]> granicus.if.org Git - python/log
python
16 years agoTrivial typo.
Eric Smith [Wed, 19 Mar 2008 12:15:10 +0000 (12:15 +0000)]
Trivial typo.

16 years agoUse test.test_support.captured_stdout instead of a custom contextmanager.
Eric Smith [Wed, 19 Mar 2008 12:09:55 +0000 (12:09 +0000)]
Use test.test_support.captured_stdout instead of a custom contextmanager.
Thanks Nick Coghlan.

16 years agoLets have another try at getting the Windows buildbots in a consistent state before...
Trent Nelson [Wed, 19 Mar 2008 09:23:08 +0000 (09:23 +0000)]
Lets have another try at getting the Windows buildbots in a consistent state before rebuilding using the new process.

16 years agoFix the tcl-8.4.18.1 path and make sure we cd into the right directory when building...
Trent Nelson [Wed, 19 Mar 2008 08:06:03 +0000 (08:06 +0000)]
Fix the tcl-8.4.18.1 path and make sure we cd into the right directory when building tcl/tk.

16 years agoAnother one.
Georg Brandl [Wed, 19 Mar 2008 07:57:57 +0000 (07:57 +0000)]
Another one.

16 years agoRemove footnote from versionchanged as it upsets LaTeX.
Georg Brandl [Wed, 19 Mar 2008 07:56:40 +0000 (07:56 +0000)]
Remove footnote from versionchanged as it upsets LaTeX.

16 years agoRemove extraneous apostrophe and semi-colon from AdditionalIncludeDirectories.
Trent Nelson [Wed, 19 Mar 2008 07:56:39 +0000 (07:56 +0000)]
Remove extraneous apostrophe and semi-colon from AdditionalIncludeDirectories.

16 years agoRefine the Visual Studio 2008 build solution in order to improve how we deal with...
Trent Nelson [Wed, 19 Mar 2008 07:45:19 +0000 (07:45 +0000)]
Refine the Visual Studio 2008 build solution in order to improve how we deal with external components, as well as fixing outstanding issues with Windows x64 build support.  Introduce two new .vcproj files, _bsddb44.vcproj and sqlite3.vcproj, which replace the previous pre-link event scripts for _bsddb and _sqlite3 respectively.  The new project files inherit from our property files as if they were any other Python module.  This has numerous benefits.  First, the components get built with exactly the same compiler flags and settings as the rest of Python.  Second, it makes it much easier to debug problems in the external components when they're part of the build system.  Third, they'll benefit from profile guided optimisation in the release builds, just like the rest of Python core.

I've also introduced a slightly new pattern for managing externals in subversion.  New components get checked in as <name>-<version>.x, where <version> matches the exact vendor version string.  After the initial import of the external component, the .x is tagged as .0 (i.e. tcl-8.4.18.x -> tcl-8.4.18.0).  Some components may not need any tweaking, whereas there are others that might (tcl/tk fall into this bucket).  In that case, the relevant modifications are made to the .x branch, which will be subsequently tagged as .1 (and then n+1 going forward) when they build successfully and all tests pass.  Buildbots will be converted to rely on these explicit tags only, which makes it easy for us to switch them over to a new version as and when required.  (Simple change to external(-amd64).bat: if we've bumped tcl to 8.4.18.1, change the .bat to rmdir 8.4.18.0 if it exists and check out a new .1 copy.)

16 years agoIssue2290: Support x64 Windows builds that live in pcbuild/amd64. Without it, sysuti...
Trent Nelson [Wed, 19 Mar 2008 06:28:24 +0000 (06:28 +0000)]
Issue2290: Support x64 Windows builds that live in pcbuild/amd64.  Without it, sysutils._python_build() returns the wrong directory, which causes the test_get_config_h_filename method in Lib/distutils/tests/test_sysconfig.py to fail.

16 years agoImport relimport using a relative import.
Martin v. Löwis [Wed, 19 Mar 2008 06:00:28 +0000 (06:00 +0000)]
Import relimport using a relative import.

