summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* For now, remove accept4() code (issue #10115)Antoine Pitrou2010-10-221-9/+0
|
* Fix test_ssl on Ubuntu buildbot with patched OpenSSLAntoine Pitrou2010-10-221-0/+1
|
* Issue #5639: Add a *server_hostname* argument to `SSLContext.wrap_socket`Antoine Pitrou2010-10-229-14/+137
| | | | | | in order to support the TLS SNI extension. `HTTPSConnection` and `urlopen()` also use this argument, so that HTTPS virtual hosts are now supported.
* Only hack cmd.library_dirs when running under Py_ENABLE_SHARED. Tested bothBarry Warsaw2010-10-221-1/+2
| | | | | with and without --enable-shared on Ubuntu 10.10. Hopefully this finally solves bug 10126. Will check 3.1 next.
* Refactor interesting use of try-finally.Georg Brandl2010-10-221-10/+9
|
* Make top_level attribute a set instead of a dict with None values.Georg Brandl2010-10-221-5/+5
|
* #10166: rewrite self-recursion to iteration in pstats.Stats.add(). Also add ↵Georg Brandl2010-10-224-27/+44
| | | | a unittest and a stats test file.
* Apply fix from r85784 on py3k too.Éric Araujo2010-10-211-2/+6
| | | | | | | Fixes bug #10126 for Python 3.2 by using $RUNSHARED to find the directory to the shared library. test_distutils now passes when Python was built with --enable-shared (Barry didn’t have the error but I did).
* Fixed a typoAlexander Belopolsky2010-10-211-1/+1
|
* Issue #7061: Fixed section titleAlexander Belopolsky2010-10-211-1/+1
|
* Add .hgeol file for the Mercurial EOL extension.Georg Brandl2010-10-211-0/+36
|
* Fix r85774 editor fail. Had a dangling try and incorrect indent.Brian Curtin2010-10-211-2/+1
|
* Fix #10162: Add try/except around _winreg.OpenKey for keys that areBrian Curtin2010-10-212-8/+13
| | | | | | | | unreadable by all users, e.g., Flash, Silverlight, and Java keys were causing errors. We don't currently have a way to grant/deny permissions for a key via winreg so there are no tests for this.
* #4829: better error message for invalid file modeGeorg Brandl2010-10-211-1/+2
|
* Issue #10089: Add support for arbitrary -X options on the command-line.Antoine Pitrou2010-10-2110-14/+136
| | | | They can be retrieved through a new attribute `sys._xoptions`.
* #1203650: allow larger list of files in windows makefile for freeze.Georg Brandl2010-10-211-5/+7
|
* #3077: fix h2py substitution of character literals.Georg Brandl2010-10-211-1/+1
|
* Fix missing import.Georg Brandl2010-10-211-2/+1
|
* #9919: fix off-by-one error in lineno command in Misc/gdbinit; also add ↵Georg Brandl2010-10-211-3/+4
| | | | newline to its output.
* #9095, #8912, #8999: add support in patchcheck for Mercurial checkouts, C ↵Georg Brandl2010-10-213-33/+68
| | | | file reindenting, and docs whitespace fixing.
* #10159: sort completion matches before comparing to dir() result.Georg Brandl2010-10-211-2/+2
|
* Issue #4388: On Mac OS X, decode command line arguments from UTF-8, instead ofVictor Stinner2010-10-204-0/+160
| | | | | | | the locale encoding. If the LANG (and LC_ALL and LC_CTYPE) environment variable is not set, the locale encoding is ISO-8859-1, whereas most programs (including Python) expect UTF-8. Python already uses UTF-8 for the filesystem encoding and to encode command line arguments on this OS.
* Move non-ascii test from test_run_code() to a new function: test_non_ascii()Victor Stinner2010-10-201-7/+9
|
* Move test_undecodable_code() from test_sys to test_cmd_lineVictor Stinner2010-10-202-38/+38
|
* test_zipimport_support: use ascii() on bytes output to avoid BytesWarningVictor Stinner2010-10-201-2/+2
| | | | The test failed with python -bb.
