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