| bd61c3a | Larry Hastings | 20 October 2013, 09:01:29 UTC | Version bump for 3.4.0a4. | 20 October 2013, 09:01:29 UTC |
| f32ce7f | Larry Hastings | 20 October 2013, 08:59:09 UTC | Rebuild pydoc topics for 3.4.0a4 release. | 20 October 2013, 08:59:09 UTC |
| ce82eb2 | Serhiy Storchaka | 20 October 2013, 06:42:26 UTC | Issue #12866: Fix bias() for 24-bit. Add more tests. | 20 October 2013, 06:42:26 UTC |
| d49c47b | Guido van Rossum | 20 October 2013, 04:26:34 UTC | Break out of loop on EOF in asyncio echo test programs. | 20 October 2013, 04:26:34 UTC |
| 0acceb7 | Nick Coghlan | 20 October 2013, 03:22:21 UTC | What's New updates prior to alpha | 20 October 2013, 03:22:21 UTC |
| 4dae27a | Nick Coghlan | 20 October 2013, 03:22:04 UTC | Tweak 'provisional' in glossary | 20 October 2013, 03:22:04 UTC |
| 92945c5 | Antoine Pitrou | 20 October 2013, 00:16:40 UTC | Un-backout e3ec6b17260c - the test fix was actually also necessary | 20 October 2013, 00:16:40 UTC |
| c1b7841 | Antoine Pitrou | 20 October 2013, 00:09:08 UTC | Backout e3ec6b17260c after Guido's fix | 20 October 2013, 00:09:08 UTC |
| 32e4685 | Guido van Rossum | 20 October 2013, 00:04:25 UTC | (Hopefully) proper fix for gentoo buildbot failure due to lacking AF_INET6 support. This should supersede revision e3ec6b17260c (but please test before removing that). | 20 October 2013, 00:04:25 UTC |
| d20afad | Antoine Pitrou | 19 October 2013, 23:51:25 UTC | Issue #19305: try to fix sporadic test_asyncio failure on FreeBSD 10.0 | 19 October 2013, 23:51:25 UTC |
| 0d9eefd | Antoine Pitrou | 19 October 2013, 23:10:52 UTC | Try to fix test_asyncio dual stack test when creating an IPv6 socket fails | 19 October 2013, 23:10:52 UTC |
| 4ca7355 | Antoine Pitrou | 19 October 2013, 22:54:10 UTC | Issue #19299: fix refleak test failures in test_asyncio | 19 October 2013, 22:54:10 UTC |
| 34d33a5 | Benjamin Peterson | 19 October 2013, 20:15:58 UTC | removal u_qualname, since compiler_scope_qualname is only ever called once | 19 October 2013, 20:15:58 UTC |
| a8a38b8 | Benjamin Peterson | 19 October 2013, 20:14:39 UTC | strengthen condition and add assertion | 19 October 2013, 20:14:39 UTC |
| ebe83f9 | Christian Heimes | 19 October 2013, 20:36:17 UTC | fix compile issue on windows. path is now a struct ptr | 19 October 2013, 20:36:17 UTC |
| e55013f | Antoine Pitrou | 19 October 2013, 20:06:26 UTC | Issue #18235: Fix the sysconfig variables LDSHARED and BLDSHARED under AIX. Patch by David Edelsohn. | 19 October 2013, 20:06:26 UTC |
| 0abb218 | Antoine Pitrou | 19 October 2013, 20:05:05 UTC | Issue #18235: Fix the sysconfig variables LDSHARED and BLDSHARED under AIX. Patch by David Edelsohn. | 19 October 2013, 20:05:05 UTC |
| 3d9e481 | Benjamin Peterson | 19 October 2013, 20:01:13 UTC | give explicitly global functions and classes a global __qualname__ (closes #19301) | 19 October 2013, 20:01:13 UTC |
| 3586673 | Ned Deily | 19 October 2013, 19:10:01 UTC | Issue #19304: Fix typos noted by Claudiu Popa. | 19 October 2013, 19:10:01 UTC |
| f5e987b | Larry Hastings | 19 October 2013, 18:50:09 UTC | Issue #18606: Add the new "statistics" module (PEP 450). Contributed by Steven D'Aprano. | 19 October 2013, 18:50:09 UTC |
| aa2b22a | Serhiy Storchaka | 19 October 2013, 18:39:31 UTC | Restore prototypes for the 'openpty' and 'forkpty' on BSDI (broken in issue #1772673). | 19 October 2013, 18:39:31 UTC |
| dd06d14 | Serhiy Storchaka | 19 October 2013, 18:15:55 UTC | Null merge | 19 October 2013, 18:15:55 UTC |
| 0e071c9 | Serhiy Storchaka | 19 October 2013, 18:14:57 UTC | Fixed tests for issue #19279. | 19 October 2013, 18:14:57 UTC |
| eaea5e9 | Serhiy Storchaka | 19 October 2013, 18:10:46 UTC | Issue #12866: The audioop module now supports 24-bit samples. | 19 October 2013, 18:10:46 UTC |
