_ 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     try_get,
  28 )
  29
  30
  31 class DailymotionBaseInfoExtractor(InfoExtractor):
  32     @staticmethod
  33     def _build_request(url):
  34         """Build a request with the family filter disabled"""
  35         request = sanitized_Request(url)
  36         request.add_header('Cookie', 'family_filter=off; ff=off')
  37         return request
  38
  39     def _download_webpage_handle_no_ff(self, url, *args, **kwargs):
  40         request = self._build_request(url)
  41         return self._download_webpage_handle(request, *args, **kwargs)
  42
  43     def _download_webpage_no_ff(self, url, *args, **kwargs):
  44         request = self._build_request(url)
  45         return self._download_webpage(request, *args, **kwargs)
  46
  47
  48 class DailymotionIE(DailymotionBaseInfoExtractor):
  49     _VALID_URL = r'(?i)https?://(?:(www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(?:(?:embed|swf|#)/)?video|swf)/(?P<id>[^/?_]+)'
  50     IE_NAME = 'dailymotion'
  51
  52     _FORMATS = [
  53         ('stream_h264_ld_url', 'ld'),
  54         ('stream_h264_url', 'standard'),
  55         ('stream_h264_hq_url', 'hq'),
  56         ('stream_h264_hd_url', 'hd'),
  57         ('stream_h264_hd1080_url', 'hd180'),
  58     ]
  59
  60     _TESTS = [{
  61         'url': 'http://www.dailymotion.com/video/x5kesuj_office-christmas-party-review-jason-bateman-olivia-munn-t-j-miller_news',
  62         'md5': '074b95bdee76b9e3654137aee9c79dfe',
  63         'info_dict': {
  64             'id': 'x5kesuj',
  65             'ext': 'mp4',
  66             'title': 'Office Christmas Party Review –  Jason Bateman, Olivia Munn, T.J. Miller',
  67             'description': 'Office Christmas Party Review -  Jason Bateman, Olivia Munn, T.J. Miller',
  68             'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  69             'duration': 187,
  70             'timestamp': 1493651285,
  71             'upload_date': '20170501',
  72             'uploader': 'Deadline',
  73             'uploader_id': 'x1xm8ri',
  74             'age_limit': 0,
  75         },
  76     }, {
  77         'url': 'https://www.dailymotion.com/video/x2iuewm_steam-machine-models-pricing-listed-on-steam-store-ign-news_videogames',
  78         'md5': '2137c41a8e78554bb09225b8eb322406',
  79         'info_dict': {
  80             'id': 'x2iuewm',
  81             'ext': 'mp4',
  82             'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News',
  83             'description': 'Several come bundled with the Steam Controller.',
  84             'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  85             'duration': 74,
  86             'timestamp': 1425657362,
  87             'upload_date': '20150306',
  88             'uploader': 'IGN',
  89             'uploader_id': 'xijv66',
  90             'age_limit': 0,
  91             'view_count': int,
  92         },
  93         'skip': 'video gone',
  94     }, {
  95         # Vevo video
  96         'url': 'http://www.dailymotion.com/video/x149uew_katy-perry-roar-official_musi',
  97         'info_dict': {
  98             'title': 'Roar (Official)',
  99             'id': 'USUV71301934',
 100             'ext': 'mp4',
 101             'uploader': 'Katy Perry',
 102             'upload_date': '20130905',
 103         },
 104         'params': {
 105             'skip_download': True,
 106         },
 107         'skip': 'VEVO is only available in some countries',
 108     }, {
 109         # age-restricted video
 110         'url': 'http://www.dailymotion.com/video/xyh2zz_leanna-decker-cyber-girl-of-the-year-desires-nude-playboy-plus_redband',
 111         'md5': '0d667a7b9cebecc3c89ee93099c4159d',
 112         'info_dict': {
 113             'id': 'xyh2zz',
 114             'ext': 'mp4',
 115             'title': 'Leanna Decker - Cyber Girl Of The Year Desires Nude [Playboy Plus]',
 116             'uploader': 'HotWaves1012',
 117             'age_limit': 18,
 118         },
 119         'skip': 'video gone',
 120     }, {
 121         # geo-restricted, player v5
