youtube_dl/extractor/dailymotion.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import json
   6 import itertools
   7
   8 from .common import InfoExtractor
   9
  10 from ..utils import (
  11     determine_ext,
  12     error_to_compat_str,
  13     ExtractorError,
  14     int_or_none,
  15     parse_iso8601,
  16     sanitized_Request,
  17     str_to_int,
  18     unescapeHTML,
  19     mimetype2ext,
  20 )
  21
  22
  23 class DailymotionBaseInfoExtractor(InfoExtractor):
  24     @staticmethod
  25     def _build_request(url):
  26         """Build a request with the family filter disabled"""
  27         request = sanitized_Request(url)
  28         request.add_header('Cookie', 'family_filter=off; ff=off')
  29         return request
  30
  31     def _download_webpage_handle_no_ff(self, url, *args, **kwargs):
  32         request = self._build_request(url)
  33         return self._download_webpage_handle(request, *args, **kwargs)
  34
  35     def _download_webpage_no_ff(self, url, *args, **kwargs):
  36         request = self._build_request(url)
  37         return self._download_webpage(request, *args, **kwargs)
  38
  39
  40 class DailymotionIE(DailymotionBaseInfoExtractor):
  41     _VALID_URL = r'(?i)(?:https?://)?(?:(www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(?:embed|swf|#)/)?video/(?P<id>[^/?_]+)'
  42     IE_NAME = 'dailymotion'
  43
  44     _FORMATS = [
  45         ('stream_h264_ld_url', 'ld'),
  46         ('stream_h264_url', 'standard'),
  47         ('stream_h264_hq_url', 'hq'),
  48         ('stream_h264_hd_url', 'hd'),
  49         ('stream_h264_hd1080_url', 'hd180'),
  50     ]
  51
  52     _TESTS = [
  53         {
  54             'url': 'http://www.dailymotion.com/video/x5kesuj_office-christmas-party-review-jason-bateman-olivia-munn-t-j-miller_news',
  55             'md5': '074b95bdee76b9e3654137aee9c79dfe',
  56             'info_dict': {
  57                 'id': 'x5kesuj',
  58                 'ext': 'mp4',
  59                 'title': 'Office Christmas Party Review –  Jason Bateman, Olivia Munn, T.J. Miller',
  60                 'description': 'Office Christmas Party Review -  Jason Bateman, Olivia Munn, T.J. Miller',
  61                 'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  62                 'duration': 187,
  63                 'timestamp': 1493651285,
  64                 'upload_date': '20170501',
  65                 'uploader': 'Deadline',
  66                 'uploader_id': 'x1xm8ri',
  67                 'age_limit': 0,
  68                 'view_count': int,
  69             },
  70         },
  71         {
  72             'url': 'https://www.dailymotion.com/video/x2iuewm_steam-machine-models-pricing-listed-on-steam-store-ign-news_videogames',
  73             'md5': '2137c41a8e78554bb09225b8eb322406',
  74             'info_dict': {
  75                 'id': 'x2iuewm',
  76                 'ext': 'mp4',
  77                 'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News',
  78                 'description': 'Several come bundled with the Steam Controller.',
  79                 'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  80                 'duration': 74,
  81                 'timestamp': 1425657362,
  82                 'upload_date': '20150306',
  83                 'uploader': 'IGN',
  84                 'uploader_id': 'xijv66',
  85                 'age_limit': 0,
  86                 'view_count': int,
  87             },
  88             'skip': 'video gone',
  89         },
  90         # Vevo video
  91         {
  92             'url': 'http://www.dailymotion.com/video/x149uew_katy-perry-roar-official_musi',
  93             'info_dict': {
  94                 'title': 'Roar (Official)',
  95                 'id': 'USUV71301934',
  96                 'ext': 'mp4',
  97                 'uploader': 'Katy Perry',
  98                 'upload_date': '20130905',
  99             },
 100             'params': {
 101                 'skip_download': True,
 102             },
