thanks for all your responses




From no_spam at umich.edu  Wed Jul  5 22:59:56 2000
From: no_spam at umich.edu (Jeff)
Date: Wed, 5 Jul 2000 13:59:56 -0700
Subject: Overloaded Operators allowed???
Message-ID: <8k07ke$j0r@news.or.intel.com>

Can you overload operators as you can in other languages?  For example, can
I make "&" give the address of something as in C++?  Or can I only overload
things like the cmp operator?  i.e. I want to say &string and get the
address of string.

thanks




From bjorn at roguewave.com  Wed Jul  5 23:38:35 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Wed, 05 Jul 2000 15:38:35 -0600
Subject: Overloaded Operators allowed???
References: <8k07ke$j0r@news.or.intel.com>
Message-ID: <3963AADB.25A6776E@roguewave.com>

Jeff wrote:
> 
> Can you overload operators as you can in other languages?  For example, can
> I make "&" give the address of something as in C++?  Or can I only overload
> things like the cmp operator?  i.e. I want to say &string and get the
> address of string.

Since "&string" is illegal syntactically in Python it wouldn't make
sense to let you overload on it.  It would be like saying "I want to
overload @ in C++ so I can use it to append lists, like list1 @
list2..."

Besides, you really don't want the address of the string (trust me
<wink>).

-- bjorn



From johannes at zellner.org  Wed Jul  5 23:28:10 2000
From: johannes at zellner.org (Johannes Zellner)
Date: Wed, 5 Jul 2000 23:28:10 +0200 (CEST)
Subject: cleanup order
Message-ID: <Pine.LNX.4.21.0007052319220.5959-100000@krispc6.physik.uni-karlsruhe.de>

Hi,

suppose I've

    # fred provides `fun'
    import fred

    class Lola:
        def __init__(self, x):
            self.x = x
        def __del__(self):
            fred.fun(self.x)

    x = Lola('hello')
    # end of script

this doesn't work, because fred is not present any more
when x is destroyed. I don't know anything about python's
internals, but I guess the dynamical loaded shared object's
function table is destroyed (somehow) before x is destroyed.

Any comments ? (or even workarounds ?)

-- 
   Johannes




From darrell at dorb.com  Thu Jul  6 02:38:34 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Wed, 5 Jul 2000 17:38:34 -0700
Subject: ASPIPython update on Starship
References: <8jvg2d$8q5$1@nnrp1.deja.com>
Message-ID: <029c01bfe6e2$84dff390$6401a8c0@home>

From: "Sam Schulenburg" <samschul at pacbell.net>
> I have made some corrections to ASPIPython, that correct memory leaks
> found in aspi.dll.
> 
> ASPIPython is a open source python interface to the (A)dvanced (S)
> system (P)programming (I)nterface for SCSI devices developed by Adaptec
> Corp. This interface allows for the development of tests under Python
> for SCSI disk drives. These routines can be found at
> http://starship.python/crew/schulen
> 

This link worked better for me :)
http://starship.python.net/crew/samschul/

--Darrell Gallion




From cut_me_out at hotmail.com  Wed Jul  5 23:38:53 2000
From: cut_me_out at hotmail.com (Alex)
Date: 5 Jul 2000 22:38:53 +0100
Subject: Python Module for sequence alignments.
Message-ID: <etdg0pppszv.fsf@w20-575-32.mit.edu>



Hi.  I made a python module that does simple sequence alignments.  If
there is enough interest, I may make it a little more elaborate.  You
can get it from 

http://puffin.lcs.mit.edu:8080/align.tgz

To use it on linux

tar zxf align.tgz
cd align
make

and make the align directory accessible to python, with a command like

setenv PYTHONPATH `pwd`:${PYTHONPATH}

It needs the Numeric extension.

Here's an example of how it works from the command line:

>>> import align                          
>>> align.align ('aneotuhaneot', 'asneouhoteuh')[0][0][:3]                
('neotuhane', 'sneouhote', '----uh--e')

The second element of the tuple it returns is the matrices used to
calculate the optimal alignments.

