Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/viki.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import time
   6 import hmac
   7 import hashlib
   8 import itertools
   9
  10 from .common import InfoExtractor
  11 from ..utils import (
  12     ExtractorError,
  13     int_or_none,
  14     parse_age_limit,
  15     parse_iso8601,
  16     sanitized_Request,
  17 )
  18
  19
  20 class VikiBaseIE(InfoExtractor):
  21     _VALID_URL_BASE = r'https?://(?:www\.)?viki\.(?:com|net|mx|jp|fr)/'
  22     _API_QUERY_TEMPLATE = '/v4/%sapp=%s&t=%s&site=www.viki.com'
  23     _API_URL_TEMPLATE = 'http://api.viki.io%s&sig=%s'
  24
  25     _APP = '65535a'
  26     _APP_VERSION = '2.2.5.1428709186'
  27     _APP_SECRET = '-$iJ}@p7!G@SyU/je1bEyWg}upLu-6V6-Lg9VD(]siH,r.,m-r|ulZ,U4LC/SeR)'
  28
  29     _NETRC_MACHINE = 'viki'
  30
  31     _token = None
  32
  33     _ERRORS = {
  34         'geo': 'Sorry, this content is not available in your region.',
  35         'upcoming': 'Sorry, this content is not yet available.',
  36         # 'paywall': 'paywall',
  37     }
  38
  39     def _prepare_call(self, path, timestamp=None, post_data=None):
  40         path += '?' if '?' not in path else '&'
  41         if not timestamp:
  42             timestamp = int(time.time())
  43         query = self._API_QUERY_TEMPLATE % (path, self._APP, timestamp)
  44         if self._token:
  45             query += '&token=%s' % self._token
  46         sig = hmac.new(
  47             self._APP_SECRET.encode('ascii'),
  48             query.encode('ascii'),
  49             hashlib.sha1
  50         ).hexdigest()
  51         url = self._API_URL_TEMPLATE % (query, sig)
  52         return sanitized_Request(
  53             url, json.dumps(post_data).encode('utf-8')) if post_data else url
  54
  55     def _call_api(self, path, video_id, note, timestamp=None, post_data=None):
  56         resp = self._download_json(
  57             self._prepare_call(path, timestamp, post_data), video_id, note)
  58
  59         error = resp.get('error')
  60         if error:
  61             if error == 'invalid timestamp':
  62                 resp = self._download_json(
  63                     self._prepare_call(path, int(resp['current_timestamp']), post_data),
  64                     video_id, '%s (retry)' % note)
  65                 error = resp.get('error')
  66             if error:
  67                 self._raise_error(resp['error'])
  68
  69         return resp
  70
  71     def _raise_error(self, error):
  72         raise ExtractorError(
  73             '%s returned error: %s' % (self.IE_NAME, error),
  74             expected=True)
  75
  76     def _check_errors(self, data):
  77         for reason, status in data.get('blocking', {}).items():
  78             if status and reason in self._ERRORS:
  79                 raise ExtractorError('%s said: %s' % (
  80                     self.IE_NAME, self._ERRORS[reason]), expected=True)
  81
  82     def _real_initialize(self):
  83         self._login()
  84
  85     def _login(self):
  86         (username, password) = self._get_login_info()
  87         if username is None:
  88             return
  89
  90         login_form = {
  91             'login_id': username,
  92             'password': password,
  93         }
  94
  95         login = self._call_api(
  96             'sessions.json', None,
  97             'Logging in as %s' % username, post_data=login_form)
  98
  99         self._token = login.get('token')
 100         if not self._token:
 101             self.report_warning('Unable to get session token, login has probably failed')
 102
 103     @staticmethod
 104     def dict_selection(dict_obj, preferred_key, allow_fallback=True):
 105         if preferred_key in dict_obj:
 106             return dict_obj.get(preferred_key)
 107
 108         if not allow_fallback:
 109             return
 110
 111         filtered_dict = list(filter(None, [dict_obj.get(k) for k in dict_obj.keys()]))
 112         return filtered_dict[0] if filtered_dict else None
 113
 114
 115 class VikiIE(VikiBaseIE):
 116     IE_NAME = 'viki'
 117     _VALID_URL = r'%s(?:videos|player)/(?P<id>[0-9]+v)' % VikiBaseIE._VALID_URL_BASE
 118     _TESTS = [{
 119         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
 120         'info_dict': {
 121             'id': '1023585v',
 122             'ext': 'mp4',
 123             'title': 'Heirs Episode 14',
 124             'uploader': 'SBS',
 125             'description': 'md5:c4b17b9626dd4b143dcc4d855ba3474e',
 126             'upload_date': '20131121',
 127             'age_limit': 13,
 128         },
 129         'skip': 'Blocked in the US',
 130     }, {
 131         # clip
 132         'url': 'http://www.viki.com/videos/1067139v-the-avengers-age-of-ultron-press-conference',
