]> git.ipfire.org Git - thirdparty/Python/cpython.git/log
thirdparty/Python/cpython.git
22 years agoBackport 1.47 and 1.7
Raymond Hettinger [Fri, 25 Apr 2003 05:31:43 +0000 (05:31 +0000)] 
Backport 1.47 and 1.7
SF Patch 549151: urllib2 POSTs on redirect
(contributed by John J Lee)

22 years agoBackport 1.18:
Raymond Hettinger [Fri, 25 Apr 2003 04:37:40 +0000 (04:37 +0000)] 
Backport 1.18:

SF bug 557704: netrc module can't handle all passwords

Let netrc handle entries with login fields (mail servers for instance)
by having login default to ''.

22 years agoAdd cross-references between urllib.urlencode() and cgi.parse_qs[l]().
Fred Drake [Thu, 24 Apr 2003 16:25:07 +0000 (16:25 +0000)] 
Add cross-references between urllib.urlencode() and cgi.parse_qs[l]().
Closes SF bug #724751.

22 years agoInclude notice that the rotor module will be deprecated in Python 2.3.
Fred Drake [Thu, 24 Apr 2003 13:47:08 +0000 (13:47 +0000)] 
Include notice that the rotor module will be deprecated in Python 2.3.

22 years agoFix documentation errors in regards to urlretrieve() and local files.
Brett Cannon [Thu, 24 Apr 2003 02:46:35 +0000 (02:46 +0000)] 
Fix documentation errors in regards to urlretrieve() and local files.

Also fix typos in urllib lib doc.

22 years agobackport note about password limitations
Skip Montanaro [Wed, 23 Apr 2003 20:36:18 +0000 (20:36 +0000)] 
backport note about password limitations

22 years agofsync(): Implemented for Windows, via calling MS _commit. This counts
Tim Peters [Wed, 23 Apr 2003 20:14:12 +0000 (20:14 +0000)] 
fsync():  Implemented for Windows, via calling MS _commit.  This counts
as "a bug" because there's no other way in core Python to ensure that
bytes written are actually on disk.  At least ZODB wants this guarantee,
for robustness against crashes.

22 years agoBackport 1.17 adding punctuation as allowable characters
Raymond Hettinger [Wed, 23 Apr 2003 19:37:42 +0000 (19:37 +0000)] 
Backport 1.17 adding punctuation as allowable characters

22 years agoPatch #681152: Support escaped unicode characters in classes. Fixes 612074.
Martin v. Löwis [Sat, 19 Apr 2003 08:39:04 +0000 (08:39 +0000)] 
Patch #681152: Support escaped unicode characters in classes. Fixes 612074.

22 years agoPatch #716969: Detect thread creation failure.
Martin v. Löwis [Sat, 19 Apr 2003 07:46:28 +0000 (07:46 +0000)] 
Patch #716969: Detect thread creation failure.

22 years agoPatch #711835: Remove unnecessary lock operations.
Martin v. Löwis [Fri, 18 Apr 2003 11:21:22 +0000 (11:21 +0000)] 
Patch #711835: Remove unnecessary lock operations.

22 years agoAdded documentation for the handle_pi() method, based on SF patch #662464.
Fred Drake [Thu, 17 Apr 2003 22:37:50 +0000 (22:37 +0000)] 
Added documentation for the handle_pi() method, based on SF patch #662464.
Closes SF bug #659188, patch #662464.

22 years agoAs discussed on python-dev, revised docs to note that
Raymond Hettinger [Wed, 16 Apr 2003 17:30:15 +0000 (17:30 +0000)] 
As discussed on python-dev, revised docs to note that
PyObject_IsTrue() can have an error result.

Also, added missing docs for PyObject_Not().

22 years agoBackport:
Neal Norwitz [Fri, 11 Apr 2003 18:21:22 +0000 (18:21 +0000)] 
Backport:
Fix SF bug #697220, string.strip implementation/doc mismatch

Attempt to make all the various string/unicode *strip methods the same.
 * Doc - add doc for when functions were added
 * UserString
 * string/unicode object methods
 * string module functions
'chars' is used for the last parameter everywhere.

