Ingests need edge (node?) knowledge source provenance fixed?

[RichardBruskiewich]

For consistency across all the ingests, we need to start tracking edge association knowledge source provenance.

• Phase 1: since the Biolink Model (and its Pydantic representation) require some revisions/updating of the provenance fields, we will initialy (over the short term) use the node/edge source field populated with properly registered infores curie values

• Phase 2: once the Biolink Model (and its Pydantic representation) are fully updated, we can change the source fields in the ingest Python code to the pertinent association slot (for edges), one of primary_knowledge_source, original_knowledge_source, aggregator_knowledge_source and/or supporting_knowledge_source, populated with their properly registered infores curie values (note: it is NOT a best practice to only use the knowledge_source slot)

[RichardBruskiewich]

I note that this issue only affects provenance tagging of biolink:Association instances, not concept nodes (which will still have the 'source' designation but using the infores of the primary information source).

[putmantime]

Koza Test Harness needs to be refactored to work with Koza 0.2