16 years agoMerged revisions 61598-61599,61601 via svnmerge from
Martin v. Löwis [Wed, 19 Mar 2008 05:22:42 +0000 (05:22 +0000)]
Merged revisions 61598-61599,61601 via svnmerge from
svn+ssh://pythondev@svn.python.org/sandbox/trunk/2to3/lib2to3

........
  r61598 | david.wolever | 2008-03-18 23:58:33 -0500 (Di, 18 Mär 2008) | 1 line

  Added fixer for zip, and refactored a bit of code in the process.  Closing #2171.
........
  r61599 | david.wolever | 2008-03-19 00:04:26 -0500 (Mi, 19 Mär 2008) | 3 lines

  Removed a bunch of duplicate code -- it's in util now.
........
  r61601 | martin.v.loewis | 2008-03-19 00:21:12 -0500 (Mi, 19 Mär 2008) | 2 lines

  Fix whitespace.
........

16 years agoInitialized merge tracking via "svnmerge" with revisions "1-61595" from
Martin v. Löwis [Wed, 19 Mar 2008 04:58:04 +0000 (04:58 +0000)]
Initialized merge tracking via "svnmerge" with revisions "1-61595" from
svn+ssh://pythondev@svn.python.org/sandbox/trunk/2to3/lib2to3

16 years agoImport lib2to3.
Martin v. Löwis [Wed, 19 Mar 2008 04:43:46 +0000 (04:43 +0000)]
Import lib2to3.

16 years agoIssue #2400: Allow relative imports to "import *".
Martin v. Löwis [Wed, 19 Mar 2008 04:39:13 +0000 (04:39 +0000)]
Issue #2400: Allow relative imports to "import *".

16 years agoFix compiler warning.
Raymond Hettinger [Wed, 19 Mar 2008 03:56:59 +0000 (03:56 +0000)]
Fix compiler warning.

16 years agoFix the struct module DeprecationWarnings that zipfile was triggering by
Gregory P. Smith [Wed, 19 Mar 2008 03:14:41 +0000 (03:14 +0000)]
Fix the struct module DeprecationWarnings that zipfile was triggering by
removing all use of signed struct values.

test_zipfile and test_zipfile64 pass.  no more warnings.

16 years agoAnother test for __future__ print_function.
Eric Smith [Wed, 19 Mar 2008 03:13:34 +0000 (03:13 +0000)]
Another test for __future__ print_function.

16 years agoAdded zip, map, filter to future_bultins (#2171)
David Wolever [Wed, 19 Mar 2008 02:35:45 +0000 (02:35 +0000)]
Added zip, map, filter to future_bultins (#2171)

16 years agoAdded my name to ACKS
David Wolever [Wed, 19 Mar 2008 02:26:57 +0000 (02:26 +0000)]
Added my name to ACKS

16 years agoFixed compiler module so __future__ print_function is compilable.
Eric Smith [Wed, 19 Mar 2008 02:11:30 +0000 (02:11 +0000)]
Fixed compiler module so __future__ print_function is compilable.

16 years agoUse zlib's crc32 routine instead of binascii when available. zlib's is faster
Gregory P. Smith [Wed, 19 Mar 2008 01:46:10 +0000 (01:46 +0000)]
Use zlib's crc32 routine instead of binascii when available.  zlib's is faster
when compiled properly optimized and about the same speed otherwise.

16 years agoMention that crc32 and adler32 are available in a different module (zlib).
Gregory P. Smith [Wed, 19 Mar 2008 01:38:35 +0000 (01:38 +0000)]
Mention that crc32 and adler32 are available in a different module (zlib).
Some people look for them in hashlib.

