yt_dlp/extractor/archiveorg.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import json
   6 from .common import InfoExtractor
   7 from .youtube import YoutubeIE, YoutubeBaseInfoExtractor
   8 from ..compat import (
   9     compat_urllib_parse_unquote,
  10     compat_urllib_parse_unquote_plus,
  11     compat_HTTPError
  12 )
  13 from ..utils import (
  14     bug_reports_message,
  15     clean_html,
  16     dict_get,
  17     extract_attributes,
  18     ExtractorError,
  19     get_element_by_id,
  20     HEADRequest,
  21     int_or_none,
  22     join_nonempty,
  23     KNOWN_EXTENSIONS,
  24     merge_dicts,
  25     mimetype2ext,
  26     orderedSet,
  27     parse_duration,
  28     parse_qs,
  29     str_to_int,
  30     str_or_none,
  31     traverse_obj,
  32     try_get,
  33     unified_strdate,
  34     unified_timestamp,
  35     urlhandle_detect_ext,
  36     url_or_none
  37 )
  38
  39
  40 class ArchiveOrgIE(InfoExtractor):
  41     IE_NAME = 'archive.org'
  42     IE_DESC = 'archive.org video and audio'
  43     _VALID_URL = r'https?://(?:www\.)?archive\.org/(?:details|embed)/(?P<id>[^?#]+)(?:[?].*)?$'
  44     _TESTS = [{
  45         'url': 'http://archive.org/details/XD300-23_68HighlightsAResearchCntAugHumanIntellect',
  46         'md5': '8af1d4cf447933ed3c7f4871162602db',
  47         'info_dict': {
  48             'id': 'XD300-23_68HighlightsAResearchCntAugHumanIntellect',
  49             'ext': 'ogv',
  50             'title': '1968 Demo - FJCC Conference Presentation Reel #1',
  51             'description': 'md5:da45c349df039f1cc8075268eb1b5c25',
  52             'release_date': '19681210',
  53             'timestamp': 1268695290,
  54             'upload_date': '20100315',
  55             'creator': 'SRI International',
  56             'uploader': 'laura@archive.org',
  57         },
  58     }, {
  59         'url': 'https://archive.org/details/Cops1922',
  60         'md5': '0869000b4ce265e8ca62738b336b268a',
  61         'info_dict': {
  62             'id': 'Cops1922',
  63             'ext': 'mp4',
  64             'title': 'Buster Keaton\'s "Cops" (1922)',
  65             'description': 'md5:43a603fd6c5b4b90d12a96b921212b9c',
  66             'uploader': 'yorkmba99@hotmail.com',
  67             'timestamp': 1387699629,
  68             'upload_date': '20131222',
  69         },
  70     }, {
  71         'url': 'http://archive.org/embed/XD300-23_68HighlightsAResearchCntAugHumanIntellect',
  72         'only_matching': True,
  73     }, {
  74         'url': 'https://archive.org/details/Election_Ads',
  75         'md5': '284180e857160cf866358700bab668a3',
  76         'info_dict': {
  77             'id': 'Election_Ads/Commercial-JFK1960ElectionAdCampaignJingle.mpg',
  78             'title': 'Commercial-JFK1960ElectionAdCampaignJingle.mpg',
  79             'ext': 'mp4',
  80         },
  81     }, {
  82         'url': 'https://archive.org/details/Election_Ads/Commercial-Nixon1960ElectionAdToughonDefense.mpg',
  83         'md5': '7915213ef02559b5501fe630e1a53f59',
  84         'info_dict': {
  85             'id': 'Election_Ads/Commercial-Nixon1960ElectionAdToughonDefense.mpg',
  86             'title': 'Commercial-Nixon1960ElectionAdToughonDefense.mpg',
  87             'ext': 'mp4',
  88             'timestamp': 1205588045,
  89             'uploader': 'mikedavisstripmaster@yahoo.com',
  90             'description': '1960 Presidential Campaign Election Commercials John F Kennedy, Richard M Nixon',
  91             'upload_date': '20080315',
