yt_dlp/extractor/vrv.py

   1 import base64
   2 import json
   3 import hashlib
   4 import hmac
   5 import random
   6 import string
   7 import time
   8
   9 from .common import InfoExtractor
  10 from ..compat import (
  11     compat_HTTPError,
  12     compat_urllib_parse_urlencode,
  13     compat_urllib_parse,
  14 )
  15 from ..utils import (
  16     ExtractorError,
  17     float_or_none,
  18     int_or_none,
  19     join_nonempty,
  20     traverse_obj,
  21 )
  22
  23
  24 class VRVBaseIE(InfoExtractor):
  25     _API_DOMAIN = None
  26     _API_PARAMS = {}
  27     _CMS_SIGNING = {}
  28     _TOKEN = None
  29     _TOKEN_SECRET = ''
  30
  31     def _call_api(self, path, video_id, note, data=None):
  32         # https://tools.ietf.org/html/rfc5849#section-3
  33         base_url = self._API_DOMAIN + '/core/' + path
  34         query = [
  35             ('oauth_consumer_key', self._API_PARAMS['oAuthKey']),
  36             ('oauth_nonce', ''.join([random.choice(string.ascii_letters) for _ in range(32)])),
  37             ('oauth_signature_method', 'HMAC-SHA1'),
  38             ('oauth_timestamp', int(time.time())),
  39         ]
  40         if self._TOKEN:
  41             query.append(('oauth_token', self._TOKEN))
  42         encoded_query = compat_urllib_parse_urlencode(query)
  43         headers = self.geo_verification_headers()
  44         if data:
  45             data = json.dumps(data).encode()
  46             headers['Content-Type'] = 'application/json'
  47         base_string = '&'.join([
  48             'POST' if data else 'GET',
  49             compat_urllib_parse.quote(base_url, ''),
  50             compat_urllib_parse.quote(encoded_query, '')])
  51         oauth_signature = base64.b64encode(hmac.new(
  52             (self._API_PARAMS['oAuthSecret'] + '&' + self._TOKEN_SECRET).encode('ascii'),
  53             base_string.encode(), hashlib.sha1).digest()).decode()
  54         encoded_query += '&oauth_signature=' + compat_urllib_parse.quote(oauth_signature, '')
  55         try:
  56             return self._download_json(
  57                 '?'.join([base_url, encoded_query]), video_id,
  58                 note='Downloading %s JSON metadata' % note, headers=headers, data=data)
  59         except ExtractorError as e:
  60             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
  61                 raise ExtractorError(json.loads(e.cause.read().decode())['message'], expected=True)
  62             raise
  63
  64     def _call_cms(self, path, video_id, note):
  65         if not self._CMS_SIGNING:
  66             index = self._call_api('index', video_id, 'CMS Signing')
  67             self._CMS_SIGNING = index.get('cms_signing') or {}
  68             if not self._CMS_SIGNING:
  69                 for signing_policy in index.get('signing_policies', []):
  70                     signing_path = signing_policy.get('path')
  71                     if signing_path and signing_path.startswith('/cms/'):
  72                         name, value = signing_policy.get('name'), signing_policy.get('value')
  73                         if name and value:
  74                             self._CMS_SIGNING[name] = value
  75         return self._download_json(
  76             self._API_DOMAIN + path, video_id, query=self._CMS_SIGNING,
  77             note='Downloading %s JSON metadata' % note, headers=self.geo_verification_headers())
  78
  79     def _get_cms_resource(self, resource_key, video_id):
  80         return self._call_api(
  81             'cms_resource', video_id, 'resource path', data={
  82                 'resource_key': resource_key,
  83             })['__links__']['cms_resource']['href']
  84
  85     def _extract_vrv_formats(self, url, video_id, stream_format, audio_lang, hardsub_lang):
  86         if not url or stream_format not in ('hls', 'dash', 'adaptive_hls'):
  87             return []
  88         format_id = join_nonempty(
  89             stream_format,
  90             audio_lang and 'audio-%s' % audio_lang,
  91             hardsub_lang and 'hardsub-%s' % hardsub_lang)
