| 7f5c22c | Eli Bendersky | 02 September 2013, 15:57:21 UTC | Refactor the main function of regrtest a bit. Moving subprocess execution of tests into a function. | 02 September 2013, 15:57:21 UTC |
| a35adf5 | Raymond Hettinger | 02 September 2013, 10:23:21 UTC | Instead of XORed indicies, switch to a hybrid of linear probing and open addressing. Modern processors tend to make consecutive memory accesses cheaper than random probes into memory. Small sets can fit into L1 cache, so they get less benefit. But they do come out ahead because the consecutive probes don't probe the same key more than once and because the randomization step occurs less frequently (or not at all). For the open addressing step, putting the perturb shift before the index calculation gets the upper bits into play sooner. | 02 September 2013, 10:23:21 UTC |
| a661f45 | Ethan Furman | 02 September 2013, 08:14:56 UTC | Close #18745: Improve enum tests in test_json for infinities and NaN. | 02 September 2013, 08:14:56 UTC |
| 51bae47 | Terry Jan Reedy | 01 September 2013, 23:04:18 UTC | Merge with 3.3 | 01 September 2013, 23:04:18 UTC |
| a0f482a | Terry Jan Reedy | 01 September 2013, 23:03:41 UTC | Remove obsolete .hgeol entry pointing to file moved elsewhere. This kine was already replaced by Lib/test/test_email/data/msg_26.txt = BIN which is just below the last line in the patch context. | 01 September 2013, 23:03:41 UTC |
| 56dadf7 | Tim Peters | 01 September 2013, 21:01:46 UTC | Merge fix from 3.3 into default. Fix issue 18889: test_sax: multiple failures on Windows desktop. "The fix" is to tell Mercurial that the test files are binary. Windows developers: to get the correct line endings in your checkout, delete Lib\test\xmltestdata, and then "hg revert" that directory. Why the Windows buildbots didn't fail test_sax remains a mystery :-( | 01 September 2013, 21:01:46 UTC |
| 9d2e3a1 | Tim Peters | 01 September 2013, 20:56:22 UTC | Fix issue 18889: test_sax: multiple failures on Windows desktop. "The fix" is to tell Mercurial that the test files are binary. Windows developers: to get the correct line endings in your checkout, delete Lib\test\xmltestdata, and then "hg revert" that directory. Why the Windows buildbots didn't fail test_sax remains a mystery :-( | 01 September 2013, 20:56:22 UTC |
| 625eed7 | Antoine Pitrou | 01 September 2013, 17:52:08 UTC | test_threading isn't rudimentary anymore | 01 September 2013, 17:52:08 UTC |
| 4c8ce84 | Antoine Pitrou | 01 September 2013, 17:51:49 UTC | test_threading isn't rudimentary anymore | 01 September 2013, 17:51:49 UTC |
| f6fa22e | Victor Stinner | 01 September 2013, 08:22:41 UTC | Issue #18571: Merge duplicate test code Merge test/subprocessdata/inherited.py into test/subprocessdata/fd_status.py | 01 September 2013, 08:22:41 UTC |
| 8913a6c | Andrew Svetlov | 01 September 2013, 04:58:41 UTC | Issue #11798: fix tests for regrtest -R : | 01 September 2013, 04:58:41 UTC |
| 59360aa | Tim Peters | 01 September 2013, 04:44:34 UTC | Back out 868ad6fa8e68 - it left all the buildbots failing. Unclear to me why it was pushed to begin with. See issue 11798. Perhaps it's because regrtest with -R was failing? Fine, but that's better than regrtest _always_ failing ;-) | 01 September 2013, 04:44:34 UTC |
| 6c3c1cc | Raymond Hettinger | 01 September 2013, 04:34:24 UTC | Update copyright. | 01 September 2013, 04:34:24 UTC |
