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

   1 # coding: utf-8
   2
   3 import json
   4 import netrc
   5 import re
   6 import socket
   7 import itertools
   8
   9 from .common import InfoExtractor, SearchInfoExtractor
  10 from ..utils import (
  11     compat_http_client,
  12     compat_parse_qs,
  13     compat_urllib_error,
  14     compat_urllib_parse,
  15     compat_urllib_request,
  16     compat_str,
  17
  18     clean_html,
  19     get_element_by_id,
  20     ExtractorError,
  21     unescapeHTML,
  22     unified_strdate,
  23     orderedSet,
  24 )
  25
  26 class YoutubeBaseInfoExtractor(InfoExtractor):
  27     """Provide base functions for Youtube extractors"""
  28     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
  29     _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
  30     _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
  31     _NETRC_MACHINE = 'youtube'
  32     # If True it will raise an error if no login info is provided
  33     _LOGIN_REQUIRED = False
  34
  35     def report_lang(self):
  36         """Report attempt to set language."""
  37         self.to_screen(u'Setting language')
  38
  39     def _set_language(self):
  40         request = compat_urllib_request.Request(self._LANG_URL)
  41         try:
  42             self.report_lang()
  43             compat_urllib_request.urlopen(request).read()
  44         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
  45             self._downloader.report_warning(u'unable to set language: %s' % compat_str(err))
  46             return False
  47         return True
  48
  49     def _login(self):
  50         (username, password) = self._get_login_info()
  51         # No authentication to be performed
  52         if username is None:
  53             if self._LOGIN_REQUIRED:
  54                 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
  55             return False
  56
  57         request = compat_urllib_request.Request(self._LOGIN_URL)
  58         try:
  59             login_page = compat_urllib_request.urlopen(request).read().decode('utf-8')
  60         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
  61             self._downloader.report_warning(u'unable to fetch login page: %s' % compat_str(err))
  62             return False
  63
  64         galx = None
  65         dsh = None
  66         match = re.search(re.compile(r'<input.+?name="GALX".+?value="(.+?)"', re.DOTALL), login_page)
  67         if match:
  68           galx = match.group(1)
  69         match = re.search(re.compile(r'<input.+?name="dsh".+?value="(.+?)"', re.DOTALL), login_page)
  70         if match:
  71           dsh = match.group(1)
  72
  73         # Log in
  74         login_form_strs = {
  75                 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
  76                 u'Email': username,
  77                 u'GALX': galx,
  78                 u'Passwd': password,
  79                 u'PersistentCookie': u'yes',
  80                 u'_utf8': u'霱',
  81                 u'bgresponse': u'js_disabled',
  82                 u'checkConnection': u'',
  83                 u'checkedDomains': u'youtube',
  84                 u'dnConn': u'',
  85                 u'dsh': dsh,
  86                 u'pstMsg': u'0',
  87                 u'rmShown': u'1',
  88                 u'secTok': u'',
  89                 u'signIn': u'Sign in',
  90                 u'timeStmp': u'',
  91                 u'service': u'youtube',
  92                 u'uilel': u'3',
  93                 u'hl': u'en_US',
  94         }
  95         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
  96         # chokes on unicode
  97         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
  98         login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
  99         request = compat_urllib_request.Request(self._LOGIN_URL, login_data)
 100         try:
 101             self.report_login()
 102             login_results = compat_urllib_request.urlopen(request).read().decode('utf-8')
 103             if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
 104                 self._downloader.report_warning(u'unable to log in: bad username or password')
 105                 return False
 106         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 107             self._downloader.report_warning(u'unable to log in: %s' % compat_str(err))
 108             return False
 109         return True
 110
 111     def _confirm_age(self):
 112         age_form = {
 113                 'next_url':     '/',
 114                 'action_confirm':   'Confirm',
 115                 }
 116         request = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
 117         try:
 118             self.report_age_confirmation()
 119             compat_urllib_request.urlopen(request).read().decode('utf-8')
 120         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 121             raise ExtractorError(u'Unable to confirm age: %s' % compat_str(err))
 122         return True
 123
 124     def _real_initialize(self):
 125         if self._downloader is None:
 126             return
 127         if not self._set_language():
 128             return
 129         if not self._login():
 130             return
 131         self._confirm_age()
 132
 133 class YoutubeIE(YoutubeBaseInfoExtractor):
 134     IE_DESC = u'YouTube.com'
 135     _VALID_URL = r"""^
 136                      (
 137                          (?:https?://)?                                       # http(s):// (optional)
 138                          (?:youtu\.be/|(?:\w+\.)?youtube(?:-nocookie)?\.com/|
 139                             tube\.majestyc\.net/)                             # the various hostnames, with wildcard subdomains
 140                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
 141                          (?:                                                  # the various things that can precede the ID:
