_ 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             self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
 213             return {}
 214         sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
 215         sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list)
 216         if not sub_lang_list:
 217             self._downloader.report_warning(u'video doesn\'t have subtitles')
 218             return {}
 219         return sub_lang_list
 220
 221     def _list_available_subtitles(self, video_id):
 222         sub_lang_list = self._get_available_subtitles(video_id)
 223         self.report_video_subtitles_available(video_id, sub_lang_list)
 224
 225     def _request_subtitle(self, sub_lang, sub_name, video_id, format):
 226         """
 227         Return the subtitle as a string or None if they are not found
 228         """
 229         self.report_video_subtitles_request(video_id, sub_lang, format)
 230         params = compat_urllib_parse.urlencode({
 231             'lang': sub_lang,
 232             'name': sub_name,
 233             'v': video_id,
 234             'fmt': format,
 235         })
 236         url = 'http://www.youtube.com/api/timedtext?' + params
 237         try:
 238             sub = compat_urllib_request.urlopen(url).read().decode('utf-8')
 239         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 240             self._downloader.report_warning(u'unable to download video subtitles for %s: %s' % (sub_lang, compat_str(err)))
 241             return
 242         if not sub:
 243             self._downloader.report_warning(u'Did not fetch video subtitles')
 244             return
 245         return sub
 246
 247     def _request_automatic_caption(self, video_id, webpage):
 248         """We need the webpage for getting the captions url, pass it as an
 249            argument to speed up the process."""
 250         sub_lang = self._downloader.params.get('subtitleslang') or 'en'
 251         sub_format = self._downloader.params.get('subtitlesformat')
 252         self.to_screen(u'%s: Looking for automatic captions' % video_id)
 253         mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
 254         err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
 255         if mobj is None:
 256             self._downloader.report_warning(err_msg)
 257             return {}
 258         player_config = json.loads(mobj.group(1))
 259         try:
 260             args = player_config[u'args']
 261             caption_url = args[u'ttsurl']
 262             timestamp = args[u'timestamp']
 263             params = compat_urllib_parse.urlencode({
 264                 'lang': 'en',
 265                 'tlang': sub_lang,
 266                 'fmt': sub_format,
 267                 'ts': timestamp,
 268                 'kind': 'asr',
 269             })
 270             subtitles_url = caption_url + '&' + params
 271             sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
 272             return {sub_lang: sub}
 273         except KeyError:
 274             self._downloader.report_warning(err_msg)
 275             return {}
 276
 277     def _extract_subtitle(self, video_id):
 278         """
 279         Return a dictionary: {language: subtitles} or {} if the subtitles
 280         couldn't be found
 281         """
 282         sub_lang_list = self._get_available_subtitles(video_id)
 283         sub_format = self._downloader.params.get('subtitlesformat')
 284         if  not sub_lang_list: #There was some error, it didn't get the available subtitles
 285             return {}
 286         if self._downloader.params.get('subtitleslang', False):
 287             sub_lang = self._downloader.params.get('subtitleslang')
 288         elif 'en' in sub_lang_list:
 289             sub_lang = 'en'
 290         else:
 291             sub_lang = list(sub_lang_list.keys())[0]
 292         if not sub_lang in sub_lang_list:
 293             self._downloader.report_warning(u'no closed captions found in the specified language "%s"' % sub_lang)
 294             return {}
 295
 296         subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
 297         if subtitle:
 298             self.to_screen('sub %s' % subtitle[:20])
 299             return {sub_lang: subtitle}
 300         else:
 301             return {}
 302
 303     def _extract_all_subtitles(self, video_id):
 304         """
 305         Return a dicitonary: {language: subtitles} or {} if the subtitles
 306         couldn't be found
 307         """
 308         sub_lang_list = self._get_available_subtitles(video_id)
 309         sub_format = self._downloader.params.get('subtitlesformat')
 310         subtitles = {}
 311         for sub_lang in sub_lang_list:
 312             subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
 313             if subtitle:
 314                 subtitles[sub_lang] = subtitle
 315         return subtitles
 316
 317     def _print_formats(self, formats):
 318         print('Available formats:')
 319         for x in formats:
 320             print('%s\t:\t%s\t[%s]' %(x, self._video_extensions.get(x, 'flv'), self._video_dimensions.get(x, '???')))