| c679227 | Serhiy Storchaka | 19 October 2013, 18:03:34 UTC | Issue #1772673: The type of `char*` arguments now changed to `const char*`. | 19 October 2013, 18:03:34 UTC |
| 80ab130 | Christian Heimes | 19 October 2013, 17:55:38 UTC | Ignore PyCharm project directory | 19 October 2013, 17:55:38 UTC |
| b649ed7 | Christian Heimes | 19 October 2013, 17:55:27 UTC | Ignore PyCharm project directory | 19 October 2013, 17:55:27 UTC |
| 0fbd94c | Christian Heimes | 19 October 2013, 17:40:49 UTC | Issue #18582: skip test of _hashlib.pbkdf2_hmac if OpenSSL is not available or too old | 19 October 2013, 17:40:49 UTC |
| 55e092f | Serhiy Storchaka | 19 October 2013, 17:39:28 UTC | Issue #19279: UTF-7 decoder no more produces illegal strings. | 19 October 2013, 17:39:28 UTC |
| 35804e4 | Serhiy Storchaka | 19 October 2013, 17:38:19 UTC | Issue #19279: UTF-7 decoder no more produces illegal strings. | 19 October 2013, 17:38:19 UTC |
| f19a6ef | Guido van Rossum | 19 October 2013, 16:52:09 UTC | Verify hostname if verify_mode is CERT_OPTIONAL too. | 19 October 2013, 16:52:09 UTC |
| 3743711 | Guido van Rossum | 19 October 2013, 16:10:13 UTC | Disable some subprocess tests that hang on AIX. See http://bugs.python.org/issue19293 | 19 October 2013, 16:10:13 UTC |
| 351f539 | Christian Heimes | 19 October 2013, 15:59:48 UTC | Issue #18582: HMAC_CTX_copy() is not available on OpenSSL < 1.0 | 19 October 2013, 15:59:48 UTC |
| 7058dad | Guido van Rossum | 19 October 2013, 15:47:26 UTC | Skip the asyncio tests when threads are not available. See http://bugs.python.org/issue19295 | 19 October 2013, 15:47:26 UTC |
| 0f5bff2 | Georg Brandl | 19 October 2013, 15:46:38 UTC | Doc markup fixes. | 19 October 2013, 15:46:38 UTC |
| 8608d26 | Nick Coghlan | 19 October 2013, 14:30:51 UTC | contextlib doc updates and refactoring - explain single use, reusable and reentrant in docs - converted suppress to a reentrant class based impl - converted redirect_stdout to a reusable impl - moved both suppress and redirect_stdout behind a functional facade - added reentrancy tests for the updated suppress - added reusability tests for the updated redirect_stdio - slightly cleaned up an exception from contextmanager | 19 October 2013, 14:30:51 UTC |
| e723622 | Christian Heimes | 19 October 2013, 12:24:44 UTC | Issue #18582: provide a faster C implementation of pbkdf2_hmac that works with OpenSSL < 1.0 | 19 October 2013, 12:24:44 UTC |
| fcd8de2 | Christian Heimes | 19 October 2013, 12:24:03 UTC | lower case pbkdf2_hmac | 19 October 2013, 12:24:03 UTC |
| 3626a50 | Christian Heimes | 19 October 2013, 12:12:02 UTC | Issue #19254: Provide an optimized Python implementation of PBKDF2_HMAC | 19 October 2013, 12:12:02 UTC |
| a412f76 | Serhiy Storchaka | 19 October 2013, 07:45:48 UTC | Fix compilation of the curses module (broken by issue #16612). | 19 October 2013, 07:45:48 UTC |
| 3182680 | Larry Hastings | 19 October 2013, 07:09:25 UTC | Issue #16612: Add "Argument Clinic", a compile-time preprocessor for C files to generate argument parsing code. (See PEP 436.) | 19 October 2013, 07:09:25 UTC |
| 5ceae41 | Ned Deily | 19 October 2013, 04:34:58 UTC | Ensure setup.py looks for zlib.h in an OS X SDK. | 19 October 2013, 04:34:58 UTC |
| 507c591 | Ned Deily | 19 October 2013, 04:32:00 UTC | Ensure setup.py looks for zlib.h in an OS X SDK. | 19 October 2013, 04:32:00 UTC |
| 29eec66 | Ned Deily | 19 October 2013, 04:16:05 UTC | Issue #19019: Change the OS X installer build script to use CFLAGS instead of OPT for special build options. By setting OPT, some compiler-specific options like -fwrapv were overridden and thus not used, which could result in broken interpreters when building with clang. | 19 October 2013, 04:16:05 UTC |
