Works
You rarely need to think about works, but they explain why Courmap can translate between providers that disagree about how to split a show. New terms are in the glossary.
A work is Courmap’s canonical unit: usually one cour, one film or one OVA, at the finest grain any provider uses. It matches how AniDB and AniList split shows, so the second half of a split season is its own work even when TVDB files it under the same season as the first half.
Every provider maps to works, never to each other. With eight providers, mapping each pair would mean 28 mapping sets to keep consistent; mapping each provider to the work means eight, and any translation is two hops: provider to work, work to provider.
Worked example: Spy x Family
Section titled “Worked example: Spy x Family”data/works/spy-x-family-part-2.yaml describes the second cour:
id: spy-x-family-part-2type: tvstatus: finishedyear: 2022episodes: 13titles: romaji: Spy x Family Part 2ids: anidb: 17061 anilist: 142838 mal: 50602 kitsu: 45619 tvdb_show: 405920 tmdb_show: 120089projections: tvdb_show: - { season: 1, episodes: 1-13, start: 13 } tmdb_show: - { season: 1, episodes: 1-13, start: 13 }prequel: spy-x-family- The id (
spy-x-family-part-2) is a slug from the romaji title. It never changes once published, so tools can store it. - Canonical episodes are the work’s own numbering: regular episodes 1 to 13, plus specials (
S1,S2and so on) and credits, trailers and parodies as separate kinds, following AniDB. - AniDB, AniList, MAL and Kitsu number episodes the way the work does, so their episode numbers need no mapping.
- TVDB and TMDB lay the episodes out in seasons, described by projections.
prequellinks the work to the one it follows. It orders sequels; it never changes how episodes are numbered.
Most works are not hand-written: about 33,000 come from seed datasets, and a YAML file with the same id overrides a seed key by key.
This product uses the TMDB API but is not endorsed or certified by TMDB.