_ Git - youtube-dl/blob - youtube_dl/extractor/dailymotion.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import base64
   5 import functools
   6 import hashlib
   7 import itertools
   8 import json
   9 import random
  10 import re
  11 import string
  12
  13 from .common import InfoExtractor
  14 from ..compat import compat_struct_pack
  15 from ..utils import (
  16     determine_ext,
  17     error_to_compat_str,
  18     ExtractorError,
  19     int_or_none,
  20     mimetype2ext,
  21     OnDemandPagedList,
  22     parse_iso8601,
  23     sanitized_Request,
  24     str_to_int,
  25     try_get,
  26     unescapeHTML,
  27     update_url_query,
  28     url_or_none,
  29     urlencode_postdata,
  30 )
  31
  32
  33 class DailymotionBaseInfoExtractor(InfoExtractor):
  34     @staticmethod
  35     def _build_request(url):
  36         """Build a request with the family filter disabled"""
  37         request = sanitized_Request(url)
  38         request.add_header('Cookie', 'family_filter=off; ff=off')
  39         return request
  40
  41     def _download_webpage_handle_no_ff(self, url, *args, **kwargs):
  42         request = self._build_request(url)
  43         return self._download_webpage_handle(request, *args, **kwargs)
  44
  45     def _download_webpage_no_ff(self, url, *args, **kwargs):
  46         request = self._build_request(url)
  47         return self._download_webpage(request, *args, **kwargs)
  48
  49
  50 class DailymotionIE(DailymotionBaseInfoExtractor):
  51     _VALID_URL = r'''(?ix)
  52                     https?://
  53                         (?:
  54                             (?:(?:www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(?:(?:embed|swf|\#)/)?video|swf)|
  55                             (?:www\.)?lequipe\.fr/video
  56                         )
  57                         /(?P<id>[^/?_]+)
  58                     '''
  59     IE_NAME = 'dailymotion'
  60
  61     _FORMATS = [
  62         ('stream_h264_ld_url', 'ld'),
  63         ('stream_h264_url', 'standard'),
  64         ('stream_h264_hq_url', 'hq'),
  65         ('stream_h264_hd_url', 'hd'),
  66         ('stream_h264_hd1080_url', 'hd180'),
  67     ]
  68
  69     _TESTS = [{
  70         'url': 'http://www.dailymotion.com/video/x5kesuj_office-christmas-party-review-jason-bateman-olivia-munn-t-j-miller_news',
  71         'md5': '074b95bdee76b9e3654137aee9c79dfe',
  72         'info_dict': {
  73             'id': 'x5kesuj',
  74             'ext': 'mp4',
  75             'title': 'Office Christmas Party Review –  Jason Bateman, Olivia Munn, T.J. Miller',
  76             'description': 'Office Christmas Party Review -  Jason Bateman, Olivia Munn, T.J. Miller',
  77             'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  78             'duration': 187,
  79             'timestamp': 1493651285,
  80             'upload_date': '20170501',
  81             'uploader': 'Deadline',
  82             'uploader_id': 'x1xm8ri',
  83             'age_limit': 0,
  84         },
  85     }, {
  86         'url': 'https://www.dailymotion.com/video/x2iuewm_steam-machine-models-pricing-listed-on-steam-store-ign-news_videogames',
  87         'md5': '2137c41a8e78554bb09225b8eb322406',
  88         'info_dict': {
  89             'id': 'x2iuewm',
  90             'ext': 'mp4',
  91             'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News',
  92             'description': 'Several come bundled with the Steam Controller.',
  93             'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  94             'duration': 74,
  95             'timestamp': 1425657362,
  96             'upload_date': '20150306',
  97             'uploader': 'IGN',
  98             'uploader_id': 'xijv66',
  99             'age_limit': 0,
 100             'view_count': int,
 101         },
 102         'skip': 'video gone',
 103     }, {
 104         # Vevo video
 105         'url': 'http://www.dailymotion.com/video/x149uew_katy-perry-roar-official_musi',
 106         'info_dict': {
 107             'title': 'Roar (Official)',
 108             'id': 'USUV71301934',
 109             'ext': 'mp4',
 110             'uploader': 'Katy Perry',
 111             'upload_date': '20130905',
 112         },
 113         'params': {
 114             'skip_download': True,
 115         },
 116         'skip': 'VEVO is only available in some countries',
 117     }, {
 118         # age-restricted video
 119         'url': 'http://www.dailymotion.com/video/xyh2zz_leanna-decker-cyber-girl-of-the-year-desires-nude-playboy-plus_redband',
