yt_dlp/extractor/nhk.py

   1 import re
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     ExtractorError,
   6     clean_html,
   7     filter_dict,
   8     get_element_by_class,
   9     int_or_none,
  10     join_nonempty,
  11     parse_duration,
  12     remove_end,
  13     traverse_obj,
  14     try_call,
  15     unescapeHTML,
  16     unified_timestamp,
  17     url_or_none,
  18     urljoin,
  19 )
  20
  21
  22 class NhkBaseIE(InfoExtractor):
  23     _API_URL_TEMPLATE = 'https://nwapi.nhk.jp/nhkworld/%sod%slist/v7b/%s/%s/%s/all%s.json'
  24     _BASE_URL_REGEX = r'https?://www3\.nhk\.or\.jp/nhkworld/(?P<lang>[a-z]{2})/'
  25
  26     def _call_api(self, m_id, lang, is_video, is_episode, is_clip):
  27         return self._download_json(
  28             self._API_URL_TEMPLATE % (
  29                 'v' if is_video else 'r',
  30                 'clip' if is_clip else 'esd',
  31                 'episode' if is_episode else 'program',
  32                 m_id, lang, '/all' if is_video else ''),
  33             m_id, query={'apikey': 'EJfK8jdS57GqlupFgAfAAwr573q01y6k'})['data']['episodes'] or []
  34
  35     def _get_api_info(self, refresh=True):
  36         if not refresh:
  37             return self.cache.load('nhk', 'api_info')
  38
  39         self.cache.store('nhk', 'api_info', {})
  40         movie_player_js = self._download_webpage(
  41             'https://movie-a.nhk.or.jp/world/player/js/movie-player.js', None,
  42             note='Downloading stream API information')
  43         api_info = {
  44             'url': self._search_regex(
  45                 r'prod:[^;]+\bapiUrl:\s*[\'"]([^\'"]+)[\'"]', movie_player_js, None, 'stream API url'),
  46             'token': self._search_regex(
  47                 r'prod:[^;]+\btoken:\s*[\'"]([^\'"]+)[\'"]', movie_player_js, None, 'stream API token'),
  48         }
  49         self.cache.store('nhk', 'api_info', api_info)
  50         return api_info
  51
  52     def _extract_stream_info(self, vod_id):
  53         for refresh in (False, True):
  54             api_info = self._get_api_info(refresh)
  55             if not api_info:
  56                 continue
  57
  58             api_url = api_info.pop('url')
  59             meta = traverse_obj(
  60                 self._download_json(
  61                     api_url, vod_id, 'Downloading stream url info', fatal=False, query={
  62                         **api_info,
  63                         'type': 'json',
  64                         'optional_id': vod_id,
  65                         'active_flg': 1,
  66                     }), ('meta', 0))
  67             stream_url = traverse_obj(
  68                 meta, ('movie_url', ('mb_auto', 'auto_sp', 'auto_pc'), {url_or_none}), get_all=False)
  69
  70             if stream_url:
  71                 formats, subtitles = self._extract_m3u8_formats_and_subtitles(stream_url, vod_id)
  72                 return {
  73                     **traverse_obj(meta, {
  74                         'duration': ('duration', {int_or_none}),
  75                         'timestamp': ('publication_date', {unified_timestamp}),
  76                         'release_timestamp': ('insert_date', {unified_timestamp}),
  77                         'modified_timestamp': ('update_date', {unified_timestamp}),
  78                     }),
  79                     'formats': formats,
  80                     'subtitles': subtitles,
  81                 }
  82         raise ExtractorError('Unable to extract stream url')
  83
  84     def _extract_episode_info(self, url, episode=None):
  85         fetch_episode = episode is None
  86         lang, m_type, episode_id = NhkVodIE._match_valid_url(url).group('lang', 'type', 'id')
  87         is_video = m_type != 'audio'
  88
  89         if is_video:
  90             episode_id = episode_id[:4] + '-' + episode_id[4:]
  91
  92         if fetch_episode:
  93             episode = self._call_api(
  94                 episode_id, lang, is_video, True, episode_id[:4] == '9999')[0]
  95
  96         def get_clean_field(key):
  97             return clean_html(episode.get(key + '_clean') or episode.get(key))
  98
  99         title = get_clean_field('sub_title')
 100         series = get_clean_field('title')
 101
 102         thumbnails = []
 103         for s, w, h in [('', 640, 360), ('_l', 1280, 720)]:
 104             img_path = episode.get('image' + s)
 105             if not img_path:
 106                 continue
 107             thumbnails.append({
 108                 'id': f'{h}p',
 109                 'height': h,
 110                 'width': w,
 111                 'url': 'https://www3.nhk.or.jp' + img_path,
 112             })
 113
 114         episode_name = title
 115         if series and title:
 116             title = f'{series} - {title}'
 117         elif series and not title:
 118             title = series
 119             series = None
 120             episode_name = None
 121         else:  # title, no series
 122             episode_name = None
 123
 124         info = {
 125             'id': episode_id + '-' + lang,
 126             'title': title,
 127             'description': get_clean_field('description'),
 128             'thumbnails': thumbnails,
 129             'series': series,
 130             'episode': episode_name,
 131         }
 132
 133         if is_video:
 134             vod_id = episode['vod_id']
 135             info.update({
