yt_dlp/extractor/radiko.py

   1 import base64
   2 import random
   3 import urllib.parse
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     ExtractorError,
   8     clean_html,
   9     time_seconds,
  10     try_call,
  11     unified_timestamp,
  12     update_url_query,
  13 )
  14
  15
  16 class RadikoBaseIE(InfoExtractor):
  17     _GEO_BYPASS = False
  18     _FULL_KEY = None
  19     _HOSTS_FOR_TIME_FREE_FFMPEG_UNSUPPORTED = (
  20         'https://c-rpaa.smartstream.ne.jp',
  21         'https://si-c-radiko.smartstream.ne.jp',
  22         'https://tf-f-rpaa-radiko.smartstream.ne.jp',
  23         'https://tf-c-rpaa-radiko.smartstream.ne.jp',
  24         'https://si-f-radiko.smartstream.ne.jp',
  25         'https://rpaa.smartstream.ne.jp',
  26     )
  27     _HOSTS_FOR_TIME_FREE_FFMPEG_SUPPORTED = (
  28         'https://rd-wowza-radiko.radiko-cf.com',
  29         'https://radiko.jp',
  30         'https://f-radiko.smartstream.ne.jp',
  31     )
  32     # Following URL forcibly connects not Time Free but Live
  33     _HOSTS_FOR_LIVE = (
  34         'https://c-radiko.smartstream.ne.jp',
  35     )
  36
  37     def _negotiate_token(self):
  38         _, auth1_handle = self._download_webpage_handle(
  39             'https://radiko.jp/v2/api/auth1', None, 'Downloading authentication page',
  40             headers={
  41                 'x-radiko-app': 'pc_html5',
  42                 'x-radiko-app-version': '0.0.1',
  43                 'x-radiko-device': 'pc',
  44                 'x-radiko-user': 'dummy_user',
  45             })
  46         auth1_header = auth1_handle.headers
  47
  48         auth_token = auth1_header['X-Radiko-AuthToken']
  49         kl = int(auth1_header['X-Radiko-KeyLength'])
  50         ko = int(auth1_header['X-Radiko-KeyOffset'])
  51         raw_partial_key = self._extract_full_key()[ko:ko + kl]
  52         partial_key = base64.b64encode(raw_partial_key).decode()
  53
  54         area_id = self._download_webpage(
  55             'https://radiko.jp/v2/api/auth2', None, 'Authenticating',
  56             headers={
  57                 'x-radiko-device': 'pc',
  58                 'x-radiko-user': 'dummy_user',
  59                 'x-radiko-authtoken': auth_token,
  60                 'x-radiko-partialkey': partial_key,
  61             }).split(',')[0]
  62
  63         if area_id == 'OUT':
  64             self.raise_geo_restricted(countries=['JP'])
  65
  66         auth_data = (auth_token, area_id)
  67         self.cache.store('radiko', 'auth_data', auth_data)
  68         return auth_data
  69
  70     def _auth_client(self):
  71         cachedata = self.cache.load('radiko', 'auth_data')
  72         if cachedata is not None:
  73             response = self._download_webpage(
  74                 'https://radiko.jp/v2/api/auth_check', None, 'Checking cached token', expected_status=401,
  75                 headers={'X-Radiko-AuthToken': cachedata[0], 'X-Radiko-AreaId': cachedata[1]})
  76             if response == 'OK':
  77                 return cachedata
  78         return self._negotiate_token()
  79
  80     def _extract_full_key(self):
  81         if self._FULL_KEY:
  82             return self._FULL_KEY
  83
  84         jscode = self._download_webpage(
  85             'https://radiko.jp/apps/js/playerCommon.js', None,
  86             note='Downloading player js code')
  87         full_key = self._search_regex(
  88             (r"RadikoJSPlayer\([^,]*,\s*(['\"])pc_html5\1,\s*(['\"])(?P<fullkey>[0-9a-f]+)\2,\s*{"),
  89             jscode, 'full key', fatal=False, group='fullkey')
  90
  91         if full_key:
  92             full_key = full_key.encode()
  93         else:  # use only full key ever known
  94             full_key = b'bcd151073c03b352e1ef2fd66c32209da9ca0afa'
  95
  96         self._FULL_KEY = full_key
  97         return full_key
  98
  99     def _find_program(self, video_id, station, cursor):
 100         station_program = self._download_xml(
 101             'https://radiko.jp/v3/program/station/weekly/%s.xml' % station, video_id,
 102             note='Downloading radio program for %s station' % station)
 103
 104         prog = None
 105         for p in station_program.findall('.//prog'):
 106             ft_str, to_str = p.attrib['ft'], p.attrib['to']
 107             ft = unified_timestamp(ft_str, False)
 108             to = unified_timestamp(to_str, False)
 109             if ft <= cursor and cursor < to:
 110                 prog = p
 111                 break
 112         if not prog:
 113             raise ExtractorError('Cannot identify radio program to download!')
