yt_dlp/extractor/viki.py

   1 import hashlib
   2 import hmac
   3 import json
   4 import time
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     ExtractorError,
   9     int_or_none,
  10     parse_age_limit,
  11     parse_iso8601,
  12     try_get,
  13 )
  14
  15
  16 class VikiBaseIE(InfoExtractor):
  17     _VALID_URL_BASE = r'https?://(?:www\.)?viki\.(?:com|net|mx|jp|fr)/'
  18     _API_URL_TEMPLATE = 'https://api.viki.io%s'
  19
  20     _DEVICE_ID = '112395910d'
  21     _APP = '100005a'
  22     _APP_VERSION = '6.11.3'
  23     _APP_SECRET = 'd96704b180208dbb2efa30fe44c48bd8690441af9f567ba8fd710a72badc85198f7472'
  24
  25     _GEO_BYPASS = False
  26     _NETRC_MACHINE = 'viki'
  27
  28     _token = None
  29
  30     _ERRORS = {
  31         'geo': 'Sorry, this content is not available in your region.',
  32         'upcoming': 'Sorry, this content is not yet available.',
  33         'paywall': 'Sorry, this content is only available to Viki Pass Plus subscribers',
  34     }
  35
  36     def _stream_headers(self, timestamp, sig):
  37         return {
  38             'X-Viki-manufacturer': 'vivo',
  39             'X-Viki-device-model': 'vivo 1606',
  40             'X-Viki-device-os-ver': '6.0.1',
  41             'X-Viki-connection-type': 'WIFI',
  42             'X-Viki-carrier': '',
  43             'X-Viki-as-id': '100005a-1625321982-3932',
  44             'timestamp': str(timestamp),
  45             'signature': str(sig),
  46             'x-viki-app-ver': self._APP_VERSION
  47         }
  48
  49     def _api_query(self, path, version=4, **kwargs):
  50         path += '?' if '?' not in path else '&'
  51         query = f'/v{version}/{path}app={self._APP}'
  52         if self._token:
  53             query += '&token=%s' % self._token
  54         return query + ''.join(f'&{name}={val}' for name, val in kwargs.items())
  55
  56     def _sign_query(self, path):
  57         timestamp = int(time.time())
  58         query = self._api_query(path, version=5)
  59         sig = hmac.new(
  60             self._APP_SECRET.encode('ascii'), f'{query}&t={timestamp}'.encode('ascii'), hashlib.sha1).hexdigest()
  61         return timestamp, sig, self._API_URL_TEMPLATE % query
  62
  63     def _call_api(
  64             self, path, video_id, note='Downloading JSON metadata', data=None, query=None, fatal=True):
  65         if query is None:
  66             timestamp, sig, url = self._sign_query(path)
  67         else:
  68             url = self._API_URL_TEMPLATE % self._api_query(path, version=4)
  69         resp = self._download_json(
  70             url, video_id, note, fatal=fatal, query=query,
  71             data=json.dumps(data).encode('utf-8') if data else None,
  72             headers=({'x-viki-app-ver': self._APP_VERSION} if data
  73                      else self._stream_headers(timestamp, sig) if query is None
  74                      else None), expected_status=400) or {}
  75
  76         self._raise_error(resp.get('error'), fatal)
  77         return resp
  78
  79     def _raise_error(self, error, fatal=True):
  80         if error is None:
  81             return
  82         msg = '%s said: %s' % (self.IE_NAME, error)
  83         if fatal:
  84             raise ExtractorError(msg, expected=True)
  85         else:
  86             self.report_warning(msg)
  87
  88     def _check_errors(self, data):
  89         for reason, status in (data.get('blocking') or {}).items():
  90             if status and reason in self._ERRORS:
  91                 message = self._ERRORS[reason]
  92                 if reason == 'geo':
  93                     self.raise_geo_restricted(msg=message)
  94                 elif reason == 'paywall':
  95                     if try_get(data, lambda x: x['paywallable']['tvod']):
  96                         self._raise_error('This video is for rent only or TVOD (Transactional Video On demand)')
  97                     self.raise_login_required(message)
  98                 self._raise_error(message)
  99
 100     def _perform_login(self, username, password):
 101         self._token = self._call_api(
 102             'sessions.json', None, 'Logging in', fatal=False,
 103             data={'username': username, 'password': password}).get('token')
 104         if not self._token:
 105             self.report_warning('Login Failed: Unable to get session token')
 106
 107     @staticmethod
 108     def dict_selection(dict_obj, preferred_key):
 109         if preferred_key in dict_obj:
 110             return dict_obj[preferred_key]
 111         return (list(filter(None, dict_obj.values())) or [None])[0]
 112
 113
 114 class VikiIE(VikiBaseIE):
 115     IE_NAME = 'viki'
 116     _VALID_URL = r'%s(?:videos|player)/(?P<id>[0-9]+v)' % VikiBaseIE._VALID_URL_BASE
 117     _TESTS = [{
 118         'note': 'Free non-DRM video with storyboards in MPD',
 119         'url': 'https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1',
 120         'info_dict': {
