_ 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                           # 3D
 159                           '85', '84', '102', '83', '101', '82', '100',
 160                           # Dash video
 161                           '138', '137', '248', '136', '247', '135', '246',
 162                           '245', '244', '134', '243', '133', '242', '160',
 163                           # Dash audio
 164                           '141', '172', '140', '171', '139',
 165                           ]
 166     _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '17', '13',
 167                                       '95', '94', '93', '92', '132', '151',
 168                                       '85', '102', '84', '101', '83', '100', '82',
 169                                       # Dash video
 170                                       '138', '248', '137', '247', '136', '246', '245',
 171                                       '244', '135', '243', '134', '242', '133', '160',
 172                                       # Dash audio
 173                                       '172', '141', '171', '140', '139',
 174                                       ]
 175     _video_extensions = {
 176         '13': '3gp',
 177         '17': 'mp4',
 178         '18': 'mp4',
 179         '22': 'mp4',
 180         '37': 'mp4',
 181         '38': 'mp4',
 182         '43': 'webm',
 183         '44': 'webm',
 184         '45': 'webm',
 185         '46': 'webm',
 186
 187         # 3d videos
 188         '82': 'mp4',
 189         '83': 'mp4',
 190         '84': 'mp4',
 191         '85': 'mp4',
 192         '100': 'webm',
 193         '101': 'webm',
 194         '102': 'webm',
 195
 196         # videos that use m3u8
 197         '92': 'mp4',
 198         '93': 'mp4',
 199         '94': 'mp4',
 200         '95': 'mp4',
 201         '96': 'mp4',
 202         '132': 'mp4',
 203         '151': 'mp4',
 204
 205         # Dash mp4
 206         '133': 'mp4',
 207         '134': 'mp4',
 208         '135': 'mp4',
 209         '136': 'mp4',
 210         '137': 'mp4',
 211         '138': 'mp4',
 212         '139': 'mp4',
 213         '140': 'mp4',
 214         '141': 'mp4',
 215         '160': 'mp4',
 216
 217         # Dash webm
 218         '171': 'webm',
 219         '172': 'webm',
 220         '242': 'webm',
 221         '243': 'webm',
 222         '244': 'webm',
 223         '245': 'webm',
 224         '246': 'webm',
 225         '247': 'webm',
 226         '248': 'webm',
 227     }
 228     _video_dimensions = {
 229         '5': '240x400',
 230         '6': '???',
 231         '13': '???',
 232         '17': '144x176',
 233         '18': '360x640',
 234         '22': '720x1280',
 235         '34': '360x640',
 236         '35': '480x854',
 237         '37': '1080x1920',
 238         '38': '3072x4096',
 239         '43': '360x640',
 240         '44': '480x854',
 241         '45': '720x1280',
 242         '46': '1080x1920',
 243         '82': '360p',
 244         '83': '480p',
 245         '84': '720p',
 246         '85': '1080p',
 247         '92': '240p',
 248         '93': '360p',
 249         '94': '480p',
 250         '95': '720p',
 251         '96': '1080p',
 252         '100': '360p',
 253         '101': '480p',
 254         '102': '720p',
 255         '132': '240p',
 256         '151': '72p',
 257         '133': '240p',
 258         '134': '360p',
 259         '135': '480p',
 260         '136': '720p',
 261         '137': '1080p',
 262         '138': '>1080p',
 263         '139': '48k',
 264         '140': '128k',
 265         '141': '256k',
 266         '160': '192p',
 267         '171': '128k',
 268         '172': '256k',
 269         '242': '240p',
 270         '243': '360p',
 271         '244': '480p',
 272         '245': '480p',
 273         '246': '480p',
 274         '247': '720p',
 275         '248': '1080p',
 276     }
 277     _special_itags = {
 278         '82': '3D',
 279         '83': '3D',
 280         '84': '3D',
 281         '85': '3D',
 282         '100': '3D',
 283         '101': '3D',
 284         '102': '3D',
 285         '133': 'DASH Video',
 286         '134': 'DASH Video',
 287         '135': 'DASH Video',
 288         '136': 'DASH Video',
 289         '137': 'DASH Video',
 290         '138': 'DASH Video',
 291         '139': 'DASH Audio',
 292         '140': 'DASH Audio',
 293         '141': 'DASH Audio',
 294         '160': 'DASH Video',
 295         '171': 'DASH Audio',
 296         '172': 'DASH Audio',
 297         '242': 'DASH Video',
 298         '243': 'DASH Video',
 299         '244': 'DASH Video',
 300         '245': 'DASH Video',
 301         '246': 'DASH Video',
 302         '247': 'DASH Video',
 303         '248': 'DASH Video',
 304     }
 305
 306     IE_NAME = u'youtube'