 103             'skip': 'VEVO is only available in some countries',
 104         },
 105         # age-restricted video
 106         {
 107             'url': 'http://www.dailymotion.com/video/xyh2zz_leanna-decker-cyber-girl-of-the-year-desires-nude-playboy-plus_redband',
 108             'md5': '0d667a7b9cebecc3c89ee93099c4159d',
 109             'info_dict': {
 110                 'id': 'xyh2zz',
 111                 'ext': 'mp4',
 112                 'title': 'Leanna Decker - Cyber Girl Of The Year Desires Nude [Playboy Plus]',
 113                 'uploader': 'HotWaves1012',
 114                 'age_limit': 18,
 115             },
 116             'skip': 'video gone',
 117         },
 118         # geo-restricted, player v5
 119         {
 120             'url': 'http://www.dailymotion.com/video/xhza0o',
 121             'only_matching': True,
 122         },
 123         # with subtitles
 124         {
 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         {
 129             'url': 'http://www.dailymotion.com/swf/video/x3n92nf',
 130             'only_matching': True,
 131         }
 132     ]
 133
 134     @staticmethod
 135     def _extract_urls(webpage):
 136         # Look for embedded Dailymotion player
 137         matches = re.findall(
 138             r'<(?:(?:embed|iframe)[^>]+?src=|input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=)(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/(?:embed|swf)/video/.+?)\1', webpage)
 139         return list(map(lambda m: unescapeHTML(m[1]), matches))
 140
 141     def _real_extract(self, url):
 142         video_id = self._match_id(url)
 143
 144         webpage = self._download_webpage_no_ff(
 145             'https://www.dailymotion.com/video/%s' % video_id, video_id)
 146
 147         age_limit = self._rta_search(webpage)
 148
 149         description = self._og_search_description(webpage) or self._html_search_meta(
 150             'description', webpage, 'description')
 151
 152         view_count_str = self._search_regex(
 153             (r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserPlays:([\s\d,.]+)"',
 154              r'video_views_count[^>]+>\s+([\s\d\,.]+)'),
 155             webpage, 'view count', fatal=False)
 156         if view_count_str:
 157             view_count_str = re.sub(r'\s', '', view_count_str)
 158         view_count = str_to_int(view_count_str)
 159         comment_count = int_or_none(self._search_regex(
 160             r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserComments:(\d+)"',
 161             webpage, 'comment count', default=None))
 162
 163         player_v5 = self._search_regex(
 164             [r'buildPlayer\(({.+?})\);\n',  # See https://github.com/rg3/youtube-dl/issues/7826
 165              r'playerV5\s*=\s*dmp\.create\([^,]+?,\s*({.+?})\);',
 166              r'buildPlayer\(({.+?})\);',
 167              r'var\s+config\s*=\s*({.+?});'],
 168             webpage, 'player v5', default=None)
 169         if player_v5:
 170             player = self._parse_json(player_v5, video_id)
 171             metadata = player['metadata']
 172
 173             self._check_error(metadata)
 174
 175             formats = []
 176             for quality, media_list in metadata['qualities'].items():
 177                 for media in media_list:
 178                     media_url = media.get('url')
 179                     if not media_url:
 180                         continue
 181                     type_ = media.get('type')
 182                     if type_ == 'application/vnd.lumberjack.manifest':
 183                         continue
 184                     ext = mimetype2ext(type_) or determine_ext(media_url)
 185                     if ext == 'm3u8':
 186                         formats.extend(self._extract_m3u8_formats(
 187                             media_url, video_id, 'mp4', preference=-1,
 188                             m3u8_id='hls', fatal=False))
 189                     elif ext == 'f4m':
 190                         formats.extend(self._extract_f4m_formats(
