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         # Look for embedded Dailymotion player
 141         matches = re.findall(
 142             r'<(?:(?:embed|iframe)[^>]+?src=|input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=)(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/(?:embed|swf)/video/.+?)\1', webpage)
 143         return list(map(lambda m: unescapeHTML(m[1]), matches))
 144
 145     def _real_extract(self, url):
 146         video_id = self._match_id(url)
 147
 148         webpage = self._download_webpage_no_ff(
 149             'https://www.dailymotion.com/video/%s' % video_id, video_id)
 150
 151         age_limit = self._rta_search(webpage)
 152
 153         description = self._og_search_description(
 154             webpage, default=None) or self._html_search_meta(
 155             'description', webpage, 'description')
 156
 157         view_count_str = self._search_regex(
 158             (r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserPlays:([\s\d,.]+)"',
 159              r'video_views_count[^>]+>\s+([\s\d\,.]+)'),
 160             webpage, 'view count', default=None)
 161         if view_count_str:
 162             view_count_str = re.sub(r'\s', '', view_count_str)
 163         view_count = str_to_int(view_count_str)
 164         comment_count = int_or_none(self._search_regex(
 165             r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserComments:(\d+)"',
 166             webpage, 'comment count', default=None))
 167
 168         player_v5 = self._search_regex(
 169             [r'buildPlayer\(({.+?})\);\n',  # See https://github.com/ytdl-org/youtube-dl/issues/7826
 170              r'playerV5\s*=\s*dmp\.create\([^,]+?,\s*({.+?})\);',
 171              r'buildPlayer\(({.+?})\);',
 172              r'var\s+config\s*=\s*({.+?});',
 173              # New layout regex (see https://github.com/ytdl-org/youtube-dl/issues/13580)
 174              r'__PLAYER_CONFIG__\s*=\s*({.+?});'],
 175             webpage, 'player v5', default=None)
 176         if player_v5:
 177             player = self._parse_json(player_v5, video_id, fatal=False) or {}
 178             metadata = try_get(player, lambda x: x['metadata'], dict)
 179             if not metadata:
 180                 metadata_url = url_or_none(try_get(
 181                     player, lambda x: x['context']['metadata_template_url1']))
 182                 if metadata_url:
 183                     metadata_url = metadata_url.replace(':videoId', video_id)
 184                 else:
 185                     metadata_url = update_url_query(
 186                         'https://www.dailymotion.com/player/metadata/video/%s'
 187                         % video_id, {
 188                             'embedder': url,
 189                             'integration': 'inline',
 190                             'GK_PV5_NEON': '1',
 191                         })
 192                 metadata = self._download_json(
 193                     metadata_url, video_id, 'Downloading metadata JSON')
 194
 195             if try_get(metadata, lambda x: x['error']['type']) == 'password_protected':
 196                 password = self._downloader.params.get('videopassword')
 197                 if password:
 198                     r = int(metadata['id'][1:], 36)
 199                     us64e = lambda x: base64.urlsafe_b64encode(x).decode().strip('=')
 200                     t = ''.join(random.choice(string.ascii_letters) for i in range(10))
 201                     n = us64e(compat_struct_pack('I', r))
 202                     i = us64e(hashlib.md5(('%s%d%s' % (password, r, t)).encode()).digest())
 203                     metadata = self._download_json(
 204                         'http://www.dailymotion.com/player/metadata/video/p' + i + t + n, video_id)
 205
 206             self._check_error(metadata)
 207
 208             formats = []
 209             for quality, media_list in metadata['qualities'].items():
 210                 for media in media_list:
 211                     media_url = media.get('url')
 212                     if not media_url:
 213                         continue
 214                     type_ = media.get('type')
 215                     if type_ == 'application/vnd.lumberjack.manifest':
 216                         continue
 217                     ext = mimetype2ext(type_) or determine_ext(media_url)
 218                     if ext == 'm3u8':
 219                         m3u8_formats = self._extract_m3u8_formats(
 220                             media_url, video_id, 'mp4', preference=-1,
 221                             m3u8_id='hls', fatal=False)
 222                         for f in m3u8_formats:
 223                             f['url'] = f['url'].split('#')[0]
 224                             formats.append(f)
 225                     elif ext == 'f4m':
 226                         formats.extend(self._extract_f4m_formats(
 227                             media_url, video_id, preference=-1, f4m_id='hds', fatal=False))
 228                     else:
 229                         f = {
 230                             'url': media_url,
 231                             'format_id': 'http-%s' % quality,
 232                             'ext': ext,
 233                         }
 234                         m = re.search(r'H264-(?P<width>\d+)x(?P<height>\d+)', media_url)
 235                         if m:
 236                             f.update({
 237                                 'width': int(m.group('width')),
