@link-assistant/web-capture 2.0.0 → 2.0.1

package/tests/integration/gdocs-public-doc.test.js

@@ -1,315 +0,0 @@
-/**
- * Integration tests for the Google Docs markdown round-trip reference document.
- *
- * The document at
- *   https://docs.google.com/document/d/1f5zI2xOFpKa90v0GjamO_t7lqSdzMlaM/edit
- * is a public document created specifically for validating web-capture's Google
- * Docs support across every URL variation and every Markdown feature category
- * documented in issue #90. The ground-truth markdown and authoring archive are
- * preserved under docs/case-studies/issue-90/reference/.
- *
- * Live capture tests are gated behind GDOCS_INTEGRATION=true so normal PR CI
- * stays hermetic. The URL-variation, document-id extraction and feature
- * checklist tests always run so regressions in the parsing surface are caught
- * even without network access.
- */
-
-import { jest } from '@jest/globals';
-import fs from 'fs';
-import path from 'path';
-import { fileURLToPath } from 'url';
-
-import {
-  isGoogleDocsUrl,
-  extractDocumentId,
-  buildExportUrl,
-  buildEditUrl,
-  buildDocsApiUrl,
-  selectGoogleDocsCaptureMethod,
-  fetchGoogleDocAsMarkdown,
-  captureGoogleDocWithBrowser,
-} from '../../src/gdocs.js';
-import { retry } from '../../src/retry.js';
-
-const LIVE =
-  process.env.GDOCS_INTEGRATION === 'true' ||
-  process.env.GDOCS_INTEGRATION === '1';
-const describeIfLive = LIVE ? describe : describe.skip;
-
-jest.setTimeout(120000);
-
-const __filename = fileURLToPath(import.meta.url);
-const __dirname = path.dirname(__filename);
-const REFERENCE_DIR = path.resolve(
-  __dirname,
-  '..',
-  '..',
-  '..',
-  'docs',
-  'case-studies',
-  'issue-90',
-  'reference'
-);
-
-// The public document referenced in issue #90.
-// https://docs.google.com/document/d/1f5zI2xOFpKa90v0GjamO_t7lqSdzMlaM/edit
-export const PUBLIC_TEST_DOCUMENT = {
-  id: '1f5zI2xOFpKa90v0GjamO_t7lqSdzMlaM',
-  canonicalUrl:
-    'https://docs.google.com/document/d/1f5zI2xOFpKa90v0GjamO_t7lqSdzMlaM/edit',
-  urlVariations: [
-    'https://docs.google.com/document/d/1f5zI2xOFpKa90v0GjamO_t7lqSdzMlaM',
-    'https://docs.google.com/document/d/1f5zI2xOFpKa90v0GjamO_t7lqSdzMlaM/edit',
-    'https://docs.google.com/document/d/1f5zI2xOFpKa90v0GjamO_t7lqSdzMlaM/edit?usp=sharing',
-    'https://docs.google.com/document/d/1f5zI2xOFpKa90v0GjamO_t7lqSdzMlaM/edit?usp=sharing&ouid=102030405060708090100&rtpof=true&sd=true',
-    'https://docs.google.com/document/d/1f5zI2xOFpKa90v0GjamO_t7lqSdzMlaM/edit?tab=t.0',
-  ],
-  // Section headings from the reference markdown. Each one should be present
-  // in any Markdown produced from the document, regardless of capture mode.
-  // See docs/case-studies/issue-90/reference/markdown-test-document.md.
-  sections: [
-    'Markdown Feature Test Document',
-    '1. Headings',
-    '2. Inline Formatting',
-    '3. Paragraphs',
-    '4. Blockquotes',
-    '5. Unordered Lists',
-    '6. Ordered Lists',
-    '7. Mixed Lists',
-    '8. Tables',
-    '9. Links',
-    '10. Images',
-    '11. Horizontal Rules',
-    '12. Special Characters',
-    '13. Nested Formatting Edge Cases',
-    '14. Empty and Minimal Table Content',
-  ],
-};
-
-// The issue #108 v2 document extends the public reference fixture with Section
-// 15, covering browser-model list type and continuation-paragraph regressions.
