]> granicus.if.org Git - python/log
python
20 years agosome platforms still need offsetof() from structmember.h
Fred Drake [Sun, 17 Oct 2004 19:36:57 +0000 (19:36 +0000)]
some platforms still need offsetof() from structmember.h

20 years agoFix and test weak referencing of itertools.tee objects.
Raymond Hettinger [Sun, 17 Oct 2004 16:40:14 +0000 (16:40 +0000)]
Fix and test weak referencing of itertools.tee objects.

20 years agoFix docstring formatting of escape sequences.
Raymond Hettinger [Sun, 17 Oct 2004 16:36:53 +0000 (16:36 +0000)]
Fix docstring formatting of escape sequences.

20 years agoSF patch #1047577, typo in liblocale.tex
Neal Norwitz [Sun, 17 Oct 2004 16:36:05 +0000 (16:36 +0000)]
SF patch #1047577, typo in liblocale.tex

20 years agoSF Patch #1048341: subprocess documentation, based on PEP/docstring by
Fredrik Lundh [Sun, 17 Oct 2004 16:29:48 +0000 (16:29 +0000)]
SF Patch #1048341: subprocess documentation, based on PEP/docstring by
Peter Astrand, with markup by Fredrik Lundh and Raymond Hettinger.

20 years agoInvalid patterns to substitute and safe_substitute would crash since pattern
Neal Norwitz [Sun, 17 Oct 2004 16:27:18 +0000 (16:27 +0000)]
Invalid patterns to substitute and safe_substitute would crash since pattern
is not a local variable.  Add a test case.

20 years agoUse proper value for False
Neal Norwitz [Sun, 17 Oct 2004 16:24:25 +0000 (16:24 +0000)]
Use proper value for False

20 years agoRemove unnecessary imports
Neal Norwitz [Sun, 17 Oct 2004 16:23:52 +0000 (16:23 +0000)]
Remove unnecessary imports

20 years agoreturn codes are available on all platforms, not just on Unix
Fredrik Lundh [Sun, 17 Oct 2004 14:47:05 +0000 (14:47 +0000)]
return codes are available on all platforms, not just on Unix

20 years agoSF Bug #1046800: Importing Tkinter correctly (thanks Vishnu).
Facundo Batista [Sat, 16 Oct 2004 21:40:35 +0000 (21:40 +0000)]
SF Bug #1046800: Importing Tkinter correctly (thanks Vishnu).

20 years agoThis commit was manufactured by cvs2svn to create tag 'r24b1'. v2.4b1
cvs2svn [Fri, 15 Oct 2004 08:07:21 +0000 (08:07 +0000)]
This commit was manufactured by cvs2svn to create tag 'r24b1'.

20 years agorelease shenanigans
Anthony Baxter [Fri, 15 Oct 2004 08:07:21 +0000 (08:07 +0000)]
release shenanigans

20 years agoApplied patch for [ 1047269 ] Buffer overwrite in PyUnicode_AsWideChar.
Marc-André Lemburg [Fri, 15 Oct 2004 07:45:05 +0000 (07:45 +0000)]
Applied patch for [ 1047269 ] Buffer overwrite in PyUnicode_AsWideChar.

Python 2.3.x candidate.

20 years agomake sure to check for this limit even if we're running with -O
Fredrik Lundh [Fri, 15 Oct 2004 06:15:08 +0000 (06:15 +0000)]
make sure to check for this limit even if we're running with -O

20 years agoBump buildno for 2.4b1
Martin v. Löwis [Fri, 15 Oct 2004 05:44:39 +0000 (05:44 +0000)]
Bump buildno for 2.4b1

20 years agoAdd _subprocess.c
Martin v. Löwis [Fri, 15 Oct 2004 04:27:51 +0000 (04:27 +0000)]
Add _subprocess.c

20 years agoPatch #1020042: Only define HAVE_UINTPTR_T for VC 7.x.
Martin v. Löwis [Fri, 15 Oct 2004 04:26:18 +0000 (04:26 +0000)]
Patch #1020042: Only define HAVE_UINTPTR_T for VC 7.x.

