yt_dlp/extractor/nhk.py

   1 import re
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     ExtractorError,
   6     int_or_none,
   7     join_nonempty,
   8     parse_duration,
   9     traverse_obj,
  10     unescapeHTML,
  11     unified_timestamp,
  12     url_or_none,
  13     urljoin,
  14 )
  15
  16
  17 class NhkBaseIE(InfoExtractor):
  18     _API_URL_TEMPLATE = 'https://nwapi.nhk.jp/nhkworld/%sod%slist/v7b/%s/%s/%s/all%s.json'
  19     _BASE_URL_REGEX = r'https?://www3\.nhk\.or\.jp/nhkworld/(?P<lang>[a-z]{2})/ondemand'
  20     _TYPE_REGEX = r'/(?P<type>video|audio)/'
  21
  22     def _call_api(self, m_id, lang, is_video, is_episode, is_clip):
  23         return self._download_json(
  24             self._API_URL_TEMPLATE % (
  25                 'v' if is_video else 'r',
  26                 'clip' if is_clip else 'esd',
  27                 'episode' if is_episode else 'program',
  28                 m_id, lang, '/all' if is_video else ''),
  29             m_id, query={'apikey': 'EJfK8jdS57GqlupFgAfAAwr573q01y6k'})['data']['episodes'] or []
  30
  31     def _get_api_info(self, refresh=True):
  32         if not refresh:
  33             return self.cache.load('nhk', 'api_info')
  34
  35         self.cache.store('nhk', 'api_info', {})
  36         movie_player_js = self._download_webpage(
  37             'https://movie-a.nhk.or.jp/world/player/js/movie-player.js', None,
  38             note='Downloading stream API information')
  39         api_info = {
  40             'url': self._search_regex(
  41                 r'prod:[^;]+\bapiUrl:\s*[\'"]([^\'"]+)[\'"]', movie_player_js, None, 'stream API url'),
  42             'token': self._search_regex(
  43                 r'prod:[^;]+\btoken:\s*[\'"]([^\'"]+)[\'"]', movie_player_js, None, 'stream API token'),
  44         }
  45         self.cache.store('nhk', 'api_info', api_info)
  46         return api_info
  47
  48     def _extract_formats_and_subtitles(self, vod_id):
  49         for refresh in (False, True):
  50             api_info = self._get_api_info(refresh)
  51             if not api_info:
  52                 continue
  53
  54             api_url = api_info.pop('url')
  55             stream_url = traverse_obj(
  56                 self._download_json(
  57                     api_url, vod_id, 'Downloading stream url info', fatal=False, query={
  58                         **api_info,
  59                         'type': 'json',
  60                         'optional_id': vod_id,
  61                         'active_flg': 1,
  62                     }),
  63                 ('meta', 0, 'movie_url', ('mb_auto', 'auto_sp', 'auto_pc'), {url_or_none}), get_all=False)
  64             if stream_url:
  65                 return self._extract_m3u8_formats_and_subtitles(stream_url, vod_id)
  66
  67         raise ExtractorError('Unable to extract stream url')
  68
  69     def _extract_episode_info(self, url, episode=None):
  70         fetch_episode = episode is None
  71         lang, m_type, episode_id = NhkVodIE._match_valid_url(url).group('lang', 'type', 'id')
  72         is_video = m_type == 'video'
  73
  74         if is_video:
  75             episode_id = episode_id[:4] + '-' + episode_id[4:]
  76
  77         if fetch_episode:
  78             episode = self._call_api(
  79                 episode_id, lang, is_video, True, episode_id[:4] == '9999')[0]
  80         title = episode.get('sub_title_clean') or episode['sub_title']
  81
  82         def get_clean_field(key):
  83             return episode.get(key + '_clean') or episode.get(key)
  84
  85         series = get_clean_field('title')
  86
  87         thumbnails = []
  88         for s, w, h in [('', 640, 360), ('_l', 1280, 720)]:
  89             img_path = episode.get('image' + s)
  90             if not img_path:
  91                 continue
  92             thumbnails.append({
  93                 'id': '%dp' % h,
  94                 'height': h,
