]> granicus.if.org Git - python/log
python
14 years agoMake decorators used in packaging preserve docstrings
Éric Araujo [Fri, 17 Jun 2011 19:10:21 +0000 (21:10 +0200)]
Make decorators used in packaging preserve docstrings

14 years agoBrange merge
Éric Araujo [Fri, 17 Jun 2011 17:39:18 +0000 (19:39 +0200)]
Brange merge

14 years agoBrange merge
Éric Araujo [Fri, 17 Jun 2011 17:38:38 +0000 (19:38 +0200)]
Brange merge

14 years agomerge #11767: use context manager to close file in __getitem__ to prevent FD leak
R David Murray [Fri, 17 Jun 2011 16:56:41 +0000 (12:56 -0400)]
merge #11767: use context manager to close file in __getitem__ to prevent FD leak

14 years ago#11767: use context manager to close file in __getitem__ to prevent FD leak
R David Murray [Fri, 17 Jun 2011 16:54:56 +0000 (12:54 -0400)]
#11767: use context manager to close file in __getitem__ to prevent FD leak

All of the other methods in mailbox that create message objects take care to
close the file descriptors they use, so it seems to make sense to have
__getitem__ do so as well.

Patch by Filip Gruszczyński.

14 years ago#12313: update Makefile.pre.in to account for email tests moving to 'test' dir
R David Murray [Fri, 17 Jun 2011 15:41:49 +0000 (11:41 -0400)]
#12313: update Makefile.pre.in to account for email tests moving to 'test' dir

14 years agoMinor tweaks to packaging tests.
Éric Araujo [Fri, 17 Jun 2011 13:47:41 +0000 (15:47 +0200)]
Minor tweaks to packaging tests.

- Move a tearDown method right after setUp
- Use assertRaises instead of reinventing it
- Skip a test instead of commenting it out, as a reminder

14 years agoPackaging tests: don’t let an internal cache grow indefinitely.
Éric Araujo [Fri, 17 Jun 2011 13:43:18 +0000 (15:43 +0200)]
Packaging tests: don’t let an internal cache grow indefinitely.

Victor Stinner diagnosed on #12167 that some reference leaks came from
util._path_created, a set used for caching; there are two tests that
cause additions to this set, so now they clear it in tearDown, avoiding
17 refleaks.  (My tests show that it’s necessary to clear the set in
only one test, clearing it in both does not stop more refleaks, but
there’s no harm in doing it.)

14 years ago(Merge 3.2) posixmodule.c: fix function name in argument parsing
Victor Stinner [Fri, 17 Jun 2011 13:18:16 +0000 (15:18 +0200)]
(Merge 3.2) posixmodule.c: fix function name in argument parsing

Fix os.fchown() and os.open()

Remove also trailing spaces and replace tabs by spaces.

14 years agoposixmodule.c: fix function name in argument parsing
Victor Stinner [Fri, 17 Jun 2011 13:15:38 +0000 (15:15 +0200)]
posixmodule.c: fix function name in argument parsing

Fix os.fchown() and os.open()

Remove also trailing spaces and replace tabs by spaces.

14 years ago(Merge 3.2) Issue #10883: test_urllib2net closes socket explicitly
Victor Stinner [Fri, 17 Jun 2011 12:53:41 +0000 (14:53 +0200)]
(Merge 3.2) Issue #10883: test_urllib2net closes socket explicitly

14 years agoIssue #10883: test_urllib2net closes socket explicitly
Victor Stinner [Fri, 17 Jun 2011 12:53:02 +0000 (14:53 +0200)]
Issue #10883: test_urllib2net closes socket explicitly

14 years ago(Merge 3.2) Issue #12133: fix a ResourceWarning in urllib.request
Victor Stinner [Fri, 17 Jun 2011 12:02:18 +0000 (14:02 +0200)]
(Merge 3.2) Issue #12133: fix a ResourceWarning in urllib.request

AbstractHTTPHandler.do_open() of urllib.request closes the HTTP connection if
its getresponse() method fails with a socket error. Patch written by Ezio
Melotti.