 142                              (?:(?:v|embed|e)/)                               # v/ or embed/ or e/
 143                              |(?:                                             # or the v= param in all its forms
 144                                  (?:watch|movie(?:_popup)?(?:\.php)?)?              # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
 145                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
 146                                  (?:.*?&)?                                    # any other preceding param (like /?s=tuff&v=xxxx)
 147                                  v=
 148                              )
 149                          )?                                                   # optional -> youtube.com/xxxx is OK
 150                      )?                                                       # all until now is optional -> you can pass the naked ID
 151                      ([0-9A-Za-z_-]+)                                         # here is it! the YouTube video ID
 152                      (?(1).+)?                                                # if we found the ID, everything can follow
 153                      $"""
 154     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
 155     # Listed in order of quality
 156     _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '17', '13',
 157                           '95', '94', '93', '92', '132', '151',
 158                           ]
 159     _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '17', '13',
 160                                       '95', '94', '93', '92', '132', '151',
 161                                       ]
 162     _video_extensions = {
 163         '13': '3gp',
 164         '17': 'mp4',
 165         '18': 'mp4',
 166         '22': 'mp4',
 167         '37': 'mp4',
 168         '38': 'mp4',
 169         '43': 'webm',
 170         '44': 'webm',
 171         '45': 'webm',
 172         '46': 'webm',
 173
 174         # videos that use m3u8
 175         '92': 'mp4',
 176         '93': 'mp4',
 177         '94': 'mp4',
 178         '95': 'mp4',
 179         '96': 'mp4',
 180         '132': 'mp4',
 181         '151': 'mp4',
 182     }
 183     _video_dimensions = {
 184         '5': '240x400',
 185         '6': '???',
 186         '13': '???',
 187         '17': '144x176',
 188         '18': '360x640',
 189         '22': '720x1280',
 190         '34': '360x640',
 191         '35': '480x854',
 192         '37': '1080x1920',
 193         '38': '3072x4096',
 194         '43': '360x640',
 195         '44': '480x854',
 196         '45': '720x1280',
 197         '46': '1080x1920',
 198         '92': '240p',
 199         '93': '360p',
 200         '94': '480p',
 201         '95': '720p',
 202         '96': '1080p',
 203         '132': '240p',
 204         '151': '72p',
 205     }
 206     IE_NAME = u'youtube'
 207     _TESTS = [
 208         {
 209             u"url":  u"http://www.youtube.com/watch?v=BaW_jenozKc",
 210             u"file":  u"BaW_jenozKc.mp4",
 211             u"info_dict": {
 212                 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
 213                 u"uploader": u"Philipp Hagemeister",
 214                 u"uploader_id": u"phihag",
 215                 u"upload_date": u"20121002",
 216                 u"description": u"test chars:  \"'/\\ä↭𝕐\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
 217             }
 218         },
 219         {
 220             u"url":  u"http://www.youtube.com/watch?v=1ltcDfZMA3U",
 221             u"file":  u"1ltcDfZMA3U.flv",
 222             u"note": u"Test VEVO video (#897)",
 223             u"info_dict": {
 224                 u"upload_date": u"20070518",
 225                 u"title": u"Maps - It Will Find You",
 226                 u"description": u"Music video by Maps performing It Will Find You.",
 227                 u"uploader": u"MuteUSA",
 228                 u"uploader_id": u"MuteUSA"
 229             }
 230         },
 231         {
 232             u"url":  u"http://www.youtube.com/watch?v=UxxajLWwzqY",
 233             u"file":  u"UxxajLWwzqY.mp4",
 234             u"note": u"Test generic use_cipher_signature video (#897)",
 235             u"info_dict": {
 236                 u"upload_date": u"20120506",
 237                 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
 238                 u"description": u"md5:b085c9804f5ab69f4adea963a2dceb3c",
 239                 u"uploader": u"IconaPop",
 240                 u"uploader_id": u"IconaPop"
 241             }
 242         },
 243         {
 244             u"url":  u"https://www.youtube.com/watch?v=07FYdnEawAQ",
 245             u"file":  u"07FYdnEawAQ.mp4",
 246             u"note": u"Test VEVO video with age protection (#956)",
 247             u"info_dict": {
 248                 u"upload_date": u"20130703",
 249                 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
 250                 u"description": u"md5:64249768eec3bc4276236606ea996373",
 251                 u"uploader": u"justintimberlakeVEVO",
 252                 u"uploader_id": u"justintimberlakeVEVO"
 253             }
 254         },
 255         {
 256             u'url': u'https://www.youtube.com/watch?v=TGi3HqYrWHE',
 257             u'file': u'TGi3HqYrWHE.mp4',
 258             u'note': u'm3u8 video',
 259             u'info_dict': {
 260                 u'title': u'Triathlon - Men - London 2012 Olympic Games',
 261                 u'description': u'- Men -  TR02 - Triathlon - 07 August 2012 - London 2012 Olympic Games',
 262                 u'uploader': u'olympic',
 263                 u'upload_date': u'20120807',
 264                 u'uploader_id': u'olympic',
 265             },
 266             u'params': {
 267                 u'skip_download': True,
 268             },
 269         },
 270     ]
 271
 272
 273     @classmethod
 274     def suitable(cls, url):
 275         """Receives a URL and returns True if suitable for this IE."""