22 years agoBackport:
Guido van Rossum [Thu, 10 Apr 2003 20:30:18 +0000 (20:30 +0000)] 
Backport:
PyErr_NormalizeException(): in the type==NULL test, we should simply
return.  Setting an exception can mess with the exception state, and
continuing is definitely wrong (since type is dereferenced later on).
Some code that calls this seems to be prepared for a NULL exception
type, so let's be safe rather than sorry and simply assume there's
nothing to normalize in this case.

22 years agoBackport reference leak fix from HEAD revision 1.79.
Fred Drake [Wed, 9 Apr 2003 18:19:24 +0000 (18:19 +0000)] 
Backport reference leak fix from HEAD revision 1.79.

22 years agoBackport comments about PyObject_Type() (HEAD revisions 1.24, 1.25).
Fred Drake [Wed, 9 Apr 2003 18:17:18 +0000 (18:17 +0000)] 
Backport comments about PyObject_Type() (HEAD revisions 1.24, 1.25).

22 years agoBackport from trunk:
Guido van Rossum [Wed, 9 Apr 2003 17:06:31 +0000 (17:06 +0000)] 
Backport from trunk:
property_traverse() should also traverse into prop_doc -- there's no
typecheck that guarantees it's a string, and BTW string subclasses
could hide references.

22 years agoFixed the gc-vs-__del__ bugs for new-style classes. That's it for this one.
Tim Peters [Tue, 8 Apr 2003 20:33:05 +0000 (20:33 +0000)] 
Fixed the gc-vs-__del__ bugs for new-style classes.  That's it for this one.

22 years agoBackporting new gc-vs-__del__ tests. These pass, but are restricted
Tim Peters [Tue, 8 Apr 2003 19:32:53 +0000 (19:32 +0000)] 
Backporting new gc-vs-__del__ tests.  These pass, but are restricted
to old-style classes.  New-style classes remain vulnerable in 2.2.

22 years agoMore backporting of gc-vs-__del__ fixes. It should be fixed for instances
Tim Peters [Tue, 8 Apr 2003 19:13:14 +0000 (19:13 +0000)] 
More backporting of gc-vs-__del__ fixes.  It should be fixed for instances
of classic classes now.  Alas, new-style classes are still a problem, and
didn't need to be fixed in 2.3 (they were already immune in 2.3 due to the
new-in-2.3 tp_del slot).

22 years agoAdded private API function _PyInstance_Lookup(). This is part of
Tim Peters [Tue, 8 Apr 2003 19:02:34 +0000 (19:02 +0000)] 
Added private API function _PyInstance_Lookup().  This is part of
backporting fixes so that garbage collection doesn't have to trigger
execution of arbitrary Python code just to figure out whether
an object has a __del__ method.

22 years agoAdded example of using positional and keyword args with atexit.register().
Fred Drake [Tue, 8 Apr 2003 17:46:33 +0000 (17:46 +0000)] 
Added example of using positional and keyword args with atexit.register().
Based on a suggestion from a reader.

22 years agoBackporting:
Barry Warsaw [Fri, 4 Apr 2003 02:48:18 +0000 (02:48 +0000)] 
Backporting:

    typed_subpart_iterator(): Fix these to use non-deprecated APIs,
    i.e. get_content_maintype() and get_content_subtype().

22 years agoBackporting email 2.5.1 patches.
Barry Warsaw [Fri, 4 Apr 2003 02:47:07 +0000 (02:47 +0000)] 
Backporting email 2.5.1 patches.

22 years agoBackporting:
Barry Warsaw [Fri, 4 Apr 2003 02:46:38 +0000 (02:46 +0000)] 
Backporting:

    revision 1.27
    date: 2003/03/30 20:46:47;  author: bwarsaw;  state: Exp;  lines: +3 -3
    __unicode__(): Fix the logic for calculating whether to add a
    separating space or not between encoded chunks.  Closes SF bug
    #710498.