| 264683e | Raymond Hettinger | 01 September 2013, 04:28:58 UTC | merge | 01 September 2013, 04:28:58 UTC |
| 95c0d67 | Raymond Hettinger | 01 September 2013, 04:27:08 UTC | Further reduce the cost of hash collisions by inspecting an additional nearby entry. | 01 September 2013, 04:27:08 UTC |
| ec15a82 | Ethan Furman | 01 September 2013, 02:17:41 UTC | Close #18738: Route __format__ calls to mixed-in type for mixed Enums (such as IntEnum). | 01 September 2013, 02:17:41 UTC |
| 34567ec | Eli Bendersky | 31 August 2013, 22:18:48 UTC | Update whatsnew/3.4.rst wrt. the socket constants switch to IntEnum [issue #18730] | 31 August 2013, 22:18:48 UTC |
| b2ff3cf | Eli Bendersky | 31 August 2013, 22:13:30 UTC | Switch the AF_* and SOCK_* constants in the socket module to IntEnum. Closes #18720. | 31 August 2013, 22:13:30 UTC |
| 7e7cf8b | Terry Jan Reedy | 31 August 2013, 21:16:45 UTC | Issue #12037: Fix test_email for desktop Windows. | 31 August 2013, 21:16:45 UTC |
| 740d6b6 | Terry Jan Reedy | 31 August 2013, 21:12:21 UTC | Issue #12037: Fix test_email for desktop Windows. | 31 August 2013, 21:12:21 UTC |
| 2e87c87 | Terry Jan Reedy | 31 August 2013, 20:28:53 UTC | Merge from 3.3 #18489 Search Engine tests | 31 August 2013, 20:28:53 UTC |
| 31e4d32 | Terry Jan Reedy | 31 August 2013, 20:27:16 UTC | Issue #18489: Add complete, gui-free tests for idlelib.SearchEngine. Patch import and initialization in SearchEngine to make testing easier. Improve docstrings, especially to clarify the double role of 'ok' parameters. Original patch by Phil Webster. | 31 August 2013, 20:27:16 UTC |
| 13bdfa7 | Ethan Furman | 31 August 2013, 19:48:51 UTC | Issue #18780: code cleanup. | 31 August 2013, 19:48:51 UTC |
| a412137 | Andrew Svetlov | 31 August 2013, 17:55:25 UTC | Temporary disable tests cleanup (issue 11798). | 31 August 2013, 17:55:25 UTC |
| fb13721 | Ethan Furman | 31 August 2013, 17:18:55 UTC | Close #18780: %-formatting now prints value for int subclasses with %d, %i, and %u codes. | 31 August 2013, 17:18:55 UTC |
| 08548f4 | Charles-François Natali | 31 August 2013, 15:32:30 UTC | Merge. | 31 August 2013, 15:32:30 UTC |
| 2c68e30 | Eli Bendersky | 31 August 2013, 14:37:23 UTC | Fix XMLPullParser documentation to say "non-blocking" instead of "asynchronous". The latter is more ambiguous. Related to issue #17741 | 31 August 2013, 14:37:23 UTC |
| 24aa041 | Charles-François Natali | 31 August 2013, 12:48:25 UTC | Use INADDR_BROADCAST instead of hard-coded value (it's part of POSIX and already appears without #ifdef a couple lines above). | 31 August 2013, 12:48:25 UTC |
| fcfb324 | Charles-François Natali | 31 August 2013, 12:40:49 UTC | Use the recent support.HOSTv6 addition. | 31 August 2013, 12:40:49 UTC |
| 4879a96 | Antoine Pitrou | 30 August 2013, 22:26:02 UTC | Issue #18756: os.urandom() now uses a lazily-opened persistent file descriptor, so as to avoid using many file descriptors when run in parallel from multiple threads. | 30 August 2013, 22:26:02 UTC |
| 267964c | Antoine Pitrou | 30 August 2013, 22:12:37 UTC | Forward port new tests from Issue #18851. | 30 August 2013, 22:12:37 UTC |
| a839271 | Antoine Pitrou | 30 August 2013, 21:38:13 UTC | Forward port new tests from Issue #18851. | 30 August 2013, 21:38:13 UTC |
| 79a53ea | Charles-François Natali | 30 August 2013, 21:34:26 UTC | Issue #18418: After fork(), reinit all threads states, not only active ones. Patch by A. Jesse Jiryu Davis. | 30 August 2013, 21:34:26 UTC |
