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).groups()
  72         if len(episode_id) == 7:
  73             episode_id = episode_id[:4] + '-' + episode_id[4:]
  74
  75         is_video = m_type == 'video'
  76         if fetch_episode:
  77             episode = self._call_api(
  78                 episode_id, lang, is_video, True, episode_id[:4] == '9999')[0]
  79         title = episode.get('sub_title_clean') or episode['sub_title']
  80
  81         def get_clean_field(key):
  82             return episode.get(key + '_clean') or episode.get(key)
  83
  84         series = get_clean_field('title')
  85
  86         thumbnails = []
  87         for s, w, h in [('', 640, 360), ('_l', 1280, 720)]:
  88             img_path = episode.get('image' + s)
  89             if not img_path:
  90                 continue
  91             thumbnails.append({
  92                 'id': '%dp' % h,
  93                 'height': h,
  94                 'width': w,
  95                 'url': 'https://www3.nhk.or.jp' + img_path,
  96             })
  97
  98         info = {
  99             'id': episode_id + '-' + lang,
 100             'title': '%s - %s' % (series, title) if series and title else title,
 101             'description': get_clean_field('description'),
 102             'thumbnails': thumbnails,
 103             'series': series,
 104             'episode': title,
 105         }
 106         if is_video:
 107             vod_id = episode['vod_id']
 108             formats, subs = self._extract_formats_and_subtitles(vod_id)
 109
 110             info.update({
 111                 'id': vod_id,
 112                 'formats': formats,
 113                 'subtitles': subs,
 114             })
 115
 116         else:
 117             if fetch_episode:
 118                 audio_path = episode['audio']['audio']
 119                 info['formats'] = self._extract_m3u8_formats(
 120                     'https://nhkworld-vh.akamaihd.net/i%s/master.m3u8' % audio_path,
 121                     episode_id, 'm4a', entry_protocol='m3u8_native',
 122                     m3u8_id='hls', fatal=False)
 123                 for f in info['formats']:
 124                     f['language'] = lang
 125             else:
 126                 info.update({
 127                     '_type': 'url_transparent',
 128                     'ie_key': NhkVodIE.ie_key(),
 129                     'url': url,
 130                 })
 131         return info
 132
 133
 134 class NhkVodIE(NhkBaseIE):
 135     # the 7-character IDs can have alphabetic chars too: assume [a-z] rather than just [a-f], eg
 136     _VALID_URL = r'%s%s(?P<id>[0-9a-z]{7}|[^/]+?-\d{8}-[0-9a-z]+)' % (NhkBaseIE._BASE_URL_REGEX, NhkBaseIE._TYPE_REGEX)
 137     # Content available only for a limited period of time. Visit
 138     # https://www3.nhk.or.jp/nhkworld/en/ondemand/ for working samples.
 139     _TESTS = [{
 140         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2061601/',
 141         'info_dict': {
 142             'id': 'yd8322ch',
 143             'ext': 'mp4',
 144             'description': 'md5:109c8b05d67a62d0592f2b445d2cd898',
 145             'title': 'GRAND SUMO Highlights - [Recap] May Tournament Day 1 (Opening Day)',
 146             'upload_date': '20230514',
 147             'timestamp': 1684083791,
 148             'series': 'GRAND SUMO Highlights',
 149             'episode': '[Recap] May Tournament Day 1 (Opening Day)',
 150             'thumbnail': 'https://mz-edge.stream.co.jp/thumbs/aid/t1684084443/4028649.jpg?w=1920&h=1080',
 151         },
 152     }, {
 153         # video clip
 154         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999011/',
 155         'md5': '7a90abcfe610ec22a6bfe15bd46b30ca',
 156         'info_dict': {
 157             'id': 'a95j5iza',
 158             'ext': 'mp4',
 159             'title': "Dining with the Chef - Chef Saito's Family recipe: MENCHI-KATSU",
 160             'description': 'md5:5aee4a9f9d81c26281862382103b0ea5',
 161             'timestamp': 1565965194,
 162             'upload_date': '20190816',
 163             'thumbnail': 'https://mz-edge.stream.co.jp/thumbs/aid/t1567086278/3715195.jpg?w=1920&h=1080',
 164             'series': 'Dining with the Chef',
 165             'episode': 'Chef Saito\'s Family recipe: MENCHI-KATSU',
 166         },
 167     }, {
 168         # audio clip
