Michael Foord [Sat, 8 May 2010 15:13:42 +0000 (15:13 +0000)]
Merged revisions 80980 via svnmerge from
svn+ssh://pythondev@svn.python.org/python/trunk
........
r80980 | michael.foord | 2010-05-08 17:09:37 +0200 (Sat, 08 May 2010) | 1 line
Documenting test discovery from package name and potential problems with test discovery importing tests from the wrong location. Issue 7780 and issue 8547.
........
Mark Dickinson [Sat, 8 May 2010 14:35:02 +0000 (14:35 +0000)]
Issue #8644: Improve accuracy of timedelta.total_seconds, by doing intermediate
computations with integer arithmetic instead of floating point.
td.total_seconds() now agrees with td / timedelta(seconds = 1).
Victor Stinner [Sat, 8 May 2010 00:36:42 +0000 (00:36 +0000)]
posix_error_with_allocated_filename() decodes the filename with
PyUnicode_DecodeFSDefaultAndSize() and call
PyErr_SetFromErrnoWithFilenameObject() instead of
PyErr_SetFromErrnoWithFilename()
Merged revisions 79911,79916-79917,80018,80418,80572-80573,80635-80639,80668,80922 via svnmerge from
svn+ssh://pythondev@svn.python.org/sandbox/trunk/2to3/lib2to3
use absolute import
........
r79916 | benjamin.peterson | 2010-04-09 16:05:21 -0500 (Fri, 09 Apr 2010) | 1 line
generalize detection of __future__ imports and attach them to the tree
........
r79917 | benjamin.peterson | 2010-04-09 16:11:44 -0500 (Fri, 09 Apr 2010) | 1 line
don't try to 'fix' relative imports when absolute_import is enabled #8858
........
r80018 | benjamin.peterson | 2010-04-12 16:12:12 -0500 (Mon, 12 Apr 2010) | 4 lines
prevent diffs from being mangled is multiprocess mode #6409
Patch by George Boutsioukis.
........
r80418 | benjamin.peterson | 2010-04-23 16:00:03 -0500 (Fri, 23 Apr 2010) | 1 line
Don't transform imports that are already relative. 2to3 turned
from . import refactor
into
from .. import refactor
which broke the transformation of 2to3 itself.
........
r80635 | benjamin.peterson | 2010-04-29 16:02:23 -0500 (Thu, 29 Apr 2010) | 1 line
pass string to Number
........
r80668 | jeffrey.yasskin | 2010-04-30 18:02:47 -0500 (Fri, 30 Apr 2010) | 4 lines
Make 2to3 run under Python 2.5 so that the benchmark suite at
http://hg.python.org/benchmarks/ can use it and still run on implementations
that haven't gotten to 2.6 yet. Fixes issue 8566.
........
r80922 | benjamin.peterson | 2010-05-07 11:06:25 -0500 (Fri, 07 May 2010) | 1 line
prevent xrange transformation from wrapping range calls it produces in list
........
................
Issue #8571: Fix an internal error when compressing or decompressing a
chunk larger than 1GB with the zlib module's compressor and decompressor
objects.
........
Victor Stinner [Thu, 6 May 2010 22:05:07 +0000 (22:05 +0000)]
Issue #8603: Create a bytes version of os.environ for Unix
Create os.environb mapping and os.getenvb() function, os.unsetenv() encodes str
argument to the file system encoding with the surrogateescape error handler
(instead of utf8/strict) and accepts bytes, and posix.environ keys and values
are bytes.
Fix asyncore issues 8573 and 8483: _strerror might throw ValueError; asyncore.__getattr__ cheap inheritance caused confusing error messages when accessing undefined class attributes; added an alias for __str__ which now is used as a fallback for __repr__
........
R. David Murray [Thu, 6 May 2010 01:41:14 +0000 (01:41 +0000)]
Merged revisions 80800 via svnmerge from
svn+ssh://pythondev@svn.python.org/python/trunk
It turns out that email5 (py3k), because it is using unicode for the
payload, doesn't do the encoding to the output character set until later
in the process. Specifically, charset.body_encode no longer does the
input-to-output charset conversion. So the if test in the exception
clause in encoders.encode_7or8bit really is needed in email5.
So, this merge only merges the test, not the removal of the 'if'.
Issue #7472: remove unused code from email.encoders.encode_7or8bit.