 191                             media_url, video_id, preference=-1, f4m_id='hds', fatal=False))
 192                     else:
 193                         f = {
 194                             'url': media_url,
 195                             'format_id': 'http-%s' % quality,
 196                             'ext': ext,
 197                         }
 198                         m = re.search(r'H264-(?P<width>\d+)x(?P<height>\d+)', media_url)
 199                         if m:
 200                             f.update({
 201                                 'width': int(m.group('width')),
 202                                 'height': int(m.group('height')),
 203                             })
 204                         formats.append(f)
 205             self._sort_formats(formats)
 206
 207             title = metadata['title']
 208             duration = int_or_none(metadata.get('duration'))
 209             timestamp = int_or_none(metadata.get('created_time'))
 210             thumbnail = metadata.get('poster_url')
 211             uploader = metadata.get('owner', {}).get('screenname')
 212             uploader_id = metadata.get('owner', {}).get('id')
 213
 214             subtitles = {}
 215             subtitles_data = metadata.get('subtitles', {}).get('data', {})
 216             if subtitles_data and isinstance(subtitles_data, dict):
 217                 for subtitle_lang, subtitle in subtitles_data.items():
 218                     subtitles[subtitle_lang] = [{
 219                         'ext': determine_ext(subtitle_url),
 220                         'url': subtitle_url,
 221                     } for subtitle_url in subtitle.get('urls', [])]
 222
 223             return {
 224                 'id': video_id,
 225                 'title': title,
 226                 'description': description,
 227                 'thumbnail': thumbnail,
 228                 'duration': duration,
 229                 'timestamp': timestamp,
 230                 'uploader': uploader,
 231                 'uploader_id': uploader_id,
 232                 'age_limit': age_limit,
 233                 'view_count': view_count,
 234                 'comment_count': comment_count,
 235                 'formats': formats,
 236                 'subtitles': subtitles,
 237             }
 238
 239         # vevo embed
 240         vevo_id = self._search_regex(
 241             r'<link rel="video_src" href="[^"]*?vevo.com[^"]*?video=(?P<id>[\w]*)',
 242             webpage, 'vevo embed', default=None)
 243         if vevo_id:
 244             return self.url_result('vevo:%s' % vevo_id, 'Vevo')
 245
 246         # fallback old player
 247         embed_page = self._download_webpage_no_ff(
 248             'https://www.dailymotion.com/embed/video/%s' % video_id,
 249             video_id, 'Downloading embed page')
 250
 251         timestamp = parse_iso8601(self._html_search_meta(
 252             'video:release_date', webpage, 'upload date'))
 253
 254         info = self._parse_json(
 255             self._search_regex(
 256                 r'var info = ({.*?}),$', embed_page,
 257                 'video info', flags=re.MULTILINE),
 258             video_id)
 259
 260         self._check_error(info)
 261
 262         formats = []
 263         for (key, format_id) in self._FORMATS:
 264             video_url = info.get(key)
 265             if video_url is not None:
 266                 m_size = re.search(r'H264-(\d+)x(\d+)', video_url)
 267                 if m_size is not None:
 268                     width, height = map(int_or_none, (m_size.group(1), m_size.group(2)))
 269                 else:
 270                     width, height = None, None
 271                 formats.append({
 272                     'url': video_url,
 273                     'ext': 'mp4',
 274                     'format_id': format_id,
 275                     'width': width,
 276                     'height': height,
 277                 })
 278         self._sort_formats(formats)
 279