 136                 **self._extract_stream_info(vod_id),
 137                 'id': vod_id,
 138             })
 139
 140         else:
 141             if fetch_episode:
 142                 # From https://www3.nhk.or.jp/nhkworld/common/player/radio/inline/rod.html
 143                 audio_path = remove_end(episode['audio']['audio'], '.m4a')
 144                 info['formats'] = self._extract_m3u8_formats(
 145                     f'{urljoin("https://vod-stream.nhk.jp", audio_path)}/index.m3u8',
 146                     episode_id, 'm4a', entry_protocol='m3u8_native',
 147                     m3u8_id='hls', fatal=False)
 148                 for f in info['formats']:
 149                     f['language'] = lang
 150             else:
 151                 info.update({
 152                     '_type': 'url_transparent',
 153                     'ie_key': NhkVodIE.ie_key(),
 154                     'url': url,
 155                 })
 156         return info
 157
 158
 159 class NhkVodIE(NhkBaseIE):
 160     _VALID_URL = [
 161         rf'{NhkBaseIE._BASE_URL_REGEX}shows/(?:(?P<type>video)/)?(?P<id>\d{{4}}[\da-z]\d+)/?(?:$|[?#])',
 162         rf'{NhkBaseIE._BASE_URL_REGEX}(?:ondemand|shows)/(?P<type>audio)/(?P<id>[^/?#]+?-\d{{8}}-[\da-z]+)',
 163         rf'{NhkBaseIE._BASE_URL_REGEX}ondemand/(?P<type>video)/(?P<id>\d{{4}}[\da-z]\d+)',  # deprecated
 164     ]
 165     # Content available only for a limited period of time. Visit
 166     # https://www3.nhk.or.jp/nhkworld/en/ondemand/ for working samples.
 167     _TESTS = [{
 168         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2049126/',
 169         'info_dict': {
 170             'id': 'nw_vod_v_en_2049_126_20230413233000_01_1681398302',
 171             'ext': 'mp4',
 172             'title': 'Japan Railway Journal - The Tohoku Shinkansen: Full Speed Ahead',
 173             'description': 'md5:49f7c5b206e03868a2fdf0d0814b92f6',
 174             'thumbnail': r're:https://.+/.+\.jpg',
 175             'episode': 'The Tohoku Shinkansen: Full Speed Ahead',
 176             'series': 'Japan Railway Journal',
 177             'modified_timestamp': 1707217907,
 178             'timestamp': 1681428600,
 179             'release_timestamp': 1693883728,
 180             'duration': 1679,
 181             'upload_date': '20230413',
 182             'modified_date': '20240206',
 183             'release_date': '20230905',
 184         },
 185     }, {
 186         # video clip
 187         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999011/',
 188         'md5': '153c3016dfd252ba09726588149cf0e7',
 189         'info_dict': {
 190             'id': 'lpZXIwaDE6_Z-976CPsFdxyICyWUzlT5',
 191             'ext': 'mp4',
 192             'title': 'Dining with the Chef - Chef Saito\'s Family recipe: MENCHI-KATSU',
 193             'description': 'md5:5aee4a9f9d81c26281862382103b0ea5',
 194             'thumbnail': r're:https://.+/.+\.jpg',
 195             'series': 'Dining with the Chef',
 196             'episode': 'Chef Saito\'s Family recipe: MENCHI-KATSU',
 197             'duration': 148,
 198             'upload_date': '20190816',
 199             'release_date': '20230902',
 200             'release_timestamp': 1693619292,
 201             'modified_timestamp': 1707217907,
 202             'modified_date': '20240206',
 203             'timestamp': 1565997540,
 204         },
 205     }, {
 206         # radio
 207         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/livinginjapan-20231001-1/',
 208         'info_dict': {
 209             'id': 'livinginjapan-20231001-1-en',
 210             'ext': 'm4a',
 211             'title': 'Living in Japan - Tips for Travelers to Japan / Ramen Vending Machines',
 212             'series': 'Living in Japan',
 213             'description': 'md5:0a0e2077d8f07a03071e990a6f51bfab',
 214             'thumbnail': r're:https://.+/.+\.jpg',
 215             'episode': 'Tips for Travelers to Japan / Ramen Vending Machines',
 216         },
 217     }, {
 218         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2015173/',
 219         'only_matching': True,
 220     }, {
 221         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/plugin-20190404-1/',
 222         'only_matching': True,
 223     }, {
 224         'url': 'https://www3.nhk.or.jp/nhkworld/fr/ondemand/audio/plugin-20190404-1/',
 225         'only_matching': True,
 226     }, {
 227         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/j_art-20150903-1/',
 228         'only_matching': True,
 229     }, {
 230         # video, alphabetic character in ID #29670
 231         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999a34/',
 232         'info_dict': {
 233             'id': 'qfjay6cg',
 234             'ext': 'mp4',
 235             'title': 'DESIGN TALKS plus - Fishermen’s Finery',
 236             'description': 'md5:8a8f958aaafb0d7cb59d38de53f1e448',
 237             'thumbnail': r're:^https?:/(/[a-z0-9.-]+)+\.jpg\?w=1920&h=1080$',
 238             'upload_date': '20210615',
 239             'timestamp': 1623722008,
 240         },
 241         'skip': '404 Not Found',
 242     }, {
 243         # japanese-language, longer id than english
 244         'url': 'https://www3.nhk.or.jp/nhkworld/ja/ondemand/video/0020271111/',
 245         'info_dict': {
 246             'id': 'nw_ja_v_jvod_ohayou_20231008',
 247             'ext': 'mp4',
 248             'title': 'おはよう日本（7時台） - 10月8日放送',