I don't really care how you use or change this, nor will I assume any
legal culpability for problems arising from its use.  For instance, if a
bug in this code tricks you into genetically engineering Giant
Super-Intelligent Man-Eating Lizards, I may cackle evilly, but I won't
explicitly own up.  I may take on the care and feeding of one or two of
them if you can prove they will co-exist peacefully with my cat, but
that's definitely as far as it goes.

But please do let me know of any problems with it, or any suggestions
you have about how I could improve it.  If you make any use or
improvements to it yourself, I would love to hear about them.

Alex.




From petersc at stud.ntnu.no  Thu Jul  6 01:44:29 2000
From: petersc at stud.ntnu.no (Peter Schneider-Kamp)
Date: Wed, 05 Jul 2000 23:44:29 +0000
Subject: readline.c & more
References: <Pine.LNX.4.21.0007052214190.4947-100000@krispc6.physik.uni-karlsruhe.de>
Message-ID: <3963C85D.3924126C@stud.ntnu.no>

Johannes Zellner wrote:
> 
>  ?) I had to modify Objects/fileobject.c because there was a
>     name clash between `getline' there and in /usr/include/stdio.h
>     (as it comes with the latest debian potato). I replaced
>     `getline' with `file_getline' in Objects/fileobject.c.

as far as I can see this has been fixed today. look for
revision 2.78 in the CVS repository (or better yet, update).

>  ?) I modified Modules/readline.c so that it allows reading and
>     writing of a history file. I also plan to implement readline
>     functions as `stifle history'. Eventually I'd also add the
>     possiblity to register a script completer which would be
>     called before the builtin completers. My question about
>     this: where would I post patches ? -- I've an account on
>     SF for other projects, but as this is my first mail to the
>     list I'd understand if you won't trust me ;-)

mailto:patches at python.org

for more details check out http://www.python.org/patches

>  ?) Is there a forum for developers? This list seems more for
>     users isn't it ?

there is. but it's hard to get in (still trying). the mailing
list python-dev is probably what you are looking for.

better-history-means-less-typing-ly y'rs
Peter
--
Peter Schneider-Kamp          ++47-7388-7331
Herman Krags veg 51-11        mailto:peter at schneider-kamp.de
N-7050 Trondheim              http://schneider-kamp.de



From bjorn at roguewave.com  Thu Jul  6 00:05:50 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Wed, 05 Jul 2000 16:05:50 -0600
Subject: cleanup order
References: <Pine.LNX.4.21.0007052319220.5959-100000@krispc6.physik.uni-karlsruhe.de>
Message-ID: <3963B13E.8E77CC19@roguewave.com>

Johannes Zellner wrote:
> 
> Hi,
> 
> suppose I've
> 
>     # fred provides `fun'
>     import fred
> 
>     class Lola:
>         def __init__(self, x):
>             self.x = x
>         def __del__(self):
>             fred.fun(self.x)
> 
>     x = Lola('hello')
>     # end of script
> 
> this doesn't work, because fred is not present any more
> when x is destroyed. I don't know anything about python's
> internals, but I guess the dynamical loaded shared object's
> function table is destroyed (somehow) before x is destroyed.
> 
> Any comments ? (or even workarounds ?)

This should work:

   ...
   def __del__(self, fun=fred.fun):
      fun(self.x)

it's using a default argument to make sure the reference count to
fred.fun doesn't reach zero until the Lola instance is deleted.

-- bjorn



From ram_garapaty at spk.agilent.com  Wed Jul  5 23:57:40 2000
From: ram_garapaty at spk.agilent.com (Ram Garapaty)
Date: Wed, 05 Jul 2000 14:57:40 -0700
Subject: serial communication in python..??
Message-ID: <3963AF53.F16F9CAA@spk.agilent.com>

Hi!
i am new to python and would like to know whether python would be any
assistance to my situation as described below:
i got a testing box which send s out the data  through a serial port, i
would like to capture that data through GUI window in PYTHON . Is this
possible in python if soo please suggest me the right topics in python
and right material to go through python asap.