 133         'md5': '86c0b5dbd4d83a6611a79987cc7a1989',
 134         'info_dict': {
 135             'id': '1067139v',
 136             'ext': 'mp4',
 137             'title': "'The Avengers: Age of Ultron' Press Conference",
 138             'description': 'md5:d70b2f9428f5488321bfe1db10d612ea',
 139             'duration': 352,
 140             'timestamp': 1430380829,
 141             'upload_date': '20150430',
 142             'uploader': 'Arirang TV',
 143             'like_count': int,
 144             'age_limit': 0,
 145         }
 146     }, {
 147         'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
 148         'info_dict': {
 149             'id': '1048879v',
 150             'ext': 'mp4',
 151             'title': 'Ankhon Dekhi',
 152             'duration': 6512,
 153             'timestamp': 1408532356,
 154             'upload_date': '20140820',
 155             'uploader': 'Spuul',
 156             'like_count': int,
 157             'age_limit': 13,
 158         },
 159         'skip': 'Blocked in the US',
 160     }, {
 161         # episode
 162         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
 163         'md5': '5fa476a902e902783ac7a4d615cdbc7a',
 164         'info_dict': {
 165             'id': '44699v',
 166             'ext': 'mp4',
 167             'title': 'Boys Over Flowers - Episode 1',
 168             'description': 'md5:b89cf50038b480b88b5b3c93589a9076',
 169             'duration': 4204,
 170             'timestamp': 1270496524,
 171             'upload_date': '20100405',
 172             'uploader': 'group8',
 173             'like_count': int,
 174             'age_limit': 13,
 175         }
 176     }, {
 177         # youtube external
 178         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
 179         'md5': '63f8600c1da6f01b7640eee7eca4f1da',
 180         'info_dict': {
 181             'id': '50562v',
 182             'ext': 'webm',
 183             'title': 'Poor Nastya [COMPLETE] - Episode 1',
 184             'description': '',
 185             'duration': 606,
 186             'timestamp': 1274949505,
 187             'upload_date': '20101213',
 188             'uploader': 'ad14065n',
 189             'uploader_id': 'ad14065n',
 190             'like_count': int,
 191             'age_limit': 13,
 192         }
 193     }, {
 194         'url': 'http://www.viki.com/player/44699v',
 195         'only_matching': True,
 196     }, {
 197         # non-English description
 198         'url': 'http://www.viki.com/videos/158036v-love-in-magic',
 199         'md5': '1713ae35df5a521b31f6dc40730e7c9c',
 200         'info_dict': {
 201             'id': '158036v',
 202             'ext': 'mp4',
 203             'uploader': 'I Planet Entertainment',
 204             'upload_date': '20111122',
 205             'timestamp': 1321985454,
 206             'description': 'md5:44b1e46619df3a072294645c770cef36',
 207             'title': 'Love In Magic',
 208             'age_limit': 13,
 209         },
 210     }]
 211
 212     def _real_extract(self, url):
 213         video_id = self._match_id(url)
 214
 215         video = self._call_api(
 216             'videos/%s.json' % video_id, video_id, 'Downloading video JSON')
 217
 218         self._check_errors(video)
 219
 220         title = self.dict_selection(video.get('titles', {}), 'en', allow_fallback=False)
 221         if not title:
 222             title = 'Episode %d' % video.get('number') if video.get('type') == 'episode' else video.get('id') or video_id
 223             container_titles = video.get('container', {}).get('titles', {})
 224             container_title = self.dict_selection(container_titles, 'en')
 225             title = '%s - %s' % (container_title, title)
 226
 227         description = self.dict_selection(video.get('descriptions', {}), 'en')
 228
 229         duration = int_or_none(video.get('duration'))
 230         timestamp = parse_iso8601(video.get('created_at'))
 231         uploader = video.get('author')
 232         like_count = int_or_none(video.get('likes', {}).get('count'))
 233         age_limit = parse_age_limit(video.get('rating'))
 234
 235         thumbnails = []
 236         for thumbnail_id, thumbnail in video.get('images', {}).items():
 237             thumbnails.append({
 238                 'id': thumbnail_id,
 239                 'url': thumbnail.get('url'),
 240             })
 241
 242         subtitles = {}
 243         for subtitle_lang, _ in video.get('subtitle_completions', {}).items():
 244             subtitles[subtitle_lang] = [{
 245                 'ext': subtitles_format,
 246                 'url': self._prepare_call(
 247                     'videos/%s/subtitles/%s.%s' % (video_id, subtitle_lang, subtitles_format)),