 169         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/r_inventions-20201104-1/',
 170         'info_dict': {
 171             'id': 'r_inventions-20201104-1-en',
 172             'ext': 'm4a',
 173             'title': "Japan's Top Inventions - Miniature Video Cameras",
 174             'description': 'md5:07ea722bdbbb4936fdd360b6a480c25b',
 175         },
 176         'skip': '404 Not Found',
 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
 204     def _real_extract(self, url):
 205         return self._extract_episode_info(url)
 206
 207
 208 class NhkVodProgramIE(NhkBaseIE):
 209     _VALID_URL = r'%s/program%s(?P<id>[0-9a-z]+)(?:.+?\btype=(?P<episode_type>clip|(?:radio|tv)Episode))?' % (NhkBaseIE._BASE_URL_REGEX, NhkBaseIE._TYPE_REGEX)
 210     _TESTS = [{
 211         # video program episodes
 212         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/sumo',
 213         'info_dict': {
 214             'id': 'sumo',
 215             'title': 'GRAND SUMO Highlights',
 216         },
 217         'playlist_mincount': 12,
 218     }, {
 219         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway',
 220         'info_dict': {
 221             'id': 'japanrailway',
 222             'title': 'Japan Railway Journal',
 223         },
 224         'playlist_mincount': 12,
 225     }, {
 226         # video program clips
 227         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway/?type=clip',
 228         'info_dict': {
 229             'id': 'japanrailway',
 230             'title': 'Japan Railway Journal',
 231         },
 232         'playlist_mincount': 5,
 233     }, {
 234         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/10yearshayaomiyazaki/',
 235         'only_matching': True,
 236     }, {
 237         # audio program
 238         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/audio/listener/',
 239         'only_matching': True,
 240     }]
 241
 242     def _real_extract(self, url):
 243         lang, m_type, program_id, episode_type = self._match_valid_url(url).groups()
 244
 245         episodes = self._call_api(
 246             program_id, lang, m_type == 'video', False, episode_type == 'clip')
 247
 248         entries = []
 249         for episode in episodes:
 250             episode_path = episode.get('url')
 251             if not episode_path:
 252                 continue
 253             entries.append(self._extract_episode_info(
 254                 urljoin(url, episode_path), episode))
 255
 256         program_title = None
 257         if entries:
 258             program_title = entries[0].get('series')
 259
 260         return self.playlist_result(entries, program_id, program_title)
 261
 262
 263 class NhkForSchoolBangumiIE(InfoExtractor):
 264     _VALID_URL = r'https?://www2\.nhk\.or\.jp/school/movie/(?P<type>bangumi|clip)\.cgi\?das_id=(?P<id>[a-zA-Z0-9_-]+)'
 265     _TESTS = [{
 266         'url': 'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id=D0005150191_00000',
 267         'info_dict': {
 268             'id': 'D0005150191_00003',
 269             'title': 'にている かな',
 270             'duration': 599.999,
 271             'timestamp': 1396414800,
 272
 273             'upload_date': '20140402',
 274             'ext': 'mp4',
 275
 276             'chapters': 'count:12'
 277         },
 278         'params': {
 279             # m3u8 download
 280             'skip_download': True,
 281         },
 282     }]
 283
 284     def _real_extract(self, url):
 285         program_type, video_id = self._match_valid_url(url).groups()
 286
 287         webpage = self._download_webpage(
 288             f'https://www2.nhk.or.jp/school/movie/{program_type}.cgi?das_id={video_id}', video_id)
 289
 290         # searches all variables
 291         base_values = {g.group(1): g.group(2) for g in re.finditer(r'var\s+([a-zA-Z_]+)\s*=\s*"([^"]+?)";', webpage)}
 292         # and programObj values too
 293         program_values = {g.group(1): g.group(3) for g in re.finditer(r'(?:program|clip)Obj\.([a-zA-Z_]+)\s*=\s*(["\'])([^"]+?)\2;', webpage)}
 294         # extract all chapters
 295         chapter_durations = [parse_duration(g.group(1)) for g in re.finditer(r'chapterTime\.push\(\'([0-9:]+?)\'\);', webpage)]
 296         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)]
 297
 298         # this is how player_core.js is actually doing (!)
