4f3abb0 | Victor Stinner | 07 October 2010, 23:29:18 UTC | copy_absolute() keeps the relative path on _Py_wgetcwd() failure .. instead of raising a fatal error. Even if the current directory was deleted, use relative paths may still work (eg. run Python with "../python"). | 07 October 2010, 23:29:18 UTC |
6672d0c | Victor Stinner | 07 October 2010, 22:53:43 UTC | fileutils.c: document which encodings are used | 07 October 2010, 22:53:43 UTC |
41c8b37 | Victor Stinner | 07 October 2010, 22:42:51 UTC | copy_absolute() raises a fatal error on _Py_wgetcwd() failure | 07 October 2010, 22:42:51 UTC |
015f4d8 | Victor Stinner | 07 October 2010, 22:29:53 UTC | _Py_wrealpath() requires the size of the output buffer | 07 October 2010, 22:29:53 UTC |
a4a7595 | Victor Stinner | 07 October 2010, 22:23:10 UTC | _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows On Windows, Py_UNICODE is wchar_t, so we can avoid the expensive Py_UNICODE* => wchar_t* conversion. | 07 October 2010, 22:23:10 UTC |
b306d75 | Victor Stinner | 07 October 2010, 22:09:40 UTC | Fix fileutils for Windows * Don't define _Py_wstat() on Windows, Windows has its own _wstat() function with a different API (the stat buffer has another type) * Include windows.h | 07 October 2010, 22:09:40 UTC |
e7c8083 | Victor Stinner | 07 October 2010, 21:55:44 UTC | Ooops, fileutils.c contains twice the same code I suppose that I reapplied my local patch creating Python/fileutils.c whereas the file already existed. | 07 October 2010, 21:55:44 UTC |
4e31443 | Victor Stinner | 07 October 2010, 21:45:39 UTC | Create fileutils.c/.h * _Py_fopen() and _Py_stat() come from Python/import.c * (_Py)_wrealpath() comes from Python/sysmodule.c * _Py_char2wchar(), _Py_wchar2char() and _Py_wfopen() come from Modules/main.c * (_Py)_wstat(), (_Py)_wgetcwd(), _Py_wreadlink() come from Modules/getpath.c | 07 October 2010, 21:45:39 UTC |
7ae7c87 | Victor Stinner | 07 October 2010, 11:06:49 UTC | _wrealpath() and _Py_wreadlink() support surrogates (PEP 383) Use _Py_wchar2char() to support surrogate characters in the input path. | 07 October 2010, 11:06:49 UTC |
afa88b5 | Brian Curtin | 07 October 2010, 01:12:19 UTC | Fix #8584. Fix the skipping of a number of tests due to incorrect importing. This undoes a previous attempt to fix some of the skips. Value and copy come from sharedctypes rather than plain ctypes, and the test skipping is then done at the class level rather than test-by-test. There are zero skipped tests after the fix. | 07 October 2010, 01:12:19 UTC |
beb4135 | Victor Stinner | 07 October 2010, 01:02:42 UTC | PyUnicode_AsWideCharString() takes a PyObject*, not a PyUnicodeObject* All unicode functions uses PyObject* except PyUnicode_AsWideChar(). Fix the prototype for the new function PyUnicode_AsWideCharString(). | 07 October 2010, 01:02:42 UTC |
ef12810 | Victor Stinner | 07 October 2010, 01:00:52 UTC | time: fix gcc warning * Create format_arg variable to use the right types * Strip trailing spaces | 07 October 2010, 01:00:52 UTC |
