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     _GEO_BYPASS = False
  31     _NETRC_MACHINE = 'viki'
  32
  33     _token = None
  34
  35     _ERRORS = {
  36         'geo': 'Sorry, this content is not available in your region.',
  37         'upcoming': 'Sorry, this content is not yet available.',
  38         # 'paywall': 'paywall',
  39     }
  40
  41     def _prepare_call(self, path, timestamp=None, post_data=None):
  42         path += '?' if '?' not in path else '&'
  43         if not timestamp:
  44             timestamp = int(time.time())
  45         query = self._API_QUERY_TEMPLATE % (path, self._APP, timestamp)
  46         if self._token:
  47             query += '&token=%s' % self._token
  48         sig = hmac.new(
  49             self._APP_SECRET.encode('ascii'),
  50             query.encode('ascii'),
  51             hashlib.sha1
  52         ).hexdigest()
  53         url = self._API_URL_TEMPLATE % (query, sig)
  54         return sanitized_Request(
  55             url, json.dumps(post_data).encode('utf-8')) if post_data else url
  56
  57     def _call_api(self, path, video_id, note, timestamp=None, post_data=None):
  58         resp = self._download_json(
  59             self._prepare_call(path, timestamp, post_data), video_id, note)
  60
  61         error = resp.get('error')
  62         if error:
  63             if error == 'invalid timestamp':
  64                 resp = self._download_json(
  65                     self._prepare_call(path, int(resp['current_timestamp']), post_data),
  66                     video_id, '%s (retry)' % note)
  67                 error = resp.get('error')
  68             if error:
  69                 self._raise_error(resp['error'])
  70
  71         return resp
  72
  73     def _raise_error(self, error):
  74         raise ExtractorError(
  75             '%s returned error: %s' % (self.IE_NAME, error),
  76             expected=True)
  77
  78     def _check_errors(self, data):
  79         for reason, status in data.get('blocking', {}).items():
  80             if status and reason in self._ERRORS:
  81                 message = self._ERRORS[reason]
  82                 if reason == 'geo':
  83                     self.raise_geo_restricted(msg=message)
  84                 raise ExtractorError('%s said: %s' % (
  85                     self.IE_NAME, message), expected=True)
  86
  87     def _real_initialize(self):
  88         self._login()
  89
  90     def _login(self):
  91         (username, password) = self._get_login_info()
  92         if username is None:
  93             return
  94
  95         login_form = {
  96             'login_id': username,
  97             'password': password,
  98         }
  99
 100         login = self._call_api(
 101             'sessions.json', None,
 102             'Logging in as %s' % username, post_data=login_form)
 103
 104         self._token = login.get('token')
 105         if not self._token:
 106             self.report_warning('Unable to get session token, login has probably failed')
 107
 108     @staticmethod
 109     def dict_selection(dict_obj, preferred_key, allow_fallback=True):
 110         if preferred_key in dict_obj:
 111             return dict_obj.get(preferred_key)
 112
 113         if not allow_fallback:
 114             return
 115
 116         filtered_dict = list(filter(None, [dict_obj.get(k) for k in dict_obj.keys()]))
 117         return filtered_dict[0] if filtered_dict else None
 118
 119
 120 class VikiIE(VikiBaseIE):
 121     IE_NAME = 'viki'
 122     _VALID_URL = r'%s(?:videos|player)/(?P<id>[0-9]+v)' % VikiBaseIE._VALID_URL_BASE
 123     _TESTS = [{
 124         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
 125         'info_dict': {
 126             'id': '1023585v',
 127             'ext': 'mp4',
 128             'title': 'Heirs Episode 14',
 129             'uploader': 'SBS',
 130             'description': 'md5:c4b17b9626dd4b143dcc4d855ba3474e',
 131             'upload_date': '20131121',
 132             'age_limit': 13,
 133         },
 134         'skip': 'Blocked in the US',
 135     }, {
 136         # clip
 137         'url': 'http://www.viki.com/videos/1067139v-the-avengers-age-of-ultron-press-conference',
 138         'md5': '86c0b5dbd4d83a6611a79987cc7a1989',
 139         'info_dict': {
 140             'id': '1067139v',
 141             'ext': 'mp4',
 142             'title': "'The Avengers: Age of Ultron' Press Conference",
 143             'description': 'md5:d70b2f9428f5488321bfe1db10d612ea',
 144             'duration': 352,
 145             'timestamp': 1430380829,
 146             'upload_date': '20150430',
 147             'uploader': 'Arirang TV',
 148             'like_count': int,
 149             'age_limit': 0,
 150         }
 151     }, {
 152         'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