20 years agoSF bug #1046855: httplib index out of range
Raymond Hettinger [Thu, 14 Oct 2004 15:23:38 +0000 (15:23 +0000)]
SF bug #1046855:  httplib index out of range

20 years agoJust remove the #include of signal.h. That it was C++-commented out
Michael W. Hudson [Thu, 14 Oct 2004 13:27:14 +0000 (13:27 +0000)]
Just remove the #include of signal.h.  That it was C++-commented out
was a mistake of mine in updating patch #975056 (I think).

20 years agoPatch 1046644 - improved distutils support for SWIG.
Anthony Baxter [Thu, 14 Oct 2004 10:02:08 +0000 (10:02 +0000)]
Patch 1046644 - improved distutils support for SWIG.

20 years agobump the version number prior to release
Fred Drake [Thu, 14 Oct 2004 05:07:17 +0000 (05:07 +0000)]
bump the version number prior to release

20 years agoremove_stderr_debug_decorations(): Always try the substitution. Else
Tim Peters [Thu, 14 Oct 2004 04:16:54 +0000 (04:16 +0000)]
remove_stderr_debug_decorations():  Always try the substitution.  Else
this test failed under the combination of passing -O to a debug-build
Python.  Now all 4 of those pass ({debug, release} x {-O, no -O}).

20 years agoUse C89 style comment for old compilers
Neal Norwitz [Thu, 14 Oct 2004 03:48:30 +0000 (03:48 +0000)]
Use C89 style comment for old compilers

20 years agoGet test to pass on amd64 (opteron). This is pretty hacky, but
Neal Norwitz [Thu, 14 Oct 2004 03:46:18 +0000 (03:46 +0000)]
Get test to pass on amd64 (opteron).  This is pretty hacky, but
rangeobject.c has an #ifdef that is reached only when LONG_MAX != INT_MAX

20 years agoSF bug 1046690: difflib.HtmlDiff doc errors.
Tim Peters [Thu, 14 Oct 2004 03:09:48 +0000 (03:09 +0000)]
SF bug 1046690:  difflib.HtmlDiff doc errors.

Easy doc corrections from Dan Gass.
Also repaired the LaTeX for optional keyword arguments throughout, which
was (incorrectly) nested instead of flat.

20 years agoRequire minimally PyXML 0.8.4.
Martin v. Löwis [Wed, 13 Oct 2004 19:57:14 +0000 (19:57 +0000)]
Require minimally PyXML 0.8.4.

20 years agoSynchronize with PyXML 1.79:
Martin v. Löwis [Wed, 13 Oct 2004 19:50:11 +0000 (19:50 +0000)]
Synchronize with PyXML 1.79:
73: Restore support for Python 2.1.
74: Factor the previous change differently
79: Conditionalize usage of PyTuple_Pack.

20 years agoReplace dynamic try/except with "if 0", to keep py2exe happy. If you
Fredrik Lundh [Wed, 13 Oct 2004 18:19:18 +0000 (18:19 +0000)]
Replace dynamic try/except with "if 0", to keep py2exe happy.  If you
want to use pywin32 instead of _subprocess, you have to edit the file.

20 years agoPatch 983206: distutils obeys LDSHARED env var. Removed the code in
Anthony Baxter [Wed, 13 Oct 2004 15:54:17 +0000 (15:54 +0000)]
Patch 983206: distutils obeys LDSHARED env var. Removed the code in
Python's own setup.py that did the same thing (and tested on Solaris,
where LDSHARED is needed...)

20 years agoPatch 977343, Solaris likes sys/loadavg.h. Added support for sys/loadavg.h
Anthony Baxter [Wed, 13 Oct 2004 15:30:56 +0000 (15:30 +0000)]
Patch 977343, Solaris likes sys/loadavg.h. Added support for sys/loadavg.h
detection to configure &c.

20 years ago969574
Anthony Baxter [Wed, 13 Oct 2004 14:55:56 +0000 (14:55 +0000)]
969574

20 years agoPatch #975056 - fixes for restartable signals on *BSD. In addition,
Anthony Baxter [Wed, 13 Oct 2004 14:48:50 +0000 (14:48 +0000)]
Patch #975056 - fixes for restartable signals on *BSD. In addition,
a few remaining calls to signal() were converted to PyOS_setsig().