 248             } for subtitles_format in ('srt', 'vtt')]
 249
 250         result = {
 251             'id': video_id,
 252             'title': title,
 253             'description': description,
 254             'duration': duration,
 255             'timestamp': timestamp,
 256             'uploader': uploader,
 257             'like_count': like_count,
 258             'age_limit': age_limit,
 259             'thumbnails': thumbnails,
 260             'subtitles': subtitles,
 261         }
 262
 263         streams = self._call_api(
 264             'videos/%s/streams.json' % video_id, video_id,
 265             'Downloading video streams JSON')
 266
 267         if 'external' in streams:
 268             result.update({
 269                 '_type': 'url_transparent',
 270                 'url': streams['external']['url'],
 271             })
 272             return result
 273
 274         formats = []
 275         for format_id, stream_dict in streams.items():
 276             height = int_or_none(self._search_regex(
 277                 r'^(\d+)[pP]$', format_id, 'height', default=None))
 278             for protocol, format_dict in stream_dict.items():
 279                 if format_id == 'm3u8':
 280                     m3u8_formats = self._extract_m3u8_formats(
 281                         format_dict['url'], video_id, 'mp4',
 282                         entry_protocol='m3u8_native', preference=-1,
 283                         m3u8_id='m3u8-%s' % protocol, fatal=False)
 284                     # Despite CODECS metadata in m3u8 all video-only formats
 285                     # are actually video+audio
 286                     for f in m3u8_formats:
 287                         if f.get('acodec') == 'none' and f.get('vcodec') != 'none':
 288                             f['acodec'] = None
 289                     formats.extend(m3u8_formats)
 290                 else:
 291                     formats.append({
 292                         'url': format_dict['url'],
 293                         'format_id': '%s-%s' % (format_id, protocol),
 294                         'height': height,
 295                     })
 296         self._sort_formats(formats)
 297
 298         result['formats'] = formats
 299         return result
 300
 301
 302 class VikiChannelIE(VikiBaseIE):
 303     IE_NAME = 'viki:channel'
 304     _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
 305     _TESTS = [{
 306         'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
 307         'info_dict': {
 308             'id': '50c',
 309             'title': 'Boys Over Flowers',
 310             'description': 'md5:ecd3cff47967fe193cff37c0bec52790',
 311         },
 312         'playlist_mincount': 71,
 313     }, {
 314         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
 315         'info_dict': {
 316             'id': '1354c',
 317             'title': 'Poor Nastya [COMPLETE]',
 318             'description': 'md5:05bf5471385aa8b21c18ad450e350525',
 319         },
 320         'playlist_count': 127,
 321     }, {
 322         'url': 'http://www.viki.com/news/24569c-showbiz-korea',
 323         'only_matching': True,
 324     }, {
 325         'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
 326         'only_matching': True,
 327     }, {
 328         'url': 'http://www.viki.com/artists/2141c-shinee',
 329         'only_matching': True,
 330     }]
 331
 332     _PER_PAGE = 25
 333
 334     def _real_extract(self, url):
 335         channel_id = self._match_id(url)
 336
 337         channel = self._call_api(
 338             'containers/%s.json' % channel_id, channel_id,
 339             'Downloading channel JSON')
 340
 341         self._check_errors(channel)
 342
 343         title = self.dict_selection(channel['titles'], 'en')
 344
 345         description = self.dict_selection(channel['descriptions'], 'en')
 346
 347         entries = []
 348         for video_type in ('episodes', 'clips', 'movies'):
 349             for page_num in itertools.count(1):
 350                 page = self._call_api(
 351                     'containers/%s/%s.json?per_page=%d&sort=number&direction=asc&with_paging=true&page=%d'
 352                     % (channel_id, video_type, self._PER_PAGE, page_num), channel_id,
 353                     'Downloading %s JSON page #%d' % (video_type, page_num))
 354                 for video in page['response']:
 355                     video_id = video['id']
 356                     entries.append(self.url_result(
 357                         'http://www.viki.com/videos/%s' % video_id, 'Viki'))
 358                 if not page['pagination']['next']:
 359                     break
 360
 361         return self.playlist_result(entries, channel_id, title, description)