This patch by Antoine Pitrou optimizes the bytecode for conditional branches by
merging the following "POP_TOP" instruction into the conditional jump.  For
example, the list comprehension "[x for x in l if not x]" produced the
following bytecode:

  1           0 BUILD_LIST               0
              3 LOAD_FAST                0 (.0)
        >>    6 FOR_ITER                23 (to 32)
              9 STORE_FAST               1 (x)
             12 LOAD_FAST                1 (x)
             15 JUMP_IF_TRUE            10 (to 28)
             18 POP_TOP
             19 LOAD_FAST                1 (x)
             22 LIST_APPEND              2
             25 JUMP_ABSOLUTE            6
        >>   28 POP_TOP
             29 JUMP_ABSOLUTE            6
        >>   32 RETURN_VALUE

but after the patch it produces the following bytecode:

  1           0 BUILD_LIST               0
              3 LOAD_FAST                0 (.0)
        >>    6 FOR_ITER                18 (to 27)
              9 STORE_FAST               1 (x)
             12 LOAD_FAST                1 (x)
             15 POP_JUMP_IF_TRUE         6
             18 LOAD_FAST                1 (x)
             21 LIST_APPEND              2
             24 JUMP_ABSOLUTE            6
        >>   27 RETURN_VALUE

Notice that not only the code is shorter, but the conditional jump
(POP_JUMP_IF_TRUE) jumps right to the start of the loop instead of going through
the JUMP_ABSOLUTE at the end. "continue" statements are helped
similarly.

Furthermore, the old jump opcodes (JUMP_IF_FALSE, JUMP_IF_TRUE) have been
replaced by two new opcodes:
- JUMP_IF_TRUE_OR_POP, which jumps if true and pops otherwise
- JUMP_IF_FALSE_OR_POP, which jumps if false and pops otherwise

be called, not just parents.  Add a comparison to getattr()
which has the same search order but also includes the type
itself.

Needs to be backported to 2.6, 2.7, and 3.0.

........
  r69870 | antoine.pitrou | 2009-02-22 11:25:52 -0600 (Sun, 22 Feb 2009) | 3 lines

  Try to make sense of the test_site buildbot failures
........
  r69871 | antoine.pitrou | 2009-02-22 12:20:46 -0600 (Sun, 22 Feb 2009) | 3 lines

  Revert debugging statements, culprit is possibly test_distutils (see #5316)
........
  r69908 | raymond.hettinger | 2009-02-23 13:32:55 -0600 (Mon, 23 Feb 2009) | 1 line

  Update itertools recipes to use next().
........

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

........
  r69902 | tarek.ziade | 2009-02-23 13:41:29 +0100 (Mon, 23 Feb 2009) | 1 line

  more test coverage
........

........
  r69896 | georg.brandl | 2009-02-23 11:24:23 +0100 (Mo, 23 Feb 2009) | 1 line

  #5348: format() converts all kinds of values.
........

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

........
  r69889 | matthias.klose | 2009-02-23 00:14:26 +0100 (Mo, 23 Feb 2009) | 2 lines

  - Link the shared python library with $(MODLIBS).
........

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

........
  r69881 | tarek.ziade | 2009-02-22 21:15:41 +0100 (Sun, 22 Feb 2009) | 1 line

  Removing unused __main__ sections
........

........
  r69878 | tarek.ziade | 2009-02-22 21:11:46 +0100 (Sun, 22 Feb 2009) | 1 line

  removing map and lambda usage, so the test is similar to py3k's branch one
........

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

........
  r69874 | tarek.ziade | 2009-02-22 20:58:12 +0100 (Sun, 22 Feb 2009) | 1 line

  moved distutils.text_file tests into a real unittest class
........

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

........
  r69861 | tarek.ziade | 2009-02-22 01:07:45 +0100 (Sun, 22 Feb 2009) | 1 line

  using versionchanged instead of versionadded for distutils doc on sdist default files
........

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

........
  r69855 | benjamin.peterson | 2009-02-21 17:09:33 -0600 (Sat, 21 Feb 2009) | 1 line

  fix compiler warnings
........

