[radiko] Fix extractor (#3655)

Authored by: Lesmiscore
This commit is contained in:
Lesmiscore 2022-05-08 00:47:51 +09:00 committed by GitHub
parent d4736fdb43
commit 1f8b4ab733
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -1,26 +1,22 @@
import re
import base64 import base64
import calendar import re
import datetime import urllib.parse
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import ( from ..utils import (
ExtractorError, ExtractorError,
update_url_query,
clean_html, clean_html,
time_seconds,
try_call,
unified_timestamp, unified_timestamp,
update_url_query,
) )
from ..compat import compat_urllib_parse
class RadikoBaseIE(InfoExtractor): class RadikoBaseIE(InfoExtractor):
_FULL_KEY = None _FULL_KEY = None
def _auth_client(self): def _auth_client(self):
auth_cache = self._downloader.cache.load('radiko', 'auth_data')
if auth_cache:
return auth_cache
_, auth1_handle = self._download_webpage_handle( _, auth1_handle = self._download_webpage_handle(
'https://radiko.jp/v2/api/auth1', None, 'Downloading authentication page', 'https://radiko.jp/v2/api/auth1', None, 'Downloading authentication page',
headers={ headers={
@ -89,8 +85,8 @@ def _find_program(self, video_id, station, cursor):
def _extract_formats(self, video_id, station, is_onair, ft, cursor, auth_token, area_id, query): def _extract_formats(self, video_id, station, is_onair, ft, cursor, auth_token, area_id, query):
m3u8_playlist_data = self._download_xml( m3u8_playlist_data = self._download_xml(
'https://radiko.jp/v3/station/stream/pc_html5/%s.xml' % station, video_id, f'https://radiko.jp/v3/station/stream/pc_html5/{station}.xml', video_id,
note='Downloading m3u8 information') note='Downloading stream information')
m3u8_urls = m3u8_playlist_data.findall('.//url') m3u8_urls = m3u8_playlist_data.findall('.//url')
formats = [] formats = []
@ -102,7 +98,7 @@ def _extract_formats(self, video_id, station, is_onair, ft, cursor, auth_token,
'station_id': station, 'station_id': station,
**query, **query,
'l': '15', 'l': '15',
'lsid': '77d0678df93a1034659c14d6fc89f018', 'lsid': '88ecea37e968c1f17d5413312d9f8003',
'type': 'b', 'type': 'b',
}) })
if playlist_url in found: if playlist_url in found:
@ -112,16 +108,17 @@ def _extract_formats(self, video_id, station, is_onair, ft, cursor, auth_token,
time_to_skip = None if is_onair else cursor - ft time_to_skip = None if is_onair else cursor - ft
domain = urllib.parse.urlparse(playlist_url).netloc
subformats = self._extract_m3u8_formats( subformats = self._extract_m3u8_formats(
playlist_url, video_id, ext='m4a', playlist_url, video_id, ext='m4a',
live=True, fatal=False, m3u8_id=None, live=True, fatal=False, m3u8_id=domain,
note=f'Downloading m3u8 information from {domain}',
headers={ headers={
'X-Radiko-AreaId': area_id, 'X-Radiko-AreaId': area_id,
'X-Radiko-AuthToken': auth_token, 'X-Radiko-AuthToken': auth_token,
}) })
for sf in subformats: for sf in subformats:
domain = sf['format_id'] = compat_urllib_parse.urlparse(sf['url']).netloc if re.fullmatch(r'[cf]-radiko\.smartstream\.ne\.jp', domain):
if re.match(r'^[cf]-radiko\.smartstream\.ne\.jp$', domain):
# Prioritize live radio vs playback based on extractor # Prioritize live radio vs playback based on extractor
sf['preference'] = 100 if is_onair else -100 sf['preference'] = 100 if is_onair else -100
if not is_onair and url_attrib['timefree'] == '1' and time_to_skip: if not is_onair and url_attrib['timefree'] == '1' and time_to_skip:
@ -151,15 +148,11 @@ class RadikoIE(RadikoBaseIE):
def _real_extract(self, url): def _real_extract(self, url):
station, video_id = self._match_valid_url(url).groups() station, video_id = self._match_valid_url(url).groups()
vid_int = unified_timestamp(video_id, False) vid_int = unified_timestamp(video_id, False)
auth_token, area_id = self._auth_client()
prog, station_program, ft, radio_begin, radio_end = self._find_program(video_id, station, vid_int) prog, station_program, ft, radio_begin, radio_end = self._find_program(video_id, station, vid_int)
title = prog.find('title').text auth_cache = self._downloader.cache.load('radiko', 'auth_data')
description = clean_html(prog.find('info').text) for attempt in range(2):
station_name = station_program.find('.//name').text auth_token, area_id = (not attempt and auth_cache) or self._auth_client()
formats = self._extract_formats( formats = self._extract_formats(
video_id=video_id, station=station, is_onair=False, video_id=video_id, station=station, is_onair=False,
ft=ft, cursor=vid_int, auth_token=auth_token, area_id=area_id, ft=ft, cursor=vid_int, auth_token=auth_token, area_id=area_id,
@ -170,12 +163,14 @@ def _real_extract(self, url):
'to': radio_end, 'to': radio_end,
'seek': video_id, 'seek': video_id,
}) })
if formats:
break
return { return {
'id': video_id, 'id': video_id,
'title': title, 'title': try_call(lambda: prog.find('title').text),
'description': description, 'description': clean_html(try_call(lambda: prog.find('info').text)),
'uploader': station_name, 'uploader': try_call(lambda: station_program.find('.//name').text),
'uploader_id': station, 'uploader_id': station,
'timestamp': vid_int, 'timestamp': vid_int,
'formats': formats, 'formats': formats,
@ -205,8 +200,7 @@ def _real_extract(self, url):
auth_token, area_id = self._auth_client() auth_token, area_id = self._auth_client()
# get current time in JST (GMT+9:00 w/o DST) # get current time in JST (GMT+9:00 w/o DST)
vid_now = datetime.datetime.now(datetime.timezone(datetime.timedelta(hours=9))) vid_now = time_seconds(hours=9)
vid_now = calendar.timegm(vid_now.timetuple())
prog, station_program, ft, _, _ = self._find_program(station, station, vid_now) prog, station_program, ft, _, _ = self._find_program(station, station, vid_now)