]> granicus.if.org Git - python/log
python
14 years agoDocument that the 'strict' error handler is used to encode/decode filenames on
Victor Stinner [Sat, 9 Oct 2010 10:34:37 +0000 (10:34 +0000)]
Document that the 'strict' error handler is used to encode/decode filenames on
Windows

14 years agoIssue #9738: Document PyErr_SetString() and PyErr_SetFromErrnoWithFilename()
Victor Stinner [Sat, 9 Oct 2010 10:12:11 +0000 (10:12 +0000)]
Issue #9738: Document PyErr_SetString() and PyErr_SetFromErrnoWithFilename()
encodings

14 years agoremove author names from prose
Benjamin Peterson [Sat, 9 Oct 2010 01:58:37 +0000 (01:58 +0000)]
remove author names from prose

14 years agoIssue #1051: Add a script (Lib/test/make_ssl_certs.py) to generate the
Antoine Pitrou [Fri, 8 Oct 2010 23:06:24 +0000 (23:06 +0000)]
Issue #1051: Add a script (Lib/test/make_ssl_certs.py) to generate the
custom certificate and private key files used by SSL-related certs.

14 years agoWhat's new entry for email 5.1.
R. David Murray [Fri, 8 Oct 2010 21:37:39 +0000 (21:37 +0000)]
What's new entry for email 5.1.

14 years agoAdd some debug output in verbose mode
Antoine Pitrou [Fri, 8 Oct 2010 18:46:09 +0000 (18:46 +0000)]
Add some debug output in verbose mode

14 years agoBetter Python spawning primitives in test.script_helper, for
Antoine Pitrou [Fri, 8 Oct 2010 18:05:42 +0000 (18:05 +0000)]
Better Python spawning primitives in test.script_helper, for
easier writing of unit tests and better error reporting.

14 years agoUpdate what's new
Antoine Pitrou [Fri, 8 Oct 2010 16:46:17 +0000 (16:46 +0000)]
Update what's new

14 years ago#4661: add bytes parsing and generation to email (email version bump to 5.1.0)
R. David Murray [Fri, 8 Oct 2010 15:55:28 +0000 (15:55 +0000)]
#4661: add bytes parsing and generation to email (email version bump to 5.1.0)

The work on this is not 100% complete, but everything is present to
allow real-world testing of the code.  The only remaining major todo
item is to (hopefully!) enhance the handling of non-ASCII bytes in headers
converted to unicode by RFC2047 encoding them rather than replacing them with
'?'s.

14 years agoIssue #1589: Add ssl.match_hostname(), to help implement server identity
Antoine Pitrou [Fri, 8 Oct 2010 10:37:08 +0000 (10:37 +0000)]
Issue #1589: Add ssl.match_hostname(), to help implement server identity
verification for higher-level protocols.

14 years agoFollows Python/fileutils.c addition.
Hirokazu Yamamoto [Fri, 8 Oct 2010 08:46:30 +0000 (08:46 +0000)]
Follows Python/fileutils.c addition.

14 years agoIssue #9978: Wait until subprocess completes initialization. (Win32KillTests in test_os)
Hirokazu Yamamoto [Fri, 8 Oct 2010 08:38:15 +0000 (08:38 +0000)]
Issue #9978: Wait until subprocess completes initialization. (Win32KillTests in test_os)

14 years agoFix handling on negative numbers in ast.literal_eval().
Raymond Hettinger [Fri, 8 Oct 2010 00:47:45 +0000 (00:47 +0000)]
Fix handling on negative numbers in ast.literal_eval().

14 years agocopy_absolute() keeps the relative path on _Py_wgetcwd() failure
Victor Stinner [Thu, 7 Oct 2010 23:29:18 +0000 (23:29 +0000)]
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").