 276         if YoutubePlaylistIE.suitable(url) or YoutubeSubscriptionsIE.suitable(url): return False
 277         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
 278
 279     def report_video_webpage_download(self, video_id):
 280         """Report attempt to download video webpage."""
 281         self.to_screen(u'%s: Downloading video webpage' % video_id)
 282
 283     def report_video_info_webpage_download(self, video_id):
 284         """Report attempt to download video info webpage."""
 285         self.to_screen(u'%s: Downloading video info webpage' % video_id)
 286
 287     def report_video_subtitles_download(self, video_id):
 288         """Report attempt to download video info webpage."""
 289         self.to_screen(u'%s: Checking available subtitles' % video_id)
 290
 291     def report_video_subtitles_request(self, video_id, sub_lang, format):
 292         """Report attempt to download video info webpage."""
 293         self.to_screen(u'%s: Downloading video subtitles for %s.%s' % (video_id, sub_lang, format))
 294
 295     def report_video_subtitles_available(self, video_id, sub_lang_list):
 296         """Report available subtitles."""
 297         sub_lang = ",".join(list(sub_lang_list.keys()))
 298         self.to_screen(u'%s: Available subtitles for video: %s' % (video_id, sub_lang))
 299
 300     def report_information_extraction(self, video_id):
 301         """Report attempt to extract video information."""
 302         self.to_screen(u'%s: Extracting video information' % video_id)
 303
 304     def report_unavailable_format(self, video_id, format):
 305         """Report extracted video URL."""
 306         self.to_screen(u'%s: Format %s not available' % (video_id, format))
 307
 308     def report_rtmp_download(self):
 309         """Indicate the download will use the RTMP protocol."""
 310         self.to_screen(u'RTMP download detected')
 311
 312     def _decrypt_signature(self, s):
 313         """Turn the encrypted s field into a working signature"""
 314
 315         if len(s) == 92:
 316             return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
 317         elif len(s) == 90:
 318             return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
 319         elif len(s) == 88:
 320             return s[48] + s[81:67:-1] + s[82] + s[66:62:-1] + s[85] + s[61:48:-1] + s[67] + s[47:12:-1] + s[3] + s[11:3:-1] + s[2] + s[12]
 321         elif len(s) == 87:
 322             return s[4:23] + s[86] + s[24:85]
 323         elif len(s) == 86:
 324             return s[83:85] + s[26] + s[79:46:-1] + s[85] + s[45:36:-1] + s[30] + s[35:30:-1] + s[46] + s[29:26:-1] + s[82] + s[25:1:-1]
 325         elif len(s) == 85:
 326             return s[2:8] + s[0] + s[9:21] + s[65] + s[22:65] + s[84] + s[66:82] + s[21]
 327         elif len(s) == 84:
 328             return s[83:36:-1] + s[2] + s[35:26:-1] + s[3] + s[25:3:-1] + s[26]
 329         elif len(s) == 83:
 330             return s[6] + s[3:6] + s[33] + s[7:24] + s[0] + s[25:33] + s[53] + s[34:53] + s[24] + s[54:]
 331         elif len(s) == 82:
 332             return s[36] + s[79:67:-1] + s[81] + s[66:40:-1] + s[33] + s[39:36:-1] + s[40] + s[35] + s[0] + s[67] + s[32:0:-1] + s[34]
 333         elif len(s) == 81:
 334             return s[56] + s[79:56:-1] + s[41] + s[55:41:-1] + s[80] + s[40:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
 335         elif len(s) == 79:
 336             return s[54] + s[77:54:-1] + s[39] + s[53:39:-1] + s[78] + s[38:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
 337
 338         else:
 339             raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
 340
 341     def _decrypt_signature_age_gate(self, s):
 342         # The videos with age protection use another player, so the algorithms
 343         # can be different.
 344         if len(s) == 86:
 345             return s[2:63] + s[82] + s[64:82] + s[63]
 346         else:
 347             # Fallback to the other algortihms
 348             return self._decrypt_signature(s)
 349
 350
 351     def _get_available_subtitles(self, video_id):
 352         self.report_video_subtitles_download(video_id)
 353         request = compat_urllib_request.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
 354         try:
 355             sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8')
 356         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 357             return (u'unable to download video subtitles: %s' % compat_str(err), None)
 358         sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
 359         sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list)
 360         if not sub_lang_list:
 361             return (u'video doesn\'t have subtitles', None)
 362         return sub_lang_list
 363
 364     def _list_available_subtitles(self, video_id):
 365         sub_lang_list = self._get_available_subtitles(video_id)
 366         self.report_video_subtitles_available(video_id, sub_lang_list)
 367
 368     def _request_subtitle(self, sub_lang, sub_name, video_id, format):
 369         """
 370         Return tuple:
 371         (error_message, sub_lang, sub)
 372         """
 373         self.report_video_subtitles_request(video_id, sub_lang, format)
 374         params = compat_urllib_parse.urlencode({
 375             'lang': sub_lang,
 376             'name': sub_name,
 377             'v': video_id,
 378             'fmt': format,
 379         })
 380         url = 'http://www.youtube.com/api/timedtext?' + params
 381         try:
 382             sub = compat_urllib_request.urlopen(url).read().decode('utf-8')
 383         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 384             return (u'unable to download video subtitles: %s' % compat_str(err), None, None)
 385         if not sub:
 386             return (u'Did not fetch video subtitles', None, None)
 387         return (None, sub_lang, sub)
 388
 389     def _request_automatic_caption(self, video_id, webpage):
 390         """We need the webpage for getting the captions url, pass it as an
 391            argument to speed up the process."""
