]> git.ipfire.org Git - thirdparty/Python/cpython.git/log
thirdparty/Python/cpython.git
11 years agoIssue 21424: Apply the nlargest() optimizations to nsmallest() as well.
Raymond Hettinger [Sun, 11 May 2014 21:21:23 +0000 (14:21 -0700)] 
Issue 21424:  Apply the nlargest() optimizations to nsmallest() as well.

11 years agomerge from 3.4 - clean up the subprocess docs warning-soup and
Gregory P. Smith [Sun, 11 May 2014 20:29:36 +0000 (13:29 -0700)] 
merge from 3.4 - clean up the subprocess docs warning-soup and
s/Unix/POSIX/.

11 years agoChange all references to Unix to POSIX in the subprocess docs. It's
Gregory P. Smith [Sun, 11 May 2014 20:28:35 +0000 (13:28 -0700)] 
Change all references to Unix to POSIX in the subprocess docs.  It's
more accurate and sounds less like a strange tale of yore.

11 years agoRemove the warning-soup from the subprocess documentation by adding
Gregory P. Smith [Sun, 11 May 2014 20:26:21 +0000 (13:26 -0700)] 
Remove the warning-soup from the subprocess documentation by adding
a Security Considerations section as preferred by both the devguide
and documentation users who do not wish to go insane.

11 years agomerge 3.4
Benjamin Peterson [Sun, 11 May 2014 18:19:17 +0000 (11:19 -0700)] 
merge 3.4

11 years agoremove confusing delete indexing (closes #21466)
Benjamin Peterson [Sun, 11 May 2014 18:18:51 +0000 (11:18 -0700)] 
remove confusing delete indexing (closes #21466)

11 years agoNull merge
Antoine Pitrou [Sun, 11 May 2014 17:39:13 +0000 (19:39 +0200)] 
Null merge

11 years agoTry to fix issue #21425 workaround for shared library builds
Antoine Pitrou [Sun, 11 May 2014 17:13:43 +0000 (19:13 +0200)] 
Try to fix issue #21425 workaround for shared library builds

11 years agoTry to fix issue #21425 workaround for shared library builds
Antoine Pitrou [Sun, 11 May 2014 17:13:43 +0000 (19:13 +0200)] 
Try to fix issue #21425 workaround for shared library builds

11 years agoFix printing out error message when test fails and run with -bb
Antoine Pitrou [Sun, 11 May 2014 17:05:50 +0000 (19:05 +0200)] 
Fix printing out error message when test fails and run with -bb

11 years agoFix printing out error message when test fails and run with -bb
Antoine Pitrou [Sun, 11 May 2014 17:05:23 +0000 (19:05 +0200)] 
Fix printing out error message when test fails and run with -bb

11 years agoNull merge
Antoine Pitrou [Sun, 11 May 2014 15:30:41 +0000 (17:30 +0200)] 
Null merge

11 years agoRemove debugging output
Antoine Pitrou [Sun, 11 May 2014 15:29:57 +0000 (17:29 +0200)] 
Remove debugging output

11 years agoTry workaround for test issues in #21425
Antoine Pitrou [Sun, 11 May 2014 14:59:16 +0000 (16:59 +0200)] 
Try workaround for test issues in #21425

11 years agoTry workaround for test issues in #21425
Antoine Pitrou [Sun, 11 May 2014 14:59:16 +0000 (16:59 +0200)] 
Try workaround for test issues in #21425

11 years agoFix debugging output to work with -W
Antoine Pitrou [Sun, 11 May 2014 14:36:22 +0000 (16:36 +0200)] 
Fix debugging output to work with -W

11 years agoFix debugging output to work with -bb
Antoine Pitrou [Sun, 11 May 2014 14:24:45 +0000 (16:24 +0200)] 
Fix debugging output to work with -bb

11 years agoAdd debugging output for #21425
Antoine Pitrou [Sun, 11 May 2014 14:09:15 +0000 (16:09 +0200)] 
Add debugging output for #21425

11 years agoIssue #21425: Fix flushing of standard streams in the interactive interpreter.
Antoine Pitrou [Sun, 11 May 2014 11:43:31 +0000 (13:43 +0200)] 
Issue #21425: Fix flushing of standard streams in the interactive interpreter.