 280         # subtitles
 281         video_subtitles = self.extract_subtitles(video_id, webpage)
 282
 283         title = self._og_search_title(webpage, default=None)
 284         if title is None:
 285             title = self._html_search_regex(
 286                 r'(?s)<span\s+id="video_title"[^>]*>(.*?)</span>', webpage,
 287                 'title')
 288
 289         return {
 290             'id': video_id,
 291             'formats': formats,
 292             'uploader': info['owner.screenname'],
 293             'timestamp': timestamp,
 294             'title': title,
 295             'description': description,
 296             'subtitles': video_subtitles,
 297             'thumbnail': info['thumbnail_url'],
 298             'age_limit': age_limit,
 299             'view_count': view_count,
 300             'duration': info['duration']
 301         }
 302
 303     def _check_error(self, info):
 304         error = info.get('error')
 305         if info.get('error') is not None:
 306             title = error['title']
 307             # See https://developer.dailymotion.com/api#access-error
 308             if error.get('code') == 'DM007':
 309                 self.raise_geo_restricted(msg=title)
 310             raise ExtractorError(
 311                 '%s said: %s' % (self.IE_NAME, title), expected=True)
 312
 313     def _get_subtitles(self, video_id, webpage):
 314         try:
 315             sub_list = self._download_webpage(
 316                 'https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id,
 317                 video_id, note=False)
 318         except ExtractorError as err:
 319             self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err))
 320             return {}
 321         info = json.loads(sub_list)
 322         if (info['total'] > 0):
 323             sub_lang_list = dict((l['language'], [{'url': l['url'], 'ext': 'srt'}]) for l in info['list'])
 324             return sub_lang_list
 325         self._downloader.report_warning('video doesn\'t have subtitles')
 326         return {}
 327
 328
 329 class DailymotionPlaylistIE(DailymotionBaseInfoExtractor):
 330     IE_NAME = 'dailymotion:playlist'
 331     _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P<id>.+?)/'
 332     _MORE_PAGES_INDICATOR = r'(?s)<div class="pages[^"]*">.*?<a\s+class="[^"]*?icon-arrow_right[^"]*?"'
 333     _PAGE_TEMPLATE = 'https://www.dailymotion.com/playlist/%s/%s'
 334     _TESTS = [{
 335         'url': 'http://www.dailymotion.com/playlist/xv4bw_nqtv_sport/1#video=xl8v3q',
 336         'info_dict': {
 337             'title': 'SPORT',
 338             'id': 'xv4bw_nqtv_sport',
 339         },
 340         'playlist_mincount': 20,
 341     }]
 342
 343     def _extract_entries(self, id):
 344         video_ids = set()
 345         processed_urls = set()
 346         for pagenum in itertools.count(1):
 347             page_url = self._PAGE_TEMPLATE % (id, pagenum)
 348             webpage, urlh = self._download_webpage_handle_no_ff(
 349                 page_url, id, 'Downloading page %s' % pagenum)
 350             if urlh.geturl() in processed_urls:
 351                 self.report_warning('Stopped at duplicated page %s, which is the same as %s' % (
 352                     page_url, urlh.geturl()), id)
 353                 break
 354
 355             processed_urls.add(urlh.geturl())
 356
 357             for video_id in re.findall(r'data-xid="(.+?)"', webpage):
 358                 if video_id not in video_ids:
 359                     yield self.url_result(
 360                         'http://www.dailymotion.com/video/%s' % video_id,
 361                         DailymotionIE.ie_key(), video_id)
 362                     video_ids.add(video_id)
 363
 364             if re.search(self._MORE_PAGES_INDICATOR, webpage) is None:
 365                 break
 366
 367     def _real_extract(self, url):
 368         mobj = re.match(self._VALID_URL, url)
 369         playlist_id = mobj.group('id')
 370         webpage = self._download_webpage(url, playlist_id)
