Log message:
py-anyio: updated to 4.8.0

4.8.0

- Added **experimental** support for running functions in subinterpreters on Python
  3.13 and later
- Added support for the ``copy()``, ``copy_into()``, ``move()`` and ``move_into()``
  methods in ``anyio.Path``, available in Python 3.14
- Changed ``TaskGroup`` on asyncio to always spawn tasks non-eagerly, even if using a
  task factory created via ``asyncio.create_eager_task_factory()``, to preserve \ 
expected
  Trio-like task scheduling semantics (PR by @agronholm and @graingert)
- Configure ``SO_RCVBUF``, ``SO_SNDBUF`` and ``TCP_NODELAY`` on the selector
  thread waker socket pair (this should improve the performance of \ 
``wait_readable()``)
  and ``wait_writable()`` when using the ``ProactorEventLoop``
- Fixed ``AssertionError`` when using ``nest-asyncio``
- Fixed return type annotation of various context managers' ``__exit__`` method

Log message:
py-anyio: updated to 4.7.0

4.7.0

- Updated ``TaskGroup`` to work with asyncio's eager task factories
- Added the ``wait_readable()`` and ``wait_writable()`` functions which will accept
  an object with a ``.fileno()`` method or an integer handle, and deprecated
  their now obsolete versions (``wait_socket_readable()`` and
  ``wait_socket_writable()``)
- Changed ``EventAdapter`` (an ``Event`` with no bound async backend) to allow \ 
``set()``
  to work even before an async backend is bound to it
- Added support for ``wait_readable()`` and ``wait_writable()`` on \ 
``ProactorEventLoop``
  (used on asyncio + Windows by default)
- Fixed a misleading ``ValueError`` in the context of DNS failures
- Fixed the return type annotations of ``readinto()`` and ``readinto1()`` \ 
methods in the
  ``anyio.AsyncFile`` class
- Fixed ``TaskInfo.has_pending_cancellation()`` on asyncio returning false \ 
positives in
  cleanup code on Python >= 3.11
- Fixed cancelled cancel scopes on asyncio calling ``asyncio.Task.uncancel`` when
  propagating a ``CancelledError`` on exit to a cancelled parent scope

Log message:
py-*: remove unused tool dependency

py-setuptools includes the py-wheel functionality nowadays

Log message:
py-anyio: update to 4.6.2.post1.

- Fixed regression caused by (`#807 \ 
<https://github.com/agronholm/anyio/pull/807>`_)
  that prevented the use of parametrized async fixtures

Log message:
py-anyio: updated to 4.6.2

4.6.2

- Fixed regression caused by
  that prevented the use of parametrized async fixtures

4.6.1

This release contains all the changes from both v4.5.1 and v4.6.0, plus:

- Fixed TaskGroup and CancelScope producing cyclic references in tracebacks
  when raising exceptions

Log message:
py-anyio: update to 4.6.0.

Dropped support for Python 3.8 (as #698 cannot be resolved without cancel \ 
message support)
Fixed 100% CPU use on asyncio while waiting for an exiting task group to finish \ 
while said task group is within a cancelled cancel scope (#695)
Fixed cancel scopes on asyncio not propagating CancelledError on exit when the \ 
enclosing cancel scope has been effectively cancelled (#698)
Fixed asyncio task groups not yielding control to the event loop at exit if \ 
there were no child tasks to wait on
Fixed inconsistent task uncancellation with asyncio cancel scopes belonging to a \ 
task group when said task group has child tasks running

Log message:
py-anyio: updated to 4.4.0

**4.4.0**

- Added the ``BlockingPortalProvider`` class to aid with constructing synchronous
  counterparts to asynchronous interfaces that would otherwise require multiple \ 
blocking
  portals
- Added ``__slots__`` to ``AsyncResource`` so that child classes can use \ 
``__slots__``
- Added the ``TaskInfo.has_pending_cancellation()`` method
- Fixed erroneous ``RuntimeError: called 'started' twice on the same task status``
  when cancelling a task in a TaskGroup created with the ``start()`` method before
  the first checkpoint is reached after calling ``task_status.started()``
- Fixed two bugs with ``TaskGroup.start()`` on asyncio:

  * Fixed erroneous ``RuntimeError: called 'started' twice on the same task status``
    when cancelling a task in a TaskGroup created with the ``start()`` method before
    the first checkpoint is reached after calling ``task_status.started()``
  * Fixed the entire task group being cancelled if a ``TaskGroup.start()`` call gets
    cancelled
- Fixed a race condition that caused crashes when multiple event loops of the same
  backend were running in separate threads and simultaneously attempted to use \ 
AnyIO for
  their first time
- Fixed cancellation delivery on asyncio incrementing the wrong cancel scope's
  cancellation counter when cascading a cancel operation to a child scope, thus \ 
failing
  to uncancel the host task
- Fixed erroneous ``TypedAttributeLookupError`` if a typed attribute getter raises
  ``KeyError``
- Fixed the asyncio backend not respecting the ``PYTHONASYNCIODEBUG`` environment
  variable when setting the ``debug`` flag in ``anyio.run()``
- Fixed ``SocketStream.receive()`` not detecting EOF on asyncio if there is also \ 
data in
  the read buffer
- Fixed ``MemoryObjectStream`` dropping an item if the item is delivered to a \ 
recipient
  that is waiting to receive an item but has a cancellation pending
- Emit a ``ResourceWarning`` for ``MemoryObjectReceiveStream`` and
  ``MemoryObjectSendStream`` that were garbage collected without being closed
- Fixed ``MemoryObjectSendStream.send()`` not raising ``BrokenResourceError`` \ 
when the
  last corresponding ``MemoryObjectReceiveStream`` is closed while waiting to send a
  falsey item

Log message:
py-anyio: updated to 4.3.0

4.3.0

Added support for the Python 3.12 walk_up keyword argument in \ 
anyio.Path.relative_to()

Fixed passing total_tokens to anyio.CapacityLimiter() as a keyword argument not \ 
working on the trio backend

Fixed Process.aclose() not performing the minimum level of necessary cleanup \ 
when cancelled. Previously:

Cancellation of Process.aclose() could leak an orphan process
Cancellation of run_process() could very briefly leak an orphan process.
Cancellation of Process.aclose() or run_process() on Trio could leave standard \ 
streams unclosed

Fixed Process.stdin.aclose(), Process.stdout.aclose(), and \ 
Process.stderr.aclose() not including a checkpoint on asyncio

Fixed documentation on how to provide your own typed attributes