  95                 'width': w,
  96                 'url': 'https://www3.nhk.or.jp' + img_path,
  97             })
  98
  99         info = {
 100             'id': episode_id + '-' + lang,
 101             'title': '%s - %s' % (series, title) if series and title else title,
 102             'description': get_clean_field('description'),
 103             'thumbnails': thumbnails,
 104             'series': series,
 105             'episode': title,
 106         }
 107         if is_video:
 108             vod_id = episode['vod_id']
 109             formats, subs = self._extract_formats_and_subtitles(vod_id)
 110
 111             info.update({
 112                 'id': vod_id,
 113                 'formats': formats,
 114                 'subtitles': subs,
 115             })
 116
 117         else:
 118             if fetch_episode:
 119                 audio_path = episode['audio']['audio']
 120                 info['formats'] = self._extract_m3u8_formats(
 121                     'https://nhkworld-vh.akamaihd.net/i%s/master.m3u8' % audio_path,
 122                     episode_id, 'm4a', entry_protocol='m3u8_native',
 123                     m3u8_id='hls', fatal=False)
 124                 for f in info['formats']:
 125                     f['language'] = lang
 126             else:
 127                 info.update({
 128                     '_type': 'url_transparent',
 129                     'ie_key': NhkVodIE.ie_key(),
 130                     'url': url,
 131                 })
 132         return info
 133
 134
 135 class NhkVodIE(NhkBaseIE):
 136     # the 7-character IDs can have alphabetic chars too: assume [a-z] rather than just [a-f], eg
 137     _VALID_URL = [rf'{NhkBaseIE._BASE_URL_REGEX}/(?P<type>video)/(?P<id>[0-9a-z]+)',
 138                   rf'{NhkBaseIE._BASE_URL_REGEX}/(?P<type>audio)/(?P<id>[^/?#]+?-\d{{8}}-[0-9a-z]+)']
 139     # Content available only for a limited period of time. Visit
 140     # https://www3.nhk.or.jp/nhkworld/en/ondemand/ for working samples.
 141     _TESTS = [{
 142         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2049126/',
 143         'info_dict': {
 144             'id': 'nw_vod_v_en_2049_126_20230413233000_01_1681398302',
 145             'ext': 'mp4',
 146             'title': 'Japan Railway Journal - The Tohoku Shinkansen: Full Speed Ahead',
 147             'description': 'md5:49f7c5b206e03868a2fdf0d0814b92f6',
 148             'thumbnail': 'md5:51bcef4a21936e7fea1ff4e06353f463',
 149             'episode': 'The Tohoku Shinkansen: Full Speed Ahead',
 150             'series': 'Japan Railway Journal',
 151         },
 152     }, {
 153         # video clip
 154         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999011/',
 155         'md5': '153c3016dfd252ba09726588149cf0e7',
 156         'info_dict': {
 157             'id': 'lpZXIwaDE6_Z-976CPsFdxyICyWUzlT5',
 158             'ext': 'mp4',
 159             'title': 'Dining with the Chef - Chef Saito\'s Family recipe: MENCHI-KATSU',
 160             'description': 'md5:5aee4a9f9d81c26281862382103b0ea5',
 161             'thumbnail': 'md5:d6a4d9b6e9be90aaadda0bcce89631ed',
 162             'series': 'Dining with the Chef',
 163             'episode': 'Chef Saito\'s Family recipe: MENCHI-KATSU',
 164         },
 165     }, {
 166         # radio
 167         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/livinginjapan-20231001-1/',
 168         'info_dict': {
 169             'id': 'livinginjapan-20231001-1-en',
 170             'ext': 'm4a',
 171             'title': 'Living in Japan - Tips for Travelers to Japan / Ramen Vending Machines',
 172             'series': 'Living in Japan',
 173             'description': 'md5:850611969932874b4a3309e0cae06c2f',
 174             'thumbnail': 'md5:960622fb6e06054a4a1a0c97ea752545',
 175             'episode': 'Tips for Travelers to Japan / Ramen Vending Machines'
 176         },
 177     }, {
 178         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2015173/',