 299         version = base_values.get('r_version') or program_values.get('version')
 300         if version:
 301             video_id = f'{video_id.split("_")[0]}_{version}'
 302
 303         formats = self._extract_m3u8_formats(
 304             f'https://nhks-vh.akamaihd.net/i/das/{video_id[0:8]}/{video_id}_V_000.f4v/master.m3u8',
 305             video_id, ext='mp4', m3u8_id='hls')
 306
 307         duration = parse_duration(base_values.get('r_duration'))
 308
 309         chapters = None
 310         if chapter_durations and chapter_titles and len(chapter_durations) == len(chapter_titles):
 311             start_time = chapter_durations
 312             end_time = chapter_durations[1:] + [duration]
 313             chapters = [{
 314                 'start_time': s,
 315                 'end_time': e,
 316                 'title': t,
 317             } for s, e, t in zip(start_time, end_time, chapter_titles)]
 318
 319         return {
 320             'id': video_id,
 321             'title': program_values.get('name'),
 322             'duration': parse_duration(base_values.get('r_duration')),
 323             'timestamp': unified_timestamp(base_values['r_upload']),
 324             'formats': formats,
 325             'chapters': chapters,
 326         }
 327
 328
 329 class NhkForSchoolSubjectIE(InfoExtractor):
 330     IE_DESC = 'Portal page for each school subjects, like Japanese (kokugo, 国語) or math (sansuu/suugaku or 算数・数学)'
 331     KNOWN_SUBJECTS = (
 332         'rika', 'syakai', 'kokugo',
 333         'sansuu', 'seikatsu', 'doutoku',
 334         'ongaku', 'taiiku', 'zukou',
 335         'gijutsu', 'katei', 'sougou',
 336         'eigo', 'tokkatsu',
 337         'tokushi', 'sonota',
 338     )
 339     _VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>%s)/?(?:[\?#].*)?$' % '|'.join(re.escape(s) for s in KNOWN_SUBJECTS)
 340
 341     _TESTS = [{
 342         'url': 'https://www.nhk.or.jp/school/sougou/',
 343         'info_dict': {
 344             'id': 'sougou',
 345             'title': '総合的な学習の時間',
 346         },
 347         'playlist_mincount': 16,
 348     }, {
 349         'url': 'https://www.nhk.or.jp/school/rika/',
 350         'info_dict': {
 351             'id': 'rika',
 352             'title': '理科',
 353         },
 354         'playlist_mincount': 15,
 355     }]
 356
 357     def _real_extract(self, url):
 358         subject_id = self._match_id(url)
 359         webpage = self._download_webpage(url, subject_id)
 360
 361         return self.playlist_from_matches(
 362             re.finditer(rf'href="((?:https?://www\.nhk\.or\.jp)?/school/{re.escape(subject_id)}/[^/]+/)"', webpage),
 363             subject_id,
 364             self._html_search_regex(r'(?s)<span\s+class="subjectName">\s*<img\s*[^<]+>\s*([^<]+?)</span>', webpage, 'title', fatal=False),
 365             lambda g: urljoin(url, g.group(1)))
 366
 367
 368 class NhkForSchoolProgramListIE(InfoExtractor):
 369     _VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>(?:%s)/[a-zA-Z0-9_-]+)' % (
 370         '|'.join(re.escape(s) for s in NhkForSchoolSubjectIE.KNOWN_SUBJECTS)
 371     )
 372     _TESTS = [{
 373         'url': 'https://www.nhk.or.jp/school/sougou/q/',
 374         'info_dict': {
 375             'id': 'sougou/q',
 376             'title': 'Ｑ～こどものための哲学',
 377         },
 378         'playlist_mincount': 20,
 379     }]
 380
 381     def _real_extract(self, url):
 382         program_id = self._match_id(url)
 383
 384         webpage = self._download_webpage(f'https://www.nhk.or.jp/school/{program_id}/', program_id)
 385
 386         title = (self._generic_title('', webpage)
 387                  or self._html_search_regex(r'<h3>([^<]+?)とは？\s*</h3>', webpage, 'title', fatal=False))
 388         title = re.sub(r'\s*\|\s*NHK\s+for\s+School\s*$', '', title) if title else None
 389         description = self._html_search_regex(
 390             r'(?s)<div\s+class="programDetail\s*">\s*<p>[^<]+</p>',
 391             webpage, 'description', fatal=False, group=0)
 392
 393         bangumi_list = self._download_json(
 394             f'https://www.nhk.or.jp/school/{program_id}/meta/program.json', program_id)
 395         # they're always bangumi
 396         bangumis = [
 397             self.url_result(f'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id={x}')
 398             for x in traverse_obj(bangumi_list, ('part', ..., 'part-video-dasid')) or []]
 399
 400         return self.playlist_result(bangumis, program_id, title, description)
 401
 402
 403 class NhkRadiruIE(InfoExtractor):
 404     _GEO_COUNTRIES = ['JP']
 405     IE_DESC = 'NHK らじる (Radiru/Rajiru)'
 406     _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]+))?'