 307     _TESTS = [
 308         {
 309             u"url":  u"http://www.youtube.com/watch?v=BaW_jenozKc",
 310             u"file":  u"BaW_jenozKc.mp4",
 311             u"info_dict": {
 312                 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
 313                 u"uploader": u"Philipp Hagemeister",
 314                 u"uploader_id": u"phihag",
 315                 u"upload_date": u"20121002",
 316                 u"description": u"test chars:  \"'/\\ä↭𝕐\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
 317             }
 318         },
 319         {
 320             u"url":  u"http://www.youtube.com/watch?v=1ltcDfZMA3U",
 321             u"file":  u"1ltcDfZMA3U.flv",
 322             u"note": u"Test VEVO video (#897)",
 323             u"info_dict": {
 324                 u"upload_date": u"20070518",
 325                 u"title": u"Maps - It Will Find You",
 326                 u"description": u"Music video by Maps performing It Will Find You.",
 327                 u"uploader": u"MuteUSA",
 328                 u"uploader_id": u"MuteUSA"
 329             }
 330         },
 331         {
 332             u"url":  u"http://www.youtube.com/watch?v=UxxajLWwzqY",
 333             u"file":  u"UxxajLWwzqY.mp4",
 334             u"note": u"Test generic use_cipher_signature video (#897)",
 335             u"info_dict": {
 336                 u"upload_date": u"20120506",
 337                 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
 338                 u"description": u"md5:b085c9804f5ab69f4adea963a2dceb3c",
 339                 u"uploader": u"Icona Pop",
 340                 u"uploader_id": u"IconaPop"
 341             }
 342         },
 343         {
 344             u"url":  u"https://www.youtube.com/watch?v=07FYdnEawAQ",
 345             u"file":  u"07FYdnEawAQ.mp4",
 346             u"note": u"Test VEVO video with age protection (#956)",
 347             u"info_dict": {
 348                 u"upload_date": u"20130703",
 349                 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
 350                 u"description": u"md5:64249768eec3bc4276236606ea996373",
 351                 u"uploader": u"justintimberlakeVEVO",
 352                 u"uploader_id": u"justintimberlakeVEVO"
 353             }
 354         },
 355         {
 356             u'url': u'https://www.youtube.com/watch?v=TGi3HqYrWHE',
 357             u'file': u'TGi3HqYrWHE.mp4',
 358             u'note': u'm3u8 video',
 359             u'info_dict': {
 360                 u'title': u'Triathlon - Men - London 2012 Olympic Games',
 361                 u'description': u'- Men -  TR02 - Triathlon - 07 August 2012 - London 2012 Olympic Games',
 362                 u'uploader': u'olympic',
 363                 u'upload_date': u'20120807',
 364                 u'uploader_id': u'olympic',
 365             },
 366             u'params': {
 367                 u'skip_download': True,
 368             },
 369         },
 370     ]
 371
 372
 373     @classmethod
 374     def suitable(cls, url):
 375         """Receives a URL and returns True if suitable for this IE."""
 376         if YoutubePlaylistIE.suitable(url) or YoutubeSubscriptionsIE.suitable(url): return False
 377         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
 378
 379     def report_video_webpage_download(self, video_id):
 380         """Report attempt to download video webpage."""
 381         self.to_screen(u'%s: Downloading video webpage' % video_id)
 382
 383     def report_video_info_webpage_download(self, video_id):
 384         """Report attempt to download video info webpage."""
 385         self.to_screen(u'%s: Downloading video info webpage' % video_id)
 386
 387     def report_video_subtitles_download(self, video_id):
 388         """Report attempt to download video info webpage."""
 389         self.to_screen(u'%s: Checking available subtitles' % video_id)
 390
 391     def report_video_subtitles_request(self, video_id, sub_lang, format):
 392         """Report attempt to download video info webpage."""
 393         self.to_screen(u'%s: Downloading video subtitles for %s.%s' % (video_id, sub_lang, format))
 394
 395     def report_video_subtitles_available(self, video_id, sub_lang_list):
 396         """Report available subtitles."""
 397         sub_lang = ",".join(list(sub_lang_list.keys()))
 398         self.to_screen(u'%s: Available subtitles for video: %s' % (video_id, sub_lang))
 399
 400     def report_information_extraction(self, video_id):
 401         """Report attempt to extract video information."""
 402         self.to_screen(u'%s: Extracting video information' % video_id)
 403
 404     def report_unavailable_format(self, video_id, format):
 405         """Report extracted video URL."""