14 years agofileutils.c: document which encodings are used
Victor Stinner [Thu, 7 Oct 2010 22:53:43 +0000 (22:53 +0000)]
fileutils.c: document which encodings are used

14 years agocopy_absolute() raises a fatal error on _Py_wgetcwd() failure
Victor Stinner [Thu, 7 Oct 2010 22:42:51 +0000 (22:42 +0000)]
copy_absolute() raises a fatal error on _Py_wgetcwd() failure

14 years ago_Py_wrealpath() requires the size of the output buffer
Victor Stinner [Thu, 7 Oct 2010 22:29:53 +0000 (22:29 +0000)]
_Py_wrealpath() requires the size of the output buffer

14 years ago_Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows
Victor Stinner [Thu, 7 Oct 2010 22:23:10 +0000 (22:23 +0000)]
_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.

14 years agoFix fileutils for Windows
Victor Stinner [Thu, 7 Oct 2010 22:09:40 +0000 (22:09 +0000)]
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

14 years agoOoops, fileutils.c contains twice the same code
Victor Stinner [Thu, 7 Oct 2010 21:55:44 +0000 (21:55 +0000)]
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.

14 years agoCreate fileutils.c/.h
Victor Stinner [Thu, 7 Oct 2010 21:45:39 +0000 (21:45 +0000)]
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

14 years ago_wrealpath() and _Py_wreadlink() support surrogates (PEP 383)
Victor Stinner [Thu, 7 Oct 2010 11:06:49 +0000 (11:06 +0000)]
_wrealpath() and _Py_wreadlink() support surrogates (PEP 383)

Use _Py_wchar2char() to support surrogate characters in the input path.

14 years agoFix #8584. Fix the skipping of a number of tests due to incorrect importing.
Brian Curtin [Thu, 7 Oct 2010 01:12:19 +0000 (01:12 +0000)]
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.

14 years agoPyUnicode_AsWideCharString() takes a PyObject*, not a PyUnicodeObject*
Victor Stinner [Thu, 7 Oct 2010 01:02:42 +0000 (01:02 +0000)]
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().

14 years agotime: fix gcc warning
Victor Stinner [Thu, 7 Oct 2010 01:00:52 +0000 (01:00 +0000)]
time: fix gcc warning

 * Create format_arg variable to use the right types
 * Strip trailing spaces

14 years agoRewrite RunMainFromImporter()
Victor Stinner [Wed, 6 Oct 2010 23:24:57 +0000 (23:24 +0000)]
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

14 years agoCreate a subfunction for PySys_SetArgvEx()
Victor Stinner [Wed, 6 Oct 2010 22:44:06 +0000 (22:44 +0000)]
Create a subfunction for PySys_SetArgvEx()

Create sys_update_path() static function. Do nothing if argc==0.

14 years agoIssue #9759: GzipFile now raises ValueError when an operation is attempted
Antoine Pitrou [Wed, 6 Oct 2010 21:21:18 +0000 (21:21 +0000)]
Issue #9759: GzipFile now raises ValueError when an operation is attempted
after the file is closed.  Patch by Jeffrey Finkelstein.

14 years agoMore new things in what's new
Antoine Pitrou [Wed, 6 Oct 2010 21:13:56 +0000 (21:13 +0000)]
More new things in what's new

14 years agoFakeSocket is (supposed to be) gone in 3.x
Antoine Pitrou [Wed, 6 Oct 2010 18:38:22 +0000 (18:38 +0000)]
FakeSocket is (supposed to be) gone in 3.x

14 years agoFixes 9903: test_concurrent_futures writes on stderr
Brian Quinlan [Wed, 6 Oct 2010 13:05:45 +0000 (13:05 +0000)]
Fixes 9903: test_concurrent_futures writes on stderr

14 years agoUpdate to Sphin 1.0.4.
Georg Brandl [Wed, 6 Oct 2010 12:29:49 +0000 (12:29 +0000)]
Update to Sphin 1.0.4.

14 years agoFix the suspicious builder not to write CRLF, update suspicious file and add instruct...
Georg Brandl [Wed, 6 Oct 2010 10:35:24 +0000 (10:35 +0000)]
Fix the suspicious builder not to write CRLF, update suspicious file and add instructions to Makefile.