 392         sub_lang = self._downloader.params.get('subtitleslang') or 'en'
 393         sub_format = self._downloader.params.get('subtitlesformat')
 394         self.to_screen(u'%s: Looking for automatic captions' % video_id)
 395         mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
 396         err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
 397         if mobj is None:
 398             return [(err_msg, None, None)]
 399         player_config = json.loads(mobj.group(1))
 400         try:
 401             args = player_config[u'args']
 402             caption_url = args[u'ttsurl']
 403             timestamp = args[u'timestamp']
 404             params = compat_urllib_parse.urlencode({
 405                 'lang': 'en',
 406                 'tlang': sub_lang,
 407                 'fmt': sub_format,
 408                 'ts': timestamp,
 409                 'kind': 'asr',
 410             })
 411             subtitles_url = caption_url + '&' + params
 412             sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
 413             return [(None, sub_lang, sub)]
 414         except KeyError:
 415             return [(err_msg, None, None)]
 416
 417     def _extract_subtitle(self, video_id):
 418         """
 419         Return a list with a tuple:
 420         [(error_message, sub_lang, sub)]
 421         """
 422         sub_lang_list = self._get_available_subtitles(video_id)
 423         sub_format = self._downloader.params.get('subtitlesformat')
 424         if  isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
 425             return [(sub_lang_list[0], None, None)]
 426         if self._downloader.params.get('subtitleslang', False):
 427             sub_lang = self._downloader.params.get('subtitleslang')
 428         elif 'en' in sub_lang_list:
 429             sub_lang = 'en'
 430         else:
 431             sub_lang = list(sub_lang_list.keys())[0]
 432         if not sub_lang in sub_lang_list:
 433             return [(u'no closed captions found in the specified language "%s"' % sub_lang, None, None)]
 434
 435         subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
 436         return [subtitle]
 437
 438     def _extract_all_subtitles(self, video_id):
 439         sub_lang_list = self._get_available_subtitles(video_id)
 440         sub_format = self._downloader.params.get('subtitlesformat')
 441         if  isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
 442             return [(sub_lang_list[0], None, None)]
 443         subtitles = []
 444         for sub_lang in sub_lang_list:
 445             subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
 446             subtitles.append(subtitle)
 447         return subtitles
 448
 449     def _print_formats(self, formats):
 450         print('Available formats:')
 451         for x in formats:
 452             print('%s\t:\t%s\t[%s]' %(x, self._video_extensions.get(x, 'flv'), self._video_dimensions.get(x, '???')))
 453
 454     def _extract_id(self, url):
 455         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 456         if mobj is None:
 457             raise ExtractorError(u'Invalid URL: %s' % url)
 458         video_id = mobj.group(2)
 459         return video_id
 460
 461     def _get_video_url_list(self, url_map):
 462         """
 463         Transform a dictionary in the format {itag:url} to a list of (itag, url)
 464         with the requested formats.
 465         """
 466         req_format = self._downloader.params.get('format', None)
 467         format_limit = self._downloader.params.get('format_limit', None)
 468         available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats
 469         if format_limit is not None and format_limit in available_formats:
 470             format_list = available_formats[available_formats.index(format_limit):]
 471         else:
 472             format_list = available_formats
 473         existing_formats = [x for x in format_list if x in url_map]
 474         if len(existing_formats) == 0:
 475             raise ExtractorError(u'no known formats available for video')
 476         if self._downloader.params.get('listformats', None):
 477             self._print_formats(existing_formats)
 478             return
 479         if req_format is None or req_format == 'best':
 480             video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality
 481         elif req_format == 'worst':
 482             video_url_list = [(existing_formats[-1], url_map[existing_formats[-1]])] # worst quality
 483         elif req_format in ('-1', 'all'):
 484             video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
 485         else:
 486             # Specific formats. We pick the first in a slash-delimeted sequence.
 487             # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'.