11 years agoIssue #21425: Fix flushing of standard streams in the interactive interpreter.
Antoine Pitrou [Sun, 11 May 2014 11:42:17 +0000 (13:42 +0200)] 
Issue #21425: Fix flushing of standard streams in the interactive interpreter.

11 years agoBy popular demand, add back a security warning.
Raymond Hettinger [Sun, 11 May 2014 09:26:23 +0000 (02:26 -0700)] 
By popular demand, add back a security warning.

11 years agoIssue #21424: Optimize heaqp.nlargest() to make fewer tuple comparisons.
Raymond Hettinger [Sun, 11 May 2014 08:55:46 +0000 (01:55 -0700)] 
Issue #21424:  Optimize heaqp.nlargest() to make fewer tuple comparisons.

Consolidates the logic for nlargest() into a single function so that
decoration tuples (elem,order) or (key, order, elem) only need to
be formed when a new element is added to the heap.  Formerly, a tuple
was created for every element regardless of whether it was added to
the heap.

The change reduces the number of tuples created, the number of ordering
integers created, and total number of tuple comparisons.

11 years agoMerge 3.4 -> default: asyncio: Upstream issue #167: remove dead code, by Marc Schlaich.
Guido van Rossum [Sat, 10 May 2014 22:48:03 +0000 (15:48 -0700)] 
Merge 3.4 -> default: asyncio: Upstream issue #167: remove dead code, by Marc Schlaich.

11 years agoasyncio: Upstream issue #167: remove dead code, by Marc Schlaich.
Guido van Rossum [Sat, 10 May 2014 22:47:15 +0000 (15:47 -0700)] 
asyncio: Upstream issue #167: remove dead code, by Marc Schlaich.

11 years agoRemove the redundant and poorly worded warning message.
Raymond Hettinger [Sat, 10 May 2014 21:05:28 +0000 (14:05 -0700)] 
Remove the redundant and poorly worded warning message.

The paragraph above already says, clearly and correctly, that
"However, being completely deterministic, it is not suitable for
all purposes, and is completely unsuitable for cryptographic purposes."

Also we should make any promises about SystemRandom or os.urandom()
being cryptographically secure (they may be, but be can't validate
that promise).  Further, those are actual random number generators
not psuedo-random number generators.

11 years agoReindent long line
Jason R. Coombs [Sat, 10 May 2014 17:24:58 +0000 (13:24 -0400)] 
Reindent long line

11 years agoReplace overly-aggressive comparison for type equality with an isinstance check.
Jason R. Coombs [Sat, 10 May 2014 17:24:18 +0000 (13:24 -0400)] 
Replace overly-aggressive comparison for type equality with an isinstance check.

11 years agoDrop support for Python 2.4 in upload command.
Jason R. Coombs [Sat, 10 May 2014 17:22:43 +0000 (13:22 -0400)] 
Drop support for Python 2.4 in upload command.

11 years agoReplace import * with explicit import
Jason R. Coombs [Sat, 10 May 2014 17:21:02 +0000 (13:21 -0400)] 
Replace import * with explicit import

11 years agoClean up style in distutils upload command
Jason R. Coombs [Sat, 10 May 2014 17:20:28 +0000 (13:20 -0400)] 
Clean up style in distutils upload command

11 years agoIssue #19655: Add tests for the new asdl parser.
Eli Bendersky [Sat, 10 May 2014 02:03:25 +0000 (19:03 -0700)] 
Issue #19655: Add tests for the new asdl parser.

This unit test runs only for source builds of Python, similarly to test_tools.

