_ 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 hashlib
   6 import itertools
   7 import json
   8 import random
   9 import re
  10 import string
  11 import struct
  12
  13 from .common import InfoExtractor
  14
  15 from ..utils import (
  16     determine_ext,
  17     error_to_compat_str,
  18     ExtractorError,
  19     int_or_none,
  20     parse_iso8601,
  21     sanitized_Request,
  22     str_to_int,
  23     unescapeHTML,
  24     mimetype2ext,
  25 )
  26
  27
  28 class DailymotionBaseInfoExtractor(InfoExtractor):
  29     @staticmethod
  30     def _build_request(url):
  31         """Build a request with the family filter disabled"""
  32         request = sanitized_Request(url)
  33         request.add_header('Cookie', 'family_filter=off; ff=off')
  34         return request
  35
  36     def _download_webpage_handle_no_ff(self, url, *args, **kwargs):
  37         request = self._build_request(url)
  38         return self._download_webpage_handle(request, *args, **kwargs)
  39
  40     def _download_webpage_no_ff(self, url, *args, **kwargs):
  41         request = self._build_request(url)
  42         return self._download_webpage(request, *args, **kwargs)
  43
  44
  45 class DailymotionIE(DailymotionBaseInfoExtractor):
  46     _VALID_URL = r'(?i)https?://(?:(www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(?:(?:embed|swf|#)/)?video|swf)/(?P<id>[^/?_]+)'
  47     IE_NAME = 'dailymotion'
  48
  49     _FORMATS = [
  50         ('stream_h264_ld_url', 'ld'),
  51         ('stream_h264_url', 'standard'),
  52         ('stream_h264_hq_url', 'hq'),
  53         ('stream_h264_hd_url', 'hd'),
  54         ('stream_h264_hd1080_url', 'hd180'),
  55     ]
  56
  57     _TESTS = [{
  58         'url': 'http://www.dailymotion.com/video/x5kesuj_office-christmas-party-review-jason-bateman-olivia-munn-t-j-miller_news',
  59         'md5': '074b95bdee76b9e3654137aee9c79dfe',
  60         'info_dict': {
  61             'id': 'x5kesuj',
  62             'ext': 'mp4',
  63             'title': 'Office Christmas Party Review –  Jason Bateman, Olivia Munn, T.J. Miller',
  64             'description': 'Office Christmas Party Review -  Jason Bateman, Olivia Munn, T.J. Miller',
  65             'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  66             'duration': 187,
  67             'timestamp': 1493651285,
  68             'upload_date': '20170501',
  69             'uploader': 'Deadline',
  70             'uploader_id': 'x1xm8ri',
  71             'age_limit': 0,
  72         },
  73     }, {
  74         'url': 'https://www.dailymotion.com/video/x2iuewm_steam-machine-models-pricing-listed-on-steam-store-ign-news_videogames',
  75         'md5': '2137c41a8e78554bb09225b8eb322406',
  76         'info_dict': {
  77             'id': 'x2iuewm',
  78             'ext': 'mp4',
  79             'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News',
  80             'description': 'Several come bundled with the Steam Controller.',
  81             'thumbnail': r're:^https?:.*\.(?:jpg|png)$',
  82             'duration': 74,
  83             'timestamp': 1425657362,
  84             'upload_date': '20150306',
  85             'uploader': 'IGN',
  86             'uploader_id': 'xijv66',
  87             'age_limit': 0,
  88             'view_count': int,
  89         },
  90         'skip': 'video gone',
  91     }, {
  92         # Vevo video
  93         'url': 'http://www.dailymotion.com/video/x149uew_katy-perry-roar-official_musi',
  94         'info_dict': {
  95             'title': 'Roar (Official)',
  96             'id': 'USUV71301934',
  97             'ext': 'mp4',
  98             'uploader': 'Katy Perry',
  99             'upload_date': '20130905',
 100         },
 101         'params': {
 102             'skip_download': True,
 103         },
 104         'skip': 'VEVO is only available in some countries',
 105     }, {
 106         # age-restricted video
 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         'url': 'http://www.dailymotion.com/video/xhza0o',
 120         'only_matching': True,
 121     }, {
 122         # with subtitles
 123         'url': 'http://www.dailymotion.com/video/x20su5f_the-power-of-nightmares-1-the-rise-of-the-politics-of-fear-bbc-2004_news',
 124         'only_matching': True,
 125     }, {
 126         'url': 'http://www.dailymotion.com/swf/video/x3n92nf',
 127         'only_matching': True,
 128     }, {
 129         'url': 'http://www.dailymotion.com/swf/x3ss1m_funny-magic-trick-barry-and-stuart_fun',
 130         'only_matching': True,
 131     }]
 132
 133     @staticmethod
 134     def _extract_urls(webpage):
 135         # Look for embedded Dailymotion player
 136         matches = re.findall(
 137             r'<(?:(?:embed|iframe)[^>]+?src=|input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=)(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/(?:embed|swf)/video/.+?)\1', webpage)
 138         return list(map(lambda m: unescapeHTML(m[1]), matches))