22 years agoFix memory corruption in garbage collection.
Jeremy Hylton [Thu, 3 Apr 2003 23:02:29 +0000 (23:02 +0000)] 
Fix memory corruption in garbage collection.

The move_finalizers() routine checks each object in the unreachable
list to see if it has a finalizer.  If it does, it is moved to the
finalizers list.  The collector checks by calling, effectively,
hasattr(obj, "__del__").  The hasattr() call can result in an
arbitrary amount of Python code being run, because it will invoke
getattr hooks on obj.

If a getattr() hook is run from move_finalizers(), it may end up
resurrecting or deallocating objects in the unreachable list.  In
fact, it's possible that the hook causes the next object in the list
to be deallocated.  That is, the object pointed to by gc->gc.gc_next
may be freed before has_finalizer() returns.

The problem with the previous revision is that it followed
gc->gc.gc_next before calling has_finalizer().  If has_finalizer()
gc->happened to deallocate the object FROM_GC(gc->gc.gc_next), then
the next time through the loop gc would point to freed memory.  The
fix is to always follow the next pointer after calling
has_finalizer().

Note that Python 2.3 does not have this problem, because
has_finalizer() checks the tp_del slot and never runs Python code.

Tim, Barry, and I peed away the better part of two days tracking this
down.

22 years agoRemove trailing newline.
Fred Drake [Tue, 1 Apr 2003 15:40:07 +0000 (15:40 +0000)] 
Remove trailing newline.

22 years agoThis was failing on Windows, due to various attempts to delete files
Tim Peters [Mon, 31 Mar 2003 22:48:29 +0000 (22:48 +0000)] 
This was failing on Windows, due to various attempts to delete files
that were still open.  Made the tail end of the test look more like
the CVS head version.

22 years agonote test_locale skip
Skip Montanaro [Mon, 31 Mar 2003 22:12:37 +0000 (22:12 +0000)] 
note test_locale skip

22 years agobackport test skip for test_locale on Mac OS X
Skip Montanaro [Mon, 31 Mar 2003 22:11:45 +0000 (22:11 +0000)] 
backport test skip for test_locale on Mac OS X

22 years agonote test_largefile restriction on Mac OS X
Skip Montanaro [Mon, 31 Mar 2003 21:59:05 +0000 (21:59 +0000)] 
note test_largefile restriction on Mac OS X

22 years agobackport largefile resource requirement for Mac OSX
Skip Montanaro [Mon, 31 Mar 2003 17:42:52 +0000 (17:42 +0000)] 
backport largefile resource requirement for Mac OSX

22 years agoMatch some of the markup changes from the trunk.
Fred Drake [Mon, 31 Mar 2003 16:45:22 +0000 (16:45 +0000)] 
Match some of the markup changes from the trunk.

22 years agoFix symbol in grammar; this should fix some hyperlinking in the HTML
Fred Drake [Mon, 31 Mar 2003 14:54:18 +0000 (14:54 +0000)] 
Fix symbol in grammar; this should fix some hyperlinking in the HTML
version.

22 years agoPatch #710576: Implement per-interpreter-state codec registries.
Martin v. Löwis [Sun, 30 Mar 2003 20:57:31 +0000 (20:57 +0000)] 
Patch #710576: Implement per-interpreter-state codec registries.

22 years agoSF patch #706338, Fix a few broken links in pydoc by Greg Chapman
Neal Norwitz [Sun, 30 Mar 2003 20:31:49 +0000 (20:31 +0000)] 
SF patch #706338, Fix a few broken links in pydoc by Greg Chapman

22 years agoBackport Patch 659834 checked in by GvR on 2002/12/30 16:25:38
Neal Norwitz [Sun, 30 Mar 2003 19:26:50 +0000 (19:26 +0000)] 
Backport Patch 659834 checked in by GvR on 2002/12/30 16:25:38

Check for readline 2.2 features.  This should make it possible to
compile readline.c again with GNU readline versions 2.0 or 2.1; this
ability was removed in readline.c rev. 2.49.  Apparently the older
versions are still in widespread deployment on older Solaris
installations.  With an older readline, completion behavior is subtly
different (a space is always added).