11 years agoIssue #19655: Replace the ASDL parser carried with CPython
Eli Bendersky [Sat, 10 May 2014 00:58:22 +0000 (17:58 -0700)] 
Issue #19655: Replace the ASDL parser carried with CPython

The new parser does not rely on Spark (which is now removed from our repo),
uses modern 3.x idioms and is significantly smaller and simpler.

It generates exactly the same AST files (.h and .c), so in practice no builds
should be affected.

11 years agoIssue #19721: Consolidate test_importlib utility code into a single
Brett Cannon [Fri, 9 May 2014 18:32:57 +0000 (14:32 -0400)] 
Issue #19721: Consolidate test_importlib utility code into a single
module.

11 years agoIssue #20776: Flesh out tests for importlib.machinery.PathFinder.
Brett Cannon [Fri, 9 May 2014 17:38:11 +0000 (13:38 -0400)] 
Issue #20776: Flesh out tests for importlib.machinery.PathFinder.

11 years agoIssue21452 Add missing backslash to build path for make_buildinfo
Tim Golden [Fri, 9 May 2014 17:19:31 +0000 (18:19 +0100)] 
Issue21452 Add missing backslash to build path for make_buildinfo

11 years agoMerge
Tim Golden [Fri, 9 May 2014 17:19:13 +0000 (18:19 +0100)] 
Merge

11 years agoIssue21452 Add missing backslash to build path for make_buildinfo
Tim Golden [Fri, 9 May 2014 17:18:11 +0000 (18:18 +0100)] 
Issue21452 Add missing backslash to build path for make_buildinfo

11 years agoIssue10752 Be more robust when finding a PERL interpreter to build OpenSSL. Initial...
Tim Golden [Fri, 9 May 2014 17:01:44 +0000 (18:01 +0100)] 
Issue10752 Be more robust when finding a PERL interpreter to build OpenSSL. Initial patch by Gabi Davar

11 years agoIssue10752 Be more robust when finding a PERL interpreter to build OpenSSL. Initial...
Tim Golden [Fri, 9 May 2014 17:01:19 +0000 (18:01 +0100)] 
Issue10752 Be more robust when finding a PERL interpreter to build OpenSSL. Initial patch by Gabi Davar

11 years agoIssue #21156: importlib.abc.InspectLoader.source_to_code() is now a
Brett Cannon [Fri, 9 May 2014 16:28:22 +0000 (12:28 -0400)] 
Issue #21156: importlib.abc.InspectLoader.source_to_code() is now a
staticmethod.

11 years agoMerge for issue #21438
Brett Cannon [Fri, 9 May 2014 15:56:07 +0000 (11:56 -0400)] 
Merge for issue #21438

11 years agoIssue #21438: Document what loaders don't require a module name for
Brett Cannon [Fri, 9 May 2014 15:55:49 +0000 (11:55 -0400)] 
Issue #21438: Document what loaders don't require a module name for
load_module().

11 years agoMerge for issue #21157
Brett Cannon [Fri, 9 May 2014 14:37:55 +0000 (10:37 -0400)] 
Merge for issue #21157

11 years agoIssue #21157: Touch up imp docs to be more explicit about importlib
Brett Cannon [Fri, 9 May 2014 14:37:31 +0000 (10:37 -0400)] 
Issue #21157: Touch up imp docs to be more explicit about importlib
alternatives.

11 years agoIssue #21141: The Windows build process no longer attempts to find Perl,
Zachary Ware [Fri, 9 May 2014 14:07:50 +0000 (09:07 -0500)] 
Issue #21141: The Windows build process no longer attempts to find Perl,
instead relying on OpenSSL source being configured and ready to build.  The
``PCbuild\build_ssl.py`` script has been re-written and re-named to
``PCbuild\prepare_ssl.py``, and takes care of configuring OpenSSL source
for both 32 and 64 bit platforms.  OpenSSL sources obtained from
svn.python.org will always be pre-configured and ready to build.

