svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71847 | thomas.heller | 2009-04-24 22:00:41 +0200 (Fr, 24 Apr 2009) | 2 lines

  Issue 5041: ctypes unwilling to allow pickling wide character.
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71842 | thomas.heller | 2009-04-24 20:10:46 +0200 (Fr, 24 Apr 2009) | 3 lines

  Issue #5161: wrong paths for ctypes cleanup when Python is built in a
  directory other than the source directory.
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71837 | mark.dickinson | 2009-04-24 17:34:14 +0100 (Fri, 24 Apr 2009) | 4 lines

  Issue #5593: Use more robust test for double-rounding in test_fsum.
  While we're at it, use new unittest.skipUnless decorator to
  implement skipping for that test.
........
  r71838 | mark.dickinson | 2009-04-24 17:37:22 +0100 (Fri, 24 Apr 2009) | 2 lines

  Remove unnecessary double negative
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71832 | mark.dickinson | 2009-04-24 14:56:07 +0100 (Fri, 24 Apr 2009) | 3 lines

  Issue #5812: The two-argument form of the Fraction constructor
  now accepts arbitrary Rational instances.
........

........
  r71827 | mark.dickinson | 2009-04-24 14:14:07 +0100 (Fri, 24 Apr 2009) | 2 lines

  Fix missing 'return NULL'
........

match for 2.7 and 3.1, and that 3.1 continues to
accept complex('j') and complex('4-j')

........
  r71824 | mark.dickinson | 2009-04-24 13:46:53 +0100 (Fri, 24 Apr 2009) | 7 lines

  Issue #5816:
   - simplify parsing and printing of complex numbers
   - make complex(repr(z)) round-tripping work for complex
     numbers involving nans, infs, or negative zeros
   - don't accept some of the stranger complex strings
     that were previously allowed---e.g., complex('1..1j')
........

nans and infs are no longer given special treatment; as a result,
repr(complex(z)) recovers z for any complex number z.

........
  r71808 | mark.dickinson | 2009-04-22 19:15:25 +0100 (Wed, 22 Apr 2009) | 2 lines

  Issue #5812: make Fraction('1e-6') valid.  Backport of r71806.
........

accepts all strings accepted by the float and Decimal constructors,
with the exception of strings representing NaNs or infinities.

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71802 | eric.smith | 2009-04-22 12:20:47 -0400 (Wed, 22 Apr 2009) | 1 line

  Fixed issue 5782: formatting with commas didn't work if no specifier type code was given.
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71799 | nick.coghlan | 2009-04-23 01:26:04 +1000 (Thu, 23 Apr 2009) | 1 line

  Issue 5354: Change API for import_fresh_module() to better support test_warnings use case (also fixes some bugs in the original implementation)
........

........
  r71796 | eric.smith | 2009-04-22 09:29:05 -0400 (Wed, 22 Apr 2009) | 20 lines

  Backport of some of the work in r71665 to trunk. This reworks much of
  int, long, and float __format__(), and it keeps their implementation
  in sync with py3k.

  Also added PyOS_double_to_string. This is the "fallback" version
  that's also available in trunk, and should be kept in sync with that
  code. I'll add an issue to document PyOS_double_to_string in the C
  API.

  There are many internal cleanups. Externally visible changes include:

  - Implement PEP 378, Format Specifier for Thousands Separator, for
    floats, ints, and longs.

  - Issue #5515: 'n' formatting for ints, longs, and floats handles
    leading zero formatting poorly.

  - Issue #5772: For float.__format__, don't add a trailing ".0" if
    we're using no type code and we have an exponent.
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71785 | r.david.murray | 2009-04-21 09:06:04 -0400 (Tue, 21 Apr 2009) | 4 lines

  Restore skips of posix and pty tests on Windows by calling the
  test_support.import_module on the appropriate modules
  before any other imports.
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71788 | eric.smith | 2009-04-21 20:47:00 -0400 (Tue, 21 Apr 2009) | 1 line

  Documentation for issue 5237, auto-numbered format fields. Contributed by Terry J. Reedy.
........

........
  r71776 | mark.dickinson | 2009-04-20 22:41:04 +0100 (Mon, 20 Apr 2009) | 2 lines

  Nit: integer division should use //, not /
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71772 | mark.dickinson | 2009-04-20 22:13:33 +0100 (Mon, 20 Apr 2009) | 5 lines

  Issue #3166: Make long -> float (and int -> float) conversions
  correctly rounded, using round-half-to-even.  This ensures that the
  value of float(n) doesn't depend on whether we're using 15-bit digits
  or 30-bit digits for Python longs.
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71766 | tarek.ziade | 2009-04-20 16:29:42 +0200 (Mon, 20 Apr 2009) | 1 line

  adding a NEWS note for #5795 (previously checked via the buildbot)
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71759 | tarek.ziade | 2009-04-20 12:33:47 +0200 (Mon, 20 Apr 2009) | 1 line

  making BuildWinInstTestCase silent in case bdist_wininst is not run under win32
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71758 | tarek.ziade | 2009-04-20 09:53:55 +0200 (Mon, 20 Apr 2009) | 1 line

  #5795 sysconfig._config_vars was shadowed in tearDown
........

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71734 | benjamin.peterson | 2009-04-18 17:15:26 -0500 (Sat, 18 Apr 2009) | 1 line

  many more types to initialize (I had to expose some of them)
........
  r71738 | benjamin.peterson | 2009-04-18 21:32:42 -0500 (Sat, 18 Apr 2009) | 1 line

  initialize weakref some weakref types
........
  r71739 | benjamin.peterson | 2009-04-18 21:40:43 -0500 (Sat, 18 Apr 2009) | 1 line

  make errors consistent
........

with the _PY_STRUCT_FLOAT_COERCE constant.  Simplify tests accordingly,
and reenable (now-fixed) broken tests.

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71750 | mark.dickinson | 2009-04-19 18:10:47 +0100 (Sun, 19 Apr 2009) | 3 lines

  Automatic conversion of floats to integers for struct.pack integer codes
  is deprecated.  Use an explicit int() instead.
........

float repr on platforms that don't currently support it, and
reorganize the defines slightly to make adding support easier.

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r71743 | ronald.oussoren | 2009-04-19 12:38:20 +0200 (Sun, 19 Apr 2009) | 2 lines

  Fix for issue5657.
........

Patch by Brian Quinlan.

........
  r71725 | benjamin.peterson | 2009-04-18 15:25:25 -0500 (Sat, 18 Apr 2009) | 1 line

  initalize -> initialize
........