14 years agoIssue #12133: fix a ResourceWarning in urllib.request
Victor Stinner [Fri, 17 Jun 2011 12:01:18 +0000 (14:01 +0200)]
Issue #12133: fix a ResourceWarning in urllib.request

AbstractHTTPHandler.do_open() of urllib.request closes the HTTP connection if
its getresponse() method fails with a socket error. Patch written by Ezio
Melotti.

14 years agoIssue #12333: run tests on the new module in a subprocess
Victor Stinner [Fri, 17 Jun 2011 11:52:56 +0000 (13:52 +0200)]
Issue #12333: run tests on the new module in a subprocess

It is not possible to unload a module written in C, so use a subprocess to run
the tests on the module compiled by test_build_ext(). Using a subprocess, we
don't have to unload the module, save/restore sys.path, and the test can be run
more than once.

This commit fixes also an access error on rmtree() on Windows: because the
module was not really unloaded, it was not possible to remove the temporary
directory (it is not possible to remove a directory on Windows if it still
contains an open file).

14 years agoIssue #12333: close files before removing the directory
Victor Stinner [Fri, 17 Jun 2011 11:25:53 +0000 (13:25 +0200)]
Issue #12333: close files before removing the directory

packaging.tests.support.TempdirManager: rmtree() fails on Windows if there are
still open files in the directory.

14 years agoMinor tweaks in packaging’s test_dist.
Éric Araujo [Fri, 17 Jun 2011 11:24:33 +0000 (13:24 +0200)]
Minor tweaks in packaging’s test_dist.

- Use different Metadata objects to write and read a PKG-INFO (METADATA)
  file, to make sure the tested values come from the file

- No need to restore methods on an instance after monkey-patching them:
  the methods are still the same on the class

- Harmonize dedent calls

14 years ago(Merge 3.2) Issue #12310: finalize the old process after _run_after_forkers()
Victor Stinner [Fri, 17 Jun 2011 10:36:26 +0000 (12:36 +0200)]
(Merge 3.2) Issue #12310: finalize the old process after _run_after_forkers()

multiprocessing: Process._bootstrap() keeps a reference to the old process to
delay its finalization until after _run_after_forkers() as been executed. This
change should fix a crash on Mac OS X Tiger when a lock is released after a
fork.

Patch written by Charles-François Nataliv and Antoine Pitrou.

14 years agoIssue #12310: finalize the old process after _run_after_forkers()
Victor Stinner [Fri, 17 Jun 2011 10:31:49 +0000 (12:31 +0200)]
Issue #12310: finalize the old process after _run_after_forkers()

multiprocessing: Process._bootstrap() keeps a reference to the old process to
delay its finalization until after _run_after_forkers() as been executed. This
change should fix a crash on Mac OS X Tiger when a lock is released after a
fork.

Patch written by Charles-François Nataliv and Antoine Pitrou.

14 years agoIssue #12333: restore the previous dir before removing the current directory
Victor Stinner [Fri, 17 Jun 2011 10:20:46 +0000 (12:20 +0200)]
Issue #12333: restore the previous dir before removing the current directory

packaging.tests.support.TempdirManager: removing the current directory is not
allowed on Windows or Solaris. Store the current directory and restore it
before removing the temporary directory (which is used as the working directory
during the tests).

