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