 120         'md5': '0d667a7b9cebecc3c89ee93099c4159d',
 121         'info_dict': {
 122             'id': 'xyh2zz',
 123             'ext': 'mp4',
 124             'title': 'Leanna Decker - Cyber Girl Of The Year Desires Nude [Playboy Plus]',
 125             'uploader': 'HotWaves1012',
 126             'age_limit': 18,
 127         },
 128         'skip': 'video gone',
 129     }, {
 130         # geo-restricted, player v5
 131         'url': 'http://www.dailymotion.com/video/xhza0o',
 132         'only_matching': True,
 133     }, {
 134         # with subtitles
 135         'url': 'http://www.dailymotion.com/video/x20su5f_the-power-of-nightmares-1-the-rise-of-the-politics-of-fear-bbc-2004_news',
 136         'only_matching': True,
 137     }, {
 138         'url': 'http://www.dailymotion.com/swf/video/x3n92nf',
 139         'only_matching': True,
 140     }, {
 141         'url': 'http://www.dailymotion.com/swf/x3ss1m_funny-magic-trick-barry-and-stuart_fun',
 142         'only_matching': True,
 143     }, {
 144         'url': 'https://www.lequipe.fr/video/x791mem',
 145         'only_matching': True,
 146     }, {
 147         'url': 'https://www.lequipe.fr/video/k7MtHciueyTcrFtFKA2',
 148         'only_matching': True,
 149     }]
 150
 151     @staticmethod
 152     def _extract_urls(webpage):
 153         urls = []
 154         # Look for embedded Dailymotion player
 155         # https://developer.dailymotion.com/player#player-parameters
 156         for mobj in re.finditer(
 157                 r'<(?:(?:embed|iframe)[^>]+?src=|input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=)(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/(?:embed|swf)/video/.+?)\1', webpage):
 158             urls.append(unescapeHTML(mobj.group('url')))
 159         for mobj in re.finditer(
 160                 r'(?s)DM\.player\([^,]+,\s*{.*?video[\'"]?\s*:\s*["\']?(?P<id>[0-9a-zA-Z]+).+?}\s*\);', webpage):
 161             urls.append('https://www.dailymotion.com/embed/video/' + mobj.group('id'))
 162         return urls
 163
 164     def _real_extract(self, url):
 165         video_id = self._match_id(url)
 166
 167         webpage = self._download_webpage_no_ff(
 168             'https://www.dailymotion.com/video/%s' % video_id, video_id)
 169
 170         age_limit = self._rta_search(webpage)
 171
 172         description = self._og_search_description(
 173             webpage, default=None) or self._html_search_meta(
 174             'description', webpage, 'description')
 175
 176         view_count_str = self._search_regex(
 177             (r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserPlays:([\s\d,.]+)"',
 178              r'video_views_count[^>]+>\s+([\s\d\,.]+)'),
 179             webpage, 'view count', default=None)
 180         if view_count_str:
 181             view_count_str = re.sub(r'\s', '', view_count_str)
 182         view_count = str_to_int(view_count_str)
 183         comment_count = int_or_none(self._search_regex(
 184             r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserComments:(\d+)"',
 185             webpage, 'comment count', default=None))
 186
 187         player_v5 = self._search_regex(
 188             [r'buildPlayer\(({.+?})\);\n',  # See https://github.com/ytdl-org/youtube-dl/issues/7826
 189              r'playerV5\s*=\s*dmp\.create\([^,]+?,\s*({.+?})\);',
 190              r'buildPlayer\(({.+?})\);',
 191              r'var\s+config\s*=\s*({.+?});',
 192              # New layout regex (see https://github.com/ytdl-org/youtube-dl/issues/13580)
 193              r'__PLAYER_CONFIG__\s*=\s*({.+?});'],
 194             webpage, 'player v5', default=None)
 195         if player_v5:
 196             player = self._parse_json(player_v5, video_id, fatal=False) or {}
 197             metadata = try_get(player, lambda x: x['metadata'], dict)
 198             if not metadata:
 199                 metadata_url = url_or_none(try_get(
 200                     player, lambda x: x['context']['metadata_template_url1']))
 201                 if metadata_url:
 202                     metadata_url = metadata_url.replace(':videoId', video_id)
 203                 else:
 204                     metadata_url = update_url_query(
 205                         'https://www.dailymotion.com/player/metadata/video/%s'
 206                         % video_id, {
 207                             'embedder': url,
 208                             'integration': 'inline',
 209                             'GK_PV5_NEON': '1',