 249             'series': 'おはよう日本（7時台）',
 250             'episode': '10月8日放送',
 251             'thumbnail': r're:https://.+/.+\.jpg',
 252             'description': 'md5:9c1d6cbeadb827b955b20e99ab920ff0',
 253         },
 254         'skip': 'expires 2023-10-15',
 255     }, {
 256         # a one-off (single-episode series). title from the api is just '<p></p>'
 257         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/3004952/',
 258         'info_dict': {
 259             'id': 'nw_vod_v_en_3004_952_20230723091000_01_1690074552',
 260             'ext': 'mp4',
 261             'title': 'Barakan Discovers - AMAMI OSHIMA: Isson\'s Treasure Isla',
 262             'description': 'md5:5db620c46a0698451cc59add8816b797',
 263             'thumbnail': r're:https://.+/.+\.jpg',
 264             'release_date': '20230905',
 265             'timestamp': 1690103400,
 266             'duration': 2939,
 267             'release_timestamp': 1693898699,
 268             'upload_date': '20230723',
 269             'modified_timestamp': 1707217907,
 270             'modified_date': '20240206',
 271             'episode': 'AMAMI OSHIMA: Isson\'s Treasure Isla',
 272             'series': 'Barakan Discovers',
 273         },
 274     }, {
 275         # /ondemand/video/ url with alphabetical character in 5th position of id
 276         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999a07/',
 277         'info_dict': {
 278             'id': 'nw_c_en_9999-a07',
 279             'ext': 'mp4',
 280             'episode': 'Mini-Dramas on SDGs: Ep 1 Close the Gender Gap [Director\'s Cut]',
 281             'series': 'Mini-Dramas on SDGs',
 282             'modified_date': '20240206',
 283             'title': 'Mini-Dramas on SDGs - Mini-Dramas on SDGs: Ep 1 Close the Gender Gap [Director\'s Cut]',
 284             'description': 'md5:3f9dcb4db22fceb675d90448a040d3f6',
 285             'timestamp': 1621962360,
 286             'duration': 189,
 287             'release_date': '20230903',
 288             'modified_timestamp': 1707217907,
 289             'upload_date': '20210525',
 290             'thumbnail': r're:https://.+/.+\.jpg',
 291             'release_timestamp': 1693713487,
 292         },
 293     }, {
 294         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999d17/',
 295         'info_dict': {
 296             'id': 'nw_c_en_9999-d17',
 297             'ext': 'mp4',
 298             'title': 'Flowers of snow blossom - The 72 Pentads of Yamato',
 299             'description': 'Today’s focus: Snow',
 300             'release_timestamp': 1693792402,
 301             'release_date': '20230904',
 302             'upload_date': '20220128',
 303             'timestamp': 1643370960,
 304             'thumbnail': r're:https://.+/.+\.jpg',
 305             'duration': 136,
 306             'series': '',
 307             'modified_date': '20240206',
 308             'modified_timestamp': 1707217907,
 309         },
 310     }, {
 311         # new /shows/ url format
 312         'url': 'https://www3.nhk.or.jp/nhkworld/en/shows/2032307/',
 313         'info_dict': {
 314             'id': 'nw_vod_v_en_2032_307_20240321113000_01_1710990282',
 315             'ext': 'mp4',
 316             'title': 'Japanology Plus - 20th Anniversary Special Part 1',
 317             'description': 'md5:817d41fc8e54339ad2a916161ea24faf',
 318             'episode': '20th Anniversary Special Part 1',
 319             'series': 'Japanology Plus',
 320             'thumbnail': r're:https://.+/.+\.jpg',
 321             'duration': 1680,
 322             'timestamp': 1711020600,
 323             'upload_date': '20240321',
 324             'release_timestamp': 1711022683,
 325             'release_date': '20240321',
 326             'modified_timestamp': 1711031012,
 327             'modified_date': '20240321',
 328         },
 329     }, {
 330         'url': 'https://www3.nhk.or.jp/nhkworld/en/shows/3020025/',
 331         'info_dict': {
 332             'id': 'nw_vod_v_en_3020_025_20230325144000_01_1679723944',
 333             'ext': 'mp4',
 334             'title': '100 Ideas to Save the World - Working Styles Evolve',
 335             'description': 'md5:9e6c7778eaaf4f7b4af83569649f84d9',
 336             'episode': 'Working Styles Evolve',
 337             'series': '100 Ideas to Save the World',
 338             'thumbnail': r're:https://.+/.+\.jpg',
 339             'duration': 899,
 340             'upload_date': '20230325',
 341             'timestamp': 1679755200,
 342             'release_date': '20230905',
 343             'release_timestamp': 1693880540,
 344             'modified_date': '20240206',
 345             'modified_timestamp': 1707217907,
 346         },
 347     }, {
 348         # new /shows/audio/ url format
 349         'url': 'https://www3.nhk.or.jp/nhkworld/en/shows/audio/livinginjapan-20231001-1/',
 350         'only_matching': True,
 351     }, {
 352         # valid url even if can't be found in wild; support needed for clip entries extraction
 353         'url': 'https://www3.nhk.or.jp/nhkworld/en/shows/9999o80/',
 354         'only_matching': True,
 355     }]
 356
 357     def _real_extract(self, url):
 358         return self._extract_episode_info(url)
 359
 360
 361 class NhkVodProgramIE(NhkBaseIE):
 362     _VALID_URL = rf'''(?x)
 363         {NhkBaseIE._BASE_URL_REGEX}(?:shows|tv)/
 364         (?:(?P<type>audio)/programs/)?(?P<id>\w+)/?
