and lifetime issues of dynamically allocated Py_buffer members (#9990)
as well as crashes (#8305, #7433). Many new features have been added
(See whatsnew/3.3), and the documentation has been updated extensively.
The ndarray test object from _testbuffer.c implements all aspects of
PEP-3118, so further development towards the complete implementation
of the PEP can proceed in a test-driven manner.
Thanks to Nick Coghlan, Antoine Pitrou and Pauli Virtanen for review
and many ideas.
- Issue #12834: Fix incorrect results of memoryview.tobytes() for
non-contiguous arrays.
- Issue #5231: Introduce memoryview.cast() method that allows changing
format and shape without making a copy of the underlying memory.
Merged revisions 57392-57619 via svnmerge from
svn+ssh://pythondev@svn.python.org/python/trunk
........
r57395 | georg.brandl | 2007-08-24 19:23:23 +0200 (Fri, 24 Aug 2007) | 2 lines
Bug #1011: fix rfc822.Message.getheader docs.
........
r57397 | georg.brandl | 2007-08-24 19:38:49 +0200 (Fri, 24 Aug 2007) | 2 lines
Patch #1006: port test_winreg to unittest.
........
r57398 | georg.brandl | 2007-08-24 19:46:54 +0200 (Fri, 24 Aug 2007) | 2 lines
Fix#1012: wrong URL to :mod:`site` in install/index.rst.
........
r57399 | georg.brandl | 2007-08-24 20:07:52 +0200 (Fri, 24 Aug 2007) | 2 lines
Patch #1008: port test_signal to unittest.
........
r57400 | georg.brandl | 2007-08-24 20:22:54 +0200 (Fri, 24 Aug 2007) | 2 lines
Port test_frozen to unittest.
........
r57401 | georg.brandl | 2007-08-24 20:27:43 +0200 (Fri, 24 Aug 2007) | 2 lines
Document new utility functions in test_support.
........
r57402 | georg.brandl | 2007-08-24 20:30:06 +0200 (Fri, 24 Aug 2007) | 2 lines
Remove test_rgbimg output file, there is no test_rgbimg.py.
........
r57403 | georg.brandl | 2007-08-24 20:35:27 +0200 (Fri, 24 Aug 2007) | 2 lines
Remove output file for test_ossaudiodev, also properly close the dsp object.
........
r57404 | georg.brandl | 2007-08-24 20:46:27 +0200 (Fri, 24 Aug 2007) | 2 lines
Convert test_linuxaudiodev to unittest. Fix a wrong finally clause in test_ossaudiodev.
........
r57406 | collin.winter | 2007-08-24 21:13:58 +0200 (Fri, 24 Aug 2007) | 1 line
Convert test_pkg to use unittest.
........
r57408 | georg.brandl | 2007-08-24 21:22:34 +0200 (Fri, 24 Aug 2007) | 2 lines
Catch the correct errors.
........
r57409 | georg.brandl | 2007-08-24 21:33:53 +0200 (Fri, 24 Aug 2007) | 2 lines
Port test_class to unittest. Patch #1671298.
........
r57415 | collin.winter | 2007-08-24 23:09:42 +0200 (Fri, 24 Aug 2007) | 1 line
Make test_structmembers pass when run with regrtests's -R flag.
........
r57455 | nick.coghlan | 2007-08-25 06:32:07 +0200 (Sat, 25 Aug 2007) | 1 line
Revert misguided attempt at fixing incompatibility between -m and -i switches (better fix coming soon)
........
r57456 | nick.coghlan | 2007-08-25 06:35:54 +0200 (Sat, 25 Aug 2007) | 1 line
Revert compile.c changes that shouldn't have been included in previous checkin
........
r57461 | nick.coghlan | 2007-08-25 12:50:41 +0200 (Sat, 25 Aug 2007) | 1 line
Fix bug 1764407 - the -i switch now does the right thing when using the -m switch
........
r57464 | guido.van.rossum | 2007-08-25 17:08:43 +0200 (Sat, 25 Aug 2007) | 4 lines
Server-side SSL and certificate validation, by Bill Janssen.
While cleaning up Bill's C style, I may have cleaned up some code
he didn't touch as well (in _ssl.c).
........
r57465 | neal.norwitz | 2007-08-25 18:41:36 +0200 (Sat, 25 Aug 2007) | 3 lines
Try to get this to build with Visual Studio by moving all the variable
declarations to the beginning of a scope.
