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 = '86085977d'  # used for android api
  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 _real_initialize(self):
 103         self._login()
 104
 105     def _login(self):
 106         username, password = self._get_login_info()
 107         if username is None:
 108             return
 109
 110         self._token = self._call_api(
 111             'sessions.json', None, 'Logging in', fatal=False,
 112             data={'username': username, 'password': password}).get('token')
 113         if not self._token:
 114             self.report_warning('Login Failed: Unable to get session token')
 115
 116     @staticmethod
 117     def dict_selection(dict_obj, preferred_key):
 118         if preferred_key in dict_obj:
 119             return dict_obj[preferred_key]
 120         return (list(filter(None, dict_obj.values())) or [None])[0]
 121
 122
 123 class VikiIE(VikiBaseIE):
 124     IE_NAME = 'viki'
 125     _VALID_URL = r'%s(?:videos|player)/(?P<id>[0-9]+v)' % VikiBaseIE._VALID_URL_BASE
 126     _TESTS = [{
 127         'note': 'Free non-DRM video with storyboards in MPD',
 128         'url': 'https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1',
 129         'info_dict': {
 130             'id': '1175236v',
 131             'ext': 'mp4',
 132             'title': 'Choosing Spouse by Lottery - Episode 1',
 133             'timestamp': 1606463239,
 134             'age_limit': 13,
 135             'uploader': 'FCC',
 136             'upload_date': '20201127',
 137         },
 138     }, {
 139         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
 140         'info_dict': {
 141             'id': '1023585v',
 142             'ext': 'mp4',
 143             'title': 'Heirs - Episode 14',
 144             'uploader': 'SBS Contents Hub',
 145             'timestamp': 1385047627,
 146             'upload_date': '20131121',
 147             'age_limit': 13,
 148             'duration': 3570,
 149             'episode_number': 14,
 150         },
 151         'skip': 'Blocked in the US',
 152     }, {
 153         # clip
 154         'url': 'http://www.viki.com/videos/1067139v-the-avengers-age-of-ultron-press-conference',
 155         'md5': '86c0b5dbd4d83a6611a79987cc7a1989',
 156         'info_dict': {
 157             'id': '1067139v',
 158             'ext': 'mp4',
 159             'title': "'The Avengers: Age of Ultron' Press Conference",
 160             'description': 'md5:d70b2f9428f5488321bfe1db10d612ea',
 161             'duration': 352,
 162             'timestamp': 1430380829,
 163             'upload_date': '20150430',
 164             'uploader': 'Arirang TV',
 165             'like_count': int,
 166             'age_limit': 0,
 167         },
 168         'skip': 'Sorry. There was an error loading this video',
 169     }, {
 170         'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
 171         'info_dict': {
 172             'id': '1048879v',
 173             'ext': 'mp4',
 174             'title': 'Ankhon Dekhi',
 175             'duration': 6512,
 176             'timestamp': 1408532356,
 177             'upload_date': '20140820',
 178             'uploader': 'Spuul',
 179             'like_count': int,
 180             'age_limit': 13,
 181         },
 182         'skip': 'Blocked in the US',
 183     }, {
 184         # episode
 185         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
 186         'md5': '0a53dc252e6e690feccd756861495a8c',
 187         'info_dict': {
 188             'id': '44699v',
 189             'ext': 'mp4',
 190             'title': 'Boys Over Flowers - Episode 1',
 191             'description': 'md5:b89cf50038b480b88b5b3c93589a9076',
 192             'duration': 4172,
 193             'timestamp': 1270496524,
 194             'upload_date': '20100405',
 195             'uploader': 'group8',
 196             'like_count': int,
 197             'age_limit': 13,
 198             'episode_number': 1,
 199         },
 200     }, {
 201         # youtube external
 202         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
 203         'md5': '63f8600c1da6f01b7640eee7eca4f1da',
 204         'info_dict': {
 205             'id': '50562v',
 206             'ext': 'webm',
 207             'title': 'Poor Nastya [COMPLETE] - Episode 1',
 208             'description': '',
 209             'duration': 606,
 210             'timestamp': 1274949505,
 211             'upload_date': '20101213',
 212             'uploader': 'ad14065n',
 213             'uploader_id': 'ad14065n',
 214             'like_count': int,
 215             'age_limit': 13,
 216         },
 217         'skip': 'Page not found!',
 218     }, {
 219         'url': 'http://www.viki.com/player/44699v',
 220         'only_matching': True,
 221     }, {
 222         # non-English description
 223         'url': 'http://www.viki.com/videos/158036v-love-in-magic',
 224         'md5': '41faaba0de90483fb4848952af7c7d0d',
 225         'info_dict': {
 226             'id': '158036v',
 227             'ext': 'mp4',
 228             'uploader': 'I Planet Entertainment',
 229             'upload_date': '20111122',
 230             'timestamp': 1321985454,
 231             'description': 'md5:44b1e46619df3a072294645c770cef36',
 232             'title': 'Love In Magic',
 233             'age_limit': 13,
 234         },
 235     }]
 236
 237     def _real_extract(self, url):
 238         video_id = self._match_id(url)
 239         video = self._call_api(f'videos/{video_id}.json', video_id, 'Downloading video JSON', query={})
