d5ba21fa76
Standalone repo that builds embeddable JAMstack map bundles (web components), composes them into a lazy-loaded SPA via layouts/, and optionally wraps a Capacitor Android/iOS app — one build, choose your output/CDN. Diátaxis docs. History reset to a single clean commit; no PII, IPs, or internal terms.
71 lines
4.1 KiB
JavaScript
71 lines
4.1 KiB
JavaScript
// verify-resource-join.mjs — prove the gospel-resource join against the RAW pools.
|
||
//
|
||
// node verify-resource-join.mjs # exits 1 on any failure
|
||
//
|
||
// ═══ WHY THIS EXISTS ════════════════════════════════════════════════════════
|
||
// `no_resource_found` is the most consequential claim in this corpus: it is what
|
||
// says a people group has NOTHING in their language. A false positive sends effort
|
||
// toward a group that already has a resource; a false negative hides a group that
|
||
// has nothing. Neither is visible by reading the table.
|
||
//
|
||
// Checking the table against language-resources.json only proves the table is
|
||
// FAITHFUL TO ITS OWN INPUT. It cannot catch an error in that input. So this checks
|
||
// BOTH DIRECTIONS against the RAW pool caches:
|
||
// * has_x=true but the ISO is NOT in the raw pool -> false positive
|
||
// * has_x=false but the ISO IS in the raw pool -> false negative (the worse one)
|
||
//
|
||
// TWO PARSING TRAPS, both of which I hit and which made a hand-check report 306
|
||
// fake mismatches before I read the schemas:
|
||
// * arclight-languages.json is a PAGINATED ENVELOPE — the records are at
|
||
// _embedded.mediaLanguages, not the top level. Reading the top level yields 0.
|
||
// * ebible-translations.csv carries a UTF-8 BOM, so row['languageCode'] is
|
||
// undefined unless the file is read as utf-8-sig / the BOM is stripped.
|
||
// A cross-check that silently parses 0 records "passes" everything, which is the
|
||
// same fail-open shape this repo keeps catching. Hence the guard asserts the raw
|
||
// pools are NON-EMPTY and near their known sizes before comparing anything.
|
||
|
||
import { readFileSync } from 'node:fs';
|
||
import { fileURLToPath } from 'node:url';
|
||
import { dirname, join } from 'node:path';
|
||
|
||
const __dirname = dirname(fileURLToPath(import.meta.url));
|
||
const OUT = join(__dirname, 'out');
|
||
|
||
let failed = 0;
|
||
const fail = (m) => { console.error(` ✗ ${m}`); failed++; };
|
||
const ok = (m) => console.log(` ✓ ${m}`);
|
||
|
||
const arc = JSON.parse(readFileSync(join(OUT, 'arclight-languages.json'), 'utf8'));
|
||
const arcIso = new Set((arc?._embedded?.mediaLanguages || []).map((r) => r.iso3).filter(Boolean));
|
||
|
||
const csv = readFileSync(join(OUT, 'ebible-translations.csv'), 'utf8').replace(/^/, '');
|
||
const [head, ...lines] = csv.split(/\r?\n/);
|
||
const col = head.split(',').findIndex((h) => h.replace(/"/g, '') === 'languageCode');
|
||
const ebIso = new Set(lines.map((l) => (l.split(',')[col] || '').replace(/"/g, '').trim()).filter(Boolean));
|
||
|
||
console.log('GOSPEL-RESOURCE JOIN VERIFICATION (against the RAW pools)');
|
||
|
||
// FAIL-CLOSED on an empty parse — an empty pool would clear every comparison below.
|
||
if (arcIso.size < 1500) fail(`arclight parsed only ${arcIso.size} ISO codes — expected ~2085. Parser broken, comparisons below are meaningless.`);
|
||
else ok(`arclight raw: ${arcIso.size} ISO codes`);
|
||
if (ebIso.size < 900) fail(`ebible parsed only ${ebIso.size} ISO codes — expected ~1241. Parser broken.`);
|
||
else ok(`ebible raw: ${ebIso.size} ISO codes`);
|
||
if (failed) { console.log('\nFAILED — raw pools did not parse; not comparing.'); process.exit(1); }
|
||
|
||
const langs = JSON.parse(readFileSync(join(OUT, 'language-resources.json'), 'utf8')).languages;
|
||
const fpFilm = langs.filter((l) => l.has_film && !arcIso.has(l.iso639_3));
|
||
const fnFilm = langs.filter((l) => !l.has_film && arcIso.has(l.iso639_3));
|
||
const fpScr = langs.filter((l) => l.has_scripture && !ebIso.has(l.iso639_3));
|
||
const fnScr = langs.filter((l) => !l.has_scripture && ebIso.has(l.iso639_3));
|
||
|
||
const rep = (label, arr) => arr.length
|
||
? fail(`${arr.length} ${label}: e.g. ${arr.slice(0, 5).map((l) => l.iso639_3).join(', ')}`)
|
||
: ok(`0 ${label}`);
|
||
|
||
rep('FALSE POSITIVE has_film (claims film, absent from raw pool)', fpFilm);
|
||
rep('FALSE NEGATIVE has_film (in raw pool, claims none) — the dangerous direction', fnFilm);
|
||
rep('FALSE POSITIVE has_scripture', fpScr);
|
||
rep('FALSE NEGATIVE has_scripture — the dangerous direction', fnScr);
|
||
|
||
console.log(`\n${failed ? `FAILED — ${failed} problem(s)` : 'JOIN VALID — both directions agree with the raw pools'}`);
|
||
process.exit(failed ? 1 : 0);
|