thanks in advance
ram



From ge at nowhere.none  Thu Jul  6 00:04:57 2000
From: ge at nowhere.none (Grant Edwards)
Date: Wed, 05 Jul 2000 22:04:57 GMT
Subject: MAPI/COM: where is the server?
Message-ID: <diO85.3111$iN5.764889@ptah.visi.com>

My MAPI/COM e-mail fetcher (running under win95) does this:

    self.session = win32com.client.Dispatch("MAPI.Session")
    self.session.Logon("MS Eschange Settings")

    [message fetching code left as an exercise for the reader]

This was all developed without any basic understanding of DCOM
or MAPI by copying other people's examples and semi-random
trial-and-error.

Question: Is it talking to Exchange server (which runs on a
          remote machine), or is it talking to a local copy of
          MS Outlook which is then acting as an intermediary?

If it's talking directly to Exchange server, then there's a
possibility I could do the same thing from Linux using either
Software AG's product or GCOM.

If my program is just talking to the local copy of MS Outlook,
then running the program under Linux isn't very useful, since I
still have to run that win95 machine that's sitting over in the
corner.

-- 
Grant Edwards                   grante             Yow!  I have the power
                                  at               to HALT PRODUCTION on all
                               visi.com            TEENAGE SEX COMEDIES!!


From thomas at xs4all.net  Thu Jul  6 00:13:48 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 6 Jul 2000 00:13:48 +0200
Subject: __getattr__ and pickle
In-Reply-To: <vzL85.16702$227.301196@nnrp1.uunet.ca>; from embed@NOSPAM.geocities.com on Wed, Jul 05, 2000 at 02:58:43PM -0400
References: <395f26a4.1832855@news.iol.ie> <395f2748.1996610@news.iol.ie> <395f27ab.2095753@news.iol.ie> <8F655BA3Egmcmhypernetcom@199.171.54.154> <vzL85.16702$227.301196@nnrp1.uunet.ca>
Message-ID: <20000706001348.E13478@xs4all.nl>

On Wed, Jul 05, 2000 at 02:58:43PM -0400, Warren Postma wrote:
> "Gordon McMillan" <gmcm at hypernet.com> wrote in message
> news:8F655BA3Egmcmhypernetcom at 199.171.54.154...

> > C'mon Sean! You've been around a while. You know when __getattr__ hacks go
> > bad it's because they go recursive. (And the stack size on Windows is off
> > by  some tiny amount, so it GPFs instead of traps.)

> That kinda says (a) Windows Sucketh, and (b) surely there must be a
> workaround for the Infinite-Recursion-Crash bug?

Well, there is. There's this little counter that keeps track, but the
recursion counter is set too high for Windows, apparently, so it chooses to
crash before Python stops itself.

> Oh yeah, there is, .... It's called Christian Tismer's Stackless Python.
> Hope that stackless makes it into the official python, by 1.7, or if not,
> there's always 3000.  Here's hoping.

There is no 1.7, the next release is going to be 2.0, not 1.6 ;) At least,
that's what the CVS tree is calling itself. As for Stackless, I agree it's a
thing of coolness, but it won't help you out in infinite recursion; it's
stackless, not limitless. You'll run out of memory eventually.

-- 
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  Thu Jul  6 00:13:46 2000
From: billtut at microsoft.com (Bill Tutt)
Date: Wed, 5 Jul 2000 15:13:46 -0700 
Subject: MAPI/COM: where is the server?
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8101D8BD26F@RED-MSG-50>

Its talking to the local MAPI COM object that talks to the Exchange MAPI
provider which uses whatever method it wants to communicate with the
Exchange server. I believe it uses Win32 RPC.

Cute diagram:
+---------------------+
|  MAPI.Session       |
|  COM Object         |
+---------------------+
|  Core MAPI          |
|     Code            |
+---------------------+  
|  Exchange MAPI      |
|     Provider        |
+---------------------+
|  Win32/DCE RPC      |
+---------------------+
|  Network            |
+---------------------+
|  Exchange Server    |
+---------------------+

Bill

 -----Original Message-----
From: 	ge at nowhere.none [mailto:ge at nowhere.none] 
Sent:	Wednesday, July 05, 2000 3:05 PM
To:	python-list at python.org
Subject:	MAPI/COM: where is the server?