14 years agoFix errors found by "make suspicious".
Georg Brandl [Wed, 6 Oct 2010 10:26:05 +0000 (10:26 +0000)]
Fix errors found by "make suspicious".

14 years agoFix spelling and markup.
Georg Brandl [Wed, 6 Oct 2010 10:14:08 +0000 (10:14 +0000)]
Fix spelling and markup.

14 years agoMigrate to Sphinx 1.0 C language constructs.
Georg Brandl [Wed, 6 Oct 2010 10:11:56 +0000 (10:11 +0000)]
Migrate to Sphinx 1.0 C language constructs.

14 years agoCopyedit of os.symlink() docs.
Georg Brandl [Wed, 6 Oct 2010 08:52:48 +0000 (08:52 +0000)]
Copyedit of os.symlink() docs.

14 years ago#10034: import readline in longer example.
Georg Brandl [Wed, 6 Oct 2010 05:08:32 +0000 (05:08 +0000)]
#10034: import readline in longer example.

14 years ago#2982 More test coverage for the pyexpat parser.
Amaury Forgeot d'Arc [Tue, 5 Oct 2010 23:14:47 +0000 (23:14 +0000)]
#2982 More test coverage for the pyexpat parser.

14 years ago#9060 Let platforms without dup2() compile the replacement fonction without error.
Amaury Forgeot d'Arc [Tue, 5 Oct 2010 22:15:37 +0000 (22:15 +0000)]
#9060 Let platforms without dup2() compile the replacement fonction without error.

14 years agofix a typo.
Senthil Kumaran [Tue, 5 Oct 2010 18:46:56 +0000 (18:46 +0000)]
fix a typo.

14 years agoRemove the references of FTPS from the urllib Documentation. There is no support...
Senthil Kumaran [Tue, 5 Oct 2010 18:35:41 +0000 (18:35 +0000)]
Remove the references of FTPS from the urllib Documentation. There is no support for it yet.

14 years agoFix test_tk under OS X with Tk 8.4. Patch by Ned Deily.
Antoine Pitrou [Tue, 5 Oct 2010 11:24:49 +0000 (11:24 +0000)]
Fix test_tk under OS X with Tk 8.4. Patch by Ned Deily.
This should fix some buildbot failures.

14 years agoIndent version tags properly.
Georg Brandl [Tue, 5 Oct 2010 07:16:01 +0000 (07:16 +0000)]
Indent version tags properly.

14 years agoFix news entry formatting nits
Éric Araujo [Tue, 5 Oct 2010 00:04:20 +0000 (00:04 +0000)]
Fix news entry formatting nits

14 years agoFix interaction of custom translation classes and caching (#9042)
Éric Araujo [Mon, 4 Oct 2010 23:52:37 +0000 (23:52 +0000)]
Fix interaction of custom translation classes and caching (#9042)

14 years agoRecode developers.txt to UTF-8, as discussed on python-dev.
Éric Araujo [Mon, 4 Oct 2010 22:32:36 +0000 (22:32 +0000)]
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

14 years agoGzipFile.peek improvements, suggested by Nir Aides.
Antoine Pitrou [Mon, 4 Oct 2010 21:55:14 +0000 (21:55 +0000)]
GzipFile.peek improvements, suggested by Nir Aides.

14 years agoFix issue 6706: adds new handle_accepted() method to asyncore.dispatcher
Giampaolo Rodolà [Mon, 4 Oct 2010 21:08:36 +0000 (21:08 +0000)]
Fix issue 6706: adds new handle_accepted() method to asyncore.dispatcher

14 years agoAdd Tal Einat.
Martin v. Löwis [Mon, 4 Oct 2010 15:48:14 +0000 (15:48 +0000)]
Add Tal Einat.

14 years agoIssue #9065: no longer use "root" as the default for the
Lars Gustäbel [Mon, 4 Oct 2010 15:18:47 +0000 (15:18 +0000)]
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.

