_ 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     unescapeHTML,
  26     urlencode_postdata,
  27 )
  28
  29
  30 class DailymotionBaseInfoExtractor(InfoExtractor):
  31     @staticmethod
  32     def _build_request(url):
  33         """Build a request with the family filter disabled"""
  34         request = sanitized_Request(url)
  35         request.add_header('Cookie', 'family_filter=off; ff=off')
  36         return request
  37
  38     def _download_webpage_handle_no_ff(self, url, *args, **kwargs):
  39         request = self._build_request(url)
  40         return self._download_webpage_handle(request, *args, **kwargs)
  41
  42     def _download_webpage_no_ff(self, url, *args, **kwargs):
  43         request = self._build_request(url)
  44         return self._download_webpage(request, *args, **kwargs)
  45
  46
  47 class DailymotionIE(DailymotionBaseInfoExtractor):
  48     _VALID_URL = r'(?i)https?://(?:(www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(?:(?:embed|swf|#)/)?video|swf)/(?P<id>[^/?_]+)'
  49     IE_NAME = 'dailymotion'
  50
  51     _FORMATS = [
  52         ('stream_h264_ld_url', 'ld'),
  53         ('stream_h264_url', 'standard'),
  54         ('stream_h264_hq_url', 'hq'),
  55         ('stream_h264_hd_url', 'hd'),
  56         ('stream_h264_hd1080_url', 'hd180'),
  57     ]
  58
  59     _TESTS = [{
  60         'url': 'http://www.dailymotion.com/video/x5kesuj_office-christmas-party-review-jason-bateman-olivia-munn-t-j-miller_news',
  61         'md5': '074b95bdee76b9e3654137aee9c79dfe',
  62         'info_dict': {
  63             'id': 'x5kesuj',
  64             'ext': 'mp4',
  65             'title': 'Office Christmas Party Review –  Jason Bateman, Olivia Munn, T.J. Miller',
  66             'description': 'Office Christmas Party Review -  Jason Bateman, Olivia Munn, T.J. Miller',
  67             'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  68             'duration': 187,
  69             'timestamp': 1493651285,
  70             'upload_date': '20170501',
  71             'uploader': 'Deadline',
  72             'uploader_id': 'x1xm8ri',
  73             'age_limit': 0,
  74         },
  75     }, {
  76         'url': 'https://www.dailymotion.com/video/x2iuewm_steam-machine-models-pricing-listed-on-steam-store-ign-news_videogames',
  77         'md5': '2137c41a8e78554bb09225b8eb322406',
  78         'info_dict': {
  79             'id': 'x2iuewm',
  80             'ext': 'mp4',
  81             'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News',
  82             'description': 'Several come bundled with the Steam Controller.',
  83             'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  84             'duration': 74,
  85             'timestamp': 1425657362,
  86             'upload_date': '20150306',
  87             'uploader': 'IGN',
  88             'uploader_id': 'xijv66',
  89             'age_limit': 0,
  90             'view_count': int,
  91         },
  92         'skip': 'video gone',
  93     }, {
  94         # Vevo video
  95         'url': 'http://www.dailymotion.com/video/x149uew_katy-perry-roar-official_musi',
  96         'info_dict': {
  97             'title': 'Roar (Official)',
  98             'id': 'USUV71301934',
  99             'ext': 'mp4',
 100             'uploader': 'Katy Perry',
 101             'upload_date': '20130905',
 102         },
 103         'params': {
 104             'skip_download': True,
 105         },
 106         'skip': 'VEVO is only available in some countries',
 107     }, {
 108         # age-restricted video
 109         'url': 'http://www.dailymotion.com/video/xyh2zz_leanna-decker-cyber-girl-of-the-year-desires-nude-playboy-plus_redband',
 110         'md5': '0d667a7b9cebecc3c89ee93099c4159d',
 111         'info_dict': {
 112             'id': 'xyh2zz',
 113             'ext': 'mp4',
 114             'title': 'Leanna Decker - Cyber Girl Of The Year Desires Nude [Playboy Plus]',
 115             'uploader': 'HotWaves1012',
 116             'age_limit': 18,
 117         },
 118         'skip': 'video gone',
 119     }, {
 120         # geo-restricted, player v5
 121         'url': 'http://www.dailymotion.com/video/xhza0o',
 122         'only_matching': True,
 123     }, {
 124         # with subtitles
 125         'url': 'http://www.dailymotion.com/video/x20su5f_the-power-of-nightmares-1-the-rise-of-the-politics-of-fear-bbc-2004_news',
 126         'only_matching': True,
 127     }, {
 128         'url': 'http://www.dailymotion.com/swf/video/x3n92nf',
 129         'only_matching': True,
 130     }, {
 131         'url': 'http://www.dailymotion.com/swf/x3ss1m_funny-magic-trick-barry-and-stuart_fun',
 132         'only_matching': True,
 133     }]
 134
 135     @staticmethod
 136     def _extract_urls(webpage):
 137         # Look for embedded Dailymotion player
 138         matches = re.findall(
 139             r'<(?:(?:embed|iframe)[^>]+?src=|input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=)(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/(?:embed|swf)/video/.+?)\1', webpage)
