Dispatch a data_source_pull job for this source.
The worker resolves the CopassIdAdapter by DataSource.provider
and drives its pull() method, dispatching a chunking job per
yielded batch. The rest of the pipeline (chunking → ontology_ingestion
→ strategy) is unchanged.
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
ISO 8601 datetime — incremental cursor, adapter-defined semantics
If true, pulled rows are written directly to the Copass-ID vault under data/{extraction_id}/raw.json and no ontology_ingestion jobs are dispatched. Use this to stage raw data for later ingestion without running extraction.