-export const PUBLIC_TEST_DOCUMENT_V2 = {
-  id: '1Rvaod_u2wgkAUNdXG-e29yQem-P36KAQ',
-  canonicalUrl:
-    'https://docs.google.com/document/d/1Rvaod_u2wgkAUNdXG-e29yQem-P36KAQ/edit',
-};
-
-function expectIssue108Markdown(markdown) {
-  expect(markdown).toContain(
-    '15. List & Blockquote Edge Cases (browser-mode reproducers)'
-  );
-  expect(markdown).toMatch(/1\.\s+Apple/u);
-  expect(markdown).toMatch(/2\.\s+Banana/u);
-  expect(markdown).toMatch(/3\.\s+Cherry/u);
-  expect(markdown).not.toMatch(/^-\s+Apple$/mu);
-  expect(markdown).toContain(
-    'Continuation paragraph that is not a list item. Same indent as Step one.'
-  );
-  expect(markdown).not.toContain(
-    '> Continuation paragraph that is not a list item.'
-  );
-  expect(markdown).toMatch(/^-\s+Red$/mu);
-  expect(markdown).toMatch(/^-\s+Green$/mu);
-  expect(markdown).toMatch(/^-\s+Blue$/mu);
-}
-
-describe('Google Docs public test document (issue #90)', () => {
-  describe('URL variations', () => {
-    it.each(PUBLIC_TEST_DOCUMENT.urlVariations)(
-      'accepts %s as a Google Docs URL',
-      (url) => {
-        expect(isGoogleDocsUrl(url)).toBe(true);
-      }
-    );
-
-    it.each(PUBLIC_TEST_DOCUMENT.urlVariations)(
-      'extracts the public document ID from %s',
-      (url) => {
-        expect(extractDocumentId(url)).toBe(PUBLIC_TEST_DOCUMENT.id);
-      }
-    );
-
-    it('builds the expected export URL for the public document', () => {
-      expect(buildExportUrl(PUBLIC_TEST_DOCUMENT.id, 'html')).toBe(
-        `https://docs.google.com/document/d/${PUBLIC_TEST_DOCUMENT.id}/export?format=html`
-      );
-    });
-
-    it('builds the expected editor URL for the public document', () => {
-      expect(buildEditUrl(PUBLIC_TEST_DOCUMENT.id)).toBe(
-        `https://docs.google.com/document/d/${PUBLIC_TEST_DOCUMENT.id}/edit`
-      );
-    });
-
-    it('builds the expected Docs API URL for the public document', () => {
-      expect(buildDocsApiUrl(PUBLIC_TEST_DOCUMENT.id)).toBe(
-        `https://docs.googleapis.com/v1/documents/${PUBLIC_TEST_DOCUMENT.id}`
-      );
-    });
-  });
-
-  describe('Capture method selection (issue #72 regression guard)', () => {
-    it('routes --capture browser to the editor-model backend', () => {
-      expect(selectGoogleDocsCaptureMethod('browser')).toBe('browser-model');
-    });
-
-    it('routes --capture api to the public-export backend when no token', () => {
-      expect(selectGoogleDocsCaptureMethod('api')).toBe('public-export');
-    });
-
-    it('routes --capture api with a token to the Docs REST API backend', () => {
-      expect(selectGoogleDocsCaptureMethod('api', 'token-123')).toBe(
-        'docs-api'
-      );
-    });
-  });
-
-  describe('Reference markdown fixture', () => {
-    const referenceMarkdownPath = path.join(
-      REFERENCE_DIR,
-      'markdown-test-document.md'
-    );
-    const referenceMarkdown = fs.readFileSync(referenceMarkdownPath, 'utf-8');
-
-    it('contains every feature category documented in issue #90', () => {
-      for (const section of PUBLIC_TEST_DOCUMENT.sections) {
-        expect(referenceMarkdown).toContain(section);
-      }
-    });
-
-    it('embeds the four test images referenced by the issue', () => {
-      expect(referenceMarkdown).toMatch(
-        /!\[Blue rectangle]\(media\/image1\.png\)/
-      );
-      expect(referenceMarkdown).toMatch(
-        /!\[Red rectangle]\(media\/image2\.png\)/
-      );
-      expect(referenceMarkdown).toMatch(
-        /!\[Green square]\(media\/image3\.png\)/
-      );
-      expect(referenceMarkdown).toMatch(
-        /!\[Yellow square]\(media\/image4\.png\)/
-      );
-    });
-  });
-
-  describeIfLive('Live capture against the public document', () => {
-    // The primary deliverable of issue #90: the reference document is
-    // downloaded from Google Docs and every feature section can be found in
-    // the captured markdown. This is our regression net for any future
-    // changes to the public-export or HTML-to-markdown pipeline.
