]> granicus.if.org Git - python/log
python
11 years agomerge from 3.3
Senthil Kumaran [Thu, 12 Sep 2013 05:55:54 +0000 (22:55 -0700)]
merge from 3.3

Improve the docstring of random.shuffle. Inform users not to provide int arg.
Addresses issue #14927

11 years agoImprove the docstring of random.shuffle. Inform users not to provide int arg.
Senthil Kumaran [Thu, 12 Sep 2013 05:54:31 +0000 (22:54 -0700)]
Improve the docstring of random.shuffle. Inform users not to provide int arg.
Addresses issue #14927

11 years agomerge from 3.3
Senthil Kumaran [Tue, 10 Sep 2013 05:40:13 +0000 (22:40 -0700)]
merge from 3.3

Clarify mmap.close method  behavior.  Addresses issue  #18815
Patch contributed by Anoop Thomas Mathew.

11 years agoClarify mmap.close method behavior. Addresses issue #18815
Senthil Kumaran [Tue, 10 Sep 2013 05:39:28 +0000 (22:39 -0700)]
Clarify mmap.close method  behavior.  Addresses issue  #18815
Patch contributed by Anoop Thomas Mathew.

11 years agomerge from 3.3
Senthil Kumaran [Tue, 10 Sep 2013 02:58:20 +0000 (19:58 -0700)]
merge from 3.3

Document Fraction's numerator and denominator properties.
Addresses issue #18800

11 years agoDocument Fraction's numerator and denominator properties.
Senthil Kumaran [Tue, 10 Sep 2013 02:57:37 +0000 (19:57 -0700)]
Document Fraction's numerator and denominator properties.
Addresses issue #18800

11 years agoGet "stopped" back into repr(Thread) when appropriate.
Tim Peters [Mon, 9 Sep 2013 23:48:24 +0000 (18:48 -0500)]
Get "stopped" back into repr(Thread) when appropriate.

Due to recent changes, a Thread doesn't know that it's over before
someone calls .join() or .is_alive().  That meant repr(Thread)
continued to include "started" (and not "stopped") before one of
those methods was called, even if hours passed since the thread
ended.

Repaired that.

11 years agoAnother stab at the thread cleanup patch.
Tim Peters [Mon, 9 Sep 2013 19:41:50 +0000 (14:41 -0500)]
Another stab at the thread cleanup patch.

Antoine Pitrou found a variation that worked for him on the
thread+fork tests, and added an important

    self._is_stopped = True

to the after-fork code.  I confess I don't know why things passed
before.  But then mixing fork with threads is insane ;-)

11 years agoBacked out changeset 1f5a7853680c
Tim Peters [Mon, 9 Sep 2013 18:47:16 +0000 (13:47 -0500)]
Backed out changeset 1f5a7853680c

Unixy buildbots were failing the thread + fork tests :-(

11 years agoMinor cleanup of the new scheme for detecting thread termination.
Tim Peters [Mon, 9 Sep 2013 17:57:10 +0000 (12:57 -0500)]
Minor cleanup of the new scheme for detecting thread termination.

Documented some obscurities, and assert'ed ._stop()'s crucial precondition.

11 years agoMerge.
Larry Hastings [Mon, 9 Sep 2013 12:12:21 +0000 (21:12 +0900)]
Merge.

11 years agoPost-3.4.0a2-release fixups.
Larry Hastings [Mon, 9 Sep 2013 12:08:52 +0000 (21:08 +0900)]
Post-3.4.0a2-release fixups.

11 years agoMerge docstring typo fix from 3.3
Mark Dickinson [Mon, 9 Sep 2013 09:34:56 +0000 (10:34 +0100)]
Merge docstring typo fix from 3.3

11 years agoDocstring typo fix: Arithmentic -> Arithmetic.
Mark Dickinson [Mon, 9 Sep 2013 09:34:24 +0000 (10:34 +0100)]
Docstring typo fix: Arithmentic -> Arithmetic.