20 years agoFixed a small bug. doctest didn't handle unicode docstrings containing
Jim Fulton [Wed, 13 Oct 2004 14:15:32 +0000 (14:15 +0000)]
Fixed a small bug. doctest didn't handle unicode docstrings containing
non-ascii characters.

20 years agooops. how did _that_ happen?
Anthony Baxter [Wed, 13 Oct 2004 13:22:34 +0000 (13:22 +0000)]
oops. how did _that_ happen?

20 years agoBacking out the basic dependency checking (from pycon sprint).
Anthony Baxter [Wed, 13 Oct 2004 12:35:28 +0000 (12:35 +0000)]
Backing out the basic dependency checking (from pycon sprint).
This support was only a first cut, and doesn't deserve to be in
a released version (where we have to support it in an ongoing
manner)

20 years agoDon't spend quite as much time looking for leaks on Windows, where
Fredrik Lundh [Wed, 13 Oct 2004 07:54:54 +0000 (07:54 +0000)]
Don't spend quite as much time looking for leaks on Windows, where
it's rather expensive to create new processes.

20 years agonormalize case when comparing directory names (problem reported by
Fredrik Lundh [Wed, 13 Oct 2004 06:55:40 +0000 (06:55 +0000)]
normalize case when comparing directory names (problem reported by
"Khalid A. B." on python-dev)

20 years agoremoved info@pythonware.com reference
Fredrik Lundh [Wed, 13 Oct 2004 06:48:37 +0000 (06:48 +0000)]
removed info@pythonware.com reference

(should probably remove all traces of _xmlrpclib, but I'll leave that
for another day.)

20 years agoAdd a comment explaining -kb.
Martin v. Löwis [Wed, 13 Oct 2004 05:29:39 +0000 (05:29 +0000)]
Add a comment explaining -kb.

20 years agoNew helper remove_stderr_debug_decorations(). This test passes in a
Tim Peters [Wed, 13 Oct 2004 04:07:12 +0000 (04:07 +0000)]
New helper remove_stderr_debug_decorations().  This test passes in a
debug build on Windows now.  More applications of the helper may be needed
on non-Windows platforms.

20 years agoKill several problems at once: test_poll() failed sometimes for me.
Tim Peters [Wed, 13 Oct 2004 03:43:40 +0000 (03:43 +0000)]
Kill several problems at once:  test_poll() failed sometimes for me.
Turns out the mysterious "expected output" file contained exactly N dots,
because test_poll() has a loop that *usually* went around N times,
printing one dot on each loop trip.  But there's no guarantee of that,
because the exact value of N depended on the vagaries of scheduling
time.sleep()s across two different processes.  So stopped printing dots,
and got rid of the expected output file.  Add a loop counter instead,
and verify that the loop goes around at least a couple of times.  Also
cut the minimum time needed for this test from 4 seconds to 1.

20 years agotest_stdout_none(): Don't print "banana" to the screen in the middle
Tim Peters [Wed, 13 Oct 2004 03:29:54 +0000 (03:29 +0000)]
test_stdout_none():  Don't print "banana" to the screen in the middle
of the test.  It's testing stdout in a different process, so it has to
print something, but I didn't find "banana" to be self-explanatory.

20 years agoWindows test_creationflags() test: print msg to stderr informing the
Tim Peters [Wed, 13 Oct 2004 03:21:35 +0000 (03:21 +0000)]
Windows test_creationflags() test:  print msg to stderr informing the
tester that a DOS box is expected to flash.  Slash the sleep from 2
seconds to a quarter second (why would we want to wait 2 seconds just
to stare at a DOS box?).

20 years agoFolded long lines.
Tim Peters [Wed, 13 Oct 2004 03:15:00 +0000 (03:15 +0000)]
Folded long lines.

20 years agoXXX about extreme expense of test_no_leaking() on Windows. I'm not sure
Tim Peters [Wed, 13 Oct 2004 03:14:40 +0000 (03:14 +0000)]
XXX about extreme expense of test_no_leaking() on Windows.  I'm not sure
what this is trying to do.  If it's necessary for it to create > 1000
processes, it should be controlled by a new resource and not run by
default on Windows.

