_ Git - youtube-dl/blob - youtube_dl/extractor/youtube.py

   1 # coding: utf-8
   2
   3 from __future__ import unicode_literals
   4
   5
   6 import itertools
   7 import json
   8 import os.path
   9 import random
  10 import re
  11 import time
  12 import traceback
  13
  14 from .common import InfoExtractor, SearchInfoExtractor
  15 from ..jsinterp import JSInterpreter
  16 from ..swfinterp import SWFInterpreter
  17 from ..compat import (
  18     compat_chr,
  19     compat_HTTPError,
  20     compat_kwargs,
  21     compat_parse_qs,
  22     compat_urllib_parse_unquote,
  23     compat_urllib_parse_unquote_plus,
  24     compat_urllib_parse_urlencode,
  25     compat_urllib_parse_urlparse,
  26     compat_urlparse,
  27     compat_str,
  28 )
  29 from ..utils import (
  30     bool_or_none,
  31     clean_html,
  32     dict_get,
  33     error_to_compat_str,
  34     ExtractorError,
  35     float_or_none,
  36     get_element_by_attribute,
  37     get_element_by_id,
  38     int_or_none,
  39     mimetype2ext,
  40     orderedSet,
  41     parse_codecs,
  42     parse_duration,
  43     qualities,
  44     remove_quotes,
  45     remove_start,
  46     smuggle_url,
  47     str_or_none,
  48     str_to_int,
  49     try_get,
  50     unescapeHTML,
  51     unified_strdate,
  52     unsmuggle_url,
  53     uppercase_escape,
  54     url_or_none,
  55     urlencode_postdata,
  56 )
  57
  58
  59 class YoutubeBaseInfoExtractor(InfoExtractor):
  60     """Provide base functions for Youtube extractors"""
  61     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
  62     _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge'
  63
  64     _LOOKUP_URL = 'https://accounts.google.com/_/signin/sl/lookup'
  65     _CHALLENGE_URL = 'https://accounts.google.com/_/signin/sl/challenge'
  66     _TFA_URL = 'https://accounts.google.com/_/signin/challenge?hl=en&TL={0}'
  67
  68     _NETRC_MACHINE = 'youtube'
  69     # If True it will raise an error if no login info is provided
  70     _LOGIN_REQUIRED = False
  71
  72     _PLAYLIST_ID_RE = r'(?:PL|LL|EC|UU|FL|RD|UL|TL|OLAK5uy_)[0-9A-Za-z-_]{10,}'
  73
  74     def _set_language(self):
  75         self._set_cookie(
  76             '.youtube.com', 'PREF', 'f1=50000000&hl=en',
  77             # YouTube sets the expire time to about two months
  78             expire_time=time.time() + 2 * 30 * 24 * 3600)
  79
  80     def _ids_to_results(self, ids):
  81         return [
  82             self.url_result(vid_id, 'Youtube', video_id=vid_id)
  83             for vid_id in ids]
  84
  85     def _login(self):
  86         """
  87         Attempt to log in to YouTube.
  88         True is returned if successful or skipped.
  89         False is returned if login failed.
  90
  91         If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
  92         """
  93         username, password = self._get_login_info()
  94         # No authentication to be performed
  95         if username is None:
  96             if self._LOGIN_REQUIRED and self._downloader.params.get('cookiefile') is None:
  97                 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
  98             return True
  99
 100         login_page = self._download_webpage(
 101             self._LOGIN_URL, None,
 102             note='Downloading login page',
 103             errnote='unable to fetch login page', fatal=False)
 104         if login_page is False:
 105             return
 106
 107         login_form = self._hidden_inputs(login_page)
 108
 109         def req(url, f_req, note, errnote):
 110             data = login_form.copy()
 111             data.update({
 112                 'pstMsg': 1,
 113                 'checkConnection': 'youtube',
 114                 'checkedDomains': 'youtube',
 115                 'hl': 'en',
 116                 'deviceinfo': '[null,null,null,[],null,"US",null,null,[],"GlifWebSignIn",null,[null,null,[]]]',
 117                 'f.req': json.dumps(f_req),
 118                 'flowName': 'GlifWebSignIn',
 119                 'flowEntry': 'ServiceLogin',
 120                 # TODO: reverse actual botguard identifier generation algo
 121                 'bgRequest': '["identifier",""]',
 122             })
 123             return self._download_json(
 124                 url, None, note=note, errnote=errnote,
 125                 transform_source=lambda s: re.sub(r'^[^[]*', '', s),
 126                 fatal=False,
 127                 data=urlencode_postdata(data), headers={
 128                     'Content-Type': 'application/x-www-form-urlencoded;charset=utf-8',
 129                     'Google-Accounts-XSRF': 1,
 130                 })
 131
 132         def warn(message):
 133             self._downloader.report_warning(message)
 134
 135         lookup_req = [
 136             username,
 137             None, [], None, 'US', None, None, 2, False, True,
 138             [
 139                 None, None,
 140                 [2, 1, None, 1,
 141                  'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn',
 142                  None, [], 4],
 143                 1, [None, None, []], None, None, None, True
 144             ],
 145             username,
 146         ]
 147
 148         lookup_results = req(
 149             self._LOOKUP_URL, lookup_req,
 150             'Looking up account info', 'Unable to look up account info')
 151
 152         if lookup_results is False:
 153             return False
 154
 155         user_hash = try_get(lookup_results, lambda x: x[0][2], compat_str)
 156         if not user_hash:
 157             warn('Unable to extract user hash')
 158             return False
 159
 160         challenge_req = [
 161             user_hash,
 162             None, 1, None, [1, None, None, None, [password, None, True]],
 163             [
 164                 None, None, [2, 1, None, 1, 'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn', None, [], 4],
 165                 1, [None, None, []], None, None, None, True
 166             ]]
 167
 168         challenge_results = req(
 169             self._CHALLENGE_URL, challenge_req,
 170             'Logging in', 'Unable to log in')
 171
 172         if challenge_results is False:
 173             return
 174
 175         login_res = try_get(challenge_results, lambda x: x[0][5], list)
 176         if login_res:
 177             login_msg = try_get(login_res, lambda x: x[5], compat_str)
 178             warn(
 179                 'Unable to login: %s' % 'Invalid password'
 180                 if login_msg == 'INCORRECT_ANSWER_ENTERED' else login_msg)
 181             return False
 182
 183         res = try_get(challenge_results, lambda x: x[0][-1], list)
 184         if not res:
 185             warn('Unable to extract result entry')
 186             return False
 187
 188         login_challenge = try_get(res, lambda x: x[0][0], list)
 189         if login_challenge:
 190             challenge_str = try_get(login_challenge, lambda x: x[2], compat_str)
 191             if challenge_str == 'TWO_STEP_VERIFICATION':
 192                 # SEND_SUCCESS - TFA code has been successfully sent to phone
 193                 # QUOTA_EXCEEDED - reached the limit of TFA codes
 194                 status = try_get(login_challenge, lambda x: x[5], compat_str)
 195                 if status == 'QUOTA_EXCEEDED':
 196                     warn('Exceeded the limit of TFA codes, try later')
 197                     return False
 198
 199                 tl = try_get(challenge_results, lambda x: x[1][2], compat_str)
 200                 if not tl:
 201                     warn('Unable to extract TL')
 202                     return False
 203
 204                 tfa_code = self._get_tfa_info('2-step verification code')
 205
 206                 if not tfa_code:
 207                     warn(
 208                         'Two-factor authentication required. Provide it either interactively or with --twofactor <code>'
 209                         '(Note that only TOTP (Google Authenticator App) codes work at this time.)')
 210                     return False
 211
 212                 tfa_code = remove_start(tfa_code, 'G-')
 213
 214                 tfa_req = [
 215                     user_hash, None, 2, None,
 216                     [
 217                         9, None, None, None, None, None, None, None,
 218                         [None, tfa_code, True, 2]
 219                     ]]
 220
 221                 tfa_results = req(
 222                     self._TFA_URL.format(tl), tfa_req,
 223                     'Submitting TFA code', 'Unable to submit TFA code')
 224
 225                 if tfa_results is False:
 226                     return False
 227
 228                 tfa_res = try_get(tfa_results, lambda x: x[0][5], list)
 229                 if tfa_res:
 230                     tfa_msg = try_get(tfa_res, lambda x: x[5], compat_str)
 231                     warn(
 232                         'Unable to finish TFA: %s' % 'Invalid TFA code'
 233                         if tfa_msg == 'INCORRECT_ANSWER_ENTERED' else tfa_msg)
 234                     return False
 235
 236                 check_cookie_url = try_get(
 237                     tfa_results, lambda x: x[0][-1][2], compat_str)
 238             else:
 239                 CHALLENGES = {
 240                     'LOGIN_CHALLENGE': "This device isn't recognized. For your security, Google wants to make sure it's really you.",
 241                     'USERNAME_RECOVERY': 'Please provide additional information to aid in the recovery process.',
 242                     'REAUTH': "There is something unusual about your activity. For your security, Google wants to make sure it's really you.",
 243                 }
 244                 challenge = CHALLENGES.get(
 245                     challenge_str,
 246                     '%s returned error %s.' % (self.IE_NAME, challenge_str))
 247                 warn('%s\nGo to https://accounts.google.com/, login and solve a challenge.' % challenge)
 248                 return False
 249         else:
 250             check_cookie_url = try_get(res, lambda x: x[2], compat_str)
 251
 252         if not check_cookie_url:
 253             warn('Unable to extract CheckCookie URL')
 254             return False
 255
 256         check_cookie_results = self._download_webpage(
 257             check_cookie_url, None, 'Checking cookie', fatal=False)
 258
 259         if check_cookie_results is False:
 260             return False
 261
 262         if 'https://myaccount.google.com/' not in check_cookie_results:
 263             warn('Unable to log in')
 264             return False
 265
 266         return True
 267
 268     def _download_webpage_handle(self, *args, **kwargs):
 269         query = kwargs.get('query', {}).copy()
 270         query['disable_polymer'] = 'true'
 271         kwargs['query'] = query
 272         return super(YoutubeBaseInfoExtractor, self)._download_webpage_handle(
 273             *args, **compat_kwargs(kwargs))
 274
 275     def _real_initialize(self):
 276         if self._downloader is None:
 277             return
 278         self._set_language()
 279         if not self._login():
 280             return
 281
 282
 283 class YoutubeEntryListBaseInfoExtractor(YoutubeBaseInfoExtractor):
 284     # Extract entries from page with "Load more" button
 285     def _entries(self, page, playlist_id):
 286         more_widget_html = content_html = page
 287         for page_num in itertools.count(1):
 288             for entry in self._process_page(content_html):
 289                 yield entry
 290
 291             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
 292             if not mobj:
 293                 break
 294
 295             count = 0
 296             retries = 3
 297             while count <= retries:
 298                 try:
 299                     # Downloading page may result in intermittent 5xx HTTP error
 300                     # that is usually worked around with a retry
 301                     more = self._download_json(
 302                         'https://youtube.com/%s' % mobj.group('more'), playlist_id,
 303                         'Downloading page #%s%s'
 304                         % (page_num, ' (retry #%d)' % count if count else ''),
 305                         transform_source=uppercase_escape)
 306                     break
 307                 except ExtractorError as e:
 308                     if isinstance(e.cause, compat_HTTPError) and e.cause.code in (500, 503):
 309                         count += 1
 310                         if count <= retries:
 311                             continue
 312                     raise
 313
 314             content_html = more['content_html']
 315             if not content_html.strip():
 316                 # Some webpages show a "Load more" button but they don't
 317                 # have more videos
 318                 break
 319             more_widget_html = more['load_more_widget_html']
 320
 321
 322 class YoutubePlaylistBaseInfoExtractor(YoutubeEntryListBaseInfoExtractor):
 323     def _process_page(self, content):
 324         for video_id, video_title in self.extract_videos_from_page(content):
 325             yield self.url_result(video_id, 'Youtube', video_id, video_title)
 326
 327     def extract_videos_from_page(self, page):
 328         ids_in_page = []
 329         titles_in_page = []
 330         for mobj in re.finditer(self._VIDEO_RE, page):
 331             # The link with index 0 is not the first video of the playlist (not sure if still actual)
 332             if 'index' in mobj.groupdict() and mobj.group('id') == '0':
 333                 continue
 334             video_id = mobj.group('id')
 335             video_title = unescapeHTML(mobj.group('title'))
 336             if video_title:
 337                 video_title = video_title.strip()
 338             try:
 339                 idx = ids_in_page.index(video_id)
 340                 if video_title and not titles_in_page[idx]:
 341                     titles_in_page[idx] = video_title
 342             except ValueError:
 343                 ids_in_page.append(video_id)
 344                 titles_in_page.append(video_title)
 345         return zip(ids_in_page, titles_in_page)
 346
 347
 348 class YoutubePlaylistsBaseInfoExtractor(YoutubeEntryListBaseInfoExtractor):
 349     def _process_page(self, content):
 350         for playlist_id in orderedSet(re.findall(
 351                 r'<h3[^>]+class="[^"]*yt-lockup-title[^"]*"[^>]*><a[^>]+href="/?playlist\?list=([0-9A-Za-z-_]{10,})"',
 352                 content)):
 353             yield self.url_result(
 354                 'https://www.youtube.com/playlist?list=%s' % playlist_id, 'YoutubePlaylist')
 355
 356     def _real_extract(self, url):
 357         playlist_id = self._match_id(url)
 358         webpage = self._download_webpage(url, playlist_id)
 359         title = self._og_search_title(webpage, fatal=False)
 360         return self.playlist_result(self._entries(webpage, playlist_id), playlist_id, title)
 361
 362
 363 class YoutubeIE(YoutubeBaseInfoExtractor):
 364     IE_DESC = 'YouTube.com'
 365     _VALID_URL = r"""(?x)^
 366                      (
 367                          (?:https?://|//)                                    # http(s):// or protocol-independent URL
 368                          (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
 369                             (?:www\.)?deturl\.com/www\.youtube\.com/|
 370                             (?:www\.)?pwnyoutube\.com/|
 371                             (?:www\.)?hooktube\.com/|
 372                             (?:www\.)?yourepeat\.com/|
 373                             tube\.majestyc\.net/|
 374                             # Invidious instances taken from https://github.com/omarroth/invidious/wiki/Invidious-Instances
 375                             (?:(?:www|dev)\.)?invidio\.us/|
 376                             (?:(?:www|no)\.)?invidiou\.sh/|
 377                             (?:(?:www|fi|de)\.)?invidious\.snopyta\.org/|
 378                             (?:www\.)?invidious\.kabi\.tk/|
 379                             (?:www\.)?invidious\.enkirton\.net/|
 380                             (?:www\.)?invidious\.13ad\.de/|
 381                             (?:www\.)?tube\.poal\.co/|
 382                             (?:www\.)?vid\.wxzm\.sx/|
 383                             youtube\.googleapis\.com/)                        # the various hostnames, with wildcard subdomains
 384                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
 385                          (?:                                                  # the various things that can precede the ID:
 386                              (?:(?:v|embed|e)/(?!videoseries))                # v/ or embed/ or e/
 387                              |(?:                                             # or the v= param in all its forms
 388                                  (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)?  # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
 389                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
 390                                  (?:.*?[&;])??                                # any other preceding param (like /?s=tuff&v=xxxx or ?s=tuff&amp;v=V36LpHqtcDY)
 391                                  v=
 392                              )
 393                          ))
 394                          |(?:
 395                             youtu\.be|                                        # just youtu.be/xxxx
 396                             vid\.plus|                                        # or vid.plus/xxxx
 397                             zwearz\.com/watch|                                # or zwearz.com/watch/xxxx
 398                          )/
 399                          |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
 400                          )
 401                      )?                                                       # all until now is optional -> you can pass the naked ID
 402                      ([0-9A-Za-z_-]{11})                                      # here is it! the YouTube video ID
 403                      (?!.*?\blist=
 404                         (?:
 405                             %(playlist_id)s|                                  # combined list/video URLs are handled by the playlist IE
 406                             WL                                                # WL are handled by the watch later IE
 407                         )
 408                      )
 409                      (?(1).+)?                                                # if we found the ID, everything can follow
 410                      $""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
 411     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
 412     _formats = {
 413         '5': {'ext': 'flv', 'width': 400, 'height': 240, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'},
 414         '6': {'ext': 'flv', 'width': 450, 'height': 270, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'},
 415         '13': {'ext': '3gp', 'acodec': 'aac', 'vcodec': 'mp4v'},
 416         '17': {'ext': '3gp', 'width': 176, 'height': 144, 'acodec': 'aac', 'abr': 24, 'vcodec': 'mp4v'},
 417         '18': {'ext': 'mp4', 'width': 640, 'height': 360, 'acodec': 'aac', 'abr': 96, 'vcodec': 'h264'},
 418         '22': {'ext': 'mp4', 'width': 1280, 'height': 720, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'},
 419         '34': {'ext': 'flv', 'width': 640, 'height': 360, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 420         '35': {'ext': 'flv', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 421         # itag 36 videos are either 320x180 (BaW_jenozKc) or 320x240 (__2ABJjxzNo), abr varies as well