11 years agomerge
Raymond Hettinger [Mon, 9 Sep 2013 07:02:05 +0000 (02:02 -0500)]
merge

11 years agoIssue 18752: Make chain.from_iterable() more visible in the documentation.
Raymond Hettinger [Mon, 9 Sep 2013 07:01:35 +0000 (02:01 -0500)]
Issue 18752:  Make chain.from_iterable() more visible in the documentation.

11 years agomerge
Raymond Hettinger [Mon, 9 Sep 2013 06:55:31 +0000 (01:55 -0500)]
merge

11 years agomerge
Raymond Hettinger [Mon, 9 Sep 2013 06:55:07 +0000 (01:55 -0500)]
merge

11 years agoIssue 18301: The classmethod decorator didn't fit well with the rough-equivalent...
Raymond Hettinger [Mon, 9 Sep 2013 06:54:27 +0000 (01:54 -0500)]
Issue 18301:  The classmethod decorator didn't fit well with the rough-equivalent example code.

11 years agoClose #18980: Enum doc fixes. Patch from Elazar Gershuni.
Ethan Furman [Mon, 9 Sep 2013 06:48:34 +0000 (23:48 -0700)]
Close #18980: Enum doc fixes.  Patch from Elazar Gershuni.

11 years agoIssue 18752: Make chain.from_iterable() more visible in the documentation.
Raymond Hettinger [Mon, 9 Sep 2013 06:29:40 +0000 (01:29 -0500)]
Issue 18752:  Make chain.from_iterable() more visible in the documentation.

11 years agoIssue 18984: Remove ._stopped Event from Thread internals.
Tim Peters [Sun, 8 Sep 2013 23:44:40 +0000 (18:44 -0500)]
Issue 18984:  Remove ._stopped Event from Thread internals.

The fix for issue 18808 left us checking two things to be sure a Thread
was done:  an Event (._stopped) and a mutex (._tstate_lock).  Clumsy &
brittle.  This patch removes the Event, leaving just a happy lock :-)

The bulk of the patch removes two excruciating tests, which were
verifying sanity of the internals of the ._stopped Event after a fork.
Thanks to Antoine Pitrou for verifying that's the only real value
these tests had.

One consequence of moving from an Event to a mutex:  waiters (threads
calling Thread.join()) used to block each on their own unique mutex
(internal to the ._stopped event), but now all contend on the same
mutex (._tstate_lock).  These approaches have different performance
characteristics on different platforms.  I don't think it matters in
this context.

11 years agoFix a typo. (closes #18953)
Serhiy Storchaka [Sun, 8 Sep 2013 17:43:02 +0000 (20:43 +0300)]
Fix a typo. (closes #18953)

11 years agoFix a typo. (closes #18953)
Serhiy Storchaka [Sun, 8 Sep 2013 17:42:13 +0000 (20:42 +0300)]
Fix a typo. (closes #18953)

11 years agoFixed tests with Tcl/Tk <8.5 (closes #18964).
Serhiy Storchaka [Sun, 8 Sep 2013 17:31:20 +0000 (20:31 +0300)]
Fixed tests with Tcl/Tk <8.5 (closes #18964).

11 years agoFixed tests with Tcl/Tk <8.5 (closes #18964).
Serhiy Storchaka [Sun, 8 Sep 2013 17:29:37 +0000 (20:29 +0300)]
Fixed tests with Tcl/Tk <8.5 (closes #18964).

11 years agoIssue #18904: test_os and test_socket use unittest.skipIf() to check if fcntl
Victor Stinner [Sun, 8 Sep 2013 12:14:38 +0000 (14:14 +0200)]
Issue #18904: test_os and test_socket use unittest.skipIf() to check if fcntl
module is present (to record skipped tests)

11 years agoIssue #18808 again: fix the after-fork logic for not-yet-started or already-stopped...
Antoine Pitrou [Sun, 8 Sep 2013 11:19:06 +0000 (13:19 +0200)]
Issue #18808 again: fix the after-fork logic for not-yet-started or already-stopped threads.
(AFAICT, in theory, we must reset all the locks, not just those in use)

11 years agoMerge.
Charles-François Natali [Sun, 8 Sep 2013 10:35:53 +0000 (12:35 +0200)]
Merge.