 121             'id': '1175236v',
 122             'ext': 'mp4',
 123             'title': 'Choosing Spouse by Lottery - Episode 1',
 124             'timestamp': 1606463239,
 125             'age_limit': 13,
 126             'uploader': 'FCC',
 127             'upload_date': '20201127',
 128         },
 129     }, {
 130         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
 131         'info_dict': {
 132             'id': '1023585v',
 133             'ext': 'mp4',
 134             'title': 'Heirs - Episode 14',
 135             'uploader': 'SBS Contents Hub',
 136             'timestamp': 1385047627,
 137             'upload_date': '20131121',
 138             'age_limit': 13,
 139             'duration': 3570,
 140             'episode_number': 14,
 141         },
 142         'skip': 'Blocked in the US',
 143     }, {
 144         # clip
 145         'url': 'http://www.viki.com/videos/1067139v-the-avengers-age-of-ultron-press-conference',
 146         'md5': '86c0b5dbd4d83a6611a79987cc7a1989',
 147         'info_dict': {
 148             'id': '1067139v',
 149             'ext': 'mp4',
 150             'title': "'The Avengers: Age of Ultron' Press Conference",
 151             'description': 'md5:d70b2f9428f5488321bfe1db10d612ea',
 152             'duration': 352,
 153             'timestamp': 1430380829,
 154             'upload_date': '20150430',
 155             'uploader': 'Arirang TV',
 156             'like_count': int,
 157             'age_limit': 0,
 158         },
 159         'skip': 'Sorry. There was an error loading this video',
 160     }, {
 161         'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
 162         'info_dict': {
 163             'id': '1048879v',
 164             'ext': 'mp4',
 165             'title': 'Ankhon Dekhi',
 166             'duration': 6512,
 167             'timestamp': 1408532356,
 168             'upload_date': '20140820',
 169             'uploader': 'Spuul',
 170             'like_count': int,
 171             'age_limit': 13,
 172         },
 173         'skip': 'Blocked in the US',
 174     }, {
 175         # episode
 176         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
 177         'md5': '0a53dc252e6e690feccd756861495a8c',
 178         'info_dict': {
 179             'id': '44699v',
 180             'ext': 'mp4',
 181             'title': 'Boys Over Flowers - Episode 1',
 182             'description': 'md5:b89cf50038b480b88b5b3c93589a9076',
 183             'duration': 4172,
 184             'timestamp': 1270496524,
 185             'upload_date': '20100405',
 186             'uploader': 'group8',
 187             'like_count': int,
 188             'age_limit': 13,
 189             'episode_number': 1,
 190         },
 191     }, {
 192         # youtube external
 193         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
 194         'md5': '63f8600c1da6f01b7640eee7eca4f1da',
 195         'info_dict': {
 196             'id': '50562v',
 197             'ext': 'webm',
 198             'title': 'Poor Nastya [COMPLETE] - Episode 1',
 199             'description': '',
 200             'duration': 606,
 201             'timestamp': 1274949505,
 202             'upload_date': '20101213',
 203             'uploader': 'ad14065n',
 204             'uploader_id': 'ad14065n',
 205             'like_count': int,
 206             'age_limit': 13,
 207         },
 208         'skip': 'Page not found!',
 209     }, {
 210         'url': 'http://www.viki.com/player/44699v',
 211         'only_matching': True,
 212     }, {
 213         # non-English description
 214         'url': 'http://www.viki.com/videos/158036v-love-in-magic',
 215         'md5': '41faaba0de90483fb4848952af7c7d0d',
 216         'info_dict': {
 217             'id': '158036v',
 218             'ext': 'mp4',
 219             'uploader': 'I Planet Entertainment',
 220             'upload_date': '20111122',
 221             'timestamp': 1321985454,
 222             'description': 'md5:44b1e46619df3a072294645c770cef36',
 223             'title': 'Love In Magic',
 224             'age_limit': 13,
 225         },
 226     }]
 227
 228     def _real_extract(self, url):
 229         video_id = self._match_id(url)
 230         video = self._call_api(f'videos/{video_id}.json', video_id, 'Downloading video JSON', query={})
 231         self._check_errors(video)
 232
 233         title = try_get(video, lambda x: x['titles']['en'], str)
 234         episode_number = int_or_none(video.get('number'))
 235         if not title:
 236             title = 'Episode %d' % episode_number if video.get('type') == 'episode' else video.get('id') or video_id
 237             container_titles = try_get(video, lambda x: x['container']['titles'], dict) or {}
 238             container_title = self.dict_selection(container_titles, 'en')
 239             title = '%s - %s' % (container_title, title)
 240
 241         thumbnails = [{
 242             'id': thumbnail_id,
 243             'url': thumbnail['url'],