  92         },
  93     }, {
  94         'url': 'https://archive.org/details/gd1977-05-08.shure57.stevenson.29303.flac16',
  95         'md5': '7d07ffb42aba6537c28e053efa4b54c9',
  96         'info_dict': {
  97             'id': 'gd1977-05-08.shure57.stevenson.29303.flac16/gd1977-05-08d01t01.flac',
  98             'title': 'Turning',
  99             'ext': 'flac',
 100         },
 101     }, {
 102         'url': 'https://archive.org/details/gd1977-05-08.shure57.stevenson.29303.flac16/gd1977-05-08d01t07.flac',
 103         'md5': 'a07cd8c6ab4ee1560f8a0021717130f3',
 104         'info_dict': {
 105             'id': 'gd1977-05-08.shure57.stevenson.29303.flac16/gd1977-05-08d01t07.flac',
 106             'title': 'Deal',
 107             'ext': 'flac',
 108             'timestamp': 1205895624,
 109             'uploader': 'mvernon54@yahoo.com',
 110             'description': 'md5:6a31f1996db0aa0fc9da6d6e708a1bb0',
 111             'upload_date': '20080319',
 112             'location': 'Barton Hall - Cornell University',
 113         },
 114     }, {
 115         'url': 'https://archive.org/details/lp_the-music-of-russia_various-artists-a-askaryan-alexander-melik',
 116         'md5': '7cb019baa9b332e82ea7c10403acd180',
 117         'info_dict': {
 118             'id': 'lp_the-music-of-russia_various-artists-a-askaryan-alexander-melik/disc1/01.01. Bells Of Rostov.mp3',
 119             'title': 'Bells Of Rostov',
 120             'ext': 'mp3',
 121         },
 122     }, {
 123         'url': 'https://archive.org/details/lp_the-music-of-russia_various-artists-a-askaryan-alexander-melik/disc1/02.02.+Song+And+Chorus+In+The+Polovetsian+Camp+From+%22Prince+Igor%22+(Act+2%2C+Scene+1).mp3',
 124         'md5': '1d0aabe03edca83ca58d9ed3b493a3c3',
 125         'info_dict': {
 126             'id': 'lp_the-music-of-russia_various-artists-a-askaryan-alexander-melik/disc1/02.02. Song And Chorus In The Polovetsian Camp From "Prince Igor" (Act 2, Scene 1).mp3',
 127             'title': 'Song And Chorus In The Polovetsian Camp From "Prince Igor" (Act 2, Scene 1)',
 128             'ext': 'mp3',
 129             'timestamp': 1569662587,
 130             'uploader': 'associate-joygen-odiongan@archive.org',
 131             'description': 'md5:012b2d668ae753be36896f343d12a236',
 132             'upload_date': '20190928',
 133         },
 134     }]
 135
 136     @staticmethod
 137     def _playlist_data(webpage):
 138         element = re.findall(r'''(?xs)
 139             <input
 140             (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
 141             \s+class=['"]?js-play8-playlist['"]?
 142             (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
 143             \s*/>
 144         ''', webpage)[0]
 145
 146         return json.loads(extract_attributes(element)['value'])
 147
 148     def _real_extract(self, url):
 149         video_id = compat_urllib_parse_unquote_plus(self._match_id(url))
 150         identifier, entry_id = (video_id.split('/', 1) + [None])[:2]
 151
 152         # Archive.org metadata API doesn't clearly demarcate playlist entries
 153         # or subtitle tracks, so we get them from the embeddable player.
 154         embed_page = self._download_webpage(f'https://archive.org/embed/{identifier}', identifier)
 155         playlist = self._playlist_data(embed_page)
 156
 157         entries = {}
 158         for p in playlist:
 159             # If the user specified a playlist entry in the URL, ignore the
 160             # rest of the playlist.