11 years agoIssue #18963: skip test_selectors.test_above_fd_setsize on older OS X versions.
Charles-François Natali [Sun, 8 Sep 2013 10:31:32 +0000 (12:31 +0200)]
Issue #18963: skip test_selectors.test_above_fd_setsize on older OS X versions.

11 years agoIssue #18935: Fix test_regrtest.test_timeout when built --without-threads (the
Charles-François Natali [Sun, 8 Sep 2013 10:27:33 +0000 (12:27 +0200)]
Issue #18935: Fix test_regrtest.test_timeout when built --without-threads (the
'--timeout' option requires faulthandler.dump_traceback_later).

11 years agoIssue #18904: test_socket: add inheritance tests using fcntl and FD_CLOEXEC
Victor Stinner [Sun, 8 Sep 2013 09:53:09 +0000 (11:53 +0200)]
Issue #18904: test_socket: add inheritance tests using fcntl and FD_CLOEXEC

11 years agoIssue #18904: Improve os.get/set_inheritable() tests
Victor Stinner [Sun, 8 Sep 2013 09:47:54 +0000 (11:47 +0200)]
Issue #18904: Improve os.get/set_inheritable() tests

11 years agoClose #18957: The PYTHONFAULTHANDLER environment variable now only enables the
Victor Stinner [Sun, 8 Sep 2013 09:36:23 +0000 (11:36 +0200)]
Close #18957: The PYTHONFAULTHANDLER environment variable now only enables the
faulthandler module if the variable is non-empty. Same behaviour than other
variables like PYTHONDONTWRITEBYTECODE.

11 years agoIssue #18963: Fix test_selectors.test_above_fd_setsize on OS X, where the
Charles-François Natali [Sun, 8 Sep 2013 09:34:42 +0000 (11:34 +0200)]
Issue #18963: Fix test_selectors.test_above_fd_setsize on OS X, where the
default RLIMIT_NOFILE hard limit can be RLIMIT_INFINITY.

11 years agoIssue #18934: Use poll/select-based selectors for multiprocessing.Connection,
Charles-François Natali [Sun, 8 Sep 2013 09:30:53 +0000 (11:30 +0200)]
Issue #18934: Use poll/select-based selectors for multiprocessing.Connection,
to avoid one extra FD per Connection.

11 years agoPut the defines in the logical section and fix indentation.
Raymond Hettinger [Sun, 8 Sep 2013 07:25:57 +0000 (00:25 -0700)]
Put the defines in the logical section and fix indentation.

11 years agomerge from 3.3
Senthil Kumaran [Sun, 8 Sep 2013 06:20:06 +0000 (23:20 -0700)]
merge from 3.3

Fix the dead link of IEEE_854-1987 standard with the Wikipedia entry.
Addresses issue #18438

11 years agoFix the dead link of IEEE_854-1987 standard with the Wikipedia entry.
Senthil Kumaran [Sun, 8 Sep 2013 06:19:29 +0000 (23:19 -0700)]
Fix the dead link of IEEE_854-1987 standard with the Wikipedia entry.
Addresses issue #18438

11 years agoMinor code beautification.
Raymond Hettinger [Sun, 8 Sep 2013 05:06:35 +0000 (22:06 -0700)]
Minor code beautification.

11 years agoImprove code clarity by removing two unattractive macros.
Raymond Hettinger [Sun, 8 Sep 2013 04:01:29 +0000 (21:01 -0700)]
Improve code clarity by removing two unattractive macros.

11 years agoRemove the freelist scheme for setobjects.
Raymond Hettinger [Sun, 8 Sep 2013 03:26:50 +0000 (20:26 -0700)]
Remove the freelist scheme for setobjects.

The setobject freelist was consuming memory but not providing much value.
Even when a freelisted setobject was available, most of the setobject
fields still needed to be initialized and the small table still required
a memset().  This meant that the custom freelisting scheme for sets was
providing almost no incremental benefit over the default Python freelist
scheme used by _PyObject_Malloc() in Objects/obmalloc.c.