My MAPI/COM e-mail fetcher (running under win95) does this:

    self.session = win32com.client.Dispatch("MAPI.Session")
    self.session.Logon("MS Eschange Settings")

    [message fetching code left as an exercise for the reader]

This was all developed without any basic understanding of DCOM
or MAPI by copying other people's examples and semi-random
trial-and-error.

Question: Is it talking to Exchange server (which runs on a
          remote machine), or is it talking to a local copy of
          MS Outlook which is then acting as an intermediary?

If it's talking directly to Exchange server, then there's a
possibility I could do the same thing from Linux using either
Software AG's product or GCOM.

If my program is just talking to the local copy of MS Outlook,
then running the program under Linux isn't very useful, since I
still have to run that win95 machine that's sitting over in the
corner.

-- 
Grant Edwards                   grante             Yow!  I have the power
                                  at               to HALT PRODUCTION on all
                               visi.com            TEENAGE SEX COMEDIES!!
-- 
http://www.python.org/mailman/listinfo/python-list



From arnold at dstc.monash.edu.au  Thu Jul  6 00:18:44 2000
From: arnold at dstc.monash.edu.au (David Arnold)
Date: Thu, 06 Jul 2000 08:18:44 +1000
Subject: MAPI/COM: where is the server? 
In-Reply-To: Your message of "Wed, 05 Jul 2000 22:04:57 GMT."
             <diO85.3111$iN5.764889@ptah.visi.com> 
Message-ID: <200007052218.IAA26977@xevious.dstc.monash.edu.au>

-->"Grant" == Grant Edwards <ge at nowhere.none> writes:

  Grant> If it's talking directly to Exchange server, then there's a
  Grant> possibility I could do the same thing from Linux using either
  Grant> Software AG's product or GCOM.

what's GCOM?  (sorry, google reveals nothing useful ...)



d



From thomas at xs4all.net  Thu Jul  6 00:25:11 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 6 Jul 2000 00:25:11 +0200
Subject: readline.c & more
In-Reply-To: <3963C85D.3924126C@stud.ntnu.no>; from petersc@stud.ntnu.no on Wed, Jul 05, 2000 at 11:44:29PM +0000
References: <Pine.LNX.4.21.0007052214190.4947-100000@krispc6.physik.uni-karlsruhe.de> <3963C85D.3924126C@stud.ntnu.no>
Message-ID: <20000706002510.F13478@xs4all.nl>

On Wed, Jul 05, 2000 at 11:44:29PM +0000, Peter Schneider-Kamp wrote:

> Johannes Zellner wrote:

> >  ?) I modified Modules/readline.c so that it allows reading and
> >     writing of a history file. I also plan to implement readline
> >     functions as `stifle history'. Eventually I'd also add the
> >     possiblity to register a script completer which would be
> >     called before the builtin completers. My question about
> >     this: where would I post patches ? -- I've an account on
> >     SF for other projects, but as this is my first mail to the
> >     list I'd understand if you won't trust me ;-)

> mailto:patches at python.org
> for more details check out http://www.python.org/patches

Actually, SourceForge is now preferred over patches at python.org.
http://www.python.org/patches also reflects this, by the way ;) Looking at
the number of patches that were submitted (mostly backlog) and processed, I
believe the SF patch manager is working out reasonably OK, even if it has
some oddities. Then again, I only submit the odd patch, dont apply them
myself ;)

Additionally, I think Skip Montanaro (skip at mojam.com) is also working on
making readline/rlcompleter write a history file, though I don't know if
calling external scripts is on his wishlist. You might want to coordinate
with him.

> >  ?) Is there a forum for developers? This list seems more for
> >     users isn't it ?