| 87adb6e | Ned Deily | 19 October 2013, 04:09:56 UTC | Issue #14499: Fix several problems with OS X universal build support: 1. ppc arch detection for extension module builds broke with Xcode 5 2. ppc arch detection in configure did not work on OS X 10.4 3. -sysroot and -arch flags were unnecessarily duplicated 4. there was no obvious way to configure an intel-32 only build. | 19 October 2013, 04:09:56 UTC |
| ea41d5f | Ned Deily | 19 October 2013, 03:49:27 UTC | Issue #15663: Update OS X installer to use Tcl/Tk 8.5.15. | 19 October 2013, 03:49:27 UTC |
| 5375b49 | Ned Deily | 19 October 2013, 03:42:32 UTC | Update OS X installer for building on 10.9. | 19 October 2013, 03:42:32 UTC |
| ac25ca1 | Ned Deily | 19 October 2013, 03:41:16 UTC | Update OS X installer for building on 10.9. | 19 October 2013, 03:41:16 UTC |
| c548660 | Victor Stinner | 19 October 2013, 00:40:16 UTC | Issue #16742: My fix on PyOS_StdioReadline() was incomplete, PyMem_FREE() was not patched | 19 October 2013, 00:40:16 UTC |
| 355491d | Guido van Rossum | 18 October 2013, 22:17:11 UTC | Write flow control for asyncio (includes asyncio.streams overhaul). | 18 October 2013, 22:17:11 UTC |
| 051a331 | Guido van Rossum | 18 October 2013, 22:15:56 UTC | Relax some asyncio test timeouts (http://bugs.python.org/issue19285). | 18 October 2013, 22:15:56 UTC |
| ee1be99 | Nadeem Vawda | 18 October 2013, 22:11:13 UTC | Issue #19222: Add support for the 'x' mode to the gzip module. Original patch by Tim Heaney. | 18 October 2013, 22:11:13 UTC |
| 8a9e99c | Nadeem Vawda | 18 October 2013, 22:11:06 UTC | Issue #19223: Add support for the 'x' mode to the bz2 module. Patch by Tim Heaney and Vajrasky Kok. | 18 October 2013, 22:11:06 UTC |
| 42ca982 | Nadeem Vawda | 18 October 2013, 22:06:19 UTC | Issue #19201: Add support for the 'x' mode to the lzma module. Patch by Tim Heaney and Vajrasky Kok. | 18 October 2013, 22:06:19 UTC |
| c516815 | Brett Cannon | 18 October 2013, 20:55:15 UTC | Fix test to not consider mixin a standalone test | 18 October 2013, 20:55:15 UTC |
| 1340049 | Brett Cannon | 18 October 2013, 19:40:11 UTC | Issue #16803: Move test_importlib.test_util to use both frozen and source code. | 18 October 2013, 19:40:11 UTC |
| 6a57dd8 | Brett Cannon | 18 October 2013, 19:12:21 UTC | Issue #16803: Have test_importlib.test_locks use frozen and source code. | 18 October 2013, 19:12:21 UTC |
| baced56 | Brett Cannon | 18 October 2013, 18:03:16 UTC | Fix a refleak in _sre | 18 October 2013, 18:03:16 UTC |
| 1830601 | Brett Cannon | 18 October 2013, 17:29:04 UTC | Add NEWS entry for issue #18810 | 18 October 2013, 17:29:04 UTC |
| 58f5680 | Brett Cannon | 18 October 2013, 17:24:13 UTC | Issue #18810: Be optimistic with stat calls when seeing if a directory exists when checking for a package. Before there was an isdir check and then various isfile checks for possible __init__ files when looking for a package. This change drops the isdir check by leaning on the assumption that a directory will not contain something named after the module being imported which is not a directory. If the module is a package then it saves a stat call. If there is nothing in the directory with the potential package name it also saves a stat call. Only if there is something in the directory named the same thing as the potential package will the number of stat calls increase (due to more wasteful __init__ checks). Semantically there is no change as the isdir check moved down so that namespace packages continue to have no chance of accidentally collecting non-existent directories. | 18 October 2013, 17:24:13 UTC |
| 2546a17 | Guido van Rossum | 18 October 2013, 17:10:36 UTC | Important race condition fix for Tulip. | 18 October 2013, 17:10:36 UTC |
