youtube_dlc/extractor/viki.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import base64
   5 import hashlib
   6 import hmac
   7 import itertools
   8 import json
   9 import re
  10 import time
  11
  12 from .common import InfoExtractor
  13 from ..compat import (
  14     compat_parse_qs,
  15     compat_urllib_parse_urlparse,
  16 )
  17 from ..utils import (
  18     ExtractorError,
  19     int_or_none,
  20     HEADRequest,
  21     parse_age_limit,
  22     parse_iso8601,
  23     sanitized_Request,
  24     std_headers,
  25 )
  26
  27
  28 class VikiBaseIE(InfoExtractor):
  29     _VALID_URL_BASE = r'https?://(?:www\.)?viki\.(?:com|net|mx|jp|fr)/'
  30     _API_QUERY_TEMPLATE = '/v4/%sapp=%s&t=%s&site=www.viki.com'
  31     _API_URL_TEMPLATE = 'https://api.viki.io%s&sig=%s'
  32
  33     _APP = '100005a'
  34     _APP_VERSION = '2.2.5.1428709186'
  35     _APP_SECRET = 'MM_d*yP@`&1@]@!AVrXf_o-HVEnoTnm$O-ti4[G~$JDI/Dc-&piU&z&5.;:}95=Iad'
  36
  37     _GEO_BYPASS = False
  38     _NETRC_MACHINE = 'viki'
  39
  40     _token = None
  41
  42     _ERRORS = {
  43         'geo': 'Sorry, this content is not available in your region.',
  44         'upcoming': 'Sorry, this content is not yet available.',
  45         # 'paywall': 'paywall',
  46     }
  47
  48     def _prepare_call(self, path, timestamp=None, post_data=None):
  49         path += '?' if '?' not in path else '&'
  50         if not timestamp:
  51             timestamp = int(time.time())
  52         query = self._API_QUERY_TEMPLATE % (path, self._APP, timestamp)
  53         if self._token:
  54             query += '&token=%s' % self._token
  55         sig = hmac.new(
  56             self._APP_SECRET.encode('ascii'),
  57             query.encode('ascii'),
  58             hashlib.sha1
  59         ).hexdigest()
  60         url = self._API_URL_TEMPLATE % (query, sig)
  61         return sanitized_Request(
  62             url, json.dumps(post_data).encode('utf-8')) if post_data else url
  63
  64     def _call_api(self, path, video_id, note, timestamp=None, post_data=None):
  65         resp = self._download_json(
  66             self._prepare_call(path, timestamp, post_data), video_id, note)
  67
  68         error = resp.get('error')
  69         if error:
  70             if error == 'invalid timestamp':
  71                 resp = self._download_json(
  72                     self._prepare_call(path, int(resp['current_timestamp']), post_data),
  73                     video_id, '%s (retry)' % note)
  74                 error = resp.get('error')
  75             if error:
  76                 self._raise_error(resp['error'])
  77
  78         return resp
  79
  80     def _raise_error(self, error):
  81         raise ExtractorError(
  82             '%s returned error: %s' % (self.IE_NAME, error),
  83             expected=True)
  84
  85     def _check_errors(self, data):
  86         for reason, status in data.get('blocking', {}).items():
  87             if status and reason in self._ERRORS:
  88                 message = self._ERRORS[reason]
  89                 if reason == 'geo':
  90                     self.raise_geo_restricted(msg=message)
  91                 raise ExtractorError('%s said: %s' % (
  92                     self.IE_NAME, message), expected=True)
  93
  94     def _real_initialize(self):
  95         self._login()
  96
  97     def _login(self):
  98         username, password = self._get_login_info()
  99         if username is None:
 100             return
 101
 102         login_form = {
 103             'login_id': username,
 104             'password': password,
 105         }
 106
 107         login = self._call_api(
 108             'sessions.json', None,
 109             'Logging in', post_data=login_form)
 110
 111         self._token = login.get('token')
 112         if not self._token:
 113             self.report_warning('Unable to get session token, login has probably failed')
 114
 115     @staticmethod
 116     def dict_selection(dict_obj, preferred_key, allow_fallback=True):
 117         if preferred_key in dict_obj:
 118             return dict_obj.get(preferred_key)
 119
 120         if not allow_fallback:
 121             return
 122
 123         filtered_dict = list(filter(None, [dict_obj.get(k) for k in dict_obj.keys()]))
 124         return filtered_dict[0] if filtered_dict else None
 125
 126
 127 class VikiIE(VikiBaseIE):
 128     IE_NAME = 'viki'