 210                         })
 211                 metadata = self._download_json(
 212                     metadata_url, video_id, 'Downloading metadata JSON')
 213
 214             if try_get(metadata, lambda x: x['error']['type']) == 'password_protected':
 215                 password = self._downloader.params.get('videopassword')
 216                 if password:
 217                     r = int(metadata['id'][1:], 36)
 218                     us64e = lambda x: base64.urlsafe_b64encode(x).decode().strip('=')
 219                     t = ''.join(random.choice(string.ascii_letters) for i in range(10))
 220                     n = us64e(compat_struct_pack('I', r))
 221                     i = us64e(hashlib.md5(('%s%d%s' % (password, r, t)).encode()).digest())
 222                     metadata = self._download_json(
 223                         'http://www.dailymotion.com/player/metadata/video/p' + i + t + n, video_id)
 224
 225             self._check_error(metadata)
 226
 227             formats = []
 228             for quality, media_list in metadata['qualities'].items():
 229                 for media in media_list:
 230                     media_url = media.get('url')
 231                     if not media_url:
 232                         continue
 233                     type_ = media.get('type')
 234                     if type_ == 'application/vnd.lumberjack.manifest':
 235                         continue
 236                     ext = mimetype2ext(type_) or determine_ext(media_url)
 237                     if ext == 'm3u8':
 238                         m3u8_formats = self._extract_m3u8_formats(
 239                             media_url, video_id, 'mp4', preference=-1,
 240                             m3u8_id='hls', fatal=False)
 241                         for f in m3u8_formats:
 242                             f['url'] = f['url'].split('#')[0]
 243                             formats.append(f)
 244                     elif ext == 'f4m':
 245                         formats.extend(self._extract_f4m_formats(
 246                             media_url, video_id, preference=-1, f4m_id='hds', fatal=False))
 247                     else:
 248                         f = {
 249                             'url': media_url,
 250                             'format_id': 'http-%s' % quality,
 251                             'ext': ext,
 252                         }
 253                         m = re.search(r'H264-(?P<width>\d+)x(?P<height>\d+)', media_url)
 254                         if m:
 255                             f.update({
 256                                 'width': int(m.group('width')),
 257                                 'height': int(m.group('height')),
 258                             })
 259                         formats.append(f)
 260             self._sort_formats(formats)
 261
 262             title = metadata['title']
 263             duration = int_or_none(metadata.get('duration'))
 264             timestamp = int_or_none(metadata.get('created_time'))
 265             thumbnail = metadata.get('poster_url')
 266             uploader = metadata.get('owner', {}).get('screenname')
 267             uploader_id = metadata.get('owner', {}).get('id')
 268
 269             subtitles = {}
 270             subtitles_data = metadata.get('subtitles', {}).get('data', {})
 271             if subtitles_data and isinstance(subtitles_data, dict):
 272                 for subtitle_lang, subtitle in subtitles_data.items():
 273                     subtitles[subtitle_lang] = [{
 274                         'ext': determine_ext(subtitle_url),
 275                         'url': subtitle_url,
 276                     } for subtitle_url in subtitle.get('urls', [])]
 277
 278             return {
 279                 'id': video_id,
 280                 'title': title,
 281                 'description': description,
 282                 'thumbnail': thumbnail,
 283                 'duration': duration,
 284                 'timestamp': timestamp,
 285                 'uploader': uploader,
 286                 'uploader_id': uploader_id,
 287                 'age_limit': age_limit,
 288                 'view_count': view_count,
 289                 'comment_count': comment_count,
 290                 'formats': formats,
 291                 'subtitles': subtitles,
 292             }
 293
 294         # vevo embed
 295         vevo_id = self._search_regex(
 296             r'<link rel="video_src" href="[^"]*?vevo\.com[^"]*?video=(?P<id>[\w]*)',