 422         '36': {'ext': '3gp', 'width': 320, 'acodec': 'aac', 'vcodec': 'mp4v'},
 423         '37': {'ext': 'mp4', 'width': 1920, 'height': 1080, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'},
 424         '38': {'ext': 'mp4', 'width': 4096, 'height': 3072, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'},
 425         '43': {'ext': 'webm', 'width': 640, 'height': 360, 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8'},
 426         '44': {'ext': 'webm', 'width': 854, 'height': 480, 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8'},
 427         '45': {'ext': 'webm', 'width': 1280, 'height': 720, 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8'},
 428         '46': {'ext': 'webm', 'width': 1920, 'height': 1080, 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8'},
 429         '59': {'ext': 'mp4', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 430         '78': {'ext': 'mp4', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 431
 432
 433         # 3D videos
 434         '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -20},
 435         '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -20},
 436         '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264', 'preference': -20},
 437         '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264', 'preference': -20},
 438         '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8', 'preference': -20},
 439         '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8', 'preference': -20},
 440         '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8', 'preference': -20},
 441
 442         # Apple HTTP Live Streaming
 443         '91': {'ext': 'mp4', 'height': 144, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10},
 444         '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10},
 445         '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -10},
 446         '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -10},
 447         '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 256, 'vcodec': 'h264', 'preference': -10},
 448         '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 256, 'vcodec': 'h264', 'preference': -10},
 449         '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10},
 450         '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 24, 'vcodec': 'h264', 'preference': -10},
 451
 452         # DASH mp4 video
 453         '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'vcodec': 'h264'},
 454         '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'vcodec': 'h264'},
 455         '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'h264'},
 456         '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'h264'},
 457         '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'h264'},
 458         '138': {'ext': 'mp4', 'format_note': 'DASH video', 'vcodec': 'h264'},  # Height can vary (https://github.com/ytdl-org/youtube-dl/issues/4559)
 459         '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'vcodec': 'h264'},
 460         '212': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'h264'},
 461         '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'h264'},
 462         '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'h264', 'fps': 60},
 463         '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'h264', 'fps': 60},
 464         '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'h264'},
 465
 466         # Dash mp4 audio
 467         '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 48, 'container': 'm4a_dash'},
 468         '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 128, 'container': 'm4a_dash'},
 469         '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 256, 'container': 'm4a_dash'},
 470         '256': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'container': 'm4a_dash'},
 471         '258': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'container': 'm4a_dash'},
 472         '325': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'dtse', 'container': 'm4a_dash'},
 473         '328': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'ec-3', 'container': 'm4a_dash'},
 474
 475         # Dash webm
 476         '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 477         '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 478         '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 479         '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 480         '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 481         '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 482         '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp9'},
 483         '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 484         '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 485         '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 486         '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 487         '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 488         '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 489         '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 490         '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 491         # itag 272 videos are either 3840x2160 (e.g. RtoitU2A-3E) or 7680x4320 (sLprVF6d7Ug)
 492         '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 493         '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 494         '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 495         '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 496         '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 497         '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 498
 499         # Dash webm audio
 500         '171': {'ext': 'webm', 'acodec': 'vorbis', 'format_note': 'DASH audio', 'abr': 128},
 501         '172': {'ext': 'webm', 'acodec': 'vorbis', 'format_note': 'DASH audio', 'abr': 256},
 502
 503         # Dash webm audio with opus inside
 504         '249': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 50},
 505         '250': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 70},
 506         '251': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 160},
 507
 508         # RTMP (unnamed)
 509         '_rtmp': {'protocol': 'rtmp'},
 510
 511         # av01 video only formats sometimes served with "unknown" codecs
 512         '394': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 513         '395': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 514         '396': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 515         '397': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 516     }
 517     _SUBTITLE_FORMATS = ('srv1', 'srv2', 'srv3', 'ttml', 'vtt')
 518
 519     _GEO_BYPASS = False
 520
 521     IE_NAME = 'youtube'
 522     _TESTS = [
 523         {
 524             'url': 'https://www.youtube.com/watch?v=BaW_jenozKc&t=1s&end=9',
 525             'info_dict': {
 526                 'id': 'BaW_jenozKc',
 527                 'ext': 'mp4',
 528                 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
 529                 'uploader': 'Philipp Hagemeister',
 530                 'uploader_id': 'phihag',
 531                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/phihag',
 532                 'channel_id': 'UCLqxVugv74EIW3VWh2NOa3Q',
 533                 'channel_url': r're:https?://(?:www\.)?youtube\.com/channel/UCLqxVugv74EIW3VWh2NOa3Q',
 534                 'upload_date': '20121002',
 535                 'description': 'test chars:  "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .',
 536                 'categories': ['Science & Technology'],
 537                 'tags': ['youtube-dl'],
 538                 'duration': 10,
 539                 'view_count': int,
 540                 'like_count': int,
 541                 'dislike_count': int,
 542                 'start_time': 1,
 543                 'end_time': 9,
 544             }
 545         },
 546         {
 547             'url': 'https://www.youtube.com/watch?v=UxxajLWwzqY',
 548             'note': 'Test generic use_cipher_signature video (#897)',
 549             'info_dict': {
 550                 'id': 'UxxajLWwzqY',
 551                 'ext': 'mp4',
 552                 'upload_date': '20120506',
 553                 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
 554                 'alt_title': 'I Love It (feat. Charli XCX)',
 555                 'description': 'md5:f3ceb5ef83a08d95b9d146f973157cc8',
 556                 'tags': ['Icona Pop i love it', 'sweden', 'pop music', 'big beat records', 'big beat', 'charli',
 557                          'xcx', 'charli xcx', 'girls', 'hbo', 'i love it', "i don't care", 'icona', 'pop',
 558                          'iconic ep', 'iconic', 'love', 'it'],
 559                 'duration': 180,
 560                 'uploader': 'Icona Pop',
 561                 'uploader_id': 'IconaPop',
 562                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/IconaPop',
 563                 'creator': 'Icona Pop',
 564                 'track': 'I Love It (feat. Charli XCX)',
 565                 'artist': 'Icona Pop',
 566             }
 567         },
 568         {
 569             'url': 'https://www.youtube.com/watch?v=07FYdnEawAQ',
 570             'note': 'Test VEVO video with age protection (#956)',
 571             'info_dict': {
 572                 'id': '07FYdnEawAQ',
 573                 'ext': 'mp4',
 574                 'upload_date': '20130703',
 575                 'title': 'Justin Timberlake - Tunnel Vision (Official Music Video) (Explicit)',
 576                 'alt_title': 'Tunnel Vision',
 577                 'description': 'md5:07dab3356cde4199048e4c7cd93471e1',
 578                 'duration': 419,
 579                 'uploader': 'justintimberlakeVEVO',
 580                 'uploader_id': 'justintimberlakeVEVO',
 581                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/justintimberlakeVEVO',
 582                 'creator': 'Justin Timberlake',
 583                 'track': 'Tunnel Vision',
 584                 'artist': 'Justin Timberlake',
 585                 'age_limit': 18,
 586             }
 587         },
 588         {
 589             'url': '//www.YouTube.com/watch?v=yZIXLfi8CZQ',
 590             'note': 'Embed-only video (#1746)',
 591             'info_dict': {
 592                 'id': 'yZIXLfi8CZQ',
 593                 'ext': 'mp4',
 594                 'upload_date': '20120608',
 595                 'title': 'Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012',
 596                 'description': 'md5:09b78bd971f1e3e289601dfba15ca4f7',
 597                 'uploader': 'SET India',
 598                 'uploader_id': 'setindia',
 599                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/setindia',
 600                 'age_limit': 18,
 601             }
 602         },
 603         {
 604             'url': 'https://www.youtube.com/watch?v=BaW_jenozKc&v=UxxajLWwzqY',
 605             'note': 'Use the first video ID in the URL',
 606             'info_dict': {
 607                 'id': 'BaW_jenozKc',
 608                 'ext': 'mp4',
 609                 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
 610                 'uploader': 'Philipp Hagemeister',
 611                 'uploader_id': 'phihag',
 612                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/phihag',
 613                 'upload_date': '20121002',
 614                 'description': 'test chars:  "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .',
 615                 'categories': ['Science & Technology'],
 616                 'tags': ['youtube-dl'],
 617                 'duration': 10,
 618                 'view_count': int,
 619                 'like_count': int,
 620                 'dislike_count': int,
 621             },
 622             'params': {
 623                 'skip_download': True,
 624             },
 625         },
 626         {
 627             'url': 'https://www.youtube.com/watch?v=a9LDPn-MO4I',
 628             'note': '256k DASH audio (format 141) via DASH manifest',
 629             'info_dict': {
 630                 'id': 'a9LDPn-MO4I',
 631                 'ext': 'm4a',
 632                 'upload_date': '20121002',
 633                 'uploader_id': '8KVIDEO',
 634                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/8KVIDEO',
 635                 'description': '',
 636                 'uploader': '8KVIDEO',
 637                 'title': 'UHDTV TEST 8K VIDEO.mp4'
 638             },
 639             'params': {
 640                 'youtube_include_dash_manifest': True,
 641                 'format': '141',
 642             },
 643             'skip': 'format 141 not served anymore',
 644         },
 645         # DASH manifest with encrypted signature
 646         {
 647             'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
 648             'info_dict': {
 649                 'id': 'IB3lcPjvWLA',
 650                 'ext': 'm4a',
 651                 'title': 'Afrojack, Spree Wilson - The Spark (Official Music Video) ft. Spree Wilson',
 652                 'description': 'md5:8f5e2b82460520b619ccac1f509d43bf',
 653                 'duration': 244,
 654                 'uploader': 'AfrojackVEVO',
 655                 'uploader_id': 'AfrojackVEVO',
 656                 'upload_date': '20131011',
 657             },
 658             'params': {
 659                 'youtube_include_dash_manifest': True,
 660                 'format': '141/bestaudio[ext=m4a]',
 661             },
 662         },
 663         # JS player signature function name containing $
 664         {
 665             'url': 'https://www.youtube.com/watch?v=nfWlot6h_JM',
 666             'info_dict': {
 667                 'id': 'nfWlot6h_JM',
 668                 'ext': 'm4a',
 669                 'title': 'Taylor Swift - Shake It Off',
 670                 'description': 'md5:bec2185232c05479482cb5a9b82719bf',
 671                 'duration': 242,
 672                 'uploader': 'TaylorSwiftVEVO',
 673                 'uploader_id': 'TaylorSwiftVEVO',
 674                 'upload_date': '20140818',
 675                 'creator': 'Taylor Swift',
 676             },
 677             'params': {
 678                 'youtube_include_dash_manifest': True,
 679                 'format': '141/bestaudio[ext=m4a]',
 680             },
 681         },
 682         # Controversy video
 683         {
 684             'url': 'https://www.youtube.com/watch?v=T4XJQO3qol8',
 685             'info_dict': {
 686                 'id': 'T4XJQO3qol8',
 687                 'ext': 'mp4',
 688                 'duration': 219,
 689                 'upload_date': '20100909',
 690                 'uploader': 'Amazing Atheist',
 691                 'uploader_id': 'TheAmazingAtheist',
 692                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/TheAmazingAtheist',
 693                 'title': 'Burning Everyone\'s Koran',
 694                 'description': 'SUBSCRIBE: http://www.youtube.com/saturninefilms\n\nEven Obama has taken a stand against freedom on this issue: http://www.huffingtonpost.com/2010/09/09/obama-gma-interview-quran_n_710282.html',
 695             }
 696         },
 697         # Normal age-gate video (No vevo, embed allowed)
 698         {
 699             'url': 'https://youtube.com/watch?v=HtVdAasjOgU',
 700             'info_dict': {
 701                 'id': 'HtVdAasjOgU',
 702                 'ext': 'mp4',
 703                 'title': 'The Witcher 3: Wild Hunt - The Sword Of Destiny Trailer',
 704                 'description': r're:(?s).{100,}About the Game\n.*?The Witcher 3: Wild Hunt.{100,}',
 705                 'duration': 142,
 706                 'uploader': 'The Witcher',
 707                 'uploader_id': 'WitcherGame',
 708                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/WitcherGame',
 709                 'upload_date': '20140605',
 710                 'age_limit': 18,
 711             },
 712         },
 713         # Age-gate video with encrypted signature
 714         {
 715             'url': 'https://www.youtube.com/watch?v=6kLq3WMV1nU',
 716             'info_dict': {
 717                 'id': '6kLq3WMV1nU',
 718                 'ext': 'mp4',
 719                 'title': 'Dedication To My Ex (Miss That) (Lyric Video)',
 720                 'description': 'md5:33765bb339e1b47e7e72b5490139bb41',
 721                 'duration': 246,
 722                 'uploader': 'LloydVEVO',
 723                 'uploader_id': 'LloydVEVO',
 724                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/LloydVEVO',
 725                 'upload_date': '20110629',
 726                 'age_limit': 18,
 727             },
 728         },
 729         # video_info is None (https://github.com/ytdl-org/youtube-dl/issues/4421)
 730         # YouTube Red ad is not captured for creator
 731         {
 732             'url': '__2ABJjxzNo',
 733             'info_dict': {
 734                 'id': '__2ABJjxzNo',
 735                 'ext': 'mp4',
 736                 'duration': 266,
 737                 'upload_date': '20100430',
 738                 'uploader_id': 'deadmau5',
 739                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/deadmau5',
 740                 'creator': 'deadmau5',
 741                 'description': 'md5:12c56784b8032162bb936a5f76d55360',
 742                 'uploader': 'deadmau5',
 743                 'title': 'Deadmau5 - Some Chords (HD)',
 744                 'alt_title': 'Some Chords',
 745             },
 746             'expected_warnings': [
 747                 'DASH manifest missing',
 748             ]
 749         },
 750         # Olympics (https://github.com/ytdl-org/youtube-dl/issues/4431)
 751         {
 752             'url': 'lqQg6PlCWgI',
 753             'info_dict': {
 754                 'id': 'lqQg6PlCWgI',
 755                 'ext': 'mp4',
 756                 'duration': 6085,
 757                 'upload_date': '20150827',
 758                 'uploader_id': 'olympic',
 759                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/olympic',
 760                 'description': 'HO09  - Women -  GER-AUS - Hockey - 31 July 2012 - London 2012 Olympic Games',
 761                 'uploader': 'Olympic',
 762                 'title': 'Hockey - Women -  GER-AUS - London 2012 Olympic Games',
 763             },
 764             'params': {
 765                 'skip_download': 'requires avconv',
 766             }
 767         },
 768         # Non-square pixels
 769         {
 770             'url': 'https://www.youtube.com/watch?v=_b-2C3KPAM0',
 771             'info_dict': {
 772                 'id': '_b-2C3KPAM0',
 773                 'ext': 'mp4',
 774                 'stretched_ratio': 16 / 9.,
 775                 'duration': 85,
 776                 'upload_date': '20110310',
 777                 'uploader_id': 'AllenMeow',
 778                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/AllenMeow',
 779                 'description': 'made by Wacom from Korea | 字幕&加油添醋 by TY\'s Allen | 感謝heylisa00cavey1001同學熱情提供梗及翻譯',
 780                 'uploader': '孫ᄋᄅ',
 781                 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人',
 782             },
 783         },
 784         # url_encoded_fmt_stream_map is empty string
 785         {
 786             'url': 'qEJwOuvDf7I',
 787             'info_dict': {
 788                 'id': 'qEJwOuvDf7I',
 789                 'ext': 'webm',
