Add fly scan plan stub

[abbiemery]

Fixes #278.

This adds a plan stub for the current fly scanning flow we have. That being:

- Stage devices and open a run
- Set up the flyer with a static sequence table and detectors with a trigger
- Declare the stream and kickoff the scan
- Collect while completing
- Close run and unstage devices.

This does the middle three points, so all you should need to do outside this is instantiate devices, stage and open a run.

[coretl]

This should probably have a more specific name.

Something like time_resolved_trigger_detectors?

[abbiemery]

I agree, I just wanted to call it something and figured i'd ask for opinions before merging.

[abbiemery]

Perhaps rewording your suggestion to improve clarity? Not sure what it should be though.

[coretl]

Requires:

• Remove ADAravis Deadtime Lookup Table #305
• Add collect_while_completing plan stub and test bluesky#1720

Test script:

# Import bluesky and ophyd
import bluesky.plans as bp
import bluesky.preprocessors as bpp
from bluesky import RunEngine
from bluesky.utils import ProgressBarManager, register_transform
from dodal.beamlines import p45
from dodal.beamlines.beamline_utils import set_directory_provider

from ophyd_async.core import HardwareTriggeredFlyable, StaticDirectoryProvider
from ophyd_async.panda._trigger import StaticSeqTableTriggerLogic
from ophyd_async.planstubs import fly_and_collect

# Create a run engine, with plotting, progressbar and transform
RE = RunEngine({}, call_returns_result=True)
RE.waiting_hook = ProgressBarManager()
register_transform("RE", prefix="<")

# Create ophyd-async devices
set_directory_provider(StaticDirectoryProvider("/dls/p45/data/2024/cm37283-2/tmp"))
det = p45.diff()
panda = p45.panda1()
flyer = HardwareTriggeredFlyable(StaticSeqTableTriggerLogic(panda.seq[1]))


@bpp.stage_decorator([panda, det])
@bpp.run_decorator()
def my_custom_plan():
    yield from fly_and_collect(
        "primary", [panda, det], flyer, 10, 0.1, 0.5, repeats=2, period=1
    )


RE(my_custom_plan(), lambda n, d: print(n, d))

[abbiemery]

@coretl I have had to change the collect_while_completing stub call back to the code used for the i22 experiment. An issue that was raised back then has not been addressed and as such the new stub is not compatible yet. The issue and following discussion is here bluesky/bluesky#1630.

[abbiemery]

I have also had to change back the complete_all and kickoff_all to just bare kickoff and complete as this are not in the latest alpha of bluesky. I have made an issue to go through ophyd-async and update these once we get the next bluesky alpha.

#328

[abbiemery]

I have also dropped the patch to 90% as it was up at 98 and was starting to get prohibitive. This will not effect the catching of overall coverage.

[abbiemery]

Made an issue to investgate the stub. #329