fix(engine,producer,lint): resolve <source> children for media extract and localize (#3238)

Parent src-only scans skipped multi-format <video>/<audio> markup, so those
elements were never extracted, downloaded, or mixed and rendered blank/silent.
Lint now accepts a child <source src> as a resolvable media src.

packages/producer/src/services/htmlCompiler.test.ts

@@ -1574,6 +1574,25 @@ describe("localizeRemoteMediaSources", () => {
     expect(remoteMediaAssets.size).toBe(0);
   });
 
+  it("localizes remote <source> children of a <video>", async () => {
+    const orig = globalThis.fetch;
+    // eslint-disable-next-line @typescript-eslint/no-explicit-any
+    (globalThis as any).fetch = async () => validTestMediaResponse();
+    try {
+      const dl = mkdtempSync(join(tmpdir(), "hf-dl-src-"));
+      const html = `<video id="rec" data-start="0" data-end="5" muted>
+<source src="https://src-ok.example.com/rec.mp4" type="video/mp4">
+<source src="https://src-ok.example.com/rec.webm" type="video/webm">
+</video>`;
+      const { html: result, remoteMediaAssets } = await localizeRemoteMediaSources(html, dl);
+      expect(result).not.toContain("https://src-ok.example.com/");
+      expect(remoteMediaAssets.size).toBe(2);
+    } finally {
+      // eslint-disable-next-line @typescript-eslint/no-explicit-any
+      (globalThis as any).fetch = orig;
+    }
+  });
+
   it("rewrites src in both double-quoted and single-quoted attributes", async () => {
     const orig = globalThis.fetch;
     // eslint-disable-next-line @typescript-eslint/no-explicit-any

packages/producer/src/services/htmlCompiler.ts

@@ -1280,6 +1280,8 @@ const REMOTE_MEDIA_SUBDIR = "_remote_media";
 // have `>` inside quoted attribute values (data-title etc.).
 const REMOTE_MEDIA_TAG_RE =
   /<(?:video|audio)\b[^>]*?\bsrc\s*=\s*["'](https?:\/\/[^"']+)["'][^>]*>/gi;
+// <source src> on media elements (picture uses srcset, not src).
+const REMOTE_SOURCE_TAG_RE = /<source\b[^>]*?\bsrc\s*=\s*["'](https?:\/\/[^"']+)["'][^>]*>/gi;
 // Match <img> tags (including agent-pipeline-emitted variants where `src` is
 // not the first attribute). Producer-side localisation is the primary fix for
 // the remote-<img> flicker; frameCapture's `pollImagesReady`/`decodeAllImages`
@@ -1350,10 +1352,11 @@ async function downloadAndRewriteUrls(
 }
 
 /**
- * Download any remote `src` URLs on `<video>` and `<audio>` elements into a
- * local subdirectory of `downloadDir`, rewrite the HTML src attributes to
- * relative paths, and return the updated HTML along with a map of
- * `{ relativePath → absoluteLocalPath }` for callers to add to `externalAssets`.
+ * Download any remote `src` URLs on `<video>` / `<audio>` elements and their
+ * `<source>` children into a local subdirectory of `downloadDir`, rewrite the
+ * HTML src attributes to relative paths, and return the updated HTML along with
+ * a map of `{ relativePath → absoluteLocalPath }` for callers to add to
+ * `externalAssets`.
  *
  * Skips URLs that fail to download (warns and preserves the original URL so
  * the browser can still attempt the remote fetch as a fallback).
@@ -1369,12 +1372,13 @@ export async function localizeRemoteMediaSources(
   html: string,
   downloadDir: string,
 ): Promise<{ html: string; remoteMediaAssets: Map<string, string> }> {
-  // Collect unique HTTP URLs from <video>/<audio> src attributes.
   const urlSet = new Set<string>();
-  const re = new RegExp(REMOTE_MEDIA_TAG_RE.source, REMOTE_MEDIA_TAG_RE.flags);
-  let m: RegExpExecArray | null;
-  while ((m = re.exec(html)) !== null) {
-    if (m[1]) urlSet.add(m[1]);
+  for (const tagRe of [REMOTE_MEDIA_TAG_RE, REMOTE_SOURCE_TAG_RE]) {
+    const re = new RegExp(tagRe.source, tagRe.flags);
+    let m: RegExpExecArray | null;
+    while ((m = re.exec(html)) !== null) {
+      if (m[1]) urlSet.add(m[1]);
+    }
   }
   return downloadAndRewriteUrls(
     urlSet,