20 years agoExperience with Zope2's tests showed it's a Bad Idea to make unittest
Tim Peters [Tue, 12 Oct 2004 22:29:54 +0000 (22:29 +0000)]
Experience with Zope2's tests showed it's a Bad Idea to make unittest
display a test's docstring as "the name" of the test.  So changed most
test docstrings to comments, and removed the clearly useless ones.  Now
unittest reports the actual names of the test methods.

20 years agoWrap long lines.
Tim Peters [Tue, 12 Oct 2004 22:19:32 +0000 (22:19 +0000)]
Wrap long lines.

20 years agoWhitespace normalization.
Tim Peters [Tue, 12 Oct 2004 21:51:32 +0000 (21:51 +0000)]
Whitespace normalization.

20 years agoSupply the _subprocess module under 7.1. I'm not sure what the status
Tim Peters [Tue, 12 Oct 2004 21:48:57 +0000 (21:48 +0000)]
Supply the _subprocess module under 7.1.  I'm not sure what the status
of this should be on non-WIN32 Windows variants.

20 years agoBefore this turns into an unreadable mess, follow PEP 7 by using
Tim Peters [Tue, 12 Oct 2004 21:38:22 +0000 (21:38 +0000)]
Before this turns into an unreadable mess, follow PEP 7 by using
hard tab indents in C code.

20 years agoImprovements when running pdb as a script.
Johannes Gijsbers [Tue, 12 Oct 2004 18:12:09 +0000 (18:12 +0000)]
Improvements when running pdb as a script.

Bug fixes:

* Use fresh copy of globals/locals so the script being debugged can't access
  the pdb namespace (e.g.: p line_prefix will no longer work).

* Remove pdb.py's path from sys.path. Having it in there is normally not a
  problem, but it could prove irritating when messing with PYTHONPATH or
  invoking pdb via /usr/bin/pdf.

* You can now set a breakpoint on the script being debugged, even if the script
  doesn't end with a '.py' extension. Also, setting breakpoints with absolute
  paths now works reliably.

Enhancements:

* Go directly to the first line of the script.

* Enter post-mortem debugging if the script being debugged doesn't catch an
  exception.

* Restart the script being debugged and preserve debugger state when the script
  being debugged exits.

Cleanup:

* Moved the __main__ method into a main() function.

* Kill the (undocumented, not in __all__) mainmodule/mainpyfile globals, add a
  mainpyfile attribute to pdb.

Thanks Ilya Sandler for the patch!

20 years agoTypo fix
Andrew M. Kuchling [Tue, 12 Oct 2004 16:38:42 +0000 (16:38 +0000)]
Typo fix

20 years agoFinish off PEP 324 section; fix Peter's last name
Andrew M. Kuchling [Tue, 12 Oct 2004 16:36:57 +0000 (16:36 +0000)]
Finish off PEP 324 section; fix Peter's last name

20 years agoStart section for PEP 324
Andrew M. Kuchling [Tue, 12 Oct 2004 15:58:02 +0000 (15:58 +0000)]
Start section for PEP 324

20 years agoFix PEP number
Andrew M. Kuchling [Tue, 12 Oct 2004 15:32:10 +0000 (15:32 +0000)]
Fix PEP number

20 years agoAdded Peter Astrand's subprocess module.
Fredrik Lundh [Tue, 12 Oct 2004 15:26:28 +0000 (15:26 +0000)]
Added Peter Astrand's subprocess module.

20 years agoDon't use mutable values for method defaults.
Raymond Hettinger [Tue, 12 Oct 2004 09:12:16 +0000 (09:12 +0000)]
Don't use mutable values for method defaults.