 790                 'title': 'Обсуждение судебной практики по выборам 14 сентября 2014 года в Санкт-Петербурге',
 791                 'description': '',
 792                 'upload_date': '20150404',
 793                 'uploader_id': 'spbelect',
 794                 'uploader': 'Наблюдатели Петербурга',
 795             },
 796             'params': {
 797                 'skip_download': 'requires avconv',
 798             },
 799             'skip': 'This live event has ended.',
 800         },
 801         # Extraction from multiple DASH manifests (https://github.com/ytdl-org/youtube-dl/pull/6097)
 802         {
 803             'url': 'https://www.youtube.com/watch?v=FIl7x6_3R5Y',
 804             'info_dict': {
 805                 'id': 'FIl7x6_3R5Y',
 806                 'ext': 'webm',
 807                 'title': 'md5:7b81415841e02ecd4313668cde88737a',
 808                 'description': 'md5:116377fd2963b81ec4ce64b542173306',
 809                 'duration': 220,
 810                 'upload_date': '20150625',
 811                 'uploader_id': 'dorappi2000',
 812                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/dorappi2000',
 813                 'uploader': 'dorappi2000',
 814                 'formats': 'mincount:31',
 815             },
 816             'skip': 'not actual anymore',
 817         },
 818         # DASH manifest with segment_list
 819         {
 820             'url': 'https://www.youtube.com/embed/CsmdDsKjzN8',
 821             'md5': '8ce563a1d667b599d21064e982ab9e31',
 822             'info_dict': {
 823                 'id': 'CsmdDsKjzN8',
 824                 'ext': 'mp4',
 825                 'upload_date': '20150501',  # According to '<meta itemprop="datePublished"', but in other places it's 20150510
 826                 'uploader': 'Airtek',
 827                 'description': 'Retransmisión en directo de la XVIII media maratón de Zaragoza.',
 828                 'uploader_id': 'UCzTzUmjXxxacNnL8I3m4LnQ',
 829                 'title': 'Retransmisión XVIII Media maratón Zaragoza 2015',
 830             },
 831             'params': {
 832                 'youtube_include_dash_manifest': True,
 833                 'format': '135',  # bestvideo
 834             },
 835             'skip': 'This live event has ended.',
 836         },
 837         {
 838             # Multifeed videos (multiple cameras), URL is for Main Camera
 839             'url': 'https://www.youtube.com/watch?v=jqWvoWXjCVs',
 840             'info_dict': {
 841                 'id': 'jqWvoWXjCVs',
 842                 'title': 'teamPGP: Rocket League Noob Stream',
 843                 'description': 'md5:dc7872fb300e143831327f1bae3af010',
 844             },
 845             'playlist': [{
 846                 'info_dict': {
 847                     'id': 'jqWvoWXjCVs',
 848                     'ext': 'mp4',
 849                     'title': 'teamPGP: Rocket League Noob Stream (Main Camera)',
 850                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 851                     'duration': 7335,
 852                     'upload_date': '20150721',
 853                     'uploader': 'Beer Games Beer',
 854                     'uploader_id': 'beergamesbeer',
 855                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 856                     'license': 'Standard YouTube License',
 857                 },
 858             }, {
 859                 'info_dict': {
 860                     'id': '6h8e8xoXJzg',
 861                     'ext': 'mp4',
 862                     'title': 'teamPGP: Rocket League Noob Stream (kreestuh)',
 863                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 864                     'duration': 7337,
 865                     'upload_date': '20150721',
 866                     'uploader': 'Beer Games Beer',
 867                     'uploader_id': 'beergamesbeer',
 868                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 869                     'license': 'Standard YouTube License',
 870                 },
 871             }, {
 872                 'info_dict': {
 873                     'id': 'PUOgX5z9xZw',
 874                     'ext': 'mp4',
 875                     'title': 'teamPGP: Rocket League Noob Stream (grizzle)',
 876                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 877                     'duration': 7337,
 878                     'upload_date': '20150721',
 879                     'uploader': 'Beer Games Beer',
 880                     'uploader_id': 'beergamesbeer',
 881                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 882                     'license': 'Standard YouTube License',
 883                 },
 884             }, {
 885                 'info_dict': {
 886                     'id': 'teuwxikvS5k',
 887                     'ext': 'mp4',
 888                     'title': 'teamPGP: Rocket League Noob Stream (zim)',
 889                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 890                     'duration': 7334,
 891                     'upload_date': '20150721',
 892                     'uploader': 'Beer Games Beer',
 893                     'uploader_id': 'beergamesbeer',
 894                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 895                     'license': 'Standard YouTube License',
 896                 },
 897             }],
 898             'params': {
 899                 'skip_download': True,
 900             },
 901             'skip': 'This video is not available.',
 902         },
 903         {
 904             # Multifeed video with comma in title (see https://github.com/ytdl-org/youtube-dl/issues/8536)
 905             'url': 'https://www.youtube.com/watch?v=gVfLd0zydlo',
 906             'info_dict': {
 907                 'id': 'gVfLd0zydlo',
 908                 'title': 'DevConf.cz 2016 Day 2 Workshops 1 14:00 - 15:30',
 909             },
 910             'playlist_count': 2,
 911             'skip': 'Not multifeed anymore',
 912         },
 913         {
 914             'url': 'https://vid.plus/FlRa-iH7PGw',
 915             'only_matching': True,
 916         },
 917         {
 918             'url': 'https://zwearz.com/watch/9lWxNJF-ufM/electra-woman-dyna-girl-official-trailer-grace-helbig.html',
 919             'only_matching': True,
 920         },
 921         {
 922             # Title with JS-like syntax "};" (see https://github.com/ytdl-org/youtube-dl/issues/7468)
 923             # Also tests cut-off URL expansion in video description (see
 924             # https://github.com/ytdl-org/youtube-dl/issues/1892,
 925             # https://github.com/ytdl-org/youtube-dl/issues/8164)
 926             'url': 'https://www.youtube.com/watch?v=lsguqyKfVQg',
 927             'info_dict': {
 928                 'id': 'lsguqyKfVQg',
 929                 'ext': 'mp4',
 930                 'title': '{dark walk}; Loki/AC/Dishonored; collab w/Elflover21',
 931                 'alt_title': 'Dark Walk - Position Music',
 932                 'description': 'md5:8085699c11dc3f597ce0410b0dcbb34a',
 933                 'duration': 133,
 934                 'upload_date': '20151119',
 935                 'uploader_id': 'IronSoulElf',
 936                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/IronSoulElf',
 937                 'uploader': 'IronSoulElf',
 938                 'creator': 'Todd Haberman,  Daniel Law Heath and Aaron Kaplan',
 939                 'track': 'Dark Walk - Position Music',
 940                 'artist': 'Todd Haberman,  Daniel Law Heath and Aaron Kaplan',
 941                 'album': 'Position Music - Production Music Vol. 143 - Dark Walk',
 942             },
 943             'params': {
 944                 'skip_download': True,
 945             },
 946         },
 947         {
 948             # Tags with '};' (see https://github.com/ytdl-org/youtube-dl/issues/7468)
 949             'url': 'https://www.youtube.com/watch?v=Ms7iBXnlUO8',
 950             'only_matching': True,
 951         },
 952         {
 953             # Video with yt:stretch=17:0
 954             'url': 'https://www.youtube.com/watch?v=Q39EVAstoRM',
 955             'info_dict': {
 956                 'id': 'Q39EVAstoRM',
 957                 'ext': 'mp4',
 958                 'title': 'Clash Of Clans#14 Dicas De Ataque Para CV 4',
 959                 'description': 'md5:ee18a25c350637c8faff806845bddee9',
 960                 'upload_date': '20151107',
 961                 'uploader_id': 'UCCr7TALkRbo3EtFzETQF1LA',
 962                 'uploader': 'CH GAMER DROID',
 963             },
 964             'params': {
 965                 'skip_download': True,
 966             },
 967             'skip': 'This video does not exist.',
 968         },
 969         {
 970             # Video licensed under Creative Commons
 971             'url': 'https://www.youtube.com/watch?v=M4gD1WSo5mA',
 972             'info_dict': {
 973                 'id': 'M4gD1WSo5mA',
 974                 'ext': 'mp4',
 975                 'title': 'md5:e41008789470fc2533a3252216f1c1d1',
 976                 'description': 'md5:a677553cf0840649b731a3024aeff4cc',
 977                 'duration': 721,
 978                 'upload_date': '20150127',
 979                 'uploader_id': 'BerkmanCenter',
 980                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/BerkmanCenter',
 981                 'uploader': 'The Berkman Klein Center for Internet & Society',
 982                 'license': 'Creative Commons Attribution license (reuse allowed)',
 983             },
 984             'params': {
 985                 'skip_download': True,
 986             },
 987         },
 988         {
 989             # Channel-like uploader_url
 990             'url': 'https://www.youtube.com/watch?v=eQcmzGIKrzg',
 991             'info_dict': {
 992                 'id': 'eQcmzGIKrzg',
 993                 'ext': 'mp4',
 994                 'title': 'Democratic Socialism and Foreign Policy | Bernie Sanders',
 995                 'description': 'md5:dda0d780d5a6e120758d1711d062a867',
 996                 'duration': 4060,
 997                 'upload_date': '20151119',
 998                 'uploader': 'Bernie Sanders',
 999                 'uploader_id': 'UCH1dpzjCEiGAt8CXkryhkZg',
1000                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCH1dpzjCEiGAt8CXkryhkZg',
1001                 'license': 'Creative Commons Attribution license (reuse allowed)',
1002             },
1003             'params': {
1004                 'skip_download': True,
1005             },
1006         },
1007         {
1008             'url': 'https://www.youtube.com/watch?feature=player_embedded&amp;amp;v=V36LpHqtcDY',
1009             'only_matching': True,
1010         },
1011         {
1012             # YouTube Red paid video (https://github.com/ytdl-org/youtube-dl/issues/10059)
1013             'url': 'https://www.youtube.com/watch?v=i1Ko8UG-Tdo',
1014             'only_matching': True,
1015         },
1016         {
1017             # Rental video preview
1018             'url': 'https://www.youtube.com/watch?v=yYr8q0y5Jfg',
1019             'info_dict': {
1020                 'id': 'uGpuVWrhIzE',
1021                 'ext': 'mp4',
1022                 'title': 'Piku - Trailer',
1023                 'description': 'md5:c36bd60c3fd6f1954086c083c72092eb',
1024                 'upload_date': '20150811',
1025                 'uploader': 'FlixMatrix',
1026                 'uploader_id': 'FlixMatrixKaravan',
1027                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/FlixMatrixKaravan',
1028                 'license': 'Standard YouTube License',
1029             },
1030             'params': {
1031                 'skip_download': True,
1032             },
1033             'skip': 'This video is not available.',
1034         },
1035         {
1036             # YouTube Red video with episode data
1037             'url': 'https://www.youtube.com/watch?v=iqKdEhx-dD4',
1038             'info_dict': {
1039                 'id': 'iqKdEhx-dD4',
1040                 'ext': 'mp4',
1041                 'title': 'Isolation - Mind Field (Ep 1)',
1042                 'description': 'md5:46a29be4ceffa65b92d277b93f463c0f',
1043                 'duration': 2085,
1044                 'upload_date': '20170118',
1045                 'uploader': 'Vsauce',
1046                 'uploader_id': 'Vsauce',
1047                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/Vsauce',
1048                 'series': 'Mind Field',
1049                 'season_number': 1,
1050                 'episode_number': 1,
1051             },
1052             'params': {
1053                 'skip_download': True,
1054             },
1055             'expected_warnings': [
1056                 'Skipping DASH manifest',
1057             ],
1058         },
1059         {
1060             # The following content has been identified by the YouTube community
1061             # as inappropriate or offensive to some audiences.
1062             'url': 'https://www.youtube.com/watch?v=6SJNVb0GnPI',
1063             'info_dict': {
1064                 'id': '6SJNVb0GnPI',
1065                 'ext': 'mp4',
1066                 'title': 'Race Differences in Intelligence',
1067                 'description': 'md5:5d161533167390427a1f8ee89a1fc6f1',
1068                 'duration': 965,
1069                 'upload_date': '20140124',
1070                 'uploader': 'New Century Foundation',
1071                 'uploader_id': 'UCEJYpZGqgUob0zVVEaLhvVg',
1072                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCEJYpZGqgUob0zVVEaLhvVg',
1073             },
1074             'params': {
1075                 'skip_download': True,
1076             },
1077         },
1078         {
1079             # itag 212
1080             'url': '1t24XAntNCY',
1081             'only_matching': True,
1082         },
1083         {
1084             # geo restricted to JP
1085             'url': 'sJL6WA-aGkQ',
1086             'only_matching': True,
1087         },
1088         {
1089             'url': 'https://www.youtube.com/watch?v=MuAGGZNfUkU&list=RDMM',
1090             'only_matching': True,
1091         },
1092         {
1093             'url': 'https://invidio.us/watch?v=BaW_jenozKc',
1094             'only_matching': True,
1095         },
1096         {
1097             # DRM protected
1098             'url': 'https://www.youtube.com/watch?v=s7_qI6_mIXc',
1099             'only_matching': True,
1100         },
1101         {
1102             # Video with unsupported adaptive stream type formats
1103             'url': 'https://www.youtube.com/watch?v=Z4Vy8R84T1U',
1104             'info_dict': {
1105                 'id': 'Z4Vy8R84T1U',
1106                 'ext': 'mp4',
1107                 'title': 'saman SMAN 53 Jakarta(Sancety) opening COFFEE4th at SMAN 53 Jakarta',
1108                 'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
1109                 'duration': 433,
1110                 'upload_date': '20130923',
1111                 'uploader': 'Amelia Putri Harwita',
1112                 'uploader_id': 'UCpOxM49HJxmC1qCalXyB3_Q',
1113                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCpOxM49HJxmC1qCalXyB3_Q',
1114                 'formats': 'maxcount:10',
1115             },
1116             'params': {
1117                 'skip_download': True,
1118                 'youtube_include_dash_manifest': False,
1119             },
1120         },
1121         {
1122             # Youtube Music Auto-generated description
1123             'url': 'https://music.youtube.com/watch?v=MgNrAu2pzNs',
1124             'info_dict': {
1125                 'id': 'MgNrAu2pzNs',
1126                 'ext': 'mp4',
1127                 'title': 'Voyeur Girl',
1128                 'description': 'md5:7ae382a65843d6df2685993e90a8628f',
1129                 'upload_date': '20190312',
1130                 'uploader': 'Various Artists - Topic',
1131                 'uploader_id': 'UCVWKBi1ELZn0QX2CBLSkiyw',
1132                 'artist': 'Stephen',
1133                 'track': 'Voyeur Girl',
1134                 'album': 'it\'s too much love to know my dear',
1135                 'release_date': '20190313',
1136                 'release_year': 2019,
1137             },
1138             'params': {
1139                 'skip_download': True,
1140             },
1141         },
1142         {
1143             # Youtube Music Auto-generated description
1144             # Retrieve 'artist' field from 'Artist:' in video description
1145             # when it is present on youtube music video
1146             'url': 'https://www.youtube.com/watch?v=k0jLE7tTwjY',
1147             'info_dict': {
1148                 'id': 'k0jLE7tTwjY',
1149                 'ext': 'mp4',
1150                 'title': 'Latch Feat. Sam Smith',
1151                 'description': 'md5:3cb1e8101a7c85fcba9b4fb41b951335',
1152                 'upload_date': '20150110',
1153                 'uploader': 'Various Artists - Topic',
1154                 'uploader_id': 'UCNkEcmYdjrH4RqtNgh7BZ9w',
1155                 'artist': 'Disclosure',
1156                 'track': 'Latch Feat. Sam Smith',
1157                 'album': 'Latch Featuring Sam Smith',
1158                 'release_date': '20121008',
1159                 'release_year': 2012,
1160             },
1161             'params': {
1162                 'skip_download': True,
1163             },
1164         },
1165         {
1166             # Youtube Music Auto-generated description
1167             # handle multiple artists on youtube music video
1168             'url': 'https://www.youtube.com/watch?v=74qn0eJSjpA',
1169             'info_dict': {
1170                 'id': '74qn0eJSjpA',
1171                 'ext': 'mp4',
1172                 'title': 'Eastside',
1173                 'description': 'md5:290516bb73dcbfab0dcc4efe6c3de5f2',
1174                 'upload_date': '20180710',
1175                 'uploader': 'Benny Blanco - Topic',
1176                 'uploader_id': 'UCzqz_ksRu_WkIzmivMdIS7A',
1177                 'artist': 'benny blanco, Halsey, Khalid',
1178                 'track': 'Eastside',
1179                 'album': 'Eastside',
1180                 'release_date': '20180713',
1181                 'release_year': 2018,
1182             },
1183             'params': {
1184                 'skip_download': True,
1185             },
1186         },
1187         {
1188             # Youtube Music Auto-generated description
1189             # handle youtube music video with release_year and no release_date
1190             'url': 'https://www.youtube.com/watch?v=-hcAI0g-f5M',
1191             'info_dict': {
1192                 'id': '-hcAI0g-f5M',
1193                 'ext': 'mp4',
1194                 'title': 'Put It On Me',
1195                 'description': 'md5:93c55acc682ae7b0c668f2e34e1c069e',
1196                 'upload_date': '20180426',
1197                 'uploader': 'Matt Maeson - Topic',
1198                 'uploader_id': 'UCnEkIGqtGcQMLk73Kp-Q5LQ',
1199                 'artist': 'Matt Maeson',
1200                 'track': 'Put It On Me',
1201                 'album': 'The Hearse',
1202                 'release_date': None,
1203                 'release_year': 2018,
1204             },
1205             'params': {
1206                 'skip_download': True,
1207             },
1208         },
1209     ]
1210
1211     def __init__(self, *args, **kwargs):
1212         super(YoutubeIE, self).__init__(*args, **kwargs)
1213         self._player_cache = {}
1214
1215     def report_video_info_webpage_download(self, video_id):
1216         """Report attempt to download video info webpage."""