| 9939cc8 | Charles-François Natali | 30 August 2013, 21:32:53 UTC | Issue #18418: After fork(), reinit all threads states, not only active ones. Patch by A. Jesse Jiryu Davis. | 30 August 2013, 21:32:53 UTC |
| b586934 | Eli Bendersky | 30 August 2013, 12:51:20 UTC | Issue #17741: Rename IncrementalParser and its methods. The new names are hopefully more descriptive and consistent. If you feel you don't agree with this change, *please* read issue 17741 first - there's a lot of discussion in there. | 30 August 2013, 12:51:20 UTC |
| 1221f6b | Philip Jenvey | 30 August 2013, 01:33:50 UTC | utilize int.from_bytes | 30 August 2013, 01:33:50 UTC |
| da0bea2 | Benjamin Peterson | 29 August 2013, 21:29:30 UTC | improve comment | 29 August 2013, 21:29:30 UTC |
| f908efb | Benjamin Peterson | 29 August 2013, 21:27:57 UTC | make lists uniformly integers for the benefit of pypy | 29 August 2013, 21:27:57 UTC |
| 6cc5039 | Gregory P. Smith | 29 August 2013, 20:39:44 UTC | Fixes Issue #15507: test_subprocess's test_send_signal could fail if the test runner were run in an environment where the process inherited an ignore setting for SIGINT. Restore the SIGINT handler to the desired KeyboardInterrupt raising one during that test. | 29 August 2013, 20:39:44 UTC |
| dee0434 | Gregory P. Smith | 29 August 2013, 20:35:27 UTC | Fixes issue #15507: test_subprocess's test_send_signal could fail if the test runner were run in an environment where the process inherited an ignore setting for SIGINT. Restore the SIGINT handler to the desired KeyboardInterrupt raising one during that test. | 29 August 2013, 20:35:27 UTC |
| 2f43d07 | Tim Peters | 29 August 2013, 17:35:37 UTC | Null merge of 3.3 into default. | 29 August 2013, 17:35:37 UTC |
| f25f80c | Charles-François Natali | 29 August 2013, 17:27:17 UTC | Merge. | 29 August 2013, 17:27:17 UTC |
| e33d5b0 | Charles-François Natali | 29 August 2013, 17:26:37 UTC | Merge. | 29 August 2013, 17:26:37 UTC |
| 74b7408 | Charles-François Natali | 29 August 2013, 17:02:23 UTC | Issue #18643: Fix some test_socket failures due to large default socket buffer sizes. | 29 August 2013, 17:02:23 UTC |
| 5fd2642 | Charles-François Natali | 29 August 2013, 17:01:40 UTC | Issue #18643: Fix some test_socket failures due to large default socket buffer sizes. | 29 August 2013, 17:01:40 UTC |
| e0d25ce | Richard Oudkerk | 29 August 2013, 13:37:47 UTC | Issue #8713: Print dangling processes/threads, if any. | 29 August 2013, 13:37:47 UTC |
| 83d7dea | Richard Oudkerk | 29 August 2013, 11:51:11 UTC | Issue #8713: Cleanup before saving process._dangling. | 29 August 2013, 11:51:11 UTC |
| 968d055 | Richard Oudkerk | 29 August 2013, 11:41:19 UTC | Issue #8713: Test should not print message about start method. | 29 August 2013, 11:41:19 UTC |
| de2800f | Serhiy Storchaka | 29 August 2013, 09:37:28 UTC | Issue #17974: Switch unittest from using getopt to using argparse. | 29 August 2013, 09:37:28 UTC |
| 64f7c4e | Serhiy Storchaka | 29 August 2013, 09:26:23 UTC | Issue #16799: Switched from getopt to argparse style in regrtest's argument parsing. Added more tests for regrtest's argument parsing. | 29 August 2013, 09:26:23 UTC |
| 48e6a8c | Serhiy Storchaka | 29 August 2013, 08:39:48 UTC | Issue #18743: Fix references to non-existant "StringIO" module in docstrings and comments. | 29 August 2013, 08:39:48 UTC |