 129     _VALID_URL = r'%s(?:videos|player)/(?P<id>[0-9]+v)' % VikiBaseIE._VALID_URL_BASE
 130     _TESTS = [{
 131         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
 132         'info_dict': {
 133             'id': '1023585v',
 134             'ext': 'mp4',
 135             'title': 'Heirs Episode 14',
 136             'uploader': 'SBS',
 137             'description': 'md5:c4b17b9626dd4b143dcc4d855ba3474e',
 138             'upload_date': '20131121',
 139             'age_limit': 13,
 140         },
 141         'skip': 'Blocked in the US',
 142     }, {
 143         # clip
 144         'url': 'http://www.viki.com/videos/1067139v-the-avengers-age-of-ultron-press-conference',
 145         'md5': '86c0b5dbd4d83a6611a79987cc7a1989',
 146         'info_dict': {
 147             'id': '1067139v',
 148             'ext': 'mp4',
 149             'title': "'The Avengers: Age of Ultron' Press Conference",
 150             'description': 'md5:d70b2f9428f5488321bfe1db10d612ea',
 151             'duration': 352,
 152             'timestamp': 1430380829,
 153             'upload_date': '20150430',
 154             'uploader': 'Arirang TV',
 155             'like_count': int,
 156             'age_limit': 0,
 157         }
 158     }, {
 159         'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
 160         'info_dict': {
 161             'id': '1048879v',
 162             'ext': 'mp4',
 163             'title': 'Ankhon Dekhi',
 164             'duration': 6512,
 165             'timestamp': 1408532356,
 166             'upload_date': '20140820',
 167             'uploader': 'Spuul',
 168             'like_count': int,
 169             'age_limit': 13,
 170         },
 171         'skip': 'Blocked in the US',
 172     }, {
 173         # episode
 174         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
 175         'md5': '94e0e34fd58f169f40c184f232356cfe',
 176         'info_dict': {
 177             'id': '44699v',
 178             'ext': 'mp4',
 179             'title': 'Boys Over Flowers - Episode 1',
 180             'description': 'md5:b89cf50038b480b88b5b3c93589a9076',
 181             'duration': 4172,
 182             'timestamp': 1270496524,
 183             'upload_date': '20100405',
 184             'uploader': 'group8',
 185             'like_count': int,
 186             'age_limit': 13,
 187         },
 188         'expected_warnings': ['Unknown MIME type image/jpeg in DASH manifest'],
 189     }, {
 190         # youtube external
 191         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
 192         'md5': '63f8600c1da6f01b7640eee7eca4f1da',
 193         'info_dict': {
 194             'id': '50562v',
 195             'ext': 'webm',
 196             'title': 'Poor Nastya [COMPLETE] - Episode 1',
 197             'description': '',
 198             'duration': 606,
 199             'timestamp': 1274949505,
 200             'upload_date': '20101213',
 201             'uploader': 'ad14065n',
 202             'uploader_id': 'ad14065n',
 203             'like_count': int,
 204             'age_limit': 13,
 205         },
 206         'skip': 'Page not found!',
 207     }, {
 208         'url': 'http://www.viki.com/player/44699v',
 209         'only_matching': True,
 210     }, {
 211         # non-English description
 212         'url': 'http://www.viki.com/videos/158036v-love-in-magic',
 213         'md5': 'adf9e321a0ae5d0aace349efaaff7691',
 214         'info_dict': {
 215             'id': '158036v',
 216             'ext': 'mp4',
 217             'uploader': 'I Planet Entertainment',
 218             'upload_date': '20111122',
 219             'timestamp': 1321985454,
 220             'description': 'md5:44b1e46619df3a072294645c770cef36',
 221             'title': 'Love In Magic',
 222             'age_limit': 13,
 223         },
 224     }]
 225
 226     def _real_extract(self, url):
 227         video_id = self._match_id(url)
 228
 229         resp = self._download_json(
 230             'https://www.viki.com/api/videos/' + video_id,
 231             video_id, 'Downloading video JSON', headers={
 232                 'x-client-user-agent': std_headers['User-Agent'],
 233                 'x-viki-app-ver': '4.0.57',
 234             })
 235         video = resp['video']
 236
 237         self._check_errors(video)
 238
 239         title = self.dict_selection(video.get('titles', {}), 'en', allow_fallback=False)
 240         if not title:
 241             title = 'Episode %d' % video.get('number') if video.get('type') == 'episode' else video.get('id') or video_id
