COSAS Mapping Script Fixes

[davidruvolo51]

I noticed that there are a couple of issues in the nightly mapping job. These occur during the import step, but there are likely minor changes in the nightly export that have effected the processing script.

Issues

• [x] Subjects table: Null ID error
• [x] Samples table: Unknown xref for attribute 'belongsToSubject' of entity 'umdm_samples'.
• [x] Clinical table: Unknown xref for attribute 'belongsToSubject' of entity 'umdm_clinical'.
• [x] Sample Preparation: `Unknown xref for attribute 'belongsToSample' of entity 'umdm_samplePreparation'.
• [x] Sequencing Data: Unknown xref for attribute 'belongsToSamplePreparation' of entity 'umdm_sequencing'.

Improvments

While I'm at it, I might as well fix a few things that have been bothering me.

• [x] Add checks for f.<id> != None throughout the script
• [x] Remove the cosastools class and rename methods to camelCase.
• [x] In the py-molgenis-client extension, remove all rest API methods in favor of the file import approach
• [x] For dt transformations, revise iteration variable name so it's clearer to read
• [x] Remove bottlenecks when collapsing values by ID

[davidruvolo51]

For Null id error when importing data into the subjects table. Add the following filter when the subjectsToRegister object is binded to the main subjects dataset.

-subjects = dt.rbind(...)[...][...]
+subjects = dt.rbind(...)[...][...][f.subjectID != None, :]

[davidruvolo51]

Errors with xref attributes were likely due to the Null id error. All identifiers exist in the raw/processed data, but the import failed as the subjects data failed to import properly.