yt_dlp/extractor/viki.py

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