........
r57466 | neal.norwitz | 2007-08-25 18:54:38 +0200 (Sat, 25 Aug 2007) | 1 line
Fix test so it is skipped properly if there is no SSL support.
........
r57467 | neal.norwitz | 2007-08-25 18:58:09 +0200 (Sat, 25 Aug 2007) | 2 lines
Fix a few more variables to try to get this to compile with Visual Studio.
........
r57473 | neal.norwitz | 2007-08-25 19:25:17 +0200 (Sat, 25 Aug 2007) | 1 line
Try to get this test to pass for systems that do not have SO_REUSEPORT
........
r57482 | gregory.p.smith | 2007-08-26 02:26:00 +0200 (Sun, 26 Aug 2007) | 7 lines
keep setup.py from listing unneeded hash modules (_md5, _sha*) as
missing when they were not built because _hashlib with openssl provided
their functionality instead.
don't build bsddb185 if bsddb was built.
........
r57483 | neal.norwitz | 2007-08-26 03:08:16 +0200 (Sun, 26 Aug 2007) | 1 line
Fix typo in docstring (missing c in reacquire)
........
r57484 | neal.norwitz | 2007-08-26 03:42:03 +0200 (Sun, 26 Aug 2007) | 2 lines
Spell check (also americanify behaviour, it's almost 3 times as common)
........
r57503 | neal.norwitz | 2007-08-26 08:29:57 +0200 (Sun, 26 Aug 2007) | 4 lines
Reap children before the test starts so hopefully SocketServer
won't find any old children left around which causes an exception
in collect_children() and the test to fail.
........
r57510 | neal.norwitz | 2007-08-26 20:50:39 +0200 (Sun, 26 Aug 2007) | 1 line
Fail gracefully if the cert files cannot be created
........
r57513 | guido.van.rossum | 2007-08-26 21:35:09 +0200 (Sun, 26 Aug 2007) | 4 lines
Bill Janssen wrote:
Here's a patch which makes test_ssl a better player in the buildbots
environment. I deep-ended on "try-except-else" clauses.
........
r57518 | neal.norwitz | 2007-08-26 23:40:16 +0200 (Sun, 26 Aug 2007) | 1 line
Get the test passing by commenting out some writes (should they be removed?)
........
r57522 | neal.norwitz | 2007-08-27 00:16:23 +0200 (Mon, 27 Aug 2007) | 3 lines
Catch IOError for when the device file doesn't exist or the user doesn't have
permission to write to the device.
........
r57524 | neal.norwitz | 2007-08-27 00:20:03 +0200 (Mon, 27 Aug 2007) | 5 lines
Another patch from Bill Janssen that:
1) Fixes the bug that two class names are initial-lower-case.
2) Replaces the poll waiting for the server to become ready with
a threading.Event signal.
........
r57536 | neal.norwitz | 2007-08-27 02:58:33 +0200 (Mon, 27 Aug 2007) | 1 line
Stop using string.join (from the module) to ease upgrade to py3k
........
r57537 | neal.norwitz | 2007-08-27 03:03:18 +0200 (Mon, 27 Aug 2007) | 1 line
Make a utility function for handling (printing) an error
........
r57538 | neal.norwitz | 2007-08-27 03:15:33 +0200 (Mon, 27 Aug 2007) | 4 lines
If we can't create a certificate, print a warning, but don't fail the test.
Modified patch from what Bill Janssen sent on python-3000.
........
r57539 | facundo.batista | 2007-08-27 03:15:34 +0200 (Mon, 27 Aug 2007) | 7 lines
Ignore test failures caused by 'resource temporarily unavailable'
exceptions raised in the test server thread, since SimpleXMLRPCServer
does not gracefully handle them. Changed number of requests handled
by tests server thread to one (was 2) because no tests require more
than one request. [GSoC - Alan McIntyre]
........
r57561 | guido.van.rossum | 2007-08-27 19:19:42 +0200 (Mon, 27 Aug 2007) | 8 lines
> Regardless, building a fixed test certificate and checking it in sounds like
> the better option. Then the openssl command in the test code can be turned
> into a comment describing how the test data was pregenerated.
Here's a patch that does that.
Bill
........
r57568 | guido.van.rossum | 2007-08-27 20:42:23 +0200 (Mon, 27 Aug 2007) | 26 lines
> Some of the code sets the error string in this directly before
> returning NULL, and other pieces of the code call PySSL_SetError,
> which creates the error string. I think some of the places which set
> the string directly probably shouldn't; instead, they should call
> PySSL_SetError to cons up the error name directly from the err code.
