open-dj conformance corpus

Seed corpus for the spec. Each version's corpus lives under corpus-<VERSION>/.

Layout

• corpus-0.2/ -- first seed corpus (Phase 15, this repo). Covers Appendix B cases 1, 2, 3, 4, 5, 8, 10, 11, 13 from the spec. 10 tracks drawn from the user's real library via the Phase 2 matched-tracks set. Phase 16 adds cases 6, 7, 9, 12, 14, 15 (needs Serato / Traktor / variable-tempo source data).

How to use

From the repo root:

python -m apps.open_dj.cli validate open-dj/conformance/corpus-0.2/case-01-classic-isrc.open-dj.json
python -m apps.open_dj.cli canon open-dj/conformance/corpus-0.2/case-01-classic-isrc.open-dj.json

The tests in tests/open_dj/test_corpus_roundtrip.py iterate the whole directory and assert every file is canon-idempotent and schema-valid.

Adding a case

1. Pick a qualifying track from the real library (see data/sync/matches.csv).
2. Export it via an adapter (cli export --adapter rekordbox --tracks <stable_id>).
3. Strip file_path to a relative path rooted at /music/.
4. Name the file case-<NN>-<slug>.open-dj.json.
5. Update corpus-0.2/README.md with the case number + source stable_id.