16 years agoAdd Jeff Rush
Andrew M. Kuchling [Wed, 19 Mar 2008 01:05:35 +0000 (01:05 +0000)]
Add Jeff Rush

16 years agoBackport of the print function, using a __future__ import.
Eric Smith [Tue, 18 Mar 2008 23:45:49 +0000 (23:45 +0000)]
Backport of the print function, using a __future__ import.
This work is substantially Anthony Baxter's, from issue
1633807.  I just freshened it, made a few minor tweaks,
and added the test cases.  I also created issue 2412,
which is to check for 2to3's behavior with the print
function.  I also added myself to ACKS.

16 years agoIssue: 2354: Add 3K warning for the cmp argument to list.sort() and sorted().
Raymond Hettinger [Tue, 18 Mar 2008 23:33:08 +0000 (23:33 +0000)]
Issue: 2354: Add 3K warning for the cmp argument to list.sort() and sorted().

16 years agoSpeed-up isinstance() for one easy case.
Raymond Hettinger [Tue, 18 Mar 2008 23:22:29 +0000 (23:22 +0000)]
Speed-up isinstance() for one easy case.

16 years agoAdd a test to make sure zlib.crc32 and binascii.crc32 return the same thing.
Gregory P. Smith [Tue, 18 Mar 2008 22:27:41 +0000 (22:27 +0000)]
Add a test to make sure zlib.crc32 and binascii.crc32 return the same thing.
Fix a buglet in binascii.crc32, the second optional argument could previously
have a signedness mismatch with the C variable its going into.

16 years agoAdd py3k warnings for code and method inequality comparisons. This should resolve...
Steven Bethard [Tue, 18 Mar 2008 22:08:20 +0000 (22:08 +0000)]
Add py3k warnings for code and method inequality comparisons. This should resolve issue 2373. The codeobject.c and methodobject.c changes are both just backports of the Python 3 code.

16 years agoHave regrtest skip test_py3kwarn when the -3 flag is missing.
Steven Bethard [Tue, 18 Mar 2008 21:30:13 +0000 (21:30 +0000)]
Have regrtest skip test_py3kwarn when the -3 flag is missing.

16 years agoAdded a warning when -3 is enabled and None is passed to filter as the first argument.
David Wolever [Tue, 18 Mar 2008 21:20:25 +0000 (21:20 +0000)]
Added a warning when -3 is enabled and None is passed to filter as the first argument.

16 years agoIgnore BIG5HKSCS-2004.TXT which is downloaded as part of a test.
Brett Cannon [Tue, 18 Mar 2008 21:12:42 +0000 (21:12 +0000)]
Ignore BIG5HKSCS-2004.TXT which is downloaded as part of a test.

16 years agonews entry for the chown fix
Gregory P. Smith [Tue, 18 Mar 2008 20:40:01 +0000 (20:40 +0000)]
news entry for the chown fix

16 years agoImport the test properly. This is especially important for py3k.
Neal Norwitz [Tue, 18 Mar 2008 20:30:38 +0000 (20:30 +0000)]
Import the test properly.  This is especially important for py3k.

16 years agoFix test_atexit so that it still passes when -3 is supplied. (It was catching the...
Steven Bethard [Tue, 18 Mar 2008 19:59:14 +0000 (19:59 +0000)]
Fix test_atexit so that it still passes when -3 is supplied. (It was catching the warning messages on stdio from using the reload() function.)

16 years agoFix chown on 64-bit linux. It needed to take a long (64-bit on 64bit linux) as
Gregory P. Smith [Tue, 18 Mar 2008 19:05:32 +0000 (19:05 +0000)]
Fix chown on 64-bit linux.  It needed to take a long (64-bit on 64bit linux) as
uid and gid input to accept values >=2**31 as valid while still accepting
negative numbers to pass -1 to chown for "no change".

Fixes issue1747858.

This should be backported to release25-maint.

16 years ago_have_soundcard() is a bad check for winsound.Beep, since you can have a soundcard...
Steven Bethard [Tue, 18 Mar 2008 19:04:32 +0000 (19:04 +0000)]
_have_soundcard() is a bad check for winsound.Beep, since you can have a soundcard but have the beep driver disabled. This revision basically disables the beep tests by wrapping them in a try/except. The Right Way To Do It is to come up with a _have_enabled_beep_driver() and use that.

