Try new upstream typing

[Tom-Willemsen]

We'll need this after bluesky/ophyd-async#594 gets merged and released upstream - draft PR at the moment.

Current issues:

Fails pyright

(.venv) c:\Instrument\dev\ibex_bluesky_core>pyright
c:\Instrument\dev\ibex_bluesky_core\tests\devices\test_block.py
  c:\Instrument\dev\ibex_bluesky_core\tests\devices\test_block.py:288:31 - error: Argument of type "float" cannot be assigned to parameter "args" of type "Tuple[Movable | NamedMovable | Any, ...]" in function "mv"
    "float" is not assignable to "Tuple[Movable | NamedMovable | Any, ...]" (reportArgumentType)
c:\Instrument\dev\ibex_bluesky_core\tests\devices\test_dae.py
  c:\Instrument\dev\ibex_bluesky_core\tests\devices\test_dae.py:943:15 - error: Argument of type "DaeSettings" cannot be assigned to parameter "args" of type "Tuple[Movable | NamedMovable | Any, ...]" in function "mv"
    "DaeSettings" is not assignable to "Tuple[Movable | NamedMovable | Any, ...]" (reportArgumentType)
  c:\Instrument\dev\ibex_bluesky_core\tests\devices\test_dae.py:943:33 - error: Argument of type "DaeSettingsData" cannot be assigned to parameter "args" of type "Tuple[Movable | NamedMovable | Any, ...]" in function "mv"
    "DaeSettingsData" is not assignable to "Tuple[Movable | NamedMovable | Any, ...]" (reportArgumentType)
3 errors, 0 warnings, 0 informations

My assumption is that's an upstream bug that will be fixed before they release the new typing.

trigger with no timeout seems broken? (fixed)

await dae.controls.end_run.trigger(wait=True, timeout=None) causes the following error, despite the fact that it should be waiting with no timeout:

Traceback (most recent call last):
  File "C:\Instrument\dev\ibex_bluesky_core\src\ibex_bluesky_core\devices\simpledae\__init__.py", line 82, in trigger
    await self.controller.stop_counting(self)
  File "C:\Instrument\dev\ibex_bluesky_core\src\ibex_bluesky_core\devices\simpledae\controllers.py", line 130, in stop_counting
    await _end_or_abort_run(dae, self._save_run)
  File "C:\Instrument\dev\ibex_bluesky_core\src\ibex_bluesky_core\devices\simpledae\controllers.py", line 22, in _end_or_abort_run
    await dae.controls.end_run.trigger(wait=True, timeout=None)
  File "c:\Instrument\dev\ibex_bluesky_core\.venv\Lib\site-packages\ophyd_async\core\_signal.py", line 262, in trigger
    await _wait_for(self._connector.backend.put(None, wait=wait), timeout, source)
  File "c:\Instrument\dev\ibex_bluesky_core\.venv\Lib\site-packages\ophyd_async\core\_signal.py", line 37, in _wait_for
    return await asyncio.wait_for(coro, timeout)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "c:\INSTRU~1\Apps\Python3\Lib\asyncio\tasks.py", line 452, in wait_for
    return await fut
           ^^^^^^^^^
  File "c:\Instrument\dev\ibex_bluesky_core\.venv\Lib\site-packages\ophyd_async\epics\signal\_aioca.py", line 283, in put
    await caput(
  File "c:\Instrument\dev\ibex_bluesky_core\.venv\Lib\site-packages\aioca\_catools.py", line 126, in call_wrapper
    return await throw_wrapper(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "c:\Instrument\dev\ibex_bluesky_core\.venv\Lib\site-packages\aioca\_catools.py", line 108, in throw_wrapper
    return await awaitable
           ^^^^^^^^^^^^^^^
  File "c:\Instrument\dev\ibex_bluesky_core\.venv\Lib\site-packages\aioca\_catools.py", line 156, in ca_timeout
    raise CANothing(name, cadef.ECA_TIMEOUT) from e
aioca._catools.CANothing: TE:NDW2922:DAE:ENDRUN: User specified timeout on IO operation expired

[coretl]

Fails pyright

That is bluesky/bluesky#1809 which @dperl-dls is working on

trigger with no timeout seems broken?

Oops, I moved the timeout code into Signal, but forgot to turn it off in aioca. Just pushed a commit that fixes.

[Tom-Willemsen]

Thanks @coretl - can confirm our tests are happier r.e. timeouts after bluesky/ophyd-async@1f288ee