11 years agoMerge from 3.4.
Tim Peters [Thu, 8 May 2014 22:43:25 +0000 (17:43 -0500)] 
Merge from 3.4.

Issue #21435: Segfault in gc with cyclic trash
Changed the iteration logic in finalize_garbage() to tolerate objects vanishing
from the list as a side effect of executing a finalizer.

11 years agoIssue #21435: Segfault in gc with cyclic trash
Tim Peters [Thu, 8 May 2014 22:42:19 +0000 (17:42 -0500)] 
Issue #21435: Segfault in gc with cyclic trash
Changed the iteration logic in finalize_garbage() to tolerate objects vanishing
from the list as a side effect of executing a finalizer.

11 years agoIssue #21396: Fix TextIOWrapper(..., write_through=True) to not force a flush() on...
Antoine Pitrou [Thu, 8 May 2014 22:31:32 +0000 (00:31 +0200)] 
Issue #21396: Fix TextIOWrapper(..., write_through=True) to not force a flush() on the underlying binary stream.

Patch by akira.

11 years agoIssue #21396: Fix TextIOWrapper(..., write_through=True) to not force a flush() on...
Antoine Pitrou [Thu, 8 May 2014 22:24:50 +0000 (00:24 +0200)] 
Issue #21396: Fix TextIOWrapper(..., write_through=True) to not force a flush() on the underlying binary stream.

Patch by akira.

11 years agoIssue #21037: Add a build option to enable AddressSanitizer support.
Charles-François Natali [Thu, 8 May 2014 22:08:51 +0000 (23:08 +0100)] 
Issue #21037: Add a build option to enable AddressSanitizer support.

11 years agoMerge #21300: fix typo
R David Murray [Thu, 8 May 2014 14:06:17 +0000 (10:06 -0400)] 
Merge #21300: fix typo

11 years ago#21300: fix typo
R David Murray [Thu, 8 May 2014 14:05:47 +0000 (10:05 -0400)] 
#21300: fix typo

11 years agoMerge with 3.4
Kristján Valur Jónsson [Thu, 8 May 2014 11:18:27 +0000 (11:18 +0000)] 
Merge with 3.4

11 years agoMerging from 3.3: The PyCOND_TIMEDWAIT must use microseconds for the timeout argument
Kristján Valur Jónsson [Thu, 8 May 2014 10:59:52 +0000 (10:59 +0000)] 
Merging from 3.3: The PyCOND_TIMEDWAIT must use microseconds for the timeout argument
in order to have the same resolution as pthreads condition variables.
At the same time, it must be large enough to accept 31 bits of
milliseconds, which is the maximum timeout value in the windows API.
A PY_LONG_LONG of microseconds fullfills both requirements.
This closes issue #20737

11 years agoThe PyCOND_TIMEDWAIT must use microseconds for the timeout argument
Kristján Valur Jónsson [Thu, 8 May 2014 10:36:27 +0000 (10:36 +0000)] 
The PyCOND_TIMEDWAIT must use microseconds for the timeout argument
in order to have the same resolution as pthreads condition variables.
At the same time, it must be large enough to accept 31 bits of
milliseconds, which is the maximum timeout value in the windows API.
A PY_LONG_LONG of microseconds fullfills both requirements.
This closes issue #20737

11 years agoThe PyCOND_TIMEDWAIT must use microseconds for the timeout argument
Kristján Valur Jónsson [Thu, 8 May 2014 10:36:27 +0000 (10:36 +0000)] 
The PyCOND_TIMEDWAIT must use microseconds for the timeout argument
in order to have the same resolution as pthreads condition variables.
At the same time, it must be large enough to accept 31 bits of
milliseconds, which is the maximum timeout value in the windows API.
A PY_LONG_LONG of microseconds fullfills both requirements.
This closes issue #20737

