Where sources disagree or names differ, a person decides. These tables record each decision and its basis.
data/manual/breed-names.json
Parker 2017 supplement-1 abbreviation -> site breed id, Wikidata QID, FCI number. Drafted with Claude Code on 2026-11-04 from the sources below; every entry is reviewed=false until Paul checks it. Size varieties share one id and carry a variant (Paul, 2026-11-04). Ids are slugs of the usual English name. Doubts are listed per entry, never resolved silently. Extended to the full Parker 2017 breed set with Claude Code on 2026-11-09: the new entries are drafts (reviewed=false), see data/reports/review.md.
33 of 164 entries reviewed
data/manual/display-names.json
English display names that replace the Wikidata label on the site (Paul, 2026-11-09: usual name and title case). Drafted with Claude Code on 2026-11-09 from the Parker 2017 supplement-1 name (AKC usage) and the FCI English name; reviewed=false until Paul checks it. The Wikidata label stays in breeds.json as names.wikidataEn.
6 of 12 entries reviewed
data/manual/families.json
Colour family (docs/DESIGN.md §2) of every breed that is not in the 30-breed table of DESIGN.md §2, which stays the source for those 30. Family is an editorial choice per breed, not a datum: the draft follows the breed type and FCI group first, then the Parker clade, and leans on the 30 choices Paul already made (Akita and Chow Chow ancient, Miniature Schnauzer and Pinscher terrier, Saint Bernard and Rottweiler mastiff). Drafted with Claude Code on 2026-11-09; every entry is reviewed=false until Paul checks it.
0 of 131 entries reviewed
data/manual/origin-places.json
FCI country (or Wikidata P495 value) -> place on the map: kind, ISO 3166-1 alpha-2, Wikidata item whose P625 gives the point and P30 the continent. Drafted with Claude Code on 2026-11-04; reviewed=false until Paul checks it. Extended to the full Parker breed set on 2026-11-09: new places are drafts (reviewed=false).
15 of 38 entries reviewed
data/manual/parker-leaf-prefixes.json
Leaf-name prefixes in Parker 2017 supplement 5 that are not breed abbreviations of supplement 1. They come from the Vaysse et al. 2011 dataset (column 'Vayasse-1' of supplement 1) plus a few Hayward/other codes. Each alias is inferred from the breed name AND from the dog counts: alias leaves + leaves under the official abbreviation = 'Total' of supplement 1. To be reviewed by Paul; set reviewed=true once checked.
31 of 31 entries reviewed