14 years agomerge heads
Benjamin Peterson [Thu, 16 Jun 2011 23:51:42 +0000 (18:51 -0500)]
merge heads

14 years agomerge 3.2
Benjamin Peterson [Thu, 16 Jun 2011 23:51:37 +0000 (18:51 -0500)]
merge 3.2

14 years agomerge heads
Benjamin Peterson [Thu, 16 Jun 2011 23:51:24 +0000 (18:51 -0500)]
merge heads

14 years agomerge 3.2
Benjamin Peterson [Thu, 16 Jun 2011 23:50:21 +0000 (18:50 -0500)]
merge 3.2

14 years agoupdate link to pycrypto (closes #12351)
Benjamin Peterson [Thu, 16 Jun 2011 23:49:46 +0000 (18:49 -0500)]
update link to pycrypto (closes #12351)

14 years agoRemove unused code in packaging.pypi.dist
Éric Araujo [Thu, 16 Jun 2011 21:50:17 +0000 (23:50 +0200)]
Remove unused code in packaging.pypi.dist

14 years agoStop binding sys.path as default parameter value in packaging.
Éric Araujo [Thu, 16 Jun 2011 21:43:15 +0000 (23:43 +0200)]
Stop binding sys.path as default parameter value in packaging.

The two public functions in database default to sys.path if the given
*paths* argument is None; the private functions don’t have default
values for their arguments anymore, which is fine as the public
functions that call them pass their arguments down.  Likewise in
install, the functions will pass down their *paths* arguments down to
database functions.

A one-line unneeded function in install was removed instead of being
changed, and the few remaining tests that used brute-force restoration
of sys.path have been cleaned up to use sys.path.remove.

14 years agoPackaging: remove last mentions and uses of setup.py in the code.
Éric Araujo [Thu, 16 Jun 2011 21:34:55 +0000 (23:34 +0200)]
Packaging: remove last mentions and uses of setup.py in the code.

Now only the compatibility layer (in create, util and install) talk
about setup.py.

14 years agoFix typo.
Raymond Hettinger [Thu, 16 Jun 2011 21:33:20 +0000 (22:33 +0100)]
Fix typo.

14 years agoFix typo.
Raymond Hettinger [Thu, 16 Jun 2011 21:32:10 +0000 (22:32 +0100)]
Fix typo.

14 years agoClean up packaging.tests.test_mixin2to3
Éric Araujo [Thu, 16 Jun 2011 21:21:01 +0000 (23:21 +0200)]
Clean up packaging.tests.test_mixin2to3

14 years agoIssue #12167: Fix a reafleak in packaging.tests.PyPIServer constructor
Victor Stinner [Wed, 15 Jun 2011 21:58:57 +0000 (23:58 +0200)]
Issue #12167: Fix a reafleak in packaging.tests.PyPIServer constructor

Don't modify mutable default arguments...

14 years agoMerge wording change from 3.2
Éric Araujo [Wed, 15 Jun 2011 15:51:05 +0000 (17:51 +0200)]
Merge wording change from 3.2

14 years agoMinor wording improvement
Éric Araujo [Wed, 15 Jun 2011 15:49:20 +0000 (17:49 +0200)]
Minor wording improvement

14 years agoMake sure all the Lib/packaging subdirectories get installed.
Barry Warsaw [Tue, 14 Jun 2011 20:51:58 +0000 (16:51 -0400)]
Make sure all the Lib/packaging subdirectories get installed.

14 years agoFix markup.
Georg Brandl [Tue, 14 Jun 2011 19:09:55 +0000 (21:09 +0200)]
Fix markup.

14 years agoMerge 3.2 - update to the fix for #12084
Brian Curtin [Tue, 14 Jun 2011 15:06:41 +0000 (10:06 -0500)]
Merge 3.2 - update to the fix for #12084

14 years agoCorrect completely broken os.stat behavior on Windows XP.
Brian Curtin [Tue, 14 Jun 2011 14:52:50 +0000 (09:52 -0500)]
Correct completely broken os.stat behavior on Windows XP.

After 1a3e8db28d49, Windows XP could not os.stat at all due to raising
immediately when GetFinalPathNameByHandle wasn't available (pre-Vista).
The proper behavior in that situation is to just not attempt a traversal
rather than outright rejecting.

This change additionally handles a failed malloc by setting the error code
and returning false.

Patch by Hirokazu Yamamoto.

14 years agobranch merge?
Brian Curtin [Mon, 13 Jun 2011 21:10:32 +0000 (16:10 -0500)]
branch merge?