 140         return list(map(lambda m: unescapeHTML(m[1]), matches))
 141
 142     def _real_extract(self, url):
 143         video_id = self._match_id(url)
 144
 145         webpage = self._download_webpage_no_ff(
 146             'https://www.dailymotion.com/video/%s' % video_id, video_id)
 147
 148         age_limit = self._rta_search(webpage)
 149
 150         description = self._og_search_description(
 151             webpage, default=None) or self._html_search_meta(
 152             'description', webpage, 'description')
 153
 154         view_count_str = self._search_regex(
 155             (r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserPlays:([\s\d,.]+)"',
 156              r'video_views_count[^>]+>\s+([\s\d\,.]+)'),
 157             webpage, 'view count', default=None)
 158         if view_count_str:
 159             view_count_str = re.sub(r'\s', '', view_count_str)
 160         view_count = str_to_int(view_count_str)
 161         comment_count = int_or_none(self._search_regex(
 162             r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserComments:(\d+)"',
 163             webpage, 'comment count', default=None))
 164
 165         player_v5 = self._search_regex(
 166             [r'buildPlayer\(({.+?})\);\n',  # See https://github.com/rg3/youtube-dl/issues/7826
 167              r'playerV5\s*=\s*dmp\.create\([^,]+?,\s*({.+?})\);',
 168              r'buildPlayer\(({.+?})\);',
 169              r'var\s+config\s*=\s*({.+?});',
 170              # New layout regex (see https://github.com/rg3/youtube-dl/issues/13580)
 171              r'__PLAYER_CONFIG__\s*=\s*({.+?});'],
 172             webpage, 'player v5', default=None)
 173         if player_v5:
 174             player = self._parse_json(player_v5, video_id)
 175             metadata = player['metadata']
 176
 177             if metadata.get('error', {}).get('type') == 'password_protected':
 178                 password = self._downloader.params.get('videopassword')
 179                 if password:
 180                     r = int(metadata['id'][1:], 36)
 181                     us64e = lambda x: base64.urlsafe_b64encode(x).decode().strip('=')
 182                     t = ''.join(random.choice(string.ascii_letters) for i in range(10))
 183                     n = us64e(compat_struct_pack('I', r))
 184                     i = us64e(hashlib.md5(('%s%d%s' % (password, r, t)).encode()).digest())
 185                     metadata = self._download_json(
 186                         'http://www.dailymotion.com/player/metadata/video/p' + i + t + n, video_id)
 187
 188             self._check_error(metadata)
 189
 190             formats = []
 191             for quality, media_list in metadata['qualities'].items():
 192                 for media in media_list:
 193                     media_url = media.get('url')
 194                     if not media_url:
 195                         continue
 196                     type_ = media.get('type')
 197                     if type_ == 'application/vnd.lumberjack.manifest':
 198                         continue
 199                     ext = mimetype2ext(type_) or determine_ext(media_url)
 200                     if ext == 'm3u8':
 201                         m3u8_formats = self._extract_m3u8_formats(
 202                             media_url, video_id, 'mp4', preference=-1,
 203                             m3u8_id='hls', fatal=False)
 204                         for f in m3u8_formats:
 205                             f['url'] = f['url'].split('#')[0]
 206                             formats.append(f)
 207                     elif ext == 'f4m':
 208                         formats.extend(self._extract_f4m_formats(
 209                             media_url, video_id, preference=-1, f4m_id='hds', fatal=False))
 210                     else:
 211                         f = {
 212                             'url': media_url,
 213                             'format_id': 'http-%s' % quality,
 214                             'ext': ext,
 215                         }
 216                         m = re.search(r'H264-(?P<width>\d+)x(?P<height>\d+)', media_url)
 217                         if m:
 218                             f.update({
 219                                 'width': int(m.group('width')),
 220                                 'height': int(m.group('height')),
 221                             })
 222                         formats.append(f)
 223             self._sort_formats(formats)
 224
 225             title = metadata['title']
 226             duration = int_or_none(metadata.get('duration'))
 227             timestamp = int_or_none(metadata.get('created_time'))
 228             thumbnail = metadata.get('poster_url')
 229             uploader = metadata.get('owner', {}).get('screenname')
 230             uploader_id = metadata.get('owner', {}).get('id')
 231
 232             subtitles = {}
 233             subtitles_data = metadata.get('subtitles', {}).get('data', {})
 234             if subtitles_data and isinstance(subtitles_data, dict):