11 years agoIssue19643 Fix whitespace
Tim Golden [Wed, 7 May 2014 17:08:08 +0000 (18:08 +0100)] 
Issue19643 Fix whitespace

11 years agoIssue19643 Add an example of shutil.rmtree which shows how to cope with readonly...
Tim Golden [Wed, 7 May 2014 17:05:45 +0000 (18:05 +0100)] 
Issue19643 Add an example of shutil.rmtree which shows how to cope with readonly files on Windows

11 years ago- Issue #17752: Fix distutils tests when run from the installed location.
doko@ubuntu.com [Wed, 7 May 2014 11:08:51 +0000 (13:08 +0200)] 
- Issue #17752: Fix distutils tests when run from the installed location.

11 years ago- Issue #17752: Fix distutils tests when run from the installed location.
doko@ubuntu.com [Wed, 7 May 2014 10:57:44 +0000 (12:57 +0200)] 
- Issue #17752: Fix distutils tests when run from the installed location.

11 years ago- Issue #17752: Fix distutils tests when run from the installed location.
doko@ubuntu.com [Wed, 7 May 2014 02:44:42 +0000 (04:44 +0200)] 
- Issue #17752: Fix distutils tests when run from the installed location.

11 years agoMerge: #21300: Clean up the docs for the email "policy" arguments.
R David Murray [Wed, 7 May 2014 01:33:50 +0000 (21:33 -0400)] 
Merge: #21300: Clean up the docs for the email "policy" arguments.

11 years ago#21300: Clean up the docs for the email "policy" arguments.
R David Murray [Wed, 7 May 2014 01:33:18 +0000 (21:33 -0400)] 
#21300: Clean up the docs for the email "policy" arguments.

11 years agoMerge 3.4->default: asyncio: Fix the second half of issue #21447: race in _write_to_s...
Guido van Rossum [Tue, 6 May 2014 21:45:39 +0000 (14:45 -0700)] 
Merge 3.4->default: asyncio: Fix the second half of issue #21447: race in _write_to_self().

11 years agoasyncio: Fix the second half of issue #21447: race in _write_to_self().
Guido van Rossum [Tue, 6 May 2014 21:42:40 +0000 (14:42 -0700)] 
asyncio: Fix the second half of issue #21447: race in _write_to_self().

11 years agoIssue #21442: Fix MSVC compiler warning introduced by issue21377.
Zachary Ware [Tue, 6 May 2014 16:42:37 +0000 (11:42 -0500)] 
Issue #21442: Fix MSVC compiler warning introduced by issue21377.

11 years agoMerge with 3.4
Zachary Ware [Tue, 6 May 2014 14:19:16 +0000 (09:19 -0500)] 
Merge with 3.4

11 years agoAdd prompts to interactive example.
Zachary Ware [Tue, 6 May 2014 14:18:17 +0000 (09:18 -0500)] 
Add prompts to interactive example.

This makes it match the new example below, and allows Sphinx's
"hide the prompts and output" feature to work.

11 years agoIssue #21366: Document the fact that ``return`` in a ``finally`` clause
Zachary Ware [Tue, 6 May 2014 14:07:51 +0000 (09:07 -0500)] 
Issue #21366: Document the fact that ``return`` in a ``finally`` clause
overrides a ``return`` in the ``try`` suite.

11 years agoIssue #21366: Document the fact that ``return`` in a ``finally`` clause
Zachary Ware [Tue, 6 May 2014 14:07:13 +0000 (09:07 -0500)] 
Issue #21366: Document the fact that ``return`` in a ``finally`` clause
overrides a ``return`` in the ``try`` suite.