 139
 140     def _real_extract(self, url):
 141         video_id = self._match_id(url)
 142
 143         webpage = self._download_webpage_no_ff(
 144             'https://www.dailymotion.com/video/%s' % video_id, video_id)
 145
 146         age_limit = self._rta_search(webpage)
 147
 148         description = self._og_search_description(webpage) or self._html_search_meta(
 149             'description', webpage, 'description')
 150
 151         view_count_str = self._search_regex(
 152             (r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserPlays:([\s\d,.]+)"',
 153              r'video_views_count[^>]+>\s+([\s\d\,.]+)'),
 154             webpage, 'view count', default=None)
 155         if view_count_str:
 156             view_count_str = re.sub(r'\s', '', view_count_str)
 157         view_count = str_to_int(view_count_str)
 158         comment_count = int_or_none(self._search_regex(
 159             r'<meta[^>]+itemprop="interactionCount"[^>]+content="UserComments:(\d+)"',
 160             webpage, 'comment count', default=None))
 161
 162         player_v5 = self._search_regex(
 163             [r'buildPlayer\(({.+?})\);\n',  # See https://github.com/rg3/youtube-dl/issues/7826
 164              r'playerV5\s*=\s*dmp\.create\([^,]+?,\s*({.+?})\);',
 165              r'buildPlayer\(({.+?})\);',
 166              r'var\s+config\s*=\s*({.+?});',
 167              # New layout regex (see https://github.com/rg3/youtube-dl/issues/13580)
 168              r'__PLAYER_CONFIG__\s*=\s*({.+?});'],
 169             webpage, 'player v5', default=None)
 170         if player_v5:
 171             player = self._parse_json(player_v5, video_id)
 172             metadata = player['metadata']
 173
 174             if metadata.get('error', {}).get('type') == 'password_protected':
 175                 password = self._downloader.params.get('videopassword')
 176                 if password:
 177                     r = int(metadata['id'][1:], 36)
 178                     us64e = lambda x: base64.urlsafe_b64encode(x).decode().strip('=')
 179                     t = ''.join(random.choice(string.ascii_letters) for i in range(10))
 180                     n = us64e(struct.pack('I', r))
 181                     i = us64e(hashlib.md5(('%s%d%s' % (password, r, t)).encode()).digest())
 182                     metadata = self._download_json(
 183                         'http://www.dailymotion.com/player/metadata/video/p' + i + t + n, video_id)
 184
 185             self._check_error(metadata)
 186
 187             formats = []
 188             for quality, media_list in metadata['qualities'].items():
 189                 for media in media_list:
 190                     media_url = media.get('url')
 191                     if not media_url:
 192                         continue
 193                     type_ = media.get('type')
 194                     if type_ == 'application/vnd.lumberjack.manifest':
 195                         continue
 196                     ext = mimetype2ext(type_) or determine_ext(media_url)
 197                     if ext == 'm3u8':
 198                         m3u8_formats = self._extract_m3u8_formats(
 199                             media_url, video_id, 'mp4', preference=-1,
 200                             m3u8_id='hls', fatal=False)
 201                         for f in m3u8_formats:
 202                             f['url'] = f['url'].split('#')[0]
 203                             formats.append(f)
 204                     elif ext == 'f4m':
 205                         formats.extend(self._extract_f4m_formats(
 206                             media_url, video_id, preference=-1, f4m_id='hds', fatal=False))
 207                     else:
 208                         f = {
 209                             'url': media_url,
 210                             'format_id': 'http-%s' % quality,
 211                             'ext': ext,
 212                         }
 213                         m = re.search(r'H264-(?P<width>\d+)x(?P<height>\d+)', media_url)
 214                         if m:
 215                             f.update({
 216                                 'width': int(m.group('width')),
 217                                 'height': int(m.group('height')),
 218                             })
 219                         formats.append(f)
 220             self._sort_formats(formats)
 221
 222             title = metadata['title']
 223             duration = int_or_none(metadata.get('duration'))
 224             timestamp = int_or_none(metadata.get('created_time'))
 225             thumbnail = metadata.get('poster_url')
 226             uploader = metadata.get('owner', {}).get('screenname')
 227             uploader_id = metadata.get('owner', {}).get('id')
 228
 229             subtitles = {}
 230             subtitles_data = metadata.get('subtitles', {}).get('data', {})
 231             if subtitles_data and isinstance(subtitles_data, dict):