  92         if 'hls' in stream_format:
  93             adaptive_formats = self._extract_m3u8_formats(
  94                 url, video_id, 'mp4', m3u8_id=format_id,
  95                 note='Downloading %s information' % format_id,
  96                 fatal=False)
  97         elif stream_format == 'dash':
  98             adaptive_formats = self._extract_mpd_formats(
  99                 url, video_id, mpd_id=format_id,
 100                 note='Downloading %s information' % format_id,
 101                 fatal=False)
 102         if audio_lang:
 103             for f in adaptive_formats:
 104                 if f.get('acodec') != 'none':
 105                     f['language'] = audio_lang
 106         return adaptive_formats
 107
 108     def _set_api_params(self):
 109         webpage = self._download_webpage(
 110             'https://vrv.co/', None, headers=self.geo_verification_headers())
 111         self._API_PARAMS = self._parse_json(self._search_regex(
 112             [
 113                 r'window\.__APP_CONFIG__\s*=\s*({.+?})(?:</script>|;)',
 114                 r'window\.__APP_CONFIG__\s*=\s*({.+})'
 115             ], webpage, 'app config'), None)['cxApiParams']
 116         self._API_DOMAIN = self._API_PARAMS.get('apiDomain', 'https://api.vrv.co')
 117
 118
 119 class VRVIE(VRVBaseIE):
 120     IE_NAME = 'vrv'
 121     _VALID_URL = r'https?://(?:www\.)?vrv\.co/watch/(?P<id>[A-Z0-9]+)'
 122     _TESTS = [{
 123         'url': 'https://vrv.co/watch/GR9PNZ396/Hidden-America-with-Jonah-Ray:BOSTON-WHERE-THE-PAST-IS-THE-PRESENT',
 124         'info_dict': {
 125             'id': 'GR9PNZ396',
 126             'ext': 'mp4',
 127             'title': 'BOSTON: WHERE THE PAST IS THE PRESENT',
 128             'description': 'md5:4ec8844ac262ca2df9e67c0983c6b83f',
 129             'uploader_id': 'seeso',
 130         },
 131         'params': {
 132             # m3u8 download
 133             'skip_download': True,
 134         },
 135     }, {
 136         # movie listing
 137         'url': 'https://vrv.co/watch/G6NQXZ1J6/Lily-CAT',
 138         'info_dict': {
 139             'id': 'G6NQXZ1J6',
 140             'title': 'Lily C.A.T',
 141             'description': 'md5:988b031e7809a6aeb60968be4af7db07',
 142         },
 143         'playlist_count': 2,
 144     }]
 145     _NETRC_MACHINE = 'vrv'
 146
 147     def _perform_login(self, username, password):
 148         token_credentials = self._call_api(
 149             'authenticate/by:credentials', None, 'Token Credentials', data={
 150                 'email': username,
 151                 'password': password,
 152             })
 153         self._TOKEN = token_credentials['oauth_token']
 154         self._TOKEN_SECRET = token_credentials['oauth_token_secret']
 155
 156     def _initialize_pre_login(self):
 157         return self._set_api_params()
 158
 159     def _real_extract(self, url):
 160         video_id = self._match_id(url)
 161
 162         object_data = self._call_cms(self._get_cms_resource(
 163             'cms:/objects/' + video_id, video_id), video_id, 'object')['items'][0]
 164         resource_path = object_data['__links__']['resource']['href']
 165         video_data = self._call_cms(resource_path, video_id, 'video')
 166         title = video_data['title']
 167         description = video_data.get('description')
 168
 169         if video_data.get('__class__') == 'movie_listing':
 170             items = self._call_cms(
 171                 video_data['__links__']['movie_listing/movies']['href'],
 172                 video_id, 'movie listing').get('items') or []
 173             if len(items) != 1:
 174                 entries = []
 175                 for item in items:
 176                     item_id = item.get('id')
 177                     if not item_id:
 178                         continue
 179                     entries.append(self.url_result(
 180                         'https://vrv.co/watch/' + item_id,
 181                         self.ie_key(), item_id, item.get('title')))
 182                 return self.playlist_result(entries, video_id, title, description)