 407     _TESTS = [{
 408         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=0449_01_3853544',
 409         'skip': 'Episode expired on 2023-04-16',
 410         'info_dict': {
 411             'channel': 'NHK-FM',
 412             'description': 'md5:94b08bdeadde81a97df4ec882acce3e9',
 413             'ext': 'm4a',
 414             'id': '0449_01_3853544',
 415             'series': 'ジャズ・トゥナイト',
 416             'thumbnail': 'https://www.nhk.or.jp/prog/img/449/g449.jpg',
 417             'timestamp': 1680969600,
 418             'title': 'ジャズ・トゥナイト　ＮＥＷジャズ特集',
 419             'upload_date': '20230408',
 420             'release_timestamp': 1680962400,
 421             'release_date': '20230408',
 422             'was_live': True,
 423         },
 424     }, {
 425         # playlist, airs every weekday so it should _hopefully_ be okay forever
 426         'url': 'https://www.nhk.or.jp/radio/ondemand/detail.html?p=0458_01',
 427         'info_dict': {
 428             'id': '0458_01',
 429             'title': 'ベストオブクラシック',
 430             '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。',
 431             'channel': 'NHK-FM',
 432             'thumbnail': 'https://www.nhk.or.jp/prog/img/458/g458.jpg',
 433         },
 434         'playlist_mincount': 3,
 435     }, {
 436         # one with letters in the id
 437         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F300_06_3738470',
 438         'note': 'Expires on 2024-03-31',
 439         'info_dict': {
 440             'id': 'F300_06_3738470',
 441             'ext': 'm4a',
 442             'title': '有島武郎「一房のぶどう」',
 443             '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）',
 444             'channel': 'NHKラジオ第1、NHK-FM',
 445             'timestamp': 1635757200,
 446             'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F300/img/corner/box_109_thumbnail.jpg',
 447             'release_date': '20161207',
 448             'series': 'らじる文庫 by ラジオ深夜便 ',
 449             'release_timestamp': 1481126700,
 450             'upload_date': '20211101',
 451         }
 452     }, {
 453         # news
 454         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F261_01_3855109',
 455         'skip': 'Expires on 2023-04-17',
 456         'info_dict': {
 457             'id': 'F261_01_3855109',
 458             'ext': 'm4a',
 459             'channel': 'NHKラジオ第1',
 460             'timestamp': 1681635900,
 461             'release_date': '20230416',
 462             'series': 'NHKラジオニュース',
 463             'title': '午後６時のNHKニュース',
 464             'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
 465             'upload_date': '20230416',
 466             'release_timestamp': 1681635600,
 467         },
 468     }]
 469
 470     def _extract_episode_info(self, headline, programme_id, series_meta):
 471         episode_id = f'{programme_id}_{headline["headline_id"]}'
 472         episode = traverse_obj(headline, ('file_list', 0, {dict}))
 473
 474         return {
 475             **series_meta,
 476             'id': episode_id,
 477             'formats': self._extract_m3u8_formats(episode.get('file_name'), episode_id, fatal=False),
 478             'container': 'm4a_dash',  # force fixup, AAC-only HLS
 479             'was_live': True,
 480             'series': series_meta.get('title'),
 481             'thumbnail': url_or_none(headline.get('headline_image')) or series_meta.get('thumbnail'),
 482             **traverse_obj(episode, {
 483                 'title': 'file_title',
 484                 'description': 'file_title_sub',
 485                 'timestamp': ('open_time', {unified_timestamp}),
 486                 'release_timestamp': ('aa_vinfo4', {lambda x: x.split('_')[0]}, {unified_timestamp}),
 487             }),
 488         }
 489
 490     def _real_extract(self, url):
 491         site_id, corner_id, headline_id = self._match_valid_url(url).group('site', 'corner', 'headline')
 492         programme_id = f'{site_id}_{corner_id}'
 493
 494         if site_id == 'F261':
 495             json_url = 'https://www.nhk.or.jp/s-media/news/news-site/list/v1/all.json'
 496         else:
 497             json_url = f'https://www.nhk.or.jp/radioondemand/json/{site_id}/bangumi_{programme_id}.json'
 498
 499         meta = self._download_json(json_url, programme_id)['main']
 500
 501         series_meta = traverse_obj(meta, {
 502             'title': 'program_name',
 503             'channel': 'media_name',
 504             'thumbnail': (('thumbnail_c', 'thumbnail_p'), {url_or_none}),
 505         }, get_all=False)
 506
