X-Git-Url: https://jfr.im/git/yt-dlp.git/blobdiff_plain/9809740ba5cc5daf53e690d104a37aa6545e53f9..5dbac313ae4e3e8521dfe2e1a6a048a98ff4b4fe:/yt_dlp/extractor/radiko.py diff --git a/yt_dlp/extractor/radiko.py b/yt_dlp/extractor/radiko.py index 498cc6be9..b0b6681c9 100644 --- a/yt_dlp/extractor/radiko.py +++ b/yt_dlp/extractor/radiko.py @@ -1,4 +1,5 @@ import base64 +import random import re import urllib.parse @@ -11,12 +12,31 @@ unified_timestamp, update_url_query, ) +from ..utils.traversal import traverse_obj class RadikoBaseIE(InfoExtractor): + _GEO_BYPASS = False _FULL_KEY = None - - def _auth_client(self): + _HOSTS_FOR_TIME_FREE_FFMPEG_UNSUPPORTED = ( + 'https://c-rpaa.smartstream.ne.jp', + 'https://si-c-radiko.smartstream.ne.jp', + 'https://tf-f-rpaa-radiko.smartstream.ne.jp', + 'https://tf-c-rpaa-radiko.smartstream.ne.jp', + 'https://si-f-radiko.smartstream.ne.jp', + 'https://rpaa.smartstream.ne.jp', + ) + _HOSTS_FOR_TIME_FREE_FFMPEG_SUPPORTED = ( + 'https://rd-wowza-radiko.radiko-cf.com', + 'https://radiko.jp', + 'https://f-radiko.smartstream.ne.jp', + ) + # Following URL forcibly connects not Time Free but Live + _HOSTS_FOR_LIVE = ( + 'https://c-radiko.smartstream.ne.jp', + ) + + def _negotiate_token(self): _, auth1_handle = self._download_webpage_handle( 'https://radiko.jp/v2/api/auth1', None, 'Downloading authentication page', headers={ @@ -25,7 +45,7 @@ def _auth_client(self): 'x-radiko-device': 'pc', 'x-radiko-user': 'dummy_user', }) - auth1_header = auth1_handle.info() + auth1_header = auth1_handle.headers auth_token = auth1_header['X-Radiko-AuthToken'] kl = int(auth1_header['X-Radiko-KeyLength']) @@ -42,10 +62,23 @@ def _auth_client(self): 'x-radiko-partialkey': partial_key, }).split(',')[0] + if area_id == 'OUT': + self.raise_geo_restricted(countries=['JP']) + auth_data = (auth_token, area_id) self.cache.store('radiko', 'auth_data', auth_data) return auth_data + def _auth_client(self): + cachedata = self.cache.load('radiko', 'auth_data') + if cachedata is not None: + response = self._download_webpage( + 'https://radiko.jp/v2/api/auth_check', None, 'Checking cached token', expected_status=401, + headers={'X-Radiko-AuthToken': cachedata[0], 'X-Radiko-AreaId': cachedata[1]}) + if response == 'OK': + return cachedata + return self._negotiate_token() + def _extract_full_key(self): if self._FULL_KEY: return self._FULL_KEY @@ -59,7 +92,7 @@ def _extract_full_key(self): if full_key: full_key = full_key.encode() - else: # use full key ever known + else: # use only full key ever known full_key = b'bcd151073c03b352e1ef2fd66c32209da9ca0afa' self._FULL_KEY = full_key @@ -67,8 +100,8 @@ def _extract_full_key(self): def _find_program(self, video_id, station, cursor): station_program = self._download_xml( - 'https://radiko.jp/v3/program/station/weekly/%s.xml' % station, video_id, - note='Downloading radio program for %s station' % station) + f'https://radiko.jp/v3/program/station/weekly/{station}.xml', video_id, + note=f'Downloading radio program for {station} station') prog = None for p in station_program.findall('.//prog'): @@ -87,24 +120,24 @@ def _extract_formats(self, video_id, station, is_onair, ft, cursor, auth_token, m3u8_playlist_data = self._download_xml( f'https://radiko.jp/v3/station/stream/pc_html5/{station}.xml', video_id, note='Downloading stream information') - m3u8_urls = m3u8_playlist_data.findall('.