 321
 322     def _real_initialize(self):
 323         if self._downloader is None:
 324             return
 325
 326         # Set language
 327         request = compat_urllib_request.Request(self._LANG_URL)
 328         try:
 329             self.report_lang()
 330             compat_urllib_request.urlopen(request).read()
 331         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 332             self._downloader.report_warning(u'unable to set language: %s' % compat_str(err))
 333             return
 334
 335         (username, password) = self._get_login_info()
 336
 337         # No authentication to be performed
 338         if username is None:
 339             return
 340
 341         request = compat_urllib_request.Request(self._LOGIN_URL)
 342         try:
 343             login_page = compat_urllib_request.urlopen(request).read().decode('utf-8')
 344         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 345             self._downloader.report_warning(u'unable to fetch login page: %s' % compat_str(err))
 346             return
 347
 348         galx = None
 349         dsh = None
 350         match = re.search(re.compile(r'<input.+?name="GALX".+?value="(.+?)"', re.DOTALL), login_page)
 351         if match:
 352           galx = match.group(1)
 353
 354         match = re.search(re.compile(r'<input.+?name="dsh".+?value="(.+?)"', re.DOTALL), login_page)
 355         if match:
 356           dsh = match.group(1)
 357
 358         # Log in
 359         login_form_strs = {
 360                 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
 361                 u'Email': username,
 362                 u'GALX': galx,
 363                 u'Passwd': password,
 364                 u'PersistentCookie': u'yes',
 365                 u'_utf8': u'霱',
 366                 u'bgresponse': u'js_disabled',
 367                 u'checkConnection': u'',
 368                 u'checkedDomains': u'youtube',
 369                 u'dnConn': u'',
 370                 u'dsh': dsh,
 371                 u'pstMsg': u'0',
 372                 u'rmShown': u'1',
 373                 u'secTok': u'',
 374                 u'signIn': u'Sign in',
 375                 u'timeStmp': u'',
 376                 u'service': u'youtube',
 377                 u'uilel': u'3',
 378                 u'hl': u'en_US',
 379         }
 380         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
 381         # chokes on unicode
 382         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
 383         login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
 384         request = compat_urllib_request.Request(self._LOGIN_URL, login_data)
 385         try:
 386             self.report_login()
 387             login_results = compat_urllib_request.urlopen(request).read().decode('utf-8')
 388             if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
 389                 self._downloader.report_warning(u'unable to log in: bad username or password')
 390                 return
 391         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 392             self._downloader.report_warning(u'unable to log in: %s' % compat_str(err))
 393             return
 394
 395         # Confirm age
 396         age_form = {
 397                 'next_url':     '/',
 398                 'action_confirm':   'Confirm',
 399                 }
 400         request = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
 401         try:
 402             self.report_age_confirmation()
 403             compat_urllib_request.urlopen(request).read().decode('utf-8')
 404         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 405             raise ExtractorError(u'Unable to confirm age: %s' % compat_str(err))
 406
 407     def _extract_id(self, url):
 408         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 409         if mobj is None:
 410             raise ExtractorError(u'Invalid URL: %s' % url)
 411         video_id = mobj.group(2)
 412         return video_id
 413
 414     def _real_extract(self, url):
 415         if re.match(r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$', url):
 416             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  ).')