14 years agobranch merge
Brian Curtin [Mon, 13 Jun 2011 21:03:01 +0000 (16:03 -0500)]
branch merge

14 years agoMerge from 3.2 for Issue #12084.
Brian Curtin [Mon, 13 Jun 2011 21:00:35 +0000 (16:00 -0500)]
Merge from 3.2 for Issue #12084.

14 years agoFix #12084. os.stat on Windows wasn't working properly with relative symlinks.
Brian Curtin [Mon, 13 Jun 2011 20:16:04 +0000 (15:16 -0500)]
Fix #12084. os.stat on Windows wasn't working properly with relative symlinks.

Use of DeviceIoControl to obtain the symlink path via the reparse tag was
removed. The code now uses GetFinalPathNameByHandle in the case of a
symbolic link and works properly given the added test which creates a symbolic
link and calls os.stat on it from multiple locations.

Victor Stinner also noticed an issue with os.lstat following the os.stat
code path when being passed bytes. The posix_lstat function was adjusted to
properly hook up win32_lstat instead of the previous STAT macro (win32_stat).

14 years agoIssue #12316: Fix sigwait() test using threads
Victor Stinner [Mon, 13 Jun 2011 14:19:06 +0000 (16:19 +0200)]
Issue #12316: Fix sigwait() test using threads

Spawn a new process instead of using fork(). Patch written by Charles-François
Natali.

14 years agoAlso specify encoding when reading setup.cfg (#12320)
Éric Araujo [Sun, 12 Jun 2011 21:02:57 +0000 (23:02 +0200)]
Also specify encoding when reading setup.cfg (#12320)

14 years agoSpecify the encoding of the setup.cfg in one packaging test (#12320)
Éric Araujo [Sun, 12 Jun 2011 20:04:58 +0000 (22:04 +0200)]
Specify the encoding of the setup.cfg in one packaging test (#12320)

14 years agoMerged documentation update for Formatter.formatTime.
Vinay Sajip [Sun, 12 Jun 2011 10:50:40 +0000 (11:50 +0100)]
Merged documentation update for Formatter.formatTime.

14 years agoUpdated Formatter.formatTime documentation.
Vinay Sajip [Sun, 12 Jun 2011 10:44:28 +0000 (11:44 +0100)]
Updated Formatter.formatTime documentation.

14 years agoMerged documentation update for issue #12206.
Vinay Sajip [Sat, 11 Jun 2011 22:04:35 +0000 (23:04 +0100)]
Merged documentation update for issue #12206.

14 years agoIssue #12206: documentation for LogRecord constructor updated re. the level argument.
Vinay Sajip [Sat, 11 Jun 2011 22:03:37 +0000 (23:03 +0100)]
Issue #12206: documentation for LogRecord constructor updated re. the level argument.

14 years agoquaint and completely out of date comment
Benjamin Peterson [Sat, 11 Jun 2011 21:33:35 +0000 (16:33 -0500)]
quaint and completely out of date comment

14 years agoallow __dir__ to return any sequence
Benjamin Peterson [Sat, 11 Jun 2011 21:12:08 +0000 (16:12 -0500)]
allow __dir__ to return any sequence

14 years agomerge 3.2 (#9284)
Benjamin Peterson [Sat, 11 Jun 2011 20:56:46 +0000 (15:56 -0500)]
merge 3.2 (#9284)

14 years agoallow "fake" filenames in findsource (closes #9284)
Benjamin Peterson [Sat, 11 Jun 2011 20:53:11 +0000 (15:53 -0500)]
allow "fake" filenames in findsource (closes #9284)

This allows findsource() to work in doctests.

A patch from Dirkjan Ochtman.