-    //
-    // Google Docs occasionally returns transient 500s on the public-export
-    // endpoint. The Habr integration suite has the same flake and solves it
-    // with retry + exponential backoff, so we reuse the shared helper here.
-    const fetchWithRetry = (url) =>
-      retry(() => fetchGoogleDocAsMarkdown(url), {
-        retries: 3,
-        baseDelay: 2000,
-      });
-
-    const browserEngine =
-      process.env.GDOCS_BROWSER_ENGINE || process.env.BROWSER_ENGINE;
-    const captureBrowserWithRetry = (url) =>
-      retry(
-        () =>
-          captureGoogleDocWithBrowser(url, {
-            ...(browserEngine ? { engine: browserEngine } : {}),
-          }),
-        {
-          retries: 2,
-          baseDelay: 2000,
-        }
-      );
-
-    it('fetches the public document via --capture api and returns markdown', async () => {
-      const { markdown, documentId, exportUrl } = await fetchWithRetry(
-        PUBLIC_TEST_DOCUMENT.canonicalUrl
-      );
-
-      expect(documentId).toBe(PUBLIC_TEST_DOCUMENT.id);
-      expect(exportUrl).toBe(buildExportUrl(PUBLIC_TEST_DOCUMENT.id, 'html'));
-      expect(markdown.length).toBeGreaterThan(1000);
-
-      for (const section of PUBLIC_TEST_DOCUMENT.sections) {
-        expect(markdown).toContain(section);
-      }
-      expect(markdown).toContain('## 1. Headings');
-      expect(markdown).toContain('**This text is bold**');
-      expect(markdown).toContain('*This text is italic*');
-      expect(markdown).toContain('~~This text has strikethrough~~');
-      expect(markdown).toContain('> This is a single-level blockquote');
-      expect(markdown).toContain('| Feature | Supported | Notes |');
-      expect(markdown).not.toContain('| Feature |  | Supported |');
-      expect(markdown).toContain('|  | x |  |');
-      expect(markdown).toContain('1.  Parent item 1');
-      expect(markdown).toContain('    1.  Child item 1.1');
-      expect(markdown).toContain('        1.  Grandchild item 1.2.1');
-      expect(markdown).not.toContain('-   Child item 1.1');
-    });
-
-    it.each(PUBLIC_TEST_DOCUMENT.urlVariations)(
-      'resolves %s to the same document when capturing',
-      async (url) => {
-        const { documentId } = await fetchWithRetry(url);
-        expect(documentId).toBe(PUBLIC_TEST_DOCUMENT.id);
-      }
-    );
-
-    it('captures the public document via --capture browser and renders DOCS_modelChunk markdown', async () => {
-      const { markdown, documentId, exportUrl, capture } =
-        await captureBrowserWithRetry(PUBLIC_TEST_DOCUMENT.canonicalUrl);
-
-      expect(documentId).toBe(PUBLIC_TEST_DOCUMENT.id);
-      expect(exportUrl).toBe(buildEditUrl(PUBLIC_TEST_DOCUMENT.id));
-      expect(capture.blocks.length).toBeGreaterThan(20);
-      expect(capture.images.length).toBeGreaterThanOrEqual(4);
-      expect(markdown.length).toBeGreaterThan(2500);
