Commit Graph

5027 Commits

Author SHA1 Message Date
Neil Schemenauer 973f8b4ca6 Make cPickle.Unpickler.noload() handle dict subclasses. noload() is
an obscure, undocumentated feature so no test was added. Closes
issue #1101399.
2009-10-14 19:33:31 +00:00
Mark Dickinson 3d830828de Issue #7078: _struct.__doc__ was being ignored. Import it into struct.
Also add description of '?' struct format character.  Thanks Gabriel
Genellina for the patch.
2009-10-08 15:54:10 +00:00
Amaury Forgeot d'Arc d0052d17b1 #1571184: makeunicodedata.py now generates the functions _PyUnicode_ToNumeric,
_PyUnicode_IsLinebreak and _PyUnicode_IsWhitespace.

It now also parses the Unihan.txt for numeric values.
2009-10-06 19:56:32 +00:00
Amaury Forgeot d'Arc 3de4647022 Fix compilation warning on Windows, where size_t is 32bit but file offsets are 64bit. 2009-10-05 20:18:05 +00:00
Philip Jenvey dbf3b258e7 #6990: clear threading.local's key only after its thread state is removed:
fixes local subclasses leaving old state around after a ref cycle GC which
could be recycled by new locals
2009-09-29 04:32:44 +00:00
Mark Dickinson b93fff0a57 Issue #3366: Add gamma function to math module.
(lgamma, erf and erfc to follow).
2009-09-28 18:54:55 +00:00
Mark Dickinson 716a9ccd3d Eliminate unnecessary get_wrapped_(u)long defines in struct module. 2009-09-27 16:39:28 +00:00
Andrew M. Kuchling b49e53ee16 #6243: fix segfault when keyname() returns a NULL pointer.
Bug noted by Trundle, patched by Trundle and Jerry Chen.
2009-09-25 22:23:54 +00:00
Kristján Valur Jónsson 2fcd03bb77 http://bugs.python.org/issue6971
Adding the SIO_KEEPALIVE_VALS command to socket.ioctl on windows
2009-09-25 15:19:51 +00:00
Brett Cannon 6ef9fd7c97 Fix whitespace. 2009-09-22 19:13:27 +00:00
Brett Cannon 4d94743c28 When range checking was added to time.strftime() a check was placed on tm_isdst
to make sure it fell within [-1, 1] just in case someone implementing
strftime() in libc was stupid enough to assume this. Turns out, though, some
OSs (e.g. zOS) are stupid enough to use values outside of this range for time
structs created by the system itself. So instead of throwing a ValueError,
tm_isdst is now normalized before being passed to strftime().

Fixes issue #6823. Thanks Robert Shapiro for diagnosing the problem and
contributing an initial patch.
2009-09-22 00:29:48 +00:00
Antoine Pitrou 2a46658bee Issue #6236, #6348: Fix various failures in the io module under AIX
and other platforms, when using a non-gcc compiler. Patch by egreen.

In addition, I made explicit the signedness of all bitfields in the
IO library.
2009-09-21 21:17:48 +00:00
Ronald Oussoren 16766d7cd6 Merge a newer version of libffi_osx, based on the
version of libffi in OSX 10.6.1.

This fixes issue6918
2009-09-20 18:54:16 +00:00
Ronald Oussoren 9f20d9d0ee Issue 6877: this patch makes it possible to link the readline extension
to the libedit emulation of the readline API on OSX 10.5 or later.

This also adds a minimal testsuite for readline to check that the
history manipuation functions have the same interface with both
C libraries.
2009-09-20 14:18:15 +00:00
Thomas Heller d25fb2ff1f Remove unused variable and static function to fix compiler warnings. 2009-09-19 10:04:54 +00:00
Georg Brandl ab849891ef #6944: the argument to PyArg_ParseTuple should be a tuple, otherwise a SystemError is set. Also clean up another usage of PyArg_ParseTuple. 2009-09-19 07:35:07 +00:00
Thomas Heller 6be522bfc2 Issue #4606: Passing 'None' if ctypes argtype is set to POINTER(...)
does now always result in NULL.
2009-09-18 20:05:44 +00:00
Thomas Heller 7a352c0ed8 Issue #5042: Structure sub-subclass does now initialize correctly with
base class positional arguments.
2009-09-18 18:55:17 +00:00
Mark Dickinson 4676048b43 Remove outdated include; this include was breaking OS X builds using
non-Apple gcc4.3 and gcc4.4 (because CoreFoundation/CoreFoundation.h
won't compile under non-Apple gcc).
2009-09-16 20:26:31 +00:00
Thomas Wouters c4dcb3801e Fix issue #1590864, multiple threads and fork() can cause deadlocks, by
acquiring the import lock around fork() calls. This prevents other threads
from having that lock while the fork happens, and is the recommended way of
dealing with such issues. There are two other locks we care about, the GIL
and the Thread Local Storage lock. The GIL is obviously held when calling
Python functions like os.fork(), and the TLS lock is explicitly reallocated
instead, while also deleting now-orphaned TLS data.

This only fixes calls to os.fork(), not extension modules or embedding
programs calling C's fork() directly. Solving that requires a new set of API
functions, and possibly a rewrite of the Python/thread_*.c mess. Add a
warning explaining the problem to the documentation in the mean time.

This also changes behaviour a little on AIX. Before, AIX (but only AIX) was
getting the import lock reallocated, seemingly to avoid this very same
problem. This is not the right approach, because the import lock is a
re-entrant one, and reallocating would do the wrong thing when forking while
holding the import lock.