 161             if entry_id and p['orig'] != entry_id:
 162                 continue
 163
 164             entries[p['orig']] = {
 165                 'formats': [],
 166                 'thumbnails': [],
 167                 'artist': p.get('artist'),
 168                 'track': p.get('title'),
 169                 'subtitles': {},
 170             }
 171
 172             for track in p.get('tracks', []):
 173                 if track['kind'] != 'subtitles':
 174                     continue
 175                 entries[p['orig']][track['label']] = {
 176                     'url': 'https://archive.org/' + track['file'].lstrip('/')
 177                 }
 178
 179         metadata = self._download_json('http://archive.org/metadata/' + identifier, identifier)
 180         m = metadata['metadata']
 181         identifier = m['identifier']
 182
 183         info = {
 184             'id': identifier,
 185             'title': m['title'],
 186             'description': clean_html(m.get('description')),
 187             'uploader': dict_get(m, ['uploader', 'adder']),
 188             'creator': m.get('creator'),
 189             'license': m.get('licenseurl'),
 190             'release_date': unified_strdate(m.get('date')),
 191             'timestamp': unified_timestamp(dict_get(m, ['publicdate', 'addeddate'])),
 192             'webpage_url': f'https://archive.org/details/{identifier}',
 193             'location': m.get('venue'),
 194             'release_year': int_or_none(m.get('year'))}
 195
 196         for f in metadata['files']:
 197             if f['name'] in entries:
 198                 entries[f['name']] = merge_dicts(entries[f['name']], {
 199                     'id': identifier + '/' + f['name'],
 200                     'title': f.get('title') or f['name'],
 201                     'display_id': f['name'],
 202                     'description': clean_html(f.get('description')),
 203                     'creator': f.get('creator'),
 204                     'duration': parse_duration(f.get('length')),
 205                     'track_number': int_or_none(f.get('track')),
 206                     'album': f.get('album'),
 207                     'discnumber': int_or_none(f.get('disc')),
 208                     'release_year': int_or_none(f.get('year'))})
 209                 entry = entries[f['name']]
 210             elif traverse_obj(f, 'original', expected_type=str) in entries:
 211                 entry = entries[f['original']]
 212             else:
 213                 continue
 214
 215             if f.get('format') == 'Thumbnail':
 216                 entry['thumbnails'].append({
 217                     'id': f['name'],
 218                     'url': 'https://archive.org/download/' + identifier + '/' + f['name'],
 219                     'width': int_or_none(f.get('width')),
 220                     'height': int_or_none(f.get('width')),
 221                     'filesize': int_or_none(f.get('size'))})
 222
 223             extension = (f['name'].rsplit('.', 1) + [None])[1]
 224             if extension in KNOWN_EXTENSIONS:
 225                 entry['formats'].append({
 226                     'url': 'https://archive.org/download/' + identifier + '/' + f['name'],
 227                     'format': f.get('format'),
 228                     'width': int_or_none(f.get('width')),
 229                     'height': int_or_none(f.get('height')),
 230                     'filesize': int_or_none(f.get('size')),
 231                     'protocol': 'https'})
 232
 233         for entry in entries.values():
 234             self._sort_formats(entry['formats'])
 235
 236         if len(entries) == 1:
 237             # If there's only one item, use it as the main info dict
 238             only_video = next(iter(entries.values()))
 239             if entry_id:
 240                 info = merge_dicts(only_video, info)
 241             else:
 242                 info = merge_dicts(info, only_video)
 243         else:
 244             # Otherwise, we have a playlist.
 245             info['_type'] = 'playlist'
 246             info['entries'] = list(entries.values())
 247
 248         if metadata.get('reviews'):
 249             info['comments'] = []
 250             for review in metadata['reviews']:
 251                 info['comments'].append({
 252                     'id': review.get('review_id'),
 253                     'author': review.get('reviewer'),
 254                     'text': str_or_none(review.get('reviewtitle'), '') + '\n\n' + review.get('reviewbody'),
 255                     'timestamp': unified_timestamp(review.get('createdate')),
 256                     'parent': 'root'})
 257
 258         return info
 259
 260
 261 class YoutubeWebArchiveIE(InfoExtractor):
 262     IE_NAME = 'web.archive:youtube'
 263     IE_DESC = 'web.archive.org saved youtube videos, "ytarchive:" prefix'
 264     _VALID_URL = r'''(?x)(?:(?P<prefix>ytarchive:)|
 265             (?:https?://)?web\.archive\.org/
 266             (?:web/)?(?:(?P<date>[0-9]{14})?[0-9A-Za-z_*]*/)?  # /web and the version index is optional
 267             (?:https?(?::|%3[Aa])//)?(?:
 268                 (?:\w+\.)?youtube\.com(?::(?:80|443))?/watch(?:\.php)?(?:\?|%3[fF])(?:[^\#]+(?:&|%26))?v(?:=|%3[dD])  # Youtube URL
 269                 |(?:wayback-fakeurl\.archive\.org/yt/)  # Or the internal fake url
 270             )
 271         )(?P<id>[0-9A-Za-z_-]{11})
 272         (?(prefix)
 273             (?::(?P<date2>[0-9]{14}))?$|
 274             (?:%26|[#&]|$)
 275         )'''
 276
 277     _TESTS = [
 278         {
 279             'url': 'https://web.archive.org/web/20150415002341/https://www.youtube.com/watch?v=aYAGB11YrSs',
 280             'info_dict': {
 281                 'id': 'aYAGB11YrSs',
 282                 'ext': 'webm',
 283                 'title': 'Team Fortress 2 - Sandviches!',
 284                 'description': 'md5:4984c0f9a07f349fc5d8e82ab7af4eaf',
 285                 'upload_date': '20110926',
 286                 'uploader': 'Zeurel',
 287                 'channel_id': 'UCukCyHaD-bK3in_pKpfH9Eg',
 288                 'duration': 32,
 289                 'uploader_id': 'Zeurel',
 290                 'uploader_url': 'http://www.youtube.com/user/Zeurel'
 291             }
 292         }, {
 293             # Internal link
 294             'url': 'https://web.archive.org/web/2oe/http://wayback-fakeurl.archive.org/yt/97t7Xj_iBv0',
 295             'info_dict': {
 296                 'id': '97t7Xj_iBv0',
 297                 'ext': 'mp4',
 298                 'title': 'Why Machines That Bend Are Better',
 299                 'description': 'md5:00404df2c632d16a674ff8df1ecfbb6c',
 300                 'upload_date': '20190312',
 301                 'uploader': 'Veritasium',
 302                 'channel_id': 'UCHnyfMqiRRG1u-2MsSQLbXA',
 303                 'duration': 771,
 304                 'uploader_id': '1veritasium',
 305                 'uploader_url': 'http://www.youtube.com/user/1veritasium'
 306             }
 307         }, {
 308             # Video from 2012, webm format itag 45. Newest capture is deleted video, with an invalid description.
