search: fix episode parsing

- `parse_top_result`: handle episode data
- `parse_search_result`: ensure `resultType` is "episode", not "video"
- `parse_search_result`: offset runs to skip the "Episode" specifier

Author: sgvictorino

tests/mixins/test_search.py

@@ -112,13 +112,36 @@ def test_search_filters(self, yt_auth):
         assert len(results) >= 3
         assert all(item["resultType"] == "episode" for item in results)
 
-    def test_search_top_result(self, yt):
+    def test_search_episode_category(self, yt):
+        """Test resultType detection for episodes by searching for a podcast without a filter."""
+        results = yt.search("Stanford Graduate School of Business")
+        episode = next(
+            item
+            for item in results
+            if item["category"] == "Episodes" and item["podcast"]["name"] == "Stanford GSB Podcasts"
+        )
+        assert episode["resultType"] == "episode"
+        assert episode["podcast"]["id"] == "MPSPPLxq_lXOUlvQDUNyoBYLkN8aVt5yAwEtG9"
+
+    def test_search_top_result_playlist(self, yt):
         results = yt.search("fdsfsfsd")  # issue 524
         assert results[0]["category"] == "Top result"
         assert results[0]["resultType"] == "playlist"
         assert results[0]["playlistId"].startswith("PL")
         assert len(results[0]["author"]) > 0
 
+    def test_search_top_result_episode(self, yt):
+        results = yt.search(
+            "Stanford GSB Podcasts 124. Making Meetings Meaningful, Pt. 1: How to Structure and Organize More Effective Gatherings"
+        )
+        assert results[0]["category"] == "Top result"
+        assert results[0]["resultType"] == "episode"
+        assert results[0]["videoId"] == "KNkyHCLOr1o"
+        assert results[0]["podcast"] == {
+            "id": "MPSPPLxq_lXOUlvQDUNyoBYLkN8aVt5yAwEtG9",
+            "name": "Stanford GSB Podcasts",
+        }
+
     def test_search_uploads(self, config, yt, yt_oauth):
         with pytest.raises(Exception, match="No filter can be set when searching uploads"):
             yt.search(

ytmusicapi/navigation.py

@@ -59,7 +59,8 @@
 THUMBNAIL = ["thumbnail", "thumbnails"]
 THUMBNAILS = ["thumbnail", "musicThumbnailRenderer", *THUMBNAIL]
 THUMBNAIL_RENDERER = ["thumbnailRenderer", "musicThumbnailRenderer", *THUMBNAIL]
-THUMBNAIL_OVERLAY = ["thumbnailOverlay", *OVERLAY_RENDERER, "playNavigationEndpoint", *WATCH_PID]
+THUMBNAIL_OVERLAY_NAVIGATION = ["thumbnailOverlay", *OVERLAY_RENDERER, "playNavigationEndpoint"]
+THUMBNAIL_OVERLAY = [*THUMBNAIL_OVERLAY_NAVIGATION, *WATCH_PID]
 THUMBNAIL_CROPPED = ["thumbnail", "croppedSquareThumbnailRenderer", *THUMBNAIL]
 FEEDBACK_TOKEN = ["feedbackEndpoint", "feedbackToken"]
 BADGE_PATH = [0, "musicInlineBadgeRenderer", "accessibilityData", "accessibilityData", "label"]

ytmusicapi/parsers/search.py

@@ -67,6 +67,13 @@ def parse_top_result(data: JsonDict, search_result_types: list[str]) -> JsonDict
         search_result["title"] = nav(data, TITLE_TEXT)
         search_result["author"] = parse_song_artists_runs(nav(data, ["subtitle", "runs"])[2:])
 
+    if result_type in ["episode"]:
+        search_result["videoId"] = nav(data, [*THUMBNAIL_OVERLAY_NAVIGATION, *WATCH_VIDEO_ID])
+        search_result["videoType"] = nav(data, [*THUMBNAIL_OVERLAY_NAVIGATION, *NAVIGATION_VIDEO_TYPE])
+        runs = nav(data, SUBTITLE_RUNS)[2:]
+        search_result["date"] = runs[0]["text"]
+        search_result["podcast"] = parse_id_name(runs[2])
+
     search_result["thumbnails"] = nav(data, THUMBNAILS, True)
     return search_result
 
@@ -96,7 +103,10 @@ def parse_search_result(
                 iter(type for prefix, type in mapping.items() if browse_id.startswith(prefix)), None
             )
         else:
-            result_type = "song" if video_type == "MUSIC_VIDEO_TYPE_ATV" else "video"
+            result_type = {
+                "MUSIC_VIDEO_TYPE_ATV": "song",
+                "MUSIC_VIDEO_TYPE_PODCAST_EPISODE": "episode",
+            }.get(video_type or "", "video")
 
     search_result["resultType"] = result_type
 
@@ -191,12 +201,13 @@ def parse_search_result(
 
     if result_type in ["episode"]:
         flex_item = get_flex_column_item(data, 1)
-        has_date = int(len(nav(flex_item, TEXT_RUNS)) > 1)
+        runs = nav(flex_item, TEXT_RUNS)[default_offset:]
+        has_date = int(len(runs) > 1)
         search_result["live"] = bool(nav(data, ["badges", 0, "liveBadgeRenderer"], True))
         if has_date:
-            search_result["date"] = nav(flex_item, TEXT_RUN_TEXT)
+            search_result["date"] = runs[0]["text"]
 
-        search_result["podcast"] = parse_id_name(nav(flex_item, ["text", "runs", has_date * 2]))
+        search_result["podcast"] = parse_id_name(runs[has_date * 2])
 
     search_result["thumbnails"] = nav(data, THUMBNAILS, True)