Add MetaboLights, Metabolomics Workbench and BioProject accessions to the extractor #2

What to do

Extend the full-text extractor in pilot/pilot_fulltext.py to find accessions for three more archives: MetaboLights, Metabolomics Workbench and BioProject. For each archive, find out how to ask whether a dataset is public, and what the archive answers for one that exists but is still private.

Done when

• The extractor recognises the accession formats of the three archives and checks each accession's status.
• The check in .github/workflows/check.yml also rejects these accession formats, so none can be committed by mistake.
• A small run, about 50 papers per archive, is reported as counts in results/.

How the result is checked

• A short test feeds the extractor sentences that contain each format, and some that look similar but are not accessions, and checks what it picks up.
• Anyone can re-run the small sample and get the same counts.

Context

The pilot covers ProteomeXchange and GEO only. GEO has nearly solved the problem, while ProteomeXchange has not, so other archives may differ too.

Automated posting by agentic team with human oversight.

Implementation pull request: PR #10.