16 years agocell_compare needs to return -2 instead of NULL.
Steven Bethard [Tue, 18 Mar 2008 19:03:50 +0000 (19:03 +0000)]
cell_compare needs to return -2 instead of NULL.

16 years agoGet regrtest working when re-running tests
Neal Norwitz [Tue, 18 Mar 2008 17:58:02 +0000 (17:58 +0000)]
Get regrtest working when re-running tests

16 years agoAdd py3k warnings for object, type, cell and dict comparisons. This should resolve...
Steven Bethard [Tue, 18 Mar 2008 17:26:10 +0000 (17:26 +0000)]
Add py3k warnings for object, type, cell and dict comparisons. This should resolve issue2342 and partly resolve issue2373.

16 years agoAdd Tools/scripts/patchcheck.py. Invoked from ``make check``, the script does
Brett Cannon [Tue, 18 Mar 2008 17:25:13 +0000 (17:25 +0000)]
Add Tools/scripts/patchcheck.py. Invoked from ``make check``, the script does
some verification:

- Runs reindent.py on all .py files.
- Checks if any changes in Doc exist.
- Whether Misc/ACKS was changed.
- Whether Misc/NEWS was changed.

The hope is that ``make check`` can become a command anybody can run to get
reminders about what all the requisite steps needed to create a proper
patch/checkin.

16 years agoIssue 1577: shutil.move() where destination is a directory was doing a
Sean Reifscheider [Tue, 18 Mar 2008 17:24:12 +0000 (17:24 +0000)]
Issue 1577: shutil.move() where destination is a directory was doing a
copy, now it is doing a os.rename() if it's on the same file-system.

16 years agoCast the arguments to PyString_AsStringAndSize() to silence compiler warnings
Brett Cannon [Tue, 18 Mar 2008 16:47:51 +0000 (16:47 +0000)]
Cast the arguments to PyString_AsStringAndSize() to silence compiler warnings
on OS X.

16 years agoUse test_support.unlink instead of os.unlink in tearDown(). (Seems to fix an occasio...
Steven Bethard [Tue, 18 Mar 2008 16:00:19 +0000 (16:00 +0000)]
Use test_support.unlink instead of os.unlink in tearDown().  (Seems to fix an occasional failure in Windows Vista.)

16 years agoFix test_errno to only check for error numbers that are defined by Standard C.
Brett Cannon [Tue, 18 Mar 2008 15:52:00 +0000 (15:52 +0000)]
Fix test_errno to only check for error numbers that are defined by Standard C.

16 years agoRemove all traces of HAVE_STRERROR.
Brett Cannon [Tue, 18 Mar 2008 15:35:58 +0000 (15:35 +0000)]
Remove all traces of HAVE_STRERROR.

The removal of strerror.c led to the function check being removed from
configure.in.

16 years agoInclude <alloca.h> on Solaris, see issue #1506.
Thomas Heller [Tue, 18 Mar 2008 15:03:17 +0000 (15:03 +0000)]
Include <alloca.h> on Solaris, see issue #1506.

It would probably be better to have a configure test for that, but
this is outside of my configure expertise.

16 years agoNote that the stderr output of the test is intentional.
Martin v. Löwis [Tue, 18 Mar 2008 13:16:05 +0000 (13:16 +0000)]
Note that the stderr output of the test is intentional.

16 years agoAdd WSA errors.
Martin v. Löwis [Tue, 18 Mar 2008 13:05:03 +0000 (13:05 +0000)]
Add WSA errors.

16 years agoAdd more Linux error codes.
Martin v. Löwis [Tue, 18 Mar 2008 12:45:37 +0000 (12:45 +0000)]
Add more Linux error codes.

16 years agonorwitz-amd64 (gentoo) has EREMOTEIO.
Martin v. Löwis [Tue, 18 Mar 2008 12:20:15 +0000 (12:20 +0000)]
norwitz-amd64 (gentoo) has EREMOTEIO.