 235                 for subtitle_lang, subtitle in subtitles_data.items():
 236                     subtitles[subtitle_lang] = [{
 237                         'ext': determine_ext(subtitle_url),
 238                         'url': subtitle_url,
 239                     } for subtitle_url in subtitle.get('urls', [])]
 240
 241             return {
 242                 'id': video_id,
 243                 'title': title,
 244                 'description': description,
 245                 'thumbnail': thumbnail,
 246                 'duration': duration,
 247                 'timestamp': timestamp,
 248                 'uploader': uploader,
 249                 'uploader_id': uploader_id,
 250                 'age_limit': age_limit,
 251                 'view_count': view_count,
 252                 'comment_count': comment_count,
 253                 'formats': formats,
 254                 'subtitles': subtitles,
 255             }
 256
 257         # vevo embed
 258         vevo_id = self._search_regex(
 259             r'<link rel="video_src" href="[^"]*?vevo\.com[^"]*?video=(?P<id>[\w]*)',
 260             webpage, 'vevo embed', default=None)
 261         if vevo_id:
 262             return self.url_result('vevo:%s' % vevo_id, 'Vevo')
 263
 264         # fallback old player
 265         embed_page = self._download_webpage_no_ff(
 266             'https://www.dailymotion.com/embed/video/%s' % video_id,
 267             video_id, 'Downloading embed page')
 268
 269         timestamp = parse_iso8601(self._html_search_meta(
 270             'video:release_date', webpage, 'upload date'))
 271
 272         info = self._parse_json(
 273             self._search_regex(
 274                 r'var info = ({.*?}),$', embed_page,
 275                 'video info', flags=re.MULTILINE),
 276             video_id)
 277
 278         self._check_error(info)
 279
 280         formats = []
 281         for (key, format_id) in self._FORMATS:
 282             video_url = info.get(key)
 283             if video_url is not None:
 284                 m_size = re.search(r'H264-(\d+)x(\d+)', video_url)
 285                 if m_size is not None:
 286                     width, height = map(int_or_none, (m_size.group(1), m_size.group(2)))
 287                 else:
 288                     width, height = None, None
 289                 formats.append({
 290                     'url': video_url,
 291                     'ext': 'mp4',
 292                     'format_id': format_id,
 293                     'width': width,
 294                     'height': height,
 295                 })
 296         self._sort_formats(formats)
 297
 298         # subtitles
 299         video_subtitles = self.extract_subtitles(video_id, webpage)
 300
 301         title = self._og_search_title(webpage, default=None)
 302         if title is None:
 303             title = self._html_search_regex(
 304                 r'(?s)<span\s+id="video_title"[^>]*>(.*?)</span>', webpage,
 305                 'title')
 306
 307         return {
 308             'id': video_id,
 309             'formats': formats,
 310             'uploader': info['owner.screenname'],
 311             'timestamp': timestamp,
 312             'title': title,
 313             'description': description,
 314             'subtitles': video_subtitles,
 315             'thumbnail': info['thumbnail_url'],
 316             'age_limit': age_limit,
 317             'view_count': view_count,
 318             'duration': info['duration']
 319         }
 320
 321     def _check_error(self, info):
 322         error = info.get('error')
 323         if error:
 324             title = error.get('title') or error['message']
 325             # See https://developer.dailymotion.com/api#access-error
 326             if error.get('code') == 'DM007':
 327                 self.raise_geo_restricted(msg=title)
 328             raise ExtractorError(
 329                 '%s said: %s' % (self.IE_NAME, title), expected=True)
 330
 331     def _get_subtitles(self, video_id, webpage):
 332         try:
 333             sub_list = self._download_webpage(
 334                 'https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id,
 335                 video_id, note=False)
 336         except ExtractorError as err:
 337             self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err))
 338             return {}
 339         info = json.loads(sub_list)
 340         if (info['total'] > 0):
 341             sub_lang_list = dict((l['language'], [{'url': l['url'], 'ext': 'srt'}]) for l in info['list'])
 342             return sub_lang_list
 343         self._downloader.report_warning('video doesn\'t have subtitles')
 344         return {}
 345
 346
 347 class DailymotionPlaylistIE(DailymotionBaseInfoExtractor):
 348     IE_NAME = 'dailymotion:playlist'
 349     _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P<id>x[0-9a-z]+)'
 350     _TESTS = [{
 351         'url': 'http://www.dailymotion.com/playlist/xv4bw_nqtv_sport/1#video=xl8v3q',
 352         'info_dict': {
 353             'title': 'SPORT',
 354             'id': 'xv4bw',