 179         'only_matching': True,
 180     }, {
 181         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/plugin-20190404-1/',
 182         'only_matching': True,
 183     }, {
 184         'url': 'https://www3.nhk.or.jp/nhkworld/fr/ondemand/audio/plugin-20190404-1/',
 185         'only_matching': True,
 186     }, {
 187         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/j_art-20150903-1/',
 188         'only_matching': True,
 189     }, {
 190         # video, alphabetic character in ID #29670
 191         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999a34/',
 192         'info_dict': {
 193             'id': 'qfjay6cg',
 194             'ext': 'mp4',
 195             'title': 'DESIGN TALKS plus - Fishermen’s Finery',
 196             'description': 'md5:8a8f958aaafb0d7cb59d38de53f1e448',
 197             'thumbnail': r're:^https?:/(/[a-z0-9.-]+)+\.jpg\?w=1920&h=1080$',
 198             'upload_date': '20210615',
 199             'timestamp': 1623722008,
 200         },
 201         'skip': '404 Not Found',
 202     }, {
 203         # japanese-language, longer id than english
 204         'url': 'https://www3.nhk.or.jp/nhkworld/ja/ondemand/video/0020271111/',
 205         'info_dict': {
 206             'id': 'nw_ja_v_jvod_ohayou_20231008',
 207             'ext': 'mp4',
 208             'title': 'おはよう日本（7時台） - 10月8日放送',
 209             'series': 'おはよう日本（7時台）',
 210             'episode': '10月8日放送',
 211             'thumbnail': 'md5:d733b1c8e965ab68fb02b2d347d0e9b4',
 212             'description': 'md5:9c1d6cbeadb827b955b20e99ab920ff0',
 213         },
 214         'skip': 'expires 2023-10-15',
 215     }]
 216
 217     def _real_extract(self, url):
 218         return self._extract_episode_info(url)
 219
 220
 221 class NhkVodProgramIE(NhkBaseIE):
 222     _VALID_URL = rf'{NhkBaseIE._BASE_URL_REGEX}/program{NhkBaseIE._TYPE_REGEX}(?P<id>\w+)(?:.+?\btype=(?P<episode_type>clip|(?:radio|tv)Episode))?'
 223     _TESTS = [{
 224         # video program episodes
 225         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/sumo',
 226         'info_dict': {
 227             'id': 'sumo',
 228             'title': 'GRAND SUMO Highlights',
 229         },
 230         'playlist_mincount': 12,
 231     }, {
 232         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway',
 233         'info_dict': {
 234             'id': 'japanrailway',
 235             'title': 'Japan Railway Journal',
 236         },
 237         'playlist_mincount': 12,
 238     }, {
 239         # video program clips
 240         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway/?type=clip',
 241         'info_dict': {
 242             'id': 'japanrailway',
 243             'title': 'Japan Railway Journal',
 244         },
 245         'playlist_mincount': 5,
 246     }, {
 247         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/10yearshayaomiyazaki/',
 248         'only_matching': True,
 249     }, {
 250         # audio program
 251         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/audio/listener/',
 252         'only_matching': True,
 253     }]
 254
 255     def _real_extract(self, url):
 256         lang, m_type, program_id, episode_type = self._match_valid_url(url).group('lang', 'type', 'id', 'episode_type')
 257         episodes = self._call_api(
 258             program_id, lang, m_type == 'video', False, episode_type == 'clip')
 259
 260         entries = []
 261         for episode in episodes:
 262             episode_path = episode.get('url')
 263             if not episode_path:
 264                 continue
 265             entries.append(self._extract_episode_info(
 266                 urljoin(url, episode_path), episode))
 267
 268         program_title = None
 269         if entries:
 270             program_title = entries[0].get('series')
 271
 272         return self.playlist_result(entries, program_id, program_title)
