Remove has_calibration from WorkflowExecution subclasses after MassSpectrometry change sheets

[brynnz22]

see issue: #1570

In issue linked above, discussed creating a Calibration class. Will need to change name of slot from has_calibration which implies a boolean and we are linking a DataObject.

This will require a migration.

[brynnz22]

After discussion with @kheal and @anastasiyaprymolenna has_calibration is fine the way its modeled on ChromatographicSeparationProcess. Calibration does not need to have its own class. There are no calibration reagents to capture.

[kheal]

See update here: #1918 . After discussing limitations of only capturing the a DataObject with regards to calibration, we will likely make a Calibration class.

[kheal]

See microbiomedata#133.

After ingestion and removal of has_calibration from WorkflowExecutionActivity

1. The range of has_calibration on CalibrationInformation class should be limited toCalibrationInformation.
2. The definitions for CalibrationInformation, has_calibration, calibration_target, CalibrationTargetEnum, internal_calibration, calibration_standard, CalibrationStandardEnum, and calibration_object can move from the workflow_execution.yaml to the a logical location of nmdc.yaml

[aclum]

Is this planned for after the rollout?

[kheal]

@aclum . As part of the rollout, @brynnz22 and I will need to make change sheets for existing GC-MS metabolomics and NOM DataGeneration records to refactor the use of the 'has_calibration' slot. After that we can remove that slot and its contents from the WorkflowExecution class. If we remove that slot before the change sheets, we'll orphan the DataObjects that are linked through that slot.

[kheal]

Note that this is dependent on the completion of microbiomedata/issues#750

[brynnz22]

@kheal after thinking about this further, I don't think we should add the has_calibration in the change sheet and write a migrator to remove the slot from the WorkflowExecution. This allows room for a lot of human error. I think we should go with what I suggested last week. Write a migrator that moves it upstream to the MassSpectrometry its on, and then remove. This allows us to do checks and make sure its on the correct instance. Whereas using a change sheet then removing it in a migrator allows for possible errors of mismatching. I can write the migrator for this. I will also do the change sheet. thanks!

[kheal]

@brynnz22 Sounds reasonable. I certainly prefer migrators over changesheets and I agree there are fewer opportunities for errors. I'll update the microbiomedata/issues#750 with this new plan. Will you make a new issue (or change this issues' name/description?) to describe the migrator? I think I understand it, but I think you'll capture it more thoroughly.