16 years agoThe behaviour of winsound.Beep() seems to differ between different versions of Window...
Trent Nelson [Tue, 18 Mar 2008 07:32:47 +0000 (07:32 +0000)]
The behaviour of winsound.Beep() seems to differ between different versions of Windows when there's either:
    a) no sound card entirely
    b) legacy beep driver has been disabled
    c) the legacy beep driver has been uninstalled
Sometimes RuntimeErrors are raised, sometimes they're not.  If _have_soundcard() returns False, don't expect winsound.Beep() to raise a RuntimeError, as this clearly isn't the case, as demonstrated by the various Win32 XP buildbots.

16 years agoIssue 2286: bump up the stack size of the 64-bit debug python_d.exe to 2100000. ...
Trent Nelson [Tue, 18 Mar 2008 07:02:12 +0000 (07:02 +0000)]
Issue 2286: bump up the stack size of the 64-bit debug python_d.exe to 2100000.  The default value of 200000 causes a stack overflow at 1965 iterations of r_object() in marshal.c, 35 iterations before the 2000 limit enforced by MAX_MARSHAL_STACK_DEPTH.

16 years agoAdd some info to the failure messages
Neal Norwitz [Tue, 18 Mar 2008 06:03:46 +0000 (06:03 +0000)]
Add some info to the failure messages

16 years agoAdd a -S/--slow flag to regrtest to have it print the 10 slowest tests with
Jeffrey Yasskin [Tue, 18 Mar 2008 05:45:40 +0000 (05:45 +0000)]
Add a -S/--slow flag to regrtest to have it print the 10 slowest tests with
their times.

16 years agoImprove the error message for a test that failed on the S-390 Debian buildbot.
Brett Cannon [Tue, 18 Mar 2008 05:43:04 +0000 (05:43 +0000)]
Improve the error message for a test that failed on the S-390 Debian buildbot.

16 years agoTry increasing the timeout to reduce the flakiness of this test.
Neal Norwitz [Tue, 18 Mar 2008 05:20:29 +0000 (05:20 +0000)]
Try increasing the timeout to reduce the flakiness of this test.

16 years agoSpeed up test_dict by about 10x by only checking selected dict literal sizes,
Jeffrey Yasskin [Tue, 18 Mar 2008 05:12:41 +0000 (05:12 +0000)]
Speed up test_dict by about 10x by only checking selected dict literal sizes,
instead of every integer from 0 to 400. Exhaustive testing wastes time without
providing enough more assurance that the code is correct.

16 years agoSpeed test_thread up from 51.328s to 0.081s by reducing its sleep times. We
Jeffrey Yasskin [Tue, 18 Mar 2008 04:56:06 +0000 (04:56 +0000)]
Speed test_thread up from 51.328s to 0.081s by reducing its sleep times. We
still sleep at all to make it likely that all threads are active at the same
time.

16 years agoIssue 2332: add new attribute names for instance method objects
Neal Norwitz [Tue, 18 Mar 2008 04:46:00 +0000 (04:46 +0000)]
Issue 2332: add new attribute names for instance method objects

16 years agoFinish backporting new buffer API to Python 2.6. Left to do: memoryview object and...
Travis E. Oliphant [Tue, 18 Mar 2008 04:44:57 +0000 (04:44 +0000)]
Finish backporting new buffer API to Python 2.6.  Left to do: memoryview object and structmodule.  But, these need to be finished in Python 3.0 first.  No objects support the new buffer API in Python 2.6 as of yet, and except for the memoryview object, I don't think they will.

16 years ago- Issue 2379: Raise a Py3K warning for __getitem__ or __getslice__ on
Guido van Rossum [Tue, 18 Mar 2008 04:42:22 +0000 (04:42 +0000)]
- Issue 2379: Raise a Py3K warning for __getitem__ or __getslice__ on
  exception instances.

