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         'params': {
 139             'format': 'bestvideo',
 140         },
 141     }, {
 142         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
 143         'info_dict': {
 144             'id': '1023585v',
 145             'ext': 'mp4',
 146             'title': 'Heirs - Episode 14',
 147             'uploader': 'SBS Contents Hub',
 148             'timestamp': 1385047627,
 149             'upload_date': '20131121',
 150             'age_limit': 13,
 151             'duration': 3570,
 152             'episode_number': 14,
 153         },
 154         'params': {
 155             'format': 'bestvideo',
 156         },
 157         'skip': 'Blocked in the US',
 158     }, {
 159         # clip
 160         'url': 'http://www.viki.com/videos/1067139v-the-avengers-age-of-ultron-press-conference',
 161         'md5': '86c0b5dbd4d83a6611a79987cc7a1989',
 162         'info_dict': {
 163             'id': '1067139v',
 164             'ext': 'mp4',
 165             'title': "'The Avengers: Age of Ultron' Press Conference",
 166             'description': 'md5:d70b2f9428f5488321bfe1db10d612ea',
 167             'duration': 352,
 168             'timestamp': 1430380829,
 169             'upload_date': '20150430',
 170             'uploader': 'Arirang TV',
 171             'like_count': int,
 172             'age_limit': 0,
 173         },
 174         'skip': 'Sorry. There was an error loading this video',
 175     }, {
 176         'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
 177         'info_dict': {
 178             'id': '1048879v',
 179             'ext': 'mp4',
 180             'title': 'Ankhon Dekhi',
 181             'duration': 6512,
 182             'timestamp': 1408532356,
 183             'upload_date': '20140820',
 184             'uploader': 'Spuul',
 185             'like_count': int,
 186             'age_limit': 13,
 187         },
 188         'skip': 'Blocked in the US',
 189     }, {
 190         # episode
 191         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
 192         'md5': '0a53dc252e6e690feccd756861495a8c',
 193         'info_dict': {
 194             'id': '44699v',
 195             'ext': 'mp4',
 196             'title': 'Boys Over Flowers - Episode 1',
 197             'description': 'md5:b89cf50038b480b88b5b3c93589a9076',
 198             'duration': 4172,
 199             'timestamp': 1270496524,
 200             'upload_date': '20100405',
 201             'uploader': 'group8',
 202             'like_count': int,
 203             'age_limit': 13,
 204             'episode_number': 1,
 205         },
 206         'params': {
 207             'format': 'bestvideo',
 208         },
 209     }, {
 210         # youtube external
 211         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
 212         'md5': '63f8600c1da6f01b7640eee7eca4f1da',
 213         'info_dict': {
 214             'id': '50562v',
 215             'ext': 'webm',
 216             'title': 'Poor Nastya [COMPLETE] - Episode 1',
 217             'description': '',
 218             'duration': 606,
 219             'timestamp': 1274949505,
 220             'upload_date': '20101213',
 221             'uploader': 'ad14065n',
 222             'uploader_id': 'ad14065n',
 223             'like_count': int,
 224             'age_limit': 13,
 225         },
 226         'skip': 'Page not found!',
 227     }, {
 228         'url': 'http://www.viki.com/player/44699v',
 229         'only_matching': True,
 230     }, {
 231         # non-English description
 232         'url': 'http://www.viki.com/videos/158036v-love-in-magic',
 233         'md5': '41faaba0de90483fb4848952af7c7d0d',
 234         'info_dict': {
 235             'id': '158036v',
 236             'ext': 'mp4',
 237             'uploader': 'I Planet Entertainment',
 238             'upload_date': '20111122',
 239             'timestamp': 1321985454,
 240             'description': 'md5:44b1e46619df3a072294645c770cef36',
 241             'title': 'Love In Magic',
 242             'age_limit': 13,
 243         },
 244         'params': {
 245             'format': 'bestvideo',
 246         },
 247     }]
 248
 249     def _real_extract(self, url):
 250         video_id = self._match_id(url)
 251         video = self._call_api(f'videos/{video_id}.json', video_id, 'Downloading video JSON', query={})
 252         self._check_errors(video)
 253
 254         title = try_get(video, lambda x: x['titles']['en'], str)
 255         episode_number = int_or_none(video.get('number'))
 256         if not title:
 257             title = 'Episode %d' % episode_number if video.get('type') == 'episode' else video.get('id') or video_id
 258             container_titles = try_get(video, lambda x: x['container']['titles'], dict) or {}
 259             container_title = self.dict_selection(container_titles, 'en')
 260             title = '%s - %s' % (container_title, title)
 261