11 years agoMerge #18952 fix from 3.3
Nick Coghlan [Sun, 8 Sep 2013 02:49:53 +0000 (12:49 +1000)]
Merge #18952 fix from 3.3

11 years agoIssue 18808: blind attempt to repair some buildbot failures.
Tim Peters [Sun, 8 Sep 2013 02:23:03 +0000 (21:23 -0500)]
Issue 18808:  blind attempt to repair some buildbot failures.

test_is_alive_after_fork is failing on some old Linux kernels, but
passing on all newer ones.  Since virtually anything can go wrong
with locks when mixing threads with fork, replace the most likely
cause with a redundant simple data member.

11 years agoClose #18952: correctly download test support data
Nick Coghlan [Sun, 8 Sep 2013 01:40:34 +0000 (11:40 +1000)]
Close #18952: correctly download test support data

When test.support was converted to a package, it started silently
skipping the tests which needed to download support data to run.

This change refactors the affected code, and also tidies up
test.support.findfile to remove the unused *here* parameter, document
the *subdir* parameter and rename the *filename* parameter to avoid
shadowing the file builtin and be consistent with the documentation.

The unexpected skips were noticed and reported by Zachary Ware

11 years agomerge from 3.3
Senthil Kumaran [Sun, 8 Sep 2013 00:52:38 +0000 (17:52 -0700)]
merge from 3.3

Correct Profile class usage example. Addresses issue #18033.
Patch contributed by Olivier Hervieu and Dmi Baranov.

11 years agoCorrect Profile class usage example. Addresses issue #18033.
Senthil Kumaran [Sun, 8 Sep 2013 00:51:58 +0000 (17:51 -0700)]
Correct Profile class usage example. Addresses issue #18033.
Patch contributed by Olivier Hervieu and Dmi Baranov.

11 years agoSmall rearrangement to bring together the three functions for probing the hash table.
Raymond Hettinger [Sun, 8 Sep 2013 00:41:01 +0000 (17:41 -0700)]
Small rearrangement to bring together the three functions for probing the hash table.

11 years agoMove the overview comment to the top of the file.
Raymond Hettinger [Sat, 7 Sep 2013 22:05:00 +0000 (15:05 -0700)]
Move the overview comment to the top of the file.

11 years agoIssue #18808: Thread.join() now waits for the underlying thread state to be destroyed...
Antoine Pitrou [Sat, 7 Sep 2013 21:38:37 +0000 (23:38 +0200)]
Issue #18808: Thread.join() now waits for the underlying thread state to be destroyed before returning.
This prevents unpredictable aborts in Py_EndInterpreter() when some non-daemon threads are still running.

11 years agoFix the merge conflict
Senthil Kumaran [Sat, 7 Sep 2013 21:12:55 +0000 (14:12 -0700)]
Fix the merge conflict

11 years agomerge from 3.3
Senthil Kumaran [Sat, 7 Sep 2013 21:09:48 +0000 (14:09 -0700)]
merge from 3.3

Fix License URL display and add test to check for license url presence.
Fixes issue #18206 Patch contributed by  Berker Peksag and py.user

11 years agoFix License URL display and add test to check for license url presence.
Senthil Kumaran [Sat, 7 Sep 2013 20:59:17 +0000 (13:59 -0700)]
Fix License URL display and add test to check for license url presence.
Fixes issue #18206 Patch contributed by  Berker Peksag and py.user

11 years agomerge from 3.3
Senthil Kumaran [Sat, 7 Sep 2013 18:30:04 +0000 (11:30 -0700)]
merge from 3.3

Removing the mention of os.isatty mention as Unix only
Correct the wrong documentation.

11 years agoRemoving the mention of os.isatty mention as Unix only
Senthil Kumaran [Sat, 7 Sep 2013 18:28:58 +0000 (11:28 -0700)]
Removing the mention of os.isatty mention as Unix only
Correct the wrong documentation.

11 years agoFix conversion from Py_ssize_t to int.
Richard Oudkerk [Sat, 7 Sep 2013 16:40:45 +0000 (17:40 +0100)]
Fix conversion from Py_ssize_t to int.

11 years ago#18895: merge with 3.3.
Ezio Melotti [Sat, 7 Sep 2013 12:24:01 +0000 (15:24 +0300)]
#18895: merge with 3.3.