 242             container_titles = video.get('container', {}).get('titles', {})
 243             container_title = self.dict_selection(container_titles, 'en')
 244             title = '%s - %s' % (container_title, title)
 245
 246         description = self.dict_selection(video.get('descriptions', {}), 'en')
 247
 248         duration = int_or_none(video.get('duration'))
 249         timestamp = parse_iso8601(video.get('created_at'))
 250         uploader = video.get('author')
 251         like_count = int_or_none(video.get('likes', {}).get('count'))
 252         age_limit = parse_age_limit(video.get('rating'))
 253
 254         thumbnails = []
 255         for thumbnail_id, thumbnail in video.get('images', {}).items():
 256             thumbnails.append({
 257                 'id': thumbnail_id,
 258                 'url': thumbnail.get('url'),
 259             })
 260
 261         subtitles = {}
 262         try:
 263             # New way to fetch subtitles
 264             new_video = self._download_json(
 265                 'https://www.viki.com/api/videos/%s' % video_id, video_id,
 266                 'Downloading new video JSON to get subtitles', fatal=False)
 267             for sub in new_video.get('streamSubtitles').get('dash'):
 268                 subtitles[sub.get('srclang')] = [{
 269                     'ext': 'vtt',
 270                     'url': sub.get('src'),
 271                     'completion': sub.get('percentage'),
 272                 }]
 273         except AttributeError:
 274             # fall-back to the old way if there isn't a streamSubtitles attribute
 275             for subtitle_lang, _ in video.get('subtitle_completions', {}).items():
 276                 subtitles[subtitle_lang] = [{
 277                     'ext': subtitles_format,
 278                     'url': self._prepare_call(
 279                         'videos/%s/subtitles/%s.%s' % (video_id, subtitle_lang, subtitles_format)),
 280                 } for subtitles_format in ('srt', 'vtt')]
 281
 282         result = {
 283             'id': video_id,
 284             'title': title,
 285             'description': description,
 286             'duration': duration,
 287             'timestamp': timestamp,
 288             'uploader': uploader,
 289             'like_count': like_count,
 290             'age_limit': age_limit,
 291             'thumbnails': thumbnails,
 292             'subtitles': subtitles,
 293         }
 294
 295         formats = []
 296
 297         def add_format(format_id, format_dict, protocol='http'):
 298             # rtmps URLs does not seem to work
 299             if protocol == 'rtmps':
 300                 return
 301             format_url = format_dict.get('url')
 302             if not format_url:
 303                 return
 304             format_drms = format_dict.get('drms')
 305             format_stream_id = format_dict.get('id')
 306             qs = compat_parse_qs(compat_urllib_parse_urlparse(format_url).query)
 307             stream = qs.get('stream', [None])[0]
 308             if stream:
 309                 format_url = base64.b64decode(stream).decode()
 310             if format_id in ('m3u8', 'hls'):
 311                 m3u8_formats = self._extract_m3u8_formats(
 312                     format_url, video_id, 'mp4',
 313                     entry_protocol='m3u8_native',
 314                     m3u8_id='m3u8-%s' % protocol, fatal=False)
 315                 # Despite CODECS metadata in m3u8 all video-only formats
 316                 # are actually video+audio
 317                 for f in m3u8_formats:
 318                     if '_drm/index_' in f['url']:
 319                         continue
 320                     if f.get('acodec') == 'none' and f.get('vcodec') != 'none':
 321                         f['acodec'] = None
 322                     formats.append(f)
 323             elif format_id in ('mpd', 'dash'):
 324                 formats.extend(self._extract_mpd_formats(
 325                     format_url, video_id, 'mpd-%s' % protocol, fatal=False))
 326             elif format_url.startswith('rtmp'):
 327                 mobj = re.search(
 328                     r'^(?P<url>rtmp://[^/]+/(?P<app>.+?))/(?P<playpath>mp4:.+)$',
 329                     format_url)
 330                 if not mobj:
 331                     return
 332                 formats.append({
 333                     'format_id': 'rtmp-%s' % format_id,
