cup26/scripts/buildPreviewData.ts

// international-results.csv + goalscorers.csv → public/data/{h2h.json,scorers.json}
// Precomputes match-preview inputs that need the full historical corpus: every
// World-Cup-team pairing's head-to-head record, and each team's top historical
// scorers. ESPN supplies live form/lineups/venue at runtime; these are the deep
// historical layers.
import { readFileSync, writeFileSync, mkdirSync } from 'node:fs';
import { fileURLToPath } from 'node:url';
import { dirname, join } from 'node:path';
import { canonicalTeam, ALL_TEAMS } from '../src/lib/teams';

const ROOT = join(dirname(fileURLToPath(import.meta.url)), '..');
const RESULTS = join(ROOT, 'data', 'raw', 'international-results.csv');
const SCORERS = join(ROOT, 'data', 'raw', 'goalscorers.csv');
const OUT_DIR = join(ROOT, 'public', 'data');

const WC = new Set(ALL_TEAMS);
const pairKey = (a: string, b: string) => [a, b].sort().join('|');

interface H2HMeeting { date: string; home: string; away: string; hs: number; as: number }
interface H2HRecord {
  a: string; b: string;
  games: number; aWins: number; bWins: number; draws: number; aGoals: number; bGoals: number;
  last: H2HMeeting[];
}

function buildH2H(): Record<string, H2HRecord> {
  const lines = readFileSync(RESULTS, 'utf8').split('\n');
  const rec: Record<string, H2HRecord> = {};
  for (let i = 1; i < lines.length; i++) {
    const c = lines[i]?.split(',');
    if (!c || c.length < 9) continue;
    const hs = Number(c[3]); const as = Number(c[4]);
    if (!Number.isFinite(hs) || !Number.isFinite(as)) continue;
    const home = canonicalTeam(c[1]!); const away = canonicalTeam(c[2]!);
    if (!WC.has(home) || !WC.has(away)) continue;
    const [a, b] = [home, away].sort() as [string, string];
    const key = pairKey(home, away);
    let r = rec[key];
    if (!r) { r = { a, b, games: 0, aWins: 0, bWins: 0, draws: 0, aGoals: 0, bGoals: 0, last: [] }; rec[key] = r; }
    // goals from a's/b's perspective
    const aIsHome = home === a;
    const aScore = aIsHome ? hs : as;
    const bScore = aIsHome ? as : hs;
    r.games++; r.aGoals += aScore; r.bGoals += bScore;
    if (aScore > bScore) r.aWins++; else if (aScore < bScore) r.bWins++; else r.draws++;
    r.last.push({ date: c[0]!, home, away, hs, as });
  }
  // keep only the 6 most recent meetings per pair
  for (const r of Object.values(rec)) {
    r.last.sort((x, y) => y.date.localeCompare(x.date));
    r.last = r.last.slice(0, 6);
  }
  return rec;
}

interface Scorer { name: string; goals: number; pens: number }

function buildScorers(): Record<string, Scorer[]> {
  const lines = readFileSync(SCORERS, 'utf8').split('\n');
  const byTeam = new Map<string, Map<string, { goals: number; pens: number }>>();
  for (let i = 1; i < lines.length; i++) {
    const c = lines[i]?.split(',');
    if (!c || c.length < 8) continue;
    const team = canonicalTeam(c[3]!);
    const scorer = c[4];
    if (!scorer || !WC.has(team)) continue;
    if ((c[6] ?? '').toUpperCase() === 'TRUE') continue; // own goal
    const pen = (c[7] ?? '').toUpperCase() === 'TRUE';
    if (!byTeam.has(team)) byTeam.set(team, new Map());
    const m = byTeam.get(team)!;
    const e = m.get(scorer) ?? { goals: 0, pens: 0 };
    e.goals++; if (pen) e.pens++;
    m.set(scorer, e);
  }
  const out: Record<string, Scorer[]> = {};
  for (const [team, m] of byTeam) {
    out[team] = [...m.entries()]
      .map(([name, v]) => ({ name, goals: v.goals, pens: v.pens }))
      .sort((x, y) => y.goals - x.goals)
      .slice(0, 10);
  }
  return out;
}

function main(): void {
  const h2h = buildH2H();
  const scorers = buildScorers();
  mkdirSync(OUT_DIR, { recursive: true });
  writeFileSync(join(OUT_DIR, 'h2h.json'), JSON.stringify(h2h));
  writeFileSync(join(OUT_DIR, 'scorers.json'), JSON.stringify(scorers));
  console.log(`preview data → h2h.json (${Object.keys(h2h).length} pairings), scorers.json (${Object.keys(scorers).length} teams)`);
}

main();