14 years agoMerge minor doc change from 3.2
Éric Araujo [Sat, 11 Jun 2011 17:56:28 +0000 (19:56 +0200)]
Merge minor doc change from 3.2

14 years agoBranch merge
Éric Araujo [Sat, 11 Jun 2011 17:56:09 +0000 (19:56 +0200)]
Branch merge

14 years agoIssue #12287: In ossaudiodev, check that the device isn't closed in several
Charles-François Natali [Sat, 11 Jun 2011 16:58:24 +0000 (18:58 +0200)]
Issue #12287: In ossaudiodev, check that the device isn't closed in several
methods.

14 years agomerge 3.2
Benjamin Peterson [Sat, 11 Jun 2011 16:34:37 +0000 (11:34 -0500)]
merge 3.2

14 years agomerge 3.1
Benjamin Peterson [Sat, 11 Jun 2011 16:34:31 +0000 (11:34 -0500)]
merge 3.1

14 years agoonto 3.1.5
Benjamin Peterson [Sat, 11 Jun 2011 16:33:01 +0000 (11:33 -0500)]
onto 3.1.5

14 years agomerge 3.2
Benjamin Peterson [Sat, 11 Jun 2011 15:01:35 +0000 (10:01 -0500)]
merge 3.2

14 years agomerge 3.1
Benjamin Peterson [Sat, 11 Jun 2011 15:01:14 +0000 (10:01 -0500)]
merge 3.1

14 years agoAdded tag v3.1.4 for changeset c918ec9f3a76
Benjamin Peterson [Sat, 11 Jun 2011 14:59:17 +0000 (09:59 -0500)]
Added tag v3.1.4 for changeset c918ec9f3a76

14 years agobump to 3.1.4 v3.1.4
Benjamin Peterson [Sat, 11 Jun 2011 14:58:58 +0000 (09:58 -0500)]
bump to 3.1.4

14 years agoAdd missing reST target to one heading in the tutorial
Éric Araujo [Sat, 11 Jun 2011 08:34:19 +0000 (10:34 +0200)]
Add missing reST target to one heading in the tutorial

14 years agoAdjust logging in packaging.util.spawn (related to #11599)
Éric Araujo [Sat, 11 Jun 2011 07:46:07 +0000 (09:46 +0200)]
Adjust logging in packaging.util.spawn (related to #11599)

14 years agoUse correct directive to document one method
Éric Araujo [Sat, 11 Jun 2011 01:27:03 +0000 (03:27 +0200)]
Use correct directive to document one method

14 years agoAllow multiple setup hooks in packaging’s setup.cfg files (#12240).
Éric Araujo [Fri, 10 Jun 2011 22:33:38 +0000 (00:33 +0200)]
Allow multiple setup hooks in packaging’s setup.cfg files (#12240).

Original patch by Erik Bray.

14 years agosetup.cfg: Document that description-file can contain more than one file
Éric Araujo [Fri, 10 Jun 2011 22:21:18 +0000 (00:21 +0200)]
setup.cfg: Document that description-file can contain more than one file

14 years agoFix assorted bugs in packaging.util.cfg_to_args (#11595).
Éric Araujo [Fri, 10 Jun 2011 21:52:26 +0000 (23:52 +0200)]
Fix assorted bugs in packaging.util.cfg_to_args (#11595).

Original patch by Erik Bray.

14 years agoMove useful function to packaging.util.
Éric Araujo [Fri, 10 Jun 2011 21:26:31 +0000 (23:26 +0200)]
Move useful function to packaging.util.

Original patch by Erik Bray as part of #11595, changed by me to improve
readability.

14 years agoUpdated Formatter.formatTime docs to indicate configuration via class attributes.
Vinay Sajip [Fri, 10 Jun 2011 18:05:16 +0000 (19:05 +0100)]
Updated Formatter.formatTime docs to indicate configuration via class attributes.

14 years agoMerged documentation change for Issue #12168.
Vinay Sajip [Fri, 10 Jun 2011 17:54:12 +0000 (18:54 +0100)]
Merged documentation change for Issue #12168.

14 years agoDocumented change for Issue #12168.
Vinay Sajip [Fri, 10 Jun 2011 17:52:50 +0000 (18:52 +0100)]
Documented change for Issue #12168.