 406         self.to_screen(u'%s: Format %s not available' % (video_id, format))
 407
 408     def report_rtmp_download(self):
 409         """Indicate the download will use the RTMP protocol."""
 410         self.to_screen(u'RTMP download detected')
 411
 412     def _decrypt_signature(self, s):
 413         """Turn the encrypted s field into a working signature"""
 414
 415         if len(s) == 92:
 416             return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
 417         elif len(s) == 90:
 418             return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
 419         elif len(s) == 89:
 420             return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1]
 421         elif len(s) == 88:
 422             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]
 423         elif len(s) == 87:
 424             return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
 425         elif len(s) == 86:
 426             return s[5:20] + s[2] + s[21:]
 427         elif len(s) == 85:
 428             return s[83:34:-1] + s[0] + s[33:27:-1] + s[3] + s[26:19:-1] + s[34] + s[18:3:-1] + s[27]
 429         elif len(s) == 84:
 430             return s[83:27:-1] + s[0] + s[26:5:-1] + s[2:0:-1] + s[27]
 431         elif len(s) == 83:
 432             return s[81:64:-1] + s[82] + s[63:52:-1] + s[45] + s[51:45:-1] + s[1] + s[44:1:-1] + s[0]
 433         elif len(s) == 82:
 434             return s[1:19] + s[0] + s[20:68] + s[19] + s[69:82]
 435         elif len(s) == 81:
 436             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]
 437         elif len(s) == 80:
 438             return s[1:19] + s[0] + s[20:68] + s[19] + s[69:80]
 439         elif len(s) == 79:
 440             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]
 441
 442         else:
 443             raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
 444
 445     def _decrypt_signature_age_gate(self, s):
 446         # The videos with age protection use another player, so the algorithms
 447         # can be different.
 448         if len(s) == 86:
 449             return s[2:63] + s[82] + s[64:82] + s[63]
 450         else:
 451             # Fallback to the other algortihms
 452             return self._decrypt_signature(s)
 453
 454
 455     def _get_available_subtitles(self, video_id):
 456         self.report_video_subtitles_download(video_id)
 457         request = compat_urllib_request.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
 458         try:
 459             sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8')
 460         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 461             self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
 462             return {}
 463         sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
 464         sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list)
 465         if not sub_lang_list:
 466             self._downloader.report_warning(u'video doesn\'t have subtitles')
 467             return {}
 468         return sub_lang_list
 469
 470     def _list_available_subtitles(self, video_id):
 471         sub_lang_list = self._get_available_subtitles(video_id)
 472         self.report_video_subtitles_available(video_id, sub_lang_list)
 473
 474     def _request_subtitle(self, sub_lang, sub_name, video_id, format):
 475         """
 476         Return the subtitle as a string or None if they are not found
 477         """
 478         self.report_video_subtitles_request(video_id, sub_lang, format)
 479         params = compat_urllib_parse.urlencode({
 480             'lang': sub_lang,
 481             'name': sub_name,
 482             'v': video_id,
 483             'fmt': format,
 484         })
 485         url = 'http://www.youtube.com/api/timedtext?' + params
 486         try:
 487             sub = compat_urllib_request.urlopen(url).read().decode('utf-8')
 488         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 489             self._downloader.report_warning(u'unable to download video subtitles for %s: %s' % (sub_lang, compat_str(err)))
 490             return
 491         if not sub:
 492             self._downloader.report_warning(u'Did not fetch video subtitles')
 493             return
 494         return sub
 495
 496     def _request_automatic_caption(self, video_id, webpage):
 497         """We need the webpage for getting the captions url, pass it as an
 498            argument to speed up the process."""