 371
 372         return {
 373             '_type': 'playlist',
 374             'id': playlist_id,
 375             'title': self._og_search_title(webpage),
 376             'entries': self._extract_entries(playlist_id),
 377         }
 378
 379
 380 class DailymotionUserIE(DailymotionPlaylistIE):
 381     IE_NAME = 'dailymotion:user'
 382     _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|swf|#|video|playlist)/)(?:(?:old/)?user/)?(?P<user>[^/]+)'
 383     _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s'
 384     _TESTS = [{
 385         'url': 'https://www.dailymotion.com/user/nqtv',
 386         'info_dict': {
 387             'id': 'nqtv',
 388             'title': 'Rémi Gaillard',
 389         },
 390         'playlist_mincount': 100,
 391     }, {
 392         'url': 'http://www.dailymotion.com/user/UnderProject',
 393         'info_dict': {
 394             'id': 'UnderProject',
 395             'title': 'UnderProject',
 396         },
 397         'playlist_mincount': 1800,
 398         'expected_warnings': [
 399             'Stopped at duplicated page',
 400         ],
 401         'skip': 'Takes too long time',
 402     }]
 403
 404     def _real_extract(self, url):
 405         mobj = re.match(self._VALID_URL, url)
 406         user = mobj.group('user')
 407         webpage = self._download_webpage(
 408             'https://www.dailymotion.com/user/%s' % user, user)
 409         full_user = unescapeHTML(self._html_search_regex(
 410             r'<a class="nav-image" title="([^"]+)" href="/%s">' % re.escape(user),
 411             webpage, 'user'))
 412
 413         return {
 414             '_type': 'playlist',
 415             'id': user,
 416             'title': full_user,
 417             'entries': self._extract_entries(user),
 418         }
 419
 420
 421 class DailymotionCloudIE(DailymotionBaseInfoExtractor):
 422     _VALID_URL_PREFIX = r'https?://api\.dmcloud\.net/(?:player/)?embed/'
 423     _VALID_URL = r'%s[^/]+/(?P<id>[^/?]+)' % _VALID_URL_PREFIX
 424     _VALID_EMBED_URL = r'%s[^/]+/[^\'"]+' % _VALID_URL_PREFIX
 425
 426     _TESTS = [{
 427         # From http://www.francetvinfo.fr/economie/entreprises/les-entreprises-familiales-le-secret-de-la-reussite_933271.html
 428         # Tested at FranceTvInfo_2
 429         'url': 'http://api.dmcloud.net/embed/4e7343f894a6f677b10006b4/556e03339473995ee145930c?auth=1464865870-0-jyhsm84b-ead4c701fb750cf9367bf4447167a3db&autoplay=1',
 430         'only_matching': True,
 431     }, {
 432         # http://www.francetvinfo.fr/societe/larguez-les-amarres-le-cobaturage-se-developpe_980101.html
 433         'url': 'http://api.dmcloud.net/player/embed/4e7343f894a6f677b10006b4/559545469473996d31429f06?auth=1467430263-0-90tglw2l-a3a4b64ed41efe48d7fccad85b8b8fda&autoplay=1',
 434         'only_matching': True,
 435     }]
 436
 437     @classmethod
 438     def _extract_dmcloud_url(cls, webpage):
 439         mobj = re.search(r'<iframe[^>]+src=[\'"](%s)[\'"]' % cls._VALID_EMBED_URL, webpage)
 440         if mobj:
 441             return mobj.group(1)
 442
 443         mobj = re.search(
 444             r'<input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=[\'"](%s)[\'"]' % cls._VALID_EMBED_URL,
 445             webpage)
 446         if mobj:
 447             return mobj.group(1)
 448
 449     def _real_extract(self, url):
 450         video_id = self._match_id(url)
 451
 452         webpage = self._download_webpage_no_ff(url, video_id)
 453
 454         title = self._html_search_regex(r'<title>([^>]+)</title>', webpage, 'title')
 455
 456         video_info = self._parse_json(self._search_regex(
 457             r'var\s+info\s*=\s*([^;]+);', webpage, 'video info'), video_id)
 458
 459         # TODO: parse ios_url, which is in fact a manifest
 460         video_url = video_info['mp4_url']
 461
 462         return {
 463             'id': video_id,
 464             'url': video_url,
 465             'title': title,
 466             'thumbnail': video_info.get('thumbnail_url'),
 467         }