Issue #11168: Remove filename debug variable from PyEval_EvalFrameEx().
It encoded the Unicode filename to UTF-8, but the encoding fails on
undecodable filename (on surrogate characters) which raises an unexpected
UnicodeEncodeError on recursion limit.
Issue #11187: Remove bootstrap code (use ASCII) of
PyUnicode_AsEncodedString(), it was replaced by a better fallback (use
the locale encoding) in PyUnicode_EncodeFSDefault().
Prepare also empty sections in NEWS.
Many tests simply didn't care if they unset a pre-existing trace function. This
made test coverage impossible. This patch fixes various tests to put back any
pre-existing trace function. It also introduces test.support.no_tracing as a
decorator which will temporarily unset the trace function for tests which
simply fail otherwise.
Thanks to Kristian Vlaardingerbroek for helping to find the cause of various
trace function unsets.
Although this patch contains API changes and is rather weighty for an
RC phase, the mailbox module was essentially unusable without the patch
since it would produce UnicodeErrors when handling non-ascii input
at arbitrary and somewhat mysterious places, and any non-trivial amount
of email processing will encounter messages with non-ascii bytes.
The release manager approved the patch application.
The changes allow binary input, and reject non-ASCII string input early
with a useful message instead of failing mysteriously later. Binary
is used internally for reading and writing the mailbox files. StringIO
and Text file input are deprecated.
Initial patch by Victor Stinner, validated and expanded by R. David Murray.
Initial patch by Kelsey Hightower. Approved by Raymond. A test was
non-trivial to write without calling the private function directly, so
we moved that for later.
the system-provided Python. Also, properly guard a new Python 3 only
installer build step so that build-installer.py can stay compatible
with the 2.7 version. (with release manager approval for 3.2rc2)
keyword-only argument. The preceding positional argument was deprecated,
so it made no sense to add filter as a positional argument.
(Patch reviewed by Brian Curtin and Anthony Long.)
(Reviewed by Georg Brandl.)
Also made similar changes to deque.reverse() though this wasn't
strictly necessary (the edge case cannot occur with two pointers
moving to meet in the middle). Making the change in reverse()
was more a matter of future-proofing.