16 years agoBlock the "socket.ssl() is deprecated" warning from test_socket_ssl.
Jeffrey Yasskin [Tue, 18 Mar 2008 04:29:35 +0000 (04:29 +0000)]
Block the "socket.ssl() is deprecated" warning from test_socket_ssl.

16 years agoIssue #2341: Add a Py3k warning when raising an exception that doesn't
Guido van Rossum [Tue, 18 Mar 2008 04:26:48 +0000 (04:26 +0000)]
Issue #2341: Add a Py3k warning when raising an exception that doesn't
derive from BaseException.

16 years agoThe output directory for tests that compare against stdout is now gone!
Brett Cannon [Tue, 18 Mar 2008 04:16:06 +0000 (04:16 +0000)]
The output directory for tests that compare against stdout is now gone!

16 years agoRemove our implementation of memmove() and strerror(); both are in the C89
Brett Cannon [Tue, 18 Mar 2008 04:09:00 +0000 (04:09 +0000)]
Remove our implementation of memmove() and strerror(); both are in the C89
standard library.

16 years agotest_errno was a no-op test; now it actually tests things and uses unittest.
Brett Cannon [Tue, 18 Mar 2008 03:46:22 +0000 (03:46 +0000)]
test_errno was a no-op test; now it actually tests things and uses unittest.

16 years agoClean up the Py3k warnings for non-BaseException-subclasses a bit. We
Guido van Rossum [Tue, 18 Mar 2008 03:15:05 +0000 (03:15 +0000)]
Clean up the Py3k warnings for non-BaseException-subclasses a bit.  We
now don't warn for some corner cases that deserve a warning, rather
than warning double or incorrectly for some other corner cases.

16 years ago- Issue #2371: Add a Py3k warning when catching an exception that
Guido van Rossum [Tue, 18 Mar 2008 02:49:46 +0000 (02:49 +0000)]
- Issue #2371: Add a Py3k warning when catching an exception that
  doesn't derive from BaseException.

16 years agoMove test_extcall to doctest.
Brett Cannon [Tue, 18 Mar 2008 01:58:56 +0000 (01:58 +0000)]
Move test_extcall to doctest.

16 years agoConvert test_dummy_threading and test_dbm to unittest.
Brett Cannon [Tue, 18 Mar 2008 01:50:25 +0000 (01:50 +0000)]
Convert test_dummy_threading and test_dbm to unittest.

16 years agoFix build on platforms that don't have intptr_t. Patch by Joseph Armbruster.
Jeffrey Yasskin [Tue, 18 Mar 2008 01:09:59 +0000 (01:09 +0000)]
Fix build on platforms that don't have intptr_t. Patch by Joseph Armbruster.

16 years agoConvert test_strftime, test_getargs, and test_pep247 to use unittest.
Brett Cannon [Tue, 18 Mar 2008 01:00:07 +0000 (01:00 +0000)]
Convert test_strftime, test_getargs, and test_pep247 to use unittest.

16 years agoFix the IOError message text when opening a file with an invalid filename.
Gregory P. Smith [Tue, 18 Mar 2008 00:20:01 +0000 (00:20 +0000)]
Fix the IOError message text when opening a file with an invalid filename.
Error reported by Ilan Schnell.

16 years agoAdd David Wolever.
Martin v. Löwis [Mon, 17 Mar 2008 21:55:30 +0000 (21:55 +0000)]
Add David Wolever.

16 years agoIssue 2321: reduce memory usage (increase the memory that is returned
Neal Norwitz [Mon, 17 Mar 2008 20:22:43 +0000 (20:22 +0000)]
Issue 2321: reduce memory usage (increase the memory that is returned
to the system) by using pymalloc for the data of unicode objects.

Will backport.

16 years agoAdded PEP 3127 support to tokenize (with tests); added PEP 3127 to NEWS.
Eric Smith [Mon, 17 Mar 2008 19:43:40 +0000 (19:43 +0000)]
Added PEP 3127 support to tokenize (with tests); added PEP 3127 to NEWS.