20 years agoAdd two items
Andrew M. Kuchling [Mon, 11 Oct 2004 19:20:06 +0000 (19:20 +0000)]
Add two items

20 years agoDocument that on Unix, the 'cmd' argument to the os.popen2/3/4 and
Johannes Gijsbers [Mon, 11 Oct 2004 18:12:20 +0000 (18:12 +0000)]
Document that on Unix, the 'cmd' argument to the os.popen2/3/4 and
popen2.popen2/3/4 functions can be a sequence. All texts are a variation on the
following:

On \UNIX, \var{cmd} may be a sequence, in which case arguments will be passed
directly to the program without shell intervention (as with
\function{os.spawnv()}). If \var{cmd} is a string it will be passed to the shell
(as with \function{os.system()}).

20 years agoThis is jiwon's patch to fix:
Michael W. Hudson [Mon, 11 Oct 2004 15:35:53 +0000 (15:35 +0000)]
This is jiwon's patch to fix:

1042238 ] Lib/compiler chokes on certain genexps

20 years agoOpen source files in universal newlines mode.
Michael W. Hudson [Mon, 11 Oct 2004 15:34:31 +0000 (15:34 +0000)]
Open source files in universal newlines mode.

20 years agoAdded note about the new usegmt argument to email.Utils.formatdate().
Barry Warsaw [Mon, 11 Oct 2004 14:32:47 +0000 (14:32 +0000)]
Added note about the new usegmt argument to email.Utils.formatdate().

20 years agoAdded a usegmt flag to email.Utils.formatdate - this allows it to be
Anthony Baxter [Mon, 11 Oct 2004 13:53:08 +0000 (13:53 +0000)]
Added a usegmt flag to email.Utils.formatdate - this allows it to be
used to replace rfc822.formatdate for protocols like HTTP (where 'GMT' must
be the timezone string).

20 years agoSF patch 1044089: New C API function PyEval_ThreadsInitialized(), by Nick
Tim Peters [Mon, 11 Oct 2004 02:40:51 +0000 (02:40 +0000)]
SF patch 1044089:  New C API function PyEval_ThreadsInitialized(), by Nick
Coghlan, for determining whether PyEval_InitThreads() has been called.
Also purged the undocumented+unused _PyThread_Started int.

20 years agoRevert rev 2.35. It was based on erroneous reasoning -- the current
Tim Peters [Sun, 10 Oct 2004 05:30:40 +0000 (05:30 +0000)]
Revert rev 2.35.  It was based on erroneous reasoning -- the current
thread's id can't get duplicated, because (of course!) the current thread
is still running.  The code should work either way, but reverting the
gratuitous change should make backporting easier, and gets the bad
reasoning out of 2.35's new comments.

20 years agoPyInterpreterState_New(), PyThreadState_New(): use malloc/free directly.
Tim Peters [Sun, 10 Oct 2004 02:47:33 +0000 (02:47 +0000)]
PyInterpreterState_New(), PyThreadState_New():  use malloc/free directly.

This appears to finish repairs for SF bug 1041645.

This is a critical bugfix.

20 years agofind_key(): This routine wasn't thread-correct, and accounts for the
Tim Peters [Sun, 10 Oct 2004 01:58:44 +0000 (01:58 +0000)]
find_key():  This routine wasn't thread-correct, and accounts for the
release-build failures noted in bug 1041645.

This is a critical bugfix.  I'm not going to backport it, though (no time).

20 years agoPyGILState_Release(): If we need to delete the TLS entry for this thread,
Tim Peters [Sat, 9 Oct 2004 23:55:36 +0000 (23:55 +0000)]
PyGILState_Release():  If we need to delete the TLS entry for this thread,
that must be done under protection of the GIL, for reasons explained in
new comments.

20 years agoAll known bugs are closed, and Python 2.4b1 is coming out soon, so bump
Barry Warsaw [Sat, 9 Oct 2004 23:01:11 +0000 (23:01 +0000)]
All known bugs are closed, and Python 2.4b1 is coming out soon, so bump
email's version number to 3.0b1.

20 years agoFix SF bug # 1030941. In _parsegen(), in the clause where we're
Barry Warsaw [Sat, 9 Oct 2004 23:00:11 +0000 (23:00 +0000)]
Fix SF bug # 1030941.  In _parsegen(), in the clause where we're
capturing_preamble but we found a StartBoundaryNotFoundDefect, we need to
consume all lines from the current position to the EOF, which we'll set as the
epilogue of the current message.  If we're not at EOF when we return from
here, the outer message's capturing_preamble assertion will fail.