 365         (?:\?(?:[^#]+&)?type=(?P<episode_type>clip|(?:radio|tv)Episode))?'''
 366     _TESTS = [{
 367         # video program episodes
 368         'url': 'https://www3.nhk.or.jp/nhkworld/en/shows/sumo/',
 369         'info_dict': {
 370             'id': 'sumo',
 371             'title': 'GRAND SUMO Highlights',
 372             'description': 'md5:fc20d02dc6ce85e4b72e0273aa52fdbf',
 373         },
 374         'playlist_mincount': 1,
 375     }, {
 376         'url': 'https://www3.nhk.or.jp/nhkworld/en/shows/japanrailway/',
 377         'info_dict': {
 378             'id': 'japanrailway',
 379             'title': 'Japan Railway Journal',
 380             'description': 'md5:ea39d93af7d05835baadf10d1aae0e3f',
 381         },
 382         'playlist_mincount': 12,
 383     }, {
 384         # video program clips
 385         'url': 'https://www3.nhk.or.jp/nhkworld/en/shows/japanrailway/?type=clip',
 386         'info_dict': {
 387             'id': 'japanrailway',
 388             'title': 'Japan Railway Journal',
 389             'description': 'md5:ea39d93af7d05835baadf10d1aae0e3f',
 390         },
 391         'playlist_mincount': 12,
 392     }, {
 393         # audio program
 394         'url': 'https://www3.nhk.or.jp/nhkworld/en/shows/audio/programs/livinginjapan/',
 395         'info_dict': {
 396             'id': 'livinginjapan',
 397             'title': 'Living in Japan',
 398             'description': 'md5:665bb36ec2a12c5a7f598ee713fc2b54',
 399         },
 400         'playlist_mincount': 12,
 401     }, {
 402         # /tv/ program url
 403         'url': 'https://www3.nhk.or.jp/nhkworld/en/tv/designtalksplus/',
 404         'info_dict': {
 405             'id': 'designtalksplus',
 406             'title': 'DESIGN TALKS plus',
 407             'description': 'md5:47b3b3a9f10d4ac7b33b53b70a7d2837',
 408         },
 409         'playlist_mincount': 20,
 410     }, {
 411         'url': 'https://www3.nhk.or.jp/nhkworld/en/shows/10yearshayaomiyazaki/',
 412         'only_matching': True,
 413     }]
 414
 415     @classmethod
 416     def suitable(cls, url):
 417         return False if NhkVodIE.suitable(url) else super().suitable(url)
 418
 419     def _extract_meta_from_class_elements(self, class_values, html):
 420         for class_value in class_values:
 421             if value := clean_html(get_element_by_class(class_value, html)):
 422                 return value
 423
 424     def _real_extract(self, url):
 425         lang, m_type, program_id, episode_type = self._match_valid_url(url).group('lang', 'type', 'id', 'episode_type')
 426         episodes = self._call_api(
 427             program_id, lang, m_type != 'audio', False, episode_type == 'clip')
 428
 429         def entries():
 430             for episode in episodes:
 431                 if episode_path := episode.get('url'):
 432                     yield self._extract_episode_info(urljoin(url, episode_path), episode)
 433
 434         html = self._download_webpage(url, program_id)
 435         program_title = self._extract_meta_from_class_elements([
 436             'p-programDetail__title',  # /ondemand/program/
 437             'pProgramHero__logoText',  # /shows/
 438             'tAudioProgramMain__title',  # /shows/audio/programs/
 439             'p-program-name'], html)  # /tv/
 440         program_description = self._extract_meta_from_class_elements([
 441             'p-programDetail__text',  # /ondemand/program/
 442             'pProgramHero__description',  # /shows/
 443             'tAudioProgramMain__info',  # /shows/audio/programs/
 444             'p-program-description'], html)  # /tv/
 445
 446         return self.playlist_result(entries(), program_id, program_title, program_description)
 447
 448
 449 class NhkForSchoolBangumiIE(InfoExtractor):
 450     _VALID_URL = r'https?://www2\.nhk\.or\.jp/school/movie/(?P<type>bangumi|clip)\.cgi\?das_id=(?P<id>[a-zA-Z0-9_-]+)'
 451     _TESTS = [{
 452         'url': 'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id=D0005150191_00000',
 453         'info_dict': {
 454             'id': 'D0005150191_00003',
 455             'title': 'にている かな',
 456             'duration': 599.999,
 457             'timestamp': 1396414800,
 458
 459             'upload_date': '20140402',
 460             'ext': 'mp4',
 461
 462             'chapters': 'count:12',
 463         },
 464         'params': {
 465             # m3u8 download
 466             'skip_download': True,
 467         },
 468     }]
 469
 470     def _real_extract(self, url):
 471         program_type, video_id = self._match_valid_url(url).groups()
 472
 473         webpage = self._download_webpage(
 474             f'https://www2.nhk.or.jp/school/movie/{program_type}.cgi?das_id={video_id}', video_id)
 475
 476         # searches all variables