4726e40 | Victor Stinner | 06 October 2010, 23:24:57 UTC | Rewrite RunMainFromImporter() * fix argv0 reference counter if PyList_SetItem() fails * don't use complex if conditions, but a simple indentation and "goto error" * simplify error handling (remove Py_XDECREF(importer) from the error label) * don't set sys_path to NULL (it's useless, sys_path is a borrowed reference and sys_path is not a static variable) * try to write only one instruction per line for better readability | 06 October 2010, 23:24:57 UTC |
c08ec9f | Victor Stinner | 06 October 2010, 22:44:06 UTC | Create a subfunction for PySys_SetArgvEx() Create sys_update_path() static function. Do nothing if argc==0. | 06 October 2010, 22:44:06 UTC |
7980eaa | Antoine Pitrou | 06 October 2010, 21:21:18 UTC | Issue #9759: GzipFile now raises ValueError when an operation is attempted after the file is closed. Patch by Jeffrey Finkelstein. | 06 October 2010, 21:21:18 UTC |
cd889af | Antoine Pitrou | 06 October 2010, 21:13:56 UTC | More new things in what's new | 06 October 2010, 21:13:56 UTC |
fa9c38d | Antoine Pitrou | 06 October 2010, 18:38:22 UTC | FakeSocket is (supposed to be) gone in 3.x | 06 October 2010, 18:38:22 UTC |
1e2ae4f | Brian Quinlan | 06 October 2010, 13:05:45 UTC | Fixes 9903: test_concurrent_futures writes on stderr | 06 October 2010, 13:05:45 UTC |
5ad8ed5 | Georg Brandl | 06 October 2010, 12:29:49 UTC | Update to Sphin 1.0.4. | 06 October 2010, 12:29:49 UTC |
19b3e00 | Georg Brandl | 06 October 2010, 10:35:24 UTC | Fix the suspicious builder not to write CRLF, update suspicious file and add instructions to Makefile. | 06 October 2010, 10:35:24 UTC |
682d7e0 | Georg Brandl | 06 October 2010, 10:26:05 UTC | Fix errors found by "make suspicious". | 06 October 2010, 10:26:05 UTC |
fa4f7f9 | Georg Brandl | 06 October 2010, 10:14:08 UTC | Fix spelling and markup. | 06 October 2010, 10:14:08 UTC |
60203b4 | Georg Brandl | 06 October 2010, 10:11:56 UTC | Migrate to Sphinx 1.0 C language constructs. | 06 October 2010, 10:11:56 UTC |
64a41ed | Georg Brandl | 06 October 2010, 08:52:48 UTC | Copyedit of os.symlink() docs. | 06 October 2010, 08:52:48 UTC |
a102ae3 | Georg Brandl | 06 October 2010, 05:08:32 UTC | #10034: import readline in longer example. | 06 October 2010, 05:08:32 UTC |
b441554 | Amaury Forgeot d'Arc | 05 October 2010, 23:14:47 UTC | #2982 More test coverage for the pyexpat parser. | 05 October 2010, 23:14:47 UTC |
90774dd | Amaury Forgeot d'Arc | 05 October 2010, 22:15:37 UTC | #9060 Let platforms without dup2() compile the replacement fonction without error. | 05 October 2010, 22:15:37 UTC |
c08d907 | Senthil Kumaran | 05 October 2010, 18:46:56 UTC | fix a typo. | 05 October 2010, 18:46:56 UTC |
242690f | Senthil Kumaran | 05 October 2010, 18:35:41 UTC | Remove the references of FTPS from the urllib Documentation. There is no support for it yet. | 05 October 2010, 18:35:41 UTC |
2f92e78 | Antoine Pitrou | 05 October 2010, 11:24:49 UTC | Fix test_tk under OS X with Tk 8.4. Patch by Ned Deily. This should fix some buildbot failures. | 05 October 2010, 11:24:49 UTC |
eeed955 | Georg Brandl | 05 October 2010, 07:16:01 UTC | Indent version tags properly. | 05 October 2010, 07:16:01 UTC |
ef777bd | Éric Araujo | 05 October 2010, 00:04:20 UTC | Fix news entry formatting nits | 05 October 2010, 00:04:20 UTC |
6108bf5 | Éric Araujo | 04 October 2010, 23:52:37 UTC | Fix interaction of custom translation classes and caching (#9042) | 04 October 2010, 23:52:37 UTC |