 488             req_formats = req_format.split('/')
 489             video_url_list = None
 490             for rf in req_formats:
 491                 if rf in url_map:
 492                     video_url_list = [(rf, url_map[rf])]
 493                     break
 494             if video_url_list is None:
 495                 raise ExtractorError(u'requested format not available')
 496         return video_url_list
 497
 498     def _extract_from_m3u8(self, manifest_url, video_id):
 499         url_map = {}
 500         def _get_urls(_manifest):
 501             lines = _manifest.split('\n')
 502             urls = filter(lambda l: l and not l.startswith('#'),
 503                             lines)
 504             return urls
 505         manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
 506         formats_urls = _get_urls(manifest)
 507         for format_url in formats_urls:
 508             itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
 509             url_map[itag] = format_url
 510         return url_map
 511
 512     def _real_extract(self, url):
 513         if re.match(r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$', url):
 514             self._downloader.report_warning(u'Did you forget to quote the URL? Remember that & is a meta-character in most shells, so you want to put the URL in quotes, like  youtube-dl \'http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc\' (or simply  youtube-dl BaW_jenozKc  ).')
 515
 516         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
 517         mobj = re.search(self._NEXT_URL_RE, url)
 518         if mobj:
 519             url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
 520         video_id = self._extract_id(url)
 521
 522         # Get video webpage
 523         self.report_video_webpage_download(video_id)
 524         url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
 525         request = compat_urllib_request.Request(url)
 526         try:
 527             video_webpage_bytes = compat_urllib_request.urlopen(request).read()
 528         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 529             raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err))
 530
 531         video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
 532
 533         # Attempt to extract SWF player URL
 534         mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
 535         if mobj is not None:
 536             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
 537         else:
 538             player_url = None
 539
 540         # Get video info
 541         self.report_video_info_webpage_download(video_id)
 542         if re.search(r'player-age-gate-content">', video_webpage) is not None:
 543             self.report_age_confirmation()
 544             age_gate = True
 545             # We simulate the access to the video from www.youtube.com/v/{video_id}
 546             # this can be viewed without login into Youtube
 547             data = compat_urllib_parse.urlencode({'video_id': video_id,
 548                                                   'el': 'embedded',
 549                                                   'gl': 'US',
 550                                                   'hl': 'en',
 551                                                   'eurl': 'https://youtube.googleapis.com/v/' + video_id,
 552                                                   'asv': 3,
 553                                                   'sts':'1588',
 554                                                   })
 555             video_info_url = 'https://www.youtube.com/get_video_info?' + data
 556             video_info_webpage = self._download_webpage(video_info_url, video_id,
 557                                     note=False,
 558                                     errnote='unable to download video info webpage')
 559             video_info = compat_parse_qs(video_info_webpage)
 560         else:
 561             age_gate = False
 562             for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
 563                 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
 564                         % (video_id, el_type))
 565                 video_info_webpage = self._download_webpage(video_info_url, video_id,
 566                                         note=False,
 567                                         errnote='unable to download video info webpage')
 568                 video_info = compat_parse_qs(video_info_webpage)
 569                 if 'token' in video_info:
 570                     break
 571         if 'token' not in video_info:
 572             if 'reason' in video_info:
 573                 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
 574             else:
 575                 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
 576
 577         # Check for "rental" videos
 578         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
 579             raise ExtractorError(u'"rental" videos not supported')
 580
 581         # Start extracting information
 582         self.report_information_extraction(video_id)
 583
 584         # uploader
 585         if 'author' not in video_info:
 586             raise ExtractorError(u'Unable to extract uploader name')
 587         video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
 588
 589         # uploader_id
 590         video_uploader_id = None
 591         mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
 592         if mobj is not None:
 593             video_uploader_id = mobj.group(1)
 594         else:
 595             self._downloader.report_warning(u'unable to extract uploader nickname')
 596
 597         # title
 598         if 'title' not in video_info:
 599             raise ExtractorError(u'Unable to extract video title')
 600         video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
 601
 602         # thumbnail image
 603         # We try first to get a high quality image:
 604         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
 605                             video_webpage, re.DOTALL)
 606         if m_thumb is not None:
 607             video_thumbnail = m_thumb.group(1)
 608         elif 'thumbnail_url' not in video_info:
 609             self._downloader.report_warning(u'unable to extract video thumbnail')
 610             video_thumbnail = ''
 611         else:   # don't panic if we can't find it
 612             video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
 613
 614         # upload date
 615         upload_date = None
 616         mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
 617         if mobj is not None:
 618             upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
 619             upload_date = unified_strdate(upload_date)
 620
 621         # description
 622         video_description = get_element_by_id("eow-description", video_webpage)
 623         if video_description:
 624             video_description = clean_html(video_description)
 625         else:
 626             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
 627             if fd_mobj:
 628                 video_description = unescapeHTML(fd_mobj.group(1))
 629             else:
 630                 video_description = u''
 631
 632         # subtitles
 633         video_subtitles = None
 634
 635         if self._downloader.params.get('writesubtitles', False):
 636             video_subtitles = self._extract_subtitle(video_id)
 637             if video_subtitles:
 638                 (sub_error, sub_lang, sub) = video_subtitles[0]
 639                 if sub_error:
 640                     self._downloader.report_warning(sub_error)
 641
 642         if self._downloader.params.get('writeautomaticsub', False):
 643             video_subtitles = self._request_automatic_caption(video_id, video_webpage)
 644             (sub_error, sub_lang, sub) = video_subtitles[0]
 645             if sub_error:
 646                 self._downloader.report_warning(sub_error)
 647
 648         if self._downloader.params.get('allsubtitles', False):
 649             video_subtitles = self._extract_all_subtitles(video_id)
 650             for video_subtitle in video_subtitles:
 651                 (sub_error, sub_lang, sub) = video_subtitle
 652                 if sub_error:
 653                     self._downloader.report_warning(sub_error)
 654
 655         if self._downloader.params.get('listsubtitles', False):
 656             self._list_available_subtitles(video_id)
 657             return
 658
 659         if 'length_seconds' not in video_info:
 660             self._downloader.report_warning(u'unable to extract video duration')
 661             video_duration = ''
 662         else:
 663             video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
 664
 665         # Decide which formats to download
 666
 667         try:
 668             mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
 669             if not mobj:
 670                 raise ValueError('Could not find vevo ID')
 671             info = json.loads(mobj.group(1))
 672             args = info['args']
 673             # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
 674             # this signatures are encrypted
 675             m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
 676             if m_s is not None:
 677                 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
 678                 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
 679         except ValueError:
 680             pass
 681
 682         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
 683             self.report_rtmp_download()
 684             video_url_list = [(None, video_info['conn'][0])]
 685         elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1:
 686             if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]:
 687                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
 688             url_map = {}
 689             for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','):
 690                 url_data = compat_parse_qs(url_data_str)
 691                 if 'itag' in url_data and 'url' in url_data:
 692                     url = url_data['url'][0]
 693                     if 'sig' in url_data:
 694                         url += '&signature=' + url_data['sig'][0]
 695                     elif 's' in url_data:
 696                         if self._downloader.params.get('verbose'):
 697                             s = url_data['s'][0]
 698                             if age_gate:
 699                                 player_version = self._search_regex(r'ad3-(.+?)\.swf',
 700                                     video_info['ad3_module'][0], 'flash player',
 701                                     fatal=False)
 702                                 player = 'flash player %s' % player_version
 703                             else:
 704                                 player = u'html5 player %s' % self._search_regex(r'html5player-(.+?)\.js', video_webpage,
 705                                     'html5 player', fatal=False)
 706                             parts_sizes = u'.'.join(compat_str(len(part)) for part in s.split('.'))
 707                             self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
 708                                 (len(s), parts_sizes, url_data['itag'][0], player))
 709                         encrypted_sig = url_data['s'][0]
 710                         if age_gate:
 711                             signature = self._decrypt_signature_age_gate(encrypted_sig)
 712                         else:
 713                             signature = self._decrypt_signature(encrypted_sig)
 714                         url += '&signature=' + signature
 715                     if 'ratebypass' not in url:
 716                         url += '&ratebypass=yes'
 717                     url_map[url_data['itag'][0]] = url
 718             video_url_list = self._get_video_url_list(url_map)
 719             if not video_url_list:
 720                 return
 721         elif video_info.get('hlsvp'):
 722             manifest_url = video_info['hlsvp'][0]
 723             url_map = self._extract_from_m3u8(manifest_url, video_id)
 724             video_url_list = self._get_video_url_list(url_map)
 725             if not video_url_list:
 726                 return
 727
 728         else:
 729             raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info')
 730
 731         results = []
 732         for format_param, video_real_url in video_url_list:
 733             # Extension
 734             video_extension = self._video_extensions.get(format_param, 'flv')
 735
 736             video_format = '{0} - {1}'.format(format_param if format_param else video_extension,
 737                                               self._video_dimensions.get(format_param, '???'))
 738
 739             results.append({
 740                 'id':       video_id,
 741                 'url':      video_real_url,
 742                 'uploader': video_uploader,
 743                 'uploader_id': video_uploader_id,
 744                 'upload_date':  upload_date,
 745                 'title':    video_title,
 746                 'ext':      video_extension,
 747                 'format':   video_format,
 748                 'thumbnail':    video_thumbnail,
 749                 'description':  video_description,
 750                 'player_url':   player_url,
 751                 'subtitles':    video_subtitles,
 752                 'duration':     video_duration
 753             })
 754         return results
 755
 756 class YoutubePlaylistIE(InfoExtractor):
 757     IE_DESC = u'YouTube.com playlists'
 758     _VALID_URL = r"""(?:
 759                         (?:https?://)?
 760                         (?:\w+\.)?
