add i18 beamline definition - WITH device_factory

[stan-dot]

Fixes #709

Instructions to reviewer on how to test:

1. Do `dodal connect i181
2. Confirm all the devices expected to work, work

Checks for reviewer

• Would the PR title make sense to a scientist on a set of release notes
• If a new device has been added does it follow the standards
• If changing the API for a pre-existing device, ensure that any beamlines using this device have updated their Bluesky plans accordingly
• Have the connection tests for the relevant beamline(s) been run via dodal connect ${BEAMLINE}

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.57%. Comparing base (aab178f) to head (de2d3c6).
Report is 2 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #722      +/-   ##
==========================================
+ Coverage   97.51%   97.57%   +0.06%     
==========================================
  Files         152      159       +7     
  Lines        6351     6523     +172     
==========================================
+ Hits         6193     6365     +172     
  Misses        158      158              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[DominicOram]

Thanks Stan! Why are there so many skipped devices here?

[stan-dot]

@DominicOram the devices are skipped as we don't need all of them at the moment, as we roll out the devices following the plans we make, so it's gradual and agile

[DominicOram]

so it's gradual and agile

The more agile way to do it would be don't put the devices in at all until you need them otherwise we now have code in that never runs and we don't know if it works.

[stan-dot]

we need them for the plans in a different repo. also aren't devices skipped all the time depending if they are off or taken out of a beamline physically?

[stan-dot]

lint error

found 0 vulnerabilities
/home/runner/work/dodal/dodal/src/dodal/devices/i18/sim_motor_set.py
  /home/runner/work/dodal/dodal/src/dodal/devices/i18/sim_motor_set.py:57:12 - error: Cannot assign to attribute "delay" for class "SynAxis"
    Expression of type "float" cannot be assigned to attribute "delay" of class "SynAxis"
      "float" is incompatible with "int" (reportAttributeAccessIssue)
1 error, 0 warnings, 0 informations 

test error:

FAILED tests/devices/i22/test_dcm.py::test_configuration - AssertionError: Expected dcm to produce
{'dcm-bragg-motor_egu': {'value': '', 'timestamp': , 'alarm_severity': }, 'dcm-bragg-velocity': {'value': 0.0, 'timestamp': , 'alarm_severity': }, 'dcm-energy-motor_egu': {'value': '', 'timestamp': , 'alarm_severity': }, 'dcm-energy-velocity': {'value': 0.0, 'timestamp': , 'alarm_severity': }, 'dcm-crystal_1_roll-motor_egu': {'value': '', 'timestamp': , 'alarm_severity': }, 'dcm-crystal_1_roll-velocity': {'value': 0.0, 'timestamp': , 'alarm_severity': }, 'dcm-offset-motor_egu': {'value': '', 'timestamp': , 'alarm_severity': }, 'dcm-offset-velocity': {'value': 0.0, 'timestamp': , 'alarm_severity': }, 'dcm-crystal_2_pitch-motor_egu': {'value': '', 'timestamp': , 'alarm_severity': }, 'dcm-crystal_2_pitch-velocity': {'value': 0.0, 'timestamp': , 'alarm_severity': }, 'dcm-crystal_2_roll-motor_egu': {'value': '', 'timestamp': , 'alarm_severity': }, 'dcm-crystal_2_roll-velocity': {'value': 0.0, 'timestamp': , 'alarm_severity': }, 'dcm-perp-motor_egu': {'value': '', 'timestamp': , 'alarm_severity': }, 'dcm-perp-velocity': {'value': 0.0, 'timestamp': , 'alarm_severity': }, 'dcm-crystal_2_d_spacing': {'value': 3.13475, 'timestamp': , 'alarm_severity': }, 'dcm-crystal_2_usage': {'value': 'Bragg', 'timestamp': , 'alarm_severity': }, 'dcm-crystal_2_type': {'value': 'silicon', 'timestamp': , 'alarm_severity': }, 'dcm-crystal_1_reflection': {'value': [1, 1, 1], 'timestamp': , 'alarm_severity': }, 'dcm-crystal_2_reflection': {'value': [1, 1, 1], 'timestamp': , 'alarm_severity': }, 'dcm-crystal_1_d_spacing': {'value': 3.13475, 'timestamp': , 'alarm_severity': }, 'dcm-crystal_1_usage': {'value': 'Bragg', 'timestamp': , 'alarm_severity': }, 'dcm-crystal_1_type': {'value': 'silicon', 'timestamp': , 'alarm_severity': }}
but actually got
{'dcm-bragg-motor_egu': {'value': '', 'timestamp': 217.034084231, 'alarm_severity': 0}, 'dcm-bragg-velocity': {'value': 0.0, 'timestamp': 217.034112413, 'alarm_severity': 0}, 'dcm-energy-motor_egu': {'value': '', 'timestamp': 217.034883664, 'alarm_severity': 0}, 'dcm-energy-velocity': {'value': 0.0, 'timestamp': 217.034907609, 'alarm_severity': 0}, 'dcm-crystal_2_pitch-motor_egu': {'value': '', 'timestamp': 217.03565736, 'alarm_severity': 0}, 'dcm-crystal_2_pitch-velocity': {'value': 0.0, 'timestamp': 217.03567862, 'alarm_severity': 0}, 'dcm-offset-motor_egu': {'value': '', 'timestamp': 217.034355621, 'alarm_severity': 0}, 'dcm-offset-velocity': {'value': 0.0, 'timestamp': 217.034377142, 'alarm_severity': 0}, 'dcm-crystal_1_roll-motor_egu': {'value': '', 'timestamp': 217.035143273, 'alarm_severity': 0}, 'dcm-crystal_1_roll-velocity': {'value': 0.0, 'timestamp': 217.035164924, 'alarm_severity': 0}, 'dcm-crystal_2_roll-motor_egu': {'value': '', 'timestamp': 217.035400046, 'alarm_severity': 0}, 'dcm-crystal_2_roll-velocity': {'value': 0.0, 'timestamp': 217.035421566, 'alarm_severity': 0}, 'dcm-perp-motor_egu': {'value': '', 'timestamp': 217.034619808, 'alarm_severity': 0}, 'dcm-perp-velocity': {'value': 0.0, 'timestamp': 217.034641078, 'alarm_severity': 0}, 'dcm-crystal_2_type': {'value': 'silicon', 'timestamp': 217.028304579, 'alarm_severity': 0}, 'dcm-crystal_1_usage': {'value': 'Bragg', 'timestamp': 217.028175165, 'alarm_severity': 0}, 'dcm-crystal_1_reflection': {'value': [1, 1, 1], 'timestamp': 217.028232263, 'alarm_severity': 0}, 'dcm-crystal_2_reflection': {'value': [1, 1, 1], 'timestamp': 217.028327742, 'alarm_severity': 0}, 'dcm-crystal_1_d_spacing': {'value': 0.31356, 'timestamp': 217.028257851, 'alarm_severity': 0}, 'dcm-crystal_2_d_spacing': {'value': 0.31356, 'timestamp': 217.028350996, 'alarm_severity': 0}, 'dcm-crystal_2_usage': {'value': 'Bragg', 'timestamp': 217.028281545, 'alarm_severity': 0}, 'dcm-crystal_1_type': {'value': 'silicon', 'timestamp': 217.028205542, 'alarm_severity': 0}}
=========== 1 failed, 578 passed, 2 skipped, 27 deselected in 21.10s ===========

