chat_web_app/llm-fe/scripts/stampCanonicals.test.js

const { test } = require('node:test');
const assert = require('node:assert/strict');
const fs = require('fs');
const os = require('os');
const path = require('path');
const { PAGE_PATHS, h1ForPath, titleForPath, descriptionForPath } = require('../src/llm-fe/utils/canonicals');
const { htmlForPath, stampCanonicals } = require('./stampCanonicals');

const HOME = `<!DOCTYPE html><html><head>
<link rel="canonical" href="https://hesychia.ai/" />
<meta property="og:title" content="Hesychia — quiet, focused AI chat" />
<meta property="og:description" content="Hesychia is a $10/month AI chat product from AI ML Operations, LLC. It stores your documents so answers can use your files." />
<meta property="og:url" content="https://hesychia.ai/" />
<meta property="og:image" content="https://hesychia.ai/og-image.png" />
<meta name="twitter:card" content="summary_large_image" />
<meta name="twitter:title" content="Hesychia — quiet, focused AI chat" />
<meta name="twitter:description" content="Hesychia is a $10/month AI chat product from AI ML Operations, LLC. It stores your documents so answers can use your files." />
<meta name="twitter:image" content="https://hesychia.ai/og-image.png" />
<meta name="description" content="Hesychia is a $10/month AI chat product from AI ML Operations, LLC. It stores your documents so answers can use your files." />
<title>Hesychia — quiet, focused AI chat</title>
</head><body><div id="root"><h1>Hesychia is a calm AI assistant.</h1></div></body></html>`;

test('each page canonical is that page, and other head tags stay', () => {
  const html = htmlForPath(HOME, '/pricing/');
  assert.match(html, /<link rel="canonical" href="https:\/\/hesychia\.ai\/pricing\/" \/>/);
  assert.equal(html.match(/rel="canonical"/g).length, 1);
  assert.match(html, /property="og:url" content="https:\/\/hesychia\.ai\/pricing\/"/);
  assert.match(html, /property="og:title" content="Hesychia pricing — \$10\/month, no ads"/);
  assert.match(html, /property="og:description" content="\$10\/month, flat\. No ads, no upsells, one price\./);
  assert.match(html, /property="og:image" content="https:\/\/hesychia\.ai\/og-image\.png"/);
  assert.match(html, /name="twitter:card" content="summary_large_image"/);
  assert.match(html, /name="twitter:title" content="Hesychia pricing — \$10\/month, no ads"/);
  const nav = html.match(/<nav aria-label="Related pages">([\s\S]*?)<\/nav>/);
  assert.ok(nav);
  const anchors = nav[1].match(/<a href=/g);
  assert.ok(anchors && anchors.length >= 2 && anchors.length <= 5);
  assert.match(nav[1], /href="\/">What Hesychia is</);
  assert.doesNotMatch(nav[1], /href="\/pricing\/"/);
  assert.doesNotMatch(nav[1], />here</);
  assert.match(html, /id="crawler-prose"/);
  assert.equal(html.match(/application\/ld\+json/g).length, 1);
  const pricingLd = JSON.parse(html.match(/<script type="application\/ld\+json">([\s\S]*?)<\/script>/)[1]);
  const pricingPage = pricingLd['@graph'].find((node) => node['@type'] === 'WebPage');
  assert.equal(pricingPage.headline, 'Hesychia pricing — $10/month, no ads');
  assert.equal(pricingPage.url, 'https://hesychia.ai/pricing/');
  assert.match(html, /\$10\/month, flat, with no ads, no upsells, and one price/);
  assert.doesNotMatch(
    html.match(/<div id="crawler-prose">([\s\S]*?)<\/div>/)[1],
    /Quiet, focused AI chat/,
  );
  assert.match(html, /<title>Hesychia pricing — \$10\/month, no ads<\/title>/);
  assert.equal(html.match(/<title\b/g).length, 1);
  assert.match(html, /name="description" content="\$10\/month, flat\. No ads, no upsells, one price\./);
  assert.equal(html.match(/name="description"/g).length, 1);
  assert.equal(html.match(/<h1\b/g).length, 1);
  assert.match(html, /<h1>Pricing<\/h1>/);
});

