]> granicus.if.org Git - python/log
python
12 years agoBranch merge
Éric Araujo [Sun, 5 Feb 2012 09:48:52 +0000 (10:48 +0100)]
Branch merge

12 years agoStop converting package_data to extra_files in pysetup create (#13712).
Éric Araujo [Sun, 5 Feb 2012 09:26:16 +0000 (10:26 +0100)]
Stop converting package_data to extra_files in pysetup create (#13712).

pysetup create, the setup.cfg creation helper, used to convert
package_data (from an existing setup.py) into extra_files, the
replacement for MANIFEST.in, but these files are only present in sdists,
not installed: they don’t have the same use case at all, so converting
one into the other did not work.

12 years agomerge with 3.2
Georg Brandl [Sun, 5 Feb 2012 08:28:39 +0000 (09:28 +0100)]
merge with 3.2

12 years agoCloses #13944: fix capitalization of class name.
Georg Brandl [Sun, 5 Feb 2012 08:25:22 +0000 (09:25 +0100)]
Closes #13944: fix capitalization of class name.

12 years agoIssue #12142: Fixed reference cycle when importing ctypes
Meador Inge [Sun, 5 Feb 2012 02:38:20 +0000 (20:38 -0600)]
Issue #12142: Fixed reference cycle when importing ctypes

12 years agoIssue #12142: Fixed reference cycle when importing ctypes
Meador Inge [Sun, 5 Feb 2012 02:36:48 +0000 (20:36 -0600)]
Issue #12142: Fixed reference cycle when importing ctypes

12 years ago_Py_Identifier are always ASCII strings
Victor Stinner [Sun, 5 Feb 2012 00:45:45 +0000 (01:45 +0100)]
_Py_Identifier are always ASCII strings

12 years agoNull merge.
Nadeem Vawda [Sat, 4 Feb 2012 21:45:40 +0000 (23:45 +0200)]
Null merge.

12 years agoIssue #1625: Document BZ2File's lack of support for multi-stream inputs.
Nadeem Vawda [Sat, 4 Feb 2012 21:44:49 +0000 (23:44 +0200)]
Issue #1625: Document BZ2File's lack of support for multi-stream inputs.

12 years agoAllow multiple values for package_data in setup.cfg (#11805).
Éric Araujo [Sat, 4 Feb 2012 20:53:07 +0000 (21:53 +0100)]
Allow multiple values for package_data in setup.cfg (#11805).

Even though the resources system obsoletes data_files and package_data
(see bug discussion), package_data still exists to allow compatibility
with distutils and thus an easier transition.  In setup.py, the values
are lists of glob patterns, so the setup.cfg syntax needed a way to
express multiple values too.

Doc for this option will be added later as part of the big packaging doc
patches.  For now, the test serves as example.

Reported by Erik Bray.

12 years agoImprove one packaging test, remove a setuptoolism in another
Éric Araujo [Sat, 4 Feb 2012 20:43:07 +0000 (21:43 +0100)]
Improve one packaging test, remove a setuptoolism in another

12 years agoIssue #13933: merge
Ned Deily [Sat, 4 Feb 2012 17:38:10 +0000 (18:38 +0100)]
Issue #13933: merge

12 years agoIssue #13933: IDLE auto-complete did not work with some imported
Ned Deily [Sat, 4 Feb 2012 17:36:43 +0000 (18:36 +0100)]
Issue #13933: IDLE auto-complete did not work with some imported
module, like hashlib.  (Patch by Roger Serwy)

12 years agoFix failing test on big-endian machines (issue #13806).
Antoine Pitrou [Sat, 4 Feb 2012 15:49:31 +0000 (16:49 +0100)]
Fix failing test on big-endian machines (issue #13806).

12 years agoFix failing test on big-endian machines (issue #13806).
Antoine Pitrou [Sat, 4 Feb 2012 15:44:21 +0000 (16:44 +0100)]
Fix failing test on big-endian machines (issue #13806).

