yt_dlp/extractor/genius.py

   1 from .common import InfoExtractor
   2 from ..utils import (
   3     ExtractorError,
   4     js_to_json,
   5     smuggle_url,
   6     str_or_none,
   7     traverse_obj,
   8     unescapeHTML,
   9 )
  10
  11
  12 class GeniusIE(InfoExtractor):
  13     _VALID_URL = r'https?://(?:www\.)?genius\.com/(?:videos|(?P<article>a))/(?P<id>[^?/#]+)'
  14     _TESTS = [{
  15         'url': 'https://genius.com/videos/Vince-staples-breaks-down-the-meaning-of-when-sparks-fly',
  16         'md5': '64c2ad98cfafcfda23bfa0ad0c512f4c',
  17         'info_dict': {
  18             'id': '6313303597112',
  19             'ext': 'mp4',
  20             'title': 'Vince Staples Breaks Down The Meaning Of “When Sparks Fly”',
  21             'description': 'md5:bc15e00342c537c0039d414423ae5752',
  22             'tags': 'count:1',
  23             'uploader_id': '4863540648001',
  24             'duration': 388.416,
  25             'upload_date': '20221005',
  26             'timestamp': 1664982341,
  27             'thumbnail': r're:^https?://.*\.jpg$',
  28         },
  29     }, {
  30         'url': 'https://genius.com/videos/Breaking-down-drakes-certified-lover-boy-kanye-beef-way-2-sexy-cudi',
  31         'md5': 'b8ed87a5efd1473bd027c20a969d4060',
  32         'info_dict': {
  33             'id': '6271792014001',
  34             'ext': 'mp4',
  35             'title': 'md5:c6355f7fa8a70bc86492a3963919fc15',
  36             'description': 'md5:1774638c31548b31b037c09e9b821393',
  37             'tags': 'count:3',
  38             'uploader_id': '4863540648001',
  39             'duration': 2685.099,
  40             'upload_date': '20210909',
  41             'timestamp': 1631209167,
  42             'thumbnail': r're:^https?://.*\.jpg$',
  43         },
  44     }, {
  45         'url': 'https://genius.com/a/cordae-anderson-paak-break-down-the-meaning-of-two-tens',
  46         'md5': 'f98a4e03b16b0a2821bd6e52fb3cc9d7',
  47         'info_dict': {
  48             'id': '6321509903112',
  49             'ext': 'mp4',
  50             'title': 'Cordae & Anderson .Paak Breaks Down The Meaning Of “Two Tens”',
  51             'description': 'md5:1255f0e1161d07342ce56a8464ac339d',
  52             'tags': ['song id: 5457554'],
  53             'uploader_id': '4863540648001',
  54             'duration': 361.813,
  55             'upload_date': '20230301',
  56             'timestamp': 1677703908,
  57             'thumbnail': r're:^https?://.*\.jpg$',
  58         },
  59     }]
  60
  61     def _real_extract(self, url):
  62         display_id, is_article = self._match_valid_url(url).group('id', 'article')
  63         webpage = self._download_webpage(url, display_id)
  64
  65         metadata = self._search_json(
  66             r'<meta content="', webpage, 'metadata', display_id,
  67             end_pattern=r'"\s+itemprop="page_data"', transform_source=unescapeHTML)
  68         video_id = traverse_obj(metadata, (
  69             (('article', 'media', ...), ('video', None)),
  70             ('provider_id', ('dfp_kv', lambda _, v: v['name'] == 'brightcove_video_id', 'values', ...))),
  71             get_all=False)
  72         if not video_id:
  73             # Not all article pages have videos, expect the error
  74             raise ExtractorError('Brightcove video ID not found in webpage', expected=bool(is_article))
  75
  76         config = self._search_json(r'var\s*APP_CONFIG\s*=', webpage, 'config', video_id, default={})
  77         account_id = config.get('brightcove_account_id', '4863540648001')
  78         player_id = traverse_obj(
  79             config, 'brightcove_standard_web_player_id', 'brightcove_standard_no_autoplay_web_player_id',
  80             'brightcove_modal_web_player_id', 'brightcove_song_story_web_player_id', default='S1ZcmcOC1x')
  81
  82         return self.url_result(
  83             smuggle_url(
  84                 f'https://players.brightcove.net/{account_id}/{player_id}_default/index.html?videoId={video_id}',
  85                 {'referrer': url}), 'BrightcoveNew', video_id)
  86
  87
  88 class GeniusLyricsIE(InfoExtractor):
  89     _VALID_URL = r'https?://(?:www\.)?genius\.com/(?P<id>[^?/#]+)-lyrics(?:[?/#]|$)'
  90     _TESTS = [{
  91         'url': 'https://genius.com/Lil-baby-heyy-lyrics',
  92         'playlist_mincount': 2,
  93         'info_dict': {
  94             'id': '8454545',
  95             'title': 'Heyy',
  96             'description': 'Heyy by Lil Baby',
  97         },
  98     }, {
  99         'url': 'https://genius.com/Outkast-two-dope-boyz-in-a-cadillac-lyrics',
 100         'playlist_mincount': 1,
 101         'info_dict': {
 102             'id': '36239',
 103             'title': 'Two Dope Boyz (In a Cadillac)',
 104             'description': 'Two Dope Boyz (In a Cadillac) by OutKast',
 105         },
 106     }, {
 107         'url': 'https://genius.com/Playboi-carti-rip-lyrics',
 108         'playlist_mincount': 1,
 109         'info_dict': {
 110             'id': '3710582',
 111             'title': 'R.I.P.',
 112             'description': 'R.I.P. by Playboi Carti',
 113         },
 114     }]
 115
 116     def _real_extract(self, url):
 117         display_id = self._match_id(url)
 118         webpage = self._download_webpage(url, display_id)
 119
 120         json_string = self._search_json(
 121             r'window\.__PRELOADED_STATE__\s*=\s*JSON\.parse\(', webpage, 'json string',
 122             display_id, transform_source=js_to_json, contains_pattern=r'\'{(?s:.+)}\'')
 123         song_info = self._parse_json(json_string, display_id)
 124         song_id = str_or_none(traverse_obj(song_info, ('songPage', 'song')))
 125         if not song_id:
 126             raise ExtractorError('Song id not found in webpage')
 127
 128         title = traverse_obj(
 129             song_info, ('songPage', 'trackingData', lambda _, x: x['key'] == 'Title', 'value'),
 130             get_all=False, default='untitled')
 131         artist = traverse_obj(
 132             song_info, ('songPage', 'trackingData', lambda _, x: x['key'] == 'Primary Artist', 'value'),
 133             get_all=False, default='unknown artist')
 134         media = traverse_obj(
 135             song_info, ('entities', 'songs', song_id, 'media'), expected_type=list, default=[])
 136
 137         entries = []
 138         for m in media:
 139             if m.get('type') in ('video', 'audio') and m.get('url'):
 140                 if m.get('provider') == 'spotify':
 141                     self.to_screen(f'{song_id}: Skipping Spotify audio embed')
 142                 else:
 143                     entries.append(self.url_result(m['url']))
 144
 145         return self.playlist_result(entries, song_id, title, f'{title} by {artist}')