 417
 418         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
 419         mobj = re.search(self._NEXT_URL_RE, url)
 420         if mobj:
 421             url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
 422         video_id = self._extract_id(url)
 423
 424         # Get video webpage
 425         self.report_video_webpage_download(video_id)
 426         url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
 427         request = compat_urllib_request.Request(url)
 428         try:
 429             video_webpage_bytes = compat_urllib_request.urlopen(request).read()
 430         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 431             raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err))
 432
 433         video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
 434
 435         # Attempt to extract SWF player URL
 436         mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
 437         if mobj is not None:
 438             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
 439         else:
 440             player_url = None
 441
 442         # Get video info
 443         self.report_video_info_webpage_download(video_id)
 444         if re.search(r'player-age-gate-content">', video_webpage) is not None:
 445             self.report_age_confirmation()
 446             age_gate = True
 447             # We simulate the access to the video from www.youtube.com/v/{video_id}
 448             # this can be viewed without login into Youtube
 449             data = compat_urllib_parse.urlencode({'video_id': video_id,
 450                                                   'el': 'embedded',
 451                                                   'gl': 'US',
 452                                                   'hl': 'en',
 453                                                   'eurl': 'https://youtube.googleapis.com/v/' + video_id,
 454                                                   'asv': 3,
 455                                                   'sts':'1588',
 456                                                   })
 457             video_info_url = 'https://www.youtube.com/get_video_info?' + data
 458             video_info_webpage = self._download_webpage(video_info_url, video_id,
 459                                     note=False,
 460                                     errnote='unable to download video info webpage')
 461             video_info = compat_parse_qs(video_info_webpage)
 462         else:
 463             age_gate = False
 464             for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
 465                 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
 466                         % (video_id, el_type))
 467                 video_info_webpage = self._download_webpage(video_info_url, video_id,
 468                                         note=False,
 469                                         errnote='unable to download video info webpage')
 470                 video_info = compat_parse_qs(video_info_webpage)
 471                 if 'token' in video_info:
 472                     break
 473         if 'token' not in video_info:
 474             if 'reason' in video_info:
 475                 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
 476             else:
 477                 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
 478
 479         # Check for "rental" videos
 480         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
 481             raise ExtractorError(u'"rental" videos not supported')
 482
 483         # Start extracting information
 484         self.report_information_extraction(video_id)
 485
 486         # uploader
 487         if 'author' not in video_info:
 488             raise ExtractorError(u'Unable to extract uploader name')
 489         video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
 490
 491         # uploader_id
 492         video_uploader_id = None
 493         mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
 494         if mobj is not None:
 495             video_uploader_id = mobj.group(1)
 496         else:
 497             self._downloader.report_warning(u'unable to extract uploader nickname')
 498
 499         # title
 500         if 'title' not in video_info:
 501             raise ExtractorError(u'Unable to extract video title')
 502         video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
 503
 504         # thumbnail image
 505         # We try first to get a high quality image:
 506         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
 507                             video_webpage, re.DOTALL)
 508         if m_thumb is not None:
 509             video_thumbnail = m_thumb.group(1)
 510         elif 'thumbnail_url' not in video_info:
 511             self._downloader.report_warning(u'unable to extract video thumbnail')
 512             video_thumbnail = ''
 513         else:   # don't panic if we can't find it
 514             video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
 515
 516         # upload date
 517         upload_date = None
 518         mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
 519         if mobj is not None:
 520             upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
 521             upload_date = unified_strdate(upload_date)
 522
 523         # description
 524         video_description = get_element_by_id("eow-description", video_webpage)
 525         if video_description:
 526             video_description = clean_html(video_description)
 527         else:
 528             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
 529             if fd_mobj:
 530                 video_description = unescapeHTML(fd_mobj.group(1))
 531             else:
 532                 video_description = u''
 533
 534         # subtitles
 535         video_subtitles = None
 536
 537         if self._downloader.params.get('writesubtitles', False):
 538             video_subtitles = self._extract_subtitle(video_id)
 539         elif self._downloader.params.get('writeautomaticsub', False):
 540             video_subtitles = self._request_automatic_caption(video_id, video_webpage)
 541
 542         if self._downloader.params.get('allsubtitles', False):
 543             video_subtitles = self._extract_all_subtitles(video_id)
 544
 545         if self._downloader.params.get('listsubtitles', False):
 546             self._list_available_subtitles(video_id)
 547             return
 548
 549         if 'length_seconds' not in video_info:
 550             self._downloader.report_warning(u'unable to extract video duration')
 551             video_duration = ''
 552         else:
 553             video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
 554
 555         # Decide which formats to download
 556         req_format = self._downloader.params.get('format', None)
 557
 558         try:
 559             mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
 560             if not mobj:
 561                 raise ValueError('Could not find vevo ID')
 562             info = json.loads(mobj.group(1))
 563             args = info['args']
 564             # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
 565             # this signatures are encrypted
 566             m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
 567             if m_s is not None:
 568                 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
 569                 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
 570         except ValueError:
 571             pass
 572
 573         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
 574             self.report_rtmp_download()
 575             video_url_list = [(None, video_info['conn'][0])]
 576         elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1:
 577             if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]:
 578                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
 579             url_map = {}
 580             for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','):
 581                 url_data = compat_parse_qs(url_data_str)
 582                 if 'itag' in url_data and 'url' in url_data:
 583                     url = url_data['url'][0]
 584                     if 'sig' in url_data:
 585                         url += '&signature=' + url_data['sig'][0]
 586                     elif 's' in url_data:
 587                         if self._downloader.params.get('verbose'):
 588                             s = url_data['s'][0]
 589                             if age_gate:
 590                                 player_version = self._search_regex(r'ad3-(.+?)\.swf',
 591                                     video_info['ad3_module'][0], 'flash player',
 592                                     fatal=False)
 593                                 player = 'flash player %s' % player_version
 594                             else:
 595                                 player = u'html5 player %s' % self._search_regex(r'html5player-(.+?)\.js', video_webpage,
 596                                     'html5 player', fatal=False)
 597                             parts_sizes = u'.'.join(compat_str(len(part)) for part in s.split('.'))
