Jarosch.

Thanks to Suman Saha for finding the bug and providing a patch.

than a Python module (e.g. "__init__.py"): don't close the file twice.

PyFile_FromFile() does also close the file if PyString_FromString() failed. It
did already close the file on fill_file_fields() error (e.g. if the file is a
directory).

failing with a UnicodeDecodeError.

ready before having been set up.

multiprocessing.reduction, which may not be available (e.g. if the OS doesn't
support FD passing over Unix domain sockets).

Issue #12973: Fix int.__pow__ overflow checks that invoked undefined behaviour, thereby producing incorrect results on Clang.

find_module() now raises a RuntimeError, instead of ImportError, on an error on
sys.path or sys.meta_path because load_package() and import_submodule() returns
None and clear the exception if a ImportError occurred.

object triggers the garbage collector.

I have tests to add in this file and it’s always nice to start from a
clean base.  I’ve also changed a test that used to write an invalid
config file ('[global]command_packages = etc.' on one line), but the
test passes before and after this change, so either it magically works
or the test is poorly written.  Sigh.

Backport from Python 3.2 (e11b4c945f7e).

sys.platform is now always 'linux2' on Linux (even on Linux 3)

merge from 3.2.  Fix closes Issue11155  - Correct the multiprocessing.Queue.put's arg (replace 'item' with 'obj') in the docs. Patch by Westley Martínez.

tarfile unnecessarily checked the existence of numerical user and group ids on
extraction. If one of them did not exist the respective id of the current user
(i.e. root) was used for the file and ownership information was lost. (Patch
by Sebastien Luttringer)

sdist used to crash with a full traceback dump instead of printing a
nice warning with the faulty line number.

a string.