 232                 for subtitle_lang, subtitle in subtitles_data.items():
 233                     subtitles[subtitle_lang] = [{
 234                         'ext': determine_ext(subtitle_url),
 235                         'url': subtitle_url,
 236                     } for subtitle_url in subtitle.get('urls', [])]
 237
 238             return {
 239                 'id': video_id,
 240                 'title': title,
 241                 'description': description,
 242                 'thumbnail': thumbnail,
 243                 'duration': duration,
 244                 'timestamp': timestamp,
 245                 'uploader': uploader,
 246                 'uploader_id': uploader_id,
 247                 'age_limit': age_limit,
 248                 'view_count': view_count,
 249                 'comment_count': comment_count,
 250                 'formats': formats,
 251                 'subtitles': subtitles,
 252             }
 253
 254         # vevo embed
 255         vevo_id = self._search_regex(
 256             r'<link rel="video_src" href="[^"]*?vevo\.com[^"]*?video=(?P<id>[\w]*)',
 257             webpage, 'vevo embed', default=None)
 258         if vevo_id:
 259             return self.url_result('vevo:%s' % vevo_id, 'Vevo')
 260
 261         # fallback old player
 262         embed_page = self._download_webpage_no_ff(
 263             'https://www.dailymotion.com/embed/video/%s' % video_id,
 264             video_id, 'Downloading embed page')
 265
 266         timestamp = parse_iso8601(self._html_search_meta(
 267             'video:release_date', webpage, 'upload date'))
 268
 269         info = self._parse_json(
 270             self._search_regex(
 271                 r'var info = ({.*?}),$', embed_page,
 272                 'video info', flags=re.MULTILINE),
 273             video_id)
 274
 275         self._check_error(info)
 276
 277         formats = []
 278         for (key, format_id) in self._FORMATS:
 279             video_url = info.get(key)
 280             if video_url is not None:
 281                 m_size = re.search(r'H264-(\d+)x(\d+)', video_url)
 282                 if m_size is not None:
 283                     width, height = map(int_or_none, (m_size.group(1), m_size.group(2)))
 284                 else:
 285                     width, height = None, None
 286                 formats.append({
 287                     'url': video_url,
 288                     'ext': 'mp4',
 289                     'format_id': format_id,
 290                     'width': width,
 291                     'height': height,
 292                 })
 293         self._sort_formats(formats)
 294
 295         # subtitles
 296         video_subtitles = self.extract_subtitles(video_id, webpage)
 297
 298         title = self._og_search_title(webpage, default=None)
 299         if title is None:
 300             title = self._html_search_regex(
 301                 r'(?s)<span\s+id="video_title"[^>]*>(.*?)</span>', webpage,
 302                 'title')
 303
 304         return {
 305             'id': video_id,
 306             'formats': formats,
 307             'uploader': info['owner.screenname'],
 308             'timestamp': timestamp,
 309             'title': title,
 310             'description': description,
 311             'subtitles': video_subtitles,
 312             'thumbnail': info['thumbnail_url'],
 313             'age_limit': age_limit,
 314             'view_count': view_count,
 315             'duration': info['duration']
 316         }
 317
 318     def _check_error(self, info):
 319         error = info.get('error')
 320         if error:
 321             title = error.get('title') or error['message']
 322             # See https://developer.dailymotion.com/api#access-error
 323             if error.get('code') == 'DM007':
 324                 self.raise_geo_restricted(msg=title)
 325             raise ExtractorError(
 326                 '%s said: %s' % (self.IE_NAME, title), expected=True)
 327
 328     def _get_subtitles(self, video_id, webpage):
 329         try:
 330             sub_list = self._download_webpage(
 331                 'https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id,
 332                 video_id, note=False)
 333         except ExtractorError as err:
 334             self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err))
 335             return {}
 336         info = json.loads(sub_list)
 337         if (info['total'] > 0):
 338             sub_lang_list = dict((l['language'], [{'url': l['url'], 'ext': 'srt'}]) for l in info['list'])
 339             return sub_lang_list
 340         self._downloader.report_warning('video doesn\'t have subtitles')
 341         return {}
 342
 343
 344 class DailymotionPlaylistIE(DailymotionBaseInfoExtractor):
 345     IE_NAME = 'dailymotion:playlist'
 346     _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P<id>[^/?#&]+)'
 347     _MORE_PAGES_INDICATOR = r'(?s)<div class="pages[^"]*">.*?<a\s+class="[^"]*?icon-arrow_right[^"]*?"'