 598                             self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
 599                                 (len(s), parts_sizes, url_data['itag'][0], player))
 600                         signature = self._decrypt_signature(url_data['s'][0])
 601                         url += '&signature=' + signature
 602                     if 'ratebypass' not in url:
 603                         url += '&ratebypass=yes'
 604                     url_map[url_data['itag'][0]] = url
 605
 606             format_limit = self._downloader.params.get('format_limit', None)
 607             available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats
 608             if format_limit is not None and format_limit in available_formats:
 609                 format_list = available_formats[available_formats.index(format_limit):]
 610             else:
 611                 format_list = available_formats
 612             existing_formats = [x for x in format_list if x in url_map]
 613             if len(existing_formats) == 0:
 614                 raise ExtractorError(u'no known formats available for video')
 615             if self._downloader.params.get('listformats', None):
 616                 self._print_formats(existing_formats)
 617                 return
 618             if req_format is None or req_format == 'best':
 619                 video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality
 620             elif req_format == 'worst':
 621                 video_url_list = [(existing_formats[-1], url_map[existing_formats[-1]])] # worst quality
 622             elif req_format in ('-1', 'all'):
 623                 video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
 624             else:
 625                 # Specific formats. We pick the first in a slash-delimeted sequence.
 626                 # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'.
 627                 req_formats = req_format.split('/')
 628                 video_url_list = None
 629                 for rf in req_formats:
 630                     if rf in url_map:
 631                         video_url_list = [(rf, url_map[rf])]
 632                         break
 633                 if video_url_list is None:
 634                     raise ExtractorError(u'requested format not available')
 635         else:
 636             raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info')
 637
 638         results = []
 639         for format_param, video_real_url in video_url_list:
 640             # Extension
 641             video_extension = self._video_extensions.get(format_param, 'flv')
 642
 643             video_format = '{0} - {1}'.format(format_param if format_param else video_extension,
 644                                               self._video_dimensions.get(format_param, '???'))
 645
 646             results.append({
 647                 'id':       video_id,
 648                 'url':      video_real_url,
 649                 'uploader': video_uploader,
 650                 'uploader_id': video_uploader_id,
 651                 'upload_date':  upload_date,
 652                 'title':    video_title,
 653                 'ext':      video_extension,
 654                 'format':   video_format,
 655                 'thumbnail':    video_thumbnail,
 656                 'description':  video_description,
 657                 'player_url':   player_url,
 658                 'subtitles':    video_subtitles,
 659                 'duration':     video_duration
 660             })
 661         return results
 662
 663 class YoutubePlaylistIE(InfoExtractor):
 664     IE_DESC = u'YouTube.com playlists'
 665     _VALID_URL = r"""(?:
 666                         (?:https?://)?
 667                         (?:\w+\.)?
 668                         youtube\.com/
 669                         (?:
 670                            (?:course|view_play_list|my_playlists|artist|playlist|watch)
 671                            \? (?:.*?&)*? (?:p|a|list)=
 672                         |  p/
 673                         )
 674                         ((?:PL|EC|UU)?[0-9A-Za-z-_]{10,})
 675                         .*
 676                      |
 677                         ((?:PL|EC|UU)[0-9A-Za-z-_]{10,})
 678                      )"""
 679     _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none'
 680     _MAX_RESULTS = 50
 681     IE_NAME = u'youtube:playlist'
 682
 683     @classmethod
 684     def suitable(cls, url):
 685         """Receives a URL and returns True if suitable for this IE."""