 499         sub_lang = self._downloader.params.get('subtitleslang') or 'en'
 500         sub_format = self._downloader.params.get('subtitlesformat')
 501         self.to_screen(u'%s: Looking for automatic captions' % video_id)
 502         mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
 503         err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
 504         if mobj is None:
 505             self._downloader.report_warning(err_msg)
 506             return {}
 507         player_config = json.loads(mobj.group(1))
 508         try:
 509             args = player_config[u'args']
 510             caption_url = args[u'ttsurl']
 511             timestamp = args[u'timestamp']
 512             params = compat_urllib_parse.urlencode({
 513                 'lang': 'en',
 514                 'tlang': sub_lang,
 515                 'fmt': sub_format,
 516                 'ts': timestamp,
 517                 'kind': 'asr',
 518             })
 519             subtitles_url = caption_url + '&' + params
 520             sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
 521             return {sub_lang: sub}
 522         # An extractor error can be raise by the download process if there are
 523         # no automatic captions but there are subtitles
 524         except (KeyError, ExtractorError):
 525             self._downloader.report_warning(err_msg)
 526             return {}
 527
 528     def _extract_subtitles(self, video_id):
 529         """
 530         Return a dictionary: {language: subtitles} or {} if the subtitles
 531         couldn't be found
 532         """
 533         sub_lang_list = self._get_available_subtitles(video_id)
 534         sub_format = self._downloader.params.get('subtitlesformat')
 535         if  not sub_lang_list: #There was some error, it didn't get the available subtitles
 536             return {}
 537         if self._downloader.params.get('allsubtitles', False):
 538             pass
 539         else:
 540             if self._downloader.params.get('subtitleslang', False):
 541                 sub_lang = self._downloader.params.get('subtitleslang')
 542             elif 'en' in sub_lang_list:
 543                 sub_lang = 'en'
 544             else:
 545                 sub_lang = list(sub_lang_list.keys())[0]
 546             if not sub_lang in sub_lang_list:
 547                 self._downloader.report_warning(u'no closed captions found in the specified language "%s"' % sub_lang)
 548                 return {}
 549             sub_lang_list = {sub_lang: sub_lang_list[sub_lang]}
 550         subtitles = {}
 551         for sub_lang in sub_lang_list:
 552             subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
 553             if subtitle:
 554                 subtitles[sub_lang] = subtitle
 555         return subtitles
 556
 557     def _print_formats(self, formats):
 558         print('Available formats:')
 559         for x in formats:
 560             print('%s\t:\t%s\t[%s]%s' %(x, self._video_extensions.get(x, 'flv'),
 561                                         self._video_dimensions.get(x, '???'),
 562                                         ' ('+self._special_itags[x]+')' if x in self._special_itags else ''))
 563
 564     def _extract_id(self, url):
 565         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 566         if mobj is None:
 567             raise ExtractorError(u'Invalid URL: %s' % url)
 568         video_id = mobj.group(2)
 569         return video_id
 570
 571     def _get_video_url_list(self, url_map):
 572         """
 573         Transform a dictionary in the format {itag:url} to a list of (itag, url)
 574         with the requested formats.
 575         """
 576         req_format = self._downloader.params.get('format', None)
 577         format_limit = self._downloader.params.get('format_limit', None)
 578         available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats
 579         if format_limit is not None and format_limit in available_formats:
 580             format_list = available_formats[available_formats.index(format_limit):]
 581         else:
 582             format_list = available_formats
 583         existing_formats = [x for x in format_list if x in url_map]
 584         if len(existing_formats) == 0:
 585             raise ExtractorError(u'no known formats available for video')
 586         if self._downloader.params.get('listformats', None):
 587             self._print_formats(existing_formats)
 588             return
 589         if req_format is None or req_format == 'best':
 590             video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality
 591         elif req_format == 'worst':
 592             video_url_list = [(existing_formats[-1], url_map[existing_formats[-1]])] # worst quality
 593         elif req_format in ('-1', 'all'):
 594             video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
 595         else:
 596             # Specific formats. We pick the first in a slash-delimeted sequence.
 597             # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'.
 598             req_formats = req_format.split('/')
 599             video_url_list = None
 600             for rf in req_formats:
 601                 if rf in url_map:
 602                     video_url_list = [(rf, url_map[rf])]
 603                     break
 604             if video_url_list is None:
 605                 raise ExtractorError(u'requested format not available')
 606         return video_url_list
 607
 608     def _extract_from_m3u8(self, manifest_url, video_id):
 609         url_map = {}
 610         def _get_urls(_manifest):
 611             lines = _manifest.split('\n')
 612             urls = filter(lambda l: l and not l.startswith('#'),
 613                             lines)
 614             return urls
 615         manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
 616         formats_urls = _get_urls(manifest)
 617         for format_url in formats_urls:
 618             itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
 619             url_map[itag] = format_url
 620         return url_map
 621
 622     def _real_extract(self, url):
 623         if re.match(r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$', url):
 624             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  ).')