7017616 | Éric Araujo | 04 October 2010, 22:32:36 UTC | Recode developers.txt to UTF-8, as discussed on python-dev. Fix Łukasz’ name and also remove trailing whitespace. See thread at http://mail.python.org/pipermail/python-dev/2010-September/103584.html | 04 October 2010, 22:32:36 UTC |
7b998e9 | Antoine Pitrou | 04 October 2010, 21:55:14 UTC | GzipFile.peek improvements, suggested by Nir Aides. | 04 October 2010, 21:55:14 UTC |
977c707 | Giampaolo Rodolà | 04 October 2010, 21:08:36 UTC | Fix issue 6706: adds new handle_accepted() method to asyncore.dispatcher | 04 October 2010, 21:08:36 UTC |
4c94c53 | Martin v. Löwis | 04 October 2010, 15:48:14 UTC | Add Tal Einat. | 04 October 2010, 15:48:14 UTC |
331b800 | Lars Gustäbel | 04 October 2010, 15:18:47 UTC | Issue #9065: no longer use "root" as the default for the uname and gname field. If tarfile creates a new archive and adds a file with a uid/gid that doesn't have a corresponding name on the system (e.g. because the user/group account was deleted) it uses the empty string in the uname/gname field now instead of "root". Using "root" as the default was a bad idea because on extraction the uname/gname fields are supposed to override the uid/gid fields. So, all archive members with nameless uids/gids belonged to the root user after extraction. | 04 October 2010, 15:18:47 UTC |
f580ade | Senthil Kumaran | 04 October 2010, 05:42:58 UTC | Remove the debugging print line from the test. | 04 October 2010, 05:42:58 UTC |
e0941c5 | Gerhard Häring | 03 October 2010, 21:47:06 UTC | Fixes #10020 | 03 October 2010, 21:47:06 UTC |
58d5dbf | Senthil Kumaran | 03 October 2010, 18:22:42 UTC | Fix Issue10012 - httplib headers, which are (sometimes mistakenly) int are explicitly cast to str (bytes - in py3k). | 03 October 2010, 18:22:42 UTC |
4271372 | Senthil Kumaran | 03 October 2010, 17:55:45 UTC | Fix Issue9272 - Change CGIHTTPServer to give the child program a copy of os.environ | 03 October 2010, 17:55:45 UTC |
388f956 | Tarek Ziadé | 03 October 2010, 14:45:06 UTC | typo in Arfrever name | 03 October 2010, 14:45:06 UTC |
4bcceef | Tarek Ziadé | 03 October 2010, 14:18:09 UTC | Fixed #8980: distutils.command.check was failing w/ docutils installed | 03 October 2010, 14:18:09 UTC |
84060b8 | Benjamin Peterson | 03 October 2010, 02:13:39 UTC | typo | 03 October 2010, 02:13:39 UTC |
970d188 | Benjamin Peterson | 02 October 2010, 17:55:47 UTC | add a test and a note about metaclasses now being abcs | 02 October 2010, 17:55:47 UTC |
5397e86 | R. David Murray | 02 October 2010, 15:58:26 UTC | #1050268: make parseaddr 'quote' the contents of quoted strings in addresses. Also made the doc string for email._parseaddr's 'quote' function more accurate; I'd love to make the function match the old docstring instead, but other code uses it according the existing semantics. | 02 October 2010, 15:58:26 UTC |
8f7bcb3 | R. David Murray | 02 October 2010, 13:29:13 UTC | Make the printing of sqlite version in verbose mode work with regrtest -w. | 02 October 2010, 13:29:13 UTC |
aef57e3 | Victor Stinner | 02 October 2010, 11:50:49 UTC | Update changelog for r85174 | 02 October 2010, 11:50:49 UTC |
46c7b3b | Victor Stinner | 02 October 2010, 11:49:31 UTC | Issue #8670: Rename testcapi unicode test methods * test_aswidechar() => unicode_aswidechar() * test_aswidecharstring() => unicode_aswidecharstring() | 02 October 2010, 11:49:31 UTC |