 273
 274
 275 class NhkForSchoolBangumiIE(InfoExtractor):
 276     _VALID_URL = r'https?://www2\.nhk\.or\.jp/school/movie/(?P<type>bangumi|clip)\.cgi\?das_id=(?P<id>[a-zA-Z0-9_-]+)'
 277     _TESTS = [{
 278         'url': 'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id=D0005150191_00000',
 279         'info_dict': {
 280             'id': 'D0005150191_00003',
 281             'title': 'にている かな',
 282             'duration': 599.999,
 283             'timestamp': 1396414800,
 284
 285             'upload_date': '20140402',
 286             'ext': 'mp4',
 287
 288             'chapters': 'count:12'
 289         },
 290         'params': {
 291             # m3u8 download
 292             'skip_download': True,
 293         },
 294     }]
 295
 296     def _real_extract(self, url):
 297         program_type, video_id = self._match_valid_url(url).groups()
 298
 299         webpage = self._download_webpage(
 300             f'https://www2.nhk.or.jp/school/movie/{program_type}.cgi?das_id={video_id}', video_id)
 301
 302         # searches all variables
 303         base_values = {g.group(1): g.group(2) for g in re.finditer(r'var\s+([a-zA-Z_]+)\s*=\s*"([^"]+?)";', webpage)}
 304         # and programObj values too
 305         program_values = {g.group(1): g.group(3) for g in re.finditer(r'(?:program|clip)Obj\.([a-zA-Z_]+)\s*=\s*(["\'])([^"]+?)\2;', webpage)}
 306         # extract all chapters
 307         chapter_durations = [parse_duration(g.group(1)) for g in re.finditer(r'chapterTime\.push\(\'([0-9:]+?)\'\);', webpage)]
 308         chapter_titles = [' '.join([g.group(1) or '', unescapeHTML(g.group(2))]).strip() for g in re.finditer(r'<div class="cpTitle"><span>(scene\s*\d+)?</span>([^<]+?)</div>', webpage)]
 309
 310         # this is how player_core.js is actually doing (!)
 311         version = base_values.get('r_version') or program_values.get('version')
 312         if version:
 313             video_id = f'{video_id.split("_")[0]}_{version}'
 314
 315         formats = self._extract_m3u8_formats(
 316             f'https://nhks-vh.akamaihd.net/i/das/{video_id[0:8]}/{video_id}_V_000.f4v/master.m3u8',
 317             video_id, ext='mp4', m3u8_id='hls')
 318
 319         duration = parse_duration(base_values.get('r_duration'))
 320
 321         chapters = None
 322         if chapter_durations and chapter_titles and len(chapter_durations) == len(chapter_titles):
 323             start_time = chapter_durations
 324             end_time = chapter_durations[1:] + [duration]
 325             chapters = [{
 326                 'start_time': s,
 327                 'end_time': e,
 328                 'title': t,
 329             } for s, e, t in zip(start_time, end_time, chapter_titles)]
 330
 331         return {
 332             'id': video_id,
 333             'title': program_values.get('name'),
 334             'duration': parse_duration(base_values.get('r_duration')),
 335             'timestamp': unified_timestamp(base_values['r_upload']),
 336             'formats': formats,
 337             'chapters': chapters,
 338         }
 339
 340
 341 class NhkForSchoolSubjectIE(InfoExtractor):
 342     IE_DESC = 'Portal page for each school subjects, like Japanese (kokugo, 国語) or math (sansuu/suugaku or 算数・数学)'
 343     KNOWN_SUBJECTS = (
 344         'rika', 'syakai', 'kokugo',
 345         'sansuu', 'seikatsu', 'doutoku',
 346         'ongaku', 'taiiku', 'zukou',
 347         'gijutsu', 'katei', 'sougou',
 348         'eigo', 'tokkatsu',
 349         'tokushi', 'sonota',
 350     )
 351     _VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>%s)/?(?:[\?#].*)?$' % '|'.join(re.escape(s) for s in KNOWN_SUBJECTS)
 352
 353     _TESTS = [{
 354         'url': 'https://www.nhk.or.jp/school/sougou/',
 355         'info_dict': {
 356             'id': 'sougou',
 357             'title': '総合的な学習の時間',
 358         },
 359         'playlist_mincount': 16,
 360     }, {
 361         'url': 'https://www.nhk.or.jp/school/rika/',