| 50254c5 | Serhiy Storchaka | 29 August 2013, 08:35:43 UTC | Issue #18743: Fix references to non-existant "StringIO" module in docstrings and comments. | 29 August 2013, 08:35:43 UTC |
| a8c2a8a | Serhiy Storchaka | 29 August 2013, 07:29:30 UTC | Issue #18760: Improved cross-references in the xml package. | 29 August 2013, 07:29:30 UTC |
| 15e6590 | Serhiy Storchaka | 29 August 2013, 07:28:44 UTC | Issue #18760: Improved cross-references in the xml package. | 29 August 2013, 07:28:44 UTC |
| afe8909 | Raymond Hettinger | 29 August 2013, 03:59:31 UTC | Tighten-up the lookkey() logic and beautify the code a bit. Use less code by moving many of the steps from the initial lookup into the main search loop. Beautify the code but keep the overall logic unchanged. | 29 August 2013, 03:59:31 UTC |
| 7c1017b | Andrew Svetlov | 28 August 2013, 22:24:39 UTC | Fix tests for #11798 | 28 August 2013, 22:24:39 UTC |
| eb97368 | Andrew Svetlov | 28 August 2013, 18:28:38 UTC | Issue #11798: TestSuite now drops references to own tests after execution. | 28 August 2013, 18:28:38 UTC |
| 6a53af8 | Richard Oudkerk | 28 August 2013, 12:50:19 UTC | Issue #18786: Don't reinstall old SIGUSR1 handler prematurely. | 28 August 2013, 12:50:19 UTC |
| d98af98 | Victor Stinner | 28 August 2013, 10:34:16 UTC | (Merge 3.3) Fix @requires_freebsd_version and @requires_linux_version decorators of test.support, run the test if the platform matchs! | 28 August 2013, 10:34:16 UTC |
| d39dca9 | Victor Stinner | 28 August 2013, 10:28:18 UTC | Fix test_socket.test_SOCK_CLOEXEC(), the test was wrong | 28 August 2013, 10:28:18 UTC |
| 3b44a40 | Victor Stinner | 28 August 2013, 10:26:28 UTC | Fix @requires_freebsd_version and @requires_linux_version decorators of test.support, run the test if the platform matchs! | 28 August 2013, 10:26:28 UTC |
| bff989e | Victor Stinner | 28 August 2013, 10:25:40 UTC | test_posix.test_pipe2() now checks that the O_NONBLOCK flag is set Use also os.get_inheritable() instead of fcntl() to check the inheritable flag (FD_CLOEXEC). | 28 August 2013, 10:25:40 UTC |
| d72fe89 | Victor Stinner | 28 August 2013, 10:22:39 UTC | select.epoll.fromfd(fd) must be not change the inheritable flag of the file descriptor | 28 August 2013, 10:22:39 UTC |
| 67973c0 | Victor Stinner | 28 August 2013, 10:21:47 UTC | Issue #18865: remove unused import from multiprocessing.util.spawnv_passfds() | 28 August 2013, 10:21:47 UTC |
| 0d097b6 | Richard Oudkerk | 28 August 2013, 10:25:34 UTC | Issue #18865: PEP 446 makes multiprocessing.util.pipe() unnecessary. | 28 August 2013, 10:25:34 UTC |
| 1fa174a | Victor Stinner | 28 August 2013, 00:06:21 UTC | Get rid of signed/unsigned comparaison in _sre.c Fix compilation warnings on Windows (Visual C++) like: "_sre.c(3121): warning C4018: '>' : signed/unsigned mismatch". _validate_outer() ensures that groups >= 0, so _validate_inner() can cast groups to size_t. | 28 August 2013, 00:06:21 UTC |
| 36a5a06 | Victor Stinner | 27 August 2013, 23:53:39 UTC | _datetimemodule.c: fix the compilation warning "conversion from 'double' to 'long', possible loss of data" in delta_new(), use an explicit cast from double to long | 27 August 2013, 23:53:39 UTC |
| b9981ba | Victor Stinner | 27 August 2013, 23:51:06 UTC | fix a compilation warning in posix_openpty() on "PPC64 AIX 3.x" buildbot | 27 August 2013, 23:51:06 UTC |
