The embed2.diff patch solves the user's problem by exporting the missing
symbols from the Python core so Python can be embedded in another Cygwin
application (well, at lest vim).

which still supports shared extensions. It takes advantage the latest
Cygwin binutils (i.e., 20030901-1) which can export symbols from
executables:

http://cygwin.com/ml/cygwin-announce/2003-09/msg00002.html

Additionally, it finally lays to rest the following mailing list
subthread:

http://mail.python.org/pipermail/python-list/2002-May/102500.html

I tested the patch under Red Hat Linux 8.0 too

Will backport to 2.3.

Backport candidate.

quotes.  Fixes SF bug #794466, with the essential patch provided by
Stuart D. Gathman.  Specifically,

_parseparam(), _get_params_preserve(): Use the parsing function that
takes quotes into account, as given (essentially) in the bug report's
test program.

Backport candidate.

option is not given.  If dbfile isn't given and can't be retrieved
from the optionsdb, just initialize it to the first element in
RGB_TXT.

Backport candidate.

figure out what the code is doing

(Contributed by Jeff Epler.)

(Contributed by Alexander Belopolsky.)

(Contributed by Alexander Belopolsky.)

Doctest would crash when encountering unbound methods:
  class A:
    def f(self): pass

  class C(A):
    g = A.f

(Contributed by Walter Dörwald.)

* Convert test_slice.py to unittest format
* Expand the test coverage.

(Contributed by George Yoshida.)

(Contributed by George Yoshida.)

(Contributed by Andrew Gaul.)

Fixes a minor leak.

(Contributed by Andrew Gaul.)

Fixes a minor leak.

(Contributed by George Yoshida.)

(Contributed by George Yoshida.)

Clarified the meaning of non-overlapping in patterns with zero length
matches.

(Contributed by Christos Georgiou.)

Reference the symbol definitions in the stat module.

For smaller datasets, it is not always true the increasing the compression
level always results in better compression.  Removed the test which made
this invalid assumption.

(Revised from the original patch contributed by Michal Pasternak.)

Also, make a couple minor fixups elsewhere.

(Contributed by Walter Dörwald).

* Convert three test modules to unittest format.
* Expanded coverage in test_structseq.py.
* Raymond added a new test in test_sets.py

When the indents were set to longer than the width and long word breaking
was enabled, an infinite loop would result because the inner loop did not
assure that at least one character was stripped off on every pass.

Avoid Armin Rigo's dastardly exercise in re-entrancy.

Rework previous fix slightly; the &0x20 test seems useless, and the isprint() check mustn't prevent the meta-bit check at the end