v0.3.0

Download the archive for your system, unpack it, and run fil. There is nothing else to
install. Start with fil add <DOI>, then fil expand; see the README for more.

Changelog

• f33afb5 Add fil IDs to identity (M3 step 1)
• b133f95 Key works on fil IDs: schema v2 and the first migration (M3 step 2)
• 8b50b71 M3 step 10: fil list, fil show, stats --why, and three MCP tools for the PDFs (#12)
• d9aca6f M3 step 11: the 20-seed validation run, five fixes it found, M3 complete (#13)
• 62ad625 M3 step 5: a legal PDF link, or the reason there is none, for every work (#5)
• 67ba495 M3 step 6: fil fetch downloads legal open-access PDFs into papers/ (#6)
• 650725d M3 step 7: read every PDF and make it searchable with fil search (#7)
• e6b2185 M3 step 8: fil import identifies, checks and files the PDFs a researcher has (#10)
• 5d67fcc M3 step 9: citation context — the sentence each citation is made in (#11)
• d72e572 Make the library a folder, with papers/ (M3 step 3)
• 86749bb Plan M3: fil IDs, the library folder, PDFs and text
• f978e4e Run spikes 9, 10 and 11: PDF availability, the extractor, reference matching (M3 step 4) (#3)
• 8c2a9a9 Run vet, tests and a no-cgo build on every PR, on three systems (#2)
• 0262da6 Search falls back to partial matches; fetch gives stubs a title before downloading (#9)
• 57ab226 Spike 9: Unpaywall adds nothing, so it is not built (D21) (#8)
• 70db164 Spikes 9 and 10: limit every redirect hop, and stop timed-out extractions (#4)