 477         base_values = {g.group(1): g.group(2) for g in re.finditer(r'var\s+([a-zA-Z_]+)\s*=\s*"([^"]+?)";', webpage)}
 478         # and programObj values too
 479         program_values = {g.group(1): g.group(3) for g in re.finditer(r'(?:program|clip)Obj\.([a-zA-Z_]+)\s*=\s*(["\'])([^"]+?)\2;', webpage)}
 480         # extract all chapters
 481         chapter_durations = [parse_duration(g.group(1)) for g in re.finditer(r'chapterTime\.push\(\'([0-9:]+?)\'\);', webpage)]
 482         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)]
 483
 484         # this is how player_core.js is actually doing (!)
 485         version = base_values.get('r_version') or program_values.get('version')
 486         if version:
 487             video_id = f'{video_id.split("_")[0]}_{version}'
 488
 489         formats = self._extract_m3u8_formats(
 490             f'https://nhks-vh.akamaihd.net/i/das/{video_id[0:8]}/{video_id}_V_000.f4v/master.m3u8',
 491             video_id, ext='mp4', m3u8_id='hls')
 492
 493         duration = parse_duration(base_values.get('r_duration'))
 494
 495         chapters = None
 496         if chapter_durations and chapter_titles and len(chapter_durations) == len(chapter_titles):
 497             start_time = chapter_durations
 498             end_time = chapter_durations[1:] + [duration]
 499             chapters = [{
 500                 'start_time': s,
 501                 'end_time': e,
 502                 'title': t,
 503             } for s, e, t in zip(start_time, end_time, chapter_titles)]
 504
 505         return {
 506             'id': video_id,
 507             'title': program_values.get('name'),
 508             'duration': parse_duration(base_values.get('r_duration')),
 509             'timestamp': unified_timestamp(base_values['r_upload']),
 510             'formats': formats,
 511             'chapters': chapters,
 512         }
 513
 514
 515 class NhkForSchoolSubjectIE(InfoExtractor):
 516     IE_DESC = 'Portal page for each school subjects, like Japanese (kokugo, 国語) or math (sansuu/suugaku or 算数・数学)'
 517     KNOWN_SUBJECTS = (
 518         'rika', 'syakai', 'kokugo',
 519         'sansuu', 'seikatsu', 'doutoku',
 520         'ongaku', 'taiiku', 'zukou',
 521         'gijutsu', 'katei', 'sougou',
 522         'eigo', 'tokkatsu',
 523         'tokushi', 'sonota',
 524     )
 525     _VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>{})/?(?:[\?#].*)?$'.format(
 526         '|'.join(re.escape(s) for s in KNOWN_SUBJECTS))
 527
 528     _TESTS = [{
 529         'url': 'https://www.nhk.or.jp/school/sougou/',
 530         'info_dict': {
 531             'id': 'sougou',
 532             'title': '総合的な学習の時間',
 533         },
 534         'playlist_mincount': 16,
 535     }, {
 536         'url': 'https://www.nhk.or.jp/school/rika/',
 537         'info_dict': {
 538             'id': 'rika',
 539             'title': '理科',
 540         },
 541         'playlist_mincount': 15,
 542     }]
 543
 544     def _real_extract(self, url):
 545         subject_id = self._match_id(url)
 546         webpage = self._download_webpage(url, subject_id)
 547
 548         return self.playlist_from_matches(
 549             re.finditer(rf'href="((?:https?://www\.nhk\.or\.jp)?/school/{re.escape(subject_id)}/[^/]+/)"', webpage),
 550             subject_id,
 551             self._html_search_regex(r'(?s)<span\s+class="subjectName">\s*<img\s*[^<]+>\s*([^<]+?)</span>', webpage, 'title', fatal=False),
 552             lambda g: urljoin(url, g.group(1)))
 553
 554
 555 class NhkForSchoolProgramListIE(InfoExtractor):
 556     _VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>(?:{})/[a-zA-Z0-9_-]+)'.format(
 557         '|'.join(re.escape(s) for s in NhkForSchoolSubjectIE.KNOWN_SUBJECTS))
 558     _TESTS = [{
 559         'url': 'https://www.nhk.or.jp/school/sougou/q/',
 560         'info_dict': {
 561             'id': 'sougou/q',
 562             'title': 'Ｑ～こどものための哲学',
 563         },
 564         'playlist_mincount': 20,
 565     }]
 566
 567     def _real_extract(self, url):
 568         program_id = self._match_id(url)
 569
 570         webpage = self._download_webpage(f'https://www.nhk.or.jp/school/{program_id}/', program_id)
 571
 572         title = (self._generic_title('', webpage)
 573                  or self._html_search_regex(r'<h3>([^<]+?)とは？\s*</h3>', webpage, 'title', fatal=False))
 574         title = re.sub(r'\s*\|\s*NHK\s+for\s+School\s*$', '', title) if title else None
 575         description = self._html_search_regex(
 576             r'(?s)<div\s+class="programDetail\s*">\s*<p>[^<]+</p>',
 577             webpage, 'description', fatal=False, group=0)
 578
 579         bangumi_list = self._download_json(
 580             f'https://www.nhk.or.jp/school/{program_id}/meta/program.json', program_id)
 581         # they're always bangumi
 582         bangumis = [