| 559ae0f | Benjamin Peterson | 18 October 2013, 16:58:17 UTC | merge 3.3 | 18 October 2013, 16:58:17 UTC |
| 910a665 | Benjamin Peterson | 18 October 2013, 16:57:55 UTC | fix description of super() behavior on descriptors | 18 October 2013, 16:57:55 UTC |
| 48df60c | Richard Oudkerk | 18 October 2013, 16:04:31 UTC | Update more configurations in pcbuild.sln. | 18 October 2013, 16:04:31 UTC |
| bb9e481 | Brett Cannon | 18 October 2013, 16:01:06 UTC | Issue #18416: Fix various os calls in importlib.machinery.FileFinder now that self.path is no longer forced to '.'. | 18 October 2013, 16:01:06 UTC |
| af002e6 | Brett Cannon | 18 October 2013, 15:39:32 UTC | merge | 18 October 2013, 15:39:32 UTC |
| 27e27f7 | Brett Cannon | 18 October 2013, 15:39:04 UTC | Issue #18416: Have importlib.machinery.PathFinder treat '' as the cwd and stop importlib.machinery.FileFinder treating '' as '.'. Previous PathFinder transformed '' into '.' which led to __file__ for modules imported from the cwd to always be relative paths. This meant the values of the attribute were wrong as soon as the cwd changed. This change now means that as long as the site module is run (which makes all entries in sys.path absolute) then all values for __file__ will also be absolute unless it's for __main__ when specified by file path in a relative way (modules imported by runpy will have an absolute path). Now that PathFinder is no longer treating '' as '.' it only makes sense for FileFinder to stop doing so as well. Now no transformation is performed for the directory given to the __init__ method. Thanks to Madison May for the initial patch. | 18 October 2013, 15:39:04 UTC |
| 44455e8 | Richard Oudkerk | 18 October 2013, 15:23:01 UTC | Update pcbuild.sln to build _overlapped. | 18 October 2013, 15:23:01 UTC |
| 57497ad | Guido van Rossum | 18 October 2013, 14:58:20 UTC | Rename Transport.pause/resume to pause_reading/pause_writing. Also relax timeout in test_call_later(). | 18 October 2013, 14:58:20 UTC |
| 40b22d0 | Brett Cannon | 18 October 2013, 14:45:59 UTC | Issue #16803: test.test_importlib.test_api now runs under frozen and source. | 18 October 2013, 14:45:59 UTC |
| 87efae2 | Serhiy Storchaka | 18 October 2013, 14:17:31 UTC | Remove redundant empty lines. | 18 October 2013, 14:17:31 UTC |
| 90b41f0 | Serhiy Storchaka | 18 October 2013, 14:16:40 UTC | Remove redundant empty lines. | 18 October 2013, 14:16:40 UTC |
| b39d247 | Serhiy Storchaka | 18 October 2013, 14:05:41 UTC | Remove a duplicate. | 18 October 2013, 14:05:41 UTC |
| 3321fb8 | Nick Coghlan | 18 October 2013, 13:59:58 UTC | Issue #16129: this should appease the buildbots | 18 October 2013, 13:59:58 UTC |
| 8acc3c5 | Richard Oudkerk | 18 October 2013, 13:42:56 UTC | Remove test_sigterm(). | 18 October 2013, 13:42:56 UTC |
| 1805a62 | Nick Coghlan | 18 October 2013, 13:11:47 UTC | Issue #16129: Py_SetStandardStreamEncoding cleanups - don't call PyErr_NoMemory with interpreter is not initialised - note that it's OK to call _PyMem_RawStrDup here - don't include this in the limited API - capitalise "IO" - be explicit that a non-zero return indicates an error - include versionadded marker in docs | 18 October 2013, 13:11:47 UTC |
| ac1a248 | Nick Coghlan | 18 October 2013, 12:39:50 UTC | Close #19284: Handle -R properly in flag helper Previously, the -R option would be specified multiple times if PYTHONHASHSEED was set. | 18 October 2013, 12:39:50 UTC |
| ef23f18 | Serhiy Storchaka | 18 October 2013, 08:55:30 UTC | Fix markup. | 18 October 2013, 08:55:30 UTC |
| 3353313 | Serhiy Storchaka | 18 October 2013, 08:55:02 UTC | Fix markup. | 18 October 2013, 08:55:02 UTC |