 686         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
 687
 688     def _real_extract(self, url):
 689         # Extract playlist id
 690         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 691         if mobj is None:
 692             raise ExtractorError(u'Invalid URL: %s' % url)
 693
 694         # Download playlist videos from API
 695         playlist_id = mobj.group(1) or mobj.group(2)
 696         page_num = 1
 697         videos = []
 698
 699         while True:
 700             url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, self._MAX_RESULTS * (page_num - 1) + 1)
 701             page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
 702
 703             try:
 704                 response = json.loads(page)
 705             except ValueError as err:
 706                 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
 707
 708             if 'feed' not in response:
 709                 raise ExtractorError(u'Got a malformed response from YouTube API')
 710             playlist_title = response['feed']['title']['$t']
 711             if 'entry' not in response['feed']:
 712                 # Number of videos is a multiple of self._MAX_RESULTS
 713                 break
 714
 715             for entry in response['feed']['entry']:
 716                 index = entry['yt$position']['$t']
 717                 if 'media$group' in entry and 'media$player' in entry['media$group']:
 718                     videos.append((index, entry['media$group']['media$player']['url']))
 719
 720             if len(response['feed']['entry']) < self._MAX_RESULTS:
 721                 break
 722             page_num += 1
 723
 724         videos = [v[1] for v in sorted(videos)]
 725
 726         url_results = [self.url_result(vurl, 'Youtube') for vurl in videos]
 727         return [self.playlist_result(url_results, playlist_id, playlist_title)]
 728
 729
 730 class YoutubeChannelIE(InfoExtractor):
 731     IE_DESC = u'YouTube.com channels'
 732     _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
 733     _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en'
 734     _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
 735     _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'
 736     IE_NAME = u'youtube:channel'
 737
 738     def extract_videos_from_page(self, page):
 739         ids_in_page = []
 740         for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
 741             if mobj.group(1) not in ids_in_page:
 742                 ids_in_page.append(mobj.group(1))
 743         return ids_in_page
 744
 745     def _real_extract(self, url):
 746         # Extract channel id
 747         mobj = re.match(self._VALID_URL, url)
 748         if mobj is None:
 749             raise ExtractorError(u'Invalid URL: %s' % url)
 750
 751         # Download channel page
 752         channel_id = mobj.group(1)
 753         video_ids = []
 754         pagenum = 1
 755
 756         url = self._TEMPLATE_URL % (channel_id, pagenum)
 757         page = self._download_webpage(url, channel_id,
 758                                       u'Downloading page #%s' % pagenum)
 759
 760         # Extract video identifiers
 761         ids_in_page = self.extract_videos_from_page(page)
 762         video_ids.extend(ids_in_page)
 763
 764         # Download any subsequent channel pages using the json-based channel_ajax query
 765         if self._MORE_PAGES_INDICATOR in page:
 766             while True:
 767                 pagenum = pagenum + 1
 768
 769                 url = self._MORE_PAGES_URL % (pagenum, channel_id)
 770                 page = self._download_webpage(url, channel_id,
 771                                               u'Downloading page #%s' % pagenum)
 772
 773                 page = json.loads(page)
 774
 775                 ids_in_page = self.extract_videos_from_page(page['content_html'])
 776                 video_ids.extend(ids_in_page)
 777
 778                 if self._MORE_PAGES_INDICATOR  not in page['load_more_widget_html']:
 779                     break
 780
 781         self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
 782
 783         urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids]
 784         url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls]
 785         return [self.playlist_result(url_entries, channel_id)]
 786
 787
 788 class YoutubeUserIE(InfoExtractor):
 789     IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
 790     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
 791     _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
 792     _GDATA_PAGE_SIZE = 50
 793     _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
 794     _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
 795     IE_NAME = u'youtube:user'
 796
 797     def _real_extract(self, url):
 798         # Extract username
 799         mobj = re.match(self._VALID_URL, url)
 800         if mobj is None:
 801             raise ExtractorError(u'Invalid URL: %s' % url)
 802
 803         username = mobj.group(1)
 804
 805         # Download video ids using YouTube Data API. Result size per
 806         # query is limited (currently to 50 videos) so we need to query
 807         # page by page until there are no video ids - it means we got
 808         # all of them.