 625
 626         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
 627         mobj = re.search(self._NEXT_URL_RE, url)
 628         if mobj:
 629             url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
 630         video_id = self._extract_id(url)
 631
 632         # Get video webpage
 633         self.report_video_webpage_download(video_id)
 634         url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
 635         request = compat_urllib_request.Request(url)
 636         try:
 637             video_webpage_bytes = compat_urllib_request.urlopen(request).read()
 638         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 639             raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err))
 640
 641         video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
 642
 643         # Attempt to extract SWF player URL
 644         mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
 645         if mobj is not None:
 646             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
 647         else:
 648             player_url = None
 649
 650         # Get video info
 651         self.report_video_info_webpage_download(video_id)
 652         if re.search(r'player-age-gate-content">', video_webpage) is not None:
 653             self.report_age_confirmation()
 654             age_gate = True
 655             # We simulate the access to the video from www.youtube.com/v/{video_id}
 656             # this can be viewed without login into Youtube
 657             data = compat_urllib_parse.urlencode({'video_id': video_id,
 658                                                   'el': 'embedded',
 659                                                   'gl': 'US',
 660                                                   'hl': 'en',
 661                                                   'eurl': 'https://youtube.googleapis.com/v/' + video_id,
 662                                                   'asv': 3,
 663                                                   'sts':'1588',
 664                                                   })
 665             video_info_url = 'https://www.youtube.com/get_video_info?' + data
 666             video_info_webpage = self._download_webpage(video_info_url, video_id,
 667                                     note=False,
 668                                     errnote='unable to download video info webpage')
 669             video_info = compat_parse_qs(video_info_webpage)
 670         else:
 671             age_gate = False
 672             for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
 673                 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
 674                         % (video_id, el_type))
 675                 video_info_webpage = self._download_webpage(video_info_url, video_id,
 676                                         note=False,
 677                                         errnote='unable to download video info webpage')
 678                 video_info = compat_parse_qs(video_info_webpage)
 679                 if 'token' in video_info:
 680                     break
 681         if 'token' not in video_info:
 682             if 'reason' in video_info:
 683                 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
 684             else:
 685                 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
 686
 687         # Check for "rental" videos
 688         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
 689             raise ExtractorError(u'"rental" videos not supported')
 690
 691         # Start extracting information
 692         self.report_information_extraction(video_id)
 693
 694         # uploader
 695         if 'author' not in video_info:
 696             raise ExtractorError(u'Unable to extract uploader name')
 697         video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
 698
 699         # uploader_id
 700         video_uploader_id = None
 701         mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
 702         if mobj is not None:
 703             video_uploader_id = mobj.group(1)
 704         else:
 705             self._downloader.report_warning(u'unable to extract uploader nickname')
 706
 707         # title
 708         if 'title' not in video_info:
 709             raise ExtractorError(u'Unable to extract video title')
 710         video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
 711
 712         # thumbnail image
 713         # We try first to get a high quality image:
 714         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
 715                             video_webpage, re.DOTALL)
 716         if m_thumb is not None:
 717             video_thumbnail = m_thumb.group(1)
 718         elif 'thumbnail_url' not in video_info:
 719             self._downloader.report_warning(u'unable to extract video thumbnail')
 720             video_thumbnail = ''
 721         else:   # don't panic if we can't find it
 722             video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
 723
 724         # upload date
 725         upload_date = None
 726         mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
 727         if mobj is not None:
 728             upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
 729             upload_date = unified_strdate(upload_date)
 730
 731         # description
 732         video_description = get_element_by_id("eow-description", video_webpage)
 733         if video_description:
 734             video_description = clean_html(video_description)
 735         else:
 736             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
 737             if fd_mobj:
 738                 video_description = unescapeHTML(fd_mobj.group(1))
 739             else:
 740                 video_description = u''
 741
 742         # subtitles
 743         video_subtitles = None
 744
 745         if self._downloader.params.get('writesubtitles', False) or self._downloader.params.get('allsubtitles', False):
 746             video_subtitles = self._extract_subtitles(video_id)
 747         elif self._downloader.params.get('writeautomaticsub', False):
 748             video_subtitles = self._request_automatic_caption(video_id, video_webpage)
 749
 750         if self._downloader.params.get('listsubtitles', False):
 751             self._list_available_subtitles(video_id)
 752             return
 753
 754         if 'length_seconds' not in video_info:
 755             self._downloader.report_warning(u'unable to extract video duration')
 756             video_duration = ''
 757         else:
 758             video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
 759
 760         # Decide which formats to download
 761
