ssl.match_hostname() has been simplified and no longer depends on re and
ipaddress module for wildcard and IP addresses. Error reporting for invalid
wildcards has been improved.
Signed-off-by: Christian Heimes <christian@python.org>

Adjust tests and user modules to match.

Previously, the ssl module stored international domain names (IDNs)
as U-labels. This is problematic for a number of reasons -- for
example, it made it impossible for users to use a different version
of IDNA than the one built into Python.

After this change, we always convert to A-labels as soon as possible,
and use them for all internal processing. In particular, server_hostname
attribute is now an A-label, and on the server side there's a new
sni_callback that receives the SNI servername as an A-label rather than
a U-label.

Changed the template to be inside a comment tag.

Exhaustively test dataclass hashing when no hash= value is provided. This is in anticipation of changing how non-default hashing is handled. (#5834)

`whilst` and `while` are both english words, `whilst` is not as commonly used.
This can be confusing to readers whose primary language is not english.

Make test.support.temp_cwd() fork-safe. The context manager test.support.temp_cwd() no longer removes the temporary directory when executing in a process other than the parent it entered from.
If a forked child exits the context manager it won't do the cleanup.

Co-authored-by: Mark Shannon <mark@hotpy.org>
Co-authored-by: Antoine Pitrou <antoine@python.org>

bpo-32556: nt._getfinalpathname, nt._getvolumepathname and nt._getdiskusage now correctly convert from bytes. (GH-5761)

dump is similar to print but less flexible.  lastopenbracketpos is now always initialized in _study2, as was stmt_bracketing, so the class settings are not needed.  get_last_open_bracket_pos is never called.

There are no code changes other than comments and docstrings.

Py_ADDRESS_IN_RANGE was renamed address_in_range in 3.6
(commit 3924f937).

It's bad form to pin to an old version of TLS. ssl.SSLContext has the right
protocol default, so let's not pass anyway.

The CPython runtime assumes that there is a one-to-one relationship (for a given interpreter) between PyThreadState and OS threads. Sending and receiving on a channel in the same interpreter was causing crashes because of this (specifically due to a check in PyThreadState_Swap()). The solution is to not switch threads if the interpreter is the same.

Initialize self._ssnd_chunk so that aifc.Error is raised as intended,
not AttributeError.

Dropped the part that says: "For objects that do not provide sequence protocol".

os.dup2() tests for dup3() system call availability at runtime,
but doesn't remember the result across calls, repeating
the test on each call with inheritable=False.

Since the caller of os.dup2() is expected to hold the GIL,
fix this by making the variable holding the test result static.

This function expects the destination buffer size to be given
in wide characters, not bytes.

This adds C versions of methods used by ABCMeta that
improve performance of various ABC operations.

Also, use PYTHON_FOR_REGEN for clinic and blake2s_impl.c rather than PYTHON_FOR_BUILD, and update .travis.yml to make use of the change.