 355         },
 356         'playlist_mincount': 20,
 357     }]
 358     _PAGE_SIZE = 100
 359
 360     def _fetch_page(self, playlist_id, authorizaion, page):
 361         page += 1
 362         videos = self._download_json(
 363             'https://graphql.api.dailymotion.com',
 364             playlist_id, 'Downloading page %d' % page,
 365             data=json.dumps({
 366                 'query': '''{
 367   collection(xid: "%s") {
 368     videos(first: %d, page: %d) {
 369       pageInfo {
 370         hasNextPage
 371         nextPage
 372       }
 373       edges {
 374         node {
 375           xid
 376           url
 377         }
 378       }
 379     }
 380   }
 381 }''' % (playlist_id, self._PAGE_SIZE, page)
 382             }).encode(), headers={
 383                 'Authorization': authorizaion,
 384                 'Origin': 'https://www.dailymotion.com',
 385             })['data']['collection']['videos']
 386         for edge in videos['edges']:
 387             node = edge['node']
 388             yield self.url_result(
 389                 node['url'], DailymotionIE.ie_key(), node['xid'])
 390
 391     def _real_extract(self, url):
 392         playlist_id = self._match_id(url)
 393         webpage = self._download_webpage(url, playlist_id)
 394         api = self._parse_json(self._search_regex(
 395             r'__PLAYER_CONFIG__\s*=\s*({.+?});',
 396             webpage, 'player config'), playlist_id)['context']['api']
 397         auth = self._download_json(
 398             api.get('auth_url', 'https://graphql.api.dailymotion.com/oauth/token'),
 399             playlist_id, data=urlencode_postdata({
 400                 'client_id': api.get('client_id', 'f1a362d288c1b98099c7'),
 401                 'client_secret': api.get('client_secret', 'eea605b96e01c796ff369935357eca920c5da4c5'),
 402                 'grant_type': 'client_credentials',
 403             }))
 404         authorizaion = '%s %s' % (auth.get('token_type', 'Bearer'), auth['access_token'])
 405         entries = OnDemandPagedList(functools.partial(
 406             self._fetch_page, playlist_id, authorizaion), self._PAGE_SIZE)
 407         return self.playlist_result(
 408             entries, playlist_id,
 409             self._og_search_title(webpage))
 410
 411
 412 class DailymotionUserIE(DailymotionBaseInfoExtractor):
 413     IE_NAME = 'dailymotion:user'
 414     _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|swf|#|video|playlist)/)(?:(?:old/)?user/)?(?P<user>[^/]+)'
 415     _MORE_PAGES_INDICATOR = r'(?s)<div class="pages[^"]*">.*?<a\s+class="[^"]*?icon-arrow_right[^"]*?"'
 416     _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s'
 417     _TESTS = [{
 418         'url': 'https://www.dailymotion.com/user/nqtv',
 419         'info_dict': {
 420             'id': 'nqtv',
 421             'title': 'Rémi Gaillard',
 422         },
 423         'playlist_mincount': 100,
 424     }, {
 425         'url': 'http://www.dailymotion.com/user/UnderProject',
 426         'info_dict': {
 427             'id': 'UnderProject',
 428             'title': 'UnderProject',
 429         },
 430         'playlist_mincount': 1800,
 431         'expected_warnings': [
 432             'Stopped at duplicated page',
 433         ],
 434         'skip': 'Takes too long time',
 435     }]
 436
 437     def _extract_entries(self, id):
 438         video_ids = set()
 439         processed_urls = set()
 440         for pagenum in itertools.count(1):
 441             page_url = self._PAGE_TEMPLATE % (id, pagenum)
 442             webpage, urlh = self._download_webpage_handle_no_ff(
 443                 page_url, id, 'Downloading page %s' % pagenum)
 444             if urlh.geturl() in processed_urls:
 445                 self.report_warning('Stopped at duplicated page %s, which is the same as %s' % (
 446                     page_url, urlh.geturl()), id)
 447                 break
 448
 449             processed_urls.add(urlh.geturl())
 450
 451             for video_id in re.findall(r'data-xid="(.+?)"', webpage):
 452                 if video_id not in video_ids:
 453                     yield self.url_result(
 454                         'http://www.dailymotion.com/video/%s' % video_id,
 455                         DailymotionIE.ie_key(), video_id)
 456                     video_ids.add(video_id)
 457
 458             if re.search(self._MORE_PAGES_INDICATOR, webpage) is None:
 459                 break
 460
 461     def _real_extract(self, url):
 462         mobj = re.match(self._VALID_URL, url)
 463         user = mobj.group('user')
 464         webpage = self._download_webpage(
 465             'https://www.dailymotion.com/user/%s' % user, user)
 466         full_user = unescapeHTML(self._html_search_regex(
 467             r'<a class="nav-image" title="([^"]+)" href="/%s">' % re.escape(user),
 468             webpage, 'user'))
 469
 470         return {
 471             '_type': 'playlist',
 472             'id': user,
 473             'title': full_user,
 474             'entries': self._extract_entries(user),
 475         }