14 years agomerge heads
Benjamin Peterson [Fri, 10 Jun 2011 17:32:07 +0000 (12:32 -0500)]
merge heads

14 years agomerge 3.2 (#12009)
Benjamin Peterson [Fri, 10 Jun 2011 17:30:16 +0000 (12:30 -0500)]
merge 3.2 (#12009)

14 years agomerge 3.1 (#12009)
Benjamin Peterson [Fri, 10 Jun 2011 17:29:40 +0000 (12:29 -0500)]
merge 3.1 (#12009)

14 years agoBranch merge
Éric Araujo [Fri, 10 Jun 2011 16:34:45 +0000 (18:34 +0200)]
Branch merge

14 years agofix regression in netrc comment handling (closes #12009)
Benjamin Peterson [Fri, 10 Jun 2011 16:32:52 +0000 (11:32 -0500)]
fix regression in netrc comment handling (closes #12009)

14 years agoDon’t try to install something when running from uninstalled source (#12246).
Éric Araujo [Fri, 10 Jun 2011 16:31:40 +0000 (18:31 +0200)]
Don’t try to install something when running from uninstalled source (#12246).

Original patch by Tshepang Lekhonkhobe.

14 years agonull merge 3.2 (patch already applied to 3.3)
Victor Stinner [Fri, 10 Jun 2011 14:38:38 +0000 (16:38 +0200)]
null merge 3.2 (patch already applied to 3.3)

14 years agonull merge 3.1 (patch already applied to 3.2)
Victor Stinner [Fri, 10 Jun 2011 14:37:41 +0000 (16:37 +0200)]
null merge 3.1 (patch already applied to 3.2)

14 years agoIssue #10801: Fix test_unicode_filenames() of test_zipfile
Victor Stinner [Fri, 10 Jun 2011 14:32:54 +0000 (16:32 +0200)]
Issue #10801: Fix test_unicode_filenames() of test_zipfile

Just try to open files from the ZIP for reading, don't extract them to avoid
UnicodeEncodeError if the filename is not encodable to the filesystem encoding
(e.g. ASCII locale encoding).

14 years agoIssue #8407: skip sigwait() tests if pthread_sigmask() is missing
Victor Stinner [Fri, 10 Jun 2011 12:02:10 +0000 (14:02 +0200)]
Issue #8407: skip sigwait() tests if pthread_sigmask() is missing

The new tests now requires pthread_sigmask(). Skip the test if the function is
missing, e.g. if Python is compiled without threads.

14 years agotest_platform: ignore DeprecationWarning on popen() test
Victor Stinner [Fri, 10 Jun 2011 11:59:59 +0000 (13:59 +0200)]
test_platform: ignore DeprecationWarning on popen() test

14 years agoIssue #8407: write error message on sigwait test failure
Victor Stinner [Fri, 10 Jun 2011 11:53:32 +0000 (13:53 +0200)]
Issue #8407: write error message on sigwait test failure

14 years agoIssue #8407: Make signal.sigwait() tests more reliable
Victor Stinner [Fri, 10 Jun 2011 10:48:13 +0000 (12:48 +0200)]
Issue #8407: Make signal.sigwait() tests more reliable

Block the signal before calling sigwait(). Use os.fork() to ensure that we have
only one thread.

Initial patch written by Charles-François Natali.

14 years agoIssue #9344: Add os.getgrouplist().
Ross Lagerwall [Fri, 10 Jun 2011 05:30:30 +0000 (07:30 +0200)]
Issue #9344: Add os.getgrouplist().

14 years agoClean up extra environment variable after packaging tests.
Éric Araujo [Fri, 10 Jun 2011 02:29:43 +0000 (04:29 +0200)]
Clean up extra environment variable after packaging tests.

packaging.util.check_environ will define HOME and PLAT if they don’t exist; for
some reason, it does not define PLAT when running the tests from a checkout (so
no regrtest warning) but does when running from an installed Python.