 309             # Should use the date in the link. Title ends with '- Youtube'. Capture has description in eow-description
 310             'url': 'https://web.archive.org/web/20120712231619/http://www.youtube.com/watch?v=AkhihxRKcrs&gl=US&hl=en',
 311             'info_dict': {
 312                 'id': 'AkhihxRKcrs',
 313                 'ext': 'webm',
 314                 'title': 'Limited Run: Mondo\'s Modern Classic 1 of 3 (SDCC 2012)',
 315                 'upload_date': '20120712',
 316                 'duration': 398,
 317                 'description': 'md5:ff4de6a7980cb65d951c2f6966a4f2f3',
 318                 'uploader_id': 'machinima',
 319                 'uploader_url': 'http://www.youtube.com/user/machinima'
 320             }
 321         }, {
 322             # FLV video. Video file URL does not provide itag information
 323             'url': 'https://web.archive.org/web/20081211103536/http://www.youtube.com/watch?v=jNQXAC9IVRw',
 324             'info_dict': {
 325                 'id': 'jNQXAC9IVRw',
 326                 'ext': 'flv',
 327                 'title': 'Me at the zoo',
 328                 'upload_date': '20050423',
 329                 'channel_id': 'UC4QobU6STFB0P71PMvOGN5A',
 330                 'duration': 19,
 331                 'description': 'md5:10436b12e07ac43ff8df65287a56efb4',
 332                 'uploader_id': 'jawed',
 333                 'uploader_url': 'http://www.youtube.com/user/jawed'
 334             }
 335         }, {
 336             'url': 'https://web.archive.org/web/20110712231407/http://www.youtube.com/watch?v=lTx3G6h2xyA',
 337             'info_dict': {
 338                 'id': 'lTx3G6h2xyA',
 339                 'ext': 'flv',
 340                 'title': 'Madeon - Pop Culture (live mashup)',
 341                 'upload_date': '20110711',
 342                 'uploader': 'Madeon',
 343                 'channel_id': 'UCqMDNf3Pn5L7pcNkuSEeO3w',
 344                 'duration': 204,
 345                 'description': 'md5:f7535343b6eda34a314eff8b85444680',
 346                 'uploader_id': 'itsmadeon',
 347                 'uploader_url': 'http://www.youtube.com/user/itsmadeon'
 348             }
 349         }, {
 350             # First capture is of dead video, second is the oldest from CDX response.
 351             'url': 'https://web.archive.org/https://www.youtube.com/watch?v=1JYutPM8O6E',
 352             'info_dict': {
 353                 'id': '1JYutPM8O6E',
 354                 'ext': 'mp4',
 355                 'title': 'Fake Teen Doctor Strikes AGAIN! - Weekly Weird News',
 356                 'upload_date': '20160218',
 357                 'channel_id': 'UCdIaNUarhzLSXGoItz7BHVA',
 358                 'duration': 1236,
 359                 'description': 'md5:21032bae736421e89c2edf36d1936947',
 360                 'uploader_id': 'MachinimaETC',
 361                 'uploader_url': 'http://www.youtube.com/user/MachinimaETC'
 362             }
 363         }, {
 364             # First capture of dead video, capture date in link links to dead capture.