| 55a1220 | Victor Stinner | 27 August 2013, 23:47:46 UTC | pythonrun.c: use MAXPATHLEN instead of PATH_MAX PATH_MAX is not available on "MIPS IRIX 6.5.30 [SB] 3.x" buildbot | 27 August 2013, 23:47:46 UTC |
| 340f712 | Victor Stinner | 27 August 2013, 23:45:39 UTC | (Merge 3.3) Fix compilation of the _sqlite module if threads are disabled | 27 August 2013, 23:45:39 UTC |
| b84fc0f | Victor Stinner | 27 August 2013, 23:44:42 UTC | Fix compilation of the _sqlite module if threads are disabled | 27 August 2013, 23:44:42 UTC |
| daf4555 | Victor Stinner | 27 August 2013, 22:53:59 UTC | Issue #18571: Implementation of the PEP 446: file descriptors and file handles are now created non-inheritable; add functions os.get/set_inheritable(), os.get/set_handle_inheritable() and socket.socket.get/set_inheritable(). | 27 August 2013, 22:53:59 UTC |
| 46e1ce2 | Serhiy Storchaka | 27 August 2013, 17:17:03 UTC | Issue #18783: Removed existing mentions of Python long type in docstrings, error messages and comments. | 27 August 2013, 17:17:03 UTC |
| 9594942 | Serhiy Storchaka | 27 August 2013, 16:40:23 UTC | Issue #18783: Removed existing mentions of Python long type in docstrings, error messages and comments. | 27 August 2013, 16:40:23 UTC |
| 14e461d | Victor Stinner | 26 August 2013, 20:28:21 UTC | Close #11619: The parser and the import machinery do not encode Unicode filenames anymore on Windows. | 26 August 2013, 20:28:21 UTC |
| 33824f6 | Victor Stinner | 26 August 2013, 12:05:19 UTC | Restore changeset 5bd9db528aed (issue #18408) "Issue #18408: PyObject_Str(), PyObject_Repr() and type_call() now fail with an assertion error if they are called with an exception set (PyErr_Occurred()). As PyEval_EvalFrameEx(), they may clear the current exception and so the caller looses its exception." | 26 August 2013, 12:05:19 UTC |
| c82bfd8 | Victor Stinner | 26 August 2013, 12:04:10 UTC | Issue #18664, #18408: Rewrite PyErr_WriteUnraisable() to handle errors * Catch PyFile_WriteString() and PyFile_WriteObject() errors * Clear the current exception on _PyObject_GetAttrId() failure * Use PyUnicode_CompareWithASCIIString() and PyFile_WriteObject() instead of _PyUnicode_AsString() and strcmp() to avoid Unicode encoding error. stderr has a more tolerant error handler than utf-8/strict. | 26 August 2013, 12:04:10 UTC |
| e513210 | Victor Stinner | 26 August 2013, 11:49:06 UTC | Issue #18408: _PyObject_Dump() now saves/restores the current exception So it can be called even if an exception was raised | 26 August 2013, 11:49:06 UTC |
| 0e1e043 | Ezio Melotti | 26 August 2013, 11:01:29 UTC | #18839: merge with 3.3. | 26 August 2013, 11:01:29 UTC |
| 4af4d27 | Ezio Melotti | 26 August 2013, 11:00:39 UTC | #18839: document that sys.exit() will not accept a non-integer numeric value as exit status. | 26 August 2013, 11:00:39 UTC |
| 8d26a56 | Eli Bendersky | 26 August 2013, 02:06:16 UTC | Clarify pyexpat documentation in StartElementHandler | 26 August 2013, 02:06:16 UTC |
| ca4d97e | Eli Bendersky | 26 August 2013, 02:05:55 UTC | Clarify pyexpat documentation in StartElementHandler | 26 August 2013, 02:05:55 UTC |
| 6206a7e | Eli Bendersky | 26 August 2013, 01:58:18 UTC | Remove the obsolete XMLParser._start/_start_list duality. XMLParser configures expat to report attributes in a list (ordered_attributes), so only _start_list is needed. Rename it to _start and kill _start. | 26 August 2013, 01:58:18 UTC |