22 years agoPatch #695250: Suppress COPYRIGHT if site.py is not read. Fixes #672614.
Martin v. Löwis [Sun, 30 Mar 2003 17:00:58 +0000 (17:00 +0000)] 
Patch #695250: Suppress COPYRIGHT if site.py is not read. Fixes #672614.

22 years agoBackport Patch 659834 checked in by GvR on 2002/12/30 16:25:38
Neal Norwitz [Sat, 29 Mar 2003 22:25:18 +0000 (22:25 +0000)] 
Backport Patch 659834 checked in by GvR on 2002/12/30 16:25:38

Check for readline 2.2 features.  This should make it possible to
compile readline.c again with GNU readline versions 2.0 or 2.1; this
ability was removed in readline.c rev. 2.49.  Apparently the older
versions are still in widespread deployment on older Solaris
installations.  With an older readline, completion behavior is subtly
different (a space is always added).

22 years agoPatch #707701: Expect '??' in events. Fixes #698517.
Martin v. Löwis [Sat, 29 Mar 2003 09:48:12 +0000 (09:48 +0000)] 
Patch #707701: Expect '??' in events. Fixes #698517.

22 years agobackport fix for missing altsep in nt (bug 709428)
Skip Montanaro [Fri, 28 Mar 2003 22:31:41 +0000 (22:31 +0000)] 
backport fix for missing altsep in nt (bug 709428)

22 years agoBackport Tim's checkin 2.218:
Neal Norwitz [Sun, 23 Mar 2003 14:36:50 +0000 (14:36 +0000)] 
Backport Tim's checkin 2.218:
    slot_sq_contains():  This leaked a reference to the result of calling
    __contains__().

22 years agoBackport 2.220,
Neal Norwitz [Sun, 23 Mar 2003 14:32:56 +0000 (14:32 +0000)] 
Backport 2.220,
SF patch #708201, unchecked return value in import.c by Jason Harper

22 years agoSynch the Windows installer with changes in the email package.
Tim Peters [Sat, 22 Mar 2003 02:08:30 +0000 (02:08 +0000)] 
Synch the Windows installer with changes in the email package.

22 years agoBackporting docs for email 2.5
Barry Warsaw [Fri, 21 Mar 2003 21:24:27 +0000 (21:24 +0000)] 
Backporting docs for email 2.5

22 years agoBackporting email 2.5 to Python 2.2 maint branch.
Barry Warsaw [Fri, 21 Mar 2003 21:09:32 +0000 (21:09 +0000)] 
Backporting email 2.5 to Python 2.2 maint branch.

22 years ago- added example of using a comparison function with list.sort(), and
Fred Drake [Thu, 20 Mar 2003 22:20:43 +0000 (22:20 +0000)] 
- added example of using a comparison function with list.sort(), and
  explained the construction of a [(key, value), ...] list as an
  alternative
- backport additional notes on list use from Python 2.3 documentation;
  mostly warnings about what not to rely on

22 years agoSF bug 705836: struct.pack of floats in non-native endian order
Tim Peters [Thu, 20 Mar 2003 18:31:20 +0000 (18:31 +0000)] 
SF bug 705836: struct.pack of floats in non-native endian order

pack_float, pack_double, save_float:  All the routines for creating
IEEE-format packed representations of floats and doubles simply ignored
that rounding can (in rare cases) propagate out of a long string of
1 bits.  At worst, the end-off carry can (by mistake) interfere with
the exponent value, and then unpacking yields a result wrong by a factor
of 2.  In less severe cases, it can end up losing more low-order bits
than intended, or fail to catch overflow *caused* by rounding.