 583             self.url_result(f'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id={x}')
 584             for x in traverse_obj(bangumi_list, ('part', ..., 'part-video-dasid')) or []]
 585
 586         return self.playlist_result(bangumis, program_id, title, description)
 587
 588
 589 class NhkRadiruIE(InfoExtractor):
 590     _GEO_COUNTRIES = ['JP']
 591     IE_DESC = 'NHK らじる (Radiru/Rajiru)'
 592     _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]+))?'
 593     _TESTS = [{
 594         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=0449_01_4003239',
 595         'skip': 'Episode expired on 2024-06-09',
 596         'info_dict': {
 597             'title': 'ジャズ・トゥナイト　ジャズ「Ｎｉｇｈｔ　ａｎｄ　Ｄａｙ」特集',
 598             'id': '0449_01_4003239',
 599             'ext': 'm4a',
 600             'uploader': 'NHK FM 東京',
 601             'description': 'md5:ad05f3c3f3f6e99b2e69f9b5e49551dc',
 602             'series': 'ジャズ・トゥナイト',
 603             'channel': 'NHK FM 東京',
 604             'thumbnail': 'https://www.nhk.or.jp/prog/img/449/g449.jpg',
 605             'upload_date': '20240601',
 606             'series_id': '0449_01',
 607             'release_date': '20240601',
 608             'timestamp': 1717257600,
 609             'release_timestamp': 1717250400,
 610         },
 611     }, {
 612         # playlist, airs every weekday so it should _hopefully_ be okay forever
 613         'url': 'https://www.nhk.or.jp/radio/ondemand/detail.html?p=0458_01',
 614         'info_dict': {
 615             'id': '0458_01',
 616             'title': 'ベストオブクラシック',
 617             '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。',
 618             'thumbnail': 'https://www.nhk.or.jp/prog/img/458/g458.jpg',
 619             'series_id': '0458_01',
 620             'uploader': 'NHK FM',
 621             'channel': 'NHK FM',
 622             'series': 'ベストオブクラシック',
 623         },
 624         'playlist_mincount': 3,
 625     }, {
 626         # one with letters in the id
 627         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F683_01_3910688',
 628         'note': 'Expires on 2025-03-31',
 629         'info_dict': {
 630             'id': 'F683_01_3910688',
 631             'ext': 'm4a',
 632             'title': '夏目漱石「文鳥」第1回',
 633             'series': '【らじる文庫】夏目漱石「文鳥」（全4回）',
 634             'series_id': 'F683_01',
 635             'description': '朗読：浅井理アナウンサー',
 636             'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F683/img/roudoku_05_rod_640.jpg',
 637             'upload_date': '20240106',
 638             'release_date': '20240106',
 639             'uploader': 'NHK R1',
 640             'release_timestamp': 1704511800,
 641             'channel': 'NHK R1',
 642             'timestamp': 1704512700,
 643         },
 644         'expected_warnings': ['Unable to download JSON metadata',
 645                               'Failed to get extended metadata. API returned Error 1: Invalid parameters'],
 646     }, {
 647         # news
 648         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F261_01_4012173',
 649         'info_dict': {
 650             'id': 'F261_01_4012173',
 651             'ext': 'm4a',
 652             'channel': 'NHKラジオ第1',
 653             'uploader': 'NHKラジオ第1',
 654             'series': 'NHKラジオニュース',
 655             'title': '午前０時のNHKニュース',
 656             'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
 657             'release_timestamp': 1718290800,
 658             'release_date': '20240613',
 659             'timestamp': 1718291400,
 660             'upload_date': '20240613',
 661         },
 662     }, {
 663         # fallback when extended metadata fails
 664         'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=2834_01_4009298',
 665         'skip': 'Expires on 2024-06-07',
 666         'info_dict': {
 667             'id': '2834_01_4009298',
 668             'title': 'まち☆キラ！開成町特集',
 669             'ext': 'm4a',
 670             'release_date': '20240531',
 671             'upload_date': '20240531',
 672             'series': 'はま☆キラ！',
 673             'thumbnail': 'https://www.nhk.or.jp/prog/img/2834/g2834.jpg',
 674             'channel': 'NHK R1,FM',
 675             'description': '',
 676             'timestamp': 1717123800,
 677             'uploader': 'NHK R1,FM',
 678             'release_timestamp': 1717120800,
 679             'series_id': '2834_01',
 680         },
 681         'expected_warnings': ['Failed to get extended metadata. API returned empty list.'],
 682     }]
 683
 684     _API_URL_TMPL = None
 685
 686     def _extract_extended_metadata(self, episode_id, aa_vinfo):
 687         service, _, area = traverse_obj(aa_vinfo, (2, {str}, {lambda x: (x or '').partition(',')}))
 688         detail_url = try_call(
 689             lambda: self._API_URL_TMPL.format(area=area, service=service, dateid=aa_vinfo[3]))