 365             'url': 'https://web.archive.org/web/20180803221945/https://www.youtube.com/watch?v=6FPhZJGvf4E',
 366             'info_dict': {
 367                 'id': '6FPhZJGvf4E',
 368                 'ext': 'mp4',
 369                 'title': 'WTF: Video Games Still Launch BROKEN?! - T.U.G.S.',
 370                 'upload_date': '20160219',
 371                 'channel_id': 'UCdIaNUarhzLSXGoItz7BHVA',
 372                 'duration': 798,
 373                 'description': 'md5:a1dbf12d9a3bd7cb4c5e33b27d77ffe7',
 374                 'uploader_id': 'MachinimaETC',
 375                 'uploader_url': 'http://www.youtube.com/user/MachinimaETC'
 376             },
 377             'expected_warnings': [
 378                 r'unable to download capture webpage \(it may not be archived\)'
 379             ]
 380         }, {   # Very old YouTube page, has - YouTube in title.
 381             'url': 'http://web.archive.org/web/20070302011044/http://youtube.com/watch?v=-06-KB9XTzg',
 382             'info_dict': {
 383                 'id': '-06-KB9XTzg',
 384                 'ext': 'flv',
 385                 'title': 'New Coin Hack!! 100% Safe!!'
 386             }
 387         }, {
 388             'url': 'web.archive.org/https://www.youtube.com/watch?v=dWW7qP423y8',
 389             'info_dict': {
 390                 'id': 'dWW7qP423y8',
 391                 'ext': 'mp4',
 392                 'title': 'It\'s Bootleg AirPods Time.',
 393                 'upload_date': '20211021',
 394                 'channel_id': 'UC7Jwj9fkrf1adN4fMmTkpug',
 395                 'channel_url': 'http://www.youtube.com/channel/UC7Jwj9fkrf1adN4fMmTkpug',
 396                 'duration': 810,
 397                 'description': 'md5:7b567f898d8237b256f36c1a07d6d7bc',
 398                 'uploader': 'DankPods',
 399                 'uploader_id': 'UC7Jwj9fkrf1adN4fMmTkpug',
 400                 'uploader_url': 'http://www.youtube.com/channel/UC7Jwj9fkrf1adN4fMmTkpug'
 401             }
 402         }, {
 403             # player response contains '};' See: https://github.com/ytdl-org/youtube-dl/issues/27093
 404             'url': 'https://web.archive.org/web/20200827003909if_/http://www.youtube.com/watch?v=6Dh-RL__uN4',
 405             'info_dict': {
 406                 'id': '6Dh-RL__uN4',
 407                 'ext': 'mp4',
 408                 'title': 'bitch lasagna',
 409                 'upload_date': '20181005',
 410                 'channel_id': 'UC-lHJZR3Gqxm24_Vd_AJ5Yw',
 411                 'channel_url': 'http://www.youtube.com/channel/UC-lHJZR3Gqxm24_Vd_AJ5Yw',
 412                 'duration': 135,
 413                 'description': 'md5:2dbe4051feeff2dab5f41f82bb6d11d0',
 414                 'uploader': 'PewDiePie',
 415                 'uploader_id': 'PewDiePie',
 416                 'uploader_url': 'http://www.youtube.com/user/PewDiePie'
 417             }
 418         }, {
 419             'url': 'https://web.archive.org/web/http://www.youtube.com/watch?v=kH-G_aIBlFw',
 420             'only_matching': True
 421         }, {
 422             'url': 'https://web.archive.org/web/20050214000000_if/http://www.youtube.com/watch?v=0altSZ96U4M',
 423             'only_matching': True
 424         }, {
 425             # Video not archived, only capture is unavailable video page
 426             'url': 'https://web.archive.org/web/20210530071008/https://www.youtube.com/watch?v=lHJTf93HL1s&spfreload=10',
 427             'only_matching': True
 428         }, {   # Encoded url
