Renamed a parameter in calc_number_widths, for consistency with the same parameter in fill_non_digits.

svn+ssh://pythondev@svn.python.org/sandbox/trunk/2to3/lib2to3

........
  r64863 | brett.cannon | 2008-07-10 19:42:32 -0500 (Thu, 10 Jul 2008) | 1 line

  Add urlparse -> urllib.parse to fix_imports.
........
  r64868 | brett.cannon | 2008-07-10 20:00:10 -0500 (Thu, 10 Jul 2008) | 1 line

  Add robotparser -> urllib.robotparser to fix_imports.
........
  r64870 | brett.cannon | 2008-07-11 00:56:27 -0500 (Fri, 11 Jul 2008) | 6 lines

  Fix the fixers for the new dbm package.

  Had to create a new fixer (fix_imports2) which did fixes in post-order. This
  because ``import anydbm`` was being translated into ``import dbm`` which was
  then subsequently changed into ``import dbm.ndbm``; one transform too many.
........
  r64942 | collin.winter | 2008-07-13 20:19:05 -0500 (Sun, 13 Jul 2008) | 1 line

  Add a comment explaining part of fix_imports.py
........
  r65001 | brett.cannon | 2008-07-16 00:11:12 -0500 (Wed, 16 Jul 2008) | 2 lines

  Remove some extraneous whitespace.
........
  r65002 | brett.cannon | 2008-07-16 00:12:04 -0500 (Wed, 16 Jul 2008) | 4 lines

  Implement a fixer for urllib(2).

  Thanks Nick Edds for the patch.
........
  r65017 | benjamin.peterson | 2008-07-16 11:04:19 -0500 (Wed, 16 Jul 2008) | 1 line

  fix 2to3 in Python 2.6
........
  r65018 | benjamin.peterson | 2008-07-16 11:55:21 -0500 (Wed, 16 Jul 2008) | 1 line

  normalize whitespace
........

Apply Amaury's patch to multiprocessing for issue 3125, removes the copy_reg and replaces it with ForkingPickler.register(), which should resolve the conflict with the global registry/ctypes

computing 2.0**-1074 accurately.  Using ldexp(1.0, -1074) should be
safer.

always set a dlerror() message.

to convert floats to and from hexadecimal strings respectively.

release build) when the format string of a pointer to an incomplete
structure is created.

Issue 2235: __hash__ is once again inherited by default, but inheritance can be blocked explicitly so that collections.Hashable remains meaningful

Added '#' formatting to integers.  This adds the 0b, 0o, or 0x prefix for bin, oct, hex.  There's still one failing case, and I need to finish the docs.  I hope to finish those today.

The fact that this was failing and went unnoticed so long seems like a good argument for being able to enable and disble py3kwarnings through Python.

process rather than both parent and child.

Does anyone actually use fork1()?  It appears to be a Solaris thing
but if Python is built with pthreads on Solaris, fork1() and fork()
should be the same.

Changed statements of the form Py_DECREF(obj), obj = 0 to Py_CLEAR(obj).

in Py_CLEAR().

Fix issue 3221 by emitting a RuntimeWarning instead of raising SystemError when the parent module can't be found during an absolute import (likely due to non-PEP 361 aware code which sets a module level __package__ attribute)

Correct a couple of errors in the updated catch_warning documentation (the Py3k version was fixed before being checked in)

Make test.test_support.catch_warnings more robust as discussed on python-dev. Also add explicit tests for itto test_warnings.