If detector misses a trigger then it waits forever without timeout

[huw-dls]

If we make our triggers too close together then the detector will only produce half the images required, then stall waiting for another trigger. This is meant to be caught by the timeout observing the number of frames written, but the following code catches that and waits for the detector to be idle:

ophyd-async/src/ophyd_async/core/_detector.py

Lines 366 to 387 in e2de04d

	try:
	async for index in indices_written:
	yield WatcherUpdate(
	name=self.name,
	current=index,
	initial=self._initial_frame,
	target=self._events_to_complete,
	unit="",
	precision=0,
	time_elapsed=time.monotonic() - self._fly_start
	if self._fly_start
	else None,
	)
	if index >= self._events_to_complete:
	break
	finally:
	await indices_written.aclose()
	if self._completable_exposures >= trigger_info.total_number_of_exposures:
	self._completable_exposures = 0
	self._events_to_complete = 0
	self._number_of_events_iter = None
	await self._controller.wait_for_idle()

This will never happen as it is waiting for a trigger.

This needs investigating as to why the finally block was added, and another way of doing this found.

Acceptance Criteria

• Raises TimeoutError if HDF writer stalls even if detector is still active