 114         assert ft, to
 115         return prog, station_program, ft, ft_str, to_str
 116
 117     def _extract_formats(self, video_id, station, is_onair, ft, cursor, auth_token, area_id, query):
 118         m3u8_playlist_data = self._download_xml(
 119             f'https://radiko.jp/v3/station/stream/pc_html5/{station}.xml', video_id,
 120             note='Downloading stream information')
 121
 122         formats = []
 123         found = set()
 124
 125         timefree_int = 0 if is_onair else 1
 126
 127         for element in m3u8_playlist_data.findall(f'.//url[@timefree="{timefree_int}"]/playlist_create_url'):
 128             pcu = element.text
 129             if pcu in found:
 130                 continue
 131             found.add(pcu)
 132             playlist_url = update_url_query(pcu, {
 133                 'station_id': station,
 134                 **query,
 135                 'l': '15',
 136                 'lsid': ''.join(random.choices('0123456789abcdef', k=32)),
 137                 'type': 'b',
 138             })
 139
 140             time_to_skip = None if is_onair else cursor - ft
 141
 142             domain = urllib.parse.urlparse(playlist_url).netloc
 143             subformats = self._extract_m3u8_formats(
 144                 playlist_url, video_id, ext='m4a',
 145                 live=True, fatal=False, m3u8_id=domain,
 146                 note=f'Downloading m3u8 information from {domain}',
 147                 headers={
 148                     'X-Radiko-AreaId': area_id,
 149                     'X-Radiko-AuthToken': auth_token,
 150                 })
 151             for sf in subformats:
 152                 if (is_onair ^ pcu.startswith(self._HOSTS_FOR_LIVE)) or (
 153                         not is_onair and pcu.startswith(self._HOSTS_FOR_TIME_FREE_FFMPEG_UNSUPPORTED)):
 154                     sf['preference'] = -100
 155                     sf['format_note'] = 'not preferred'
 156                 if not is_onair and timefree_int == 1 and time_to_skip:
 157                     sf['downloader_options'] = {'ffmpeg_args': ['-ss', str(time_to_skip)]}
 158             formats.extend(subformats)
 159
 160         return formats
 161
 162
 163 class RadikoIE(RadikoBaseIE):
 164     _VALID_URL = r'https?://(?:www\.)?radiko\.jp/#!/ts/(?P<station>[A-Z0-9-]+)/(?P<id>\d+)'
 165
 166     _TESTS = [{
 167         # QRR (文化放送) station provides <desc>
 168         'url': 'https://radiko.jp/#!/ts/QRR/20210425101300',
 169         'only_matching': True,
 170     }, {
 171         # FMT (TOKYO FM) station does not provide <desc>
 172         'url': 'https://radiko.jp/#!/ts/FMT/20210810150000',
 173         'only_matching': True,
 174     }, {
 175         'url': 'https://radiko.jp/#!/ts/JOAK-FM/20210509090000',
 176         'only_matching': True,
 177     }]
 178
 179     def _real_extract(self, url):
 180         station, video_id = self._match_valid_url(url).groups()
 181         vid_int = unified_timestamp(video_id, False)
 182         prog, station_program, ft, radio_begin, radio_end = self._find_program(video_id, station, vid_int)
 183
 184         auth_token, area_id = self._auth_client()
 185
 186         return {
 187             'id': video_id,
 188             'title': try_call(lambda: prog.find('title').text),
 189             'description': clean_html(try_call(lambda: prog.find('info').text)),
 190             'uploader': try_call(lambda: station_program.find('.//name').text),
 191             'uploader_id': station,
 192             'timestamp': vid_int,
 193             'is_live': True,
 194             'formats': self._extract_formats(
 195                 video_id=video_id, station=station, is_onair=False,
 196                 ft=ft, cursor=vid_int, auth_token=auth_token, area_id=area_id,
 197                 query={
 198                     'start_at': radio_begin,
 199                     'ft': radio_begin,
 200                     'end_at': radio_end,
 201                     'to': radio_end,
 202                     'seek': video_id
 203                 }
 204             ),
 205         }
 206
 207
 208 class RadikoRadioIE(RadikoBaseIE):
 209     _VALID_URL = r'https?://(?:www\.)?radiko\.jp/#!/live/(?P<id>[A-Z0-9-]+)'
 210
 211     _TESTS = [{
 212         # QRR (文化放送) station provides <desc>
 213         'url': 'https://radiko.jp/#!/live/QRR',
 214         'only_matching': True,
 215     }, {
 216         # FMT (TOKYO FM) station does not provide <desc>
 217         'url': 'https://radiko.jp/#!/live/FMT',
 218         'only_matching': True,
 219     }, {
 220         'url': 'https://radiko.jp/#!/live/JOAK-FM',
 221         'only_matching': True,
 222     }]
 223
 224     def _real_extract(self, url):
 225         station = self._match_id(url)
 226         self.report_warning('Downloader will not stop at the end of the program! Press Ctrl+C to stop')
 227
 228         auth_token, area_id = self._auth_client()
 229         # get current time in JST (GMT+9:00 w/o DST)
 230         vid_now = time_seconds(hours=9)
 231
 232         prog, station_program, ft, _, _ = self._find_program(station, station, vid_now)
 233
 234         title = prog.find('title').text
 235         description = clean_html(prog.find('info').text)
 236         station_name = station_program.find('.//name').text
 237
 238         formats = self._extract_formats(
 239             video_id=station, station=station, is_onair=True,
 240             ft=ft, cursor=vid_now, auth_token=auth_token, area_id=area_id,
 241             query={})
 242
 243         return {
 244             'id': station,
 245             'title': title,
 246             'description': description,
 247             'uploader': station_name,
 248             'uploader_id': station,
 249             'timestamp': ft,
 250             'formats': formats,
 251             'is_live': True,
 252         }