yt_dlp/extractor/jiosaavn.py

@@ -1,23 +1,30 @@
 from .common import InfoExtractor
 from ..utils import (
     js_to_json,
-    url_or_none,
     urlencode_postdata,
-    urljoin,
 )
-from ..utils.traversal import traverse_obj
 
 
-class JioSaavnBaseIE(InfoExtractor):
+class JioSaavnBaseInfoExtractor(InfoExtractor):
-    def _extract_initial_data(self, url, audio_id):
+    def extract_initial_data_as_json(self, url, id):
-        webpage = self._download_webpage(url, audio_id)
+        webpage = self._download_webpage(url, id)
-        return self._search_json(
+        init_data_re = r'window.__INITIAL_DATA__\s*=\s*(?P<json>.+?);*\s*\</script>'
-            r'window\.__INITIAL_DATA__\s*=', webpage,
+
-            'init json', audio_id, transform_source=js_to_json)
+        init_data = self._parse_json(self._search_regex(init_data_re, webpage,
+                                                        'init-json'),
+                                     id, transform_source=js_to_json)
+        return init_data
 
 
-class JioSaavnSongIE(JioSaavnBaseIE):
+class JioSaavnSongIE(JioSaavnBaseInfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?(?:jiosaavn\.com/song/[^/?#]+/|saavn\.com/s/song/(?:[^/?#]+/){3})(?P<id>[^/?#]+)'
+    _VALID_URL = r'''(?x)
+                    https?://(?:www\.)?
+                        (?:
+                            jiosaavn\.com/song/[^/]+/|
+                            saavn.com/s/song/(?:[^/]+/){3}
+                        )
+                        (?P<id>[^/]+)
+                   '''
     _TESTS = [{
         'url': 'https://www.jiosaavn.com/song/leja-re/OQsEfQFVUXk',
         'md5': '7b1f70de088ede3a152ea34aece4df42',
@@ -35,30 +42,43 @@ class JioSaavnSongIE(JioSaavnBaseIE):
 
     def _real_extract(self, url):
         audio_id = self._match_id(url)
-        song_data = self._extract_initial_data(url, audio_id)['song']['song']
+        song_data = self.extract_initial_data_as_json(url, audio_id)['song']['song']
-        media_data = self._download_json(
+
-            'https://www.jiosaavn.com/api.php', audio_id, data=urlencode_postdata({
+        data = urlencode_postdata({'__call': 'song.generateAuthToken',
-                '__call': 'song.generateAuthToken',
+                                   '_format': 'json',
-                '_format': 'json',
+                                   'bitrate': '128',
-                'bitrate': '128',
+                                   'url': song_data['encrypted_media_url'],
-                'url': song_data['encrypted_media_url'],
+                                   })
-            }))
+
+        def clean_api_json(resp):
+            return self._search_regex(r'(?P<json>\{.+?})', resp, 'api-json')
+
+        media_url = self._download_json('https://www.jiosaavn.com/api.php',
+                                        audio_id, data=data,
+                                        transform_source=clean_api_json,
+                                        )['auth_url']
 
         return {
             'id': audio_id,
-            'url': media_data['auth_url'],
+            'title': song_data['title']['text'],
-            'ext': media_data.get('type'),
+            'formats': [{
-            'vcodec': 'none',
+                'url': media_url,
-            **traverse_obj(song_data, {
+                'ext': 'mp3',
-                'title': ('title', 'text'),
+                'format_note': 'MPEG audio',
-                'album': ('album', 'text'),
+                'format_id': '128',
-                'thumbnail': ('image', 0, {url_or_none}),
+                'vcodec': 'none',
-            }),
+            }],
+            'album': song_data.get('album', {}).get('text'),
+            'thumbnail': song_data.get('image', [None])[0],
         }
 
 
-class JioSaavnAlbumIE(JioSaavnBaseIE):
+class JioSaavnAlbumIE(JioSaavnBaseInfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?(?:jio)?saavn\.com/album/[^/?#]+/(?P<id>[^/?#]+)'
+    _VALID_URL = r'''(?x)
+                    https?://(?:www\.)?
+                        (?:(?:jio)?saavn\.com/album/[^/]+/)
+                        (?P<id>[^/]+)
+                   '''
     _TESTS = [{
         'url': 'https://www.jiosaavn.com/album/96/buIOjYZDrNA_',
         'info_dict': {
@@ -70,10 +90,9 @@ class JioSaavnAlbumIE(JioSaavnBaseIE):
 
     def _real_extract(self, url):
         album_id = self._match_id(url)
-        album_view = self._extract_initial_data(url, album_id)['albumView']
+        json_data = self.extract_initial_data_as_json(url, album_id)
+        album_data = json_data['albumView']['album']
+        songs = json_data['albumView']['modules'][0]['data']
+        songs = [self.url_result(f'https://www.jiosaavn.com{song["title"]["action"]}') for song in songs]
 
-        return self.playlist_from_matches(
+        return self.playlist_result(songs, album_id, album_data.get('title', {}).get('text'))
-            traverse_obj(album_view, (
-                'modules', lambda _, x: x['key'] == 'list', 'data', ..., 'title', 'action', {str})),
-            album_id, traverse_obj(album_view, ('album', 'title', 'text', {str})), ie=JioSaavnSongIE,
-            getter=lambda x: urljoin('https://www.jiosaavn.com/', x))