 238                                 'height': int(m.group('height')),
 239                             })
 240                         formats.append(f)
 241             self._sort_formats(formats)
 242
 243             title = metadata['title']
 244             duration = int_or_none(metadata.get('duration'))
 245             timestamp = int_or_none(metadata.get('created_time'))
 246             thumbnail = metadata.get('poster_url')
 247             uploader = metadata.get('owner', {}).get('screenname')
 248             uploader_id = metadata.get('owner', {}).get('id')
 249
 250             subtitles = {}
 251             subtitles_data = metadata.get('subtitles', {}).get('data', {})
 252             if subtitles_data and isinstance(subtitles_data, dict):
 253                 for subtitle_lang, subtitle in subtitles_data.items():
 254                     subtitles[subtitle_lang] = [{
 255                         'ext': determine_ext(subtitle_url),
 256                         'url': subtitle_url,
 257                     } for subtitle_url in subtitle.get('urls', [])]
 258
 259             return {
 260                 'id': video_id,
 261                 'title': title,
 262                 'description': description,
 263                 'thumbnail': thumbnail,
 264                 'duration': duration,
 265                 'timestamp': timestamp,
 266                 'uploader': uploader,
 267                 'uploader_id': uploader_id,
 268                 'age_limit': age_limit,
 269                 'view_count': view_count,
 270                 'comment_count': comment_count,
 271                 'formats': formats,
 272                 'subtitles': subtitles,
 273             }
 274
 275         # vevo embed
 276         vevo_id = self._search_regex(
 277             r'<link rel="video_src" href="[^"]*?vevo\.com[^"]*?video=(?P<id>[\w]*)',
 278             webpage, 'vevo embed', default=None)
 279         if vevo_id:
 280             return self.url_result('vevo:%s' % vevo_id, 'Vevo')
 281
 282         # fallback old player
 283         embed_page = self._download_webpage_no_ff(
 284             'https://www.dailymotion.com/embed/video/%s' % video_id,
 285             video_id, 'Downloading embed page')
 286
 287         timestamp = parse_iso8601(self._html_search_meta(
 288             'video:release_date', webpage, 'upload date'))
 289
 290         info = self._parse_json(
 291             self._search_regex(
 292                 r'var info = ({.*?}),$', embed_page,
 293                 'video info', flags=re.MULTILINE),
 294             video_id)
 295
 296         self._check_error(info)
 297
 298         formats = []
 299         for (key, format_id) in self._FORMATS:
 300             video_url = info.get(key)
 301             if video_url is not None:
 302                 m_size = re.search(r'H264-(\d+)x(\d+)', video_url)
 303                 if m_size is not None:
 304                     width, height = map(int_or_none, (m_size.group(1), m_size.group(2)))
 305                 else:
 306                     width, height = None, None
 307                 formats.append({
 308                     'url': video_url,
 309                     'ext': 'mp4',
 310                     'format_id': format_id,
 311                     'width': width,
 312                     'height': height,
 313                 })
 314         self._sort_formats(formats)
 315
 316         # subtitles
 317         video_subtitles = self.extract_subtitles(video_id, webpage)
 318
 319         title = self._og_search_title(webpage, default=None)
 320         if title is None:
 321             title = self._html_search_regex(
 322                 r'(?s)<span\s+id="video_title"[^>]*>(.*?)</span>', webpage,
 323                 'title')
 324
 325         return {
 326             'id': video_id,
 327             'formats': formats,
 328             'uploader': info['owner.screenname'],
 329             'timestamp': timestamp,
 330             'title': title,
 331             'description': description,
 332             'subtitles': video_subtitles,
 333             'thumbnail': info['thumbnail_url'],
 334             'age_limit': age_limit,
 335             'view_count': view_count,
 336             'duration': info['duration']
 337         }
 338
 339     def _check_error(self, info):
 340         error = info.get('error')
 341         if error:
 342             title = error.get('title') or error['message']
 343             # See https://developer.dailymotion.com/api#access-error
 344             if error.get('code') == 'DM007':
 345                 self.raise_geo_restricted(msg=title)
 346             raise ExtractorError(
 347                 '%s said: %s' % (self.IE_NAME, title), expected=True)
 348
 349     def _get_subtitles(self, video_id, webpage):
 350         try:
 351             sub_list = self._download_webpage(
 352                 'https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id,
 353                 video_id, note=False)
 354         except ExtractorError as err:
 355             self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err))
 356             return {}
 357         info = json.loads(sub_list)
 358         if (info['total'] > 0):
 359             sub_lang_list = dict((l['language'], [{'url': l['url'], 'ext': 'srt'}]) for l in info['list'])
 360             return sub_lang_list
 361         self._downloader.report_warning('video doesn\'t have subtitles')
 362         return {}
