26c03bd7d5
Fix memory leak in path_converter()
...
Issue #28200 : Replace PyUnicode_AsWideCharString() with
PyUnicode_AsUnicodeAndSize().
2016-09-19 11:55:44 +02:00
286987bbac
delete dead code
2016-09-18 23:49:51 -07:00
f5e8e8fc2b
merge 3.5 ( #24022 )
2016-09-18 23:44:02 -07:00
57bda335e1
merge 3.4
2016-09-18 23:43:18 -07:00
26d998cfdd
properly handle the single null-byte file ( closes #24022 )
2016-09-18 23:41:11 -07:00
6692f01c91
merge
2016-09-18 21:46:08 -07:00
7eb1becc25
Issue #28189 : dictitems_contains no longer swallows compare errors.
...
(Patch by Xiang Zhang)
2016-09-18 21:45:11 -07:00
5820f3a381
Issue #28193 : Use lru_cache in the re module.
2016-09-18 20:17:21 -07:00
5d4b09c106
correct silly spelling problem
2016-09-18 19:24:52 -07:00
19517e4da7
replace obmalloc's homegrown uptr and uchar types with standard ones
2016-09-18 19:22:22 -07:00
3924f93794
improvements to code that checks whether Python (obmalloc) allocated an address
...
- Rename Py_ADDRESS_IN_RANGE to address_in_range and make it a static
function instead of macro. Any compiler worth its salt will inline this
function.
- Remove the duplicated function version of Py_ADDRESS_IN_RANGE used when memory
analysis was active. Instead, we can simply mark address_in_range as allergic
to dynamic memory checking. We can now remove the
__attribute__((no_address_safety_analysis)) from _PyObject_Free and
_PyObject_Realloc. All the badness is contained in address_in_range now.
- Fix the code that tried to only read pool->arenaindex once. Putting something
in a variable is no guarantee that it won't be read multiple times. We must
use volatile for that.
2016-09-18 19:12:48 -07:00
ac965ca16c
stop using Py_LL and Py_ULL
2016-09-18 18:12:21 -07:00
41e35f37ed
always define HAVE_LONG_LONG ( #27961 )
2016-09-18 18:02:58 -07:00
9ac11a752a
properly free memory in pgen
2016-09-18 18:00:25 -07:00
254ad58c07
Fix test_huntrleaks_fd_leak() of test_regrtest
...
Issue #28195 : Don't expect the fd leak message to be on a specific line number,
just make sure that the line is present in the output.
2016-09-19 00:11:30 +02:00
3515dcce80
issue23591: fix flag decomposition and repr
2016-09-18 13:15:41 -07:00
9a7bbb2e3f
Issue #25400 : RobotFileParser now correctly returns default values for crawl_delay and request_rate
...
Initial patch by Peter Wirtz.
2016-09-18 20:17:58 +03:00
85c98bf968
Issue #26661 : setup.py now detects system libffi with multiarch wrapper.
2016-09-18 14:34:13 +02:00
96b2dd5997
Issue #26661 : setup.py now detects system libffi with multiarch wrapper.
2016-09-18 14:33:30 +02:00
44efa8aa0b
Null merge from 3.5
2016-09-18 14:53:12 +03:00
5a5ae745ff
Issue #26610 : Skip test_with_pip if _ctypes is not available in OpenIndiana
2016-09-18 14:52:25 +03:00
b66df15a86
Null merge from 3.5
2016-09-18 13:57:40 +03:00
bf3c1c3235
Issue #28075 : Fix test_access_denied in Python 3.5
...
I forgot there two variations of os.stat() in Python 3.5.
2016-09-18 13:56:29 +03:00
2a8d7f1c47
Issue #28151 : Use pythontest.net in test_robotparser
2016-09-18 11:21:57 +03:00
a2365c1de2
Null merge from 3.5
2016-09-17 17:29:52 -07:00
17a564ecc9
Add file missed from backport
2016-09-17 17:29:37 -07:00
74f4af7ac3
Issue #27932 : Prevent memory leak in win32_ver().
2016-09-17 17:27:48 -07:00
1374053127
Null merge from 3.5
2016-09-17 17:27:41 -07:00
1ec262be80
Issue #27932 : Prevent memory leak in win32_ver().
2016-09-17 17:25:42 -07:00
7e3dfd1548
Fixes bad merge for issue #28110
2016-09-17 16:34:38 -07:00
7358efa50b
Issue #28110 : launcher.msi has different product codes between 32-bit and 64-bit
2016-09-17 16:13:59 -07:00
1c75c18ed2
Issue #28110 : launcher.msi has different product codes between 32-bit and 64-bit
2016-09-17 16:12:55 -07:00
a7b1058c17
Issue #28176 : Merge from 3.5
2016-09-18 01:07:30 +03:00
bdcc8f9d1b
Issue #28176 : Increase timeout for test_sock_connect_sock_write_race
2016-09-18 01:06:50 +03:00
6dd8eca4a9
Issue #28192 : Adds tests for hook in isolated mode
2016-09-17 14:35:32 -07:00
27f26ad962
Issue #28161 : Opening CON for write access fails
...
Issue #28162 : WindowsConsoleIO readall() fails if first line starts with Ctrl+Z
Issue #28163 : WindowsConsoleIO fileno() passes wrong flags to _open_osfhandle
Issue #28164 : _PyIO_get_console_type fails for various paths
2016-09-17 13:51:23 -07:00
5d625cff07
Issue #26384 : Merge from 3.5
2016-09-17 23:23:13 +03:00
bcfb35f80d
Issue #26384 : Fix UnboundLocalError in socket._sendfile_use_sendfile
2016-09-17 23:22:06 +03:00
ed51b26858
Issue #28137 : Renames Windows path file to ._pth
...
Issue #28138 : Windows ._pth file should allow import site
2016-09-17 12:54:06 -07:00
313523ce2d
Issue #28192 : Don't import readline in isolated mode
2016-09-17 12:22:41 -07:00
d2154ab0b7
Merge from 3.5
2016-09-17 16:21:42 +03:00
9a1c91a10d
Fix test_access_denied in 3.5
2016-09-17 16:20:55 +03:00
052e4f18c4
Issue #28075 : Merge from 3.5
2016-09-17 15:51:14 +03:00
0b4dc4846b
Issue #28075 : Check for ERROR_ACCESS_DENIED in Windows implementation of os.stat()
...
Patch by Eryk Sun.
2016-09-17 15:49:59 +03:00
d508d00919
Issue #28139 : Merge indentation fixes from 3.5 into 3.6
2016-09-17 07:59:14 +00:00
6d57fe1c23
Issue #28139 : Fix messed up indentation
...
Also update the classmethod and staticmethod doc strings and comments to
match the RST documentation.
2016-09-17 03:26:16 +00:00
abf275af58
Issue #22493 : Warning message emitted by using inline flags in the middle of
...
regular expression now contains a (truncated) regex pattern.
Patch by Tim Graham.
2016-09-17 01:29:58 +03:00
8761e59f36
Pending final editing of 3.6 whatsnew, add a list of all PEPs implemented.
2016-09-16 15:46:55 -04:00
7b4bcd2004
Issue #25270 : Merge from 3.5
2016-09-16 17:32:06 +03:00
4a72a7b6c4
Issue #25270 : Prevent codecs.escape_encode() from raising SystemError when an empty bytestring is passed
2016-09-16 17:31:06 +03:00
Victor Stinner
Benjamin Peterson
Raymond Hettinger
Ethan Furman
Berker Peksag
Christian Heimes
Steve Dower
Martin Panter
Serhiy Storchaka
Ned Deily