test('stamp writes a shell per route and leaves the homepage self-referencing', () => {
  const dir = fs.mkdtempSync(path.join(os.tmpdir(), 'canon-'));
  fs.writeFileSync(path.join(dir, 'index.html'), HOME);
  stampCanonicals(dir);

  const home = fs.readFileSync(path.join(dir, 'index.html'), 'utf8');
  assert.match(home, /href="https:\/\/hesychia\.ai\/"/);
  const homeLd = JSON.parse(home.match(/<script type="application\/ld\+json">([\s\S]*?)<\/script>/)[1]);
  const homeOrg = homeLd['@graph'].find((node) => node['@type'] === 'Organization' && node.name === 'Hesychia');
  assert.equal(homeOrg.url, 'https://hesychia.ai');
  assert.equal(homeOrg.logo, 'https://hesychia.ai/apple-touch-icon.png');
  assert.doesNotMatch(home, /href="https:\/\/hesychia\.ai\/pricing\/"/);
  assert.equal(home.match(/<h1\b/g).length, 1);
  assert.match(home, /<h1>Hesychia is a calm AI assistant.<\/h1>/);

  const titles = PAGE_PATHS.map((pathname) => titleForPath(pathname));
  assert.equal(new Set(titles).size, titles.length);
  const descriptions = PAGE_PATHS.map((pathname) => descriptionForPath(pathname));
  assert.equal(new Set(descriptions).size, descriptions.length);

  for (const pathname of PAGE_PATHS) {
    if (pathname === '/') continue;
    const rel = pathname.endsWith('/')
      ? path.join(pathname.slice(1), 'index.html')
      : pathname.slice(1);
    const html = fs.readFileSync(path.join(dir, rel), 'utf8');
    const title = titleForPath(pathname);
    assert.match(html, new RegExp(`href="https://hesychia\\.ai${pathname.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}"`));
    assert.match(html, /name="twitter:card" content="summary_large_image"/);
    const related = html.match(/<nav aria-label="Related pages">([\s\S]*?)<\/nav>/);
    assert.ok(related, `${pathname} missing related nav`);
    const relatedAnchors = related[1].match(/<a href=/g);
    assert.ok(
      relatedAnchors && relatedAnchors.length >= 2 && relatedAnchors.length <= 5,
      `${pathname} related link count`,
    );
    assert.doesNotMatch(related[1], new RegExp(`href="${pathname.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}"`));
    assert.match(html, /property="og:image" content="https:\/\/hesychia\.ai\/og-image\.png"/);
    assert.match(
      html,
      new RegExp(`property="og:title" content="${title.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}"`),
    );
    assert.equal(html.match(/rel="canonical"/g).length, 1);
    assert.equal(html.match(/<h1\b/g).length, 1);
    assert.match(html, new RegExp(`<h1>${h1ForPath(pathname).replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}</h1>`));
    assert.ok(title.length >= 30 && title.length <= 60, `${pathname} title length ${title.length}`);
    assert.match(html, new RegExp(`<title>${title.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}</title>`));
    assert.equal(html.match(/application\/ld\+json/g).length, 1);
    const ld = JSON.parse(html.match(/<script type="application\/ld\+json">([\s\S]*?)<\/script>/)[1]);
    const webPage = ld['@graph'].find((node) => node['@type'] === 'WebPage');
    assert.equal(webPage.headline, title, `${pathname} JSON-LD headline`);
    const org = ld['@graph'].find((node) => node['@type'] === 'Organization' && node.name === 'Hesychia');
    assert.ok(org, `${pathname} missing Organization`);
    assert.equal(org.url, 'https://hesychia.ai');
    assert.equal(org.logo, 'https://hesychia.ai/apple-touch-icon.png');
    if (pathname === '/changelog/') {
      const posts = ld['@graph'].filter((node) => node['@type'] === 'BlogPosting');
      assert.ok(posts.length >= 2);
      for (const post of posts) {
        assert.ok(post.headline);
        assert.equal(post.author.name, 'AI ML Operations, LLC');
        assert.match(post.datePublished, /^\d{4}-\d{2}-\d{2}$/);
        assert.equal(post.dateModified, post.datePublished);
      }
    }
    const description = descriptionForPath(pathname);
    assert.ok(
      description.length >= 110 && description.length <= 160,
      `${pathname} description length ${description.length}`,
    );
    assert.match(
      html,
      new RegExp(`content="${description.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}"`),
    );
  }
});