 363
 364
 365 class DailymotionPlaylistIE(DailymotionBaseInfoExtractor):
 366     IE_NAME = 'dailymotion:playlist'
 367     _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P<id>x[0-9a-z]+)'
 368     _TESTS = [{
 369         'url': 'http://www.dailymotion.com/playlist/xv4bw_nqtv_sport/1#video=xl8v3q',
 370         'info_dict': {
 371             'title': 'SPORT',
 372             'id': 'xv4bw',
 373         },
 374         'playlist_mincount': 20,
 375     }]
 376     _PAGE_SIZE = 100
 377
 378     def _fetch_page(self, playlist_id, authorizaion, page):
 379         page += 1
 380         videos = self._download_json(
 381             'https://graphql.api.dailymotion.com',
 382             playlist_id, 'Downloading page %d' % page,
 383             data=json.dumps({
 384                 'query': '''{
 385   collection(xid: "%s") {
 386     videos(first: %d, page: %d) {
 387       pageInfo {
 388         hasNextPage
 389         nextPage
 390       }
 391       edges {
 392         node {
 393           xid
 394           url
 395         }
 396       }
 397     }
 398   }
 399 }''' % (playlist_id, self._PAGE_SIZE, page)
 400             }).encode(), headers={
 401                 'Authorization': authorizaion,
 402                 'Origin': 'https://www.dailymotion.com',
 403             })['data']['collection']['videos']
 404         for edge in videos['edges']:
 405             node = edge['node']
 406             yield self.url_result(
 407                 node['url'], DailymotionIE.ie_key(), node['xid'])
 408
 409     def _real_extract(self, url):
 410         playlist_id = self._match_id(url)
 411         webpage = self._download_webpage(url, playlist_id)
 412         api = self._parse_json(self._search_regex(
 413             r'__PLAYER_CONFIG__\s*=\s*({.+?});',
 414             webpage, 'player config'), playlist_id)['context']['api']
 415         auth = self._download_json(
 416             api.get('auth_url', 'https://graphql.api.dailymotion.com/oauth/token'),
 417             playlist_id, data=urlencode_postdata({
 418                 'client_id': api.get('client_id', 'f1a362d288c1b98099c7'),
 419                 'client_secret': api.get('client_secret', 'eea605b96e01c796ff369935357eca920c5da4c5'),
 420                 'grant_type': 'client_credentials',
 421             }))
 422         authorizaion = '%s %s' % (auth.get('token_type', 'Bearer'), auth['access_token'])
 423         entries = OnDemandPagedList(functools.partial(
 424             self._fetch_page, playlist_id, authorizaion), self._PAGE_SIZE)
 425         return self.playlist_result(
 426             entries, playlist_id,
 427             self._og_search_title(webpage))
 428
 429
 430 class DailymotionUserIE(DailymotionBaseInfoExtractor):
 431     IE_NAME = 'dailymotion:user'
 432     _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|swf|#|video|playlist)/)(?:(?:old/)?user/)?(?P<user>[^/]+)'
 433     _MORE_PAGES_INDICATOR = r'(?s)<div class="pages[^"]*">.*?<a\s+class="[^"]*?icon-arrow_right[^"]*?"'
 434     _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s'
 435     _TESTS = [{
 436         'url': 'https://www.dailymotion.com/user/nqtv',
 437         'info_dict': {
 438             'id': 'nqtv',
 439             'title': 'Rémi Gaillard',
 440         },
 441         'playlist_mincount': 100,
 442     }, {
 443         'url': 'http://www.dailymotion.com/user/UnderProject',
 444         'info_dict': {
 445             'id': 'UnderProject',
 446             'title': 'UnderProject',
 447         },
 448         'playlist_mincount': 1800,
 449         'expected_warnings': [
 450             'Stopped at duplicated page',
 451         ],
 452         'skip': 'Takes too long time',
 453     }]
 454
 455     def _extract_entries(self, id):
 456         video_ids = set()
 457         processed_urls = set()
 458         for pagenum in itertools.count(1):
 459             page_url = self._PAGE_TEMPLATE % (id, pagenum)
 460             webpage, urlh = self._download_webpage_handle_no_ff(
 461                 page_url, id, 'Downloading page %s' % pagenum)
 462             if urlh.geturl() in processed_urls:
 463                 self.report_warning('Stopped at duplicated page %s, which is the same as %s' % (
 464                     page_url, urlh.geturl()), id)
 465                 break
 466
 467             processed_urls.add(urlh.geturl())
 468
 469             for video_id in re.findall(r'data-xid="(.+?)"', webpage):
 470                 if video_id not in video_ids:
 471                     yield self.url_result(
 472                         'http://www.dailymotion.com/video/%s' % video_id,
 473                         DailymotionIE.ie_key(), video_id)
 474                     video_ids.add(video_id)
 475
 476             if re.search(self._MORE_PAGES_INDICATOR, webpage) is None:
 477                 break
 478
 479     def _real_extract(self, url):
 480         mobj = re.match(self._VALID_URL, url)
 481         user = mobj.group('user')
 482         webpage = self._download_webpage(
 483             'https://www.dailymotion.com/user/%s' % user, user)
 484         full_user = unescapeHTML(self._html_search_regex(
 485             r'<a class="nav-image" title="([^"]+)" href="/%s">' % re.escape(user),
 486             webpage, 'user'))
 487
 488         return {
 489             '_type': 'playlist',
 490             'id': user,
 491             'title': full_user,
 492             'entries': self._extract_entries(user),
 493         }