 362         'info_dict': {
 363             'id': 'rika',
 364             'title': '理科',
 365         },
 366         'playlist_mincount': 15,
 367     }]
 368
 369     def _real_extract(self, url):
 370         subject_id = self._match_id(url)
 371         webpage = self._download_webpage(url, subject_id)
 372
 373         return self.playlist_from_matches(
 374             re.finditer(rf'href="((?:https?://www\.nhk\.or\.jp)?/school/{re.escape(subject_id)}/[^/]+/)"', webpage),
 375             subject_id,
 376             self._html_search_regex(r'(?s)<span\s+class="subjectName">\s*<img\s*[^<]+>\s*([^<]+?)</span>', webpage, 'title', fatal=False),
 377             lambda g: urljoin(url, g.group(1)))
 378
 379
 380 class NhkForSchoolProgramListIE(InfoExtractor):
 381     _VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>(?:%s)/[a-zA-Z0-9_-]+)' % (
 382         '|'.join(re.escape(s) for s in NhkForSchoolSubjectIE.KNOWN_SUBJECTS)
 383     )
 384     _TESTS = [{
 385         'url': 'https://www.nhk.or.jp/school/sougou/q/',
 386         'info_dict': {
 387             'id': 'sougou/q',
 388             'title': 'Ｑ～こどものための哲学',
 389         },
 390         'playlist_mincount': 20,
 391     }]
 392
 393     def _real_extract(self, url):
 394         program_id = self._match_id(url)
 395
 396         webpage = self._download_webpage(f'https://www.nhk.or.jp/school/{program_id}/', program_id)
 397
 398         title = (self._generic_title('', webpage)
 399                  or self._html_search_regex(r'<h3>([^<]+?)とは？\s*</h3>', webpage, 'title', fatal=False))
 400         title = re.sub(r'\s*\|\s*NHK\s+for\s+School\s*$', '', title) if title else None
 401         description = self._html_search_regex(
 402             r'(?s)<div\s+class="programDetail\s*">\s*<p>[^<]+</p>',
 403             webpage, 'description', fatal=False, group=0)
 404
 405         bangumi_list = self._download_json(
 406             f'https://www.nhk.or.jp/school/{program_id}/meta/program.json', program_id)
 407         # they're always bangumi
 408         bangumis = [
 409             self.url_result(f'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id={x}')
 410             for x in traverse_obj(bangumi_list, ('part', ..., 'part-video-dasid')) or []]
 411
 412         return self.playlist_result(bangumis, program_id, title, description)
 413
 414
 415 class NhkRadiruIE(InfoExtractor):
 416     _GEO_COUNTRIES = ['JP']
 417     IE_DESC = 'NHK らじる (Radiru/Rajiru)'
 418     _VALID_URL = r'https?://www\.nhk\.or\.jp/radio/(?:player/ondemand|ondemand/detail)\.html\?p=(?P<site>[\da-zA-Z]+)_(?P<corner>[\da-zA-Z]+)(?:_(?P<headline>[\da-zA-Z]+))?'
 419     _TESTS = [{
 420         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=0449_01_3853544',
 421         'skip': 'Episode expired on 2023-04-16',
 422         'info_dict': {
 423             'channel': 'NHK-FM',
 424             'description': 'md5:94b08bdeadde81a97df4ec882acce3e9',
 425             'ext': 'm4a',
 426             'id': '0449_01_3853544',
 427             'series': 'ジャズ・トゥナイト',
 428             'thumbnail': 'https://www.nhk.or.jp/prog/img/449/g449.jpg',
 429             'timestamp': 1680969600,
 430             'title': 'ジャズ・トゥナイト　ＮＥＷジャズ特集',
 431             'upload_date': '20230408',
 432             'release_timestamp': 1680962400,
 433             'release_date': '20230408',
 434             'was_live': True,
 435         },
 436     }, {
 437         # playlist, airs every weekday so it should _hopefully_ be okay forever
 438         'url': 'https://www.nhk.or.jp/radio/ondemand/detail.html?p=0458_01',
 439         'info_dict': {
 440             'id': '0458_01',
 441             'title': 'ベストオブクラシック',
 442             'description': '世界中の上質な演奏会をじっくり堪能する本格派クラシック番組。',
 443             'channel': 'NHK-FM',
 444             'thumbnail': 'https://www.nhk.or.jp/prog/img/458/g458.jpg',
 445         },
 446         'playlist_mincount': 3,
 447     }, {