 297             webpage, 'vevo embed', default=None)
 298         if vevo_id:
 299             return self.url_result('vevo:%s' % vevo_id, 'Vevo')
 300
 301         # fallback old player
 302         embed_page = self._download_webpage_no_ff(
 303             'https://www.dailymotion.com/embed/video/%s' % video_id,
 304             video_id, 'Downloading embed page')
 305
 306         timestamp = parse_iso8601(self._html_search_meta(
 307             'video:release_date', webpage, 'upload date'))
 308
 309         info = self._parse_json(
 310             self._search_regex(
 311                 r'var info = ({.*?}),$', embed_page,
 312                 'video info', flags=re.MULTILINE),
 313             video_id)
 314
 315         self._check_error(info)
 316
 317         formats = []
 318         for (key, format_id) in self._FORMATS:
 319             video_url = info.get(key)
 320             if video_url is not None:
 321                 m_size = re.search(r'H264-(\d+)x(\d+)', video_url)
 322                 if m_size is not None:
 323                     width, height = map(int_or_none, (m_size.group(1), m_size.group(2)))
 324                 else:
 325                     width, height = None, None
 326                 formats.append({
 327                     'url': video_url,
 328                     'ext': 'mp4',
 329                     'format_id': format_id,
 330                     'width': width,
 331                     'height': height,
 332                 })
 333         self._sort_formats(formats)
 334
 335         # subtitles
 336         video_subtitles = self.extract_subtitles(video_id, webpage)
 337
 338         title = self._og_search_title(webpage, default=None)
 339         if title is None:
 340             title = self._html_search_regex(
 341                 r'(?s)<span\s+id="video_title"[^>]*>(.*?)</span>', webpage,
 342                 'title')
 343
 344         return {
 345             'id': video_id,
 346             'formats': formats,
 347             'uploader': info['owner.screenname'],
 348             'timestamp': timestamp,
 349             'title': title,
 350             'description': description,
 351             'subtitles': video_subtitles,
 352             'thumbnail': info['thumbnail_url'],
 353             'age_limit': age_limit,
 354             'view_count': view_count,
 355             'duration': info['duration']
 356         }
 357
 358     def _check_error(self, info):
 359         error = info.get('error')
 360         if error:
 361             title = error.get('title') or error['message']
 362             # See https://developer.dailymotion.com/api#access-error
 363             if error.get('code') == 'DM007':
 364                 self.raise_geo_restricted(msg=title)
 365             raise ExtractorError(
 366                 '%s said: %s' % (self.IE_NAME, title), expected=True)
 367
 368     def _get_subtitles(self, video_id, webpage):
 369         try:
 370             sub_list = self._download_webpage(
 371                 'https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id,
 372                 video_id, note=False)
 373         except ExtractorError as err:
 374             self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err))
 375             return {}
 376         info = json.loads(sub_list)
 377         if (info['total'] > 0):
 378             sub_lang_list = dict((l['language'], [{'url': l['url'], 'ext': 'srt'}]) for l in info['list'])
 379             return sub_lang_list
 380         self._downloader.report_warning('video doesn\'t have subtitles')
 381         return {}
 382
 383
 384 class DailymotionPlaylistIE(DailymotionBaseInfoExtractor):
 385     IE_NAME = 'dailymotion:playlist'
 386     _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P<id>x[0-9a-z]+)'
 387     _TESTS = [{
 388         'url': 'http://www.dailymotion.com/playlist/xv4bw_nqtv_sport/1#video=xl8v3q',
 389         'info_dict': {
 390             'title': 'SPORT',
 391             'id': 'xv4bw',
 392         },
 393         'playlist_mincount': 20,
 394     }]
 395     _PAGE_SIZE = 100
 396
 397     def _fetch_page(self, playlist_id, authorizaion, page):
 398         page += 1
 399         videos = self._download_json(
 400             'https://graphql.api.dailymotion.com',
 401             playlist_id, 'Downloading page %d' % page,
 402             data=json.dumps({
 403                 'query': '''{
 404   collection(xid: "%s") {
 405     videos(first: %d, page: %d) {
 406       pageInfo {
 407         hasNextPage
 408         nextPage