 153         'info_dict': {
 154             'id': '1048879v',
 155             'ext': 'mp4',
 156             'title': 'Ankhon Dekhi',
 157             'duration': 6512,
 158             'timestamp': 1408532356,
 159             'upload_date': '20140820',
 160             'uploader': 'Spuul',
 161             'like_count': int,
 162             'age_limit': 13,
 163         },
 164         'skip': 'Blocked in the US',
 165     }, {
 166         # episode
 167         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
 168         'md5': '5fa476a902e902783ac7a4d615cdbc7a',
 169         'info_dict': {
 170             'id': '44699v',
 171             'ext': 'mp4',
 172             'title': 'Boys Over Flowers - Episode 1',
 173             'description': 'md5:b89cf50038b480b88b5b3c93589a9076',
 174             'duration': 4204,
 175             'timestamp': 1270496524,
 176             'upload_date': '20100405',
 177             'uploader': 'group8',
 178             'like_count': int,
 179             'age_limit': 13,
 180         }
 181     }, {
 182         # youtube external
 183         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
 184         'md5': '63f8600c1da6f01b7640eee7eca4f1da',
 185         'info_dict': {
 186             'id': '50562v',
 187             'ext': 'webm',
 188             'title': 'Poor Nastya [COMPLETE] - Episode 1',
 189             'description': '',
 190             'duration': 606,
 191             'timestamp': 1274949505,
 192             'upload_date': '20101213',
 193             'uploader': 'ad14065n',
 194             'uploader_id': 'ad14065n',
 195             'like_count': int,
 196             'age_limit': 13,
 197         }
 198     }, {
 199         'url': 'http://www.viki.com/player/44699v',
 200         'only_matching': True,
 201     }, {
 202         # non-English description
 203         'url': 'http://www.viki.com/videos/158036v-love-in-magic',
 204         'md5': '1713ae35df5a521b31f6dc40730e7c9c',
 205         'info_dict': {
 206             'id': '158036v',
 207             'ext': 'mp4',
 208             'uploader': 'I Planet Entertainment',
 209             'upload_date': '20111122',
 210             'timestamp': 1321985454,
 211             'description': 'md5:44b1e46619df3a072294645c770cef36',
 212             'title': 'Love In Magic',
 213             'age_limit': 13,
 214         },
 215     }]
 216
 217     def _real_extract(self, url):
 218         video_id = self._match_id(url)
 219
 220         video = self._call_api(
 221             'videos/%s.json' % video_id, video_id, 'Downloading video JSON')
 222
 223         self._check_errors(video)
 224
 225         title = self.dict_selection(video.get('titles', {}), 'en', allow_fallback=False)
 226         if not title:
 227             title = 'Episode %d' % video.get('number') if video.get('type') == 'episode' else video.get('id') or video_id
 228             container_titles = video.get('container', {}).get('titles', {})
 229             container_title = self.dict_selection(container_titles, 'en')
 230             title = '%s - %s' % (container_title, title)
 231
 232         description = self.dict_selection(video.get('descriptions', {}), 'en')
 233
 234         duration = int_or_none(video.get('duration'))
 235         timestamp = parse_iso8601(video.get('created_at'))
 236         uploader = video.get('author')
 237         like_count = int_or_none(video.get('likes', {}).get('count'))
 238         age_limit = parse_age_limit(video.get('rating'))
 239
 240         thumbnails = []
 241         for thumbnail_id, thumbnail in video.get('images', {}).items():
 242             thumbnails.append({
 243                 'id': thumbnail_id,
 244                 'url': thumbnail.get('url'),
 245             })
 246
 247         subtitles = {}
 248         for subtitle_lang, _ in video.get('subtitle_completions', {}).items():
 249             subtitles[subtitle_lang] = [{
 250                 'ext': subtitles_format,
 251                 'url': self._prepare_call(
 252                     'videos/%s/subtitles/%s.%s' % (video_id, subtitle_lang, subtitles_format)),
 253             } for subtitles_format in ('srt', 'vtt')]
 254
 255         result = {
 256             'id': video_id,
 257             'title': title,
 258             'description': description,
 259             'duration': duration,
 260             'timestamp': timestamp,
 261             'uploader': uploader,
 262             'like_count': like_count,
 263             'age_limit': age_limit,
 264             'thumbnails': thumbnails,
 265             'subtitles': subtitles,
 266         }
 267
 268         streams = self._call_api(
 269             'videos/%s/streams.json' % video_id, video_id,
 270             'Downloading video streams JSON')
 271
 272         if 'external' in streams:
 273             result.update({