14 years agoRemove the debugging print line from the test.
Senthil Kumaran [Mon, 4 Oct 2010 05:42:58 +0000 (05:42 +0000)]
Remove the debugging print line from the test.

14 years agoFixes #10020
Gerhard Häring [Sun, 3 Oct 2010 21:47:06 +0000 (21:47 +0000)]
Fixes #10020

14 years agoFix Issue10012 - httplib headers, which are (sometimes mistakenly) int are explicitly...
Senthil Kumaran [Sun, 3 Oct 2010 18:22:42 +0000 (18:22 +0000)]
Fix Issue10012 - httplib headers, which are (sometimes mistakenly) int are explicitly cast to str (bytes - in py3k).

14 years agoFix Issue9272 - Change CGIHTTPServer to give the child program a copy of os.environ
Senthil Kumaran [Sun, 3 Oct 2010 17:55:45 +0000 (17:55 +0000)]
Fix Issue9272 - Change CGIHTTPServer to give the child program a copy of os.environ

14 years agotypo in Arfrever name
Tarek Ziadé [Sun, 3 Oct 2010 14:45:06 +0000 (14:45 +0000)]
typo in Arfrever name

14 years agoFixed #8980: distutils.command.check was failing w/ docutils installed
Tarek Ziadé [Sun, 3 Oct 2010 14:18:09 +0000 (14:18 +0000)]
Fixed #8980: distutils.command.check was failing w/ docutils installed

14 years agotypo
Benjamin Peterson [Sun, 3 Oct 2010 02:13:39 +0000 (02:13 +0000)]
typo

14 years agoadd a test and a note about metaclasses now being abcs
Benjamin Peterson [Sat, 2 Oct 2010 17:55:47 +0000 (17:55 +0000)]
add a test and a note about metaclasses now being abcs

14 years ago#1050268: make parseaddr 'quote' the contents of quoted strings in addresses.
R. David Murray [Sat, 2 Oct 2010 15:58:26 +0000 (15:58 +0000)]
#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.

14 years agoMake the printing of sqlite version in verbose mode work with regrtest -w.
R. David Murray [Sat, 2 Oct 2010 13:29:13 +0000 (13:29 +0000)]
Make the printing of sqlite version in verbose mode work with regrtest -w.

14 years agoUpdate changelog for r85174
Victor Stinner [Sat, 2 Oct 2010 11:50:49 +0000 (11:50 +0000)]
Update changelog for r85174

14 years agoIssue #8670: Rename testcapi unicode test methods
Victor Stinner [Sat, 2 Oct 2010 11:49:31 +0000 (11:49 +0000)]
Issue #8670: Rename testcapi unicode test methods

 * test_aswidechar() => unicode_aswidechar()
 * test_aswidecharstring() => unicode_aswidecharstring()

14 years agoOops, revert unwanted _testcapi changes of r85174
Victor Stinner [Sat, 2 Oct 2010 11:46:20 +0000 (11:46 +0000)]
Oops, revert unwanted _testcapi changes of r85174

14 years agoIssue #8670: ctypes.c_wchar supports non-BMP characters with 32 bits wchar_t
Victor Stinner [Sat, 2 Oct 2010 11:25:35 +0000 (11:25 +0000)]
Issue #8670: ctypes.c_wchar supports non-BMP characters with 32 bits wchar_t

14 years agoIssue #8670: PyUnicode_AsWideChar() and PyUnicode_AsWideCharString() replace
Victor Stinner [Sat, 2 Oct 2010 11:11:27 +0000 (11:11 +0000)]
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).

14 years agoIssue #8870: PyUnicode_AsWideCharString() doesn't count the trailing nul character
Victor Stinner [Sat, 2 Oct 2010 11:03:13 +0000 (11:03 +0000)]
Issue #8870: PyUnicode_AsWideCharString() doesn't count the trailing nul character

And write unit tests for PyUnicode_AsWideChar() and PyUnicode_AsWideCharString().

14 years agoUse proper variable name 'data' instead of 'str' in the send method.
Senthil Kumaran [Sat, 2 Oct 2010 10:33:13 +0000 (10:33 +0000)]
Use proper variable name 'data' instead of 'str' in the send method.

