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