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