 409       }
 410       edges {
 411         node {
 412           xid
 413           url
 414         }
 415       }
 416     }
 417   }
 418 }''' % (playlist_id, self._PAGE_SIZE, page)
 419             }).encode(), headers={
 420                 'Authorization': authorizaion,
 421                 'Origin': 'https://www.dailymotion.com',
 422             })['data']['collection']['videos']
 423         for edge in videos['edges']:
 424             node = edge['node']
 425             yield self.url_result(
 426                 node['url'], DailymotionIE.ie_key(), node['xid'])
 427
 428     def _real_extract(self, url):
 429         playlist_id = self._match_id(url)
 430         webpage = self._download_webpage(url, playlist_id)
 431         api = self._parse_json(self._search_regex(
 432             r'__PLAYER_CONFIG__\s*=\s*({.+?});',
 433             webpage, 'player config'), playlist_id)['context']['api']
 434         auth = self._download_json(
 435             api.get('auth_url', 'https://graphql.api.dailymotion.com/oauth/token'),
 436             playlist_id, data=urlencode_postdata({
 437                 'client_id': api.get('client_id', 'f1a362d288c1b98099c7'),
 438                 'client_secret': api.get('client_secret', 'eea605b96e01c796ff369935357eca920c5da4c5'),
 439                 'grant_type': 'client_credentials',
 440             }))
 441         authorizaion = '%s %s' % (auth.get('token_type', 'Bearer'), auth['access_token'])
 442         entries = OnDemandPagedList(functools.partial(
 443             self._fetch_page, playlist_id, authorizaion), self._PAGE_SIZE)
 444         return self.playlist_result(
 445             entries, playlist_id,
 446             self._og_search_title(webpage))
 447
 448
 449 class DailymotionUserIE(DailymotionBaseInfoExtractor):
 450     IE_NAME = 'dailymotion:user'
 451     _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|swf|#|video|playlist)/)(?:(?:old/)?user/)?(?P<user>[^/]+)'
 452     _MORE_PAGES_INDICATOR = r'(?s)<div class="pages[^"]*">.*?<a\s+class="[^"]*?icon-arrow_right[^"]*?"'
 453     _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s'
 454     _TESTS = [{
 455         'url': 'https://www.dailymotion.com/user/nqtv',
 456         'info_dict': {
 457             'id': 'nqtv',
 458             'title': 'Rémi Gaillard',
 459         },
 460         'playlist_mincount': 100,
 461     }, {
 462         'url': 'http://www.dailymotion.com/user/UnderProject',
 463         'info_dict': {
 464             'id': 'UnderProject',
 465             'title': 'UnderProject',
 466         },
 467         'playlist_mincount': 1800,
 468         'expected_warnings': [
 469             'Stopped at duplicated page',
 470         ],
 471         'skip': 'Takes too long time',
 472     }]
 473
 474     def _extract_entries(self, id):
 475         video_ids = set()
 476         processed_urls = set()
 477         for pagenum in itertools.count(1):
 478             page_url = self._PAGE_TEMPLATE % (id, pagenum)
 479             webpage, urlh = self._download_webpage_handle_no_ff(
 480                 page_url, id, 'Downloading page %s' % pagenum)
 481             if urlh.geturl() in processed_urls:
 482                 self.report_warning('Stopped at duplicated page %s, which is the same as %s' % (
 483                     page_url, urlh.geturl()), id)
 484                 break
 485
 486             processed_urls.add(urlh.geturl())
 487
 488             for video_id in re.findall(r'data-xid="(.+?)"', webpage):
 489                 if video_id not in video_ids:
 490                     yield self.url_result(
 491                         'http://www.dailymotion.com/video/%s' % video_id,
 492                         DailymotionIE.ie_key(), video_id)
 493                     video_ids.add(video_id)
 494
 495             if re.search(self._MORE_PAGES_INDICATOR, webpage) is None:
 496                 break
 497
 498     def _real_extract(self, url):
 499         mobj = re.match(self._VALID_URL, url)
 500         user = mobj.group('user')
 501         webpage = self._download_webpage(
 502             'https://www.dailymotion.com/user/%s' % user, user)
 503         full_user = unescapeHTML(self._html_search_regex(
 504             r'<a class="nav-image" title="([^"]+)" href="/%s">' % re.escape(user),
 505             webpage, 'user'))
 506
 507         return {
 508             '_type': 'playlist',
 509             'id': user,
 510             'title': full_user,
 511             'entries': self._extract_entries(user),
 512         }