16 years agoDocument unicode.isnumeric() and unicode.isdecimal() (issue2326)
Steven Bethard [Mon, 17 Mar 2008 19:33:11 +0000 (19:33 +0000)]
Document unicode.isnumeric() and unicode.isdecimal() (issue2326)

16 years agoUse a buffer large enough to ensure we don't overrun, even if the value
Neal Norwitz [Mon, 17 Mar 2008 19:02:45 +0000 (19:02 +0000)]
Use a buffer large enough to ensure we don't overrun, even if the value
is outside the range we expect.

16 years agoForce zlib.crc32 and zlib.adler32 to return a signed integer on all platforms
Gregory P. Smith [Mon, 17 Mar 2008 18:48:05 +0000 (18:48 +0000)]
Force zlib.crc32 and zlib.adler32 to return a signed integer on all platforms
regardless of the native sizeof(long) used in the integer object.

This somewhat odd behavior of returning a signed is maintained in 2.x for
compatibility reasons of always returning an integer rather than a long object.

Fixes Issue1202 for Python 2.6

16 years agoAdd necessary headers to back-port new buffer protocol to Python 2.6
Travis E. Oliphant [Mon, 17 Mar 2008 17:36:12 +0000 (17:36 +0000)]
Add necessary headers to back-port new buffer protocol to Python 2.6

16 years agoFinished backporting PEP 3127, Integer Literal Support and Syntax.
Eric Smith [Mon, 17 Mar 2008 17:32:20 +0000 (17:32 +0000)]
Finished backporting PEP 3127, Integer Literal Support and Syntax.

Added 0b and 0o literals to tokenizer.
Modified PyOS_strtoul to support 0b and 0o inputs.
Modified PyLong_FromString to support guessing 0b and 0o inputs.
Renamed test_hexoct.py to test_int_literal.py and added binary tests.
Added upper and lower case 0b, 0O, and 0X tests to test_int_literal.py

16 years agoAdd Trent Nelson.
Martin v. Löwis [Mon, 17 Mar 2008 16:31:57 +0000 (16:31 +0000)]
Add Trent Nelson.

16 years agoMake isinstance(OldstyleClass, NewstyleClass) return False instead of raising
Jeffrey Yasskin [Mon, 17 Mar 2008 16:31:21 +0000 (16:31 +0000)]
Make isinstance(OldstyleClass, NewstyleClass) return False instead of raising
an exception. Issue reported by Joseph Armbruster.

16 years agoAllow Gnu gcc's to build python on OSX by removing -Wno-long-double,
Jeffrey Yasskin [Mon, 17 Mar 2008 14:40:53 +0000 (14:40 +0000)]
Allow Gnu gcc's to build python on OSX by removing -Wno-long-double,
-no-cpp-precomp, and -mno-fused-madd from configure.
 * r22183 added -no-cpp-precomp, which
   http://gcc.gnu.org/ml/gcc/2005-12/msg00368.html claims hasn't been
   needed since gcc-3.1.
 * r25607 added -Wno-long-double to avoid a warning in
   Include/objimpl.h (issue 525481). The long double is still there,
   but OSX 10.4's gcc no longer warns about it.
 * r33666 fixed issue 775892 on OSX 10.3 by adding -mno-fused-madd,
   which changed the sign of some float 0s. Tim Peters said it wasn't
   a real issue anyway, and it no longer causes test failures.
Fixes issue #1779871.

16 years agoReformated lines > 79 chars.
Eric Smith [Mon, 17 Mar 2008 12:14:29 +0000 (12:14 +0000)]
Reformated lines > 79 chars.
Deleted unused macro ISXDIGIT.

16 years agoIssue 2264: empty float presentation type needs to have at least one digit past the...
Eric Smith [Mon, 17 Mar 2008 11:01:01 +0000 (11:01 +0000)]
Issue 2264: empty float presentation type needs to have at least one digit past the decimal point.

Added "Z" format_char to PyOS_ascii_formatd to support empty float presentation type.
Renamed buf_size in PyOS_ascii_formatd to more accurately reflect it's meaning.
Modified format.__float__ to use the new "Z" format as the default.
Added test cases.

