From e960c3c223acadb2fac81fb68595d902cf21e349 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sergey=20M=E2=80=A4?= Date: Wed, 4 May 2016 22:25:39 +0600 Subject: [PATCH] [yandexmusic:playlist] Improve extraction (Closes #6801) --- youtube_dl/extractor/yandexmusic.py | 40 +++++++++++++++++++---------- 1 file changed, 26 insertions(+), 14 deletions(-) diff --git a/youtube_dl/extractor/yandexmusic.py b/youtube_dl/extractor/yandexmusic.py index ce3723b55..22050add3 100644 --- a/youtube_dl/extractor/yandexmusic.py +++ b/youtube_dl/extractor/yandexmusic.py @@ -177,7 +177,7 @@ class YandexMusicAlbumIE(YandexMusicPlaylistBaseIE): class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE): IE_NAME = 'yandexmusic:playlist' IE_DESC = 'Яндекс.Музыка - Плейлист' - _VALID_URL = r'https?://music\.yandex\.(?:ru|kz|ua|by)/users/[^/]+/playlists/(?P\d+)' + _VALID_URL = r'https?://music\.yandex\.(?Pru|kz|ua|by)/users/(?P[^/]+)/playlists/(?P\d+)' _TESTS = [{ 'url': 'http://music.yandex.ru/users/music.partners/playlists/1245', @@ -201,19 +201,32 @@ class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE): }] def _real_extract(self, url): - playlist_id = self._match_id(url) - - webpage = self._download_webpage(url, playlist_id) - - mu = self._parse_json( - self._search_regex( - r'var\s+Mu\s*=\s*({.+?});\s*', webpage, 'player'), - playlist_id) + mobj = re.match(self._VALID_URL, url) + tld = mobj.group('tld') + user = mobj.group('user') + playlist_id = mobj.group('id') + + playlist = self._download_json( + 'https://music.yandex.%s/handlers/playlist.jsx' % tld, + playlist_id, 'Downloading missing tracks JSON', + fatal=False, + headers={ + 'Referer': url, + 'X-Requested-With': 'XMLHttpRequest', + 'X-Retpath-Y': url, + }, + query={ + 'owner': user, + 'kinds': playlist_id, + 'light': 'true', + 'lang': tld, + 'external-domain': 'music.yandex.%s' % tld, + 'overembed': 'false', + })['playlist'] - playlist = mu['pageData']['playlist'] tracks, track_ids = playlist['tracks'], playlist['trackIds'] - # tracks dictionary shipped with webpage is limited to 150 tracks, + # tracks dictionary shipped with playlist.jsx API is limited to 150 tracks, # missing tracks should be retrieved manually. if len(tracks) < len(track_ids): present_track_ids = set([compat_str(track['id']) for track in tracks if track.get('id')]) @@ -222,10 +235,9 @@ class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE): 'https://music.yandex.ru/handlers/track-entries.jsx', urlencode_postdata({ 'entries': ','.join(missing_track_ids), - 'lang': mu.get('settings', {}).get('lang', 'en'), - 'external-domain': 'music.yandex.ru', + 'lang': tld, + 'external-domain': 'music.yandex.%s' % tld, 'overembed': 'false', - 'sign': mu.get('authData', {}).get('user', {}).get('sign'), 'strict': 'true', })) request.add_header('Referer', url)