 690         if not detail_url:
 691             return {}
 692
 693         response = self._download_json(
 694             detail_url, episode_id, 'Downloading extended metadata',
 695             'Failed to download extended metadata', fatal=False, expected_status=400)
 696         if not response:
 697             return {}
 698
 699         if error := traverse_obj(response, ('error', {dict})):
 700             self.report_warning(
 701                 'Failed to get extended metadata. API returned '
 702                 f'Error {join_nonempty("code", "message", from_dict=error, delim=": ")}')
 703             return {}
 704
 705         full_meta = traverse_obj(response, ('list', service, 0, {dict}))
 706         if not full_meta:
 707             self.report_warning('Failed to get extended metadata. API returned empty list.')
 708             return {}
 709
 710         station = ' '.join(traverse_obj(full_meta, (('service', 'area'), 'name', {str}))) or None
 711         thumbnails = [{
 712             'id': str(id_),
 713             'preference': 1 if id_.startswith('thumbnail') else -2 if id_.startswith('logo') else -1,
 714             **traverse_obj(thumb, {
 715                 'url': 'url',
 716                 'width': ('width', {int_or_none}),
 717                 'height': ('height', {int_or_none}),
 718             }),
 719         } for id_, thumb in traverse_obj(full_meta, ('images', {dict.items}, lambda _, v: v[1]['url']))]
 720
 721         return filter_dict({
 722             'channel': station,
 723             'uploader': station,
 724             'description': join_nonempty(
 725                 'subtitle', 'content', 'act', 'music', delim='\n\n', from_dict=full_meta),
 726             'thumbnails': thumbnails,
 727             **traverse_obj(full_meta, {
 728                 'title': ('title', {str}),
 729                 'timestamp': ('end_time', {unified_timestamp}),
 730                 'release_timestamp': ('start_time', {unified_timestamp}),
 731             }),
 732         })
 733
 734     def _extract_episode_info(self, episode, programme_id, series_meta):
 735         episode_id = f'{programme_id}_{episode["id"]}'
 736         aa_vinfo = traverse_obj(episode, ('aa_contents_id', {lambda x: x.split(';')}))
 737         extended_metadata = self._extract_extended_metadata(episode_id, aa_vinfo)
 738         fallback_start_time, _, fallback_end_time = traverse_obj(
 739             aa_vinfo, (4, {str}, {lambda x: (x or '').partition('_')}))
 740
 741         return {
 742             **series_meta,
 743             'id': episode_id,
 744             'formats': self._extract_m3u8_formats(episode.get('stream_url'), episode_id, fatal=False),
 745             'container': 'm4a_dash',  # force fixup, AAC-only HLS
 746             'was_live': True,
 747             'title': episode.get('program_title'),
 748             'description': episode.get('program_sub_title'),  # fallback
 749             'timestamp': unified_timestamp(fallback_end_time),
 750             'release_timestamp': unified_timestamp(fallback_start_time),
 751             **extended_metadata,
 752         }
 753
 754     def _extract_news_info(self, headline, programme_id, series_meta):
 755         episode_id = f'{programme_id}_{headline["headline_id"]}'
 756         episode = traverse_obj(headline, ('file_list', 0, {dict}))
 757
 758         return {
 759             **series_meta,
 760             'id': episode_id,
 761             'formats': self._extract_m3u8_formats(episode.get('file_name'), episode_id, fatal=False),
 762             'container': 'm4a_dash',  # force fixup, AAC-only HLS
 763             'was_live': True,
 764             'series': series_meta.get('title'),
 765             'thumbnail': url_or_none(headline.get('headline_image')) or series_meta.get('thumbnail'),
 766             **traverse_obj(episode, {
 767                 'title': ('file_title', {str}),
 768                 'description': ('file_title_sub', {str}),
 769                 'timestamp': ('open_time', {unified_timestamp}),
 770                 'release_timestamp': ('aa_vinfo4', {lambda x: x.split('_')[0]}, {unified_timestamp}),
 771             }),
 772         }
 773
 774     def _real_initialize(self):
 775         if self._API_URL_TMPL:
 776             return
 777         api_config = self._download_xml(
 778             'https://www.nhk.or.jp/radio/config/config_web.xml', None, 'Downloading API config', fatal=False)
 779         NhkRadiruIE._API_URL_TMPL = try_call(lambda: f'https:{api_config.find(".//url_program_detail").text}')
 780
 781     def _real_extract(self, url):
 782         site_id, corner_id, headline_id = self._match_valid_url(url).group('site', 'corner', 'headline')
 783         programme_id = f'{site_id}_{corner_id}'
 784
 785         if site_id == 'F261':  # XXX: News programmes use old API (for now?)