* fix uninitialized struct member #10152Benjamin Peterson2010-10-201-0/+1
|
* logging: Improved StreamHandler documentation.Vinay Sajip2010-10-201-11/+13
|
* logging: Made StreamHandler terminator configurable.Vinay Sajip2010-10-202-3/+7
|
* Fix for issue #7473.Ronald Oussoren2010-10-202-1/+9
| | | | | | | Without this patch you'll get link errors in some extensions (in particular the _curses_panel) one when you try to build a 3-way universal framework when you previously installed a 2-way universal framework using the python.org installer.
* Don't lie about the supported architectures in the OSX installerRonald Oussoren2010-10-201-1/+1
| | | | | | | | Without this patch the i386/x86_64 installer for OSX 10.6 lies in the ReadMe file and the "Important Information" screen of the installer (that is, the installer claims it supports the i386 and ppc architectures insetead of the ones it really supports)
* logging: clarified Filter documentation.Vinay Sajip2010-10-201-6/+10
|
* Fix r85728: use "" to mean the system default locale, which should work on ↵Georg Brandl2010-10-201-7/+4
| | | | more systems.
* Correct signature of BytesGenerator in docs.R. David Murray2010-10-191-1/+1
|
* logging: Documented usage of callables as filters.Vinay Sajip2010-10-192-12/+16
|
* Issue #7061: Fixed some of the issues in turtle module documentationAlexander Belopolsky2010-10-191-11/+13
| | | | reported by Terry J. Reedy.
* Be consistent in the spelling of thread-safe(ty).Georg Brandl2010-10-195-7/+7
|
* logging: Allowed filters to be just callables.Vinay Sajip2010-10-193-1/+45
|
* logging: Added tests for _logRecordClass changes.Vinay Sajip2010-10-193-2/+51
|
* #10092: Properly reset locale in Locale*Calendar classes. The context ↵Georg Brandl2010-10-194-5/+30
| | | | manager was buggy because setlocale() returns the *new* locale, not the old. Also add a test for this.
* Fixed the docstring for calendar.isleap() function.Alexander Belopolsky2010-10-191-1/+1
| | | | Thanks Boštjan Mejak for the patch.
* logging: Added _logRecordClass, getLogRecordClass, setLogRecordClass to ↵Vinay Sajip2010-10-192-3/+26
| | | | increase flexibility of LogRecord creation.
* Fix Issue10140 - Tools/scripts/pathfix.py: add option to preserve timestampsSenthil Kumaran2010-10-191-3/+18
|
* initfsencoding(): get_codeset() failure is now a fatal errorVictor Stinner2010-10-191-13/+6
| | | | | Don't fallback to utf-8 anymore to avoid mojibake. I never got any error from his function.
* zipimport: remove arbitrary length limit from message formatsVictor Stinner2010-10-181-14/+9
| | | | | PyErr_Format() and PyUnicode_FromFormat() are able to allocate the right buffer size and to catch memory allocation failures.
* zipimport: document archive encoding; fix indentationVictor Stinner2010-10-181-3/+4
|
* Revert r85699 and r85701 (zipimport): fullname is a module name, not a pathVictor Stinner2010-10-181-42/+18
| | | | UTF-8 is just fine for module names.
* zipimport: fix "can't find module ..." error messageVictor Stinner2010-10-181-4/+16
| | | | | | I cannot use %U: fullname is a bytes object, not an unicode object. %A format cannot be used, it adds 'b' (bytes) prefix. So create cant_find_module() function to decode the filename and raise the error message.
* PyUnicode_FromFormatV(): Fix %A formatVictor Stinner2010-10-182-3/+7
| | | | It was not completly implemented. Add a test.
* zipimport: find_module(), is_package() and get_source() supports surrogatesVictor Stinner2010-10-181-16/+28
| | | | Use PyUnicode_FSConverter to support surrogates in the full name.
* zipimport: pass path size to make_filename()Victor Stinner2010-10-181-5/+5
| | | | Don't hardcode path size in make_filename().