11 years agoIssue21393 Use CryptReleaseContext to release Crypt handle on Windows
Tim Golden [Tue, 6 May 2014 12:29:45 +0000 (13:29 +0100)] 
Issue21393 Use CryptReleaseContext to release Crypt handle on Windows

11 years agoIssue21440 Use support.rmtree in test_zipfile & test_tarfile
Tim Golden [Tue, 6 May 2014 12:24:26 +0000 (13:24 +0100)] 
Issue21440 Use support.rmtree in test_zipfile & test_tarfile

11 years agoIssue #21233: Fix _PyObject_Alloc() when compiled with WITH_VALGRIND defined
Victor Stinner [Tue, 6 May 2014 09:32:29 +0000 (11:32 +0200)] 
Issue #21233: Fix _PyObject_Alloc() when compiled with WITH_VALGRIND defined

11 years agoIssue18314 ACKS & NEWS
Tim Golden [Mon, 5 May 2014 19:59:50 +0000 (20:59 +0100)] 
Issue18314 ACKS & NEWS

11 years agoIssue18314 Allow unlink to remove junctions. Includes support for creating junctions...
Tim Golden [Mon, 5 May 2014 18:46:17 +0000 (19:46 +0100)] 
Issue18314 Allow unlink to remove junctions. Includes support for creating junctions. Patch by Kim Gräsman

11 years agoMerge 3.4->default: asyncio docs: ProactorEventLoop does not support SSL.
Guido van Rossum [Mon, 5 May 2014 14:35:29 +0000 (07:35 -0700)] 
Merge 3.4->default: asyncio docs: ProactorEventLoop does not support SSL.

11 years agoasyncio docs: ProactorEventLoop does not support SSL.
Guido van Rossum [Mon, 5 May 2014 14:34:56 +0000 (07:34 -0700)] 
asyncio docs: ProactorEventLoop does not support SSL.

11 years agoIssue #21088: Merge from 3.4.
Larry Hastings [Sun, 4 May 2014 11:45:57 +0000 (04:45 -0700)] 
Issue #21088: Merge from 3.4.

11 years agoIssue #21088: Bugfix for curses.window.addch() regression in 3.4.0.
Larry Hastings [Sun, 4 May 2014 11:41:18 +0000 (04:41 -0700)] 
Issue #21088: Bugfix for curses.window.addch() regression in 3.4.0.
In porting to Argument Clinic, the first two arguments were reversed.

11 years agoIssue #19414: Have the OrderedDict mark deleted links as unusable.
Raymond Hettinger [Sun, 4 May 2014 04:58:45 +0000 (21:58 -0700)] 
Issue #19414: Have the OrderedDict mark deleted links as unusable.
This gives an earlier and more visible failure if a link is deleted
during iteration.

11 years agoIssue #18604: merge from 3.4
Ned Deily [Sun, 4 May 2014 02:25:34 +0000 (19:25 -0700)] 
Issue #18604: merge from 3.4

11 years agoIssue #18604: Skip the Tk instantiation test on OS X because it can
Ned Deily [Sun, 4 May 2014 02:24:05 +0000 (19:24 -0700)] 
Issue #18604: Skip the Tk instantiation test on OS X because it can
cause GUI tests to segfault in Cocoa Tk when run under regrtest -j
(multiple threads running subprocesses).

11 years agoIssue #21421: Add __slots__ to the MappingViews ABCs.
Raymond Hettinger [Sun, 4 May 2014 02:06:32 +0000 (19:06 -0700)] 
Issue #21421:  Add __slots__ to the MappingViews ABCs.

11 years agoNeaten-up a bit add add missing size change check.
Raymond Hettinger [Sun, 4 May 2014 01:45:54 +0000 (18:45 -0700)] 
Neaten-up a bit add add missing size change check.

11 years agoSimplify and speedup the internals of the heapq module.
Raymond Hettinger [Sun, 4 May 2014 01:36:48 +0000 (18:36 -0700)] 
Simplify and speedup the internals of the heapq module.

11 years agoimprove idioms (closes #20642)
Benjamin Peterson [Sun, 4 May 2014 00:22:00 +0000 (20:22 -0400)] 
improve idioms (closes #20642)

Patch by Claudiu Popa.

