valknar/worldcup
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: normalize Bosnia & Herzegovina and USA team name variants

Browse Source 
Add TEAM_ALIASES to lib/wiki-scraper.ts applied at extraction time so both
scraper and sync consistently produce canonical names. Removes the duplicate
alias map from seed.ts in favour of the shared normalizeTeam() export.

Aliases added:
  Bosnia & Herzegovina  → Bosnia and Herzegovina
  USA                   → United States

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Sebastian Krüger
2026-06-15 17:33:05 +02:00
parent f885e4312c
commit b832b62f5e
2 changed files with 19 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files
lib/wiki-scraper.ts
+15 -1
View File
@@ -75,6 +75,20 @@ export async function fetchWikiHtml(page: string, retries = 5): Promise<string |
return null return null
} }
// ── Team name normalisation ────────────────────────────────────────────────
const TEAM_ALIASES: Record<string, string> = {
'West Germany': 'Germany',
'Korea Republic': 'South Korea',
'IR Iran': 'Iran',
'Bosnia & Herzegovina': 'Bosnia and Herzegovina',
'USA': 'United States',
}
export function normalizeTeam(name: string): string {
return TEAM_ALIASES[name] ?? name
}
// ── Parsing helpers ──────────────────────────────────────────────────────── // ── Parsing helpers ────────────────────────────────────────────────────────
function parseScoreText(text: string): [number, number] | null { function parseScoreText(text: string): [number, number] | null {
@@ -92,7 +106,7 @@ function extractTeam($: CheerioAPI, $cell: Cheerio<Element>): string {
return false return false
} }
}) })
return name return normalizeTeam(name)
} }
function parseGoals($: CheerioAPI, $td: Cheerio<Element>): Goal[] { function parseGoals($: CheerioAPI, $td: Cheerio<Element>): Goal[] {
scripts/seed.ts
+4 -13
View File
@@ -5,6 +5,7 @@ import { readFileSync, existsSync } from 'fs'
import path from 'path' import path from 'path'
import { fileURLToPath } from 'url' import { fileURLToPath } from 'url'
import { getIso } from '../lib/iso-codes' import { getIso } from '../lib/iso-codes'
import { normalizeTeam } from '../lib/wiki-scraper'
const DATABASE_URL = process.env.DATABASE_URL ?? 'postgres://wc:wc@localhost:5432/worldcup' const DATABASE_URL = process.env.DATABASE_URL ?? 'postgres://wc:wc@localhost:5432/worldcup'
const __dirname = path.dirname(fileURLToPath(import.meta.url)) const __dirname = path.dirname(fileURLToPath(import.meta.url))
@@ -15,16 +16,6 @@ const YEARS = [
1978,1982,1986,1990,1994,1998,2002,2006,2010,2014,2018,2022, 1978,1982,1986,1990,1994,1998,2002,2006,2010,2014,2018,2022,
] ]
// Normalize team names from Wikipedia to canonical DB names
const TEAM_ALIASES: Record<string, string> = {
'West Germany': 'Germany',
'Korea Republic': 'South Korea',
'IR Iran': 'Iran',
}
function normTeam(name: string): string {
return TEAM_ALIASES[name] ?? name
}
function readJson<T>(filePath: string): T | null { function readJson<T>(filePath: string): T | null {
if (!existsSync(filePath)) return null if (!existsSync(filePath)) return null
@@ -174,7 +165,7 @@ async function run() {
const teamCache = new Map<string, number>() const teamCache = new Map<string, number>()
async function upsertTeam(rawName: string): Promise<number> { async function upsertTeam(rawName: string): Promise<number> {
const name = normTeam(rawName) const name = normalizeTeam(rawName)
if (teamCache.has(name)) return teamCache.get(name)! if (teamCache.has(name)) return teamCache.get(name)!
const iso2 = getIso(name) const iso2 = getIso(name)
const [row] = await db.execute(sql` const [row] = await db.execute(sql`
@@ -207,8 +198,8 @@ async function run() {
INSERT INTO tournaments (year, host, winner, runner_up, third_place, fourth_place, teams_count) INSERT INTO tournaments (year, host, winner, runner_up, third_place, fourth_place, teams_count)
VALUES ( VALUES (
${year}, ${meta.host || null}, ${year}, ${meta.host || null},
${normTeam(meta.winner ?? '') || null}, ${normTeam(meta.runner_up ?? '') || null}, ${normalizeTeam(meta.winner ?? '') || null}, ${normalizeTeam(meta.runner_up ?? '') || null},
${normTeam(meta.third_place ?? '') || null}, ${normTeam(meta.fourth_place ?? '') || null}, ${normalizeTeam(meta.third_place ?? '') || null}, ${normalizeTeam(meta.fourth_place ?? '') || null},
${meta.teams_count ?? null} ${meta.teams_count ?? null}
) )
ON CONFLICT (year) DO UPDATE SET ON CONFLICT (year) DO UPDATE SET