11 years ago#18895: split a sentence in unittest docs.
Ezio Melotti [Sat, 7 Sep 2013 12:23:36 +0000 (15:23 +0300)]
#18895: split a sentence in unittest docs.

11 years ago#18894: merge with 3.3.
Ezio Melotti [Sat, 7 Sep 2013 12:20:03 +0000 (15:20 +0300)]
#18894: merge with 3.3.

11 years ago#18894: remove mention of deprecated fail* methods.
Ezio Melotti [Sat, 7 Sep 2013 12:19:30 +0000 (15:19 +0300)]
#18894: remove mention of deprecated fail* methods.

11 years agoAdded tag v3.4.0a2 for changeset 9265a2168e2c
Larry Hastings [Sat, 7 Sep 2013 11:42:24 +0000 (23:42 +1200)]
Added tag v3.4.0a2 for changeset 9265a2168e2c

11 years agoVersion number bump for Python 3.4.0a2. v3.4.0a2
Larry Hastings [Sat, 7 Sep 2013 11:42:07 +0000 (23:42 +1200)]
Version number bump for Python 3.4.0a2.

11 years agoUpdate pydoc topic index for Python 3.4.0a2.
Larry Hastings [Sat, 7 Sep 2013 11:41:12 +0000 (23:41 +1200)]
Update pydoc topic index for Python 3.4.0a2.

11 years agoClose #18954: Fix some typo in fileutils.c comments
Victor Stinner [Sat, 7 Sep 2013 08:36:04 +0000 (10:36 +0200)]
Close #18954: Fix some typo in fileutils.c comments

Patch written by Vajrasky Kok.

11 years agoAdded Elazar to Misc/ACKS.
Ethan Furman [Sat, 7 Sep 2013 02:58:01 +0000 (19:58 -0700)]
Added Elazar to Misc/ACKS.

11 years agoClose #18908: Keep Enum docs in their own section. Patch by Elazar Gershuni.
Ethan Furman [Sat, 7 Sep 2013 02:53:30 +0000 (19:53 -0700)]
Close #18908: Keep Enum docs in their own section.  Patch by Elazar Gershuni.

11 years agoIssue #18458: Prevent crashes with newer versions of libedit. Its readline
Ned Deily [Fri, 6 Sep 2013 22:16:19 +0000 (15:16 -0700)]
Issue #18458: Prevent crashes with newer versions of libedit.  Its readline
emulation has changed from 0-based indexing to 1-based like gnu readline.
Original patch by Ronald Oussoren.

11 years agoMerge 3.3 into default.
Tim Peters [Fri, 6 Sep 2013 20:42:47 +0000 (15:42 -0500)]
Merge 3.3 into default.

Issue 18944:  fix a 1-character typo in test_set.py.

The error caused test_inline_methods() to test much less than intended.
Caught (& fixed) by Armin Rigo.

11 years agoIssue 18944: fix a 1-character typo in test_set.py.
Tim Peters [Fri, 6 Sep 2013 20:41:30 +0000 (15:41 -0500)]
Issue 18944:  fix a 1-character typo in test_set.py.

The error caused test_inline_methods() to test much less than intended.
Caught (& fixed) by Armin Rigo.

11 years agoRemove old-school inheritance
Antoine Pitrou [Fri, 6 Sep 2013 19:18:25 +0000 (21:18 +0200)]
Remove old-school inheritance

11 years agoIssue #18934: Relax test_multiprocessing.test_invalid_handles a bit: we just
Charles-François Natali [Fri, 6 Sep 2013 19:12:22 +0000 (21:12 +0200)]
Issue #18934: Relax test_multiprocessing.test_invalid_handles a bit: we just
want to check that Connection.poll() doesn't crash.

11 years agoIssue #18623: Factor out the _SuppressCoreFiles context manager into test.support.
Antoine Pitrou [Fri, 6 Sep 2013 18:50:00 +0000 (20:50 +0200)]
Issue #18623: Factor out the _SuppressCoreFiles context manager into test.support.
Patch by Valerie Lambert.