1217         self.to_screen('%s: Downloading video info webpage' % video_id)
1218
1219     def report_information_extraction(self, video_id):
1220         """Report attempt to extract video information."""
1221         self.to_screen('%s: Extracting video information' % video_id)
1222
1223     def report_unavailable_format(self, video_id, format):
1224         """Report extracted video URL."""
1225         self.to_screen('%s: Format %s not available' % (video_id, format))
1226
1227     def report_rtmp_download(self):
1228         """Indicate the download will use the RTMP protocol."""
1229         self.to_screen('RTMP download detected')
1230
1231     def _signature_cache_id(self, example_sig):
1232         """ Return a string representation of a signature """
1233         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
1234
1235     def _extract_signature_function(self, video_id, player_url, example_sig):
1236         id_m = re.match(
1237             r'.*?-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player(?:-new)?|(?:/[a-z]{2,3}_[A-Z]{2})?/base)?\.(?P<ext>[a-z]+)$',
1238             player_url)
1239         if not id_m:
1240             raise ExtractorError('Cannot identify player %r' % player_url)
1241         player_type = id_m.group('ext')
1242         player_id = id_m.group('id')
1243
1244         # Read from filesystem cache
1245         func_id = '%s_%s_%s' % (
1246             player_type, player_id, self._signature_cache_id(example_sig))
1247         assert os.path.basename(func_id) == func_id
1248
1249         cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
1250         if cache_spec is not None:
1251             return lambda s: ''.join(s[i] for i in cache_spec)
1252
1253         download_note = (
1254             'Downloading player %s' % player_url
1255             if self._downloader.params.get('verbose') else
1256             'Downloading %s player %s' % (player_type, player_id)
1257         )
1258         if player_type == 'js':
1259             code = self._download_webpage(
1260                 player_url, video_id,
1261                 note=download_note,
1262                 errnote='Download of %s failed' % player_url)
1263             res = self._parse_sig_js(code)
1264         elif player_type == 'swf':
1265             urlh = self._request_webpage(
1266                 player_url, video_id,
1267                 note=download_note,
1268                 errnote='Download of %s failed' % player_url)
1269             code = urlh.read()
1270             res = self._parse_sig_swf(code)
1271         else:
1272             assert False, 'Invalid player type %r' % player_type
1273
1274         test_string = ''.join(map(compat_chr, range(len(example_sig))))
1275         cache_res = res(test_string)
1276         cache_spec = [ord(c) for c in cache_res]
1277
1278         self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
1279         return res
1280
1281     def _print_sig_code(self, func, example_sig):
1282         def gen_sig_code(idxs):
1283             def _genslice(start, end, step):
1284                 starts = '' if start == 0 else str(start)
1285                 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
1286                 steps = '' if step == 1 else (':%d' % step)
1287                 return 's[%s%s%s]' % (starts, ends, steps)
1288
1289             step = None
1290             # Quelch pyflakes warnings - start will be set when step is set
1291             start = '(Never used)'
1292             for i, prev in zip(idxs[1:], idxs[:-1]):
1293                 if step is not None:
1294                     if i - prev == step:
1295                         continue
1296                     yield _genslice(start, prev, step)
1297                     step = None
1298                     continue
1299                 if i - prev in [-1, 1]:
1300                     step = i - prev
1301                     start = prev
1302                     continue
1303                 else:
1304                     yield 's[%d]' % prev
1305             if step is None:
1306                 yield 's[%d]' % i
1307             else:
1308                 yield _genslice(start, i, step)
1309
1310         test_string = ''.join(map(compat_chr, range(len(example_sig))))
1311         cache_res = func(test_string)
1312         cache_spec = [ord(c) for c in cache_res]
1313         expr_code = ' + '.join(gen_sig_code(cache_spec))
1314         signature_id_tuple = '(%s)' % (
1315             ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
1316         code = ('if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
1317                 '    return %s\n') % (signature_id_tuple, expr_code)
1318         self.to_screen('Extracted signature function:\n' + code)
1319
1320     def _parse_sig_js(self, jscode):
1321         funcname = self._search_regex(
1322             (r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1323              r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1324              r'(?P<sig>[a-zA-Z0-9$]+)\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)',
1325              # Obsolete patterns
1326              r'(["\'])signature\1\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1327              r'\.sig\|\|(?P<sig>[a-zA-Z0-9$]+)\(',
1328              r'yt\.akamaized\.net/\)\s*\|\|\s*.*?\s*[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?:encodeURIComponent\s*\()?\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1329              r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1330              r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1331              r'\bc\s*&&\s*a\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1332              r'\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1333              r'\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\('),
1334             jscode, 'Initial JS player signature function name', group='sig')
1335
1336         jsi = JSInterpreter(jscode)
1337         initial_function = jsi.extract_function(funcname)
1338         return lambda s: initial_function([s])
1339
1340     def _parse_sig_swf(self, file_contents):
1341         swfi = SWFInterpreter(file_contents)
1342         TARGET_CLASSNAME = 'SignatureDecipher'
1343         searched_class = swfi.extract_class(TARGET_CLASSNAME)
1344         initial_function = swfi.extract_function(searched_class, 'decipher')
1345         return lambda s: initial_function([s])
1346
1347     def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
1348         """Turn the encrypted s field into a working signature"""
1349
1350         if player_url is None:
1351             raise ExtractorError('Cannot decrypt signature without player_url')
1352
1353         if player_url.startswith('//'):
1354             player_url = 'https:' + player_url
1355         elif not re.match(r'https?://', player_url):
1356             player_url = compat_urlparse.urljoin(
1357                 'https://www.youtube.com', player_url)
1358         try:
1359             player_id = (player_url, self._signature_cache_id(s))
1360             if player_id not in self._player_cache:
1361                 func = self._extract_signature_function(
1362                     video_id, player_url, s
1363                 )
1364                 self._player_cache[player_id] = func
1365             func = self._player_cache[player_id]
1366             if self._downloader.params.get('youtube_print_sig_code'):
1367                 self._print_sig_code(func, s)
1368             return func(s)
1369         except Exception as e:
1370             tb = traceback.format_exc()
1371             raise ExtractorError(
1372                 'Signature extraction failed: ' + tb, cause=e)
1373
1374     def _get_subtitles(self, video_id, webpage):
1375         try:
1376             subs_doc = self._download_xml(
1377                 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
1378                 video_id, note=False)
1379         except ExtractorError as err:
1380             self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err))
1381             return {}
1382
1383         sub_lang_list = {}
1384         for track in subs_doc.findall('track'):
1385             lang = track.attrib['lang_code']
1386             if lang in sub_lang_list:
1387                 continue
1388             sub_formats = []
1389             for ext in self._SUBTITLE_FORMATS:
1390                 params = compat_urllib_parse_urlencode({
1391                     'lang': lang,
1392                     'v': video_id,
1393                     'fmt': ext,
1394                     'name': track.attrib['name'].encode('utf-8'),
1395                 })
1396                 sub_formats.append({
1397                     'url': 'https://www.youtube.com/api/timedtext?' + params,
1398                     'ext': ext,
1399                 })
1400             sub_lang_list[lang] = sub_formats
1401         if not sub_lang_list:
1402             self._downloader.report_warning('video doesn\'t have subtitles')
1403             return {}
1404         return sub_lang_list
1405
1406     def _get_ytplayer_config(self, video_id, webpage):
1407         patterns = (
1408             # User data may contain arbitrary character sequences that may affect
1409             # JSON extraction with regex, e.g. when '};' is contained the second
1410             # regex won't capture the whole JSON. Yet working around by trying more
1411             # concrete regex first keeping in mind proper quoted string handling
1412             # to be implemented in future that will replace this workaround (see
1413             # https://github.com/ytdl-org/youtube-dl/issues/7468,
1414             # https://github.com/ytdl-org/youtube-dl/pull/7599)
1415             r';ytplayer\.config\s*=\s*({.+?});ytplayer',
1416             r';ytplayer\.config\s*=\s*({.+?});',
1417         )
1418         config = self._search_regex(
1419             patterns, webpage, 'ytplayer.config', default=None)
1420         if config:
1421             return self._parse_json(
1422                 uppercase_escape(config), video_id, fatal=False)
1423
1424     def _get_automatic_captions(self, video_id, webpage):
1425         """We need the webpage for getting the captions url, pass it as an
1426            argument to speed up the process."""
1427         self.to_screen('%s: Looking for automatic captions' % video_id)
1428         player_config = self._get_ytplayer_config(video_id, webpage)
1429         err_msg = 'Couldn\'t find automatic captions for %s' % video_id
1430         if not player_config:
1431             self._downloader.report_warning(err_msg)
1432             return {}
1433         try:
1434             args = player_config['args']
1435             caption_url = args.get('ttsurl')
1436             if caption_url:
1437                 timestamp = args['timestamp']
1438                 # We get the available subtitles
1439                 list_params = compat_urllib_parse_urlencode({
1440                     'type': 'list',
1441                     'tlangs': 1,
1442                     'asrs': 1,
1443                 })
1444                 list_url = caption_url + '&' + list_params
1445                 caption_list = self._download_xml(list_url, video_id)
1446                 original_lang_node = caption_list.find('track')
1447                 if original_lang_node is None:
1448                     self._downloader.report_warning('Video doesn\'t have automatic captions')
1449                     return {}
1450                 original_lang = original_lang_node.attrib['lang_code']
1451                 caption_kind = original_lang_node.attrib.get('kind', '')
1452
1453                 sub_lang_list = {}
1454                 for lang_node in caption_list.findall('target'):
1455                     sub_lang = lang_node.attrib['lang_code']
1456                     sub_formats = []
1457                     for ext in self._SUBTITLE_FORMATS:
1458                         params = compat_urllib_parse_urlencode({
1459                             'lang': original_lang,
1460                             'tlang': sub_lang,
1461                             'fmt': ext,
1462                             'ts': timestamp,
1463                             'kind': caption_kind,
1464                         })
1465                         sub_formats.append({
1466                             'url': caption_url + '&' + params,
1467                             'ext': ext,
1468                         })
1469                     sub_lang_list[sub_lang] = sub_formats
1470                 return sub_lang_list
1471
1472             def make_captions(sub_url, sub_langs):
1473                 parsed_sub_url = compat_urllib_parse_urlparse(sub_url)
1474                 caption_qs = compat_parse_qs(parsed_sub_url.query)
1475                 captions = {}
1476                 for sub_lang in sub_langs:
1477                     sub_formats = []
1478                     for ext in self._SUBTITLE_FORMATS:
1479                         caption_qs.update({
1480                             'tlang': [sub_lang],
1481                             'fmt': [ext],
1482                         })
1483                         sub_url = compat_urlparse.urlunparse(parsed_sub_url._replace(
1484                             query=compat_urllib_parse_urlencode(caption_qs, True)))
1485                         sub_formats.append({
1486                             'url': sub_url,
1487                             'ext': ext,
1488                         })
1489                     captions[sub_lang] = sub_formats
1490                 return captions
1491
1492             # New captions format as of 22.06.2017
1493             player_response = args.get('player_response')
1494             if player_response and isinstance(player_response, compat_str):
1495                 player_response = self._parse_json(
1496                     player_response, video_id, fatal=False)
1497                 if player_response:
1498                     renderer = player_response['captions']['playerCaptionsTracklistRenderer']
1499                     base_url = renderer['captionTracks'][0]['baseUrl']
1500                     sub_lang_list = []
1501                     for lang in renderer['translationLanguages']:
1502                         lang_code = lang.get('languageCode')
1503                         if lang_code:
1504                             sub_lang_list.append(lang_code)
1505                     return make_captions(base_url, sub_lang_list)
1506
1507             # Some videos don't provide ttsurl but rather caption_tracks and
1508             # caption_translation_languages (e.g. 20LmZk1hakA)
1509             # Does not used anymore as of 22.06.2017
1510             caption_tracks = args['caption_tracks']
1511             caption_translation_languages = args['caption_translation_languages']
1512             caption_url = compat_parse_qs(caption_tracks.split(',')[0])['u'][0]
1513             sub_lang_list = []
1514             for lang in caption_translation_languages.split(','):
1515                 lang_qs = compat_parse_qs(compat_urllib_parse_unquote_plus(lang))
1516                 sub_lang = lang_qs.get('lc', [None])[0]
1517                 if sub_lang:
1518                     sub_lang_list.append(sub_lang)
1519             return make_captions(caption_url, sub_lang_list)
1520         # An extractor error can be raise by the download process if there are
1521         # no automatic captions but there are subtitles
1522         except (KeyError, IndexError, ExtractorError):
1523             self._downloader.report_warning(err_msg)
1524             return {}
1525
1526     def _mark_watched(self, video_id, video_info, player_response):
1527         playback_url = url_or_none(try_get(
1528             player_response,
1529             lambda x: x['playbackTracking']['videostatsPlaybackUrl']['baseUrl']) or try_get(
1530             video_info, lambda x: x['videostats_playback_base_url'][0]))
1531         if not playback_url:
1532             return
1533         parsed_playback_url = compat_urlparse.urlparse(playback_url)
1534         qs = compat_urlparse.parse_qs(parsed_playback_url.query)
1535
1536         # cpn generation algorithm is reverse engineered from base.js.
1537         # In fact it works even with dummy cpn.
1538         CPN_ALPHABET = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_'
1539         cpn = ''.join((CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(0, 16)))
1540
1541         qs.update({
1542             'ver': ['2'],
1543             'cpn': [cpn],
1544         })
1545         playback_url = compat_urlparse.urlunparse(
1546             parsed_playback_url._replace(query=compat_urllib_parse_urlencode(qs, True)))
1547
1548         self._download_webpage(
1549             playback_url, video_id, 'Marking watched',
1550             'Unable to mark watched', fatal=False)
1551
1552     @staticmethod
1553     def _extract_urls(webpage):
1554         # Embedded YouTube player
1555         entries = [
1556             unescapeHTML(mobj.group('url'))
1557             for mobj in re.finditer(r'''(?x)
1558             (?:
1559                 <iframe[^>]+?src=|
1560                 data-video-url=|
1561                 <embed[^>]+?src=|
1562                 embedSWF\(?:\s*|
1563                 <object[^>]+data=|
1564                 new\s+SWFObject\(
1565             )
1566             (["\'])
1567                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
1568                 (?:embed|v|p)/[0-9A-Za-z_-]{11}.*?)