Yukihiro Nakadaira noticed a typo in encode_7or8bit that was trying
to special case iso-2022 codecs. It turns out that the code in
question is never used, because whereas it was designed to trigger
if the payload encoding was eight bit but its output encoding was
7 bit, in practice the payload is always converted to the 7bit
encoding before encode_7or8bit is called. Patch by Shawat Anand.
........
Have the serve.py script announce the directory it is
serving and which port it is serving it on (I can
never remember the default port number it uses...)
........
In a number of places code still revers
to "sys.platform == 'mac'" and that is
dead code because it refers to a platform
that is no longer supported (and hasn't been
supported for several releases).
Barry Warsaw [Wed, 5 May 2010 16:18:31 +0000 (16:18 +0000)]
Bug 7755: audiotest.au is arguably copyrighted material, but definitely makes
Debian unhappy. The actual contents of the audio clip are unimportant, so
replace it with something that we know is okay. Guido likes woodpeckers.
On Windows, ctypes does no longer check the stack before and after
calling a foreign function. This allows to use the unmodified libffi
library.
Remove most files from _ctypes/libffi_msvc, only two include files
stay (updated from _ctypes/libffi/...). Other files are used in the
cross-platform _ctypes/libffi directory.
........
r80766 | thomas.heller | 2010-05-04 21:08:18 +0200 (Di, 04 Mai 2010) | 2 lines
Issue #1533: fix inconsistency in range function argument processing:
any non-float non-integer argument is now converted to an integer (if
possible) using its __int__ method. Previously, only small arguments
were treated this way; larger arguments (those whose __int__ was
outside the range of a C long) would produce a TypeError.
Patch by Alexander Belopolsky (with minor modifications).
........
r80759 | mark.dickinson | 2010-05-04 17:19:06 +0100 (Tue, 04 May 2010) | 1 line
Issue #8567: Fix incorrect precedence of signals in Decimal module.
When a Decimal operation raises multiple signals and more than one of
those signals is trapped, the specification determines the order in
which the signals should be handled. In many cases this order wasn't
being followed, leading to the wrong Python exception being raised.
This commit fixes those cases, and adds extra tests. The tests are
only enabled when EXTENDEDERRORTESTS is True, since they involve
rerunning each Decimal testcase several times.
........
Jeffrey Yasskin [Mon, 3 May 2010 19:29:34 +0000 (19:29 +0000)]
Make (most of) Python's tests pass under Thread Sanitizer.
http://code.google.com/p/data-race-test/wiki/ThreadSanitizer is a dynamic data
race detector that runs on top of valgrind. With this patch, the binaries at
http://code.google.com/p/data-race-test/wiki/ThreadSanitizer#Binaries pass many
but not all of the Python tests. All of regrtest still passes outside of tsan.
I've implemented part of the C1x atomic types so that we can explicitly mark
variables that are used across threads, and get defined behavior as compilers
advance.
I've added tsan's client header and implementation to the codebase in
dynamic_annotations.{h,c} (docs at
http://code.google.com/p/data-race-test/wiki/DynamicAnnotations).
Unfortunately, I haven't been able to get helgrind and drd to give sensible
error messages, even when I use their client annotations, so I'm not supporting
them.
Issue #7865: The close() method of :mod:`io` objects should not swallow
exceptions raised by the implicit flush(). Also ensure that calling
close() several times is supported. Patch by Pascal Chambon.
........
Victor Stinner [Sun, 2 May 2010 17:24:51 +0000 (17:24 +0000)]
Issue #8533: revert r80694; try a different fix: regrtest uses backslashreplace
error handler for stdout to avoid UnicodeEncodeError (write non-ASCII character
to stdout using ASCII encoding)
Only override the AC_PROG_CC determined CFLAGS if they were set by the user.
This restores the default behavior in the common case of not having CFLAGS
defined when running configure.
........
This function is only used to decode Python module filenames, but Python
doesn't support surrogates in modules filenames yet. So nobody noticed this
minor bug.
Fix for issue #3646: with this patch it is possible to do a
framework install of Python in your home directory (on OSX):
$ configure --enable-framework=${HOME}/Library/Frameworks
$ make && make install
Without this patch the framework would get installed just fine,
but 'make install' would try to install the application bundles
and command-line tools outside the user's home, which doesn't work
for non-admin users (and is bad form anyway).
........
projects / python / log