14 years agoFix - issue10010 .. index:: position in the docs.
Senthil Kumaran [Sat, 2 Oct 2010 03:16:04 +0000 (03:16 +0000)]
Fix - issue10010 .. index:: position in the docs.

14 years agotype.__abstractmethods__ should raise an AttributeError #10006
Benjamin Peterson [Sat, 2 Oct 2010 00:03:31 +0000 (00:03 +0000)]
type.__abstractmethods__ should raise an AttributeError #10006

14 years agoLower the tone of the warning about SSL certificate validation.
Amaury Forgeot d'Arc [Fri, 1 Oct 2010 23:42:24 +0000 (23:42 +0000)]
Lower the tone of the warning about SSL certificate validation.

14 years agoFix docstring typo.
R. David Murray [Fri, 1 Oct 2010 20:38:33 +0000 (20:38 +0000)]
Fix docstring typo.

14 years ago#10004: in Q encoded word ignore '=xx' when xx is not valid hex.
R. David Murray [Fri, 1 Oct 2010 15:40:20 +0000 (15:40 +0000)]
#10004: in Q encoded word ignore '=xx' when xx is not valid hex.

Bug report and fix by Thomas Guettler.

14 years agoFix #10003. Add SIGBREAK to the set of valid signals on Windows.
Brian Curtin [Fri, 1 Oct 2010 14:49:24 +0000 (14:49 +0000)]
Fix #10003. Add SIGBREAK to the set of valid signals on Windows.

This fixes a regression noticed by bzr, introduced by issue #9324.

14 years agoIssue #6608: time.asctime is now checking struct tm fields its input
Alexander Belopolsky [Fri, 1 Oct 2010 14:18:49 +0000 (14:18 +0000)]
Issue #6608: time.asctime is now checking struct tm fields its input
before passing it to the system asctime.  Patch by MunSic Jeong.

14 years agoCosmetic fix for PCBuild/readme.txt.
Hirokazu Yamamoto [Fri, 1 Oct 2010 10:40:49 +0000 (10:40 +0000)]
Cosmetic fix for PCBuild/readme.txt.

14 years agoFix weird markup that caused latex to fail.
Georg Brandl [Fri, 1 Oct 2010 05:38:10 +0000 (05:38 +0000)]
Fix weird markup that caused latex to fail.

14 years agoFix a couple spelling errors in comments and delete redundant __len__ def.
R. David Murray [Fri, 1 Oct 2010 02:08:02 +0000 (02:08 +0000)]
Fix a couple spelling errors in comments and delete redundant __len__ def.

14 years agoIssue1491 - BaseHTTPServer incorrectly implements response code 100
Senthil Kumaran [Thu, 30 Sep 2010 06:09:18 +0000 (06:09 +0000)]
Issue1491 - BaseHTTPServer incorrectly implements response code 100

14 years agoRemove an unreferenced variable. len is no longer needed.
Brian Curtin [Wed, 29 Sep 2010 19:09:33 +0000 (19:09 +0000)]
Remove an unreferenced variable. len is no longer needed.

14 years agoIssue #9599: Further accuracy tweaks to loghelper. For an integer n that's small...
Mark Dickinson [Wed, 29 Sep 2010 19:06:36 +0000 (19:06 +0000)]
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.

14 years agoAdd testcases from bug report #9980.
Mark Dickinson [Wed, 29 Sep 2010 18:41:54 +0000 (18:41 +0000)]
Add testcases from bug report #9980.

14 years agoFix PyUnicode_AsWideCharString(): set *size if size is not NULL
Victor Stinner [Wed, 29 Sep 2010 17:55:12 +0000 (17:55 +0000)]
Fix PyUnicode_AsWideCharString(): set *size if size is not NULL

14 years agotest_imp: getfilesystemencoding() cannot be None anymore
Victor Stinner [Wed, 29 Sep 2010 16:59:46 +0000 (16:59 +0000)]
test_imp: getfilesystemencoding() cannot be None anymore

And the codec name is normalized.

