[tubitv] Fix/improve TV series extraction (#2829)

Authored by: bbepis

yt_dlp/extractor/tubitv.py

@@ -107,6 +107,9 @@ def _real_extract(self, url):
                 'url': self._proto_relative_url(sub_url),
             })
 
+        season_number, episode_number, episode_title = self._search_regex(
+            r'^S(\d+):E(\d+) - (.+)', title, 'episode info', fatal=False, group=(1, 2, 3), default=(None, None, None))
+
         return {
             'id': video_id,
             'title': title,
@@ -117,6 +120,9 @@ def _real_extract(self, url):
             'duration': int_or_none(video_data.get('duration')),
             'uploader_id': video_data.get('publisher_id'),
             'release_year': int_or_none(video_data.get('year')),
+            'season_number': int_or_none(season_number),
+            'episode_number': int_or_none(episode_number),
+            'episode_title': episode_title
         }
 
 
@@ -132,9 +138,11 @@ class TubiTvShowIE(InfoExtractor):
 
     def _entries(self, show_url, show_name):
         show_webpage = self._download_webpage(show_url, show_name)
+
         show_json = self._parse_json(self._search_regex(
-            r"window\.__data\s*=\s*({.+?});\s*</script>",
-            show_webpage, 'data',), show_name, transform_source=js_to_json)['video']
+            r'window\.__data\s*=\s*({[^<]+});\s*</script>',
+            show_webpage, 'data'), show_name, transform_source=js_to_json)['video']
+
         for episode_id in show_json['fullContentById'].keys():
             yield self.url_result(
                 'tubitv:%s' % episode_id,

yt_dlp/utils.py

@@ -3143,6 +3143,8 @@ def fix_kv(m):
 
         return '"%s"' % v
 
+    code = re.sub(r'new Date\((".+")\)', r'\g<1>', code)
+
     return re.sub(r'''(?sx)
         "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
         '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|