 334                     'ext': 'flv',
 335                     'url': mobj.group('url'),
 336                     'play_path': mobj.group('playpath'),
 337                     'app': mobj.group('app'),
 338                     'page_url': url,
 339                     'drms': format_drms,
 340                     'stream_id': format_stream_id,
 341                 })
 342             else:
 343                 urlh = self._request_webpage(
 344                     HEADRequest(format_url), video_id, 'Checking file size', fatal=False)
 345                 formats.append({
 346                     'url': format_url,
 347                     'format_id': '%s-%s' % (format_id, protocol),
 348                     'height': int_or_none(self._search_regex(
 349                         r'^(\d+)[pP]$', format_id, 'height', default=None)),
 350                     'drms': format_drms,
 351                     'stream_id': format_stream_id,
 352                     'filesize': int_or_none(urlh.headers.get('Content-Length')),
 353                 })
 354
 355         for format_id, format_dict in (resp.get('streams') or {}).items():
 356             add_format(format_id, format_dict)
 357         if not formats:
 358             streams = self._call_api(
 359                 'videos/%s/streams.json' % video_id, video_id,
 360                 'Downloading video streams JSON')
 361
 362             if 'external' in streams:
 363                 result.update({
 364                     '_type': 'url_transparent',
 365                     'url': streams['external']['url'],
 366                 })
 367                 return result
 368
 369             for format_id, stream_dict in streams.items():
 370                 for protocol, format_dict in stream_dict.items():
 371                     add_format(format_id, format_dict, protocol)
 372         self._sort_formats(formats)
 373
 374         result['formats'] = formats
 375         return result
 376
 377
 378 class VikiChannelIE(VikiBaseIE):
 379     IE_NAME = 'viki:channel'
 380     _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
 381     _TESTS = [{
 382         'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
 383         'info_dict': {
 384             'id': '50c',
 385             'title': 'Boys Over Flowers',
 386             'description': 'md5:ecd3cff47967fe193cff37c0bec52790',
 387         },
 388         'playlist_mincount': 71,
 389     }, {
 390         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
 391         'info_dict': {
 392             'id': '1354c',
 393             'title': 'Poor Nastya [COMPLETE]',
 394             'description': 'md5:05bf5471385aa8b21c18ad450e350525',
 395         },
 396         'playlist_count': 127,
 397     }, {
 398         'url': 'http://www.viki.com/news/24569c-showbiz-korea',
 399         'only_matching': True,
 400     }, {
 401         'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
 402         'only_matching': True,
 403     }, {
 404         'url': 'http://www.viki.com/artists/2141c-shinee',
 405         'only_matching': True,
 406     }]
 407
 408     _PER_PAGE = 25
 409
 410     def _real_extract(self, url):
 411         channel_id = self._match_id(url)
 412
 413         channel = self._call_api(
 414             'containers/%s.json' % channel_id, channel_id,
 415             'Downloading channel JSON')
 416
 417         self._check_errors(channel)
 418
 419         title = self.dict_selection(channel['titles'], 'en')
 420
 421         description = self.dict_selection(channel['descriptions'], 'en')
 422
 423         entries = []
 424         for video_type in ('episodes', 'clips', 'movies'):
 425             for page_num in itertools.count(1):
 426                 page = self._call_api(
 427                     'containers/%s/%s.json?per_page=%d&sort=number&direction=asc&with_paging=true&page=%d'
 428                     % (channel_id, video_type, self._PER_PAGE, page_num), channel_id,
 429                     'Downloading %s JSON page #%d' % (video_type, page_num))
 430                 for video in page['response']:
 431                     video_id = video['id']
 432                     entries.append(self.url_result(
 433                         'https://www.viki.com/videos/%s' % video_id, 'Viki'))
 434                 if not page['pagination']['next']:
 435                     break
 436
 437         return self.playlist_result(entries, channel_id, title, description)