 761                         youtube\.com/
 762                         (?:
 763                            (?:course|view_play_list|my_playlists|artist|playlist|watch)
 764                            \? (?:.*?&)*? (?:p|a|list)=
 765                         |  p/
 766                         )
 767                         ((?:PL|EC|UU|FL)?[0-9A-Za-z-_]{10,})
 768                         .*
 769                      |
 770                         ((?:PL|EC|UU|FL)[0-9A-Za-z-_]{10,})
 771                      )"""
 772     _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none'
 773     _MAX_RESULTS = 50
 774     IE_NAME = u'youtube:playlist'
 775
 776     @classmethod
 777     def suitable(cls, url):
 778         """Receives a URL and returns True if suitable for this IE."""
 779         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
 780
 781     def _real_extract(self, url):
 782         # Extract playlist id
 783         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 784         if mobj is None:
 785             raise ExtractorError(u'Invalid URL: %s' % url)
 786
 787         # Download playlist videos from API
 788         playlist_id = mobj.group(1) or mobj.group(2)
 789         videos = []
 790
 791         for page_num in itertools.count(1):
 792             start_index = self._MAX_RESULTS * (page_num - 1) + 1
 793             if start_index >= 1000:
 794                 self._downloader.report_warning(u'Max number of results reached')
 795                 break
 796             url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, start_index)
 797             page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
 798
 799             try:
 800                 response = json.loads(page)
 801             except ValueError as err:
 802                 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
 803
 804             if 'feed' not in response:
 805                 raise ExtractorError(u'Got a malformed response from YouTube API')
 806             playlist_title = response['feed']['title']['$t']
 807             if 'entry' not in response['feed']:
 808                 # Number of videos is a multiple of self._MAX_RESULTS
 809                 break
 810
 811             for entry in response['feed']['entry']:
 812                 index = entry['yt$position']['$t']
 813                 if 'media$group' in entry and 'media$player' in entry['media$group']:
 814                     videos.append((index, entry['media$group']['media$player']['url']))
 815
 816         videos = [v[1] for v in sorted(videos)]
 817
 818         url_results = [self.url_result(vurl, 'Youtube') for vurl in videos]
 819         return [self.playlist_result(url_results, playlist_id, playlist_title)]
 820
 821
 822 class YoutubeChannelIE(InfoExtractor):
 823     IE_DESC = u'YouTube.com channels'
 824     _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
 825     _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en'
 826     _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
 827     _MORE_PAGES_URL = 'http://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
 828     IE_NAME = u'youtube:channel'
 829
 830     def extract_videos_from_page(self, page):
 831         ids_in_page = []
 832         for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
 833             if mobj.group(1) not in ids_in_page:
 834                 ids_in_page.append(mobj.group(1))
 835         return ids_in_page
 836
 837     def _real_extract(self, url):
 838         # Extract channel id
 839         mobj = re.match(self._VALID_URL, url)
 840         if mobj is None:
 841             raise ExtractorError(u'Invalid URL: %s' % url)
 842
 843         # Download channel page
 844         channel_id = mobj.group(1)
 845         video_ids = []
 846         pagenum = 1
 847
 848         url = self._TEMPLATE_URL % (channel_id, pagenum)
 849         page = self._download_webpage(url, channel_id,
 850                                       u'Downloading page #%s' % pagenum)
 851
 852         # Extract video identifiers
 853         ids_in_page = self.extract_videos_from_page(page)
 854         video_ids.extend(ids_in_page)
 855
 856         # Download any subsequent channel pages using the json-based channel_ajax query
 857         if self._MORE_PAGES_INDICATOR in page:
 858             for pagenum in itertools.count(1):
 859                 url = self._MORE_PAGES_URL % (pagenum, channel_id)
 860                 page = self._download_webpage(url, channel_id,
 861                                               u'Downloading page #%s' % pagenum)
 862
 863                 page = json.loads(page)
 864
 865                 ids_in_page = self.extract_videos_from_page(page['content_html'])
 866                 video_ids.extend(ids_in_page)
 867
 868                 if self._MORE_PAGES_INDICATOR  not in page['load_more_widget_html']:
 869                     break
 870
 871         self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
 872
 873         urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids]
 874         url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls]
 875         return [self.playlist_result(url_entries, channel_id)]
 876
 877
 878 class YoutubeUserIE(InfoExtractor):
 879     IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
 880     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
 881     _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
 882     _GDATA_PAGE_SIZE = 50
 883     _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
 884     _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
 885     IE_NAME = u'youtube:user'
 886
 887     def _real_extract(self, url):
 888         # Extract username
 889         mobj = re.match(self._VALID_URL, url)
 890         if mobj is None:
 891             raise ExtractorError(u'Invalid URL: %s' % url)
 892
 893         username = mobj.group(1)
 894
 895         # Download video ids using YouTube Data API. Result size per
 896         # query is limited (currently to 50 videos) so we need to query
 897         # page by page until there are no video ids - it means we got
 898         # all of them.
