🌐 Mailman 3 July 2001 - Pythonmac-SIG - python.org

Re: [Pythonmac-SIG] MacPy 2.1: re module error in IDE
by Jack Jansen July 2, 2001

July 2, 2001

I tested your snippet on MacPython as from the current CVS repository, and while it doesn't crash with the recursion error it also doesn't work: it returns None for the match object (and while I'm not a computer I'm under the distinct impression that it should return a match:-). I've added a comment to your bug report, and I think we should wait to see what the sre guru's have to say about the matter. -- Jack Jansen | ++++ stop the execution of Mumia Abu-Jamal ++++ Jack.Jansen(a)oratrix.com | ++++ if you agree copy these lines to your sig ++++ www.oratrix.nl/~jack | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm

1 0

Re: [Pythonmac-SIG] MacPy 2.1: re module error in IDE
by Damon Butler July 2, 2001

July 2, 2001

Reply to: Re: [Pythonmac-SIG] MacPy 2.1: re module error in IDE Hi, Jack, >- Python 2.0.1 is much newer than Python 2.1, so that it's fixed in the latter >doesn't necessarily mean the fix should be in 2.1. This is I did not know. >Did you check 2.1 on other platforms? I've been trusting my Windows-using Python programmer next cube over, who's regex scripts with sre apparently started working under 2.1 where they didn't under 2.0. I just ran a simple test on WinPy 2.1 myself, however, and was surprised at the results. See below. >- There's two recursion limits in sre. One is the hard limit enforced by the >engine, the other is a check for stack overflow, which will also cause a >recursion limit. So which recursion limit is it that I'm encountering here? (I typed the following lines into the interpreter on both Mac and Win version of Py 2.1.) >>> import sre, pre, string >>> l = ["XXX", "%"*20000, "XXX"] >>> >>> sre_regex = sre.compile(r"XXX.*?XXX") >>> match_object = sre_regex.search(string.join(l)) Traceback (most recent call last): File "<stdin>", line 1, in ? RuntimeError: maximum recursion limit exceeded >>> >>> pre_regex = pre.compile(r"XXX.*XXX") >>> match_object = pre_regex.search(string.join(l)) >>> match_object <pre.MatchObject instance at 0x04619c90> It was this discrepancy between sre and pre that I reported as a bug to SourceForge (#437472), and was the recursion limit error I thought had been fixed. Apparently not in 2.1 on either Mac or Windows, anyway, but is it fixed in 2.0.1? --Damon

1 0

MacPy 2.1: re module error in IDE
by Damon Butler July 2, 2001

July 2, 2001

I just reported a fatal bug (well, fatal for me in my line of work) in the sre module over at SourceForge (#437472). The short of it is that the recursion limit bug present in sre as of Python 2.0 is still present in the MacPython implementation, despite being fixed as of Python 2.0.1 for Windows (and other platforms?). Because of this, I'm forced to revert to use of the pre module instead. Unfortunately, I cannot use the pre module and Just Van Rossum's wonderful IDE at the same time. When pre is used instead of sre, the python interpreter generates an error when booting up the IDE that requires a force-quit before I can proceed to use my Mac in any fashion whatsoever. I've attached a screenshot which lists the entire spiel (MacPyIDE_pre_error.jpg), but the most salient bit (I assume) is the tail end of that spiel, which is as follows: import PyFontify File "EP_Mac_R&D_1:Applications:Python 2.1:Mac:Tools:IDE:Splash.py", line 60, in my__import__ return _real__import__(name) File "EP_Mac_R&D_1:Applications:Python 2.1:Mac:Tools:IDE:PyFontify.py", line 90, in ? matchRE = re.compile(matchPat) File "ep_mac_r&d_1:applications:python 2.1:lib:pre.py", line 245, in compile code=pcre_compile(pattern, flags, groupindex) TypeError: pcre_compile() argument 1 must be string without null bytes, not string I haven't reported this bug to SourceForge yet. It's not clear to me if I should, since the cause of the IDE bug is really a bug with sre that limits its usefulness. The IDE loads fine with sre active, but sre is broken. pre is not broken, but when it's active the IDE does not function. If anyone can respond to this issue, I'd appreciate it, but please note that I'm very much a self-taught, amateur programmer who's really only starting to use Python. And I haven't joined this list yet because the detail of the majority of archived threads are generally beyond my ken. Thanks in advance. Damon Butler ------------ Impressions Book and Journal Services, Inc. (608)244-6218 dbutler(a)impressions.com

2 1

Re: [Pythonmac-SIG] MacPy 2.1: re module error in IDE
by Jack Jansen July 2, 2001

July 2, 2001

Damon, there's two things I can say initially: - Python 2.0.1 is much newer than Python 2.1, so that it's fixed in the latter doesn't necessarily mean the fix should be in 2.1. Did you check 2.1 on other platforms? - There's two recursion limits in sre. One is the hard limit enforced by the engine, the other is a check for stack overflow, which will also cause a recursion limit. There is very little you can do about this, except for using ResEdit or Resorcerer to edit the stack size in the "cfrg" resource of PythonInterpreter and PythonIDE. I suspenc that you are running against the stack overflow. The pre/IDE problem is strange. I'll leave it to Just to comment on this. -- Jack Jansen | ++++ stop the execution of Mumia Abu-Jamal ++++ Jack.Jansen(a)oratrix.com | ++++ if you agree copy these lines to your sig ++++ www.oratrix.nl/~jack | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm

1 0

Re: No os._exit() in MacOS? (fwd)
by David Eppstein July 1, 2001

July 1, 2001

This is regarding some problems I've been having with quitting W apps in MacPython -- Just found a workaround and suggested forwarding to this list. I've put the problematic source online at http://www.ics.uci.edu/~eppstein/noquit.py -- David Eppstein UC Irvine Dept. of Information & Computer Science eppstein(a)ics.uci.edu http://www.ics.uci.edu/~eppstein/ ---------- Forwarded Message ---------- Date: Sunday, July 1, 2001 11:18 PM +0200 From: Just van Rossum <just(a)letterror.com> To: David Eppstein <eppstein(a)ics.uci.edu> Subject: Re: No os._exit() in MacOS? Hi David, > See attached source noquit.py, slightly modified from the example in > http://www.nevada.edu/~cwebster/Python/WWidgets/usingW.html#applications > I also made a file noquit.rsrc copied from > http://www.nevada.edu/~cwebster/Python/WWidgets/MyApp.rsrc.hqx > and saved as an applet from the IDE. > > When I run the resulting applet, cmd-Q quits normally, > but pushing the "Quit" button (which just calls _quit) > results in a frozen applet. Ok, I managed to work around the problem: it seems to be a bug related to the SIOUX output window. Please post a followup to the macpython list, as I have no idea what exactly causes it -- it *is* a bug, and Jack should know about it. Here's the workaround recipe: Drop the resource file belonging to your app (<appname>.rsrc) onto EditPythonPrefs. Choose "Default Startup Options". Choose "Delay console window until needed" as well as "Keep stdio window open on: *never*". (Those are good settings for W apps anyway.) Now build your applet again. Good luck! Just ---------- End Forwarded Message ----------

1 0