12 years agomerge 3.2
Benjamin Peterson [Sat, 4 Feb 2012 14:56:43 +0000 (09:56 -0500)]
merge 3.2

12 years agothreading primitives now have timeouts
Benjamin Peterson [Sat, 4 Feb 2012 14:55:52 +0000 (09:55 -0500)]
threading primitives now have timeouts

12 years agoIssue #8184: Fix a potential file descriptor leak when a
Charles-François Natali [Sat, 4 Feb 2012 14:12:08 +0000 (15:12 +0100)]
Issue #8184: Fix a potential file descriptor leak when a
multiprocessing.Connection socket can't be bound.

12 years agoIssue #8184: Fix a potential file descriptor leak when a
Charles-François Natali [Sat, 4 Feb 2012 13:55:53 +0000 (14:55 +0100)]
Issue #8184: Fix a potential file descriptor leak when a
multiprocessing.Connection socket can't be bound.

12 years agoDocument that some of LZMAFile.__init__()'s args are keyword-only.
Nadeem Vawda [Sat, 4 Feb 2012 12:06:07 +0000 (14:06 +0200)]
Document that some of LZMAFile.__init__()'s args are keyword-only.

12 years agoMake BZ2File.__init__()'s fileobj argument keyword-only.
Nadeem Vawda [Sat, 4 Feb 2012 11:58:07 +0000 (13:58 +0200)]
Make BZ2File.__init__()'s fileobj argument keyword-only.

12 years agoUpdate docstring for BZ2File.__init__().
Nadeem Vawda [Sat, 4 Feb 2012 11:08:11 +0000 (13:08 +0200)]
Update docstring for BZ2File.__init__().

12 years agoSmall grammar fixes by Mark Summerfield.
Georg Brandl [Sat, 4 Feb 2012 07:55:56 +0000 (08:55 +0100)]
Small grammar fixes by Mark Summerfield.

12 years agomerge 3.2
Benjamin Peterson [Sat, 4 Feb 2012 00:25:01 +0000 (19:25 -0500)]
merge 3.2

12 years agoput returns on their own lines
Benjamin Peterson [Sat, 4 Feb 2012 00:22:31 +0000 (19:22 -0500)]
put returns on their own lines

12 years agomerge 3.2
Benjamin Peterson [Sat, 4 Feb 2012 00:09:38 +0000 (19:09 -0500)]
merge 3.2

12 years agoremove unused import
Benjamin Peterson [Sat, 4 Feb 2012 00:07:30 +0000 (19:07 -0500)]
remove unused import

12 years agoIssue #13861: merge
Ned Deily [Fri, 3 Feb 2012 22:16:11 +0000 (23:16 +0100)]
Issue #13861: merge

12 years agoIssue #13861: Prevent test_apropos* test case failures in test_pydoc.
Ned Deily [Fri, 3 Feb 2012 22:14:37 +0000 (23:14 +0100)]
Issue #13861: Prevent test_apropos* test case failures in test_pydoc.

12 years agoMerge
Brett Cannon [Fri, 3 Feb 2012 17:08:32 +0000 (12:08 -0500)]
Merge

12 years agoCheck for errors in creating sub-interpreters when testing the C API.
Brett Cannon [Fri, 3 Feb 2012 17:08:03 +0000 (12:08 -0500)]
Check for errors in creating sub-interpreters when testing the C API.

12 years agoIssue #13777: Add PF_SYSTEM sockets on OS X.
Martin v. Löwis [Fri, 3 Feb 2012 16:44:58 +0000 (17:44 +0100)]
Issue #13777: Add PF_SYSTEM sockets on OS X.
Patch by Michael Goderbauer.

13 years agoIssue #13901: Prevent test_packaging failures on OS X with --enable-shared.
Ned Deily [Fri, 3 Feb 2012 01:46:37 +0000 (02:46 +0100)]
Issue #13901: Prevent test_packaging failures on OS X with --enable-shared.

13 years agomerge
Ned Deily [Fri, 3 Feb 2012 01:45:05 +0000 (02:45 +0100)]
merge

13 years agoIssue #13901: Prevent test_distutils failures on OS X with --enable-shared.
Ned Deily [Fri, 3 Feb 2012 01:42:16 +0000 (02:42 +0100)]
Issue #13901: Prevent test_distutils failures on OS X with --enable-shared.

13 years agoFix small grammatical inconsistency.
Georg Brandl [Thu, 2 Feb 2012 21:22:19 +0000 (22:22 +0100)]
Fix small grammatical inconsistency.