> there is. but it's hard to get in (still trying). the mailing
> list python-dev is probably what you are looking for.

python-dev at python.org and patches at python.org are both invite-only, but the
archives are public (and they're up to date again, too :):

http://www.python.org/pipermail/python-dev
http://www.python.org/pipermail/patches

If you want to talk shop about a patch you wrote, and you don't feel
confident enough to upload it to SourceForge, python-dev at python.org is the
right email address. Unless there is an active SIG or such on the subject,
of course.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From thomas at xs4all.net  Thu Jul  6 00:46:46 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 6 Jul 2000 00:46:46 +0200
Subject: Overloaded Operators allowed???
In-Reply-To: <8k07ke$j0r@news.or.intel.com>; from no_spam@umich.edu on Wed, Jul 05, 2000 at 01:59:56PM -0700
References: <8k07ke$j0r@news.or.intel.com>
Message-ID: <20000706004646.G13478@xs4all.nl>

On Wed, Jul 05, 2000 at 01:59:56PM -0700, Jeff wrote:

> Can you overload operators as you can in other languages?  For example, can
> I make "&" give the address of something as in C++?  Or can I only overload
> things like the cmp operator?  i.e. I want to say &string and get the
> address of string.

Well, you have two issues here. There is no 'address-of' operator in Python,
so there is nothing to overload. Secondly, overloading in python is more
like overloading _behaviour_ than actually writing your own entirely
different operators. You can overload a number of operations, but you won't
be able to tell in what context they were called. For instance, if you
define a __pow__ overloader, it can be called by 'x ** y' or by 'pow(x,y)',
or by 'pow(x,y,z)'.

Also, because Python doesn't have augmented assignment (+=, >>=, etc) your
overloaders always have to return a new result object, instead of modifying
themselves in-place. So it's theoretically possible to make a file object
that takes arguments like the C++ I/O mechanisms:

x << "spam"

But because << is right-associative, you can't do something like:

x << "spam" << int(count) << "eggs"

unless you provide wrapper objects for all arguments that provide proper
__lshift__ methods ;-P As for augmented assignments, I've written a patch
for that, but it won't be going into Python for a while, if ever. See
http://www.xs4all.nl/~thomas/python/

This isn't really a problem, though, even in large programs: If what you
want to overload an operator with isn't really the same as the operator was
intended for, you shouldn't use an operator, you should use a method. I
remember how scared I was when I read about C++'s operator->*() ;-P

The 'special method names' are explained in the Language Reference:

http://www.python.org/doc/current/ref/specialnames.html

This includes operator overloading, indexing, etcetera. The part about the
methods that overload the usual operators is:

http://www.python.org/doc/current/ref/numeric-types.html

One thing that is missing from Python 1.5.2 is the ability to override the
'in' operator. That's added in Python 2.0 though, using __contains__. In
Python 1.5.2 and earlier, and in Python 2.0 when there is no __contains__
method, 'in' works by repeatedly calling '__getitem__'.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From aaron.ginn at motorola.com  Thu Jul  6 00:09:24 2000
From: aaron.ginn at motorola.com (Aaron Ginn)
Date: 05 Jul 2000 15:09:24 -0700
Subject: Two Tkinter questions (long)
Message-ID: <sn66qkgrfv.fsf@motorola.com>

I have two questions about Tkinter.  First, how do you supress a
command that is executed by a button when a widget is created?  I'm
trying to create a file browser with three buttons, an OK button which
applies the selection, a filter button which filters the current
directory for a matching pattern and a cancel button which does the
obvious.  The problem is that when I create the file browser in a
toplevel widget, the command that my buttons are supposed to execute
are executed immediately.  For instance, here's a portion of code:

class FileBrowser:

    def __init__(self, obj, dir):

	self.list_of_files = []
	self.list_of_dirs  = []

	# Set the default directory to dir.
	self.filter_directory = StringVar()
	self.filter_directory.set(dir)

	# Set defualt filter value to all files in dir.
	self.filter_pattern = StringVar()
	self.filter_pattern.set('*')

	self.selection_value = StringVar()
	self.selection_value.set('None')

	self.top = Toplevel()

< SNIP a lot of wigdets for brevity >

	self.Filter = Button(self.button_frame, text = 'Filter', \
		      command = self.do_filter())
	self.Filter.pack(side = LEFT, padx = 10)