22 years ago- backport portions of SF patch #700798: fixes and cleanups for descriptor info
Fred Drake [Thu, 20 Mar 2003 18:22:50 +0000 (18:22 +0000)] 
- backport portions of SF patch #700798: fixes and cleanups for descriptor info
- use a TeX "tie" to prevent word-wrapping in "section x.y"-like text

22 years ago- explain what a UNC path is in the makedirs() description, since
Fred Drake [Thu, 20 Mar 2003 17:42:48 +0000 (17:42 +0000)] 
- explain what a UNC path is in the makedirs() description, since
  they're actually mentioned there

22 years agoThis commit was manufactured by cvs2svn to create branch
cvs2svn [Mon, 17 Mar 2003 18:35:42 +0000 (18:35 +0000)] 
This commit was manufactured by cvs2svn to create branch
'release22-maint'.

22 years agobinascii_a2b_base64: Properly return an empty string if the input was all
Thomas Wouters [Mon, 17 Mar 2003 11:34:43 +0000 (11:34 +0000)] 
binascii_a2b_base64: Properly return an empty string if the input was all
    invalid, rather than returning a string of random garbage of the
    estimated result length. Closes SF patch #703471 by Hye-Shik Chang.

Backport from 2.3.

22 years agoAdd some more Distutil changes
Andrew M. Kuchling [Wed, 12 Mar 2003 14:28:21 +0000 (14:28 +0000)] 
Add some more Distutil changes

22 years agoAdd item
Andrew M. Kuchling [Wed, 12 Mar 2003 14:17:38 +0000 (14:17 +0000)] 
Add item

22 years ago[Backport patch #649762] Fix for asynchat endless loop
Andrew M. Kuchling [Wed, 12 Mar 2003 14:11:00 +0000 (14:11 +0000)] 
[Backport patch #649762] Fix for asynchat endless loop
When the null string is used as the terminator, it used to be the same
as None, meaning "collect all the data".  In the current code, however, it
falls into an endless loop; this change reverts to the old behavior.

22 years agoBackport patch from revision 2.80:
Fred Drake [Thu, 6 Mar 2003 16:27:58 +0000 (16:27 +0000)] 
Backport patch from revision 2.80:
Fix memory leak: free memory storing the content model passed to the
ElementDeclHandler by Expat.
Fixes SF bug #676990.

22 years agoSF bug #697220, string.strip implementation/doc mismatch
Neal Norwitz [Tue, 4 Mar 2003 17:44:34 +0000 (17:44 +0000)] 
SF bug #697220, string.strip implementation/doc mismatch

22 years agoBackported fix to [521782] unreliable file.read() error handling.
Gustavo Niemeyer [Tue, 4 Mar 2003 00:50:24 +0000 (00:50 +0000)] 
Backported fix to [521782] unreliable file.read() error handling.

22 years agoget_completer() takes no args
Neal Norwitz [Sat, 1 Mar 2003 15:19:49 +0000 (15:19 +0000)] 
get_completer() takes no args

22 years ago- Backported SF patch #676342: after using pdb, the readline command
Guido van Rossum [Sat, 1 Mar 2003 02:14:53 +0000 (02:14 +0000)] 
- Backported SF patch #676342: after using pdb, the readline command
  completion was botched.

22 years agoBackport rev. 1.73: Translate spaces in the machine name to underscores
Andrew M. Kuchling [Wed, 26 Feb 2003 22:26:03 +0000 (22:26 +0000)] 
Backport rev. 1.73: Translate spaces in the machine name to underscores

22 years agoPartial backport of 2.11: better error messages on import failures.
Jack Jansen [Tue, 25 Feb 2003 14:20:44 +0000 (14:20 +0000)] 
Partial backport of 2.11: better error messages on import failures.
Fixes #652590.

