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 ..utils import (
  11     ExtractorError,
  12     int_or_none,
  13     parse_age_limit,
  14     parse_iso8601,
  15 )
  16 from ..compat import compat_urllib_request
  17 from .common import InfoExtractor
  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     def _prepare_call(self, path, timestamp=None, post_data=None):
  32         path += '?' if '?' not in path else '&'
  33         if not timestamp:
  34             timestamp = int(time.time())
  35         query = self._API_QUERY_TEMPLATE % (path, self._APP, timestamp)
  36         sig = hmac.new(
  37             self._APP_SECRET.encode('ascii'),
  38             query.encode('ascii'),
  39             hashlib.sha1
  40         ).hexdigest()
  41         url = self._API_URL_TEMPLATE % (query, sig)
  42         return compat_urllib_request.Request(
  43             url, json.dumps(post_data).encode('utf-8')) if post_data else url
  44
  45     def _call_api(self, path, video_id, note, timestamp=None, post_data=None):
  46         resp = self._download_json(
  47             self._prepare_call(path, timestamp, post_data), video_id, note)
  48
  49         error = resp.get('error')
  50         if error:
  51             if error == 'invalid timestamp':
  52                 resp = self._download_json(
  53                     self._prepare_call(path, int(resp['current_timestamp']), post_data),
  54                     video_id, '%s (retry)' % note)
  55                 error = resp.get('error')
  56             if error:
  57                 self._raise_error(resp['error'])
  58
  59         return resp
  60
  61     def _raise_error(self, error):
  62         raise ExtractorError(
  63             '%s returned error: %s' % (self.IE_NAME, error),
  64             expected=True)
  65
  66     def _real_initialize(self):
  67         self._login()
  68
  69     def _login(self):
  70         (username, password) = self._get_login_info()
  71         if username is None:
  72             return
  73
  74         login_form = {
  75             'login_id': username,
  76             'password': password,
  77         }
  78
  79         self._call_api(
  80             'sessions.json', None,
  81             'Logging in as %s' % username, post_data=login_form)
  82
  83
  84 class VikiIE(VikiBaseIE):
  85     IE_NAME = 'viki'
  86     _VALID_URL = r'%s(?:videos|player)/(?P<id>[0-9]+v)' % VikiBaseIE._VALID_URL_BASE
  87     _TESTS = [{
  88         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
  89         'info_dict': {
  90             'id': '1023585v',
  91             'ext': 'mp4',
  92             'title': 'Heirs Episode 14',
  93             'uploader': 'SBS',
  94             'description': 'md5:c4b17b9626dd4b143dcc4d855ba3474e',
  95             'upload_date': '20131121',
  96             'age_limit': 13,
  97         },
  98         'skip': 'Blocked in the US',
  99     }, {
 100         # clip
 101         'url': 'http://www.viki.com/videos/1067139v-the-avengers-age-of-ultron-press-conference',
 102         'md5': '86c0b5dbd4d83a6611a79987cc7a1989',
 103         'info_dict': {
 104             'id': '1067139v',
 105             'ext': 'mp4',
 106             'title': "'The Avengers: Age of Ultron' Press Conference",
 107             'description': 'md5:d70b2f9428f5488321bfe1db10d612ea',
 108             'duration': 352,
 109             'timestamp': 1430380829,
 110             'upload_date': '20150430',
 111             'uploader': 'Arirang TV',
 112             'like_count': int,
 113             'age_limit': 0,
 114         }
 115     }, {
 116         'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
 117         'info_dict': {
 118             'id': '1048879v',
 119             'ext': 'mp4',
 120             'title': 'Ankhon Dekhi',
 121             'duration': 6512,
 122             'timestamp': 1408532356,
 123             'upload_date': '20140820',
 124             'uploader': 'Spuul',
 125             'like_count': int,
 126             'age_limit': 13,
 127         },
 128         'params': {
 129             # m3u8 download
 130             'skip_download': True,
 131         }
 132     }, {
 133         # episode
 134         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
 135         'md5': '190f3ef426005ba3a080a63325955bc3',
 136         'info_dict': {
 137             'id': '44699v',
 138             'ext': 'mp4',
 139             'title': 'Boys Over Flowers - Episode 1',
 140             'description': 'md5:52617e4f729c7d03bfd4bcbbb6e946f2',
 141             'duration': 4155,
 142             'timestamp': 1270496524,
 143             'upload_date': '20100405',
 144             'uploader': 'group8',
 145             'like_count': int,
 146             'age_limit': 13,
 147         }
 148     }, {
 149         # youtube external
 150         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
 151         'md5': '216d1afdc0c64d1febc1e9f2bd4b864b',
 152         'info_dict': {
 153             'id': '50562v',
 154             'ext': 'mp4',
 155             'title': 'Poor Nastya [COMPLETE] - Episode 1',
 156             'description': '',
 157             'duration': 607,
 158             'timestamp': 1274949505,
 159             'upload_date': '20101213',
 160             'uploader': 'ad14065n',
 161             'uploader_id': 'ad14065n',
 162             'like_count': int,
 163             'age_limit': 13,
 164         }
 165     }, {
 166         'url': 'http://www.viki.com/player/44699v',
 167         'only_matching': True,
 168     }]
 169
 170     def _real_extract(self, url):
 171         video_id = self._match_id(url)
 172
 173         video = self._call_api(
 174             'videos/%s.json' % video_id, video_id, 'Downloading video JSON')
 175
 176         title = None
 177         titles = video.get('titles')
 178         if titles:
 179             title = titles.get('en') or titles[titles.keys()[0]]
 180         if not title:
 181             title = 'Episode %d' % video.get('number') if video.get('type') == 'episode' else video.get('id') or video_id
 182             container_titles = video.get('container', {}).get('titles')
 183             if container_titles:
 184                 container_title = container_titles.get('en') or container_titles[container_titles.keys()[0]]
 185                 title = '%s - %s' % (container_title, title)
 186
 187         descriptions = video.get('descriptions')
 188         description = descriptions.get('en') or descriptions[titles.keys()[0]] if descriptions else None
 189
 190         duration = int_or_none(video.get('duration'))
 191         timestamp = parse_iso8601(video.get('created_at'))
 192         uploader = video.get('author')
 193         like_count = int_or_none(video.get('likes', {}).get('count'))
 194         age_limit = parse_age_limit(video.get('rating'))
 195
 196         thumbnails = []
 197         for thumbnail_id, thumbnail in video.get('images', {}).items():
 198             thumbnails.append({
 199                 'id': thumbnail_id,
 200                 'url': thumbnail.get('url'),
 201             })
 202
 203         subtitles = {}
 204         for subtitle_lang, _ in video.get('subtitle_completions', {}).items():
 205             subtitles[subtitle_lang] = [{
 206                 'ext': subtitles_format,
 207                 'url': self._prepare_call(
 208                     'videos/%s/subtitles/%s.%s' % (video_id, subtitle_lang, subtitles_format)),
 209             } for subtitles_format in ('srt', 'vtt')]
 210
 211         result = {
 212             'id': video_id,
 213             'title': title,
 214             'description': description,
 215             'duration': duration,
 216             'timestamp': timestamp,
 217             'uploader': uploader,
 218             'like_count': like_count,
 219             'age_limit': age_limit,
 220             'thumbnails': thumbnails,
 221             'subtitles': subtitles,
 222         }
 223
 224         streams = self._call_api(
 225             'videos/%s/streams.json' % video_id, video_id,
 226             'Downloading video streams JSON')
 227
 228         if 'external' in streams:
 229             result.update({
 230                 '_type': 'url_transparent',
 231                 'url': streams['external']['url'],
 232             })
 233             return result
 234
 235         formats = []
 236         for format_id, stream_dict in streams.items():
 237             height = self._search_regex(
 238                 r'^(\d+)[pP]$', format_id, 'height', default=None)
 239             for protocol, format_dict in stream_dict.items():
 240                 if format_id == 'm3u8':
 241                     formats = self._extract_m3u8_formats(
 242                         format_dict['url'], video_id, 'mp4', m3u8_id='m3u8-%s' % protocol)
 243                 else:
 244                     formats.append({
 245                         'url': format_dict['url'],
 246                         'format_id': '%s-%s' % (format_id, protocol),
 247                         'height': height,
 248                     })
 249         self._sort_formats(formats)
 250
 251         result['formats'] = formats
 252         return result
 253
 254
 255 class VikiChannelIE(VikiBaseIE):
 256     IE_NAME = 'viki:channel'
 257     _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
 258     _TESTS = [{
 259         'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
 260         'info_dict': {
 261             'id': '50c',
 262             'title': 'Boys Over Flowers',
 263             'description': 'md5:ecd3cff47967fe193cff37c0bec52790',
 264         },
 265         'playlist_count': 70,
 266     }, {
 267         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
 268         'info_dict': {
 269             'id': '1354c',
 270             'title': 'Poor Nastya [COMPLETE]',
 271             'description': 'md5:05bf5471385aa8b21c18ad450e350525',
 272         },
 273         'playlist_count': 127,
 274     }, {
 275         'url': 'http://www.viki.com/news/24569c-showbiz-korea',
 276         'only_matching': True,
 277     }, {
 278         'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
 279         'only_matching': True,
 280     }, {
 281         'url': 'http://www.viki.com/artists/2141c-shinee',
 282         'only_matching': True,
 283     }]
 284
 285     _PER_PAGE = 25
 286
 287     def _real_extract(self, url):
 288         channel_id = self._match_id(url)
 289
 290         channel = self._call_api(
 291             'containers/%s.json' % channel_id, channel_id,
 292             'Downloading channel JSON')
 293
 294         titles = channel['titles']
 295         title = titles.get('en') or titles[titles.keys()[0]]
 296
 297         descriptions = channel['descriptions']
 298         description = descriptions.get('en') or descriptions[descriptions.keys()[0]]
 299
 300         entries = []
 301         for video_type in ('episodes', 'clips', 'movies'):
 302             for page_num in itertools.count(1):
 303                 page = self._call_api(
 304                     'containers/%s/%s.json?per_page=%d&sort=number&direction=asc&with_paging=true&page=%d'
 305                     % (channel_id, video_type, self._PER_PAGE, page_num), channel_id,
 306                     'Downloading %s JSON page #%d' % (video_type, page_num))
 307                 for video in page['response']:
 308                     video_id = video['id']
 309                     entries.append(self.url_result(
 310                         'http://www.viki.com/videos/%s' % video_id, 'Viki'))
 311                 if not page['pagination']['next']:
 312                     break
 313
 314         return self.playlist_result(entries, channel_id, title, description)