Turtle generator

ods_to_rdf.py reads src/ontology/<COMPASS_USE_CASE>/source-data.ods, validates the result against src/ontology/shapes.ttl, and writes compass.ttl and vocab.ttl next to the workbook. COMPASS_USE_CASE (default oceancare) is read from the environment, then from the repository-root .env.

just data::generate    # regenerate the Turtle
just data::check       # fail if the committed Turtle differs from a fresh run

Output is deterministic: unchanged input gives byte-identical files.

Test

cd src/turtle-generator && uv run pytest