-      expect(markdown).toContain('# Markdown Feature Test Document');
-      expect(markdown).toContain('## 1. Headings');
-      expect(markdown).toContain('**This text is bold**');
-      expect(markdown).toContain('*This text is italic*');
-      expect(markdown).toContain('~~This text has strikethrough~~');
-      expect(markdown).toContain('> This is a single-level blockquote');
-      expect(markdown).toContain('[Regular link](https://example.com)');
-      expect(markdown).toContain('| Feature | Supported | Notes |');
-      expect(markdown).not.toContain('| Feature |  | Supported |');
-      expect(markdown).toContain('|  | x |  |');
-      expect(markdown).toContain('1. Parent item 1');
-      expect(markdown).toContain('    1. Child item 1.1');
-      expect(markdown).toContain('        1. Grandchild item 1.2.1');
-      expect(markdown).not.toContain('- Child item 1.1');
-      expect(markdown).not.toContain('1. Parent item 1\n\n');
-      expect(markdown).toContain('![Blue rectangle](');
-      expect(markdown).toContain('![Red rectangle](');
-      expect(markdown).toContain('![Green square](');
-      expect(markdown).toContain('![Yellow square](');
-      expect(markdown).toContain('docs-images-rt/');
-      expect(markdown).toContain('---');
-    });
-
-    it('fetches the issue #108 v2 document via --capture api and preserves Section 15 semantics', async () => {
-      const { markdown, documentId, exportUrl } = await fetchWithRetry(
-        PUBLIC_TEST_DOCUMENT_V2.canonicalUrl
-      );
-
-      expect(documentId).toBe(PUBLIC_TEST_DOCUMENT_V2.id);
-      expect(exportUrl).toBe(
-        buildExportUrl(PUBLIC_TEST_DOCUMENT_V2.id, 'html')
-      );
-      expectIssue108Markdown(markdown);
-    });
-
-    it('captures the issue #108 v2 document via --capture browser and preserves Section 15 semantics', async () => {
-      const { markdown, documentId, exportUrl, capture } =
-        await captureBrowserWithRetry(PUBLIC_TEST_DOCUMENT_V2.canonicalUrl);
-
-      expect(documentId).toBe(PUBLIC_TEST_DOCUMENT_V2.id);
-      expect(exportUrl).toBe(buildEditUrl(PUBLIC_TEST_DOCUMENT_V2.id));
-      expect(capture.blocks.length).toBeGreaterThan(30);
-      expectIssue108Markdown(markdown);
-      expect(markdown).not.toContain('*Apple*');
-      expect(markdown).not.toContain('*Red*');
-    });
-  });
-});

package/tests/integration/github-readme.test.js

@@ -1,133 +0,0 @@
-/**
- * Integration tests for capturing GitHub repository pages (issue #5).
- *
- * Live tests are gated behind GITHUB_REPOSITORY_INTEGRATION so default/offline
- * runs stay deterministic. The mocked API endpoint tests cover the default
- * behavior without network access; this suite proves the real GitHub repository
- * page can still be captured as compact text/markdown, original HTML, and PNG
- * screenshots in every supported JavaScript browser engine.