1569             \1''', webpage)]
1570
1571         # lazyYT YouTube embed
1572         entries.extend(list(map(
1573             unescapeHTML,
1574             re.findall(r'class="lazyYT" data-youtube-id="([^"]+)"', webpage))))
1575
1576         # Wordpress "YouTube Video Importer" plugin
1577         matches = re.findall(r'''(?x)<div[^>]+
1578             class=(?P<q1>[\'"])[^\'"]*\byvii_single_video_player\b[^\'"]*(?P=q1)[^>]+
1579             data-video_id=(?P<q2>[\'"])([^\'"]+)(?P=q2)''', webpage)
1580         entries.extend(m[-1] for m in matches)
1581
1582         return entries
1583
1584     @staticmethod
1585     def _extract_url(webpage):
1586         urls = YoutubeIE._extract_urls(webpage)
1587         return urls[0] if urls else None
1588
1589     @classmethod
1590     def extract_id(cls, url):
1591         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
1592         if mobj is None:
1593             raise ExtractorError('Invalid URL: %s' % url)
1594         video_id = mobj.group(2)
1595         return video_id
1596
1597     def _extract_annotations(self, video_id):
1598         return self._download_webpage(
1599             'https://www.youtube.com/annotations_invideo', video_id,
1600             note='Downloading annotations',
1601             errnote='Unable to download video annotations', fatal=False,
1602             query={
1603                 'features': 1,
1604                 'legacy': 1,
1605                 'video_id': video_id,
1606             })
1607
1608     @staticmethod
1609     def _extract_chapters(description, duration):
1610         if not description:
1611             return None
1612         chapter_lines = re.findall(
1613             r'(?:^|<br\s*/>)([^<]*<a[^>]+onclick=["\']yt\.www\.watch\.player\.seekTo[^>]+>(\d{1,2}:\d{1,2}(?::\d{1,2})?)</a>[^>]*)(?=$|<br\s*/>)',
1614             description)
1615         if not chapter_lines:
1616             return None
1617         chapters = []
1618         for next_num, (chapter_line, time_point) in enumerate(
1619                 chapter_lines, start=1):
1620             start_time = parse_duration(time_point)
1621             if start_time is None:
1622                 continue
1623             if start_time > duration:
1624                 break
1625             end_time = (duration if next_num == len(chapter_lines)
1626                         else parse_duration(chapter_lines[next_num][1]))
1627             if end_time is None:
1628                 continue
1629             if end_time > duration:
1630                 end_time = duration
1631             if start_time > end_time:
1632                 break
1633             chapter_title = re.sub(
1634                 r'<a[^>]+>[^<]+</a>', '', chapter_line).strip(' \t-')
1635             chapter_title = re.sub(r'\s+', ' ', chapter_title)
1636             chapters.append({
1637                 'start_time': start_time,
1638                 'end_time': end_time,
1639                 'title': chapter_title,
1640             })
1641         return chapters
1642
1643     def _real_extract(self, url):
1644         url, smuggled_data = unsmuggle_url(url, {})
1645
1646         proto = (
1647             'http' if self._downloader.params.get('prefer_insecure', False)
1648             else 'https')
1649
1650         start_time = None
1651         end_time = None
1652         parsed_url = compat_urllib_parse_urlparse(url)
1653         for component in [parsed_url.fragment, parsed_url.query]:
1654             query = compat_parse_qs(component)
1655             if start_time is None and 't' in query:
1656                 start_time = parse_duration(query['t'][0])
1657             if start_time is None and 'start' in query:
1658                 start_time = parse_duration(query['start'][0])
1659             if end_time is None and 'end' in query:
1660                 end_time = parse_duration(query['end'][0])
1661
1662         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1663         mobj = re.search(self._NEXT_URL_RE, url)
1664         if mobj:
1665             url = proto + '://www.youtube.com/' + compat_urllib_parse_unquote(mobj.group(1)).lstrip('/')
1666         video_id = self.extract_id(url)
1667
1668         # Get video webpage
1669         url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id
1670         video_webpage = self._download_webpage(url, video_id)
1671
1672         # Attempt to extract SWF player URL
1673         mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1674         if mobj is not None:
1675             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1676         else:
1677             player_url = None
1678
1679         dash_mpds = []
1680
1681         def add_dash_mpd(video_info):
1682             dash_mpd = video_info.get('dashmpd')
1683             if dash_mpd and dash_mpd[0] not in dash_mpds:
1684                 dash_mpds.append(dash_mpd[0])
1685
1686         def add_dash_mpd_pr(pl_response):
1687             dash_mpd = url_or_none(try_get(
1688                 pl_response, lambda x: x['streamingData']['dashManifestUrl'],
1689                 compat_str))
1690             if dash_mpd and dash_mpd not in dash_mpds:
1691                 dash_mpds.append(dash_mpd)
1692
1693         is_live = None
1694         view_count = None
1695
1696         def extract_view_count(v_info):
1697             return int_or_none(try_get(v_info, lambda x: x['view_count'][0]))
1698
1699         def extract_token(v_info):
1700             return dict_get(v_info, ('account_playback_token', 'accountPlaybackToken', 'token'))
1701
1702         player_response = {}
1703
1704         # Get video info
1705         embed_webpage = None
1706         if re.search(r'player-age-gate-content">', video_webpage) is not None:
1707             age_gate = True
1708             # We simulate the access to the video from www.youtube.com/v/{video_id}
1709             # this can be viewed without login into Youtube
1710             url = proto + '://www.youtube.com/embed/%s' % video_id
1711             embed_webpage = self._download_webpage(url, video_id, 'Downloading embed webpage')
1712             data = compat_urllib_parse_urlencode({
1713                 'video_id': video_id,
1714                 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1715                 'sts': self._search_regex(
1716                     r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''),
1717             })
1718             video_info_url = proto + '://www.youtube.com/get_video_info?' + data
1719             video_info_webpage = self._download_webpage(
1720                 video_info_url, video_id,
1721                 note='Refetching age-gated info webpage',
1722                 errnote='unable to download video info webpage')
1723             video_info = compat_parse_qs(video_info_webpage)
1724             add_dash_mpd(video_info)
1725         else:
1726             age_gate = False
1727             video_info = None
1728             sts = None
1729             # Try looking directly into the video webpage
1730             ytplayer_config = self._get_ytplayer_config(video_id, video_webpage)
1731             if ytplayer_config:
1732                 args = ytplayer_config['args']
1733                 if args.get('url_encoded_fmt_stream_map') or args.get('hlsvp'):
1734                     # Convert to the same format returned by compat_parse_qs
1735                     video_info = dict((k, [v]) for k, v in args.items())
1736                     add_dash_mpd(video_info)
1737                 # Rental video is not rented but preview is available (e.g.
1738                 # https://www.youtube.com/watch?v=yYr8q0y5Jfg,
1739                 # https://github.com/ytdl-org/youtube-dl/issues/10532)
1740                 if not video_info and args.get('ypc_vid'):
1741                     return self.url_result(
1742                         args['ypc_vid'], YoutubeIE.ie_key(), video_id=args['ypc_vid'])
1743                 if args.get('livestream') == '1' or args.get('live_playback') == 1:
1744                     is_live = True
1745                 sts = ytplayer_config.get('sts')
1746                 if not player_response:
1747                     pl_response = str_or_none(args.get('player_response'))
1748                     if pl_response:
1749                         pl_response = self._parse_json(pl_response, video_id, fatal=False)
1750                         if isinstance(pl_response, dict):
1751                             player_response = pl_response
1752             if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True):
1753                 add_dash_mpd_pr(player_response)
1754                 # We also try looking in get_video_info since it may contain different dashmpd
1755                 # URL that points to a DASH manifest with possibly different itag set (some itags
1756                 # are missing from DASH manifest pointed by webpage's dashmpd, some - from DASH
1757                 # manifest pointed by get_video_info's dashmpd).
1758                 # The general idea is to take a union of itags of both DASH manifests (for example
1759                 # video with such 'manifest behavior' see https://github.com/ytdl-org/youtube-dl/issues/6093)
1760                 self.report_video_info_webpage_download(video_id)
1761                 for el in ('embedded', 'detailpage', 'vevo', ''):
1762                     query = {
1763                         'video_id': video_id,
1764                         'ps': 'default',
1765                         'eurl': '',
1766                         'gl': 'US',
1767                         'hl': 'en',
1768                     }
1769                     if el:
1770                         query['el'] = el
1771                     if sts:
1772                         query['sts'] = sts
1773                     video_info_webpage = self._download_webpage(
1774                         '%s://www.youtube.com/get_video_info' % proto,
1775                         video_id, note=False,
1776                         errnote='unable to download video info webpage',
1777                         fatal=False, query=query)
1778                     if not video_info_webpage:
1779                         continue
1780                     get_video_info = compat_parse_qs(video_info_webpage)
1781                     if not player_response:
1782                         pl_response = get_video_info.get('player_response', [None])[0]
1783                         if isinstance(pl_response, dict):
1784                             player_response = pl_response
1785                             add_dash_mpd_pr(player_response)
1786                     add_dash_mpd(get_video_info)
1787                     if view_count is None:
1788                         view_count = extract_view_count(get_video_info)
1789                     if not video_info:
1790                         video_info = get_video_info
1791                     get_token = extract_token(get_video_info)
1792                     if get_token:
1793                         # Different get_video_info requests may report different results, e.g.
1794                         # some may report video unavailability, but some may serve it without
1795                         # any complaint (see https://github.com/ytdl-org/youtube-dl/issues/7362,
1796                         # the original webpage as well as el=info and el=embedded get_video_info
1797                         # requests report video unavailability due to geo restriction while
1798                         # el=detailpage succeeds and returns valid data). This is probably
1799                         # due to YouTube measures against IP ranges of hosting providers.
1800                         # Working around by preferring the first succeeded video_info containing
1801                         # the token if no such video_info yet was found.
1802                         token = extract_token(video_info)
1803                         if not token:
1804                             video_info = get_video_info
1805                         break
1806
1807         def extract_unavailable_message():
1808             return self._html_search_regex(
1809                 r'(?s)<h1[^>]+id="unavailable-message"[^>]*>(.+?)</h1>',
1810                 video_webpage, 'unavailable message', default=None)
1811
1812         if not video_info:
1813             unavailable_message = extract_unavailable_message()
1814             if not unavailable_message:
1815                 unavailable_message = 'Unable to extract video data'
1816             raise ExtractorError(
1817                 'YouTube said: %s' % unavailable_message, expected=True, video_id=video_id)
1818
1819         video_details = try_get(
1820             player_response, lambda x: x['videoDetails'], dict) or {}
1821
1822         # title
1823         if 'title' in video_info:
1824             video_title = video_info['title'][0]
1825         elif 'title' in player_response:
1826             video_title = video_details['title']
1827         else:
1828             self._downloader.report_warning('Unable to extract video title')
1829             video_title = '_'
1830
1831         # description
1832         description_original = video_description = get_element_by_id("eow-description", video_webpage)
1833         if video_description:
1834
1835             def replace_url(m):
1836                 redir_url = compat_urlparse.urljoin(url, m.group(1))
1837                 parsed_redir_url = compat_urllib_parse_urlparse(redir_url)
1838                 if re.search(r'^(?:www\.)?(?:youtube(?:-nocookie)?\.com|youtu\.be)$', parsed_redir_url.netloc) and parsed_redir_url.path == '/redirect':
1839                     qs = compat_parse_qs(parsed_redir_url.query)
1840                     q = qs.get('q')
1841                     if q and q[0]:
1842                         return q[0]
1843                 return redir_url
1844
1845             description_original = video_description = re.sub(r'''(?x)
1846                 <a\s+
1847                     (?:[a-zA-Z-]+="[^"]*"\s+)*?
1848                     (?:title|href)="([^"]+)"\s+
1849                     (?:[a-zA-Z-]+="[^"]*"\s+)*?
1850                     class="[^"]*"[^>]*>
1851                 [^<]+\.{3}\s*
1852                 </a>
1853             ''', replace_url, video_description)
1854             video_description = clean_html(video_description)
1855         else:
1856             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
1857             if fd_mobj:
1858                 video_description = unescapeHTML(fd_mobj.group(1))
1859             else:
1860                 video_description = ''
1861
1862         if not smuggled_data.get('force_singlefeed', False):
1863             if not self._downloader.params.get('noplaylist'):
1864                 multifeed_metadata_list = try_get(
1865                     player_response,
1866                     lambda x: x['multicamera']['playerLegacyMulticameraRenderer']['metadataList'],
1867                     compat_str) or try_get(
1868                     video_info, lambda x: x['multifeed_metadata_list'][0], compat_str)
1869                 if multifeed_metadata_list:
1870                     entries = []
1871                     feed_ids = []
1872                     for feed in multifeed_metadata_list.split(','):
1873                         # Unquote should take place before split on comma (,) since textual
1874                         # fields may contain comma as well (see
1875                         # https://github.com/ytdl-org/youtube-dl/issues/8536)
1876                         feed_data = compat_parse_qs(compat_urllib_parse_unquote_plus(feed))
1877                         entries.append({
1878                             '_type': 'url_transparent',
1879                             'ie_key': 'Youtube',
1880                             'url': smuggle_url(
1881                                 '%s://www.youtube.com/watch?v=%s' % (proto, feed_data['id'][0]),
1882                                 {'force_singlefeed': True}),
1883                             'title': '%s (%s)' % (video_title, feed_data['title'][0]),
1884                         })
1885                         feed_ids.append(feed_data['id'][0])
1886                     self.to_screen(
1887                         'Downloading multifeed video (%s) - add --no-playlist to just download video %s'
1888                         % (', '.join(feed_ids), video_id))
1889                     return self.playlist_result(entries, video_id, video_title, video_description)
1890             else:
1891                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1892
1893         if view_count is None:
1894             view_count = extract_view_count(video_info)
1895         if view_count is None and video_details:
1896             view_count = int_or_none(video_details.get('viewCount'))
1897
1898         if is_live is None:
1899             is_live = bool_or_none(dict_get(
1900                 video_details, ('isLive', 'isLiveContent'),
1901                 skip_false_values=False))
1902
1903         # Check for "rental" videos
1904         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1905             raise ExtractorError('"rental" videos not supported. See https://github.com/ytdl-org/youtube-dl/issues/359 for more information.', expected=True)
1906
1907         def _extract_filesize(media_url):
1908             return int_or_none(self._search_regex(
1909                 r'\bclen[=/](\d+)', media_url, 'filesize', default=None))
1910
1911         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1912             self.report_rtmp_download()
1913             formats = [{
1914                 'format_id': '_rtmp',
1915                 'protocol': 'rtmp',
1916                 'url': video_info['conn'][0],
1917                 'player_url': player_url,
1918             }]
1919         elif not is_live and (len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1):
1920             encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
1921             if 'rtmpe%3Dyes' in encoded_url_map:
1922                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/ytdl-org/youtube-dl/issues/343 for more information.', expected=True)
1923             formats_spec = {}
1924             fmt_list = video_info.get('fmt_list', [''])[0]
1925             if fmt_list:
1926                 for fmt in fmt_list.split(','):
1927                     spec = fmt.split('/')
1928                     if len(spec) > 1:
1929                         width_height = spec[1].split('x')
1930                         if len(width_height) == 2:
1931                             formats_spec[spec[0]] = {
1932                                 'resolution': spec[1],
1933                                 'width': int_or_none(width_height[0]),
1934                                 'height': int_or_none(width_height[1]),
1935                             }
1936             q = qualities(['small', 'medium', 'hd720'])
1937             streaming_formats = try_get(player_response, lambda x: x['streamingData']['formats'], list)
1938             if streaming_formats:
1939                 for fmt in streaming_formats:
1940                     itag = str_or_none(fmt.get('itag'))
1941                     if not itag:
1942                         continue
1943                     quality = fmt.get('quality')
1944                     quality_label = fmt.get('qualityLabel') or quality
1945                     formats_spec[itag] = {
1946                         'asr': int_or_none(fmt.get('audioSampleRate')),
1947                         'filesize': int_or_none(fmt.get('contentLength')),
1948                         'format_note': quality_label,
1949                         'fps': int_or_none(fmt.get('fps')),
1950                         'height': int_or_none(fmt.get('height')),
1951                         'quality': q(quality),
1952                         # bitrate for itag 43 is always 2147483647
1953                         'tbr': float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000) if itag != '43' else None,
1954                         'width': int_or_none(fmt.get('width')),
1955                     }
1956             formats = []
1957             for url_data_str in encoded_url_map.split(','):
1958                 url_data = compat_parse_qs(url_data_str)
1959                 if 'itag' not in url_data or 'url' not in url_data or url_data.get('drm_families'):
1960                     continue
1961                 stream_type = int_or_none(try_get(url_data, lambda x: x['stream_type'][0]))
1962                 # Unsupported FORMAT_STREAM_TYPE_OTF
1963                 if stream_type == 3:
1964                     continue
1965                 format_id = url_data['itag'][0]
1966                 url = url_data['url'][0]
1967
1968                 if 's' in url_data or self._downloader.params.get('youtube_include_dash_manifest', True):
1969                     ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
1970                     jsplayer_url_json = self._search_regex(
1971                         ASSETS_RE,
1972                         embed_webpage if age_gate else video_webpage,
1973                         'JS player URL (1)', default=None)
1974                     if not jsplayer_url_json and not age_gate:
1975                         # We need the embed website after all
1976                         if embed_webpage is None:
1977                             embed_url = proto + '://www.youtube.com/embed/%s' % video_id
1978                             embed_webpage = self._download_webpage(
1979                                 embed_url, video_id, 'Downloading embed webpage')
1980                         jsplayer_url_json = self._search_regex(
1981                             ASSETS_RE, embed_webpage, 'JS player URL')
1982
1983                     player_url = json.loads(jsplayer_url_json)
1984                     if player_url is None:
1985                         player_url_json = self._search_regex(
1986                             r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
1987                             video_webpage, 'age gate player URL')
1988                         player_url = json.loads(player_url_json)
1989
1990                 if 'sig' in url_data:
1991                     url += '&signature=' + url_data['sig'][0]
1992                 elif 's' in url_data:
1993                     encrypted_sig = url_data['s'][0]
1994
1995                     if self._downloader.params.get('verbose'):
1996                         if player_url is None:
1997                             player_version = 'unknown'
1998                             player_desc = 'unknown'
1999                         else:
2000                             if player_url.endswith('swf'):
2001                                 player_version = self._search_regex(
2002                                     r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
2003                                     'flash player', fatal=False)
2004                                 player_desc = 'flash player %s' % player_version
2005                             else:
2006                                 player_version = self._search_regex(
2007                                     [r'html5player-([^/]+?)(?:/html5player(?:-new)?)?\.js',
2008                                      r'(?:www|player(?:_ias)?)-([^/]+)(?:/[a-z]{2,3}_[A-Z]{2})?/base\.js'],
2009                                     player_url,
2010                                     'html5 player', fatal=False)
2011                                 player_desc = 'html5 player %s' % player_version
2012
2013                         parts_sizes = self._signature_cache_id(encrypted_sig)
2014                         self.to_screen('{%s} signature length %s, %s' %
2015                                        (format_id, parts_sizes, player_desc))
2016
2017                     signature = self._decrypt_signature(
2018                         encrypted_sig, video_id, player_url, age_gate)
2019                     sp = try_get(url_data, lambda x: x['sp'][0], compat_str) or 'signature'
2020                     url += '&%s=%s' % (sp, signature)
2021                 if 'ratebypass' not in url:
2022                     url += '&ratebypass=yes'
2023
2024                 dct = {
2025                     'format_id': format_id,
2026                     'url': url,
2027                     'player_url': player_url,
2028                 }
2029                 if format_id in self._formats:
2030                     dct.update(self._formats[format_id])
2031                 if format_id in formats_spec:
2032                     dct.update(formats_spec[format_id])
2033
2034                 # Some itags are not included in DASH manifest thus corresponding formats will
2035                 # lack metadata (see https://github.com/ytdl-org/youtube-dl/pull/5993).