 809
 810         video_ids = []
 811         pagenum = 0
 812
 813         while True:
 814             start_index = pagenum * self._GDATA_PAGE_SIZE + 1
 815
 816             gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
 817             page = self._download_webpage(gdata_url, username,
 818                                           u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
 819
 820             # Extract video identifiers
 821             ids_in_page = []
 822
 823             for mobj in re.finditer(self._VIDEO_INDICATOR, page):
 824                 if mobj.group(1) not in ids_in_page:
 825                     ids_in_page.append(mobj.group(1))
 826
 827             video_ids.extend(ids_in_page)
 828
 829             # A little optimization - if current page is not
 830             # "full", ie. does not contain PAGE_SIZE video ids then
 831             # we can assume that this page is the last one - there
 832             # are no more ids on further pages - no need to query
 833             # again.
 834
 835             if len(ids_in_page) < self._GDATA_PAGE_SIZE:
 836                 break
 837
 838             pagenum += 1
 839
 840         urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
 841         url_results = [self.url_result(rurl, 'Youtube') for rurl in urls]
 842         return [self.playlist_result(url_results, playlist_title = username)]
 843
 844 class YoutubeSearchIE(SearchInfoExtractor):
 845     IE_DESC = u'YouTube.com searches'
 846     _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
 847     _MAX_RESULTS = 1000
 848     IE_NAME = u'youtube:search'
 849     _SEARCH_KEY = 'ytsearch'
 850
 851     def report_download_page(self, query, pagenum):
 852         """Report attempt to download search page with given number."""
 853         self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
 854
 855     def _get_n_results(self, query, n):
 856         """Get a specified number of results for a query"""
 857
 858         video_ids = []
 859         pagenum = 0
 860         limit = n
 861
 862         while (50 * pagenum) < limit:
 863             self.report_download_page(query, pagenum+1)
 864             result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
 865             request = compat_urllib_request.Request(result_url)
 866             try:
 867                 data = compat_urllib_request.urlopen(request).read().decode('utf-8')
 868             except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 869                 raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
 870             api_response = json.loads(data)['data']
 871
 872             if not 'items' in api_response:
 873                 raise ExtractorError(u'[youtube] No video results')
 874
 875             new_ids = list(video['id'] for video in api_response['items'])
 876             video_ids += new_ids
 877
 878             limit = min(n, api_response['totalItems'])
 879             pagenum += 1
 880
 881         if len(video_ids) > n:
 882             video_ids = video_ids[:n]
 883         videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
 884         return self.playlist_result(videos, query)
 885
 886
 887 class YoutubeShowIE(InfoExtractor):
 888     IE_DESC = u'YouTube.com (multi-season) shows'
 889     _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
 890     IE_NAME = u'youtube:show'
 891
 892     def _real_extract(self, url):
 893         mobj = re.match(self._VALID_URL, url)
 894         show_name = mobj.group(1)
 895         webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
 896         # There's one playlist for each season of the show
 897         m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
 898         self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
 899         return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
 900
 901
 902 class YoutubeSubscriptionsIE(YoutubeIE):
 903     """It's a subclass of YoutubeIE because we need to login"""
 904     IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
 905     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
 906     IE_NAME = u'youtube:subscriptions'
 907     _FEED_TEMPLATE = 'http://www.youtube.com/feed_ajax?action_load_system_feed=1&feed_name=subscriptions&paging=%s'
 908     _PAGING_STEP = 30
 909
 910     # Overwrite YoutubeIE properties we don't want
 911     _TESTS = []
 912     @classmethod
 913     def suitable(cls, url):
 914         return re.match(cls._VALID_URL, url) is not None
 915
 916     def _real_initialize(self):
 917         (username, password) = self._get_login_info()
 918         if username is None:
 919             raise ExtractorError(u'No login info available, needed for downloading the Youtube subscriptions.', expected=True)
 920         super(YoutubeSubscriptionsIE, self)._real_initialize()
 921
 922     def _real_extract(self, url):
 923         feed_entries = []
 924         # The step argument is available only in 2.7 or higher
 925         for i in itertools.count(0):
 926             paging = i*self._PAGING_STEP
 927             info = self._download_webpage(self._FEED_TEMPLATE % paging, 'feed',
 928                                           u'Downloading page %s' % i)
 929             info = json.loads(info)
 930             feed_html = info['feed_html']
 931             m_ids = re.finditer(r'"/watch\?v=(.*?)"', feed_html)
 932             ids = orderedSet(m.group(1) for m in m_ids)
 933             feed_entries.extend(self.url_result(id, 'Youtube') for id in ids)
 934             if info['paging'] is None:
 935                 break
 936         return self.playlist_result(feed_entries, playlist_title='Youtube Subscriptions')