 762         try:
 763             mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
 764             if not mobj:
 765                 raise ValueError('Could not find vevo ID')
 766             info = json.loads(mobj.group(1))
 767             args = info['args']
 768             # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
 769             # this signatures are encrypted
 770             m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
 771             if m_s is not None:
 772                 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
 773                 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
 774             m_s = re.search(r'[&,]s=', args.get('adaptive_fmts', u''))
 775             if m_s is not None:
 776                 if 'url_encoded_fmt_stream_map' in video_info:
 777                     video_info['url_encoded_fmt_stream_map'][0] += ',' + args['adaptive_fmts']
 778                 else:
 779                     video_info['url_encoded_fmt_stream_map'] = [args['adaptive_fmts']]
 780             elif 'adaptive_fmts' in video_info:
 781                 if 'url_encoded_fmt_stream_map' in video_info:
 782                     video_info['url_encoded_fmt_stream_map'][0] += ',' + video_info['adaptive_fmts'][0]
 783                 else:
 784                     video_info['url_encoded_fmt_stream_map'] = video_info['adaptive_fmts']
 785         except ValueError:
 786             pass
 787
 788         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
 789             self.report_rtmp_download()
 790             video_url_list = [(None, video_info['conn'][0])]
 791         elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1:
 792             if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]:
 793                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
 794             url_map = {}
 795             for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','):
 796                 url_data = compat_parse_qs(url_data_str)
 797                 if 'itag' in url_data and 'url' in url_data:
 798                     url = url_data['url'][0]
 799                     if 'sig' in url_data:
 800                         url += '&signature=' + url_data['sig'][0]
 801                     elif 's' in url_data:
 802                         if self._downloader.params.get('verbose'):
 803                             s = url_data['s'][0]
 804                             if age_gate:
 805                                 player_version = self._search_regex(r'ad3-(.+?)\.swf',
 806                                     video_info['ad3_module'][0] if 'ad3_module' in video_info else 'NOT FOUND',
 807                                     'flash player', fatal=False)
 808                                 player = 'flash player %s' % player_version
 809                             else:
 810                                 player = u'html5 player %s' % self._search_regex(r'html5player-(.+?)\.js', video_webpage,
 811                                     'html5 player', fatal=False)
 812                             parts_sizes = u'.'.join(compat_str(len(part)) for part in s.split('.'))
 813                             self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
 814                                 (len(s), parts_sizes, url_data['itag'][0], player))
 815                         encrypted_sig = url_data['s'][0]
 816                         if age_gate:
 817                             signature = self._decrypt_signature_age_gate(encrypted_sig)
 818                         else:
 819                             signature = self._decrypt_signature(encrypted_sig)
 820                         url += '&signature=' + signature
 821                     if 'ratebypass' not in url:
 822                         url += '&ratebypass=yes'
 823                     url_map[url_data['itag'][0]] = url
 824             video_url_list = self._get_video_url_list(url_map)
 825             if not video_url_list:
 826                 return
 827         elif video_info.get('hlsvp'):
 828             manifest_url = video_info['hlsvp'][0]
 829             url_map = self._extract_from_m3u8(manifest_url, video_id)
 830             video_url_list = self._get_video_url_list(url_map)
 831             if not video_url_list:
 832                 return
 833
 834         else:
 835             raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info')
 836
 837         results = []
 838         for format_param, video_real_url in video_url_list:
 839             # Extension
 840             video_extension = self._video_extensions.get(format_param, 'flv')
 841
 842             video_format = '{0} - {1}{2}'.format(format_param if format_param else video_extension,
 843                                               self._video_dimensions.get(format_param, '???'),
 844                                               ' ('+self._special_itags[format_param]+')' if format_param in self._special_itags else '')
 845
 846             results.append({
 847                 'id':       video_id,
 848                 'url':      video_real_url,
 849                 'uploader': video_uploader,
 850                 'uploader_id': video_uploader_id,
 851                 'upload_date':  upload_date,
 852                 'title':    video_title,
 853                 'ext':      video_extension,
 854                 'format':   video_format,
 855                 'thumbnail':    video_thumbnail,
 856                 'description':  video_description,
 857                 'player_url':   player_url,
 858                 'subtitles':    video_subtitles,
 859                 'duration':     video_duration
 860             })
 861         return results
 862
 863 class YoutubePlaylistIE(InfoExtractor):
 864     IE_DESC = u'YouTube.com playlists'
 865     _VALID_URL = r"""(?:
 866                         (?:https?://)?
 867                         (?:\w+\.)?
 868                         youtube\.com/
 869                         (?:
 870                            (?:course|view_play_list|my_playlists|artist|playlist|watch)
 871                            \? (?:.*?&)*? (?:p|a|list)=
 872                         |  p/
 873                         )
 874                         ((?:PL|EC|UU|FL)?[0-9A-Za-z-_]{10,})
 875                         .*
 876                      |
 877                         ((?:PL|EC|UU|FL)[0-9A-Za-z-_]{10,})
 878                      )"""
 879     _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none'
 880     _MAX_RESULTS = 50
 881     IE_NAME = u'youtube:playlist'
 882
 883     @classmethod
 884     def suitable(cls, url):
 885         """Receives a URL and returns True if suitable for this IE."""