 183             video_data = items[0]
 184
 185         streams_path = video_data['__links__'].get('streams', {}).get('href')
 186         if not streams_path:
 187             self.raise_login_required()
 188         streams_json = self._call_cms(streams_path, video_id, 'streams')
 189
 190         audio_locale = streams_json.get('audio_locale')
 191         formats = []
 192         for stream_type, streams in streams_json.get('streams', {}).items():
 193             if stream_type in ('adaptive_hls', 'adaptive_dash'):
 194                 for stream in streams.values():
 195                     formats.extend(self._extract_vrv_formats(
 196                         stream.get('url'), video_id, stream_type.split('_')[1],
 197                         audio_locale, stream.get('hardsub_locale')))
 198         self._sort_formats(formats)
 199
 200         subtitles = {}
 201         for k in ('captions', 'subtitles'):
 202             for subtitle in streams_json.get(k, {}).values():
 203                 subtitle_url = subtitle.get('url')
 204                 if not subtitle_url:
 205                     continue
 206                 subtitles.setdefault(subtitle.get('locale', 'en-US'), []).append({
 207                     'url': subtitle_url,
 208                     'ext': subtitle.get('format', 'ass'),
 209                 })
 210
 211         thumbnails = []
 212         for thumbnail in traverse_obj(video_data, ('images', 'thumbnail', ..., ...)) or []:
 213             thumbnail_url = thumbnail.get('source')
 214             if not thumbnail_url:
 215                 continue
 216             thumbnails.append({
 217                 'url': thumbnail_url,
 218                 'width': int_or_none(thumbnail.get('width')),
 219                 'height': int_or_none(thumbnail.get('height')),
 220             })
 221
 222         return {
 223             'id': video_id,
 224             'title': title,
 225             'formats': formats,
 226             'subtitles': subtitles,
 227             'thumbnails': thumbnails,
 228             'description': description,
 229             'duration': float_or_none(video_data.get('duration_ms'), 1000),
 230             'uploader_id': video_data.get('channel_id'),
 231             'series': video_data.get('series_title'),
 232             'season': video_data.get('season_title'),
 233             'season_number': int_or_none(video_data.get('season_number')),
 234             'season_id': video_data.get('season_id'),
 235             'episode': title,
 236             'episode_number': int_or_none(video_data.get('episode_number')),
 237             'episode_id': video_data.get('production_episode_id'),
 238         }
 239
 240
 241 class VRVSeriesIE(VRVBaseIE):
 242     IE_NAME = 'vrv:series'
 243     _VALID_URL = r'https?://(?:www\.)?vrv\.co/series/(?P<id>[A-Z0-9]+)'
 244     _TEST = {
 245         'url': 'https://vrv.co/series/G68VXG3G6/The-Perfect-Insider',
 246         'info_dict': {
 247             'id': 'G68VXG3G6',
 248         },
 249         'playlist_mincount': 11,
 250     }
 251
 252     def _initialize_pre_login(self):
 253         return self._set_api_params()
 254
 255     def _real_extract(self, url):
 256         series_id = self._match_id(url)
 257
 258         seasons_path = self._get_cms_resource(
 259             'cms:/seasons?series_id=' + series_id, series_id)
 260         seasons_data = self._call_cms(seasons_path, series_id, 'seasons')
 261
 262         entries = []
 263         for season in seasons_data.get('items', []):
 264             episodes_path = season['__links__']['season/episodes']['href']
 265             episodes = self._call_cms(episodes_path, series_id, 'episodes')
 266             for episode in episodes.get('items', []):
 267                 episode_id = episode['id']
 268                 entries.append(self.url_result(
 269                     'https://vrv.co/watch/' + episode_id,
 270                     'VRV', episode_id, episode.get('title')))
 271
 272         return self.playlist_result(entries, series_id)