 240         self._check_errors(video)
 241
 242         title = try_get(video, lambda x: x['titles']['en'], str)
 243         episode_number = int_or_none(video.get('number'))
 244         if not title:
 245             title = 'Episode %d' % episode_number if video.get('type') == 'episode' else video.get('id') or video_id
 246             container_titles = try_get(video, lambda x: x['container']['titles'], dict) or {}
 247             container_title = self.dict_selection(container_titles, 'en')
 248             title = '%s - %s' % (container_title, title)
 249
 250         thumbnails = [{
 251             'id': thumbnail_id,
 252             'url': thumbnail['url'],
 253         } for thumbnail_id, thumbnail in (video.get('images') or {}).items() if thumbnail.get('url')]
 254
 255         resp = self._call_api(
 256             'playback_streams/%s.json?drms=dt1,dt2&device_id=%s' % (video_id, self._DEVICE_ID),
 257             video_id, 'Downloading video streams JSON')['main'][0]
 258
 259         stream_id = try_get(resp, lambda x: x['properties']['track']['stream_id'])
 260         subtitles = dict((lang, [{
 261             'ext': ext,
 262             'url': self._API_URL_TEMPLATE % self._api_query(
 263                 f'videos/{video_id}/auth_subtitles/{lang}.{ext}', stream_id=stream_id)
 264         } for ext in ('srt', 'vtt')]) for lang in (video.get('subtitle_completions') or {}).keys())
 265
 266         mpd_url = resp['url']
 267         # 1080p is hidden in another mpd which can be found in the current manifest content
 268         mpd_content = self._download_webpage(mpd_url, video_id, note='Downloading initial MPD manifest')
 269         mpd_url = self._search_regex(
 270             r'(?mi)<BaseURL>(http.+.mpd)', mpd_content, 'new manifest', default=mpd_url)
 271         formats = self._extract_mpd_formats(mpd_url, video_id)
 272         self._sort_formats(formats)
 273
 274         return {
 275             'id': video_id,
 276             'formats': formats,
 277             'title': title,
 278             'description': self.dict_selection(video.get('descriptions', {}), 'en'),
 279             'duration': int_or_none(video.get('duration')),
 280             'timestamp': parse_iso8601(video.get('created_at')),
 281             'uploader': video.get('author'),
 282             'uploader_url': video.get('author_url'),
 283             'like_count': int_or_none(try_get(video, lambda x: x['likes']['count'])),
 284             'age_limit': parse_age_limit(video.get('rating')),
 285             'thumbnails': thumbnails,
 286             'subtitles': subtitles,
 287             'episode_number': episode_number,
 288         }
 289
 290
 291 class VikiChannelIE(VikiBaseIE):
 292     IE_NAME = 'viki:channel'
 293     _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
 294     _TESTS = [{
 295         'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
 296         'info_dict': {
 297             'id': '50c',
 298             'title': 'Boys Over Flowers',
 299             'description': 'md5:804ce6e7837e1fd527ad2f25420f4d59',
 300         },
 301         'playlist_mincount': 51,
 302     }, {
 303         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
 304         'info_dict': {
 305             'id': '1354c',
 306             'title': 'Poor Nastya [COMPLETE]',
 307             'description': 'md5:05bf5471385aa8b21c18ad450e350525',
 308         },
 309         'playlist_count': 127,
 310         'skip': 'Page not found',
 311     }, {
 312         'url': 'http://www.viki.com/news/24569c-showbiz-korea',
 313         'only_matching': True,
 314     }, {
 315         'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
 316         'only_matching': True,
 317     }, {
 318         'url': 'http://www.viki.com/artists/2141c-shinee',
 319         'only_matching': True,
 320     }]
 321
 322     _video_types = ('episodes', 'movies', 'clips', 'trailers')
 323
 324     def _entries(self, channel_id):
 325         params = {
 326             'app': self._APP, 'token': self._token, 'only_ids': 'true',
 327             'direction': 'asc', 'sort': 'number', 'per_page': 30
 328         }
 329         video_types = self._configuration_arg('video_types') or self._video_types
 330         for video_type in video_types:
 331             if video_type not in self._video_types:
 332                 self.report_warning(f'Unknown video_type: {video_type}')
 333             page_num = 0
 334             while True:
 335                 page_num += 1
 336                 params['page'] = page_num
 337                 res = self._call_api(
 338                     f'containers/{channel_id}/{video_type}.json', channel_id, query=params, fatal=False,
 339                     note='Downloading %s JSON page %d' % (video_type.title(), page_num))
 340
 341                 for video_id in res.get('response') or []:
 342                     yield self.url_result(f'https://www.viki.com/videos/{video_id}', VikiIE.ie_key(), video_id)
 343                 if not res.get('more'):
 344                     break
 345
 346     def _real_extract(self, url):
 347         channel_id = self._match_id(url)
 348         channel = self._call_api('containers/%s.json' % channel_id, channel_id, 'Downloading channel JSON')
 349         self._check_errors(channel)
 350         return self.playlist_result(
 351             self._entries(channel_id), channel_id,
 352             self.dict_selection(channel['titles'], 'en'),
 353             self.dict_selection(channel['descriptions'], 'en'))