ea3f305 | Victor Stinner | 02 October 2010, 11:46:20 UTC | Oops, revert unwanted _testcapi changes of r85174 | 02 October 2010, 11:46:20 UTC |
749261e | Victor Stinner | 02 October 2010, 11:25:35 UTC | Issue #8670: ctypes.c_wchar supports non-BMP characters with 32 bits wchar_t | 02 October 2010, 11:25:35 UTC |
5593d8a | Victor Stinner | 02 October 2010, 11:11:27 UTC | Issue #8670: PyUnicode_AsWideChar() and PyUnicode_AsWideCharString() replace UTF-16 surrogate pairs by single non-BMP characters for 16 bits Py_UNICODE and 32 bits wchar_t (eg. Linux in narrow build). | 02 October 2010, 11:11:27 UTC |
1c24bd0 | Victor Stinner | 02 October 2010, 11:03:13 UTC | Issue #8870: PyUnicode_AsWideCharString() doesn't count the trailing nul character And write unit tests for PyUnicode_AsWideChar() and PyUnicode_AsWideCharString(). | 02 October 2010, 11:03:13 UTC |
5a2da3b | Senthil Kumaran | 02 October 2010, 10:33:13 UTC | Use proper variable name 'data' instead of 'str' in the send method. | 02 October 2010, 10:33:13 UTC |
7cafd26 | Senthil Kumaran | 02 October 2010, 03:16:04 UTC | Fix - issue10010 .. index:: position in the docs. | 02 October 2010, 03:16:04 UTC |
aec5fd1 | Benjamin Peterson | 02 October 2010, 00:03:31 UTC | type.__abstractmethods__ should raise an AttributeError #10006 | 02 October 2010, 00:03:31 UTC |
ea8676b | Amaury Forgeot d'Arc | 01 October 2010, 23:42:24 UTC | Lower the tone of the warning about SSL certificate validation. | 01 October 2010, 23:42:24 UTC |
70a9993 | R. David Murray | 01 October 2010, 20:38:33 UTC | Fix docstring typo. | 01 October 2010, 20:38:33 UTC |
31e984c | R. David Murray | 01 October 2010, 15:40:20 UTC | #10004: in Q encoded word ignore '=xx' when xx is not valid hex. Bug report and fix by Thomas Guettler. | 01 October 2010, 15:40:20 UTC |
9e88b5a | Brian Curtin | 01 October 2010, 14:49:24 UTC | Fix #10003. Add SIGBREAK to the set of valid signals on Windows. This fixes a regression noticed by bzr, introduced by issue #9324. | 01 October 2010, 14:49:24 UTC |
38e2996 | Alexander Belopolsky | 01 October 2010, 14:18:49 UTC | Issue #6608: time.asctime is now checking struct tm fields its input before passing it to the system asctime. Patch by MunSic Jeong. | 01 October 2010, 14:18:49 UTC |
0b0ebb4 | Hirokazu Yamamoto | 01 October 2010, 10:40:49 UTC | Cosmetic fix for PCBuild/readme.txt. | 01 October 2010, 10:40:49 UTC |
d0ffa4c | Georg Brandl | 01 October 2010, 05:38:10 UTC | Fix weird markup that caused latex to fail. | 01 October 2010, 05:38:10 UTC |
d2c310f | R. David Murray | 01 October 2010, 02:08:02 UTC | Fix a couple spelling errors in comments and delete redundant __len__ def. | 01 October 2010, 02:08:02 UTC |
0f476d4 | Senthil Kumaran | 30 September 2010, 06:09:18 UTC | Issue1491 - BaseHTTPServer incorrectly implements response code 100 | 30 September 2010, 06:09:18 UTC |
748cace | Brian Curtin | 29 September 2010, 19:09:33 UTC | Remove an unreferenced variable. len is no longer needed. | 29 September 2010, 19:09:33 UTC |
c603717 | Mark Dickinson | 29 September 2010, 19:06:36 UTC | Issue #9599: Further accuracy tweaks to loghelper. For an integer n that's small enough to be converted to a float without overflow, log(n) is now computed as log(float(n)), and similarly for log10. | 29 September 2010, 19:06:36 UTC |
