* Modules/bz2module.c
  (BZ2FileObject): Now the structure includes a pointer to a file object,
   instead of "inheriting" one. Also, some members were copied from the
   PyFileObject structure to avoid dealing with the internals of that
   structure from outside fileobject.c.

  (Util_GetLine,Util_DropReadAhead,Util_ReadAhead,Util_ReadAheadGetLineSkip,
   BZ2File_write,BZ2File_writelines,BZ2File_init,BZ2File_dealloc,
   BZ2Comp_dealloc,BZ2Decomp_dealloc):
   	These functions were adapted to the change above.

  (BZ2File_seek,BZ2File_close): Use PyObject_CallMethod instead of
   getting the function attribute locally.

  (BZ2File_notsup): Removed, since it's not necessary anymore to overload
   truncate(), and readinto() with dummy functions.

  (BZ2File_methods): Added xreadlines() as an alias to BZ2File_getiter,
   and removed truncate() and readinto().

  (BZ2File_get_newlines,BZ2File_get_closed,BZ2File_get_mode,BZ2File_get_name,
   BZ2File_getset):
   	Implemented getters for "newlines", "mode", and "name".

  (BZ2File_members): Implemented "softspace" member.

  (BZ2File_init): Reworked to create a file instance instead of initializing
   itself as a file subclass. Also, pass "name" object untouched to the
   file constructor, and use PyObject_CallFunction instead of building the
   argument tuple locally.

  (BZ2File_Type): Set tp_new to PyType_GenericNew, tp_members to
   BZ2File_members, and tp_getset to BZ2File_getset.

  (initbz2): Do not set BZ2File_Type.tp_base nor BZ2File_Type.tp_new.


* Doc/lib/libbz2.tex
  Do not mention that BZ2File inherits from the file type.

descr_check(); it wasn't useful.  Change the type argument of the
various _get() methods to PyObject * because the call signature of
tp_descr_get doesn't guarantee its type.

in the light of weird args, and (c) not to expect None (which is now
changed to NULL by slot_tp_descr_get()).

select_select() didn't set an exception in the SELECT_USES_HEAP case when
malloc() returned NULL.

now, it was only called once, and its existence merely obfuscates the
control flow.

getting done.  Since this isn't yet implemented in cPickle, the
new tests are in TempAbstractPickleTests (which cPickle doesn't
run).

  expected.
- Fixed resizing of multi-column lists, somewhat.

when Python code calls a descriptor's __get__ method.  It should
translate None to NULL in both argument positions, and insist that at
least one of the argument positions is not NULL after this
transformation.

"make lib" takes a while, doesn't it?

[ 680429 ] __module__ broken for extension classes

was used.  This simplifies some logic in copy_reg.py (used by
pickling).  It also broke a test, but this was rewritten to test the
new feature. :-)

The attached patch enables the _iconv_codec
module to build cleanly under Cygwin.

The attached patch enables the array module
to build cleanly under Cygwin again.

Add \refs to GC section that were XXXed out.

Object/unicodeobject.c::unicode_count().

instead of PyTuple_GetItem, so an overwritten __getitem__
in a tuple subclass works. SF bug #665835.

- Allow for "manual:" pseudo-scheme in downloadURL to signal that
  the download should be done manually.

split into OPT and BASECFLAGS (Makefile.pre.in rev. 1.108), because
now there are essential CFLAGS in BASECFLAGS.

object is not a real str or unicode but an instance
of a subclass, construct the output via looping
over __getitem__. This guarantees that the result
is the same for function==None and function==lambda x:x

This doesn't happen for tuples, because filtertuple()
uses PyTuple_GetItem().

(This was discussed on SF bug #665835).

(preInstall, at least).

my MAL. Fixed. (Btw. eval() still doesn't take buffers, but that was so
even before my patch.)

Incorporated nnorwitz's comment re. Py__USING_UNICODE.

When subclassing from an int but not overriding __new__,
long values were not converted properly.  Try to convert
longs into an int.