 429             'url': 'https://web.archive.org/web/20120712231619/http%3A//www.youtube.com/watch%3Fgl%3DUS%26v%3DAkhihxRKcrs%26hl%3Den',
 430             'only_matching': True
 431         }, {
 432             'url': 'https://web.archive.org/web/20120712231619/http%3A//www.youtube.com/watch%3Fv%3DAkhihxRKcrs%26gl%3DUS%26hl%3Den',
 433             'only_matching': True
 434         }, {
 435             'url': 'https://web.archive.org/web/20060527081937/http://www.youtube.com:80/watch.php?v=ELTFsLT73fA&amp;search=soccer',
 436             'only_matching': True
 437         }, {
 438             'url': 'https://web.archive.org/http://www.youtube.com:80/watch?v=-05VVye-ffg',
 439             'only_matching': True
 440         }, {
 441             'url': 'ytarchive:BaW_jenozKc:20050214000000',
 442             'only_matching': True
 443         }, {
 444             'url': 'ytarchive:BaW_jenozKc',
 445             'only_matching': True
 446         },
 447     ]
 448     _YT_INITIAL_DATA_RE = r'(?:(?:(?:window\s*\[\s*["\']ytInitialData["\']\s*\]|ytInitialData)\s*=\s*({.+?})\s*;)|%s)' % YoutubeBaseInfoExtractor._YT_INITIAL_DATA_RE
 449     _YT_INITIAL_PLAYER_RESPONSE_RE = r'(?:(?:(?:window\s*\[\s*["\']ytInitialPlayerResponse["\']\s*\]|ytInitialPlayerResponse)\s*=[(\s]*({.+?})[)\s]*;)|%s)' % YoutubeBaseInfoExtractor._YT_INITIAL_PLAYER_RESPONSE_RE
 450     _YT_INITIAL_BOUNDARY_RE = r'(?:(?:var\s+meta|</script|\n)|%s)' % YoutubeBaseInfoExtractor._YT_INITIAL_BOUNDARY_RE
 451
 452     _YT_DEFAULT_THUMB_SERVERS = ['i.ytimg.com']  # thumbnails most likely archived on these servers
 453     _YT_ALL_THUMB_SERVERS = orderedSet(
 454         _YT_DEFAULT_THUMB_SERVERS + ['img.youtube.com', *[f'{c}{n or ""}.ytimg.com' for c in ('i', 's') for n in (*range(0, 5), 9)]])
 455
 456     _WAYBACK_BASE_URL = 'https://web.archive.org/web/%sif_/'
 457     _OLDEST_CAPTURE_DATE = 20050214000000
 458     _NEWEST_CAPTURE_DATE = 20500101000000
 459
 460     def _call_cdx_api(self, item_id, url, filters: list = None, collapse: list = None, query: dict = None, note=None, fatal=False):
 461         # CDX docs: https://github.com/internetarchive/wayback/blob/master/wayback-cdx-server/README.md
 462         query = {
 463             'url': url,
 464             'output': 'json',
 465             'fl': 'original,mimetype,length,timestamp',
 466             'limit': 500,
 467             'filter': ['statuscode:200'] + (filters or []),
 468             'collapse': collapse or [],
 469             **(query or {})
 470         }
 471         res = self._download_json(
 472             'https://web.archive.org/cdx/search/cdx', item_id,
 473             note or 'Downloading CDX API JSON', query=query, fatal=fatal)
 474         if isinstance(res, list) and len(res) >= 2:
 475             # format response to make it easier to use
 476             return list(dict(zip(res[0], v)) for v in res[1:])
 477         elif not isinstance(res, list) or len(res) != 0:
 478             self.report_warning('Error while parsing CDX API response' + bug_reports_message())
 479
 480     def _extract_yt_initial_variable(self, webpage, regex, video_id, name):
 481         return self._parse_json(self._search_regex(
 482             (r'%s\s*%s' % (regex, self._YT_INITIAL_BOUNDARY_RE),
 483              regex), webpage, name, default='{}'), video_id, fatal=False)
 484
 485     def _extract_webpage_title(self, webpage):
 486         page_title = self._html_extract_title(webpage, default='')
 487         # YouTube video pages appear to always have either 'YouTube -' as prefix or '- YouTube' as suffix.