 348     _PAGE_TEMPLATE = 'https://www.dailymotion.com/playlist/%s/%s'
 349     _TESTS = [{
 350         'url': 'http://www.dailymotion.com/playlist/xv4bw_nqtv_sport/1#video=xl8v3q',
 351         'info_dict': {
 352             'title': 'SPORT',
 353             'id': 'xv4bw_nqtv_sport',
 354         },
 355         'playlist_mincount': 20,
 356     }]
 357
 358     def _extract_entries(self, id):
 359         video_ids = set()
 360         processed_urls = set()
 361         for pagenum in itertools.count(1):
 362             page_url = self._PAGE_TEMPLATE % (id, pagenum)
 363             webpage, urlh = self._download_webpage_handle_no_ff(
 364                 page_url, id, 'Downloading page %s' % pagenum)
 365             if urlh.geturl() in processed_urls:
 366                 self.report_warning('Stopped at duplicated page %s, which is the same as %s' % (
 367                     page_url, urlh.geturl()), id)
 368                 break
 369
 370             processed_urls.add(urlh.geturl())
 371
 372             for video_id in re.findall(r'data-xid="(.+?)"', webpage):
 373                 if video_id not in video_ids:
 374                     yield self.url_result(
 375                         'http://www.dailymotion.com/video/%s' % video_id,
 376                         DailymotionIE.ie_key(), video_id)
 377                     video_ids.add(video_id)
 378
 379             if re.search(self._MORE_PAGES_INDICATOR, webpage) is None:
 380                 break
 381
 382     def _real_extract(self, url):
 383         mobj = re.match(self._VALID_URL, url)
 384         playlist_id = mobj.group('id')
 385         webpage = self._download_webpage(url, playlist_id)
 386
 387         return {
 388             '_type': 'playlist',
 389             'id': playlist_id,
 390             'title': self._og_search_title(webpage),
 391             'entries': self._extract_entries(playlist_id),
 392         }
 393
 394
 395 class DailymotionUserIE(DailymotionPlaylistIE):
 396     IE_NAME = 'dailymotion:user'
 397     _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|swf|#|video|playlist)/)(?:(?:old/)?user/)?(?P<user>[^/]+)'
 398     _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s'
 399     _TESTS = [{
 400         'url': 'https://www.dailymotion.com/user/nqtv',
 401         'info_dict': {
 402             'id': 'nqtv',
 403             'title': 'Rémi Gaillard',
 404         },
 405         'playlist_mincount': 100,
 406     }, {
 407         'url': 'http://www.dailymotion.com/user/UnderProject',
 408         'info_dict': {
 409             'id': 'UnderProject',
 410             'title': 'UnderProject',
 411         },
 412         'playlist_mincount': 1800,
 413         'expected_warnings': [
 414             'Stopped at duplicated page',
 415         ],
 416         'skip': 'Takes too long time',
 417     }]
 418
 419     def _real_extract(self, url):
 420         mobj = re.match(self._VALID_URL, url)
 421         user = mobj.group('user')
 422         webpage = self._download_webpage(
 423             'https://www.dailymotion.com/user/%s' % user, user)
 424         full_user = unescapeHTML(self._html_search_regex(
 425             r'<a class="nav-image" title="([^"]+)" href="/%s">' % re.escape(user),
 426             webpage, 'user'))
 427
 428         return {
 429             '_type': 'playlist',
 430             'id': user,
 431             'title': full_user,
 432             'entries': self._extract_entries(user),
 433         }