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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import hashlib
   5 import hmac
   6 import itertools
   7 import json
   8 import re
   9 import time
  10
  11 from .common import InfoExtractor
  12 from ..utils import (
  13     ExtractorError,
  14     int_or_none,
  15     parse_age_limit,
  16     parse_iso8601,
  17     sanitized_Request,
  18 )
  19
  20
  21 class VikiBaseIE(InfoExtractor):
  22     _VALID_URL_BASE = r'https?://(?:www\.)?viki\.(?:com|net|mx|jp|fr)/'
  23     _API_QUERY_TEMPLATE = '/v4/%sapp=%s&t=%s&site=www.viki.com'
  24     _API_URL_TEMPLATE = 'http://api.viki.io%s&sig=%s'
  25
  26     _APP = '65535a'
  27     _APP_VERSION = '2.2.5.1428709186'
  28     _APP_SECRET = '-$iJ}@p7!G@SyU/je1bEyWg}upLu-6V6-Lg9VD(]siH,r.,m-r|ulZ,U4LC/SeR)'
  29
  30     _NETRC_MACHINE = 'viki'
  31
  32     _token = None
  33
  34     _ERRORS = {
  35         'geo': 'Sorry, this content is not available in your region.',
  36         'upcoming': 'Sorry, this content is not yet available.',
  37         # 'paywall': 'paywall',
  38     }
  39
  40     def _prepare_call(self, path, timestamp=None, post_data=None):
  41         path += '?' if '?' not in path else '&'
  42         if not timestamp:
  43             timestamp = int(time.time())
  44         query = self._API_QUERY_TEMPLATE % (path, self._APP, timestamp)
  45         if self._token:
  46             query += '&token=%s' % self._token
  47         sig = hmac.new(
  48             self._APP_SECRET.encode('ascii'),
  49             query.encode('ascii'),
  50             hashlib.sha1
  51         ).hexdigest()
  52         url = self._API_URL_TEMPLATE % (query, sig)
  53         return sanitized_Request(
  54             url, json.dumps(post_data).encode('utf-8')) if post_data else url
  55
  56     def _call_api(self, path, video_id, note, timestamp=None, post_data=None):
  57         resp = self._download_json(
  58             self._prepare_call(path, timestamp, post_data), video_id, note)
  59
  60         error = resp.get('error')
  61         if error:
  62             if error == 'invalid timestamp':
  63                 resp = self._download_json(
  64                     self._prepare_call(path, int(resp['current_timestamp']), post_data),
  65                     video_id, '%s (retry)' % note)
  66                 error = resp.get('error')
  67             if error:
  68                 self._raise_error(resp['error'])
  69
  70         return resp
  71
  72     def _raise_error(self, error):
  73         raise ExtractorError(
  74             '%s returned error: %s' % (self.IE_NAME, error),
  75             expected=True)
  76
  77     def _check_errors(self, data):
  78         for reason, status in data.get('blocking', {}).items():
  79             if status and reason in self._ERRORS:
  80                 raise ExtractorError('%s said: %s' % (
  81                     self.IE_NAME, self._ERRORS[reason]), expected=True)
  82
  83     def _real_initialize(self):
  84         self._login()
  85
  86     def _login(self):
  87         (username, password) = self._get_login_info()
  88         if username is None:
  89             return
  90
  91         login_form = {
  92             'login_id': username,
  93             'password': password,
  94         }
  95
  96         login = self._call_api(
  97             'sessions.json', None,
  98             'Logging in as %s' % username, post_data=login_form)
  99
 100         self._token = login.get('token')
 101         if not self._token:
 102             self.report_warning('Unable to get session token, login has probably failed')
 103
 104     @staticmethod
 105     def dict_selection(dict_obj, preferred_key, allow_fallback=True):
 106         if preferred_key in dict_obj:
 107             return dict_obj.get(preferred_key)
 108
 109         if not allow_fallback:
 110             return
 111
 112         filtered_dict = list(filter(None, [dict_obj.get(k) for k in dict_obj.keys()]))
 113         return filtered_dict[0] if filtered_dict else None
 114
 115
 116 class VikiIE(VikiBaseIE):
 117     IE_NAME = 'viki'
 118     _VALID_URL = r'%s(?:videos|player)/(?P<id>[0-9]+v)' % VikiBaseIE._VALID_URL_BASE
 119     _TESTS = [{
 120         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
 121         'info_dict': {
 122             'id': '1023585v',
 123             'ext': 'mp4',
 124             'title': 'Heirs Episode 14',
 125             'uploader': 'SBS',
 126             'description': 'md5:c4b17b9626dd4b143dcc4d855ba3474e',
 127             'upload_date': '20131121',
 128             'age_limit': 13,
 129         },
 130         'skip': 'Blocked in the US',
 131     }, {
 132         # clip
 133         'url': 'http://www.viki.com/videos/1067139v-the-avengers-age-of-ultron-press-conference',
 134         'md5': '86c0b5dbd4d83a6611a79987cc7a1989',
 135         'info_dict': {
 136             'id': '1067139v',
 137             'ext': 'mp4',
 138             'title': "'The Avengers: Age of Ultron' Press Conference",