13 years agoNull merge.
Charles-François Natali [Thu, 2 Feb 2012 19:38:46 +0000 (20:38 +0100)]
Null merge.

13 years agoMerge.
Charles-François Natali [Thu, 2 Feb 2012 19:38:10 +0000 (20:38 +0100)]
Merge.

13 years agoMerge.
Charles-François Natali [Thu, 2 Feb 2012 19:37:29 +0000 (20:37 +0100)]
Merge.

13 years agoIssue #13817: After fork(), reinit the ad-hoc TLS implementation earlier to fix
Charles-François Natali [Thu, 2 Feb 2012 19:32:48 +0000 (20:32 +0100)]
Issue #13817: After fork(), reinit the ad-hoc TLS implementation earlier to fix
a random deadlock when fork() is called in a multithreaded process in debug
mode, and make PyOS_AfterFork() more robust.

13 years agoIssue #13817: After fork(), reinit the ad-hoc TLS implementation earlier to fix
Charles-François Natali [Thu, 2 Feb 2012 19:31:42 +0000 (20:31 +0100)]
Issue #13817: After fork(), reinit the ad-hoc TLS implementation earlier to fix
a random deadlock when fork() is called in a multithreaded process in debug
mode, and make PyOS_AfterFork() more robust.

13 years agoMerge branch 3.2
Petri Lehtinen [Thu, 2 Feb 2012 19:23:15 +0000 (21:23 +0200)]
Merge branch 3.2

Closes #13402.

13 years agoDocument absoluteness of sys.executable
Petri Lehtinen [Thu, 2 Feb 2012 18:59:48 +0000 (20:59 +0200)]
Document absoluteness of sys.executable

Closes #13402.

13 years agoMake sqlite3 tests editable with Emacs
Petri Lehtinen [Thu, 2 Feb 2012 15:17:36 +0000 (17:17 +0200)]
Make sqlite3 tests editable with Emacs

Change the coding declaration from ISO-8859-1 to iso-8859-1. Emacs
doesn't understand the former.

13 years agoIssue #1813: merge changeset that reverts a glibc workaround for the
Stefan Krah [Thu, 2 Feb 2012 15:50:10 +0000 (16:50 +0100)]
Issue #1813: merge changeset that reverts a glibc workaround for the
Fedora buildbot.

13 years agoIssue #1813: Revert workaround for a glibc bug on the Fedora buildbot.
Stefan Krah [Thu, 2 Feb 2012 15:40:52 +0000 (16:40 +0100)]
Issue #1813: Revert workaround for a glibc bug on the Fedora buildbot.

13 years agomerge. again.
Brian Curtin [Wed, 1 Feb 2012 21:24:10 +0000 (15:24 -0600)]
merge. again.

13 years agomerge updates
Brian Curtin [Wed, 1 Feb 2012 21:20:03 +0000 (15:20 -0600)]
merge updates

13 years agomerge from 3.2
Brian Curtin [Wed, 1 Feb 2012 21:15:14 +0000 (15:15 -0600)]
merge from 3.2

13 years agoAdd a hint that CSD == Service Pack.
Brian Curtin [Wed, 1 Feb 2012 21:14:00 +0000 (15:14 -0600)]
Add a hint that CSD == Service Pack.

People searcing for the way to get a "service pack" will never find that we
provide it here, and people that find this function won't know what CSD is
until they run the function. On top of this, they won't know what the value
means unless they really have a service pack installed.

CSD, or Customer Service Diagnostics, is apparently no longer used, and was
rarely used term at that. Most references to it online are from
universities making Windows 2000 and XP service packs available to students.

13 years agoMerge branch 3.2
Petri Lehtinen [Wed, 1 Feb 2012 20:20:41 +0000 (22:20 +0200)]
Merge branch 3.2

Closes #13676.

13 years agosqlite3: Handle strings with embedded zeros correctly
Petri Lehtinen [Wed, 1 Feb 2012 20:18:19 +0000 (22:18 +0200)]
sqlite3: Handle strings with embedded zeros correctly

Closes #13676.