22 years agoRemoved debug output. Fixes #635570
Jack Jansen [Tue, 25 Feb 2003 14:19:07 +0000 (14:19 +0000)] 
Removed debug output. Fixes #635570

22 years agoBackport:
Guido van Rossum [Mon, 24 Feb 2003 01:23:03 +0000 (01:23 +0000)] 
Backport:
Fix from SF patch #633359 by Greg Chapman for SF bug #610299:
    The problem is in sre_compile.py: the call to
    _compile_charset near the end of _compile_info forgets to
    pass in the flags, so that the info charset is not compiled
    with re.U. (The info charset is used when searching to find
    the first character at which a match could start; it is not
    generated for patterns beginning with a repeat like '\w{1}'.)

22 years agoFix SF bug #691793, Python 2.3a2 build fails on Tru64
Neal Norwitz [Sun, 23 Feb 2003 23:45:16 +0000 (23:45 +0000)] 
Fix SF bug #691793, Python 2.3a2 build fails on Tru64

22 years agoBackport relevant portions of:
Neal Norwitz [Sun, 23 Feb 2003 23:34:37 +0000 (23:34 +0000)] 
Backport relevant portions of:

Fix SF bug #691793, Python 2.3a2 build fails on Tru64

Need to make sure that preprocessor directives start in first column.
This means we can't indent code which has preprocessor directives,
nor have a space between [ #include for example.

22 years agoPatch #491107: Cygwin setup.py import workaround patch
Jason Tishler [Fri, 21 Feb 2003 12:18:17 +0000 (12:18 +0000)] 
Patch #491107: Cygwin setup.py import workaround patch

Back patched to the 2.2.x branch.

22 years agoBackport 1.38:
Raymond Hettinger [Fri, 21 Feb 2003 07:16:37 +0000 (07:16 +0000)] 
Backport 1.38:
 Fix an old bug in poll().  When a signal is handled while we're
 blocked in select(), this will raise select.error with errno set to
 EINTR.  The except clauses correctly ignores this error, but the rest
 of the logic will then call read() for all objects in select's *input*
 list of read file descriptors.  Then when an object's read_handler()
 is naive, it will call recv() on its socket, which will raise an
 IOError, and then asyncore decides to close the socket.  To fix this,
 we simply return in this case.
 Backport candidate.

Backport 1.40:
 Fix spelling error

22 years agoSpelling error
Raymond Hettinger [Fri, 21 Feb 2003 06:59:07 +0000 (06:59 +0000)] 
Spelling error

22 years agoTypo
Raymond Hettinger [Fri, 21 Feb 2003 06:05:45 +0000 (06:05 +0000)] 
Typo

22 years agoBackport libtime.tex 1.53:
Raymond Hettinger [Fri, 21 Feb 2003 05:59:16 +0000 (05:59 +0000)] 
Backport libtime.tex 1.53:

    Fixed longstanding bug in the description of strftime's %W code.

22 years agoBackport 1.7: Clarify when a read-only object is created.
Raymond Hettinger [Fri, 21 Feb 2003 04:22:47 +0000 (04:22 +0000)] 
Backport 1.7:  Clarify when a read-only object is created.

22 years agoBackport 1.45: Failed module import was dumping to the console instead
Raymond Hettinger [Fri, 21 Feb 2003 04:18:32 +0000 (04:18 +0000)] 
Backport 1.45:  Failed module import was dumping to the console instead
of creating a nice dialog window.

22 years agoSF bug #685775: turtle circle() documentation error
Raymond Hettinger [Fri, 21 Feb 2003 03:42:28 +0000 (03:42 +0000)] 
SF bug #685775: turtle circle() documentation error

22 years agoBackport of rev 2.199 from trunk.
Guido van Rossum [Wed, 19 Feb 2003 03:52:47 +0000 (03:52 +0000)] 
Backport of rev 2.199 from trunk.

PyObject_Generic{Get,Set}Attr:

Don't access tp_descr_{get,set} of a descriptor without checking the
flag bits of the descriptor's type.  While we know that the main type
(the type of the object whose attribute is being accessed) has all the
right flag bits (or else PyObject_Generic{Get,Set}Attr wouldn't be
called), we don't know that for its class attributes!