0c0714f | Mark Dickinson | 29 September 2010, 18:41:54 UTC | Add testcases from bug report #9980. | 29 September 2010, 18:41:54 UTC |
71e91a3 | Victor Stinner | 29 September 2010, 17:55:12 UTC | Fix PyUnicode_AsWideCharString(): set *size if size is not NULL | 29 September 2010, 17:55:12 UTC |
15d597a | Victor Stinner | 29 September 2010, 16:59:46 UTC | test_imp: getfilesystemencoding() cannot be None anymore And the codec name is normalized. | 29 September 2010, 16:59:46 UTC |
16004ac | Victor Stinner | 29 September 2010, 16:59:18 UTC | Use os.fsencode() to support surrogates | 29 September 2010, 16:59:18 UTC |
c39211f | Victor Stinner | 29 September 2010, 16:35:47 UTC | Issue #9630: Redecode filenames when setting the filesystem encoding Redecode the filenames of: - all modules: __file__ and __path__ attributes - all code objects: co_filename attribute - sys.path - sys.meta_path - sys.executable - sys.path_importer_cache (keys) Keep weak references to all code objects until initfsencoding() is called, to be able to redecode co_filename attribute of all code objects. | 29 September 2010, 16:35:47 UTC |
a5785b1 | Antoine Pitrou | 29 September 2010, 16:19:50 UTC | Fix NNTP when there's a ".netrc" file | 29 September 2010, 16:19:50 UTC |
2620d81 | Antoine Pitrou | 29 September 2010, 16:08:29 UTC | Disable the usenetrc option when testing nntplib | 29 September 2010, 16:08:29 UTC |
69ab951 | Antoine Pitrou | 29 September 2010, 15:03:40 UTC | Issue #9360: Cleanup and improvements to the nntplib module. The API now conforms to the philosophy of bytes and unicode separation in Python 3. A test suite has also been added. | 29 September 2010, 15:03:40 UTC |
926f0da | Brian Curtin | 29 September 2010, 14:51:42 UTC | Fix #9978. WINFUNCTYPE is from ctypes, not from ctypes.wintypes. r85073 changed the importing in wintypes to not use *, so the previous usage here became even more incorrect. | 29 September 2010, 14:51:42 UTC |
79ecd76 | Antoine Pitrou | 29 September 2010, 11:24:21 UTC | Issue #9983: warn that urllib and httplib don't perform SSL certificate validation. | 29 September 2010, 11:24:21 UTC |
c3ed2e7 | Antoine Pitrou | 29 September 2010, 10:49:46 UTC | Issue #9962: GzipFile now has the peek() method. | 29 September 2010, 10:49:46 UTC |
4c2e4fa | Victor Stinner | 29 September 2010, 10:37:16 UTC | Issue #9979: Use PyUnicode_AsWideCharString() in _ctypes module * Convert unicode to wide character string before creating the PyCapsule object * Catch integer overflow * Avoid useless memset() * Prepare the support of surrogates | 29 September 2010, 10:37:16 UTC |
b290478 | Victor Stinner | 29 September 2010, 10:34:19 UTC | Issue #9979: Use PyUnicode_AsWideCharString() in time.strftime() Allocate memory with PyMem_Alloc() instead of the PyBytes API. Prepare the surrogates support. | 29 September 2010, 10:34:19 UTC |
449057f | Victor Stinner | 29 September 2010, 10:30:43 UTC | Issue #9979: Use PyUnicode_AsWideCharString() for _locale.strcoll() It simplifies the code and prepare the surrogates support. | 29 September 2010, 10:30:43 UTC |