 274                 '_type': 'url_transparent',
 275                 'url': streams['external']['url'],
 276             })
 277             return result
 278
 279         formats = []
 280         for format_id, stream_dict in streams.items():
 281             height = int_or_none(self._search_regex(
 282                 r'^(\d+)[pP]$', format_id, 'height', default=None))
 283             for protocol, format_dict in stream_dict.items():
 284                 # rtmps URLs does not seem to work
 285                 if protocol == 'rtmps':
 286                     continue
 287                 format_url = format_dict['url']
 288                 if format_id == 'm3u8':
 289                     m3u8_formats = self._extract_m3u8_formats(
 290                         format_url, video_id, 'mp4',
 291                         entry_protocol='m3u8_native',
 292                         m3u8_id='m3u8-%s' % protocol, fatal=False)
 293                     # Despite CODECS metadata in m3u8 all video-only formats
 294                     # are actually video+audio
 295                     for f in m3u8_formats:
 296                         if f.get('acodec') == 'none' and f.get('vcodec') != 'none':
 297                             f['acodec'] = None
 298                     formats.extend(m3u8_formats)
 299                 elif format_url.startswith('rtmp'):
 300                     mobj = re.search(
 301                         r'^(?P<url>rtmp://[^/]+/(?P<app>.+?))/(?P<playpath>mp4:.+)$',
 302                         format_url)
 303                     if not mobj:
 304                         continue
 305                     formats.append({
 306                         'format_id': 'rtmp-%s' % format_id,
 307                         'ext': 'flv',
 308                         'url': mobj.group('url'),
 309                         'play_path': mobj.group('playpath'),
 310                         'app': mobj.group('app'),
 311                         'page_url': url,
 312                     })
 313                 else:
 314                     formats.append({
 315                         'url': format_url,
 316                         'format_id': '%s-%s' % (format_id, protocol),
 317                         'height': height,
 318                     })
 319         self._sort_formats(formats)
 320
 321         result['formats'] = formats
 322         return result
 323
 324
 325 class VikiChannelIE(VikiBaseIE):
 326     IE_NAME = 'viki:channel'
 327     _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
 328     _TESTS = [{
 329         'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
 330         'info_dict': {
 331             'id': '50c',
 332             'title': 'Boys Over Flowers',
 333             'description': 'md5:ecd3cff47967fe193cff37c0bec52790',
 334         },
 335         'playlist_mincount': 71,
 336     }, {
 337         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
 338         'info_dict': {
 339             'id': '1354c',
 340             'title': 'Poor Nastya [COMPLETE]',
 341             'description': 'md5:05bf5471385aa8b21c18ad450e350525',
 342         },
 343         'playlist_count': 127,
 344     }, {
 345         'url': 'http://www.viki.com/news/24569c-showbiz-korea',
 346         'only_matching': True,
 347     }, {
 348         'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
 349         'only_matching': True,
 350     }, {
 351         'url': 'http://www.viki.com/artists/2141c-shinee',
 352         'only_matching': True,
 353     }]
 354
 355     _PER_PAGE = 25
 356
 357     def _real_extract(self, url):
 358         channel_id = self._match_id(url)
 359
 360         channel = self._call_api(
 361             'containers/%s.json' % channel_id, channel_id,
 362             'Downloading channel JSON')
 363
 364         self._check_errors(channel)
 365
 366         title = self.dict_selection(channel['titles'], 'en')
 367
 368         description = self.dict_selection(channel['descriptions'], 'en')
 369
 370         entries = []
 371         for video_type in ('episodes', 'clips', 'movies'):
 372             for page_num in itertools.count(1):
 373                 page = self._call_api(
 374                     'containers/%s/%s.json?per_page=%d&sort=number&direction=asc&with_paging=true&page=%d'
 375                     % (channel_id, video_type, self._PER_PAGE, page_num), channel_id,
 376                     'Downloading %s JSON page #%d' % (video_type, page_num))
 377                 for video in page['response']:
 378                     video_id = video['id']
 379                     entries.append(self.url_result(
 380                         'http://www.viki.com/videos/%s' % video_id, 'Viki'))
 381                 if not page['pagination']['next']:
 382                     break
 383
 384         return self.playlist_result(entries, channel_id, title, description)