Mention frame.f_trace in sys.settrace docs, as well as the fact you still
need to call `sys.settrace` to enable the tracing machinery before setting
`frame.f_trace` will have any effect.

* Check intersection of two sets explicitly

Comparing ``len(a) > ``len(a - b)`` is essentially looking for an
intersection between the two sets. If set ``b`` does not intersect ``a``
then ``len(a - b)`` will be equal to ``len(a)``. This logic is more
clearly expressed as ``a & b``.

* Change while/pop to a for-loop

Copying the list, then repeatedly popping the first element was
unnecessarily slow. I also cleaned up a couple other inefficiencies.
There's no need to unpack a tuple, then re-pack and append it. The list
can be created with the first element instead of empty. Secondly, the
``endswith`` method returns a bool, so there's no need for an if-
statement to set ``encoding`` to True or False.

* Use set.intersection to check for intersections

``a.intersection(b)`` method is more clear of purpose than ``not
a.isdisjoint(b)`` and avoids an unnecessary set construction that ``a &
set(b)`` performs.

* Use not isdisjoint instead of intersection

While it reads slightly worse, the isdisjoint method will stop when it
finds a counterexample and returns a bool, rather than looping over the
entire iterable and constructing a new set.

* Make dict and weakref offsets opaque for C heap types

* Add news

https://bugs.python.org/issue38155

Signed-off-by: Jason Plurad <pluradj@us.ibm.com>

The function return type is void, not int.

dump_traceback_later() and cancel_dump_traceback_later() functions of
the faulthandler module are always available since Python 3.7.

It is a known and tracked bug: disable the test until it's fixed.

Add _PyRuntimeState.preinitializing field: set to 1 while
Py_PreInitialize() is running.

_PyRuntimeState: rename also pre_initialized field to preinitialized.

_PyObject_Dump() now dumps the object address for freed objects and
objects with ob_type=NULL.

When a Python thread has no frame, _Py_DumpTraceback() and
_Py_DumpTracebackThreads() now write "<no Python frame>", rather than
writing nothing.

This includes such names as "cls", "self", "typename", "_typename",
"fields" and "_fields".
Passing positional arguments by keyword is deprecated.

A little change on first paragraph of python tutorial to be more clearly



https://bugs.python.org/issue37904



Automerge-Triggered-By: @ericvsmith

Even when the helper is not started yet.

This behavior follows conventional generator one.
There is no reason for `async_generator_athrow` to handle `gen.throw()` differently.



https://bugs.python.org/issue38013

In the format string for assert_called the evaluation order is incorrect and hence for mock's without name, 'None' is printed whereas it should be 'mock' like for other messages. The error message is ("Expected '%s' to have been called." % self._mock_name or 'mock').

When using multiprocesss (-jN), the main process now uses a timeout
of 60 seconds instead of the double of the --timeout value. The
buildbot server stops a job which does not produce any output in 1200
seconds.

See https://bugs.python.org/issue38192 .



https://bugs.python.org/issue38192

One happens when EditorWindow.close is called twice.
Printing a traceback, when IDLE is run from a terminal,
is useless and annoying.

It no longer tries to create or access .idlerc or any files within.
Users must run IDLE to discover problems with saving settings.

This PR replaces the old note mentioning that `typing` is a provisional module with a new one mentioning types are not enforced at runtime. I am not sure if there was any official announcement about making `typing` non-provisional, but _de-facto_ no new features were added during Python 3.7, and no backwards incompatible changes were made except for few small things that were considered bugs.

``OPENSSL_VERSION_1_1`` was never defined in ``_hashopenssl.c``. 


https://bugs.python.org/issue33936