13 years agomerge
Raymond Hettinger [Wed, 1 Feb 2012 17:08:08 +0000 (09:08 -0800)]
merge

13 years agoAdd usage note
Raymond Hettinger [Wed, 1 Feb 2012 17:07:40 +0000 (09:07 -0800)]
Add usage note

13 years agoIssue #13706: Add assertions to detect bugs earlier
Victor Stinner [Tue, 31 Jan 2012 23:22:23 +0000 (00:22 +0100)]
Issue #13706: Add assertions to detect bugs earlier

13 years agoFix a minor output typo as found by Terry Reedy.
Brett Cannon [Tue, 31 Jan 2012 22:02:10 +0000 (17:02 -0500)]
Fix a minor output typo as found by Terry Reedy.

13 years agoMerge
Brett Cannon [Tue, 31 Jan 2012 19:58:07 +0000 (14:58 -0500)]
Merge

13 years agoMerge with 3.2 whitespace
Terry Jan Reedy [Tue, 31 Jan 2012 07:58:39 +0000 (02:58 -0500)]
Merge with 3.2 whitespace

13 years agowhitespace
Terry Jan Reedy [Tue, 31 Jan 2012 07:57:29 +0000 (02:57 -0500)]
whitespace

13 years agoMerge 3.2
Terry Jan Reedy [Tue, 31 Jan 2012 07:42:19 +0000 (02:42 -0500)]
Merge 3.2
- Issue #13506: Add '' to path for IDLE Shell when started and restarted with Restart Shell.
  Original patches by Marco Scataglini and Roger Serwy.
Merge 3.2 #13506 Add '' to path for interactive interpreter by adding with_cwd
parameter to PyShell.PyShell.transfer_path() and changing elsewhere as needed.
Original patches by Marco Scataglini and Roger Serwy.

13 years ago#13506 Add '' to path for interactive interpreter by adding with_cwd parameter
Terry Jan Reedy [Tue, 31 Jan 2012 07:26:32 +0000 (02:26 -0500)]
#13506 Add '' to path for interactive interpreter by adding with_cwd parameter
to PyShell.PyShell.transfer_path() and changing elsewhere as needed.
Original patches by Marco Scataglini and Roger Serwy.

13 years agoLet importlib.test.benchmark take a specific benchmark name to run.
Brett Cannon [Tue, 31 Jan 2012 00:27:51 +0000 (19:27 -0500)]
Let importlib.test.benchmark take a specific benchmark name to run.

13 years agoAllow for the specification of a file to dump importlib benchmark
Brett Cannon [Tue, 31 Jan 2012 00:12:29 +0000 (19:12 -0500)]
Allow for the specification of a file to dump importlib benchmark
results to (and to compare against previous runs).
* * *
Move importlib.test.benchmark to argparse.

13 years agoFix zipimport.c's read_directory() to use appropriate types for the values
Gregory P. Smith [Mon, 30 Jan 2012 23:55:29 +0000 (15:55 -0800)]
Fix zipimport.c's read_directory() to use appropriate types for the values
being read from the header vs the values being used by fseek and ftell
(Py_ssize_t for those).  Updates the Py_BuildValue format string to match
(including several existing wrong 'i's that should have been 'l's).

13 years agoFix zip_import.c's read_directory() to use appropriate types for the values
Gregory P. Smith [Mon, 30 Jan 2012 23:17:33 +0000 (15:17 -0800)]
Fix zip_import.c's read_directory() to use appropriate types for the values
being read from the header vs the values being used by fseek and ftell
(Py_ssize_t for those) and how they are computed.  Py_ssize_t is used for
actual file offsets so that files greater than 2gigs could be supported.
Updates the Py_BuildValue format string to match (including several existing
wrong 'i's that should have been 'l's).

13 years agoIssue #8828: Add new function os.replace(), for cross-platform renaming with overwriting.
Antoine Pitrou [Mon, 30 Jan 2012 21:08:52 +0000 (22:08 +0100)]
Issue #8828: Add new function os.replace(), for cross-platform renaming with overwriting.

13 years agoIssue #13890: Also fix for extension module tests for case-insensitivity.
Brett Cannon [Mon, 30 Jan 2012 17:51:49 +0000 (12:51 -0500)]
Issue #13890: Also fix for extension module tests for case-insensitivity.