//url') formats = [] found = set() - for url_tag in m3u8_urls: - pcu = url_tag.find('playlist_create_url') - url_attrib = url_tag.attrib - playlist_url = update_url_query(pcu.text, { + + timefree_int = 0 if is_onair else 1 + + for element in m3u8_playlist_data.findall(f'.//url[@timefree="{timefree_int}"]/playlist_create_url'): + pcu = element.text + if pcu in found: + continue + found.add(pcu) + playlist_url = update_url_query(pcu, { 'station_id': station, **query, 'l': '15', - 'lsid': '88ecea37e968c1f17d5413312d9f8003', + 'lsid': ''.join(random.choices('0123456789abcdef', k=32)), 'type': 'b', }) - if playlist_url in found: - continue - else: - found.add(playlist_url) time_to_skip = None if is_onair else cursor - ft @@ -118,16 +151,20 @@ def _extract_formats(self, video_id, station, is_onair, ft, cursor, auth_token, 'X-Radiko-AuthToken': auth_token, }) for sf in subformats: - if re.fullmatch(r'[cf]-radiko\.smartstream\.ne\.jp', domain): - # Prioritize live radio vs playback based on extractor - sf['preference'] = 100 if is_onair else -100 - if not is_onair and url_attrib['timefree'] == '1' and time_to_skip: - sf['downloader_options'] = {'ffmpeg_args': ['-ss', time_to_skip]} + if (is_onair ^ pcu.startswith(self._HOSTS_FOR_LIVE)) or ( + not is_onair and pcu.startswith(self._HOSTS_FOR_TIME_FREE_FFMPEG_UNSUPPORTED)): + sf['preference'] = -100 + sf['format_note'] = 'not preferred' + if not is_onair and timefree_int == 1 and time_to_skip: + sf['downloader_options'] = {'ffmpeg_args': ['-ss', str(time_to_skip)]} formats.extend(subformats) - self._sort_formats(formats) return formats + def _extract_performers(self, prog): + return traverse_obj(prog, ( + 'pfm/text()', ..., {lambda x: re.split(r'[/／、　,，]', x)}, ..., {str.strip})) or None + class RadikoIE(RadikoBaseIE): _VALID_URL = r'https?://(?:www\.)?radiko\.jp/#!/ts/(?P[A-Z0-9-]+)/(?P\d+)' @@ -150,31 +187,29 @@ def _real_extract(self, url): vid_int = unified_timestamp(video_id, False) prog, station_program, ft, radio_begin, radio_end = self._find_program(video_id, station, vid_int) - auth_cache = self.cache.load('radiko', 'auth_data') - for attempt in range(2): - auth_token, area_id = (not attempt and auth_cache) or self._auth_client() - formats = self._extract_formats( - video_id=video_id, station=station, is_onair=False, - ft=ft, cursor=vid_int, auth_token=auth_token, area_id=area_id, - query={ - 'start_at': radio_begin, - 'ft': radio_begin, - 'end_at': radio_end, - 'to': radio_end, - 'seek': video_id, - }) - if formats: - break + auth_token, area_id = self._auth_client() return { 'id': video_id, 'title': try_call(lambda: prog.find('title').text), + 'cast': self._extract_performers(prog), 'description': clean_html(try_call(lambda: prog.find('info').text)), 'uploader': try_call(lambda: station_program.find('.//name').text), 'uploader_id': station, 'timestamp': vid_int, - 'formats': formats, + 'duration': try_call(lambda: unified_timestamp(radio_end, False) - unified_timestamp(radio_begin, False)), 'is_live': True, + 'formats': self._extract_formats( + video_id=video_id, station=station, is_onair=False, + ft=ft, cursor=vid_int, auth_token=auth_token, area_id=area_id, + query={ + 'start_at': radio_begin, + 'ft': radio_begin, + 'end_at': radio_end, + 'to': radio_end, + 'seek': video_id, + }, + ), } @@ -216,6 +251,7 @@ def _real_extract(self, url): return { 'id': station, 'title': title, + 'cast': self._extract_performers(prog), 'description': description, 'uploader': station_name, 'uploader_id': station,