 886         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
 887
 888     def _real_extract(self, url):
 889         # Extract playlist id
 890         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 891         if mobj is None:
 892             raise ExtractorError(u'Invalid URL: %s' % url)
 893
 894         # Download playlist videos from API
 895         playlist_id = mobj.group(1) or mobj.group(2)
 896         videos = []
 897
 898         for page_num in itertools.count(1):
 899             start_index = self._MAX_RESULTS * (page_num - 1) + 1
 900             if start_index >= 1000:
 901                 self._downloader.report_warning(u'Max number of results reached')
 902                 break
 903             url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, start_index)
 904             page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
 905
 906             try:
 907                 response = json.loads(page)
 908             except ValueError as err:
 909                 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
 910
 911             if 'feed' not in response:
 912                 raise ExtractorError(u'Got a malformed response from YouTube API')
 913             playlist_title = response['feed']['title']['$t']
 914             if 'entry' not in response['feed']:
 915                 # Number of videos is a multiple of self._MAX_RESULTS
 916                 break
 917
 918             for entry in response['feed']['entry']:
 919                 index = entry['yt$position']['$t']
 920                 if 'media$group' in entry and 'media$player' in entry['media$group']:
 921                     videos.append((index, entry['media$group']['media$player']['url']))
 922
 923         videos = [v[1] for v in sorted(videos)]
 924
 925         url_results = [self.url_result(vurl, 'Youtube') for vurl in videos]
 926         return [self.playlist_result(url_results, playlist_id, playlist_title)]
 927
 928
 929 class YoutubeChannelIE(InfoExtractor):
 930     IE_DESC = u'YouTube.com channels'
 931     _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
 932     _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en'
 933     _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
 934     _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'
 935     IE_NAME = u'youtube:channel'
 936
 937     def extract_videos_from_page(self, page):
 938         ids_in_page = []
 939         for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
 940             if mobj.group(1) not in ids_in_page:
 941                 ids_in_page.append(mobj.group(1))
 942         return ids_in_page
 943
 944     def _real_extract(self, url):
 945         # Extract channel id
 946         mobj = re.match(self._VALID_URL, url)
 947         if mobj is None:
 948             raise ExtractorError(u'Invalid URL: %s' % url)
 949
 950         # Download channel page
 951         channel_id = mobj.group(1)
 952         video_ids = []
 953         pagenum = 1
 954
 955         url = self._TEMPLATE_URL % (channel_id, pagenum)
 956         page = self._download_webpage(url, channel_id,
 957                                       u'Downloading page #%s' % pagenum)
 958
 959         # Extract video identifiers
 960         ids_in_page = self.extract_videos_from_page(page)
 961         video_ids.extend(ids_in_page)
 962
 963         # Download any subsequent channel pages using the json-based channel_ajax query
 964         if self._MORE_PAGES_INDICATOR in page:
 965             for pagenum in itertools.count(1):
 966                 url = self._MORE_PAGES_URL % (pagenum, channel_id)
 967                 page = self._download_webpage(url, channel_id,
 968                                               u'Downloading page #%s' % pagenum)
 969
 970                 page = json.loads(page)
 971
 972                 ids_in_page = self.extract_videos_from_page(page['content_html'])
 973                 video_ids.extend(ids_in_page)
 974
 975                 if self._MORE_PAGES_INDICATOR  not in page['load_more_widget_html']:
 976                     break
 977
 978         self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
 979
 980         urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids]
 981         url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls]
 982         return [self.playlist_result(url_entries, channel_id)]
 983
 984
 985 class YoutubeUserIE(InfoExtractor):
 986     IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
 987     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
 988     _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
 989     _GDATA_PAGE_SIZE = 50
 990     _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
 991     _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
 992     IE_NAME = u'youtube:user'
 993
 994     def _real_extract(self, url):
 995         # Extract username
 996         mobj = re.match(self._VALID_URL, url)
 997         if mobj is None:
 998             raise ExtractorError(u'Invalid URL: %s' % url)
 999
1000         username = mobj.group(1)
1001
1002         # Download video ids using YouTube Data API. Result size per
1003         # query is limited (currently to 50 videos) so we need to query
1004         # page by page until there are no video ids - it means we got
1005         # all of them.
1006
1007         video_ids = []
1008
1009         for pagenum in itertools.count(0):
1010             start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1011
1012             gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1013             page = self._download_webpage(gdata_url, username,
1014                                           u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
1015
1016             # Extract video identifiers
1017             ids_in_page = []
1018
1019             for mobj in re.finditer(self._VIDEO_INDICATOR, page):
1020                 if mobj.group(1) not in ids_in_page:
1021                     ids_in_page.append(mobj.group(1))
1022
1023             video_ids.extend(ids_in_page)
1024
1025             # A little optimization - if current page is not
1026             # "full", ie. does not contain PAGE_SIZE video ids then
1027             # we can assume that this page is the last one - there
1028             # are no more ids on further pages - no need to query
1029             # again.