2036                 # Trying to extract metadata from url_encoded_fmt_stream_map entry.
2037                 mobj = re.search(r'^(?P<width>\d+)[xX](?P<height>\d+)$', url_data.get('size', [''])[0])
2038                 width, height = (int(mobj.group('width')), int(mobj.group('height'))) if mobj else (None, None)
2039
2040                 filesize = int_or_none(url_data.get(
2041                     'clen', [None])[0]) or _extract_filesize(url)
2042
2043                 quality = url_data.get('quality', [None])[0]
2044
2045                 more_fields = {
2046                     'filesize': filesize,
2047                     'tbr': float_or_none(url_data.get('bitrate', [None])[0], 1000),
2048                     'width': width,
2049                     'height': height,
2050                     'fps': int_or_none(url_data.get('fps', [None])[0]),
2051                     'format_note': url_data.get('quality_label', [None])[0] or quality,
2052                     'quality': q(quality),
2053                 }
2054                 for key, value in more_fields.items():
2055                     if value:
2056                         dct[key] = value
2057                 type_ = url_data.get('type', [None])[0]
2058                 if type_:
2059                     type_split = type_.split(';')
2060                     kind_ext = type_split[0].split('/')
2061                     if len(kind_ext) == 2:
2062                         kind, _ = kind_ext
2063                         dct['ext'] = mimetype2ext(type_split[0])
2064                         if kind in ('audio', 'video'):
2065                             codecs = None
2066                             for mobj in re.finditer(
2067                                     r'(?P<key>[a-zA-Z_-]+)=(?P<quote>["\']?)(?P<val>.+?)(?P=quote)(?:;|$)', type_):
2068                                 if mobj.group('key') == 'codecs':
2069                                     codecs = mobj.group('val')
2070                                     break
2071                             if codecs:
2072                                 dct.update(parse_codecs(codecs))
2073                 if dct.get('acodec') == 'none' or dct.get('vcodec') == 'none':
2074                     dct['downloader_options'] = {
2075                         # Youtube throttles chunks >~10M
2076                         'http_chunk_size': 10485760,
2077                     }
2078                 formats.append(dct)
2079         else:
2080             manifest_url = (
2081                 url_or_none(try_get(
2082                     player_response,
2083                     lambda x: x['streamingData']['hlsManifestUrl'],
2084                     compat_str))
2085                 or url_or_none(try_get(
2086                     video_info, lambda x: x['hlsvp'][0], compat_str)))
2087             if manifest_url:
2088                 formats = []
2089                 m3u8_formats = self._extract_m3u8_formats(
2090                     manifest_url, video_id, 'mp4', fatal=False)
2091                 for a_format in m3u8_formats:
2092                     itag = self._search_regex(
2093                         r'/itag/(\d+)/', a_format['url'], 'itag', default=None)
2094                     if itag:
2095                         a_format['format_id'] = itag
2096                         if itag in self._formats:
2097                             dct = self._formats[itag].copy()
2098                             dct.update(a_format)
2099                             a_format = dct
2100                     a_format['player_url'] = player_url
2101                     # Accept-Encoding header causes failures in live streams on Youtube and Youtube Gaming
2102                     a_format.setdefault('http_headers', {})['Youtubedl-no-compression'] = 'True'
2103                     formats.append(a_format)
2104             else:
2105                 error_message = clean_html(video_info.get('reason', [None])[0])
2106                 if not error_message:
2107                     error_message = extract_unavailable_message()
2108                 if error_message:
2109                     raise ExtractorError(error_message, expected=True)
2110                 raise ExtractorError('no conn, hlsvp, hlsManifestUrl or url_encoded_fmt_stream_map information found in video info')
2111
2112         # uploader
2113         video_uploader = try_get(
2114             video_info, lambda x: x['author'][0],
2115             compat_str) or str_or_none(video_details.get('author'))
2116         if video_uploader:
2117             video_uploader = compat_urllib_parse_unquote_plus(video_uploader)
2118         else:
2119             self._downloader.report_warning('unable to extract uploader name')
2120
2121         # uploader_id
2122         video_uploader_id = None
2123         video_uploader_url = None
2124         mobj = re.search(
2125             r'<link itemprop="url" href="(?P<uploader_url>https?://www\.youtube\.com/(?:user|channel)/(?P<uploader_id>[^"]+))">',
2126             video_webpage)
2127         if mobj is not None:
2128             video_uploader_id = mobj.group('uploader_id')
2129             video_uploader_url = mobj.group('uploader_url')
2130         else:
2131             self._downloader.report_warning('unable to extract uploader nickname')
2132
2133         channel_id = (
2134             str_or_none(video_details.get('channelId'))
2135             or self._html_search_meta(
2136                 'channelId', video_webpage, 'channel id', default=None)
2137             or self._search_regex(
2138                 r'data-channel-external-id=(["\'])(?P<id>(?:(?!\1).)+)\1',
2139                 video_webpage, 'channel id', default=None, group='id'))
2140         channel_url = 'http://www.youtube.com/channel/%s' % channel_id if channel_id else None
2141
2142         # thumbnail image
2143         # We try first to get a high quality image:
2144         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
2145                             video_webpage, re.DOTALL)
2146         if m_thumb is not None:
2147             video_thumbnail = m_thumb.group(1)
2148         elif 'thumbnail_url' not in video_info:
2149             self._downloader.report_warning('unable to extract video thumbnail')
2150             video_thumbnail = None
2151         else:   # don't panic if we can't find it
2152             video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0])
2153
2154         # upload date
2155         upload_date = self._html_search_meta(
2156             'datePublished', video_webpage, 'upload date', default=None)
2157         if not upload_date:
2158             upload_date = self._search_regex(
2159                 [r'(?s)id="eow-date.*?>(.*?)</span>',
2160                  r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']'],
2161                 video_webpage, 'upload date', default=None)
2162         upload_date = unified_strdate(upload_date)
2163
2164         video_license = self._html_search_regex(
2165             r'<h4[^>]+class="title"[^>]*>\s*License\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li',
2166             video_webpage, 'license', default=None)
2167
2168         m_music = re.search(
2169             r'''(?x)
2170                 <h4[^>]+class="title"[^>]*>\s*Music\s*</h4>\s*
2171                 <ul[^>]*>\s*
2172                 <li>(?P<title>.+?)
2173                 by (?P<creator>.+?)
2174                 (?:
2175                     \(.+?\)|
2176                     <a[^>]*
2177                         (?:
2178                             \bhref=["\']/red[^>]*>|             # drop possible
2179                             >\s*Listen ad-free with YouTube Red # YouTube Red ad
2180                         )
2181                     .*?
2182                 )?</li
2183             ''',
2184             video_webpage)
2185         if m_music:
2186             video_alt_title = remove_quotes(unescapeHTML(m_music.group('title')))
2187             video_creator = clean_html(m_music.group('creator'))
2188         else:
2189             video_alt_title = video_creator = None
2190
2191         def extract_meta(field):
2192             return self._html_search_regex(
2193                 r'<h4[^>]+class="title"[^>]*>\s*%s\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li>\s*' % field,
2194                 video_webpage, field, default=None)
2195
2196         track = extract_meta('Song')
2197         artist = extract_meta('Artist')
2198         album = extract_meta('Album')
2199
2200         # Youtube Music Auto-generated description
2201         release_date = release_year = None
2202         if video_description:
2203             mobj = re.search(r'(?s)Provided to YouTube by [^\n]+\n+(?P<track>[^·]+)·(?P<artist>[^\n]+)\n+(?P<album>[^\n]+)(?:.+?℗\s*(?P<release_year>\d{4})(?!\d))?(?:.+?Released on\s*:\s*(?P<release_date>\d{4}-\d{2}-\d{2}))?(.+?\nArtist\s*:\s*(?P<clean_artist>[^\n]+))?', video_description)
2204             if mobj:
2205                 if not track:
2206                     track = mobj.group('track').strip()
2207                 if not artist:
2208                     artist = mobj.group('clean_artist') or ', '.join(a.strip() for a in mobj.group('artist').split('·'))
2209                 if not album:
2210                     album = mobj.group('album'.strip())
2211                 release_year = mobj.group('release_year')
2212                 release_date = mobj.group('release_date')
2213                 if release_date:
2214                     release_date = release_date.replace('-', '')
2215                     if not release_year:
2216                         release_year = int(release_date[:4])
2217                 if release_year:
2218                     release_year = int(release_year)
2219
2220         m_episode = re.search(
2221             r'<div[^>]+id="watch7-headline"[^>]*>\s*<span[^>]*>.*?>(?P<series>[^<]+)</a></b>\s*S(?P<season>\d+)\s*•\s*E(?P<episode>\d+)</span>',
2222             video_webpage)
2223         if m_episode:
2224             series = unescapeHTML(m_episode.group('series'))
2225             season_number = int(m_episode.group('season'))
2226             episode_number = int(m_episode.group('episode'))
2227         else:
2228             series = season_number = episode_number = None
2229
2230         m_cat_container = self._search_regex(
2231             r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
2232             video_webpage, 'categories', default=None)
2233         if m_cat_container:
2234             category = self._html_search_regex(
2235                 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
2236                 default=None)
2237             video_categories = None if category is None else [category]
2238         else:
2239             video_categories = None
2240
2241         video_tags = [
2242             unescapeHTML(m.group('content'))
2243             for m in re.finditer(self._meta_regex('og:video:tag'), video_webpage)]
2244
2245         def _extract_count(count_name):
2246             return str_to_int(self._search_regex(
2247                 r'-%s-button[^>]+><span[^>]+class="yt-uix-button-content"[^>]*>([\d,]+)</span>'
2248                 % re.escape(count_name),
2249                 video_webpage, count_name, default=None))
2250
2251         like_count = _extract_count('like')
2252         dislike_count = _extract_count('dislike')
2253
2254         if view_count is None:
2255             view_count = str_to_int(self._search_regex(
2256                 r'<[^>]+class=["\']watch-view-count[^>]+>\s*([\d,\s]+)', video_webpage,
2257                 'view count', default=None))
2258
2259         average_rating = (
2260             float_or_none(video_details.get('averageRating'))
2261             or try_get(video_info, lambda x: float_or_none(x['avg_rating'][0])))
2262
2263         # subtitles
2264         video_subtitles = self.extract_subtitles(video_id, video_webpage)
2265         automatic_captions = self.extract_automatic_captions(video_id, video_webpage)
2266
2267         video_duration = try_get(
2268             video_info, lambda x: int_or_none(x['length_seconds'][0]))
2269         if not video_duration:
2270             video_duration = int_or_none(video_details.get('lengthSeconds'))
2271         if not video_duration:
2272             video_duration = parse_duration(self._html_search_meta(
2273                 'duration', video_webpage, 'video duration'))
2274
2275         # annotations
2276         video_annotations = None
2277         if self._downloader.params.get('writeannotations', False):
2278             video_annotations = self._extract_annotations(video_id)
2279
2280         chapters = self._extract_chapters(description_original, video_duration)
2281
2282         # Look for the DASH manifest
2283         if self._downloader.params.get('youtube_include_dash_manifest', True):
2284             dash_mpd_fatal = True
2285             for mpd_url in dash_mpds:
2286                 dash_formats = {}
2287                 try:
2288                     def decrypt_sig(mobj):
2289                         s = mobj.group(1)
2290                         dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
2291                         return '/signature/%s' % dec_s
2292
2293                     mpd_url = re.sub(r'/s/([a-fA-F0-9\.]+)', decrypt_sig, mpd_url)
2294
2295                     for df in self._extract_mpd_formats(
2296                             mpd_url, video_id, fatal=dash_mpd_fatal,
2297                             formats_dict=self._formats):
2298                         if not df.get('filesize'):
2299                             df['filesize'] = _extract_filesize(df['url'])
2300                         # Do not overwrite DASH format found in some previous DASH manifest
2301                         if df['format_id'] not in dash_formats:
2302                             dash_formats[df['format_id']] = df
2303                         # Additional DASH manifests may end up in HTTP Error 403 therefore
2304                         # allow them to fail without bug report message if we already have
2305                         # some DASH manifest succeeded. This is temporary workaround to reduce
2306                         # burst of bug reports until we figure out the reason and whether it
2307                         # can be fixed at all.
2308                         dash_mpd_fatal = False
2309                 except (ExtractorError, KeyError) as e:
2310                     self.report_warning(
2311                         'Skipping DASH manifest: %r' % e, video_id)
2312                 if dash_formats:
2313                     # Remove the formats we found through non-DASH, they
2314                     # contain less info and it can be wrong, because we use
2315                     # fixed values (for example the resolution). See
2316                     # https://github.com/ytdl-org/youtube-dl/issues/5774 for an
2317                     # example.
2318                     formats = [f for f in formats if f['format_id'] not in dash_formats.keys()]
2319                     formats.extend(dash_formats.values())
2320
2321         # Check for malformed aspect ratio
2322         stretched_m = re.search(
2323             r'<meta\s+property="og:video:tag".*?content="yt:stretch=(?P<w>[0-9]+):(?P<h>[0-9]+)">',
2324             video_webpage)
2325         if stretched_m:
2326             w = float(stretched_m.group('w'))
2327             h = float(stretched_m.group('h'))
2328             # yt:stretch may hold invalid ratio data (e.g. for Q39EVAstoRM ratio is 17:0).
2329             # We will only process correct ratios.
2330             if w > 0 and h > 0:
2331                 ratio = w / h
2332                 for f in formats:
2333                     if f.get('vcodec') != 'none':
2334                         f['stretched_ratio'] = ratio
2335
2336         if not formats:
2337             token = extract_token(video_info)
2338             if not token:
2339                 if 'reason' in video_info:
2340                     if 'The uploader has not made this video available in your country.' in video_info['reason']:
2341                         regions_allowed = self._html_search_meta(
2342                             'regionsAllowed', video_webpage, default=None)
2343                         countries = regions_allowed.split(',') if regions_allowed else None
2344                         self.raise_geo_restricted(
2345                             msg=video_info['reason'][0], countries=countries)
2346                     reason = video_info['reason'][0]
2347                     if 'Invalid parameters' in reason:
2348                         unavailable_message = extract_unavailable_message()
2349                         if unavailable_message:
2350                             reason = unavailable_message
2351                     raise ExtractorError(
2352                         'YouTube said: %s' % reason,
2353                         expected=True, video_id=video_id)
2354                 else:
2355                     raise ExtractorError(
2356                         '"token" parameter not in video info for unknown reason',
2357                         video_id=video_id)
2358
2359         if not formats and (video_info.get('license_info') or try_get(player_response, lambda x: x['streamingData']['licenseInfos'])):
2360             raise ExtractorError('This video is DRM protected.', expected=True)
2361
2362         self._sort_formats(formats)
2363
2364         self.mark_watched(video_id, video_info, player_response)
2365
2366         return {
2367             'id': video_id,
2368             'uploader': video_uploader,
2369             'uploader_id': video_uploader_id,
2370             'uploader_url': video_uploader_url,
2371             'channel_id': channel_id,
2372             'channel_url': channel_url,
2373             'upload_date': upload_date,
2374             'license': video_license,
2375             'creator': video_creator or artist,
2376             'title': video_title,
2377             'alt_title': video_alt_title or track,
2378             'thumbnail': video_thumbnail,
2379             'description': video_description,
2380             'categories': video_categories,
2381             'tags': video_tags,
2382             'subtitles': video_subtitles,
2383             'automatic_captions': automatic_captions,
2384             'duration': video_duration,
2385             'age_limit': 18 if age_gate else 0,
2386             'annotations': video_annotations,
2387             'chapters': chapters,
2388             'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
2389             'view_count': view_count,
2390             'like_count': like_count,
2391             'dislike_count': dislike_count,
2392             'average_rating': average_rating,
2393             'formats': formats,
2394             'is_live': is_live,
2395             'start_time': start_time,
2396             'end_time': end_time,
2397             'series': series,
2398             'season_number': season_number,
2399             'episode_number': episode_number,
2400             'track': track,
2401             'artist': artist,
2402             'album': album,
2403             'release_date': release_date,
2404             'release_year': release_year,
2405         }
2406
2407
2408 class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
2409     IE_DESC = 'YouTube.com playlists'
2410     _VALID_URL = r"""(?x)(?:
2411                         (?:https?://)?