Cleaning up the envvar in test_dist fixes the warning on my machine, but I
suspect that a test runner using a different order to run files or running them
in parallel may have PLAT defined in its environment because of another test.
Quite a lot of code ends up calling check_environ; maybe we should just clean
up PLAT in every test.  For now I’m doing this simple fix, we’ll see if we get
bug reports.

14 years agoFix omission in test for packaging install_distinfo command.
Éric Araujo [Fri, 10 Jun 2011 01:53:49 +0000 (03:53 +0200)]
Fix omission in test for packaging install_distinfo command.

The code does not write checksum or file length for .pyc and .pyo in the RECORD
file, in compliance with PEP 376, but the test forgot to take .pyo into
account.  This was not caught because there were no .pyo in the checkout, but
after installing there are .pyo files created by compileall, and the test picks
them up.

14 years agoRemove *.egg-info from Makefile.pre.in (follow-up to d615eb7bce33, #12218)
Éric Araujo [Fri, 10 Jun 2011 01:48:04 +0000 (03:48 +0200)]
Remove *.egg-info from Makefile.pre.in (follow-up to d615eb7bce33, #12218)

14 years agoPackaging: use repr to display projects name (3ebabfbf6fe3 followup)
Éric Araujo [Fri, 10 Jun 2011 01:10:53 +0000 (03:10 +0200)]
Packaging: use repr to display projects name (3ebabfbf6fe3 followup)

14 years agoIssue #8407: signal.sigwait() releases the GIL
Victor Stinner [Thu, 9 Jun 2011 23:39:53 +0000 (01:39 +0200)]
Issue #8407: signal.sigwait() releases the GIL

Initial patch by Charles-François Natali.

14 years agoRemoved a Windows 9x trick used before LoadLibraryExW.
Brian Curtin [Thu, 9 Jun 2011 22:55:54 +0000 (17:55 -0500)]
Removed a Windows 9x trick used before LoadLibraryExW.

Windows 9x has long been unsupported and the result of GetFullPathName
was not even being used in the first place.

14 years agoMerge #10694: zipfile now ignores garbage at the end of a zipfile.
R David Murray [Thu, 9 Jun 2011 19:52:31 +0000 (15:52 -0400)]
Merge #10694: zipfile now ignores garbage at the end of a zipfile.

14 years ago#10694: zipfile now ignores garbage at the end of a zipfile.
R David Murray [Thu, 9 Jun 2011 19:50:51 +0000 (15:50 -0400)]
#10694: zipfile now ignores garbage at the end of a zipfile.

Original fix by 'rep', final patch (with tests) by Xuanji Li.

14 years agomerge #12283: Fixed regression in smtplib quoting of leading dots in DATA.
R David Murray [Thu, 9 Jun 2011 19:21:02 +0000 (15:21 -0400)]
merge #12283: Fixed regression in smtplib quoting of leading dots in DATA.

14 years ago#12283: Fixed regression in smtplib quoting of leading dots in DATA.
R David Murray [Thu, 9 Jun 2011 19:05:57 +0000 (15:05 -0400)]
#12283: Fixed regression in smtplib quoting of leading dots in DATA.

I unfortunately introduced the regression when I refactored the code,
and there were no tests of quoting so it wasn't caught.  Now there
is one.

14 years agoMerged upstream changes.
Vinay Sajip [Thu, 9 Jun 2011 17:43:11 +0000 (18:43 +0100)]
Merged upstream changes.

14 years agoMade time formats in Formatter more configurable.
Vinay Sajip [Thu, 9 Jun 2011 17:42:19 +0000 (18:42 +0100)]
Made time formats in Formatter more configurable.

14 years ago#10424: argument names are now included in the missing argument message
R David Murray [Thu, 9 Jun 2011 16:34:07 +0000 (12:34 -0400)]
#10424: argument names are now included in the missing argument message

Fix and initial test patch by Michele Orrù.