255dfdb | Victor Stinner | 29 September 2010, 10:28:51 UTC | Issue #9979: Use PyUnicode_AsWideCharString() in import.c Don't truncate path if it is too long anymore, and allocate fewer memory (but allocate it on the heap, not on the stack). | 29 September 2010, 10:28:51 UTC |
137c34c | Victor Stinner | 29 September 2010, 10:25:54 UTC | Issue #9979: Create function PyUnicode_AsWideCharString(). | 29 September 2010, 10:25:54 UTC |
41a64a5 | Victor Stinner | 29 September 2010, 01:30:45 UTC | linecache.updatecache(): don't the lines into the cache on IOError Use the same behaviour than Python 2.7. | 29 September 2010, 01:30:45 UTC |
78f43a7 | Victor Stinner | 29 September 2010, 01:24:59 UTC | linecache.updatecache() returns an empty list on error lines variable was not set on IOError | 29 September 2010, 01:24:59 UTC |
d68d4e8 | R. David Murray | 29 September 2010, 01:22:20 UTC | Have test_sqlite print version info when run in verbose mode. | 29 September 2010, 01:22:20 UTC |
11cabcf | R. David Murray | 29 September 2010, 01:08:05 UTC | #7110: have regrtest print test failures and tracebacks to stderr not stdout. Patch by Sandro Tosi. | 29 September 2010, 01:08:05 UTC |
c663b58 | Antoine Pitrou | 28 September 2010, 23:59:51 UTC | Give a dedicated page to memoryview objects, so that they can be part of the concrete objects layer, while the buffer protocol is part of the abstract objects layer. | 28 September 2010, 23:59:51 UTC |
f7ba2fa | Antoine Pitrou | 28 September 2010, 23:39:41 UTC | Buffers are not sequence objects (!). Put them in the abstract objects layers instead. | 28 September 2010, 23:39:41 UTC |
99a00a4 | Antoine Pitrou | 28 September 2010, 23:04:04 UTC | Various improvements to the docs of the buffer API | 28 September 2010, 23:04:04 UTC |
b78b489 | R. David Murray | 28 September 2010, 22:25:18 UTC | #9628: fix runtests.sh -x option so more than one test can be excluded. | 28 September 2010, 22:25:18 UTC |
c8f0196 | Antoine Pitrou | 28 September 2010, 22:03:27 UTC | Fix compilation under Windows | 28 September 2010, 22:03:27 UTC |
a0b6873 | Antoine Pitrou | 28 September 2010, 21:52:30 UTC | Discourage use of the old buffer API funcs | 28 September 2010, 21:52:30 UTC |
3e1fd27 | Antoine Pitrou | 28 September 2010, 21:23:11 UTC | Issue #9090: When a socket with a timeout fails with EWOULDBLOCK or EAGAIN, retry the select() loop instead of bailing out. This is because select() can incorrectly report a socket as ready for reading (for example, if it received some data with an invalid checksum). | 28 September 2010, 21:23:11 UTC |
cc868d4 | Hirokazu Yamamoto | 28 September 2010, 21:08:38 UTC | Issue #3612: Added new types to ctypes.wintypes. (CHAR and pointers) | 28 September 2010, 21:08:38 UTC |
7405c20 | Hirokazu Yamamoto | 28 September 2010, 18:36:04 UTC | Updated PC/VC6 openssl build script. (for openssl-1.0.0a) | 28 September 2010, 18:36:04 UTC |
313dab4 | Hirokazu Yamamoto | 28 September 2010, 18:29:57 UTC | Now perl path with spaces can be used. | 28 September 2010, 18:29:57 UTC |
57b4163 | Hirokazu Yamamoto | 28 September 2010, 17:55:53 UTC | Reverted to r70917. | 28 September 2010, 17:55:53 UTC |
826903e | Antoine Pitrou | 28 September 2010, 15:29:16 UTC | Issue #9970: improve C API documentation for memoryview objects | 28 September 2010, 15:29:16 UTC |