13 years agoIssue #13890: Fix importlib case-sensitivity tests to not run on Windows.
Brett Cannon [Mon, 30 Jan 2012 17:48:16 +0000 (12:48 -0500)]
Issue #13890: Fix importlib case-sensitivity tests to not run on Windows.
Thanks to os.environ under Windows only updating the dict and not the
environment itself (as exposed by nt.environ), tests using
PYTHONCASEOK always fail. Now the tests are skipped when os.environ
does not do what is expected.

13 years agomerge 3.2 (closes #13908)
Benjamin Peterson [Mon, 30 Jan 2012 01:17:07 +0000 (20:17 -0500)]
merge 3.2 (closes #13908)

13 years agoready types returned from PyType_FromSpec
Benjamin Peterson [Mon, 30 Jan 2012 01:16:37 +0000 (20:16 -0500)]
ready types returned from PyType_FromSpec

13 years agoadjust declaration
Benjamin Peterson [Mon, 30 Jan 2012 01:13:18 +0000 (20:13 -0500)]
adjust declaration

13 years agoIssue #13847: Fix test_time, time.gmtime() doesn't use localtime()
Victor Stinner [Sun, 29 Jan 2012 23:23:32 +0000 (00:23 +0100)]
Issue #13847: Fix test_time, time.gmtime() doesn't use localtime()

On Windows, localtime(-1) fails, but not gmtime(1).

13 years agoIssue #13874: read_null() of faulthandler uses volatile to avoid optimisation
Victor Stinner [Sun, 29 Jan 2012 23:07:43 +0000 (00:07 +0100)]
Issue #13874: read_null() of faulthandler uses volatile to avoid optimisation

Clang 3.0 removes "y = *x;" instruction if the optimisation level is 3.

13 years agoIssue #13848: open() and the FileIO constructor now check for NUL characters in the...
Antoine Pitrou [Sun, 29 Jan 2012 17:43:36 +0000 (18:43 +0100)]
Issue #13848: open() and the FileIO constructor now check for NUL characters in the file name.
Patch by Hynek Schlawack.

13 years agoIssue #13848: open() and the FileIO constructor now check for NUL characters in the...
Antoine Pitrou [Sun, 29 Jan 2012 17:36:34 +0000 (18:36 +0100)]
Issue #13848: open() and the FileIO constructor now check for NUL characters in the file name.
Patch by Hynek Schlawack.

13 years agoIssue #6774: Back out c8b77efe8b56, which only brings confusion.
Charles-François Natali [Sun, 29 Jan 2012 15:42:54 +0000 (16:42 +0100)]
Issue #6774: Back out c8b77efe8b56, which only brings confusion.

13 years agomerge with 3.2
Georg Brandl [Sun, 29 Jan 2012 14:38:55 +0000 (15:38 +0100)]
merge with 3.2

13 years agoFix #13900: resolve self-referential description of a parameter.
Georg Brandl [Sun, 29 Jan 2012 14:38:47 +0000 (15:38 +0100)]
Fix #13900: resolve self-referential description of a parameter.

13 years agomerge 3.2
Benjamin Peterson [Sun, 29 Jan 2012 01:33:52 +0000 (20:33 -0500)]
merge 3.2

13 years agoremove tests from really old regex module
Benjamin Peterson [Sun, 29 Jan 2012 01:33:21 +0000 (20:33 -0500)]
remove tests from really old regex module

13 years agoIssue #13806: The size check in audioop decompression functions was too strict and...
Antoine Pitrou [Sat, 28 Jan 2012 21:02:47 +0000 (22:02 +0100)]
Issue #13806: The size check in audioop decompression functions was too strict and could reject valid compressed data.
Patch by Oleg Plakhotnyuk.

13 years agoIssue #13806: The size check in audioop decompression functions was too strict and...
Antoine Pitrou [Sat, 28 Jan 2012 21:01:59 +0000 (22:01 +0100)]
Issue #13806: The size check in audioop decompression functions was too strict and could reject valid compressed data.
Patch by Oleg Plakhotnyuk.

13 years agoIssue #13895: fix test_ssl hanging under Ubuntu
Antoine Pitrou [Sat, 28 Jan 2012 16:40:23 +0000 (17:40 +0100)]
Issue #13895: fix test_ssl hanging under Ubuntu

13 years agoIssue #13895: fix test_ssl hanging under Ubuntu
Antoine Pitrou [Sat, 28 Jan 2012 16:38:34 +0000 (17:38 +0100)]
Issue #13895: fix test_ssl hanging under Ubuntu

13 years agoFix typo in comment.
Nadeem Vawda [Sat, 28 Jan 2012 15:32:47 +0000 (17:32 +0200)]
Fix typo in comment.