11 years ago#18852: Handle readline.__doc__ being None in site.py readline activation.
R David Murray [Fri, 6 Sep 2013 17:08:08 +0000 (13:08 -0400)]
#18852: Handle readline.__doc__ being None in site.py readline activation.

Patch by Berker Peksag.

11 years agoClose #18924: Block naive attempts to change an Enum member.
Ethan Furman [Fri, 6 Sep 2013 14:16:48 +0000 (07:16 -0700)]
Close #18924:  Block naive attempts to change an Enum member.

11 years agoUpdate whatsnew/3.4 wrt. --version going to stdout. #18338, #18920, #18922
Eli Bendersky [Fri, 6 Sep 2013 13:55:58 +0000 (06:55 -0700)]
Update whatsnew/3.4 wrt. --version going to stdout. #18338, #18920, #18922

11 years agoIssue #18920: argparse's default version action (for -v, --version) should
Eli Bendersky [Fri, 6 Sep 2013 13:49:15 +0000 (06:49 -0700)]
Issue #18920: argparse's default version action (for -v, --version) should
output to stdout, matching the 'python -v'

Reported by Wolfgang Maier

11 years agoIssue #18849: Fixed a Windows-specific tempfile bug where collision with an
Eli Bendersky [Fri, 6 Sep 2013 13:14:16 +0000 (06:14 -0700)]
Issue #18849: Fixed a Windows-specific tempfile bug where collision with an
existing directory caused mkstemp and related APIs to fail instead of
retrying. Report and fix by Vlad Shcherbina.

11 years agoIssue #18849: Fixed a Windows-specific tempfile bug where collision with an
Eli Bendersky [Fri, 6 Sep 2013 13:11:19 +0000 (06:11 -0700)]
Issue #18849: Fixed a Windows-specific tempfile bug where collision with an
existing directory caused mkstemp and related APIs to fail instead of
retrying. Report and fix by Vlad Shcherbina.

11 years agoCloses #18941: Merged fix from 3.3.
Vinay Sajip [Fri, 6 Sep 2013 09:26:48 +0000 (10:26 +0100)]
Closes #18941: Merged fix from 3.3.

11 years agoIssue #18941: Respected delay when doing rollover.
Vinay Sajip [Fri, 6 Sep 2013 09:25:31 +0000 (10:25 +0100)]
Issue #18941: Respected delay when doing rollover.

11 years agoCloses #18940: Merged fix from 3.3.
Vinay Sajip [Fri, 6 Sep 2013 09:11:37 +0000 (10:11 +0100)]
Closes #18940: Merged fix from 3.3.

11 years agoIssue #18940: Handled low-volume logging when delay is True.
Vinay Sajip [Fri, 6 Sep 2013 09:10:22 +0000 (10:10 +0100)]
Issue #18940: Handled low-volume logging when delay is True.

11 years agoCloses #18939: Merged documentation update from 3.3.
Vinay Sajip [Fri, 6 Sep 2013 08:51:27 +0000 (09:51 +0100)]
Closes #18939: Merged documentation update from 3.3.

11 years agoIssue #18939: Updated venv documentation with some clarifications.
Vinay Sajip [Fri, 6 Sep 2013 08:50:43 +0000 (09:50 +0100)]
Issue #18939: Updated venv documentation with some clarifications.

11 years agoIssue #15663: Tcl/Tk 8.5.14 is now included with the OS X 10.6+
Ned Deily [Fri, 6 Sep 2013 08:18:36 +0000 (01:18 -0700)]
Issue #15663: Tcl/Tk 8.5.14 is now included with the OS X 10.6+
64-bit/32-bit installer for 10.6+.  It is no longer necessary
to install a third-party version of Tcl/Tk 8.5 to work around the
problems in the Apple-supplied Tcl/Tk 8.5 shipped in OS X 10.6
and later releases.