- */
-
-import { jest } from '@jest/globals';
-import { createBrowser } from '../../src/browser.js';
-import { fetchHtml } from '../../src/lib.js';
-import { retry } from '../../src/retry.js';
-import { expectPngScreenshot } from '../helpers/screenshot.js';
-import {
-  fetchGithubRepositorySnapshot,
-  formatGithubRepositoryMarkdown,
-  formatGithubRepositoryText,
-  parseGithubRepositoryUrl,
-} from '../../src/github.js';
-
-const GITHUB_REPOSITORY_URL =
-  process.env.GITHUB_REPOSITORY_URL ||
-  'https://github.com/link-assistant/web-capture';
-
-const SKIP_LIVE =
-  !process.env.GITHUB_REPOSITORY_INTEGRATION ||
-  process.env.GITHUB_REPOSITORY_INTEGRATION === 'false';
-const describeIfLive = SKIP_LIVE ? describe.skip : describe;
-
-jest.setTimeout(120000);
-
-async function navigateWithRetry(page, url) {
-  await retry(
-    async () => {
-      await page.goto(url, {
-        waitUntil: 'domcontentloaded',
-        timeout: 30000,
-      });
-    },
-    {
-      retries: 3,
-      baseDelay: 2000,
-      onRetry: (err, attempt, delay) => {
-        console.log(
-          `Navigation retry ${attempt} for ${url} after ${delay}ms: ${err.message}`
-        );
-      },
-    }
-  );
-  await page.waitForTimeout(5000);
-}
-
-describeIfLive('GitHub repository capture', () => {
-  it('downloads compact repository text and markdown with README plus file tree', async () => {
-    const parsed = parseGithubRepositoryUrl(GITHUB_REPOSITORY_URL);
-    expect(parsed).not.toBeNull();
-
-    const snapshot = await retry(
-      () => fetchGithubRepositorySnapshot(GITHUB_REPOSITORY_URL),
-      {
-        retries: 3,
-        baseDelay: 2000,
-      }
-    );
-
-    const text = formatGithubRepositoryText(snapshot);
-    const markdown = formatGithubRepositoryMarkdown(snapshot);
-
-    expect(text).toContain(`Repository: ${parsed.fullName}`);
-    expect(text).toContain('Files:');
-    expect(text).toMatch(/README/i);
-    expect(text.length).toBeGreaterThan(500);
-
-    expect(markdown).toContain(`# ${parsed.fullName}`);
-    expect(markdown).toContain('## Files');
-    expect(markdown).toMatch(/## .*README/i);
-    expect(markdown.length).toBeGreaterThan(500);
-    expect(markdown).not.toMatch(/<html/i);
-  });
-
-  it('downloads original GitHub repository HTML', async () => {
-    const parsed = parseGithubRepositoryUrl(GITHUB_REPOSITORY_URL);
-    expect(parsed).not.toBeNull();
-
-    const html = await retry(() => fetchHtml(GITHUB_REPOSITORY_URL), {
-      retries: 3,
-      baseDelay: 2000,
-    });
-
-    expect(html).toMatch(/<html/i);
-    expect(html).toContain(parsed.owner);
-    expect(html).toContain(parsed.repo);
-    expect(html.length).toBeGreaterThan(1000);
-  });
-});
-
-describe.each(['puppeteer', 'playwright'])(
-  'GitHub repository screenshot (%s engine)',
-  (engine) => {
-    describeIfLive(`${engine}`, () => {
-      let browser;
-
-      beforeAll(async () => {
-        browser = await createBrowser(engine);
-      });
-
-      afterAll(async () => {
-        if (browser) {
-          await browser.close();
-        }
-      });
-
-      it('captures the GitHub repository page as a PNG screenshot', async () => {
-        const page = await browser.newPage();
-        await page.setExtraHTTPHeaders({
-          'Accept-Language': 'en-US,en;q=0.9',
-          'Accept-Charset': 'utf-8',
-        });
-        await page.setUserAgent(
-          'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36'
-        );
-        await page.setViewport({ width: 1280, height: 800 });
-        await navigateWithRetry(page, GITHUB_REPOSITORY_URL);
-
-        const screenshot = await page.screenshot({ type: 'png' });
-
-        expectPngScreenshot(screenshot);