 448         # one with letters in the id
 449         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F300_06_3738470',
 450         'note': 'Expires on 2024-03-31',
 451         'info_dict': {
 452             'id': 'F300_06_3738470',
 453             'ext': 'm4a',
 454             'title': '有島武郎「一房のぶどう」',
 455             'description': '朗読：川野一宇（ラジオ深夜便アンカー）\r\n\r\n（2016年12月8日放送「ラジオ深夜便『アンカー朗読シリーズ』」より）',
 456             'channel': 'NHKラジオ第1、NHK-FM',
 457             'timestamp': 1635757200,
 458             'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F300/img/corner/box_109_thumbnail.jpg',
 459             'release_date': '20161207',
 460             'series': 'らじる文庫 by ラジオ深夜便 ',
 461             'release_timestamp': 1481126700,
 462             'upload_date': '20211101',
 463         }
 464     }, {
 465         # news
 466         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F261_01_3855109',
 467         'skip': 'Expires on 2023-04-17',
 468         'info_dict': {
 469             'id': 'F261_01_3855109',
 470             'ext': 'm4a',
 471             'channel': 'NHKラジオ第1',
 472             'timestamp': 1681635900,
 473             'release_date': '20230416',
 474             'series': 'NHKラジオニュース',
 475             'title': '午後６時のNHKニュース',
 476             'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
 477             'upload_date': '20230416',
 478             'release_timestamp': 1681635600,
 479         },
 480     }]
 481
 482     def _extract_episode_info(self, headline, programme_id, series_meta):
 483         episode_id = f'{programme_id}_{headline["headline_id"]}'
 484         episode = traverse_obj(headline, ('file_list', 0, {dict}))
 485
 486         return {
 487             **series_meta,
 488             'id': episode_id,
 489             'formats': self._extract_m3u8_formats(episode.get('file_name'), episode_id, fatal=False),
 490             'container': 'm4a_dash',  # force fixup, AAC-only HLS
 491             'was_live': True,
 492             'series': series_meta.get('title'),
 493             'thumbnail': url_or_none(headline.get('headline_image')) or series_meta.get('thumbnail'),
 494             **traverse_obj(episode, {
 495                 'title': 'file_title',
 496                 'description': 'file_title_sub',
 497                 'timestamp': ('open_time', {unified_timestamp}),
 498                 'release_timestamp': ('aa_vinfo4', {lambda x: x.split('_')[0]}, {unified_timestamp}),
 499             }),
 500         }
 501
 502     def _real_extract(self, url):
 503         site_id, corner_id, headline_id = self._match_valid_url(url).group('site', 'corner', 'headline')
 504         programme_id = f'{site_id}_{corner_id}'
 505
 506         if site_id == 'F261':
 507             json_url = 'https://www.nhk.or.jp/s-media/news/news-site/list/v1/all.json'
 508         else:
 509             json_url = f'https://www.nhk.or.jp/radioondemand/json/{site_id}/bangumi_{programme_id}.json'
 510
 511         meta = self._download_json(json_url, programme_id)['main']
 512
 513         series_meta = traverse_obj(meta, {
 514             'title': 'program_name',
 515             'channel': 'media_name',
 516             'thumbnail': (('thumbnail_c', 'thumbnail_p'), {url_or_none}),
 517         }, get_all=False)
 518
 519         if headline_id:
 520             return self._extract_episode_info(
 521                 traverse_obj(meta, (
 522                     'detail_list', lambda _, v: v['headline_id'] == headline_id), get_all=False),
 523                 programme_id, series_meta)
 524
 525         def entries():
 526             for headline in traverse_obj(meta, ('detail_list', ..., {dict})):
 527                 yield self._extract_episode_info(headline, programme_id, series_meta)
 528
 529         return self.playlist_result(
 530             entries(), programme_id, playlist_description=meta.get('site_detail'), **series_meta)
 531
 532