2412                         (?:\w+\.)?
2413                         (?:
2414                             (?:
2415                                 youtube\.com|
2416                                 invidio\.us
2417                             )
2418                             /
2419                             (?:
2420                                (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/(?:videoseries|[0-9A-Za-z_-]{11}))
2421                                \? (?:.*?[&;])*? (?:p|a|list)=
2422                             |  p/
2423                             )|
2424                             youtu\.be/[0-9A-Za-z_-]{11}\?.*?\blist=
2425                         )
2426                         (
2427                             (?:PL|LL|EC|UU|FL|RD|UL|TL|OLAK5uy_)?[0-9A-Za-z-_]{10,}
2428                             # Top tracks, they can also include dots
2429                             |(?:MC)[\w\.]*
2430                         )
2431                         .*
2432                      |
2433                         (%(playlist_id)s)
2434                      )""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
2435     _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
2436     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)(?:[^>]+>(?P<title>[^<]+))?'
2437     IE_NAME = 'youtube:playlist'
2438     _TESTS = [{
2439         'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
2440         'info_dict': {
2441             'title': 'ytdl test PL',
2442             'id': 'PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
2443         },
2444         'playlist_count': 3,
2445     }, {
2446         'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
2447         'info_dict': {
2448             'id': 'PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
2449             'title': 'YDL_Empty_List',
2450         },
2451         'playlist_count': 0,
2452         'skip': 'This playlist is private',
2453     }, {
2454         'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
2455         'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
2456         'info_dict': {
2457             'title': '29C3: Not my department',
2458             'id': 'PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
2459         },
2460         'playlist_count': 95,
2461     }, {
2462         'note': 'issue #673',
2463         'url': 'PLBB231211A4F62143',
2464         'info_dict': {
2465             'title': '[OLD]Team Fortress 2 (Class-based LP)',
2466             'id': 'PLBB231211A4F62143',
2467         },
2468         'playlist_mincount': 26,
2469     }, {
2470         'note': 'Large playlist',
2471         'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
2472         'info_dict': {
2473             'title': 'Uploads from Cauchemar',
2474             'id': 'UUBABnxM4Ar9ten8Mdjj1j0Q',
2475         },
2476         'playlist_mincount': 799,
2477     }, {
2478         'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
2479         'info_dict': {
2480             'title': 'YDL_safe_search',
2481             'id': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
2482         },
2483         'playlist_count': 2,
2484         'skip': 'This playlist is private',
2485     }, {
2486         'note': 'embedded',
2487         'url': 'https://www.youtube.com/embed/videoseries?list=PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
2488         'playlist_count': 4,
2489         'info_dict': {
2490             'title': 'JODA15',
2491             'id': 'PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
2492         }
2493     }, {
2494         'url': 'http://www.youtube.com/embed/_xDOZElKyNU?list=PLsyOSbh5bs16vubvKePAQ1x3PhKavfBIl',
2495         'playlist_mincount': 485,
2496         'info_dict': {
2497             'title': '2017 華語最新單曲 (2/24更新)',
2498             'id': 'PLsyOSbh5bs16vubvKePAQ1x3PhKavfBIl',
2499         }
2500     }, {
2501         'note': 'Embedded SWF player',
2502         'url': 'https://www.youtube.com/p/YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ?hl=en_US&fs=1&rel=0',
2503         'playlist_count': 4,
2504         'info_dict': {
2505             'title': 'JODA7',
2506             'id': 'YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ',
2507         }
2508     }, {
2509         'note': 'Buggy playlist: the webpage has a "Load more" button but it doesn\'t have more videos',
2510         'url': 'https://www.youtube.com/playlist?list=UUXw-G3eDE9trcvY2sBMM_aA',
2511         'info_dict': {
2512             'title': 'Uploads from Interstellar Movie',
2513             'id': 'UUXw-G3eDE9trcvY2sBMM_aA',
2514         },
2515         'playlist_mincount': 21,
2516     }, {
2517         # Playlist URL that does not actually serve a playlist
2518         'url': 'https://www.youtube.com/watch?v=FqZTN594JQw&list=PLMYEtVRpaqY00V9W81Cwmzp6N6vZqfUKD4',
2519         'info_dict': {
2520             'id': 'FqZTN594JQw',
2521             'ext': 'webm',
2522             'title': "Smiley's People 01 detective, Adventure Series, Action",
2523             'uploader': 'STREEM',
2524             'uploader_id': 'UCyPhqAZgwYWZfxElWVbVJng',
2525             'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCyPhqAZgwYWZfxElWVbVJng',
2526             'upload_date': '20150526',
2527             'license': 'Standard YouTube License',
2528             'description': 'md5:507cdcb5a49ac0da37a920ece610be80',
2529             'categories': ['People & Blogs'],
2530             'tags': list,
2531             'view_count': int,
2532             'like_count': int,
2533             'dislike_count': int,
2534         },
2535         'params': {
2536             'skip_download': True,
2537         },
2538         'add_ie': [YoutubeIE.ie_key()],
2539     }, {
2540         'url': 'https://youtu.be/yeWKywCrFtk?list=PL2qgrgXsNUG5ig9cat4ohreBjYLAPC0J5',
2541         'info_dict': {
2542             'id': 'yeWKywCrFtk',
2543             'ext': 'mp4',
2544             'title': 'Small Scale Baler and Braiding Rugs',
2545             'uploader': 'Backus-Page House Museum',
2546             'uploader_id': 'backuspagemuseum',
2547             'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/backuspagemuseum',
2548             'upload_date': '20161008',
2549             'license': 'Standard YouTube License',
2550             'description': 'md5:800c0c78d5eb128500bffd4f0b4f2e8a',
2551             'categories': ['Nonprofits & Activism'],
2552             'tags': list,
2553             'like_count': int,
2554             'dislike_count': int,
2555         },
2556         'params': {
2557             'noplaylist': True,
2558             'skip_download': True,
2559         },
2560     }, {
2561         'url': 'https://youtu.be/uWyaPkt-VOI?list=PL9D9FC436B881BA21',
2562         'only_matching': True,
2563     }, {
2564         'url': 'TLGGrESM50VT6acwMjAyMjAxNw',
2565         'only_matching': True,
2566     }, {
2567         # music album playlist
2568         'url': 'OLAK5uy_m4xAFdmMC5rX3Ji3g93pQe3hqLZw_9LhM',
2569         'only_matching': True,
2570     }, {
2571         'url': 'https://invidio.us/playlist?list=PLDIoUOhQQPlXr63I_vwF9GD8sAKh77dWU',
2572         'only_matching': True,
2573     }]
2574
2575     def _real_initialize(self):
2576         self._login()
2577
2578     def _extract_mix(self, playlist_id):
2579         # The mixes are generated from a single video
2580         # the id of the playlist is just 'RD' + video_id
2581         ids = []
2582         last_id = playlist_id[-11:]
2583         for n in itertools.count(1):
2584             url = 'https://youtube.com/watch?v=%s&list=%s' % (last_id, playlist_id)
2585             webpage = self._download_webpage(
2586                 url, playlist_id, 'Downloading page {0} of Youtube mix'.format(n))
2587             new_ids = orderedSet(re.findall(
2588                 r'''(?xs)data-video-username=".*?".*?
2589                            href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
2590                 webpage))
2591             # Fetch new pages until all the videos are repeated, it seems that
2592             # there are always 51 unique videos.
2593             new_ids = [_id for _id in new_ids if _id not in ids]
2594             if not new_ids:
2595                 break
2596             ids.extend(new_ids)
2597             last_id = ids[-1]
2598
2599         url_results = self._ids_to_results(ids)
2600
2601         search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
2602         title_span = (
2603             search_title('playlist-title')
2604             or search_title('title long-title')
2605             or search_title('title'))
2606         title = clean_html(title_span)
2607
2608         return self.playlist_result(url_results, playlist_id, title)
2609
2610     def _extract_playlist(self, playlist_id):
2611         url = self._TEMPLATE_URL % playlist_id
2612         page = self._download_webpage(url, playlist_id)
2613
2614         # the yt-alert-message now has tabindex attribute (see https://github.com/ytdl-org/youtube-dl/issues/11604)
2615         for match in re.findall(r'<div class="yt-alert-message"[^>]*>([^<]+)</div>', page):
2616             match = match.strip()
2617             # Check if the playlist exists or is private
2618             mobj = re.match(r'[^<]*(?:The|This) playlist (?P<reason>does not exist|is private)[^<]*', match)
2619             if mobj:
2620                 reason = mobj.group('reason')
2621                 message = 'This playlist %s' % reason
2622                 if 'private' in reason:
2623                     message += ', use --username or --netrc to access it'
2624                 message += '.'
2625                 raise ExtractorError(message, expected=True)
2626             elif re.match(r'[^<]*Invalid parameters[^<]*', match):
2627                 raise ExtractorError(
2628                     'Invalid parameters. Maybe URL is incorrect.',
2629                     expected=True)
2630             elif re.match(r'[^<]*Choose your language[^<]*', match):
2631                 continue
2632             else:
2633                 self.report_warning('Youtube gives an alert message: ' + match)
2634
2635         playlist_title = self._html_search_regex(
2636             r'(?s)<h1 class="pl-header-title[^"]*"[^>]*>\s*(.*?)\s*</h1>',
2637             page, 'title', default=None)
2638
2639         _UPLOADER_BASE = r'class=["\']pl-header-details[^>]+>\s*<li>\s*<a[^>]+\bhref='
2640         uploader = self._search_regex(
2641             r'%s["\']/(?:user|channel)/[^>]+>([^<]+)' % _UPLOADER_BASE,
2642             page, 'uploader', default=None)
2643         mobj = re.search(
2644             r'%s(["\'])(?P<path>/(?:user|channel)/(?P<uploader_id>.+?))\1' % _UPLOADER_BASE,
2645             page)
2646         if mobj:
2647             uploader_id = mobj.group('uploader_id')
2648             uploader_url = compat_urlparse.urljoin(url, mobj.group('path'))
2649         else:
2650             uploader_id = uploader_url = None
2651
2652         has_videos = True
2653
2654         if not playlist_title:
2655             try:
2656                 # Some playlist URLs don't actually serve a playlist (e.g.
2657                 # https://www.youtube.com/watch?v=FqZTN594JQw&list=PLMYEtVRpaqY00V9W81Cwmzp6N6vZqfUKD4)
2658                 next(self._entries(page, playlist_id))
2659             except StopIteration:
2660                 has_videos = False
2661
2662         playlist = self.playlist_result(
2663             self._entries(page, playlist_id), playlist_id, playlist_title)
2664         playlist.update({
2665             'uploader': uploader,
2666             'uploader_id': uploader_id,
2667             'uploader_url': uploader_url,
2668         })
2669
2670         return has_videos, playlist
2671
2672     def _check_download_just_video(self, url, playlist_id):
2673         # Check if it's a video-specific URL
2674         query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
2675         video_id = query_dict.get('v', [None])[0] or self._search_regex(
2676             r'(?:(?:^|//)youtu\.be/|youtube\.com/embed/(?!videoseries))([0-9A-Za-z_-]{11})', url,
2677             'video id', default=None)
2678         if video_id:
2679             if self._downloader.params.get('noplaylist'):
2680                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
2681                 return video_id, self.url_result(video_id, 'Youtube', video_id=video_id)
2682             else:
2683                 self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
2684                 return video_id, None
2685         return None, None
2686
2687     def _real_extract(self, url):
2688         # Extract playlist id
2689         mobj = re.match(self._VALID_URL, url)
2690         if mobj is None:
2691             raise ExtractorError('Invalid URL: %s' % url)
2692         playlist_id = mobj.group(1) or mobj.group(2)
2693
2694         video_id, video = self._check_download_just_video(url, playlist_id)
2695         if video:
2696             return video
2697
2698         if playlist_id.startswith(('RD', 'UL', 'PU')):
2699             # Mixes require a custom extraction process
2700             return self._extract_mix(playlist_id)
2701
2702         has_videos, playlist = self._extract_playlist(playlist_id)
2703         if has_videos or not video_id:
2704             return playlist
2705
2706         # Some playlist URLs don't actually serve a playlist (see
2707         # https://github.com/ytdl-org/youtube-dl/issues/10537).
2708         # Fallback to plain video extraction if there is a video id
2709         # along with playlist id.
2710         return self.url_result(video_id, 'Youtube', video_id=video_id)
2711
2712
2713 class YoutubeChannelIE(YoutubePlaylistBaseInfoExtractor):
2714     IE_DESC = 'YouTube.com channels'
2715     _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com|(?:www\.)?invidio\.us)/channel/(?P<id>[0-9A-Za-z_-]+)'
2716     _TEMPLATE_URL = 'https://www.youtube.com/channel/%s/videos'
2717     _VIDEO_RE = r'(?:title="(?P<title>[^"]+)"[^>]+)?href="/watch\?v=(?P<id>[0-9A-Za-z_-]+)&?'
2718     IE_NAME = 'youtube:channel'
2719     _TESTS = [{
2720         'note': 'paginated channel',
2721         'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w',
2722         'playlist_mincount': 91,
2723         'info_dict': {
2724             'id': 'UUKfVa3S1e4PHvxWcwyMMg8w',
2725             'title': 'Uploads from lex will',
2726         }
2727     }, {
2728         'note': 'Age restricted channel',
2729         # from https://www.youtube.com/user/DeusExOfficial
2730         'url': 'https://www.youtube.com/channel/UCs0ifCMCm1icqRbqhUINa0w',
2731         'playlist_mincount': 64,
2732         'info_dict': {
2733             'id': 'UUs0ifCMCm1icqRbqhUINa0w',
2734             'title': 'Uploads from Deus Ex',
2735         },
2736     }, {
2737         'url': 'https://invidio.us/channel/UC23qupoDRn9YOAVzeoxjOQA',
2738         'only_matching': True,
2739     }]
2740
2741     @classmethod
2742     def suitable(cls, url):
2743         return (False if YoutubePlaylistsIE.suitable(url) or YoutubeLiveIE.suitable(url)
2744                 else super(YoutubeChannelIE, cls).suitable(url))
2745
2746     def _build_template_url(self, url, channel_id):
2747         return self._TEMPLATE_URL % channel_id
2748
2749     def _real_extract(self, url):
2750         channel_id = self._match_id(url)
2751
2752         url = self._build_template_url(url, channel_id)
2753
2754         # Channel by page listing is restricted to 35 pages of 30 items, i.e. 1050 videos total (see #5778)
2755         # Workaround by extracting as a playlist if managed to obtain channel playlist URL
2756         # otherwise fallback on channel by page extraction
2757         channel_page = self._download_webpage(
2758             url + '?view=57', channel_id,
2759             'Downloading channel page', fatal=False)
2760         if channel_page is False:
2761             channel_playlist_id = False
2762         else:
2763             channel_playlist_id = self._html_search_meta(
2764                 'channelId', channel_page, 'channel id', default=None)
2765             if not channel_playlist_id:
2766                 channel_url = self._html_search_meta(
2767                     ('al:ios:url', 'twitter:app:url:iphone', 'twitter:app:url:ipad'),
2768                     channel_page, 'channel url', default=None)
2769                 if channel_url:
2770                     channel_playlist_id = self._search_regex(
2771                         r'vnd\.youtube://user/([0-9A-Za-z_-]+)',
2772                         channel_url, 'channel id', default=None)
2773         if channel_playlist_id and channel_playlist_id.startswith('UC'):
2774             playlist_id = 'UU' + channel_playlist_id[2:]
2775             return self.url_result(
2776                 compat_urlparse.urljoin(url, '/playlist?list=%s' % playlist_id), 'YoutubePlaylist')
2777
2778         channel_page = self._download_webpage(url, channel_id, 'Downloading page #1')
2779         autogenerated = re.search(r'''(?x)
2780                 class="[^"]*?(?:
2781                     channel-header-autogenerated-label|
2782                     yt-channel-title-autogenerated
2783                 )[^"]*"''', channel_page) is not None
2784
2785         if autogenerated:
2786             # The videos are contained in a single page
2787             # the ajax pages can't be used, they are empty
2788             entries = [
2789                 self.url_result(
2790                     video_id, 'Youtube', video_id=video_id,
2791                     video_title=video_title)
2792                 for video_id, video_title in self.extract_videos_from_page(channel_page)]
2793             return self.playlist_result(entries, channel_id)
2794
2795         try:
2796             next(self._entries(channel_page, channel_id))
2797         except StopIteration:
2798             alert_message = self._html_search_regex(
2799                 r'(?s)<div[^>]+class=(["\']).*?\byt-alert-message\b.*?\1[^>]*>(?P<alert>[^<]+)</div>',
2800                 channel_page, 'alert', default=None, group='alert')
2801             if alert_message:
2802                 raise ExtractorError('Youtube said: %s' % alert_message, expected=True)
2803
2804         return self.playlist_result(self._entries(channel_page, channel_id), channel_id)
2805
2806
2807 class YoutubeUserIE(YoutubeChannelIE):
2808     IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
2809     _VALID_URL = r'(?:(?:https?://(?:\w+\.)?youtube\.com/(?:(?P<user>user|c)/)?(?!(?:attribution_link|watch|results|shared)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)'
2810     _TEMPLATE_URL = 'https://www.youtube.com/%s/%s/videos'
2811     IE_NAME = 'youtube:user'
2812
2813     _TESTS = [{
2814         'url': 'https://www.youtube.com/user/TheLinuxFoundation',
2815         'playlist_mincount': 320,
2816         'info_dict': {
2817             'id': 'UUfX55Sx5hEFjoC3cNs6mCUQ',
2818             'title': 'Uploads from The Linux Foundation',
2819         }
2820     }, {
2821         # Only available via https://www.youtube.com/c/12minuteathlete/videos
2822         # but not https://www.youtube.com/user/12minuteathlete/videos
2823         'url': 'https://www.youtube.com/c/12minuteathlete/videos',
2824         'playlist_mincount': 249,
2825         'info_dict': {
2826             'id': 'UUVjM-zV6_opMDx7WYxnjZiQ',
2827             'title': 'Uploads from 12 Minute Athlete',
2828         }
2829     }, {
2830         'url': 'ytuser:phihag',
2831         'only_matching': True,
2832     }, {
2833         'url': 'https://www.youtube.com/c/gametrailers',
2834         'only_matching': True,
2835     }, {
2836         'url': 'https://www.youtube.com/gametrailers',
2837         'only_matching': True,
2838     }, {
2839         # This channel is not available, geo restricted to JP
2840         'url': 'https://www.youtube.com/user/kananishinoSMEJ/videos',
2841         'only_matching': True,
2842     }]
2843
2844     @classmethod
2845     def suitable(cls, url):
2846         # Don't return True if the url can be extracted with other youtube
2847         # extractor, the regex would is too permissive and it would match.