-      });
-    });
-  }
-);

package/tests/integration/google-drive-image.test.js

@@ -1,246 +0,0 @@
-import request from 'supertest';
-import nock from '../helpers/nock.js';
-import unzipper from 'unzipper';
-import { app } from '../../src/index.js';
-
-const FIRST_FILE_ID = '1Cxkx6-428EQAX0-eiaq66H829ohnPp7q';
-const SECOND_FILE_ID = '1fgJaftjv53xCN7vgiJOaQlbfWkUqPgyd';
-const JPEG_BYTES = Buffer.from(
-  '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',
-  'base64'
-);
-
-const DRIVE_URLS = [
-  {
-    name: 'first issue URL with /view',
-    fileId: FIRST_FILE_ID,
-    url: `https://drive.google.com/file/d/${FIRST_FILE_ID}/view`,
-  },
-  {
-    name: 'first issue URL without /view',
-    fileId: FIRST_FILE_ID,
-    url: `https://drive.google.com/file/d/${FIRST_FILE_ID}`,
-  },
-  {
-    name: 'second issue URL with /view',
-    fileId: SECOND_FILE_ID,
-    url: `https://drive.google.com/file/d/${SECOND_FILE_ID}/view`,
-  },
-  {
-    name: 'second issue URL without /view',
-    fileId: SECOND_FILE_ID,
-    url: `https://drive.google.com/file/d/${SECOND_FILE_ID}`,
-  },
-];
-
-function directDownloadUrl(fileId) {
-  return `https://drive.google.com/uc?export=download&id=${fileId}`;
-}
-
-function mockDriveHead(fileId) {
-  return nock('https://drive.google.com')
-    .head('/uc')
-    .query({ export: 'download', id: fileId })
-    .reply(200, '', { 'content-type': 'image/jpeg' });
-}
-
-function mockDriveGet(fileId) {
-  return nock('https://drive.google.com')
-    .get('/uc')
-    .query({ export: 'download', id: fileId })
-    .reply(200, JPEG_BYTES, {
-      'content-type': 'image/jpeg',
-      'content-length': String(JPEG_BYTES.length),
-    });
-}
-
-function bufferParser(res, callback) {
-  const chunks = [];
-  res.on('data', (chunk) => chunks.push(Buffer.from(chunk)));
-  res.on('end', () => callback(null, Buffer.concat(chunks)));
-}
-
-afterEach(() => {
-  nock.cleanAll();
-});
-
-describe('Google Drive public image links', () => {
-  it.each(DRIVE_URLS)(
-    '/image downloads image bytes for $name',
-    async (item) => {
-      mockDriveGet(item.fileId);
-
-      const res = await request(app)
-        .get('/image')
-        .query({ url: item.url })
-        .buffer(true)
-        .parse(bufferParser)
-        .expect(200);
-
-      expect(res.headers['content-type']).toMatch(/^image\/jpeg/);
-      expect(res.body.subarray(0, 3)).toEqual(Buffer.from([0xff, 0xd8, 0xff]));
-      expect(nock.isDone()).toBe(true);
-    }
-  );
-
-  it.each(DRIVE_URLS)(
-    '/fetch proxies direct image bytes for $name',
-    async (item) => {
-      mockDriveGet(item.fileId);
-
-      const res = await request(app)
-        .get('/fetch')
-        .query({ url: item.url })
-        .buffer(true)
-        .parse(bufferParser)
-        .expect(200);
-
-      expect(res.headers['content-type']).toMatch(/^image\/jpeg/);
-      expect(res.body.subarray(0, 3)).toEqual(Buffer.from([0xff, 0xd8, 0xff]));
-      expect(nock.isDone()).toBe(true);
-    }
-  );
-
-  it.each(DRIVE_URLS)(
-    '/stream proxies direct image bytes for $name',
-    async (item) => {
-      mockDriveGet(item.fileId);
-
-      const res = await request(app)
-        .get('/stream')
-        .query({ url: item.url })
-        .buffer(true)
-        .parse(bufferParser)
-        .expect(200);
-
-      expect(res.headers['content-type']).toMatch(/^image\/jpeg/);
-      expect(res.body.subarray(0, 3)).toEqual(Buffer.from([0xff, 0xd8, 0xff]));
-      expect(nock.isDone()).toBe(true);
-    }
-  );
-
-  it.each(DRIVE_URLS)(
-    '/markdown includes a direct image link for $name',
-    async (item) => {
-      mockDriveHead(item.fileId);
-
-      const res = await request(app)
-        .get('/markdown')