 488         return self._html_search_regex(
 489             r'(?:YouTube\s*-\s*(.*)$)|(?:(.*)\s*-\s*YouTube$)',
 490             page_title, 'title', default='')
 491
 492     def _extract_metadata(self, video_id, webpage):
 493         search_meta = ((lambda x: self._html_search_meta(x, webpage, default=None)) if webpage else (lambda x: None))
 494         player_response = self._extract_yt_initial_variable(
 495             webpage, self._YT_INITIAL_PLAYER_RESPONSE_RE, video_id, 'initial player response') or {}
 496         initial_data = self._extract_yt_initial_variable(
 497             webpage, self._YT_INITIAL_DATA_RE, video_id, 'initial player response') or {}
 498
 499         initial_data_video = traverse_obj(
 500             initial_data, ('contents', 'twoColumnWatchNextResults', 'results', 'results', 'contents', ..., 'videoPrimaryInfoRenderer'),
 501             expected_type=dict, get_all=False, default={})
 502
 503         video_details = traverse_obj(
 504             player_response, 'videoDetails', expected_type=dict, get_all=False, default={})
 505
 506         microformats = traverse_obj(
 507             player_response, ('microformat', 'playerMicroformatRenderer'), expected_type=dict, get_all=False, default={})
 508
 509         video_title = (
 510             video_details.get('title')
 511             or YoutubeBaseInfoExtractor._get_text(microformats, 'title')
 512             or YoutubeBaseInfoExtractor._get_text(initial_data_video, 'title')
 513             or self._extract_webpage_title(webpage)
 514             or search_meta(['og:title', 'twitter:title', 'title']))
 515
 516         channel_id = str_or_none(
 517             video_details.get('channelId')
 518             or microformats.get('externalChannelId')
 519             or search_meta('channelId')
 520             or self._search_regex(
 521                 r'data-channel-external-id=(["\'])(?P<id>(?:(?!\1).)+)\1',  # @b45a9e6
 522                 webpage, 'channel id', default=None, group='id'))
 523         channel_url = f'http://www.youtube.com/channel/{channel_id}' if channel_id else None
 524
 525         duration = int_or_none(
 526             video_details.get('lengthSeconds')
 527             or microformats.get('lengthSeconds')
 528             or parse_duration(search_meta('duration')))
 529         description = (
 530             video_details.get('shortDescription')
 531             or YoutubeBaseInfoExtractor._get_text(microformats, 'description')
 532             or clean_html(get_element_by_id('eow-description', webpage))  # @9e6dd23
 533             or search_meta(['description', 'og:description', 'twitter:description']))
 534
 535         uploader = video_details.get('author')
 536
 537         # Uploader ID and URL
 538         uploader_mobj = re.search(
 539             r'<link itemprop="url" href="(?P<uploader_url>https?://www\.youtube\.com/(?:user|channel)/(?P<uploader_id>[^"]+))">',  # @fd05024
 540             webpage)
 541         if uploader_mobj is not None:
 542             uploader_id, uploader_url = uploader_mobj.group('uploader_id'), uploader_mobj.group('uploader_url')
 543         else:
 544             # @a6211d2
 545             uploader_url = url_or_none(microformats.get('ownerProfileUrl'))
 546             uploader_id = self._search_regex(
 547                 r'(?:user|channel)/([^/]+)', uploader_url or '', 'uploader id', default=None)
 548
 549         upload_date = unified_strdate(
 550             dict_get(microformats, ('uploadDate', 'publishDate'))
 551             or search_meta(['uploadDate', 'datePublished'])
 552             or self._search_regex(
 553                 [r'(?s)id="eow-date.*?>(.*?)</span>',
 554                  r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']'],  # @7998520
 555                 webpage, 'upload date', default=None))
 556
 557         return {
 558             'title': video_title,
 559             'description': description,
 560             'upload_date': upload_date,
 561             'uploader': uploader,
 562             'channel_id': channel_id,
 563             'channel_url': channel_url,
 564             'duration': duration,
 565             'uploader_url': uploader_url,
 566             'uploader_id': uploader_id,
 567         }
 568
 569     def _extract_thumbnails(self, video_id):
 570         try_all = 'thumbnails' in self._configuration_arg('check_all')
 571         thumbnail_base_urls = ['http://{server}/vi{webp}/{video_id}'.format(
 572             webp='_webp' if ext == 'webp' else '', video_id=video_id, server=server)
 573             for server in (self._YT_ALL_THUMB_SERVERS if try_all else self._YT_DEFAULT_THUMB_SERVERS) for ext in (('jpg', 'webp') if try_all else ('jpg',))]
 574
 575         thumbnails = []
 576         for url in thumbnail_base_urls:
 577             response = self._call_cdx_api(
 578                 video_id, url, filters=['mimetype:image/(?:webp|jpeg)'],
 579                 collapse=['urlkey'], query={'matchType': 'prefix'})
 580             if not response:
 581                 continue
 582             thumbnails.extend(
 583                 {
 584                     'url': (self._WAYBACK_BASE_URL % (int_or_none(thumbnail_dict.get('timestamp')) or self._OLDEST_CAPTURE_DATE)) + thumbnail_dict.get('original'),
 585                     'filesize': int_or_none(thumbnail_dict.get('length')),
 586                     'preference': int_or_none(thumbnail_dict.get('length'))
 587                 } for thumbnail_dict in response)
 588             if not try_all:
 589                 break
 590
 591         self._remove_duplicate_formats(thumbnails)
 592         return thumbnails
 593
 594     def _get_capture_dates(self, video_id, url_date):
 595         capture_dates = []
 596         # Note: CDX API will not find watch pages with extra params in the url.