> However, PySSL_SetError only works after the construction of an ssl
> object, which means it can't be used there... I'll take a longer look
> at it and see if there's a reasonable fix.
Here's a patch which addresses this. It also fixes the indentation in
PySSL_SetError, bringing it into line with PEP 7, fixes a compile warning
about one of the OpenSSL macros, and makes the namespace a bit more
consistent. I've tested it on FC 7 and OS X 10.4.
% ./python ./Lib/test/regrtest.py -R :1: -u all test_ssl
test_ssl
beginning 6 repetitions
123456
......
1 test OK.
[29244 refs]
%
[GvR: slightly edited to enforce 79-char line length, even if it required
violating the style guide.]
........
r57570 | guido.van.rossum | 2007-08-27 21:11:11 +0200 (Mon, 27 Aug 2007) | 2 lines
Patch 10124 by Bill Janssen, docs for the new ssl code.
........
r57574 | guido.van.rossum | 2007-08-27 22:51:00 +0200 (Mon, 27 Aug 2007) | 3 lines
Patch # 1739906 by Christian Heimes -- add reduce to functools (importing
it from __builtin__).
........
r57575 | guido.van.rossum | 2007-08-27 22:52:10 +0200 (Mon, 27 Aug 2007) | 2 lines
News about functools.reduce.
........
r57611 | georg.brandl | 2007-08-28 10:29:08 +0200 (Tue, 28 Aug 2007) | 2 lines
Document rev. 57574.
........
r57612 | sean.reifschneider | 2007-08-28 11:07:54 +0200 (Tue, 28 Aug 2007) | 2 lines
Adding basic imputil documentation.
........
r57614 | georg.brandl | 2007-08-28 12:48:18 +0200 (Tue, 28 Aug 2007) | 2 lines
Fix some glitches.
........
r57616 | lars.gustaebel | 2007-08-28 14:31:09 +0200 (Tue, 28 Aug 2007) | 5 lines
TarFile.__init__() no longer fails if no name argument is passed and
the fileobj argument has no usable name attribute (e.g. StringIO).
(will backport to 2.5)
........
r57619 | thomas.wouters | 2007-08-28 17:28:19 +0200 (Tue, 28 Aug 2007) | 22 lines
Improve extended slicing support in builtin types and classes. Specifically:
- Specialcase extended slices that amount to a shallow copy the same way as
is done for simple slices, in the tuple, string and unicode case.
- Specialcase step-1 extended slices to optimize the common case for all
involved types.
- For lists, allow extended slice assignment of differing lengths as long
as the step is 1. (Previously, 'l[:2:1] = []' failed even though
'l[:2] = []' and 'l[:2:None] = []' do not.)
- Implement extended slicing for buffer, array, structseq, mmap and
UserString.UserString.
- Implement slice-object support (but not non-step-1 slice assignment) for
UserString.MutableString.
- Add tests for all new functionality.
........
*ordering* between objects; there is only a default equality test
(defined by an object being equal to itself only). Read the comment
in object.c. The current implementation never uses a three-way
comparison to compute a rich comparison, but it does use a rich
comparison to compute a three-way comparison. I'm not quite done
ripping out all the calls to PyObject_Compare/Cmp, or replacing
tp_compare implementations with tp_richcompare implementations;
but much of that has happened (to make most unit tests pass).
The following tests still fail, because I need help deciding
or understanding:
test_codeop -- depends on comparing code objects
test_datetime -- need Tim Peters' opinion
test_marshal -- depends on comparing code objects
test_mutants -- need help understanding it
The problem with test_codeop and test_marshal is this: these tests
compare two different code objects and expect them to be equal.
Is that still a feature we'd like to support? I've temporarily
removed the comparison and hash code from code objects, so they
use the default (equality by pointer only) comparison.
For the other two tests, run them to see for yourself.
(There may be more failing test with "-u all".)
A general problem with getting lots of these tests to pass is
the reality that for object types that have a natural total ordering,
implementing __cmp__ is much more convenient than implementing
__eq__, __ne__, __lt__, and so on. Should we go back to allowing
__cmp__ to provide a total ordering? Should we provide some other
way to implement rich comparison with a single method override?
Alex proposed a __key__() method; I've considered a __richcmp__()
method. Or perhaps __cmp__() just shouldn't be killed off...
Nick Coghlan
Stefan Krah
Guido van Rossum
Thomas Wouters