| a8b0707 | Ethan Furman | 18 October 2013, 08:22:08 UTC | Issue #19030: special-cased __dict__ as the actual dict is not returned, a proxy is. | 18 October 2013, 08:22:08 UTC |
| 2498d9e | Ethan Furman | 18 October 2013, 07:45:40 UTC | Issue #19272: slight clarification of pickle docs with regard to lambda. | 18 October 2013, 07:45:40 UTC |
| 63c141c | Ethan Furman | 18 October 2013, 07:27:39 UTC | Close #19030: inspect.getmembers and inspect.classify_class_attrs Order of search is now: 1. Try getattr 2. If that throws an exception, check __dict__ directly 3. If still not found, walk the mro looking for the eldest class that has the attribute (e.g. things returned by __getattr__) 4. If none of that works (e.g. due to a buggy __dir__, __getattr__, etc. method or missing __slot__ attribute), ignore the attribute entirely. | 18 October 2013, 07:27:39 UTC |
| 0e0cd46 | Ethan Furman | 18 October 2013, 02:34:12 UTC | Catching up on NEWS entries. I'll make sure and include them in future patches. | 18 October 2013, 02:34:12 UTC |
| 3c97e46 | Ned Deily | 18 October 2013, 01:08:00 UTC | Issue #19262: Install asyncio and test_asyncio directories. | 18 October 2013, 01:08:00 UTC |
| fc29e0f | Guido van Rossum | 17 October 2013, 22:39:45 UTC | Rename the logger to plain "logger". | 17 October 2013, 22:39:45 UTC |
| b795aa8 | Ned Deily | 17 October 2013, 22:21:40 UTC | Issue #19275: Fix test_site failure on OS X due to typo. | 17 October 2013, 22:21:40 UTC |
| 5ea7f93 | Guido van Rossum | 17 October 2013, 21:23:17 UTC | Make asyncio tests run on Windows. | 17 October 2013, 21:23:17 UTC |
| 27b7c7e | Guido van Rossum | 17 October 2013, 20:40:50 UTC | Initial checkin of asyncio package (== Tulip, == PEP 3156). | 17 October 2013, 20:40:50 UTC |
| 5b37f97 | Serhiy Storchaka | 17 October 2013, 20:05:19 UTC | Issue #19276: Fixed the wave module on 64-bit big-endian platforms. | 17 October 2013, 20:05:19 UTC |
| d3b7505 | Serhiy Storchaka | 17 October 2013, 20:04:04 UTC | Issue #19276: Fixed the wave module on 64-bit big-endian platforms. | 17 October 2013, 20:04:04 UTC |
| 2bb0ac0 | Georg Brandl | 17 October 2013, 17:52:33 UTC | merge with 3.3 | 17 October 2013, 17:52:33 UTC |
| 9ff06dc | Georg Brandl | 17 October 2013, 17:51:34 UTC | Fix inaccurate versionchanged tag for compressobj(): most parameter were there, just got kwarg support. | 17 October 2013, 17:51:34 UTC |
| 8f9571e | Nick Coghlan | 17 October 2013, 15:46:19 UTC | Skip #16129 test until I debug cross-platform issues | 17 October 2013, 15:46:19 UTC |
| 6508dc5 | Nick Coghlan | 17 October 2013, 15:44:22 UTC | Try to debug overspecified test :( | 17 October 2013, 15:44:22 UTC |
| cc5c728 | Richard Oudkerk | 17 October 2013, 14:22:10 UTC | Fix signal handler in test. | 17 October 2013, 14:22:10 UTC |
| 240f86d | Nick Coghlan | 17 October 2013, 13:40:57 UTC | Close #19266: contextlib.ignore -> contextlib.suppress Patch by Zero Piraeus. | 17 October 2013, 13:40:57 UTC |
| 1eb509a | Richard Oudkerk | 17 October 2013, 13:31:51 UTC | Merge. | 17 October 2013, 13:31:51 UTC |
| a0f074f | Nick Coghlan | 17 October 2013, 13:27:17 UTC | Issue #16129: Move Py_SetStandardStreamEncoding declaration | 17 October 2013, 13:27:17 UTC |
| 1e2f67c | Richard Oudkerk | 17 October 2013, 13:24:06 UTC | Try doing a raw test of os.fork()/os.kill(). | 17 October 2013, 13:24:06 UTC |
| 7d270ee | Nick Coghlan | 17 October 2013, 12:35:35 UTC | Issue #16129: Add `Py_SetStandardStreamEncoding` This new pre-initialization API allows embedding applications like Blender to force a particular encoding and error handler for the standard IO streams. Also refactors Modules/_testembed.c to let us start testing multiple embedding scenarios. (Initial patch by Bastien Montagne) | 17 October 2013, 12:35:35 UTC |