22 years agoBackport of rev 2.199 from trunk.
Guido van Rossum [Wed, 19 Feb 2003 03:21:21 +0000 (03:21 +0000)] 
Backport of rev 2.199 from trunk.

PyObject_Generic{Get,Set}Attr:

Don't access tp_descr_{get,set} of a descriptor without checking the
flag bits of the descriptor's type.  While we know that the main type
(the type of the object whose attribute is being accessed) has all the
right flag bits (or else PyObject_Generic{Get,Set}Attr wouldn't be
called), we don't know that for its class attributes!

22 years agoFix SF bug #687655, String formatting conversions misleading
Neal Norwitz [Mon, 17 Feb 2003 18:59:54 +0000 (18:59 +0000)] 
Fix SF bug #687655, String formatting conversions misleading

22 years agobackporting Fred's doc fix.
Skip Montanaro [Fri, 14 Feb 2003 06:46:24 +0000 (06:46 +0000)] 
backporting Fred's doc fix.

22 years agoadd missing description of os.extsep
Skip Montanaro [Fri, 14 Feb 2003 05:46:52 +0000 (05:46 +0000)] 
add missing description of os.extsep

22 years agoBackport 2.217 and 2.218:
Guido van Rossum [Thu, 13 Feb 2003 17:06:02 +0000 (17:06 +0000)] 
Backport 2.217 and 2.218:
Provide access to the import lock, fixing SF bug #580952.  This is
mostly from SF patch #683257, but I had to change unlock_import() to
return an error value to avoid fatal error.

22 years agoBackport SF #660455 fix.
Guido van Rossum [Wed, 12 Feb 2003 19:09:45 +0000 (19:09 +0000)] 
Backport SF #660455 fix.

22 years agoBackport: Add more missing PyErr_NoMemory() after failled memory allocs
Neal Norwitz [Tue, 11 Feb 2003 23:19:35 +0000 (23:19 +0000)] 
Backport: Add more missing PyErr_NoMemory() after failled memory allocs

22 years agoSF bug 684667: Modules/selectmodule.c returns NULL without exception set.
Tim Peters [Tue, 11 Feb 2003 18:05:44 +0000 (18:05 +0000)] 
SF bug 684667: Modules/selectmodule.c returns NULL without exception set.
Backport of fix from head.

22 years agoBackport 1.16: Fix printing plural (s or "").
Guido van Rossum [Mon, 10 Feb 2003 20:48:19 +0000 (20:48 +0000)] 
Backport 1.16: Fix printing plural (s or "").

22 years agoPartial backport for changes to fix SF bug #678518 (assert & global).
Neal Norwitz [Mon, 10 Feb 2003 01:57:51 +0000 (01:57 +0000)] 
Partial backport for changes to fix SF bug #678518 (assert & global).

22 years agoFix SF bug #642168, help() fails for some builtin topics
Neal Norwitz [Fri, 7 Feb 2003 20:51:20 +0000 (20:51 +0000)] 
Fix SF bug #642168, help() fails for some builtin topics

Fix pydoc when doing help for:  and, or, not, UNICODE.

22 years agoSF patch #682514, mmapmodule.c write fix for LP64 executables
Neal Norwitz [Fri, 7 Feb 2003 19:46:44 +0000 (19:46 +0000)] 
SF patch #682514, mmapmodule.c write fix for LP64 executables

Make length an int so we get the right value from
PyArg_ParseTuple(args, "s#", &str, &length)

22 years agoBUILDEXE setting in out-of-tree builds
Skip Montanaro [Fri, 7 Feb 2003 15:59:28 +0000 (15:59 +0000)] 
BUILDEXE setting in out-of-tree builds

22 years agobackport of a tiny part of patch 557719 - just enough to allow the BUILDEXE
Skip Montanaro [Fri, 7 Feb 2003 15:54:52 +0000 (15:54 +0000)] 
backport of a tiny part of patch 557719 - just enough to allow the BUILDEXE
Makefile variable to be set properly when doing out-of-tree builds.