1030
1031             if len(ids_in_page) < self._GDATA_PAGE_SIZE:
1032                 break
1033
1034         urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
1035         url_results = [self.url_result(rurl, 'Youtube') for rurl in urls]
1036         return [self.playlist_result(url_results, playlist_title = username)]
1037
1038 class YoutubeSearchIE(SearchInfoExtractor):
1039     IE_DESC = u'YouTube.com searches'
1040     _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1041     _MAX_RESULTS = 1000
1042     IE_NAME = u'youtube:search'
1043     _SEARCH_KEY = 'ytsearch'
1044
1045     def report_download_page(self, query, pagenum):
1046         """Report attempt to download search page with given number."""
1047         self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
1048
1049     def _get_n_results(self, query, n):
1050         """Get a specified number of results for a query"""
1051
1052         video_ids = []
1053         pagenum = 0
1054         limit = n
1055
1056         while (50 * pagenum) < limit:
1057             self.report_download_page(query, pagenum+1)
1058             result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
1059             request = compat_urllib_request.Request(result_url)
1060             try:
1061                 data = compat_urllib_request.urlopen(request).read().decode('utf-8')
1062             except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1063                 raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
1064             api_response = json.loads(data)['data']
1065
1066             if not 'items' in api_response:
1067                 raise ExtractorError(u'[youtube] No video results')
1068
1069             new_ids = list(video['id'] for video in api_response['items'])
1070             video_ids += new_ids
1071
1072             limit = min(n, api_response['totalItems'])
1073             pagenum += 1
1074
1075         if len(video_ids) > n:
1076             video_ids = video_ids[:n]
1077         videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
1078         return self.playlist_result(videos, query)
1079
1080
1081 class YoutubeShowIE(InfoExtractor):
1082     IE_DESC = u'YouTube.com (multi-season) shows'
1083     _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1084     IE_NAME = u'youtube:show'
1085
1086     def _real_extract(self, url):
1087         mobj = re.match(self._VALID_URL, url)
1088         show_name = mobj.group(1)
1089         webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1090         # There's one playlist for each season of the show
1091         m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1092         self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1093         return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1094
1095
1096 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1097     """
1098     Base class for extractors that fetch info from
1099     http://www.youtube.com/feed_ajax
1100     Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1101     """
1102     _LOGIN_REQUIRED = True
1103     _PAGING_STEP = 30
1104     # use action_load_personal_feed instead of action_load_system_feed
1105     _PERSONAL_FEED = False
1106
1107     @property
1108     def _FEED_TEMPLATE(self):
1109         action = 'action_load_system_feed'
1110         if self._PERSONAL_FEED:
1111             action = 'action_load_personal_feed'
1112         return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1113
1114     @property
1115     def IE_NAME(self):
1116         return u'youtube:%s' % self._FEED_NAME
1117
1118     def _real_initialize(self):
1119         self._login()
1120
1121     def _real_extract(self, url):
1122         feed_entries = []
1123         # The step argument is available only in 2.7 or higher
1124         for i in itertools.count(0):
1125             paging = i*self._PAGING_STEP
1126             info = self._download_webpage(self._FEED_TEMPLATE % paging,
1127                                           u'%s feed' % self._FEED_NAME,
1128                                           u'Downloading page %s' % i)
1129             info = json.loads(info)
1130             feed_html = info['feed_html']
1131             m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1132             ids = orderedSet(m.group(1) for m in m_ids)
1133             feed_entries.extend(self.url_result(id, 'Youtube') for id in ids)
1134             if info['paging'] is None:
1135                 break
1136         return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1137
1138 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1139     IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1140     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1141     _FEED_NAME = 'subscriptions'
1142     _PLAYLIST_TITLE = u'Youtube Subscriptions'
1143
1144 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1145     IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1146     _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1147     _FEED_NAME = 'recommended'
1148     _PLAYLIST_TITLE = u'Youtube Recommended videos'
1149
1150 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1151     IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1152     _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1153     _FEED_NAME = 'watch_later'
1154     _PLAYLIST_TITLE = u'Youtube Watch Later'
1155     _PAGING_STEP = 100
1156     _PERSONAL_FEED = True
1157
1158 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1159     IE_NAME = u'youtube:favorites'
1160     IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1161     _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:o?rites)?'
1162     _LOGIN_REQUIRED = True
1163
1164     def _real_extract(self, url):
1165         webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1166         playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1167         return self.url_result(playlist_id, 'YoutubePlaylist')