 139             'description': 'md5:d70b2f9428f5488321bfe1db10d612ea',
 140             'duration': 352,
 141             'timestamp': 1430380829,
 142             'upload_date': '20150430',
 143             'uploader': 'Arirang TV',
 144             'like_count': int,
 145             'age_limit': 0,
 146         }
 147     }, {
 148         'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
 149         'info_dict': {
 150             'id': '1048879v',
 151             'ext': 'mp4',
 152             'title': 'Ankhon Dekhi',
 153             'duration': 6512,
 154             'timestamp': 1408532356,
 155             'upload_date': '20140820',
 156             'uploader': 'Spuul',
 157             'like_count': int,
 158             'age_limit': 13,
 159         },
 160         'skip': 'Blocked in the US',
 161     }, {
 162         # episode
 163         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
 164         'md5': '5fa476a902e902783ac7a4d615cdbc7a',
 165         'info_dict': {
 166             'id': '44699v',
 167             'ext': 'mp4',
 168             'title': 'Boys Over Flowers - Episode 1',
 169             'description': 'md5:b89cf50038b480b88b5b3c93589a9076',
 170             'duration': 4204,
 171             'timestamp': 1270496524,
 172             'upload_date': '20100405',
 173             'uploader': 'group8',
 174             'like_count': int,
 175             'age_limit': 13,
 176         }
 177     }, {
 178         # youtube external
 179         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
 180         'md5': '63f8600c1da6f01b7640eee7eca4f1da',
 181         'info_dict': {
 182             'id': '50562v',
 183             'ext': 'webm',
 184             'title': 'Poor Nastya [COMPLETE] - Episode 1',
 185             'description': '',
 186             'duration': 606,
 187             'timestamp': 1274949505,
 188             'upload_date': '20101213',
 189             'uploader': 'ad14065n',
 190             'uploader_id': 'ad14065n',
 191             'like_count': int,
 192             'age_limit': 13,
 193         }
 194     }, {
 195         'url': 'http://www.viki.com/player/44699v',
 196         'only_matching': True,
 197     }, {
 198         # non-English description
 199         'url': 'http://www.viki.com/videos/158036v-love-in-magic',
 200         'md5': '1713ae35df5a521b31f6dc40730e7c9c',
 201         'info_dict': {
 202             'id': '158036v',
 203             'ext': 'mp4',
 204             'uploader': 'I Planet Entertainment',
 205             'upload_date': '20111122',
 206             'timestamp': 1321985454,
 207             'description': 'md5:44b1e46619df3a072294645c770cef36',
 208             'title': 'Love In Magic',
 209             'age_limit': 13,
 210         },
 211     }]
 212
 213     def _real_extract(self, url):
 214         video_id = self._match_id(url)
 215
 216         video = self._call_api(
 217             'videos/%s.json' % video_id, video_id, 'Downloading video JSON')
 218
 219         self._check_errors(video)
 220
 221         title = self.dict_selection(video.get('titles', {}), 'en', allow_fallback=False)
 222         if not title:
 223             title = 'Episode %d' % video.get('number') if video.get('type') == 'episode' else video.get('id') or video_id
 224             container_titles = video.get('container', {}).get('titles', {})
 225             container_title = self.dict_selection(container_titles, 'en')
 226             title = '%s - %s' % (container_title, title)
 227
 228         description = self.dict_selection(video.get('descriptions', {}), 'en')
 229
 230         duration = int_or_none(video.get('duration'))
 231         timestamp = parse_iso8601(video.get('created_at'))
 232         uploader = video.get('author')
 233         like_count = int_or_none(video.get('likes', {}).get('count'))
 234         age_limit = parse_age_limit(video.get('rating'))
 235
 236         thumbnails = []
 237         for thumbnail_id, thumbnail in video.get('images', {}).items():
 238             thumbnails.append({
 239                 'id': thumbnail_id,
 240                 'url': thumbnail.get('url'),
 241             })
 242
 243         subtitles = {}
 244         for subtitle_lang, _ in video.get('subtitle_completions', {}).items():
 245             subtitles[subtitle_lang] = [{
 246                 'ext': subtitles_format,
 247                 'url': self._prepare_call(
 248                     'videos/%s/subtitles/%s.%s' % (video_id, subtitle_lang, subtitles_format)),
 249             } for subtitles_format in ('srt', 'vtt')]
 250
 251         result = {
 252             'id': video_id,
 253             'title': title,
 254             'description': description,
 255             'duration': duration,
 256             'timestamp': timestamp,
 257             'uploader': uploader,
 258             'like_count': like_count,
 259             'age_limit': age_limit,
 260             'thumbnails': thumbnails,
 261             'subtitles': subtitles,
 262         }
 263
 264         streams = self._call_api(
