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