Detect RSS podcast media enclosures

Author: lazyGPT07

plugins/feed-discovery/src/feed-discovery.test.ts

@@ -36,6 +36,26 @@ describe("feed parsing", () => {
     expect(atom.ok).toBe(true);
     expect(json.ok).toBe(true);
   });
+
+  it("classifies RSS feeds with audio enclosures as podcasts", () => {
+    const result = parseFeedDocument(
+      "https://example.com/podcast.xml",
+      `<?xml version="1.0"?>
+      <rss version="2.0" xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">
+        <channel>
+          <title>Example Podcast</title>
+          <itunes:author>Example Host</itunes:author>
+          <item>
+            <title>Episode 1</title>
+            <enclosure url="https://example.com/episode-1.mp3" type="audio/mpeg" />
+          </item>
+        </channel>
+      </rss>`
+    );
+
+    expect(result.ok).toBe(true);
+    expect(result.kind).toBe("podcast");
+  });
 });
 
 describe("site probing helpers", () => {

plugins/feed-discovery/src/feed-parsing.ts

@@ -96,7 +96,7 @@ function parseRss(feedUrl: string, rss: Record<string, unknown>): ValidationResu
     title: title || "Untitled RSS Feed",
     description: stringValue(channel.description),
     homepageUrl: linkValue(channel.link),
-    kind: detectRssKind(channel),
+    kind: detectRssKind(channel, items),
     language: stringValue(channel.language),
     imageUrl: imageValue(channel.image),
     lastPublishedAt: newestDate([stringValue(channel.lastBuildDate), stringValue(channel.pubDate), ...sampleItems.map((item) => item.publishedAt)]),
@@ -132,10 +132,21 @@ function parseAtom(feedUrl: string, feed: Record<string, unknown>): ValidationRe
   };
 }
 
-function detectRssKind(channel: Record<string, unknown>): FeedKind {
+function detectRssKind(channel: Record<string, unknown>, items: Record<string, unknown>[]): FeedKind {
   if (channel.itunes || channel["itunes:author"] || channel.enclosure) {
     return "podcast";
   }
+  const hasMediaEnclosure = items.some((item) =>
+    asArray(item.enclosure)
+      .filter(isRecord)
+      .some((enclosure) => {
+        const type = stringValue(enclosure.type)?.toLowerCase();
+        return type?.startsWith("audio/") || type?.startsWith("video/");
+      })
+  );
+  if (hasMediaEnclosure) {
+    return "podcast";
+  }
   return "blog";
 }