-        .query({ url: item.url })
-        .expect(200);
-
-      expect(res.headers['content-type']).toMatch(/text\/markdown/);
-      expect(res.text).toContain(
-        `![Google Drive image](${directDownloadUrl(item.fileId)})`
-      );
-      expect(nock.isDone()).toBe(true);
-    }
-  );
-
-  it.each(DRIVE_URLS)(
-    '/txt includes the original and direct image links for $name',
-    async (item) => {
-      mockDriveHead(item.fileId);
-
-      const res = await request(app)
-        .get('/txt')
-        .query({ url: item.url })
-        .expect(200);
-
-      expect(res.headers['content-type']).toMatch(/text\/plain/);
-      expect(res.text).toContain('Google Drive image');
-      expect(res.text).toContain(`Original Google Drive file: ${item.url}`);
-      expect(res.text).toContain(`Image: ${directDownloadUrl(item.fileId)}`);
-      expect(nock.isDone()).toBe(true);
-    }
-  );
-
-  it.each(DRIVE_URLS)(
-    '/html includes a direct image tag for $name',
-    async (item) => {
-      mockDriveHead(item.fileId);
-
-      const res = await request(app)
-        .get('/html')
-        .query({ url: item.url })
-        .expect(200);
-
-      expect(res.headers['content-type']).toMatch(/text\/html/);
-      expect(res.text).toContain(
-        `src="${directDownloadUrl(item.fileId).replace('&', '&')}"`
-      );
-      expect(nock.isDone()).toBe(true);
-    }
-  );
-
-  it('archive includes markdown and the downloaded Drive image', async () => {
-    const item = DRIVE_URLS[0];
-    mockDriveHead(item.fileId);
-    mockDriveGet(item.fileId);
-
-    const res = await request(app)
-      .get('/archive')
-      .query({ url: item.url })
-      .buffer(true)
-      .parse(bufferParser)
-      .expect(200);
-
-    const zip = await unzipper.Open.buffer(res.body);
-    const document = zip.files.find((file) => file.path === 'document.md');
-    const image = zip.files.find((file) => file.path.startsWith('images/'));
-
-    expect(document).toBeDefined();
-    expect(image).toBeDefined();
-    expect((await document.buffer()).toString('utf8')).toContain(
-      '![Google Drive image](images/image-1.jpg)'
-    );
-    expect((await image.buffer()).subarray(0, 3)).toEqual(
-      Buffer.from([0xff, 0xd8, 0xff])
-    );
-    expect(nock.isDone()).toBe(true);
-  });
-
-  it('docx includes the downloaded Drive image', async () => {
-    const item = DRIVE_URLS[0];
-    mockDriveHead(item.fileId);
-    mockDriveGet(item.fileId);
-
-    const res = await request(app)
-      .get('/docx')
-      .query({ url: item.url })
-      .buffer(true)
-      .parse(bufferParser)
-      .expect(200);
-
-    expect(res.headers['content-type']).toMatch(
-      /application\/vnd\.openxmlformats-officedocument\.wordprocessingml\.document/
-    );
-    expect(res.body.subarray(0, 2)).toEqual(Buffer.from('PK'));
-
-    const docx = await unzipper.Open.buffer(res.body);
-    const mediaEntry = docx.files.find(
-      (file) => file.path.startsWith('word/media/') && !file.path.endsWith('/')
-    );
-    expect(mediaEntry).toBeDefined();
-    expect((await mediaEntry.buffer()).subarray(0, 3)).toEqual(
-      Buffer.from([0xff, 0xd8, 0xff])
-    );
-    expect(nock.isDone()).toBe(true);
-  });
-
-  it('pdf renders the downloaded Drive image document', async () => {
-    const item = DRIVE_URLS[0];
-    mockDriveGet(item.fileId);
-
-    const res = await request(app)
-      .get('/pdf')
-      .query({ url: item.url })
-      .buffer(true)
-      .parse(bufferParser)
-      .expect(200);
-
-    expect(res.headers['content-type']).toMatch(/application\/pdf/);
-    expect(res.body.subarray(0, 4).toString('utf8')).toBe('%PDF');
-    expect(nock.isDone()).toBe(true);
-  });
-});