11 years agoIssue #1584: Provide options to override default search paths for Tcl and Tk
Ned Deily [Fri, 6 Sep 2013 08:07:05 +0000 (01:07 -0700)]
Issue #1584: Provide options to override default search paths for Tcl and Tk
when building _tkinter.  configure has two new options; if used, both must
be specified:

  ./configure \
      --with-tcltk-includes="-I/opt/local/include" \
      --with-tcltk-libs="-L/opt/local/lib -ltcl8.5 -ltk8.5"

In addition, the options can be overridden with make:

   make \
       TCLTK_INCLUDES="-I/opt/local/include" \
       TCLTK_LIBS="-L/opt/local/lib -ltcl8.6 -ltk8.6"

11 years agomerge from 3.3
Senthil Kumaran [Fri, 6 Sep 2013 04:43:53 +0000 (21:43 -0700)]
merge from 3.3

Improve urlencode docstring. Patch by Brian Brazil.
Closes issue #15350

11 years agoImprove urlencode docstring. Patch by Brian Brazil.
Senthil Kumaran [Fri, 6 Sep 2013 04:42:38 +0000 (21:42 -0700)]
Improve urlencode docstring. Patch by Brian Brazil.

11 years agoNerge 3.3 into default.
Tim Peters [Fri, 6 Sep 2013 04:04:26 +0000 (23:04 -0500)]
Nerge 3.3 into default.

Issue #18942: sys._debugmallocstats() output was damaged on Windows.

_PyDebugAllocatorStats() called PyOS_snprintf() with a %zd format
code, but MS doesn't support that code.  Interpolated
PY_FORMAT_SIZE_T in place of the "z".

11 years agoIssue #18942: sys._debugmallocstats() output was damaged on Windows.
Tim Peters [Fri, 6 Sep 2013 03:57:04 +0000 (22:57 -0500)]
Issue #18942: sys._debugmallocstats() output was damaged on Windows.

_PyDebugAllocatorStats() called PyOS_snprintf() with a %zd format
code, but MS doesn't support that code.  Interpolated
PY_FORMAT_SIZE_T in place of the "z".

11 years agoCloses #18933: Merged update from 3.3.
Vinay Sajip [Thu, 5 Sep 2013 22:02:45 +0000 (23:02 +0100)]
Closes #18933: Merged update from 3.3.

11 years agoIssue #18933: Added links to source code.
Vinay Sajip [Thu, 5 Sep 2013 22:01:07 +0000 (23:01 +0100)]
Issue #18933: Added links to source code.

11 years agoIssue #18934: multiprocessing: use selectors module.
Charles-François Natali [Thu, 5 Sep 2013 18:46:49 +0000 (20:46 +0200)]
Issue #18934: multiprocessing: use selectors module.

11 years agoIssue #18672: Fixed format specifiers for Py_ssize_t in debugging output in
Serhiy Storchaka [Thu, 5 Sep 2013 15:02:31 +0000 (18:02 +0300)]
Issue #18672: Fixed format specifiers for Py_ssize_t in debugging output in
the _sre moduel.

11 years agoIssue #18672: Fixed format specifiers for Py_ssize_t in debugging output in
Serhiy Storchaka [Thu, 5 Sep 2013 15:01:15 +0000 (18:01 +0300)]
Issue #18672: Fixed format specifiers for Py_ssize_t in debugging output in
the _sre moduel.

11 years agoIssue #18922: Now The Lib/smtpd.py and Tools/i18n/msgfmt.py scripts write
Serhiy Storchaka [Thu, 5 Sep 2013 14:44:53 +0000 (17:44 +0300)]
Issue #18922: Now The Lib/smtpd.py and Tools/i18n/msgfmt.py scripts write
their version strings to stdout, and not to sderr.

11 years agoNull merge
Serhiy Storchaka [Thu, 5 Sep 2013 14:33:04 +0000 (17:33 +0300)]
Null merge

11 years agoMerge heads
Serhiy Storchaka [Thu, 5 Sep 2013 14:32:15 +0000 (17:32 +0300)]
Merge heads

11 years agoMerge heads
Serhiy Storchaka [Thu, 5 Sep 2013 14:31:37 +0000 (17:31 +0300)]
Merge heads

11 years agoIssue #18830: inspect.getclasstree() no more produces duplicated entries even
Serhiy Storchaka [Thu, 5 Sep 2013 14:16:12 +0000 (17:16 +0300)]
Issue #18830: inspect.getclasstree() no more produces duplicated entries even
when input list contains duplicates.