| d640fe2 | Ezio Melotti | 25 August 2013, 22:33:30 UTC | #18803: merge with 3.3. | 25 August 2013, 22:33:30 UTC |
| 7c4a7e6 | Ezio Melotti | 25 August 2013, 22:32:56 UTC | #18803: fix more typos. Patch by Févry Thibault. | 25 August 2013, 22:32:56 UTC |
| c020e7f | Eli Bendersky | 25 August 2013, 22:30:39 UTC | Merge doc fix from 3.3 | 25 August 2013, 22:30:39 UTC |
| bf8ab77 | Eli Bendersky | 25 August 2013, 22:27:36 UTC | Update XMLParser.close documentation and fix formatting. Using ``method`` markup because the method is on a callback object, not an explicitly documented method. :meth: markup creates links within the current class which is incorrect. In addition, indent the paragraph correctly. | 25 August 2013, 22:27:36 UTC |
| 74eba65 | Ezio Melotti | 25 August 2013, 22:05:57 UTC | Fix markup in elementtree docs. | 25 August 2013, 22:05:57 UTC |
| ec42975 | Ezio Melotti | 25 August 2013, 21:17:21 UTC | #18833: merge with 3.3. | 25 August 2013, 21:17:21 UTC |
| 4f1353a | Ezio Melotti | 25 August 2013, 20:56:43 UTC | #18833: add a test for test_telnetlib. Patch by Alex Volkov. | 25 August 2013, 20:56:43 UTC |
| 7eaf3f7 | Antoine Pitrou | 25 August 2013, 17:48:18 UTC | Issue #18808: Non-daemon threads are now automatically joined when a sub-interpreter is shutdown (it would previously dump a fatal error). | 25 August 2013, 17:48:18 UTC |
| 0bb766b | Charles-François Natali | 25 August 2013, 16:29:16 UTC | Merge. | 25 August 2013, 16:29:16 UTC |
| a439b32 | Charles-François Natali | 25 August 2013, 16:28:44 UTC | Merge. | 25 August 2013, 16:28:44 UTC |
| c20a7b9 | Charles-François Natali | 25 August 2013, 16:27:59 UTC | Merge. | 25 August 2013, 16:27:59 UTC |
| d317f4c | Charles-François Natali | 25 August 2013, 16:25:38 UTC | Issue #18763: subprocess: The file descriptors are now closed after calling the preexec_fn callback, which may open file descriptors. | 25 August 2013, 16:25:38 UTC |
| 249cdc3 | Charles-François Natali | 25 August 2013, 16:24:45 UTC | Issue #18763: subprocess: The file descriptors are now closed after calling the preexec_fn callback, which may open file descriptors. | 25 August 2013, 16:24:45 UTC |
| 58b3ebf | Serhiy Storchaka | 25 August 2013, 16:16:01 UTC | Issue #18817: Fix a resource warning in Lib/aifc.py demo. Patch by Vajrasky Kok. | 25 August 2013, 16:16:01 UTC |
| b33baf1 | Serhiy Storchaka | 25 August 2013, 16:12:56 UTC | Issue #18817: Fix a resource warning in Lib/aifc.py demo. | 25 August 2013, 16:12:56 UTC |
| 0cb8e51 | R David Murray | 25 August 2013, 15:09:45 UTC | Merge #16611: BaseCookie now parses 'secure' and 'httponly' flags. | 25 August 2013, 15:09:45 UTC |
| cd0f74b | R David Murray | 25 August 2013, 15:09:02 UTC | #16611: BaseCookie now parses 'secure' and 'httponly' flags. Previously it generated them if they were given a value, but completely ignored them if they were present in the string passed in to be parsed. Now if the flag appears on a cookie, the corresponding Morsel key will reference a True value. Other pre-existing behavior is retained in this maintenance patch: if the source contains something like 'secure=foo', morsel['secure'] will return 'foo'. Since such a value doesn't round trip and never did (and would be a surprising occurrence) a subsequent non-bug-fix patch may change this behavior. Inspired by a patch from Julien Phalip, who reviewed this one. | 25 August 2013, 15:09:02 UTC |