20 years agoAn example message for SF bug # 1030941.
Barry Warsaw [Sat, 9 Oct 2004 22:57:57 +0000 (22:57 +0000)]
An example message for SF bug # 1030941.

20 years agotest_missing_start_boundary(): A test for SF bug # 1030941.
Barry Warsaw [Sat, 9 Oct 2004 22:57:33 +0000 (22:57 +0000)]
test_missing_start_boundary(): A test for SF bug # 1030941.

20 years ago_PyGILState_Init(), PyGILState_Ensure(): Since PyThread_set_key_value()
Tim Peters [Sat, 9 Oct 2004 22:47:13 +0000 (22:47 +0000)]
_PyGILState_Init(), PyGILState_Ensure():  Since PyThread_set_key_value()
can fail, check its return value, and die if it does fail.

_PyGILState_Init():  Assert that the thread doesn't already have an
association for autoTLSkey.  If it does, PyThread_set_key_value() will
ignore the attempt to (re)set the association, which the code clearly
doesn't want.

20 years agoDocument the results of painful reverse-engineering of the "portable TLS"
Tim Peters [Sat, 9 Oct 2004 22:33:09 +0000 (22:33 +0000)]
Document the results of painful reverse-engineering of the "portable TLS"
code.

PyThread_set_key_value():  It's clear that this code assumes the passed-in
value isn't NULL, so document that it must not be, and assert that it
isn't.  It remains unclear whether existing callers want the odd semantics
actually implemented by this function.

20 years agoFix for SF bug # 1010102. The default is PureProxy not SMTPProxy.
Barry Warsaw [Sat, 9 Oct 2004 21:44:13 +0000 (21:44 +0000)]
Fix for SF bug # 1010102.  The default is PureProxy not SMTPProxy.

20 years ago__init__(): Coerce the input_charset to unicode (with ascii encoding) before
Barry Warsaw [Sat, 9 Oct 2004 21:08:30 +0000 (21:08 +0000)]
__init__(): Coerce the input_charset to unicode (with ascii encoding) before
calling .lower() on it.  This fixes the problem described in SF patch # 866982
where in the tr_TR.ISO-8859-9 locale, 'I'.lower() isn't 'i'.  unicodes are
locale insensitive.

20 years agoStyle guide & consistency changes. No semantic changes.
Tim Peters [Sat, 9 Oct 2004 17:38:29 +0000 (17:38 +0000)]
Style guide & consistency changes.  No semantic changes.

20 years agoTrim trailing whitespace.
Tim Peters [Sat, 9 Oct 2004 17:25:05 +0000 (17:25 +0000)]
Trim trailing whitespace.

20 years agoSimplify delitem() code by calling rotate() directly instead of using
Raymond Hettinger [Sat, 9 Oct 2004 16:02:18 +0000 (16:02 +0000)]
Simplify delitem() code by calling rotate() directly instead of using
arguments passed through tuples.

20 years agoPatch #1042969: add closing brace for \token {generator_expression}. Thanks
Johannes Gijsbers [Sat, 9 Oct 2004 15:52:04 +0000 (15:52 +0000)]
Patch #1042969: add closing brace for \token {generator_expression}. Thanks
George Yoshida!

20 years agoSF patch #1043218
Raymond Hettinger [Sat, 9 Oct 2004 07:10:44 +0000 (07:10 +0000)]
SF patch #1043218
Simplify internal calls and logic for _fix() and _fixexponents().
(Contributed by Facundo Batista.)

20 years ago[Bug #1022311] curses module uses y,x ordering of arguments, not x,y
Andrew M. Kuchling [Fri, 8 Oct 2004 18:48:43 +0000 (18:48 +0000)]
[Bug #1022311] curses module uses y,x ordering of arguments, not x,y

20 years ago[Bug #1041501] Fix example code
Andrew M. Kuchling [Fri, 8 Oct 2004 18:34:47 +0000 (18:34 +0000)]
[Bug #1041501] Fix example code

20 years ago[Bug #1031897] Fix order of decorator application
Andrew M. Kuchling [Fri, 8 Oct 2004 18:29:29 +0000 (18:29 +0000)]
[Bug #1031897] Fix order of decorator application

20 years agoFix bug in _checkquote that raised an exception on empty "arg".
Piers Lauder [Fri, 8 Oct 2004 04:05:39 +0000 (04:05 +0000)]
Fix bug in _checkquote that raised an exception on empty "arg".