14 years agoUse os.fsencode() to support surrogates
Victor Stinner [Wed, 29 Sep 2010 16:59:18 +0000 (16:59 +0000)]
Use os.fsencode() to support surrogates

14 years agoIssue #9630: Redecode filenames when setting the filesystem encoding
Victor Stinner [Wed, 29 Sep 2010 16:35:47 +0000 (16:35 +0000)]
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.

14 years agoFix NNTP when there's a ".netrc" file
Antoine Pitrou [Wed, 29 Sep 2010 16:19:50 +0000 (16:19 +0000)]
Fix NNTP when there's a ".netrc" file

14 years agoDisable the usenetrc option when testing nntplib
Antoine Pitrou [Wed, 29 Sep 2010 16:08:29 +0000 (16:08 +0000)]
Disable the usenetrc option when testing nntplib

14 years agoIssue #9360: Cleanup and improvements to the nntplib module. The API
Antoine Pitrou [Wed, 29 Sep 2010 15:03:40 +0000 (15:03 +0000)]
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.

14 years agoFix #9978. WINFUNCTYPE is from ctypes, not from ctypes.wintypes.
Brian Curtin [Wed, 29 Sep 2010 14:51:42 +0000 (14:51 +0000)]
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.

14 years agoIssue #9983: warn that urllib and httplib don't perform SSL certificate validation.
Antoine Pitrou [Wed, 29 Sep 2010 11:24:21 +0000 (11:24 +0000)]
Issue #9983: warn that urllib and httplib don't perform SSL certificate validation.

14 years agoIssue #9962: GzipFile now has the peek() method.
Antoine Pitrou [Wed, 29 Sep 2010 10:49:46 +0000 (10:49 +0000)]
Issue #9962: GzipFile now has the peek() method.

14 years agoIssue #9979: Use PyUnicode_AsWideCharString() in _ctypes module
Victor Stinner [Wed, 29 Sep 2010 10:37:16 +0000 (10:37 +0000)]
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

14 years agoIssue #9979: Use PyUnicode_AsWideCharString() in time.strftime()
Victor Stinner [Wed, 29 Sep 2010 10:34:19 +0000 (10:34 +0000)]
Issue #9979: Use PyUnicode_AsWideCharString() in time.strftime()

Allocate memory with PyMem_Alloc() instead of the PyBytes API. Prepare the
surrogates support.

14 years agoIssue #9979: Use PyUnicode_AsWideCharString() for _locale.strcoll()
Victor Stinner [Wed, 29 Sep 2010 10:30:43 +0000 (10:30 +0000)]
Issue #9979: Use PyUnicode_AsWideCharString() for _locale.strcoll()

It simplifies the code and prepare the surrogates support.

14 years agoIssue #9979: Use PyUnicode_AsWideCharString() in import.c
Victor Stinner [Wed, 29 Sep 2010 10:28:51 +0000 (10:28 +0000)]
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).

14 years agoIssue #9979: Create function PyUnicode_AsWideCharString().
Victor Stinner [Wed, 29 Sep 2010 10:25:54 +0000 (10:25 +0000)]
Issue #9979: Create function PyUnicode_AsWideCharString().

14 years agolinecache.updatecache(): don't the lines into the cache on IOError
Victor Stinner [Wed, 29 Sep 2010 01:30:45 +0000 (01:30 +0000)]
linecache.updatecache(): don't the lines into the cache on IOError

Use the same behaviour than Python 2.7.

14 years agolinecache.updatecache() returns an empty list on error
Victor Stinner [Wed, 29 Sep 2010 01:24:59 +0000 (01:24 +0000)]
linecache.updatecache() returns an empty list on error

lines variable was not set on IOError

14 years agoHave test_sqlite print version info when run in verbose mode.
R. David Murray [Wed, 29 Sep 2010 01:22:20 +0000 (01:22 +0000)]
Have test_sqlite print version info when run in verbose mode.