 533 class NhkRadioNewsPageIE(InfoExtractor):
 534     _VALID_URL = r'https?://www\.nhk\.or\.jp/radionews/?(?:$|[?#])'
 535     _TESTS = [{
 536         # airs daily, on-the-hour most hours
 537         'url': 'https://www.nhk.or.jp/radionews/',
 538         'playlist_mincount': 5,
 539         'info_dict': {
 540             'id': 'F261_01',
 541             'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
 542             'description': 'md5:bf2c5b397e44bc7eb26de98d8f15d79d',
 543             'channel': 'NHKラジオ第1',
 544             'title': 'NHKラジオニュース',
 545         }
 546     }]
 547
 548     def _real_extract(self, url):
 549         return self.url_result('https://www.nhk.or.jp/radio/ondemand/detail.html?p=F261_01', NhkRadiruIE)
 550
 551
 552 class NhkRadiruLiveIE(InfoExtractor):
 553     _GEO_COUNTRIES = ['JP']
 554     _VALID_URL = r'https?://www\.nhk\.or\.jp/radio/player/\?ch=(?P<id>r[12]|fm)'
 555     _TESTS = [{
 556         # radio 1, no area specified
 557         'url': 'https://www.nhk.or.jp/radio/player/?ch=r1',
 558         'info_dict': {
 559             'id': 'r1-tokyo',
 560             'title': 're:^ＮＨＫネットラジオ第1 東京.+$',
 561             'ext': 'm4a',
 562             'thumbnail': 'https://www.nhk.or.jp/common/img/media/r1-200x200.png',
 563             'live_status': 'is_live',
 564         },
 565     }, {
 566         # radio 2, area specified
 567         # (the area doesnt actually matter, r2 is national)
 568         'url': 'https://www.nhk.or.jp/radio/player/?ch=r2',
 569         'params': {'extractor_args': {'nhkradirulive': {'area': ['fukuoka']}}},
 570         'info_dict': {
 571             'id': 'r2-fukuoka',
 572             'title': 're:^ＮＨＫネットラジオ第2 福岡.+$',
 573             'ext': 'm4a',
 574             'thumbnail': 'https://www.nhk.or.jp/common/img/media/r2-200x200.png',
 575             'live_status': 'is_live',
 576         },
 577     }, {
 578         # fm, area specified
 579         'url': 'https://www.nhk.or.jp/radio/player/?ch=fm',
 580         'params': {'extractor_args': {'nhkradirulive': {'area': ['sapporo']}}},
 581         'info_dict': {
 582             'id': 'fm-sapporo',
 583             'title': 're:^ＮＨＫネットラジオＦＭ 札幌.+$',
 584             'ext': 'm4a',
 585             'thumbnail': 'https://www.nhk.or.jp/common/img/media/fm-200x200.png',
 586             'live_status': 'is_live',
 587         }
 588     }]
 589
 590     _NOA_STATION_IDS = {'r1': 'n1', 'r2': 'n2', 'fm': 'n3'}
 591
 592     def _real_extract(self, url):
 593         station = self._match_id(url)
 594         area = self._configuration_arg('area', ['tokyo'])[0]
 595
 596         config = self._download_xml(
 597             'https://www.nhk.or.jp/radio/config/config_web.xml', station, 'Downloading area information')
 598         data = config.find(f'.//data//area[.="{area}"]/..')
 599
 600         if not data:
 601             raise ExtractorError('Invalid area. Valid areas are: %s' % ', '.join(
 602                 [i.text for i in config.findall('.//data//area')]), expected=True)
 603
 604         noa_info = self._download_json(
 605             f'https:{config.find(".//url_program_noa").text}'.format(area=data.find('areakey').text),
 606             station, note=f'Downloading {area} station metadata')
 607         present_info = traverse_obj(noa_info, ('nowonair_list', self._NOA_STATION_IDS.get(station), 'present'))
 608
 609         return {
 610             'title': ' '.join(traverse_obj(present_info, (('service', 'area',), 'name', {str}))),
 611             'id': join_nonempty(station, area),
 612             'thumbnails': traverse_obj(present_info, ('service', 'images', ..., {
 613                 'url': 'url',
 614                 'width': ('width', {int_or_none}),
 615                 'height': ('height', {int_or_none}),
 616             })),
 617             'formats': self._extract_m3u8_formats(data.find(f'{station}hls').text, station),
 618             'is_live': True,
 619         }