// Post-build audit of dist/client: one h1, alt text, SEO tags, unique ids, internal links/anchors.
// Usage: npm run build && npm run check:site
import { existsSync, readdirSync, readFileSync, statSync } from 'node:fs';
import { join, relative } from 'node:path';

const ROOT = 'dist/client';
const SERVER_ROUTES = [/^\/api\//]; // on-demand routes, not files

const walk = (dir) =>
  readdirSync(dir).flatMap((f) => {
    const p = join(dir, f);
    return statSync(p).isDirectory() ? walk(p) : [p];
  });

const pages = walk(ROOT).filter((f) => f.endsWith('.html') && !f.includes('/pagefind/'));
const idsByPage = new Map();
const problems = [];
const report = (page, msg) => problems.push(`${page}: ${msg}`);

const urlOf = (file) =>
  '/' +
  relative(ROOT, file)
    .replace(/index\.html$/, '')
    .replace(/\.html$/, '');
const html = new Map(pages.map((f) => [urlOf(f), readFileSync(f, 'utf8')]));

for (const [url, src] of html) {
  const ids = [...src.matchAll(/\sid="([^"]+)"/g)].map((m) => m[1]);
  idsByPage.set(url, new Set(ids));
  const dupes = ids.filter((id, i) => ids.indexOf(id) !== i);
  if (dupes.length) report(url, `duplicate id(s): ${[...new Set(dupes)].join(', ')}`);

  const h1 = (src.match(/<h1[\s>]/g) ?? []).length;
  if (h1 !== 1) report(url, `${h1} <h1> elements`);
  if (!/<html[^>]+lang="/.test(src)) report(url, 'missing <html lang>');
  if (!/<title>[^<]+<\/title>/.test(src)) report(url, 'missing <title>');
  if (url !== '/404' && !/<meta name="description" content="[^"]+"/.test(src))
    report(url, 'missing meta description');
  if (url !== '/404' && !/<link rel="canonical" href="https:\/\/ta1sph\.info\//.test(src))
    report(url, 'missing canonical');
  for (const img of src.match(/<img\b[^>]*>/g) ?? []) {
    if (!/\salt="/.test(img)) report(url, `img without alt: ${img.slice(0, 80)}`);
  }
}

for (const [url, src] of html) {
  const refs = [...src.matchAll(/\s(?:href|src)="([^"]+)"/g)].map((m) => m[1]);
  for (const ref of refs) {
    if (/^(https?:|mailto:|tel:|data:|javascript:)/.test(ref) || ref.startsWith('//')) {
      if (!ref.startsWith('https://ta1sph.info')) continue;
    }
    const target = new URL(ref.replaceAll('&amp;', '&'), `https://ta1sph.info${url}`);
    if (target.host !== 'ta1sph.info') continue;
    const path = decodeURIComponent(target.pathname);
    if (SERVER_ROUTES.some((r) => r.test(path))) continue;
    const file = path.endsWith('/')
      ? join(ROOT, path, 'index.html')
      : ([join(ROOT, path), join(ROOT, `${path}.html`)].find(existsSync) ?? join(ROOT, path));
    if (!existsSync(file)) {
      report(url, `broken link → ${path}`);
      continue;
    }
    if (target.hash && idsByPage.has(path)) {
      const id = decodeURIComponent(target.hash.slice(1));
      if (!idsByPage.get(path).has(id)) report(url, `missing anchor → ${path}${target.hash}`);
    }
  }
}

console.log(`Checked ${html.size} pages.`);
if (problems.length) {
  console.error(problems.map((p) => `  ✗ ${p}`).join('\n'));
  process.exit(1);
}
console.log('  ✓ no problems found');