 899
 900         video_ids = []
 901
 902         for pagenum in itertools.count(0):
 903             start_index = pagenum * self._GDATA_PAGE_SIZE + 1
 904
 905             gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
 906             page = self._download_webpage(gdata_url, username,
 907                                           u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
 908
 909             # Extract video identifiers
 910             ids_in_page = []
 911
 912             for mobj in re.finditer(self._VIDEO_INDICATOR, page):
 913                 if mobj.group(1) not in ids_in_page:
 914                     ids_in_page.append(mobj.group(1))
 915
 916             video_ids.extend(ids_in_page)
 917
 918             # A little optimization - if current page is not
 919             # "full", ie. does not contain PAGE_SIZE video ids then
 920             # we can assume that this page is the last one - there
 921             # are no more ids on further pages - no need to query
 922             # again.
 923
 924             if len(ids_in_page) < self._GDATA_PAGE_SIZE:
 925                 break
 926
 927         urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
 928         url_results = [self.url_result(rurl, 'Youtube') for rurl in urls]
 929         return [self.playlist_result(url_results, playlist_title = username)]
 930
 931 class YoutubeSearchIE(SearchInfoExtractor):
 932     IE_DESC = u'YouTube.com searches'
 933     _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
 934     _MAX_RESULTS = 1000
 935     IE_NAME = u'youtube:search'
 936     _SEARCH_KEY = 'ytsearch'
 937
 938     def report_download_page(self, query, pagenum):
 939         """Report attempt to download search page with given number."""
 940         self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
 941
 942     def _get_n_results(self, query, n):
 943         """Get a specified number of results for a query"""
 944
 945         video_ids = []
 946         pagenum = 0
 947         limit = n
 948
 949         while (50 * pagenum) < limit:
 950             self.report_download_page(query, pagenum+1)
 951             result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
 952             request = compat_urllib_request.Request(result_url)
 953             try:
 954                 data = compat_urllib_request.urlopen(request).read().decode('utf-8')
 955             except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 956                 raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
 957             api_response = json.loads(data)['data']
 958
 959             if not 'items' in api_response:
 960                 raise ExtractorError(u'[youtube] No video results')
 961
 962             new_ids = list(video['id'] for video in api_response['items'])
 963             video_ids += new_ids
 964
 965             limit = min(n, api_response['totalItems'])
 966             pagenum += 1
 967
 968         if len(video_ids) > n:
 969             video_ids = video_ids[:n]
 970         videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
 971         return self.playlist_result(videos, query)
 972
 973
 974 class YoutubeShowIE(InfoExtractor):
 975     IE_DESC = u'YouTube.com (multi-season) shows'
 976     _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
 977     IE_NAME = u'youtube:show'
 978
 979     def _real_extract(self, url):
 980         mobj = re.match(self._VALID_URL, url)
 981         show_name = mobj.group(1)
 982         webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
 983         # There's one playlist for each season of the show
 984         m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
 985         self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
 986         return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
 987
 988
 989 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
 990     """
 991     Base class for extractors that fetch info from
 992     http://www.youtube.com/feed_ajax
 993     Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
 994     """
 995     _LOGIN_REQUIRED = True
 996     _PAGING_STEP = 30
 997     # use action_load_personal_feed instead of action_load_system_feed
 998     _PERSONAL_FEED = False
 999
1000     @property
1001     def _FEED_TEMPLATE(self):
1002         action = 'action_load_system_feed'
1003         if self._PERSONAL_FEED:
1004             action = 'action_load_personal_feed'
1005         return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1006
1007     @property
1008     def IE_NAME(self):
1009         return u'youtube:%s' % self._FEED_NAME
1010
1011     def _real_initialize(self):
1012         self._login()
1013
1014     def _real_extract(self, url):
1015         feed_entries = []
1016         # The step argument is available only in 2.7 or higher
1017         for i in itertools.count(0):
1018             paging = i*self._PAGING_STEP
1019             info = self._download_webpage(self._FEED_TEMPLATE % paging,
1020                                           u'%s feed' % self._FEED_NAME,
1021                                           u'Downloading page %s' % i)
1022             info = json.loads(info)
1023             feed_html = info['feed_html']
1024             m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1025             ids = orderedSet(m.group(1) for m in m_ids)
1026             feed_entries.extend(self.url_result(id, 'Youtube') for id in ids)
1027             if info['paging'] is None:
1028                 break
1029         return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1030
1031 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1032     IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1033     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1034     _FEED_NAME = 'subscriptions'
1035     _PLAYLIST_TITLE = u'Youtube Subscriptions'
1036
1037 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1038     IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1039     _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1040     _FEED_NAME = 'recommended'
1041     _PLAYLIST_TITLE = u'Youtube Recommended videos'
1042
1043 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1044     IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1045     _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1046     _FEED_NAME = 'watch_later'
1047     _PLAYLIST_TITLE = u'Youtube Watch Later'
1048     _PAGING_STEP = 100
1049     _PERSONAL_FEED = True
1050
1051 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1052     IE_NAME = u'youtube:favorites'
1053     IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1054     _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:o?rites)?'
1055     _LOGIN_REQUIRED = True
1056
1057     def _real_extract(self, url):
1058         webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1059         playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1060         return self.url_result(playlist_id, 'YoutubePlaylist')