@iain-hall I'll try tackling the second one

[DiamondJoseph]

lint error

found 0 vulnerabilities
/home/runner/work/dodal/dodal/src/dodal/devices/i18/sim_motor_set.py
  /home/runner/work/dodal/dodal/src/dodal/devices/i18/sim_motor_set.py:57:12 - error: Cannot assign to attribute "delay" for class "SynAxis"

I would not use the Ophyd (i.e. not ophyd-async) SynAxis

[stan-dot]

the test_dcm error comes from a failure to correctly read in the 'ANY' mock value.

[stan-dot]

 src/dodal/beamlines/i18.py:17: in <module>
    from dodal.devices.i18.sim_detector import SimDetector
/opt/hostedtoolcache/Python/3.10.14/x64/lib/python3.10/site-packages/dodal/devices/i18/sim_detector.py:2: in <module>
    from ophyd_async.sim import PatternGenerator
/opt/hostedtoolcache/Python/3.10.14/x64/lib/python3.10/site-packages/ophyd_async/sim/__init__.py:1: in <module>
    from .pattern_generator import PatternGenerator
/opt/hostedtoolcache/Python/3.10.14/x64/lib/python3.10/site-packages/ophyd_async/sim/pattern_generator.py:14: in <module>
    import h5py
E   ModuleNotFoundError: No module named 'h5py'
____________ ERROR collecting src/dodal/devices/i18/sim_detector.py ____________
ImportError while importing test module '/home/runner/work/dodal/dodal/src/dodal/devices/i18/sim_detector.py'.

that is because in ophyd-async h5py was added only as a dev dependency

[stan-dot]

@DiamondJoseph how do I ask for the ophyd-async to be installed with the [sim] dependencies?

the error in question: ModuleNotFoundError: No module named 'h5py'

[DiamondJoseph]

I'll answer on sSlack to avoid everyone getting emails

[stan-dot]

waiting until pydantic 2

[DominicOram]

codecov/patch — 95.27% of diff hit (target 97.51%)
[Details](https://github.com/DiamondLightSource/dodal/pull/722/checks?check_run_id=36218650010)
@codecov
codecov/project Failing after 1s — 97.46% (-0.05%) compared to 497f390

@DiamondJoseph if the beamlines files are config not code, they should be excluded from code coverage, right?

I still think a smoke test would be good for them. i.e. go through and instantiate every device as a mock. I think we have similar code that does this for MX beamlins.

[DiamondJoseph]

@DiamondJoseph if the beamlines files are config not code, they should be excluded from code coverage, right?

I still think a smoke test would be good for them. i.e. go through and instantiate every device as a mock. I think we have similar code that does this for MX beamlins.

We do have that, it just is respecting skip_device so isn't checking the devices that are skipped. Removing the skips from the devices that are functional will help get the numbers back up. There's also 2 missed lines in the ThorLabs table set method, the other new device classes have tests but that does not.

[DiamondJoseph]

The device_factory suggestions are nits, but unskipping the devices will

a. get the code coverage up
b. mean you don't have to document why they are skipped when they are fully functional
c. give you a wider range of devices when you get to test on the beamline

[DiamondJoseph]

Just gotta get the i22 changes that snuck in with the rebase out again then it should be ready

[DiamondJoseph]

Link the issue for it #1009

[stan-dot]

added

[DiamondJoseph]

Once the test coverage is fixed this can go in- as discussed, can either add tests to the set methods on those classes, or (since they're not marked as Movables, and the axes are still available as children devices) remove the set method altogether.

[stan-dot]

96.92% of diff hit (target 97.51%)
97.50% (-0.02%) compared to aab178f

seriously?

[DiamondJoseph]

96.92% of diff hit (target 97.51%)
97.50% (-0.02%) compared to aab178f

seriously?

Seriously- the target is a moving target that only goes up. All new code has to be at least as tested as everything already in.

[stan-dot]

moving target that only goes up

that's very perfectionist

[DiamondJoseph]

Must: tests need asserts.

[stan-dot]

rebasing manually atm

[stan-dot]

superceded by #1030