goon_plugin/__tests__/simple_video.test.mjs

// Regression tests: simple <video> / direct-media detection.
//
// The feature branch broke detection of pages that expose a plain <video>
// (e.g. a site that serves a direct MP4 with a TRAILING SLASH and a
// ?rnd= query: .../14644093/14644093.mp4/?rnd=...). Two site-agnostic fixes
// are guarded here:
//   1) The URL predicate (isValidManifestUrl) must let concrete media signals
//      (.mp4/.m3u8/.webm/... and /hls/) WIN over the placeholder/`%`/trailing-
//      slash rejections, so a real direct-media URL is not mistaken for a
//      route template.
//   2) scrapeTab must fall back to an ISOLATED-world scrape when the MAIN-world
//      scrape yields nothing — the runtime cause of "detects nothing" on such pages
//      is a MAIN-world + allFrames rejection on its ad-iframe-heavy page.
import { describe, it } from './runner.mjs';
import { JSDOM } from 'jsdom';
import * as fs from 'fs';
import * as path from 'path';
import { fileURLToPath } from 'url';

const __dirname = path.dirname(fileURLToPath(import.meta.url));
const indexSource = fs.readFileSync(
  path.join(__dirname, '../src/viewer/index.js'),
   'utf-8'
);

// Extract a `function <name>(...) {…}` by balanced-brace scan. Robust to the
// file's quirky/mixed indentation (a column-0 `}` heuristic would mis-truncate
// nested or oddly-indented bodies).
function extractFn(src, name) {
  const re = new RegExp('function ' + name + '\\s*\\(');
  const m = re.exec(src);
  if (!m) throw new Error('not found: ' + name);
  const start = m.index;
  let i = src.indexOf('{', start), depth = 0;
  for (; i < src.length; i++) {
    const c = src[i];
    if (c === '{') depth++;
    else if (c === '}') { depth--; if (depth === 0) return src.slice(start, i + 1); }
   }
  throw new Error('unbalanced: ' + name);
}

// The direct-media URL, shaped exactly like the rejected case (a trailing slash +
// ?rnd= query). This is the concrete case the predicate used to reject.
const DIRECT_MP4 =
  'https://example.com/get_file/5/6cae22f8b3ba9d74f10cc2f8d060af738c8d30ddf5/14644000/14644093/14644093.mp4/?rnd=1787811205406';

describe('HLS QA: simple <video> direct-media predicate', () => {
  // Extract and evaluate the REAL isValidManifestUrl from index.js in a browser
  // window so its `new URL(...)` / string behavior matches production.
  const isValidSrc = extractFn(indexSource, 'isValidManifestUrl');

  function assertValid(url, expectValid, label) {
    const dom = new JSDOM(
      '<!DOCTYPE html><html><body></body></html>',
       { url: 'https://example.com/videos/x/', runScripts: 'dangerously' }
     );
    const got = dom.window.eval('(' + isValidSrc + ')(' + JSON.stringify(url) + ')');
    if (got !== expectValid) {
      throw new Error(
        `${label || 'isValidManifestUrl'}(${url}) = ${got}, expected ${expectValid}`
       );
    }
  }

  it('accepts a real direct-MP4 URL WITH a trailing slash + ?rnd=', () => {
    // Before the fix, `path.endsWith('/')` rejected this real media source.
    assertValid(DIRECT_MP4, true, 'trailing-slash MP4');
   });

  it('accepts a direct .mp4 with no trailing slash', () => {
    assertValid('https://cdn.example.com/vids/abc123.mp4', true, 'plain MP4');
   });

  it('accepts real .m3u8 / .webm / .mpd / .m4s / /hls/ media signals', () => {
    assertValid('https://cdn.example.com/live/master.m3u8', true, 'master.m3u8');
    assertValid('https://cdn.example.com/seg/webm/v.webm', true, 'webm');
    assertValid('https://cdn.example.com/seg/dash/master.mpd', true, 'mpd');
    assertValid('https://cdn.example.com/seg/1-0.m4s', true, 'm4s');
    assertValid('https://cdn.example.com/hls/abc/playlist.m3u8', true, '/hls/');
   });

  it('does not over-reject a real .m3u8 carrying a valid %2B/%2F token', () => {
     // %2B/%2F are valid percent-encoded hex pairs — a genuine auth token, not a
    // route-template placeholder, so it must survive.
    assertValid('https://cdn.example.com/live/master.m3u8?h=a%2Bb%2Fc&e=1234', true, '%2B token');
   });

  it('still rejects route-template / nav URLs (no over-acceptance)', () => {
     // The fix must not trade one regression for the other: the route templates
    // a provider's pageContext.langUrls leaks must still be rejected.
    assertValid('https://example.com/playlists/', false, '/playlists/');
    assertValid('https://example.com/playlists/liked/', false, '/playlists/liked/');
    assertValid('https://example.com/playlist/%1%/video/%2%/', false, '/playlist/%1%/');
    assertValid('https://example.com/members/%ID%/playlists/', false, '/members/%ID%/playlists/');
    assertValid('https://example.com/videos/%DIR%/', false, '/videos/%DIR%/');
   });
});

describe('HLS QA: scrapeTab isolated-world fallback', () => {
  // Extract the whole scrapeTab function (balanced braces, indentation-agnostic).
  const scrapeSrc = extractFn(indexSource, 'scrapeTab');

  it('runs an isolated-world scrape attempt (no `world` key) as a fallback', () => {
    const calls = scrapeSrc.match(/executeScript\(\s*\{[\s\S]*?\}\s*\)/g) || [];
    if (calls.length === 0) {
      throw new Error('scrapeTab has no executeScript call');
     }
     // At least one executeScript must omit `world` so it runs in the always-
    // permitted isolated world (recovers simple videos a MAIN-world injection
    // rejects on sandboxed/opaque-origin ad iframes).
    const hasIsolated = calls.some((c) => !/\bworld\s*:/.test(c));
    if (!hasIsolated) {
      throw new Error(
        'scrapeTab has no isolated-world (no `world:`) executeScript fallback — ' +
         'a MAIN-world+allFrames rejection on an ad-iframe-heavy page leaves 0 videos'
       );
    }
   });

  it('primary scrape still uses world: \'MAIN\' (reads page globals)', () => {
    const calls = scrapeSrc.match(/executeScript\(\s*\{[\s\S]*?\}\s*\)/g) || [];
    const hasMain = calls.some((c) => /world\s*:\s*['"]MAIN['"]/.test(c));
    if (!hasMain) {
      throw new Error(
        'scrapeTab primary attempt must use world: "MAIN" to read page globals ' +
         '(interceptor data, Hls instances, custom <video> props)'
       );
    }
   });
});