20 years agoSF patch #1041364: True/False instead of 1/0 in libstdtypes.tex
Raymond Hettinger [Fri, 8 Oct 2004 01:52:15 +0000 (01:52 +0000)]
SF patch #1041364: True/False instead of 1/0 in libstdtypes.tex
(Contributed by Gerrit Holl.  Reviewed by Jeff Epler.)

20 years agoRewrite rmtree using os.walk to fix bug #1025127:
Johannes Gijsbers [Thu, 7 Oct 2004 21:10:08 +0000 (21:10 +0000)]
Rewrite rmtree using os.walk to fix bug #1025127:

The shutils.rmtree() implementation uses an excessive amount of memory when
deleting large directory hierarchies. Before actually deleting any files, it
builds up a list of (function, filename) tuples for all the files that it is
going to remove.

20 years agoThis is Michiel de Hoon's patch, as attached to the bug report:
Michael W. Hudson [Thu, 7 Oct 2004 13:46:33 +0000 (13:46 +0000)]
This is Michiel de Hoon's patch, as attached to the bug report:

1030629 ] PyOS_InputHook broken

with a couple of utterly inconsequential changes by me.

20 years agoAdd helpful comment
Andrew M. Kuchling [Thu, 7 Oct 2004 12:30:54 +0000 (12:30 +0000)]
Add helpful comment

20 years agoTake out Guido's name and put in the PSF
Andrew M. Kuchling [Thu, 7 Oct 2004 12:27:31 +0000 (12:27 +0000)]
Take out Guido's name and put in the PSF

20 years agoUpdate two URLs
Andrew M. Kuchling [Thu, 7 Oct 2004 12:23:12 +0000 (12:23 +0000)]
Update two URLs

20 years agoAdd -m to man page
Andrew M. Kuchling [Thu, 7 Oct 2004 12:04:50 +0000 (12:04 +0000)]
Add -m to man page

20 years agoAdd item
Andrew M. Kuchling [Thu, 7 Oct 2004 12:00:33 +0000 (12:00 +0000)]
Add item

20 years agoSF patch #1035498: -m option to run a module as a script
Raymond Hettinger [Thu, 7 Oct 2004 06:46:25 +0000 (06:46 +0000)]
SF patch #1035498:  -m option to run a module as a script
(Contributed by Nick Coghlan.)

20 years agoFinalize the freelist of list objects.
Raymond Hettinger [Thu, 7 Oct 2004 03:58:07 +0000 (03:58 +0000)]
Finalize the freelist of list objects.

20 years agoFix bug introduced by the previous patch by changing TimeRE.__seqToRe() to
Brett Cannon [Wed, 6 Oct 2004 22:48:58 +0000 (22:48 +0000)]
Fix bug introduced by the previous patch by changing TimeRE.__seqToRe() to
accept any iterable instead of only a sliceable object.

20 years agoArmin's patch to prevent overflows.
Raymond Hettinger [Wed, 6 Oct 2004 17:51:54 +0000 (17:51 +0000)]
Armin's patch to prevent overflows.

20 years agoConvert a listcomp to a gencomp (was already editing code).
Brett Cannon [Wed, 6 Oct 2004 02:23:14 +0000 (02:23 +0000)]
Convert a listcomp to a gencomp (was already editing code).

20 years agoLocale data that contains regex metacharacters are now properly escaped.
Brett Cannon [Wed, 6 Oct 2004 02:11:37 +0000 (02:11 +0000)]
Locale data that contains regex metacharacters are now properly escaped.

Closes bug #1039270.

20 years agoAdd some items
Andrew M. Kuchling [Tue, 5 Oct 2004 20:23:34 +0000 (20:23 +0000)]
Add some items