2848         other_yt_ies = iter(klass for (name, klass) in globals().items() if name.startswith('Youtube') and name.endswith('IE') and klass is not cls)
2849         if any(ie.suitable(url) for ie in other_yt_ies):
2850             return False
2851         else:
2852             return super(YoutubeUserIE, cls).suitable(url)
2853
2854     def _build_template_url(self, url, channel_id):
2855         mobj = re.match(self._VALID_URL, url)
2856         return self._TEMPLATE_URL % (mobj.group('user') or 'user', mobj.group('id'))
2857
2858
2859 class YoutubeLiveIE(YoutubeBaseInfoExtractor):
2860     IE_DESC = 'YouTube.com live streams'
2861     _VALID_URL = r'(?P<base_url>https?://(?:\w+\.)?youtube\.com/(?:(?:user|channel|c)/)?(?P<id>[^/]+))/live'
2862     IE_NAME = 'youtube:live'
2863
2864     _TESTS = [{
2865         'url': 'https://www.youtube.com/user/TheYoungTurks/live',
2866         'info_dict': {
2867             'id': 'a48o2S1cPoo',
2868             'ext': 'mp4',
2869             'title': 'The Young Turks - Live Main Show',
2870             'uploader': 'The Young Turks',
2871             'uploader_id': 'TheYoungTurks',
2872             'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/TheYoungTurks',
2873             'upload_date': '20150715',
2874             'license': 'Standard YouTube License',
2875             'description': 'md5:438179573adcdff3c97ebb1ee632b891',
2876             'categories': ['News & Politics'],
2877             'tags': ['Cenk Uygur (TV Program Creator)', 'The Young Turks (Award-Winning Work)', 'Talk Show (TV Genre)'],
2878             'like_count': int,
2879             'dislike_count': int,
2880         },
2881         'params': {
2882             'skip_download': True,
2883         },
2884     }, {
2885         'url': 'https://www.youtube.com/channel/UC1yBKRuGpC1tSM73A0ZjYjQ/live',
2886         'only_matching': True,
2887     }, {
2888         'url': 'https://www.youtube.com/c/CommanderVideoHq/live',
2889         'only_matching': True,
2890     }, {
2891         'url': 'https://www.youtube.com/TheYoungTurks/live',
2892         'only_matching': True,
2893     }]
2894
2895     def _real_extract(self, url):
2896         mobj = re.match(self._VALID_URL, url)
2897         channel_id = mobj.group('id')
2898         base_url = mobj.group('base_url')
2899         webpage = self._download_webpage(url, channel_id, fatal=False)
2900         if webpage:
2901             page_type = self._og_search_property(
2902                 'type', webpage, 'page type', default='')
2903             video_id = self._html_search_meta(
2904                 'videoId', webpage, 'video id', default=None)
2905             if page_type.startswith('video') and video_id and re.match(
2906                     r'^[0-9A-Za-z_-]{11}$', video_id):
2907                 return self.url_result(video_id, YoutubeIE.ie_key())
2908         return self.url_result(base_url)
2909
2910
2911 class YoutubePlaylistsIE(YoutubePlaylistsBaseInfoExtractor):
2912     IE_DESC = 'YouTube.com user/channel playlists'
2913     _VALID_URL = r'https?://(?:\w+\.)?youtube\.com/(?:user|channel)/(?P<id>[^/]+)/playlists'
2914     IE_NAME = 'youtube:playlists'
2915
2916     _TESTS = [{
2917         'url': 'https://www.youtube.com/user/ThirstForScience/playlists',
2918         'playlist_mincount': 4,
2919         'info_dict': {
2920             'id': 'ThirstForScience',
2921             'title': 'Thirst for Science',
2922         },
2923     }, {
2924         # with "Load more" button
2925         'url': 'https://www.youtube.com/user/igorkle1/playlists?view=1&sort=dd',
2926         'playlist_mincount': 70,
2927         'info_dict': {
2928             'id': 'igorkle1',
2929             'title': 'Игорь Клейнер',
2930         },
2931     }, {
2932         'url': 'https://www.youtube.com/channel/UCiU1dHvZObB2iP6xkJ__Icw/playlists',
2933         'playlist_mincount': 17,
2934         'info_dict': {
2935             'id': 'UCiU1dHvZObB2iP6xkJ__Icw',
2936             'title': 'Chem Player',
2937         },
2938     }]
2939
2940
2941 class YoutubeSearchBaseInfoExtractor(YoutubePlaylistBaseInfoExtractor):
2942     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})(?:[^"]*"[^>]+\btitle="(?P<title>[^"]+))?'
2943
2944
2945 class YoutubeSearchIE(SearchInfoExtractor, YoutubeSearchBaseInfoExtractor):
2946     IE_DESC = 'YouTube.com searches'
2947     # there doesn't appear to be a real limit, for example if you search for
2948     # 'python' you get more than 8.000.000 results
2949     _MAX_RESULTS = float('inf')
2950     IE_NAME = 'youtube:search'
2951     _SEARCH_KEY = 'ytsearch'
2952     _EXTRA_QUERY_ARGS = {}
2953     _TESTS = []
2954
2955     def _get_n_results(self, query, n):
2956         """Get a specified number of results for a query"""
2957
2958         videos = []
2959         limit = n
2960
2961         url_query = {
2962             'search_query': query.encode('utf-8'),
2963         }
2964         url_query.update(self._EXTRA_QUERY_ARGS)
2965         result_url = 'https://www.youtube.com/results?' + compat_urllib_parse_urlencode(url_query)
2966
2967         for pagenum in itertools.count(1):
2968             data = self._download_json(
2969                 result_url, video_id='query "%s"' % query,
2970                 note='Downloading page %s' % pagenum,
2971                 errnote='Unable to download API page',
2972                 query={'spf': 'navigate'})
2973             html_content = data[1]['body']['content']
2974
2975             if 'class="search-message' in html_content:
2976                 raise ExtractorError(
2977                     '[youtube] No video results', expected=True)
2978
2979             new_videos = list(self._process_page(html_content))
2980             videos += new_videos
2981             if not new_videos or len(videos) > limit:
2982                 break
2983             next_link = self._html_search_regex(
2984                 r'href="(/results\?[^"]*\bsp=[^"]+)"[^>]*>\s*<span[^>]+class="[^"]*\byt-uix-button-content\b[^"]*"[^>]*>Next',
2985                 html_content, 'next link', default=None)
2986             if next_link is None:
2987                 break
2988             result_url = compat_urlparse.urljoin('https://www.youtube.com/', next_link)
2989
2990         if len(videos) > n:
2991             videos = videos[:n]
2992         return self.playlist_result(videos, query)
2993
2994
2995 class YoutubeSearchDateIE(YoutubeSearchIE):
2996     IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
2997     _SEARCH_KEY = 'ytsearchdate'
2998     IE_DESC = 'YouTube.com searches, newest videos first'
2999     _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'}
3000
3001
3002 class YoutubeSearchURLIE(YoutubeSearchBaseInfoExtractor):
3003     IE_DESC = 'YouTube.com search URLs'
3004     IE_NAME = 'youtube:search_url'
3005     _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?(?:search_query|q)=(?P<query>[^&]+)(?:[&]|$)'
3006     _TESTS = [{
3007         'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
3008         'playlist_mincount': 5,
3009         'info_dict': {
3010             'title': 'youtube-dl test video',
3011         }
3012     }, {
3013         'url': 'https://www.youtube.com/results?q=test&sp=EgQIBBgB',
3014         'only_matching': True,
3015     }]
3016
3017     def _real_extract(self, url):
3018         mobj = re.match(self._VALID_URL, url)
3019         query = compat_urllib_parse_unquote_plus(mobj.group('query'))
3020         webpage = self._download_webpage(url, query)
3021         return self.playlist_result(self._process_page(webpage), playlist_title=query)
3022
3023
3024 class YoutubeShowIE(YoutubePlaylistsBaseInfoExtractor):
3025     IE_DESC = 'YouTube.com (multi-season) shows'
3026     _VALID_URL = r'https?://(?:www\.)?youtube\.com/show/(?P<id>[^?#]*)'
3027     IE_NAME = 'youtube:show'
3028     _TESTS = [{
3029         'url': 'https://www.youtube.com/show/airdisasters',
3030         'playlist_mincount': 5,
3031         'info_dict': {
3032             'id': 'airdisasters',
3033             'title': 'Air Disasters',
3034         }
3035     }]
3036
3037     def _real_extract(self, url):
3038         playlist_id = self._match_id(url)
3039         return super(YoutubeShowIE, self)._real_extract(
3040             'https://www.youtube.com/show/%s/playlists' % playlist_id)
3041
3042
3043 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
3044     """
3045     Base class for feed extractors
3046     Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
3047     """
3048     _LOGIN_REQUIRED = True
3049
3050     @property
3051     def IE_NAME(self):
3052         return 'youtube:%s' % self._FEED_NAME
3053
3054     def _real_initialize(self):
3055         self._login()
3056
3057     def _entries(self, page):
3058         # The extraction process is the same as for playlists, but the regex
3059         # for the video ids doesn't contain an index
3060         ids = []
3061         more_widget_html = content_html = page
3062         for page_num in itertools.count(1):
3063             matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
3064
3065             # 'recommended' feed has infinite 'load more' and each new portion spins
3066             # the same videos in (sometimes) slightly different order, so we'll check
3067             # for unicity and break when portion has no new videos
3068             new_ids = list(filter(lambda video_id: video_id not in ids, orderedSet(matches)))
3069             if not new_ids:
3070                 break
3071
3072             ids.extend(new_ids)
3073
3074             for entry in self._ids_to_results(new_ids):
3075                 yield entry
3076
3077             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
3078             if not mobj:
3079                 break
3080
3081             more = self._download_json(
3082                 'https://youtube.com/%s' % mobj.group('more'), self._PLAYLIST_TITLE,
3083                 'Downloading page #%s' % page_num,
3084                 transform_source=uppercase_escape)
3085             content_html = more['content_html']
3086             more_widget_html = more['load_more_widget_html']
3087
3088     def _real_extract(self, url):
3089         page = self._download_webpage(
3090             'https://www.youtube.com/feed/%s' % self._FEED_NAME,
3091             self._PLAYLIST_TITLE)
3092         return self.playlist_result(
3093             self._entries(page), playlist_title=self._PLAYLIST_TITLE)
3094
3095
3096 class YoutubeWatchLaterIE(YoutubePlaylistIE):
3097     IE_NAME = 'youtube:watchlater'
3098     IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)'
3099     _VALID_URL = r'https?://(?:www\.)?youtube\.com/(?:feed/watch_later|(?:playlist|watch)\?(?:.+&)?list=WL)|:ytwatchlater'
3100
3101     _TESTS = [{
3102         'url': 'https://www.youtube.com/playlist?list=WL',
3103         'only_matching': True,
3104     }, {
3105         'url': 'https://www.youtube.com/watch?v=bCNU9TrbiRk&index=1&list=WL',
3106         'only_matching': True,
3107     }]
3108
3109     def _real_extract(self, url):
3110         _, video = self._check_download_just_video(url, 'WL')
3111         if video:
3112             return video
3113         _, playlist = self._extract_playlist('WL')
3114         return playlist
3115
3116
3117 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
3118     IE_NAME = 'youtube:favorites'
3119     IE_DESC = 'YouTube.com favourite videos, ":ytfav" for short (requires authentication)'
3120     _VALID_URL = r'https?://(?:www\.)?youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
3121     _LOGIN_REQUIRED = True
3122
3123     def _real_extract(self, url):
3124         webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
3125         playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
3126         return self.url_result(playlist_id, 'YoutubePlaylist')
3127
3128
3129 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
3130     IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)'
3131     _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/recommended|:ytrec(?:ommended)?'
3132     _FEED_NAME = 'recommended'
3133     _PLAYLIST_TITLE = 'Youtube Recommended videos'
3134
3135
3136 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
3137     IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
3138     _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
3139     _FEED_NAME = 'subscriptions'
3140     _PLAYLIST_TITLE = 'Youtube Subscriptions'
3141
3142
3143 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
3144     IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)'
3145     _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/history|:ythistory'
3146     _FEED_NAME = 'history'
3147     _PLAYLIST_TITLE = 'Youtube History'
3148
3149
3150 class YoutubeTruncatedURLIE(InfoExtractor):
3151     IE_NAME = 'youtube:truncated_url'
3152     IE_DESC = False  # Do not list
3153     _VALID_URL = r'''(?x)
3154         (?:https?://)?
3155         (?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/
3156         (?:watch\?(?:
3157             feature=[a-z_]+|
3158             annotation_id=annotation_[^&]+|
3159             x-yt-cl=[0-9]+|
3160             hl=[^&]*|
3161             t=[0-9]+
3162         )?
3163         |
3164             attribution_link\?a=[^&]+
3165         )
3166         $
3167     '''
3168
3169     _TESTS = [{
3170         'url': 'https://www.youtube.com/watch?annotation_id=annotation_3951667041',
3171         'only_matching': True,
3172     }, {
3173         'url': 'https://www.youtube.com/watch?',
3174         'only_matching': True,
3175     }, {
3176         'url': 'https://www.youtube.com/watch?x-yt-cl=84503534',
3177         'only_matching': True,
3178     }, {
3179         'url': 'https://www.youtube.com/watch?feature=foo',
3180         'only_matching': True,
3181     }, {
3182         'url': 'https://www.youtube.com/watch?hl=en-GB',
3183         'only_matching': True,
3184     }, {
3185         'url': 'https://www.youtube.com/watch?t=2372',
3186         'only_matching': True,
3187     }]
3188
3189     def _real_extract(self, url):
3190         raise ExtractorError(
3191             'Did you forget to quote the URL? Remember that & is a meta '
3192             'character in most shells, so you want to put the URL in quotes, '
3193             'like  youtube-dl '
3194             '"https://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
3195             ' or simply  youtube-dl BaW_jenozKc  .',
3196             expected=True)
3197
3198
3199 class YoutubeTruncatedIDIE(InfoExtractor):
3200     IE_NAME = 'youtube:truncated_id'
3201     IE_DESC = False  # Do not list
3202     _VALID_URL = r'https?://(?:www\.)?youtube\.com/watch\?v=(?P<id>[0-9A-Za-z_-]{1,10})$'
3203
3204     _TESTS = [{
3205         'url': 'https://www.youtube.com/watch?v=N_708QY7Ob',
3206         'only_matching': True,
3207     }]
3208
3209     def _real_extract(self, url):
3210         video_id = self._match_id(url)
3211         raise ExtractorError(
3212             'Incomplete YouTube ID %s. URL %s looks truncated.' % (video_id, url),
3213             expected=True)