13 years agoIssue #13894: Skip test_threading.CRLockTests if _CRLock isn't available. Patch
Charles-François Natali [Sat, 28 Jan 2012 10:37:42 +0000 (11:37 +0100)]
Issue #13894: Skip test_threading.CRLockTests if _CRLock isn't available. Patch
by Matt Joiner.

13 years agoIssue #13894: Skip test_threading.CRLockTests if _CRLock isn't available. Patch
Charles-François Natali [Sat, 28 Jan 2012 10:36:04 +0000 (11:36 +0100)]
Issue #13894: Skip test_threading.CRLockTests if _CRLock isn't available. Patch
by Matt Joiner.

13 years agoMerge 3.2 -> default (issue 13889)
Mark Dickinson [Fri, 27 Jan 2012 21:17:04 +0000 (21:17 +0000)]
Merge 3.2 -> default (issue 13889)

13 years agoIssue #13889: Add missing _Py_SET_53BIT_PRECISION_* calls around uses of dtoa.c funct...
Mark Dickinson [Fri, 27 Jan 2012 21:16:01 +0000 (21:16 +0000)]
Issue #13889: Add missing _Py_SET_53BIT_PRECISION_* calls around uses of dtoa.c functions in float round.

13 years agoFix intermittent test_ssl failure.
Antoine Pitrou [Fri, 27 Jan 2012 16:34:56 +0000 (17:34 +0100)]
Fix intermittent test_ssl failure.

13 years agoFix intermittent test_ssl failure.
Antoine Pitrou [Fri, 27 Jan 2012 16:33:01 +0000 (17:33 +0100)]
Fix intermittent test_ssl failure.

13 years agomerge heads (#13887)
Benjamin Peterson [Fri, 27 Jan 2012 14:14:29 +0000 (09:14 -0500)]
merge heads (#13887)

13 years agonote that get() is not affected by default_factory (closes #13887)
Benjamin Peterson [Fri, 27 Jan 2012 14:14:01 +0000 (09:14 -0500)]
note that get() is not affected by default_factory (closes #13887)

13 years agoFix error handling in examples of C API use.
Antoine Pitrou [Fri, 27 Jan 2012 13:08:04 +0000 (14:08 +0100)]
Fix error handling in examples of C API use.

13 years agoFix error handling in examples of C API use.
Antoine Pitrou [Fri, 27 Jan 2012 13:07:29 +0000 (14:07 +0100)]
Fix error handling in examples of C API use.

13 years agoIssue #13812: When a multiprocessing Process child raises an exception, flush stderr...
Antoine Pitrou [Fri, 27 Jan 2012 09:53:35 +0000 (10:53 +0100)]
Issue #13812: When a multiprocessing Process child raises an exception, flush stderr after printing the exception traceback.

13 years agoIssue #13812: When a multiprocessing Process child raises an exception, flush stderr...
Antoine Pitrou [Fri, 27 Jan 2012 09:52:37 +0000 (10:52 +0100)]
Issue #13812: When a multiprocessing Process child raises an exception, flush stderr after printing the exception traceback.

13 years agoClarify that ssl.OP_ALL can be different from OpenSSL's SSL_OP_ALL.
Antoine Pitrou [Fri, 27 Jan 2012 09:03:23 +0000 (10:03 +0100)]
Clarify that ssl.OP_ALL can be different from OpenSSL's SSL_OP_ALL.

13 years agoClarify that ssl.OP_ALL can be different from OpenSSL's SSL_OP_ALL.
Antoine Pitrou [Fri, 27 Jan 2012 09:02:55 +0000 (10:02 +0100)]
Clarify that ssl.OP_ALL can be different from OpenSSL's SSL_OP_ALL.

13 years agoIssue #13885: CVE-2011-3389: the _ssl module would always disable the CBC IV attack...
Antoine Pitrou [Fri, 27 Jan 2012 08:53:29 +0000 (09:53 +0100)]
Issue #13885: CVE-2011-3389: the _ssl module would always disable the CBC IV attack countermeasure.