The filter button is bound to a method called do_filter which is as
follows:


    def do_filter(self):
	"""
	Method that generates a list of files and a list of
	directories in a directory given a filter pattern,
	and add them to a listbox for each.
	"""

	filter_path = '%s/%s' %(self.filter_directory.get(), self.filter_pattern.get())
	files_matching_pattern = glob.glob(filter_path)

	# Clear the listboxes
	self.filelist.delete(0,END)
	self.dirlist.delete(0,END)

	# Add . and .. to list of directories.
	self.dirlist.insert(END, '.')
	self.dirlist.insert(END, '..')

	# Fill the listboxes
	for file in files_matching_pattern:
	    [head, tail] = os.path.split(file)
	    if os.path.isfile(file):
		self.filelist.insert(END, tail)
		self.list_of_files.append(tail)
	    elif os.path.isdir(file):
		self.dirlist.insert(END, tail)
		self.list_of_dirs.append(tail)


This works fine, but the command is executed as soon as the widget is
created, and then I can't execute it again.  It is as if the script
has been executed and terminated but the GUI still exists.  Why is the
command that button is bound to executing before I invoke the
button?

This leads to my second question.  I put the command in a lambda
statement to supress it's execution.  When I do this the filter
function works great, and I can execute it multiple times.  However,
when I try to bind the method do_filter to an Event, the event fails
with the following error:


coronado <189> semiDRC.gui -gui
Exception in Tkinter callback
Traceback (innermost last):
  File "/home/ginn/python_tk/lib/python1.5/lib-tk/Tkinter.py", line 764, in __call__
    return apply(self.func, args)
  File "/home/ginn/bin/semiDRC.gui", line 450, in <lambda>
    self.filter_entry.bind('<Return>', lambda s=self: s.do_filter())
AttributeError: do_filter


The bind statement is


self.filter_entry.bind('<Return>', lambda s=self: s.do_filter())


Is there a problem with binding an event to a lambda function?  How do
you do this?  I'm very confused right now!

Aaron

-- 
Aaron J. Ginn                     Motorola SPS
Phone: (480) 814-4463             SemiCustom Solutions
Fax:   (480) 814-4058             1300 N. Alma School Rd.
mailto:aaron.ginn at motorola.com    Chandler, AZ 85226


From ge at nowhere.none  Thu Jul  6 00:49:20 2000
From: ge at nowhere.none (Grant Edwards)
Date: Wed, 05 Jul 2000 22:49:20 GMT
Subject: Two Tkinter questions (long)
References: <sn66qkgrfv.fsf@motorola.com>
Message-ID: <QXO85.3118$iN5.767799@ptah.visi.com>

In article <sn66qkgrfv.fsf at motorola.com>, Aaron Ginn wrote:

>The problem is that when I create the file browser in a
>toplevel widget, the command that my buttons are supposed to execute
>are executed immediately.  For instance, here's a portion of code:
>
>class FileBrowser:
>
>    def __init__(self, obj, dir):
>
>< SNIP a lot of wigdets for brevity >
>
>	self.Filter = Button(self.button_frame, text = 'Filter', \
>		      command = self.do_filter())

    self.Filter = Button(self.button_frame, text = 'Filter', \
                      command = self.do_filter)

>self.filter_entry.bind('<Return>', lambda s=self: s.do_filter())

Functions bound to events need to take an event as a parameter
(if you don't want the event, you can use lambda to discard it):

self.filter_entry.bin('<Return>', labda event: self.do_filter())

or you can define another function that will accept an event
parameter:

def do_filter(self, *args)

Now you can just do

self.filter_entry.bin('<Return>', self.do_filter)


>Is there a problem with binding an event to a lambda function?  

No.  The problem is that normal button command functions are
passed 0 arguments and event functions are passed one argument
(the event).

-- 
Grant Edwards                   grante             Yow!  I am a traffic light,
                                  at               and Alan Ginzberg kidnapped
                               visi.com            my laundry in 1927!