22 years agoFix SF bug #675259, os.environ leaks under FreeBSD and Mac OS X
Neal Norwitz [Fri, 7 Feb 2003 02:29:01 +0000 (02:29 +0000)] 
Fix SF bug #675259, os.environ leaks under FreeBSD and Mac OS X

Even with the extra work to cleanup the env, *BSD still leaks.  Add a note.

22 years agoUpdate to better reflect the usage of struct_time instances throughout;
Fred Drake [Tue, 4 Feb 2003 15:13:25 +0000 (15:13 +0000)] 
Update to better reflect the usage of struct_time instances throughout;
continuing to call these "time tuples" is misleading at best.
Closes SF bug #671731; backported from rev 1.55.

22 years agobackport:
Neal Norwitz [Sun, 2 Feb 2003 19:59:59 +0000 (19:59 +0000)] 
backport:

revision 2.196
date: 2002/12/07 21:39:16;  author: tim_one;  state: Exp;  lines: +27 -28
slot_nb_nonzero():  Another leak uncovered by the sandbox datetime
tests.  I found the logic too confusing to follow here, so rewrote more
than was likely absolutely necessary.

22 years agobackport:
Neal Norwitz [Sun, 2 Feb 2003 19:38:19 +0000 (19:38 +0000)] 
backport:

revision 1.10
date: 2002/10/29 19:08:29;  author: gvanrossum;  state: Exp;  lines: +27 -0
Added test for this fix to classobject.c:

Since properties are supported here, is possible that
instance_getattr2() raises an exception.  Fix all code that made this
assumption.

22 years agobackport:
Neal Norwitz [Sun, 2 Feb 2003 19:37:32 +0000 (19:37 +0000)] 
backport:

revision 2.164
date: 2002/10/29 18:36:40;  author: gvanrossum;  state: Exp;  lines: +12 -13
Since properties are supported here, is possible that
instance_getattr2() raises an exception.  Fix all code that made this
assumption.

22 years agobackport:
Neal Norwitz [Sun, 2 Feb 2003 19:34:40 +0000 (19:34 +0000)] 
backport:

revision 1.14
date: 2003/01/29 14:20:23;  author: mwh;  state: Exp;  lines: +5 -0
Teach the parsermodule about floor division.  Fixes

[ 676521 ] parser module validation failure

22 years agobackport:
Neal Norwitz [Sun, 2 Feb 2003 19:34:14 +0000 (19:34 +0000)] 
backport:

revision 2.75
date: 2003/01/29 14:20:22;  author: mwh;  state: Exp;  lines: +2 -0
Teach the parsermodule about floor division.  Fixes

[ 676521 ] parser module validation failure

22 years agobackport:
Neal Norwitz [Sun, 2 Feb 2003 19:26:25 +0000 (19:26 +0000)] 
backport:

revision 1.14
date: 2003/01/15 23:38:15;  author: mhammond;  state: Exp;  lines: +1 -1
Correct docstring for SetValueEx()

revision 1.13
date: 2002/12/20 20:13:35;  author: theller;  state: Exp;  lines: +1 -1
Fix an error message in the _winreg module. The error message referred
to a constant in the 'win32con' module, but this constant is also
defined in the _winreg module itself.

22 years agobackport:
Neal Norwitz [Sun, 2 Feb 2003 19:25:22 +0000 (19:25 +0000)] 
backport:

revision 1.11
date: 2002/12/18 23:20:39;  author: nnorwitz;  state: Exp;  lines: +6 -2
SF # 654974, fix unchecked return values in structseq

Check return values after memory allocation.

22 years agobackport:
Neal Norwitz [Sun, 2 Feb 2003 19:24:38 +0000 (19:24 +0000)] 
backport:

revision 1.29
date: 2002/11/21 21:08:39;  author: gvanrossum;  state: Exp;  lines: +14 -8
The _Event class should be more careful with releasing its lock when
interrupted.  A try/finally will do nicely.  Maybe other classes need
this too, but since they manipulate more state it's less clear that
that is always the right thing, and I'm in a hurry.