scripts: add extractTournamentIdentity helper for backfill

Pure function that derives canonical (name, year) identity from a
scoring_events row, stripping trailing 4-digit years from the name or
falling back to eventDate. Used by the Phase 2 backfill to group
per-window events into canonical tournaments.
This commit is contained in:
Chris Parsons 2026-05-01 20:47:23 +00:00
parent 7168c85ae2
commit 85bca8bb77
No known key found for this signature in database
2 changed files with 101 additions and 0 deletions

View file

@ -0,0 +1,44 @@
import { describe, it, expect } from "vitest";
import { extractTournamentIdentity } from "../match-tournament";
describe("extractTournamentIdentity", () => {
it("uses eventDate year when the name has no trailing year", () => {
const identity = extractTournamentIdentity({
name: "Masters Tournament",
eventDate: "2026-04-09",
eventType: "tournament",
});
expect(identity).toEqual({ name: "Masters Tournament", year: 2026 });
});
it("derives the year from eventDate for a simple tournament name", () => {
const identity = extractTournamentIdentity({
name: "Wimbledon",
eventDate: "2026-07-01",
eventType: "tournament",
});
expect(identity).toEqual({ name: "Wimbledon", year: 2026 });
});
it("strips a trailing year from the name and uses it as the canonical year", () => {
const identity = extractTournamentIdentity({
name: "Wimbledon 2026",
eventDate: "2026-07-01",
eventType: "tournament",
});
expect(identity).toEqual({ name: "Wimbledon", year: 2026 });
});
it("throws when neither the name nor eventDate supply a year", () => {
expect(() =>
extractTournamentIdentity({
name: "Wimbledon",
eventDate: null,
eventType: "tournament",
}),
).toThrow(/cannot determine year/);
});
});

View file

@ -0,0 +1,57 @@
/**
* Pure function(s) for extracting canonical tournament identity
* (name + year) from a `scoring_events` row.
*
* Used by the Phase 2 backfill to group per-window events into
* canonical `tournaments` rows.
*/
export interface ScoringEventInput {
name: string;
eventDate: string | null;
eventType: string;
}
export interface TournamentIdentity {
/** Tournament name with any trailing 4-digit year stripped. */
name: string;
year: number;
}
const TRAILING_YEAR_RE = / (\d{4})$/;
/**
* Extracts the canonical `(name, year)` identity for a tournament
* from a scoring event.
*
* Resolution order for year:
* 1. A trailing 4-digit year on the event name (e.g. "Wimbledon 2026").
* The year is stripped from the returned name.
* 2. The first 4 characters of `eventDate` (format `YYYY-MM-DD`).
*
* Throws if neither source supplies a year.
*/
export function extractTournamentIdentity(
ev: ScoringEventInput,
): TournamentIdentity {
const trimmedName = ev.name.trim();
const match = trimmedName.match(TRAILING_YEAR_RE);
if (match) {
const yearFromName = Number(match[1]);
const nameWithoutYear = trimmedName.slice(0, match.index).trim();
return { name: nameWithoutYear, year: yearFromName };
}
if (ev.eventDate) {
const yearStr = ev.eventDate.slice(0, 4);
const yearFromDate = Number(yearStr);
if (Number.isFinite(yearFromDate) && yearStr.length === 4) {
return { name: trimmedName, year: yearFromDate };
}
}
throw new Error(
`cannot determine year for scoring event "${ev.name}" (eventDate=${ev.eventDate ?? "null"})`,
);
}