Will backport to 2.6, minus the tiny AIX behaviour change.
2009-09-16 19:55:54 +00:00
Benjamin Peterson 8246968b12 tabbify 2009-09-09 11:40:54 +00:00
Benjamin Peterson 7f6d0834f9 #6865 fix ref counting in initialization of pwd module 2009-09-08 23:04:22 +00:00
Thomas Heller 81809a42c1 Make ctypes compile again with older Python versions. 2009-09-08 19:24:36 +00:00
Amaury Forgeot d'Arc fff896b309 #6750: TextIOWrapped could duplicate output when several threads write to it.
this affect text files opened with io.open(), and the print() function of py3k
2009-08-29 18:14:40 +00:00
Gregory P. Smith 4e63d54b36 Add weakref support to the thread.lock type. 2009-08-20 09:39:38 +00:00
Brett Cannon 764465f315 Expat could crash if given the wrong kind of input by never stopping its
tokenizing step.

Thanks to Ivan Krstić for the patch.
2009-08-13 19:27:12 +00:00
Gregory P. Smith c4ad0345cf Fix issue1628205: Socket file objects returned by socket.socket.makefile() now
properly handles EINTR within the read, readline, write & flush methods.
The socket.sendall() method now properly handles interrupted system calls.
2009-08-13 18:54:50 +00:00
Gregory P. Smith f466642c66 comment typo fix 2009-08-12 17:02:37 +00:00
Antoine Pitrou 20e1f932fa Issue #6629: Fix a data corruption issue in the new `io` package, which could
occur when writing to a BufferedRandom object (e.g. a file opened in "rb+" or
"wb+" mode) after having buffered a certain amount of data for reading. This
bug was not present in the pure Python implementation.

Yes, this is a serious issue.
2009-08-06 20:18:29 +00:00
Mark Dickinson 8d5a6522a9 Issue #6644: Fix compile error on AIX. 2009-08-04 19:22:35 +00:00
Raymond Hettinger 8fdab95260 Issue 6637: defaultdict.copy() failed with an empty factory. 2009-08-04 19:08:05 +00:00
Mark Dickinson fe67bd9168 Issue #6561: '\d' regular expression should not match characters of
category [No]; only those of category [Nd].  (Backport of r74237
from py3k.)
2009-07-28 20:35:03 +00:00
Amaury Forgeot d'Arc 74b3016783 #6553: crash in cPickle.load(), when given a StringIO with incomplete data.
Will backport to 2.6, 3.x already fixed a similar issue with issue4298.
2009-07-23 19:26:02 +00:00
Thomas Heller d8431ae3bd Revert rev 74134, as it does not completely fixx issue #6493. 2009-07-21 19:04:02 +00:00
Thomas Heller 6adda9641d Issue #6493: Fix a ctypes problem setting bitfields more than 31 bits
wide.
2009-07-21 06:27:14 +00:00
Alexandre Vassalotti 0b5b8af435 Fix GCC warning about fprintf used without a string literal and
without format arguments.
2009-07-17 08:09:04 +00:00
Alexandre Vassalotti 6d78736177 Initialize variables in PyCurses_getsyx() to avoid compiler warnings. 2009-07-17 07:54:23 +00:00
Georg Brandl d77faaf48f #5910: fix kqueue for calls with more than one event. 2009-07-16 07:18:07 +00:00
Alexandre Vassalotti 999ecc0eaf Issue #2389: Pickle array objects using a list representation for portability
across different machine architectures and compatibility with Python 3.x.
2009-07-15 18:19:47 +00:00
Amaury Forgeot d'Arc ce32eb7406 #6416: Fix compilation of the select module on Windows, as well as test_subprocess:
PIPE_BUF is not defined on Windows, and probably has no meaning there.

Anyway the subprocess module uses another way to perform non-blocking reads (with a thread)
2009-07-09 22:37:22 +00:00
Mark Dickinson 5fd3af24a2 Issue #1523: Remove deprecated overflow masking in struct module, and
make sure that out-of-range values consistently raise struct.error.
2009-07-07 15:08:28 +00:00
Mark Dickinson 24766ba6af Typo in error message 2009-07-07 10:18:22 +00:00
Amaury Forgeot d'Arc e93eee7d22 #6420: Fix a compilation warning in the nis module, for OpenBSD and FreeBSD. 2009-07-07 06:49:41 +00:00
Mark Dickinson 463dc4bf26 Issues #1530559, #1741130: Fix various inconsistencies in struct.pack
integer packing, and reenable some previously broken tests.
2009-07-05 10:01:24 +00:00
Alexandre Vassalotti 0fe799151f Issue 2370: Add Python 3 warnings for the removal of operator.isCallable and
operator.sequenceIncludes.

Patch contributed by Jeff Balogh (and updated slightly by me).
2009-07-05 04:22:40 +00:00
Gregory P. Smith 9d36fd2acb Adds the select.PIPE_BUF attribute to expose the system constant. 2009-07-03 20:48:31 +00:00
Jesus Cea 869321484d multiprocessing doesn't compile in Solaris because a typo 2009-07-02 14:30:18 +00:00
Hirokazu Yamamoto 8839fd7372 Issue #6344: Fixed a crash of mmap.read() when passed a negative argument.
Reviewed by Amaury Forgeot d'Arc.
2009-06-29 13:25:16 +00:00
Hirokazu Yamamoto bcff47a6ec Issue #4856: Py_GetFileAttributesEx[AW] are not needed because GetFileAttributesEx[AW]
won't fail with ERROR_CALL_NOT_IMPLEMENTED on win NT.
Reviewed by Amaury Forgeot d'Arc.
2009-06-29 11:27:03 +00:00
Hirokazu Yamamoto a3c5609079 Issue #4856: Remove checks for win NT. 2009-06-28 10:23:00 +00:00