 507         if headline_id:
 508             return self._extract_episode_info(
 509                 traverse_obj(meta, (
 510                     'detail_list', lambda _, v: v['headline_id'] == headline_id), get_all=False),
 511                 programme_id, series_meta)
 512
 513         def entries():
 514             for headline in traverse_obj(meta, ('detail_list', ..., {dict})):
 515                 yield self._extract_episode_info(headline, programme_id, series_meta)
 516
 517         return self.playlist_result(
 518             entries(), programme_id, playlist_description=meta.get('site_detail'), **series_meta)
 519
 520
 521 class NhkRadioNewsPageIE(InfoExtractor):
 522     _VALID_URL = r'https?://www\.nhk\.or\.jp/radionews/?(?:$|[?#])'
 523     _TESTS = [{
 524         # airs daily, on-the-hour most hours
 525         'url': 'https://www.nhk.or.jp/radionews/',
 526         'playlist_mincount': 5,
 527         'info_dict': {
 528             'id': 'F261_01',
 529             'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
 530             'description': 'md5:bf2c5b397e44bc7eb26de98d8f15d79d',
 531             'channel': 'NHKラジオ第1',
 532             'title': 'NHKラジオニュース',
 533         }
 534     }]
 535
 536     def _real_extract(self, url):
 537         return self.url_result('https://www.nhk.or.jp/radio/ondemand/detail.html?p=F261_01', NhkRadiruIE)
 538
 539
 540 class NhkRadiruLiveIE(InfoExtractor):
 541     _GEO_COUNTRIES = ['JP']
 542     _VALID_URL = r'https?://www\.nhk\.or\.jp/radio/player/\?ch=(?P<id>r[12]|fm)'
 543     _TESTS = [{
 544         # radio 1, no area specified
 545         'url': 'https://www.nhk.or.jp/radio/player/?ch=r1',
 546         'info_dict': {
 547             'id': 'r1-tokyo',
 548             'title': 're:^ＮＨＫネットラジオ第1 東京.+$',
 549             'ext': 'm4a',
 550             'thumbnail': 'https://www.nhk.or.jp/common/img/media/r1-200x200.png',
 551             'live_status': 'is_live',
 552         },
 553     }, {
 554         # radio 2, area specified
 555         # (the area doesnt actually matter, r2 is national)
 556         'url': 'https://www.nhk.or.jp/radio/player/?ch=r2',
 557         'params': {'extractor_args': {'nhkradirulive': {'area': ['fukuoka']}}},
 558         'info_dict': {
 559             'id': 'r2-fukuoka',
 560             'title': 're:^ＮＨＫネットラジオ第2 福岡.+$',
 561             'ext': 'm4a',
 562             'thumbnail': 'https://www.nhk.or.jp/common/img/media/r2-200x200.png',
 563             'live_status': 'is_live',
 564         },
 565     }, {
 566         # fm, area specified
 567         'url': 'https://www.nhk.or.jp/radio/player/?ch=fm',
 568         'params': {'extractor_args': {'nhkradirulive': {'area': ['sapporo']}}},
 569         'info_dict': {
 570             'id': 'fm-sapporo',
 571             'title': 're:^ＮＨＫネットラジオＦＭ 札幌.+$',
 572             'ext': 'm4a',
 573             'thumbnail': 'https://www.nhk.or.jp/common/img/media/fm-200x200.png',
 574             'live_status': 'is_live',
 575         }
 576     }]
 577
 578     _NOA_STATION_IDS = {'r1': 'n1', 'r2': 'n2', 'fm': 'n3'}
 579
 580     def _real_extract(self, url):
 581         station = self._match_id(url)
 582         area = self._configuration_arg('area', ['tokyo'])[0]
 583
 584         config = self._download_xml(
 585             'https://www.nhk.or.jp/radio/config/config_web.xml', station, 'Downloading area information')
 586         data = config.find(f'.//data//area[.="{area}"]/..')
 587
 588         if not data:
 589             raise ExtractorError('Invalid area. Valid areas are: %s' % ', '.join(
 590                 [i.text for i in config.findall('.//data//area')]), expected=True)
 591
 592         noa_info = self._download_json(
 593             f'https:{config.find(".//url_program_noa").text}'.format(area=data.find('areakey').text),
 594             station, note=f'Downloading {area} station metadata')
 595         present_info = traverse_obj(noa_info, ('nowonair_list', self._NOA_STATION_IDS.get(station), 'present'))
 596
 597         return {
 598             'title': ' '.join(traverse_obj(present_info, (('service', 'area',), 'name', {str}))),
 599             'id': join_nonempty(station, area),
 600             'thumbnails': traverse_obj(present_info, ('service', 'images', ..., {
 601                 'url': 'url',
 602                 'width': ('width', {int_or_none}),
 603                 'height': ('height', {int_or_none}),
 604             })),
 605             'formats': self._extract_m3u8_formats(data.find(f'{station}hls').text, station),
 606             'is_live': True,
 607         }