Port Slits to ophyd-async and expand fields

[callumforrester]

Inspired by #377

An ophyd-async version of Slits is required for the upcoming I22 experiment in June 2024 in order to populate the NeXus file with relevant metadata. It seems sensible to port the existing version into an ophyd-async class that remains compatible with hyperion rather than create two devices.

Additional fields need to be added to populate the NeXus file correctly (see below)

Acceptance Criteria

• The following ophyd devices are ophyd-async: S4SlitGaps
• Slits config available in i22 beamline module
• Fake slits config available in p38 beamline module
• @DominicOram and the hyperion team are happy with their compatiblity
• The attributes required by the NeXus files are exposed: X_centre, Y_centre, X_size, Y_size

[callumforrester]

Assumption to be confirmed is that I22 slits are the same as I03 slits, or at least that the IOCs can be modified to have the same naming conventions.

[callumforrester]

Have just realised there's no overlap in fields needed by I22 and hyperion

• Hyperion
  • xgap
  • ygap
• I22
  • X_centre, Y_centre, X_size, Y_size

so two options:

1. Have one device with all fields, extra config to customise output of .read()
2. Have two devices

Any thoughts @DominicOram or @olliesilvester?

[DominicOram]

Ok, had a chat to controls, xgap is the same as X_size but PVs are annoyingly not standard. We will ask them to move to the I22 standard given that the other slits on i03 use this standard. So have two devices for now, we will move when we have the standard. However, I still don't think these slits will have individual blade control so we may need the common device to have that as optional

[DominicOram]

FYI https://jira.diamond.ac.uk/browse/I03-956

[DominicOram]

I think we should be able to switch to these by the time they're written, controls are making the changes now