11 years agomerge 3.4
Benjamin Peterson [Sun, 4 May 2014 00:18:56 +0000 (20:18 -0400)] 
merge 3.4

11 years agouse with blocks to make sure files are closed
Benjamin Peterson [Sun, 4 May 2014 00:18:50 +0000 (20:18 -0400)] 
use with blocks to make sure files are closed

11 years agomerge 3.4 (#21357)
Benjamin Peterson [Sun, 4 May 2014 00:16:59 +0000 (20:16 -0400)] 
merge 3.4 (#21357)

11 years agoimprove test coverage of filecmp (closes #21357)
Benjamin Peterson [Sun, 4 May 2014 00:07:16 +0000 (20:07 -0400)] 
improve test coverage of filecmp (closes #21357)

Patch by Diana Clarke.

11 years agodo not expose known hash api in stable API
Benjamin Peterson [Sat, 3 May 2014 23:39:15 +0000 (19:39 -0400)] 
do not expose known hash api in stable API

11 years agoIssue #21101: Eliminate double hashing in the C code for collections.Counter().
Raymond Hettinger [Sat, 3 May 2014 23:41:19 +0000 (16:41 -0700)] 
Issue #21101:  Eliminate double hashing in the C code for collections.Counter().

11 years agoIssue 21101: Internal API for dict getitem and setitem where the hash value is known.
Raymond Hettinger [Sat, 3 May 2014 23:32:11 +0000 (16:32 -0700)] 
Issue 21101:  Internal API for dict getitem and setitem where the hash value is known.

11 years agomerge
Raymond Hettinger [Sat, 3 May 2014 22:26:17 +0000 (15:26 -0700)] 
merge

11 years agoIssue 21375: Fix possible Py_ssizet overflow in heapq.
Raymond Hettinger [Sat, 3 May 2014 22:22:07 +0000 (15:22 -0700)] 
Issue 21375:  Fix possible Py_ssizet overflow in heapq.

11 years agoIssue #21233: Oops, Fix _PyObject_Alloc(): initialize nbytes before going to
Victor Stinner [Fri, 2 May 2014 21:26:03 +0000 (23:26 +0200)] 
Issue #21233: Oops, Fix _PyObject_Alloc(): initialize nbytes before going to
redirect.

11 years agoIssue #21233: Add new C functions: PyMem_RawCalloc(), PyMem_Calloc(),
Victor Stinner [Fri, 2 May 2014 20:31:14 +0000 (22:31 +0200)] 
Issue #21233: Add new C functions: PyMem_RawCalloc(), PyMem_Calloc(),
PyObject_Calloc(), _PyObject_GC_Calloc(). bytes(int) and bytearray(int) are now
using ``calloc()`` instead of ``malloc()`` for large objects which is faster
and use less memory (until the bytearray buffer is filled with data).

11 years agoIssue #21393: random.c: on Windows, close the hCryptProv handle at exit
Victor Stinner [Fri, 2 May 2014 20:06:44 +0000 (22:06 +0200)] 
Issue #21393: random.c: on Windows, close the hCryptProv handle at exit

11 years agoCloses #18604: Merge with 3.4
Zachary Ware [Fri, 2 May 2014 15:52:12 +0000 (10:52 -0500)] 
Closes #18604: Merge with 3.4

11 years agoIssue #18604: Consolidated checks for GUI availability.
Zachary Ware [Fri, 2 May 2014 15:51:07 +0000 (10:51 -0500)] 
Issue #18604: Consolidated checks for GUI availability.

test_support._is_gui_available is now defined the same way on every
platform, and now includes the Windows-specific check that had been in the
Windows version of _is_gui_available and the OSX-specific check that was
in tkinter.test.support.check_tk_availability.  Also, every platform
checks whether Tk can be instantiated (if the platform-specific checks
passed).