16 years agoAdded a footnote to each pointing out that for XML output if an encoding
Mark Summerfield [Mon, 17 Mar 2008 08:28:15 +0000 (08:28 +0000)]
Added a footnote to each pointing out that for XML output if an encoding
string is given it should conform to the appropriate XML standards---for
example, "UTF-8" is okay, but "UTF8" is not.

16 years agoClarified documentation on use of shutdown().
Vinay Sajip [Sun, 16 Mar 2008 21:35:58 +0000 (21:35 +0000)]
Clarified documentation on use of shutdown().

16 years ago#2299: typos in newtypes.rst.
Georg Brandl [Sun, 16 Mar 2008 08:00:19 +0000 (08:00 +0000)]
#2299: typos in newtypes.rst.

16 years agoUpdate docs to reflect removal of Exact/Inexact
Raymond Hettinger [Sun, 16 Mar 2008 05:20:42 +0000 (05:20 +0000)]
Update docs to reflect removal of Exact/Inexact

16 years agoHandle memory allocation failure. Found by Adam Olsen
Neal Norwitz [Sat, 15 Mar 2008 22:36:01 +0000 (22:36 +0000)]
Handle memory allocation failure.  Found by Adam Olsen

16 years agoAdd a warning for code like:
Neal Norwitz [Sat, 15 Mar 2008 22:03:18 +0000 (22:03 +0000)]
Add a warning for code like:
  assert (0, 'message')

An empty tuple does not create a warning.  While questionable usage:
  assert (), 'message'

should not display a warning.  Tested manually.
The warning message could be improved.  Feel free to update it.

16 years agoZap one more use of Exact/Inexact.
Raymond Hettinger [Sat, 15 Mar 2008 20:37:50 +0000 (20:37 +0000)]
Zap one more use of Exact/Inexact.

16 years agoRemoved Exact/Inexact after discussion with Yasskin.
Raymond Hettinger [Sat, 15 Mar 2008 20:02:04 +0000 (20:02 +0000)]
Removed Exact/Inexact after discussion with Yasskin.

Unlike Scheme where exactness is implemented as taints, the Python
implementation associated exactness with data types.  This created
inheritance issues (making an exact subclass of floats would result
in the subclass having both an explicit Exact registration and an
inherited Inexact registration).  This was a problem for the
decimal module which was designed to span both exact and inexact
arithmetic.  There was also a question of use cases and no examples
were found where ABCs for exactness could be used to improve code.
One other issue was having separate tags for both the affirmative
and negative cases.  This is at odds with the approach taken
elsewhere in the Python (i.e. we don't have an ABC both Hashable
and Unhashable).

16 years ago.
Skip Montanaro [Sat, 15 Mar 2008 16:07:11 +0000 (16:07 +0000)]
.

16 years agoadd %f format to datetime - issue 1158
Skip Montanaro [Sat, 15 Mar 2008 16:04:45 +0000 (16:04 +0000)]
add %f format to datetime - issue 1158

16 years agonote that fork and forkpty raise OSError on failure
Skip Montanaro [Sat, 15 Mar 2008 02:32:49 +0000 (02:32 +0000)]
note that fork and forkpty raise OSError on failure

17 years agoFix lots of broken links in the docs, found by Sphinx' external link checker.
Georg Brandl [Sat, 15 Mar 2008 00:20:19 +0000 (00:20 +0000)]
Fix lots of broken links in the docs, found by Sphinx' external link checker.

17 years agoRemove obsolete paragraph. #2288.
Georg Brandl [Fri, 14 Mar 2008 23:10:34 +0000 (23:10 +0000)]
Remove obsolete paragraph. #2288.

17 years agoRun debug version, cd to PCbuild.
Martin v. Löwis [Fri, 14 Mar 2008 21:19:28 +0000 (21:19 +0000)]
Run debug version, cd to PCbuild.