 122         'url': 'http://www.dailymotion.com/video/xhza0o',
 123         'only_matching': True,
 124     }, {
 125         # with subtitles
 126         'url': 'http://www.dailymotion.com/video/x20su5f_the-power-of-nightmares-1-the-rise-of-the-politics-of-fear-bbc-2004_news',
 127         'only_matching': True,
 128     }, {
 129         'url': 'http://www.dailymotion.com/swf/video/x3n92nf',
 130         'only_matching': True,
 131     }, {
 132         'url': 'http://www.dailymotion.com/swf/x3ss1m_funny-magic-trick-barry-and-stuart_fun',
 133         'only_matching': True,
 134     }]
 135
 136     @staticmethod
 137     def _extract_urls(webpage):
 138         # Look for embedded Dailymotion player
 139         matches = re.findall(
 140             r'<(?:(?:embed|iframe)[^>]+?src=|input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=)(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/(?:embed|swf)/video/.+?)\1', webpage)
 141         return list(map(lambda m: unescapeHTML(m[1]), matches))
 142
 143     def _real_extract(self, url):
 144         video_id = self._match_id(url)
 145
 146         webpage = self._download_webpage_no_ff(
 147             'https://www.dailymotion.com/video/%s' % video_id, video_id)
 148
 149         age_limit = self._rta_search(webpage)
 150
 151         description = self._og_search_description(
 152             webpage, default=None) or self._html_search_meta(
 153             'description', webpage, 'description')
 154
 155         view_count_str = self._search_regex(
 156             (r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserPlays:([\s\d,.]+)"',
 157              r'video_views_count[^>]+>\s+([\s\d\,.]+)'),
 158             webpage, 'view count', default=None)
 159         if view_count_str:
 160             view_count_str = re.sub(r'\s', '', view_count_str)
 161         view_count = str_to_int(view_count_str)
 162         comment_count = int_or_none(self._search_regex(
 163             r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserComments:(\d+)"',
 164             webpage, 'comment count', default=None))
 165
 166         player_v5 = self._search_regex(
 167             [r'buildPlayer\(({.+?})\);\n',  # See https://github.com/rg3/youtube-dl/issues/7826
 168              r'playerV5\s*=\s*dmp\.create\([^,]+?,\s*({.+?})\);',
 169              r'buildPlayer\(({.+?})\);',
 170              r'var\s+config\s*=\s*({.+?});',
 171              # New layout regex (see https://github.com/rg3/youtube-dl/issues/13580)
 172              r'__PLAYER_CONFIG__\s*=\s*({.+?});'],
 173             webpage, 'player v5', default=None)
 174         if player_v5:
 175             player = self._parse_json(player_v5, video_id)
 176             metadata = try_get(
 177                 player, lambda x: x['metadata'], dict) or self._download_json(
 178                 'http://www.dailymotion.com/player/metadata/video/%s' % video_id, video_id, query={
 179                     'integration': 'inline',
 180                     'GK_PV5_NEON': '1',
 181                 })
 182
 183             if metadata.get('error', {}).get('type') == 'password_protected':
 184                 password = self._downloader.params.get('videopassword')
 185                 if password:
 186                     r = int(metadata['id'][1:], 36)
 187                     us64e = lambda x: base64.urlsafe_b64encode(x).decode().strip('=')
 188                     t = ''.join(random.choice(string.ascii_letters) for i in range(10))
 189                     n = us64e(compat_struct_pack('I', r))
 190                     i = us64e(hashlib.md5(('%s%d%s' % (password, r, t)).encode()).digest())
 191                     metadata = self._download_json(
 192                         'http://www.dailymotion.com/player/metadata/video/p' + i + t + n, video_id)
 193
 194             self._check_error(metadata)
 195
 196             formats = []
 197             for quality, media_list in metadata['qualities'].items():
 198                 for media in media_list:
 199                     media_url = media.get('url')
 200                     if not media_url:
 201                         continue
 202                     type_ = media.get('type')
 203                     if type_ == 'application/vnd.lumberjack.manifest':
 204                         continue
 205                     ext = mimetype2ext(type_) or determine_ext(media_url)
 206                     if ext == 'm3u8':
 207                         m3u8_formats = self._extract_m3u8_formats(
 208                             media_url, video_id, 'mp4', preference=-1,
 209                             m3u8_id='hls', fatal=False)
 210                         for f in m3u8_formats:
 211                             f['url'] = f['url'].split('#')[0]
 212                             formats.append(f)
 213                     elif ext == 'f4m':
 214                         formats.extend(self._extract_f4m_formats(
 215                             media_url, video_id, preference=-1, f4m_id='hds', fatal=False))
 216                     else:
 217                         f = {
 218                             'url': media_url,
 219                             'format_id': 'http-%s' % quality,
 220                             'ext': ext,
 221                         }
 222                         m = re.search(r'H264-(?P<width>\d+)x(?P<height>\d+)', media_url)
 223                         if m:
 224                             f.update({
 225                                 'width': int(m.group('width')),
 226                                 'height': int(m.group('height')),
 227                             })
 228                         formats.append(f)
 229             self._sort_formats(formats)
 230
 231             title = metadata['title']
 232             duration = int_or_none(metadata.get('duration'))
 233             timestamp = int_or_none(metadata.get('created_time'))
 234             thumbnail = metadata.get('poster_url')
 235             uploader = metadata.get('owner', {}).get('screenname')
 236             uploader_id = metadata.get('owner', {}).get('id')
 237
 238             subtitles = {}
 239             subtitles_data = metadata.get('subtitles', {}).get('data', {})
 240             if subtitles_data and isinstance(subtitles_data, dict):
 241                 for subtitle_lang, subtitle in subtitles_data.items():
 242                     subtitles[subtitle_lang] = [{
 243                         'ext': determine_ext(subtitle_url),
 244                         'url': subtitle_url,
 245                     } for subtitle_url in subtitle.get('urls', [])]
 246
 247             return {
 248                 'id': video_id,
 249                 'title': title,
 250                 'description': description,
 251                 'thumbnail': thumbnail,
 252                 'duration': duration,
 253                 'timestamp': timestamp,
 254                 'uploader': uploader,
 255                 'uploader_id': uploader_id,
 256                 'age_limit': age_limit,
 257                 'view_count': view_count,
 258                 'comment_count': comment_count,
 259                 'formats': formats,
 260                 'subtitles': subtitles,
 261             }
 262
 263         # vevo embed
 264         vevo_id = self._search_regex(
 265             r'<link rel="video_src" href="[^"]*?vevo\.com[^"]*?video=(?P<id>[\w]*)',
 266             webpage, 'vevo embed', default=None)
 267         if vevo_id:
 268             return self.url_result('vevo:%s' % vevo_id, 'Vevo')
 269
 270         # fallback old player
 271         embed_page = self._download_webpage_no_ff(
 272             'https://www.dailymotion.com/embed/video/%s' % video_id,
 273             video_id, 'Downloading embed page')
 274
 275         timestamp = parse_iso8601(self._html_search_meta(
 276             'video:release_date', webpage, 'upload date'))
 277
 278         info = self._parse_json(
 279             self._search_regex(
 280                 r'var info = ({.*?}),$', embed_page,
 281                 'video info', flags=re.MULTILINE),
 282             video_id)
 283
 284         self._check_error(info)
 285
 286         formats = []
 287         for (key, format_id) in self._FORMATS:
 288             video_url = info.get(key)
 289             if video_url is not None:
 290                 m_size = re.search(r'H264-(\d+)x(\d+)', video_url)
 291                 if m_size is not None:
 292                     width, height = map(int_or_none, (m_size.group(1), m_size.group(2)))
 293                 else:
 294                     width, height = None, None
 295                 formats.append({
 296                     'url': video_url,
 297                     'ext': 'mp4',
 298                     'format_id': format_id,
 299                     'width': width,
 300                     'height': height,
 301                 })
 302         self._sort_formats(formats)
 303
 304         # subtitles
 305         video_subtitles = self.extract_subtitles(video_id, webpage)
 306
 307         title = self._og_search_title(webpage, default=None)
 308         if title is None:
 309             title = self._html_search_regex(
 310                 r'(?s)<span\s+id="video_title"[^>]*>(.*?)</span>', webpage,
 311                 'title')
 312
 313         return {
 314             'id': video_id,
 315             'formats': formats,
 316             'uploader': info['owner.screenname'],
 317             'timestamp': timestamp,
 318             'title': title,
 319             'description': description,
 320             'subtitles': video_subtitles,
 321             'thumbnail': info['thumbnail_url'],
 322             'age_limit': age_limit,
 323             'view_count': view_count,
 324             'duration': info['duration']
 325         }
 326
 327     def _check_error(self, info):
 328         error = info.get('error')
 329         if error:
 330             title = error.get('title') or error['message']
 331             # See https://developer.dailymotion.com/api#access-error
 332             if error.get('code') == 'DM007':
 333                 self.raise_geo_restricted(msg=title)
 334             raise ExtractorError(
 335                 '%s said: %s' % (self.IE_NAME, title), expected=True)
 336
 337     def _get_subtitles(self, video_id, webpage):
 338         try:
 339             sub_list = self._download_webpage(
 340                 'https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id,
 341                 video_id, note=False)
 342         except ExtractorError as err:
 343             self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err))
 344             return {}
 345         info = json.loads(sub_list)
 346         if (info['total'] > 0):
 347             sub_lang_list = dict((l['language'], [{'url': l['url'], 'ext': 'srt'}]) for l in info['list'])
 348             return sub_lang_list
 349         self._downloader.report_warning('video doesn\'t have subtitles')
 350         return {}
 351
 352
 353 class DailymotionPlaylistIE(DailymotionBaseInfoExtractor):
 354     IE_NAME = 'dailymotion:playlist'
 355     _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P<id>x[0-9a-z]+)'
 356     _TESTS = [{
 357         'url': 'http://www.dailymotion.com/playlist/xv4bw_nqtv_sport/1#video=xl8v3q',
 358         'info_dict': {
 359             'title': 'SPORT',
 360             'id': 'xv4bw',
 361         },
 362         'playlist_mincount': 20,
 363     }]
 364     _PAGE_SIZE = 100
 365
 366     def _fetch_page(self, playlist_id, authorizaion, page):
 367         page += 1
 368         videos = self._download_json(
 369             'https://graphql.api.dailymotion.com',
 370             playlist_id, 'Downloading page %d' % page,
 371             data=json.dumps({
 372                 'query': '''{
 373   collection(xid: "%s") {
 374     videos(first: %d, page: %d) {
 375       pageInfo {
 376         hasNextPage
 377         nextPage
 378       }
 379       edges {
 380         node {
 381           xid
 382           url
 383         }
 384       }
 385     }
 386   }
 387 }''' % (playlist_id, self._PAGE_SIZE, page)
 388             }).encode(), headers={
 389                 'Authorization': authorizaion,
 390                 'Origin': 'https://www.dailymotion.com',
 391             })['data']['collection']['videos']
 392         for edge in videos['edges']:
 393             node = edge['node']
 394             yield self.url_result(
 395                 node['url'], DailymotionIE.ie_key(), node['xid'])
 396
 397     def _real_extract(self, url):
 398         playlist_id = self._match_id(url)
 399         webpage = self._download_webpage(url, playlist_id)
 400         api = self._parse_json(self._search_regex(
 401             r'__PLAYER_CONFIG__\s*=\s*({.+?});',
 402             webpage, 'player config'), playlist_id)['context']['api']
 403         auth = self._download_json(
 404             api.get('auth_url', 'https://graphql.api.dailymotion.com/oauth/token'),
 405             playlist_id, data=urlencode_postdata({
 406                 'client_id': api.get('client_id', 'f1a362d288c1b98099c7'),
 407                 'client_secret': api.get('client_secret', 'eea605b96e01c796ff369935357eca920c5da4c5'),
 408                 'grant_type': 'client_credentials',
 409             }))
 410         authorizaion = '%s %s' % (auth.get('token_type', 'Bearer'), auth['access_token'])
 411         entries = OnDemandPagedList(functools.partial(
 412             self._fetch_page, playlist_id, authorizaion), self._PAGE_SIZE)
 413         return self.playlist_result(
 414             entries, playlist_id,
 415             self._og_search_title(webpage))
 416
 417
 418 class DailymotionUserIE(DailymotionBaseInfoExtractor):
 419     IE_NAME = 'dailymotion:user'
 420     _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|swf|#|video|playlist)/)(?:(?:old/)?user/)?(?P<user>[^/]+)'
 421     _MORE_PAGES_INDICATOR = r'(?s)<div class="pages[^"]*">.*?<a\s+class="[^"]*?icon-arrow_right[^"]*?"'
 422     _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s'
 423     _TESTS = [{
 424         'url': 'https://www.dailymotion.com/user/nqtv',
 425         'info_dict': {
 426             'id': 'nqtv',
 427             'title': 'Rémi Gaillard',
 428         },
 429         'playlist_mincount': 100,
 430     }, {
 431         'url': 'http://www.dailymotion.com/user/UnderProject',
 432         'info_dict': {
 433             'id': 'UnderProject',
 434             'title': 'UnderProject',
 435         },
 436         'playlist_mincount': 1800,
 437         'expected_warnings': [
 438             'Stopped at duplicated page',
 439         ],
 440         'skip': 'Takes too long time',
 441     }]
 442
 443     def _extract_entries(self, id):
 444         video_ids = set()
 445         processed_urls = set()
 446         for pagenum in itertools.count(1):
 447             page_url = self._PAGE_TEMPLATE % (id, pagenum)
 448             webpage, urlh = self._download_webpage_handle_no_ff(
 449                 page_url, id, 'Downloading page %s' % pagenum)
 450             if urlh.geturl() in processed_urls:
 451                 self.report_warning('Stopped at duplicated page %s, which is the same as %s' % (
 452                     page_url, urlh.geturl()), id)
 453                 break
 454
 455             processed_urls.add(urlh.geturl())
 456
 457             for video_id in re.findall(r'data-xid="(.+?)"', webpage):
 458                 if video_id not in video_ids:
 459                     yield self.url_result(
 460                         'http://www.dailymotion.com/video/%s' % video_id,
 461                         DailymotionIE.ie_key(), video_id)
 462                     video_ids.add(video_id)
 463
 464             if re.search(self._MORE_PAGES_INDICATOR, webpage) is None:
 465                 break
 466
 467     def _real_extract(self, url):
 468         mobj = re.match(self._VALID_URL, url)
 469         user = mobj.group('user')
 470         webpage = self._download_webpage(
 471             'https://www.dailymotion.com/user/%s' % user, user)
 472         full_user = unescapeHTML(self._html_search_regex(
 473             r'<a class="nav-image" title="([^"]+)" href="/%s">' % re.escape(user),
 474             webpage, 'user'))
 475
 476         return {
 477             '_type': 'playlist',
 478             'id': user,
 479             'title': full_user,
 480             'entries': self._extract_entries(user),
 481         }