........
  r69837 | raymond.hettinger | 2009-02-21 01:17:22 -0600 (Sat, 21 Feb 2009) | 4 lines

  Fix keyword arguments for itertools.count().
  Step arg without a start arg was ignored.
........
  r69838 | raymond.hettinger | 2009-02-21 02:58:42 -0600 (Sat, 21 Feb 2009) | 1 line

  Speedup and simplify negative counter using count's new step argument.
........

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

........
  r69846 | mark.dickinson | 2009-02-21 20:27:01 +0000 (Sat, 21 Feb 2009) | 2 lines

  Issue #5341: Fix a variety of spelling errors.
........

are source-only and source/bytecode loaders.

protocol API.

+ Ditch using arguments to super().
+ Ditch subclassing from object directly.
+ Move directory check out of chaining path hook to file path hook/finder.
+ Rename some classes to better reflect they are finders, not importers.

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

........
  r69816 | mark.dickinson | 2009-02-20 20:42:53 +0000 (Fri, 20 Feb 2009) | 2 lines

  Issue #5295: Typos in turtle.py
........

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

........
  r69806 | eric.smith | 2009-02-20 09:02:36 -0500 (Fri, 20 Feb 2009) | 1 line

  Issue #5247: Improve error message when unknown format codes are used when using str.format() with str, int, and float arguments.
........

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

........
  r69415 | benjamin.peterson | 2009-02-07 13:08:22 -0600 (Sat, 07 Feb 2009) | 1 line

  make destinsrc private
........
  r69591 | martin.v.loewis | 2009-02-13 14:26:16 -0600 (Fri, 13 Feb 2009) | 1 line

  Update Tix build procedure.
........
  r69593 | martin.v.loewis | 2009-02-13 14:51:48 -0600 (Fri, 13 Feb 2009) | 1 line

  Add optional code signing after merging.
........

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

........
  r69769 | georg.brandl | 2009-02-19 02:30:06 -0600 (Thu, 19 Feb 2009) | 1 line

  #5310, #3558: fix operator precedence table.
........
  r69776 | georg.brandl | 2009-02-19 10:34:51 -0600 (Thu, 19 Feb 2009) | 2 lines

  #5317: update IronPython URL.
........

........
  r69268 | kristjan.jonsson | 2009-02-04 04:05:25 -0600 (Wed, 04 Feb 2009) | 1 line

  issue 4804:  Provide checks for the format string of strftime, and for the "mode" string of fopen on Windows.  These strings are user provided from python and so we can avoid invoking the C runtime invalid parameter handler by first checking that they are valid.
........
  r69516 | hirokazu.yamamoto | 2009-02-10 22:13:06 -0600 (Tue, 10 Feb 2009) | 2 lines

  Issue #5204: Define _PyVerify_fd on VC6 to make
  test_fdopen (test_os.py) pass.
........
  r69757 | raymond.hettinger | 2009-02-18 23:34:35 -0600 (Wed, 18 Feb 2009) | 1 line

  Add some cross-references to the docs.  Simplify the python code equivalent for izip().  Supply an optional argument for the nth() recipe.
........
  r69761 | raymond.hettinger | 2009-02-18 23:51:41 -0600 (Wed, 18 Feb 2009) | 1 line

  Add an example for math.fsum() and elaborate on the accurary note.
........
  r69765 | raymond.hettinger | 2009-02-19 00:55:03 -0600 (Thu, 19 Feb 2009) | 1 line

  Add links to helpful external resources.
........
  r69770 | raymond.hettinger | 2009-02-19 03:50:24 -0600 (Thu, 19 Feb 2009) | 1 line

  Inline coefficients in gamma().  Add reflection formula.  Add comments.
........
  r69772 | vinay.sajip | 2009-02-19 06:31:32 -0600 (Thu, 19 Feb 2009) | 1 line

  #5287: Add exception handling around findCaller() call to help out IronPython.
........
  r69777 | jeroen.ruigrok | 2009-02-19 12:52:21 -0600 (Thu, 19 Feb 2009) | 3 lines

  Since we recommend one module per import line, reflect this also in the
  documentation.
........
  r69795 | benjamin.peterson | 2009-02-19 21:31:23 -0600 (Thu, 19 Feb 2009) | 1 line

  revert r69777 since all the experts agree that extra import lines distract from the code
........