From tim.hochberg at ieee.org  Thu Jul  6 01:06:17 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Wed, 05 Jul 2000 23:06:17 GMT
Subject: Two Tkinter questions (long)
References: <sn66qkgrfv.fsf@motorola.com> <QXO85.3118$iN5.767799@ptah.visi.com>
Message-ID: <JbP85.13738$ll.84983@news1.rdc1.az.home.com>

"Grant Edwards" <ge at nowhere.none> wrote in message
news:QXO85.3118$iN5.767799 at ptah.visi.com...
> In article <sn66qkgrfv.fsf at motorola.com>, Aaron Ginn wrote:
[SNIP]
> Functions bound to events need to take an event as a parameter
> (if you don't want the event, you can use lambda to discard it):
>
> self.filter_entry.bin('<Return>', labda event: self.do_filter())

You still needs to bind self to a default variable in order for it to be
accesible inside the lambda, e.g,

self.filter_entry.bin('<Return>', lambda event, self=self :
self.do_filter())

(I suspect this is just a typo on Grant's part, but I figure I'd clarify it
for Aaron...)

-tim




From kellyk at my-deja.com  Thu Jul  6 00:57:39 2000
From: kellyk at my-deja.com (kellyk at my-deja.com)
Date: Wed, 05 Jul 2000 22:57:39 GMT
Subject: Accessing ASP-IIS objects in modules
Message-ID: <8k0egu$ujn$1@nnrp1.deja.com>

I am using Python in ASP under Microsoft IIS. I can use the IIS objects
like Server, Request and Response easily in my ASP script but cannot
access them from a module.

For example I want to use Response.Write("blah") from within a module
(called from an ASP script) but Reponse is not a known object from
there.

Any ideas?


Sent via Deja.com http://www.deja.com/
Before you buy.


From tim.hochberg at ieee.org  Thu Jul  6 01:11:22 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Wed, 05 Jul 2000 23:11:22 GMT
Subject: Two Tkinter questions (long)
References: <sn66qkgrfv.fsf@motorola.com>
Message-ID: <ugP85.13739$ll.84948@news1.rdc1.az.home.com>

"Aaron Ginn" <aaron.ginn at motorola.com> wrote in message
news:sn66qkgrfv.fsf at motorola.com...
>
> I have two questions about Tkinter.  First, how do you supress a
> command that is executed by a button when a widget is created?  I'm
> trying to create a file browser with three buttons, an OK button which
> applies the selection, a filter button which filters the current
> directory for a matching pattern and a cancel button which does the
> obvious.  The problem is that when I create the file browser in a
> toplevel widget, the command that my buttons are supposed to execute
> are executed immediately.  For instance, here's a portion of code:

[SNIP]

> self.Filter = Button(self.button_frame, text = 'Filter', \
>       command = self.do_filter())

Changing this to:

self.Filter = Button(self.button_frame, text = 'Filter', # Backslash is not
nesc. here.
       command = self.do_filter) # Note no ()

should go a long way toward fixing your problem. What you were doing was
setting command to the result of a call of the function do_filter which
returnd None. Thus the function is executed once when this line is first
executed, but then never again since it is not bound to command, None is.

[SNIP]


-tim




From cjensen at bioeng.ucsd.edu  Thu Jul  6 01:19:58 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Wed, 05 Jul 2000 16:19:58 -0700
Subject: Hello Word in CGI
Message-ID: <3963C29E.76211EB7@bioeng.ucsd.edu>

We get this error in our server log:
Wed Jul  5 16:12:52 2000] [error] [client 132.239.236.47] Premature end
of script headers:
/usr/local/httpd/cgi-bin/test_jmsun/cont_forms/helloworld.py


The browser gives us this error:
Internal Server Error

The server encountered an internal error or misconfiguration and was
unable to complete your request.

Please contact the server administrator, webmaster at bioeng.ucsd.edu and
inform them of the time the error occurred, and
anything you might have done that may have caused the error.

More information about this error may be available in the server error
log.


Apache/1.3.6 Server at cmrg.ucsd.edu Po