 244         } for thumbnail_id, thumbnail in (video.get('images') or {}).items() if thumbnail.get('url')]
 245
 246         resp = self._call_api(
 247             'playback_streams/%s.json?drms=dt3&device_id=%s' % (video_id, self._DEVICE_ID),
 248             video_id, 'Downloading video streams JSON')['main'][0]
 249
 250         stream_id = try_get(resp, lambda x: x['properties']['track']['stream_id'])
 251         subtitles = dict((lang, [{
 252             'ext': ext,
 253             'url': self._API_URL_TEMPLATE % self._api_query(
 254                 f'videos/{video_id}/auth_subtitles/{lang}.{ext}', stream_id=stream_id)
 255         } for ext in ('srt', 'vtt')]) for lang in (video.get('subtitle_completions') or {}).keys())
 256
 257         mpd_url = resp['url']
 258         # 720p is hidden in another MPD which can be found in the current manifest content
 259         mpd_content = self._download_webpage(mpd_url, video_id, note='Downloading initial MPD manifest')
 260         mpd_url = self._search_regex(
 261             r'(?mi)<BaseURL>(http.+.mpd)', mpd_content, 'new manifest', default=mpd_url)
 262         if 'mpdhd_high' not in mpd_url and 'sig=' not in mpd_url:
 263             # Modify the URL to get 1080p
 264             mpd_url = mpd_url.replace('mpdhd', 'mpdhd_high')
 265         formats = self._extract_mpd_formats(mpd_url, video_id)
 266
 267         return {
 268             'id': video_id,
 269             'formats': formats,
 270             'title': title,
 271             'description': self.dict_selection(video.get('descriptions', {}), 'en'),
 272             'duration': int_or_none(video.get('duration')),
 273             'timestamp': parse_iso8601(video.get('created_at')),
 274             'uploader': video.get('author'),
 275             'uploader_url': video.get('author_url'),
 276             'like_count': int_or_none(try_get(video, lambda x: x['likes']['count'])),
 277             'age_limit': parse_age_limit(video.get('rating')),
 278             'thumbnails': thumbnails,
 279             'subtitles': subtitles,
 280             'episode_number': episode_number,
 281         }
 282
 283
 284 class VikiChannelIE(VikiBaseIE):
 285     IE_NAME = 'viki:channel'
 286     _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
 287     _TESTS = [{
 288         'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
 289         'info_dict': {
 290             'id': '50c',
 291             'title': 'Boys Over Flowers',
 292             'description': 'md5:804ce6e7837e1fd527ad2f25420f4d59',
 293         },
 294         'playlist_mincount': 51,
 295     }, {
 296         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
 297         'info_dict': {
 298             'id': '1354c',
 299             'title': 'Poor Nastya [COMPLETE]',
 300             'description': 'md5:05bf5471385aa8b21c18ad450e350525',
 301         },
 302         'playlist_count': 127,
 303         'skip': 'Page not found',
 304     }, {
 305         'url': 'http://www.viki.com/news/24569c-showbiz-korea',
 306         'only_matching': True,
 307     }, {
 308         'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
 309         'only_matching': True,
 310     }, {
 311         'url': 'http://www.viki.com/artists/2141c-shinee',
 312         'only_matching': True,
 313     }]
 314
 315     _video_types = ('episodes', 'movies', 'clips', 'trailers')
 316
 317     def _entries(self, channel_id):
 318         params = {
 319             'app': self._APP, 'token': self._token, 'only_ids': 'true',
 320             'direction': 'asc', 'sort': 'number', 'per_page': 30
 321         }
 322         video_types = self._configuration_arg('video_types') or self._video_types
 323         for video_type in video_types:
 324             if video_type not in self._video_types:
 325                 self.report_warning(f'Unknown video_type: {video_type}')
 326             page_num = 0
 327             while True:
 328                 page_num += 1
 329                 params['page'] = page_num
 330                 res = self._call_api(
 331                     f'containers/{channel_id}/{video_type}.json', channel_id, query=params, fatal=False,
 332                     note='Downloading %s JSON page %d' % (video_type.title(), page_num))
 333
 334                 for video_id in res.get('response') or []:
 335                     yield self.url_result(f'https://www.viki.com/videos/{video_id}', VikiIE.ie_key(), video_id)
 336                 if not res.get('more'):
 337                     break
 338
 339     def _real_extract(self, url):
 340         channel_id = self._match_id(url)
 341         channel = self._call_api('containers/%s.json' % channel_id, channel_id, 'Downloading channel JSON')
 342         self._check_errors(channel)
 343         return self.playlist_result(
 344             self._entries(channel_id), channel_id,
 345             self.dict_selection(channel['titles'], 'en'),
 346             self.dict_selection(channel['descriptions'], 'en'))