 597         response = self._call_cdx_api(
 598             video_id, f'https://www.youtube.com/watch?v={video_id}',
 599             filters=['mimetype:text/html'], collapse=['timestamp:6', 'digest'], query={'matchType': 'prefix'}) or []
 600         all_captures = sorted([int_or_none(r['timestamp']) for r in response if int_or_none(r['timestamp']) is not None])
 601
 602         # Prefer the new polymer UI captures as we support extracting more metadata from them
 603         # WBM captures seem to all switch to this layout ~July 2020
 604         modern_captures = [x for x in all_captures if x >= 20200701000000]
 605         if modern_captures:
 606             capture_dates.append(modern_captures[0])
 607         capture_dates.append(url_date)
 608         if all_captures:
 609             capture_dates.append(all_captures[0])
 610
 611         if 'captures' in self._configuration_arg('check_all'):
 612             capture_dates.extend(modern_captures + all_captures)
 613
 614         # Fallbacks if any of the above fail
 615         capture_dates.extend([self._OLDEST_CAPTURE_DATE, self._NEWEST_CAPTURE_DATE])
 616         return orderedSet(filter(None, capture_dates))
 617
 618     def _real_extract(self, url):
 619         video_id, url_date, url_date_2 = self._match_valid_url(url).group('id', 'date', 'date2')
 620         url_date = url_date or url_date_2
 621
 622         urlh = None
 623         try:
 624             urlh = self._request_webpage(
 625                 HEADRequest('https://web.archive.org/web/2oe_/http://wayback-fakeurl.archive.org/yt/%s' % video_id),
 626                 video_id, note='Fetching archived video file url', expected_status=True)
 627         except ExtractorError as e:
 628             # HTTP Error 404 is expected if the video is not saved.
 629             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 404:
 630                 self.raise_no_formats(
 631                     'The requested video is not archived, indexed, or there is an issue with web.archive.org',
 632                     expected=True)
 633             else:
 634                 raise
 635
 636         capture_dates = self._get_capture_dates(video_id, int_or_none(url_date))
 637         self.write_debug('Captures to try: ' + join_nonempty(*capture_dates, delim=', '))
 638         info = {'id': video_id}
 639         for capture in capture_dates:
 640             webpage = self._download_webpage(
 641                 (self._WAYBACK_BASE_URL + 'http://www.youtube.com/watch?v=%s') % (capture, video_id),
 642                 video_id=video_id, fatal=False, errnote='unable to download capture webpage (it may not be archived)',
 643                 note='Downloading capture webpage')
 644             current_info = self._extract_metadata(video_id, webpage or '')
 645             # Try avoid getting deleted video metadata
 646             if current_info.get('title'):
 647                 info = merge_dicts(info, current_info)
 648                 if 'captures' not in self._configuration_arg('check_all'):
 649                     break
 650
 651         info['thumbnails'] = self._extract_thumbnails(video_id)
 652
 653         if urlh:
 654             url = compat_urllib_parse_unquote(urlh.geturl())
 655             video_file_url_qs = parse_qs(url)
 656             # Attempt to recover any ext & format info from playback url & response headers
 657             format = {'url': url, 'filesize': int_or_none(urlh.headers.get('x-archive-orig-content-length'))}
 658             itag = try_get(video_file_url_qs, lambda x: x['itag'][0])
 659             if itag and itag in YoutubeIE._formats:
 660                 format.update(YoutubeIE._formats[itag])
 661                 format.update({'format_id': itag})
 662             else:
 663                 mime = try_get(video_file_url_qs, lambda x: x['mime'][0])
 664                 ext = (mimetype2ext(mime)
 665                        or urlhandle_detect_ext(urlh)
 666                        or mimetype2ext(urlh.headers.get('x-archive-guessed-content-type')))
 667                 format.update({'ext': ext})
 668             info['formats'] = [format]
 669             if not info.get('duration'):
 670                 info['duration'] = str_to_int(try_get(video_file_url_qs, lambda x: x['dur'][0]))
 671
 672         if not info.get('title'):
 673             info['title'] = video_id
 674         return info