 262         thumbnails = [{
 263             'id': thumbnail_id,
 264             'url': thumbnail['url'],
 265         } for thumbnail_id, thumbnail in (video.get('images') or {}).items() if thumbnail.get('url')]
 266
 267         resp = self._call_api(
 268             'playback_streams/%s.json?drms=dt1,dt2&device_id=%s' % (video_id, self._DEVICE_ID),
 269             video_id, 'Downloading video streams JSON')['main'][0]
 270
 271         stream_id = try_get(resp, lambda x: x['properties']['track']['stream_id'])
 272         subtitles = dict((lang, [{
 273             'ext': ext,
 274             'url': self._API_URL_TEMPLATE % self._api_query(
 275                 f'videos/{video_id}/auth_subtitles/{lang}.{ext}', stream_id=stream_id)
 276         } for ext in ('srt', 'vtt')]) for lang in (video.get('subtitle_completions') or {}).keys())
 277
 278         mpd_url = resp['url']
 279         # 1080p is hidden in another mpd which can be found in the current manifest content
 280         mpd_content = self._download_webpage(mpd_url, video_id, note='Downloading initial MPD manifest')
 281         mpd_url = self._search_regex(
 282             r'(?mi)<BaseURL>(http.+.mpd)', mpd_content, 'new manifest', default=mpd_url)
 283         formats = self._extract_mpd_formats(mpd_url, video_id)
 284         self._sort_formats(formats)
 285
 286         return {
 287             'id': video_id,
 288             'formats': formats,
 289             'title': title,
 290             'description': self.dict_selection(video.get('descriptions', {}), 'en'),
 291             'duration': int_or_none(video.get('duration')),
 292             'timestamp': parse_iso8601(video.get('created_at')),
 293             'uploader': video.get('author'),
 294             'uploader_url': video.get('author_url'),
 295             'like_count': int_or_none(try_get(video, lambda x: x['likes']['count'])),
 296             'age_limit': parse_age_limit(video.get('rating')),
 297             'thumbnails': thumbnails,
 298             'subtitles': subtitles,
 299             'episode_number': episode_number,
 300         }
 301
 302
 303 class VikiChannelIE(VikiBaseIE):
 304     IE_NAME = 'viki:channel'
 305     _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
 306     _TESTS = [{
 307         'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
 308         'info_dict': {
 309             'id': '50c',
 310             'title': 'Boys Over Flowers',
 311             'description': 'md5:804ce6e7837e1fd527ad2f25420f4d59',
 312         },
 313         'playlist_mincount': 51,
 314     }, {
 315         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
 316         'info_dict': {
 317             'id': '1354c',
 318             'title': 'Poor Nastya [COMPLETE]',
 319             'description': 'md5:05bf5471385aa8b21c18ad450e350525',
 320         },
 321         'playlist_count': 127,
 322         'skip': 'Page not found',
 323     }, {
 324         'url': 'http://www.viki.com/news/24569c-showbiz-korea',
 325         'only_matching': True,
 326     }, {
 327         'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
 328         'only_matching': True,
 329     }, {
 330         'url': 'http://www.viki.com/artists/2141c-shinee',
 331         'only_matching': True,
 332     }]
 333
 334     _video_types = ('episodes', 'movies', 'clips', 'trailers')
 335
 336     def _entries(self, channel_id):
 337         params = {
 338             'app': self._APP, 'token': self._token, 'only_ids': 'true',
 339             'direction': 'asc', 'sort': 'number', 'per_page': 30
 340         }
 341         video_types = self._configuration_arg('video_types') or self._video_types
 342         for video_type in video_types:
 343             if video_type not in self._video_types:
 344                 self.report_warning(f'Unknown video_type: {video_type}')
 345             page_num = 0
 346             while True:
 347                 page_num += 1
 348                 params['page'] = page_num
 349                 res = self._call_api(
 350                     f'containers/{channel_id}/{video_type}.json', channel_id, query=params, fatal=False,
 351                     note='Downloading %s JSON page %d' % (video_type.title(), page_num))
 352
 353                 for video_id in res.get('response') or []:
 354                     yield self.url_result(f'https://www.viki.com/videos/{video_id}', VikiIE.ie_key(), video_id)
 355                 if not res.get('more'):
 356                     break
 357
 358     def _real_extract(self, url):
 359         channel_id = self._match_id(url)
 360         channel = self._call_api('containers/%s.json' % channel_id, channel_id, 'Downloading channel JSON')
 361         self._check_errors(channel)
 362         return self.playlist_result(
 363             self._entries(channel_id), channel_id,
 364             self.dict_selection(channel['titles'], 'en'),
 365             self.dict_selection(channel['descriptions'], 'en'))