 265             'videos/%s/streams.json' % video_id, video_id,
 266             'Downloading video streams JSON')
 267
 268         if 'external' in streams:
 269             result.update({
 270                 '_type': 'url_transparent',
 271                 'url': streams['external']['url'],
 272             })
 273             return result
 274
 275         formats = []
 276         for format_id, stream_dict in streams.items():
 277             height = int_or_none(self._search_regex(
 278                 r'^(\d+)[pP]$', format_id, 'height', default=None))
 279             for protocol, format_dict in stream_dict.items():
 280                 # rtmps URLs does not seem to work
 281                 if protocol == 'rtmps':
 282                     continue
 283                 format_url = format_dict['url']
 284                 if format_id == 'm3u8':
 285                     m3u8_formats = self._extract_m3u8_formats(
 286                         format_url, video_id, 'mp4',
 287                         entry_protocol='m3u8_native',
 288                         m3u8_id='m3u8-%s' % protocol, fatal=False)
 289                     # Despite CODECS metadata in m3u8 all video-only formats
 290                     # are actually video+audio
 291                     for f in m3u8_formats:
 292                         if f.get('acodec') == 'none' and f.get('vcodec') != 'none':
 293                             f['acodec'] = None
 294                     formats.extend(m3u8_formats)
 295                 elif format_url.startswith('rtmp'):
 296                     mobj = re.search(
 297                         r'^(?P<url>rtmp://[^/]+/(?P<app>.+?))/(?P<playpath>mp4:.+)$',
 298                         format_url)
 299                     if not mobj:
 300                         continue
 301                     formats.append({
 302                         'format_id': 'rtmp-%s' % format_id,
 303                         'ext': 'flv',
 304                         'url': mobj.group('url'),
 305                         'play_path': mobj.group('playpath'),
 306                         'app': mobj.group('app'),
 307                         'page_url': url,
 308                     })
 309                 else:
 310                     formats.append({
 311                         'url': format_url,
 312                         'format_id': '%s-%s' % (format_id, protocol),
 313                         'height': height,
 314                     })
 315         self._sort_formats(formats)
 316
 317         result['formats'] = formats
 318         return result
 319
 320
 321 class VikiChannelIE(VikiBaseIE):
 322     IE_NAME = 'viki:channel'
 323     _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
 324     _TESTS = [{
 325         'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
 326         'info_dict': {
 327             'id': '50c',
 328             'title': 'Boys Over Flowers',
 329             'description': 'md5:ecd3cff47967fe193cff37c0bec52790',
 330         },
 331         'playlist_mincount': 71,
 332     }, {
 333         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
 334         'info_dict': {
 335             'id': '1354c',
 336             'title': 'Poor Nastya [COMPLETE]',
 337             'description': 'md5:05bf5471385aa8b21c18ad450e350525',
 338         },
 339         'playlist_count': 127,
 340     }, {
 341         'url': 'http://www.viki.com/news/24569c-showbiz-korea',
 342         'only_matching': True,
 343     }, {
 344         'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
 345         'only_matching': True,
 346     }, {
 347         'url': 'http://www.viki.com/artists/2141c-shinee',
 348         'only_matching': True,
 349     }]
 350
 351     _PER_PAGE = 25
 352
 353     def _real_extract(self, url):
 354         channel_id = self._match_id(url)
 355
 356         channel = self._call_api(
 357             'containers/%s.json' % channel_id, channel_id,
 358             'Downloading channel JSON')
 359
 360         self._check_errors(channel)
 361
 362         title = self.dict_selection(channel['titles'], 'en')
 363
 364         description = self.dict_selection(channel['descriptions'], 'en')
 365
 366         entries = []
 367         for video_type in ('episodes', 'clips', 'movies'):
 368             for page_num in itertools.count(1):
 369                 page = self._call_api(
 370                     'containers/%s/%s.json?per_page=%d&sort=number&direction=asc&with_paging=true&page=%d'
 371                     % (channel_id, video_type, self._PER_PAGE, page_num), channel_id,
 372                     'Downloading %s JSON page #%d' % (video_type, page_num))
 373                 for video in page['response']:
 374                     video_id = video['id']
 375                     entries.append(self.url_result(
 376                         'http://www.viki.com/videos/%s' % video_id, 'Viki'))
 377                 if not page['pagination']['next']:
 378                     break
 379
 380         return self.playlist_result(entries, channel_id, title, description)