 786             meta = self._download_json(
 787                 'https://www.nhk.or.jp/s-media/news/news-site/list/v1/all.json', programme_id)['main']
 788             series_meta = traverse_obj(meta, {
 789                 'title': ('program_name', {str}),
 790                 'channel': ('media_name', {str}),
 791                 'uploader': ('media_name', {str}),
 792                 'thumbnail': (('thumbnail_c', 'thumbnail_p'), {url_or_none}),
 793             }, get_all=False)
 794
 795             if headline_id:
 796                 headline = traverse_obj(
 797                     meta, ('detail_list', lambda _, v: v['headline_id'] == headline_id, any))
 798                 if not headline:
 799                     raise ExtractorError('Content not found; it has most likely expired', expected=True)
 800                 return self._extract_news_info(headline, programme_id, series_meta)
 801
 802             def news_entries():
 803                 for headline in traverse_obj(meta, ('detail_list', ..., {dict})):
 804                     yield self._extract_news_info(headline, programme_id, series_meta)
 805
 806             return self.playlist_result(
 807                 news_entries(), programme_id, description=meta.get('site_detail'), **series_meta)
 808
 809         meta = self._download_json(
 810             'https://www.nhk.or.jp/radio-api/app/v1/web/ondemand/series', programme_id, query={
 811                 'site_id': site_id,
 812                 'corner_site_id': corner_id,
 813             })
 814
 815         fallback_station = join_nonempty('NHK', traverse_obj(meta, ('radio_broadcast', {str})), delim=' ')
 816         series_meta = {
 817             'series': join_nonempty('title', 'corner_name', delim=' ', from_dict=meta),
 818             'series_id': programme_id,
 819             'thumbnail': traverse_obj(meta, ('thumbnail_url', {url_or_none})),
 820             'channel': fallback_station,
 821             'uploader': fallback_station,
 822         }
 823
 824         if headline_id:
 825             episode = traverse_obj(meta, ('episodes', lambda _, v: v['id'] == int(headline_id), any))
 826             if not episode:
 827                 raise ExtractorError('Content not found; it has most likely expired', expected=True)
 828             return self._extract_episode_info(episode, programme_id, series_meta)
 829
 830         def entries():
 831             for episode in traverse_obj(meta, ('episodes', ..., {dict})):
 832                 yield self._extract_episode_info(episode, programme_id, series_meta)
 833
 834         return self.playlist_result(
 835             entries(), programme_id, title=series_meta.get('series'),
 836             description=meta.get('series_description'), **series_meta)
 837
 838
 839 class NhkRadioNewsPageIE(InfoExtractor):
 840     _VALID_URL = r'https?://www\.nhk\.or\.jp/radionews/?(?:$|[?#])'
 841     _TESTS = [{
 842         # airs daily, on-the-hour most hours
 843         'url': 'https://www.nhk.or.jp/radionews/',
 844         'playlist_mincount': 5,
 845         'info_dict': {
 846             'id': 'F261_01',
 847             'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
 848             'description': 'md5:bf2c5b397e44bc7eb26de98d8f15d79d',
 849             'channel': 'NHKラジオ第1',
 850             'uploader': 'NHKラジオ第1',
 851             'title': 'NHKラジオニュース',
 852         },
 853     }]
 854
 855     def _real_extract(self, url):
 856         return self.url_result('https://www.nhk.or.jp/radio/ondemand/detail.html?p=F261_01', NhkRadiruIE)
 857
 858
 859 class NhkRadiruLiveIE(InfoExtractor):
 860     _GEO_COUNTRIES = ['JP']
 861     _VALID_URL = r'https?://www\.nhk\.or\.jp/radio/player/\?ch=(?P<id>r[12]|fm)'
 862     _TESTS = [{
 863         # radio 1, no area specified
 864         'url': 'https://www.nhk.or.jp/radio/player/?ch=r1',
 865         'info_dict': {
 866             'id': 'r1-tokyo',
 867             'title': 're:^ＮＨＫネットラジオ第1 東京.+$',
 868             'ext': 'm4a',
 869             'thumbnail': 'https://www.nhk.or.jp/common/img/media/r1-200x200.png',
 870             'live_status': 'is_live',
 871         },
 872     }, {
 873         # radio 2, area specified
 874         # (the area doesnt actually matter, r2 is national)
 875         'url': 'https://www.nhk.or.jp/radio/player/?ch=r2',
 876         'params': {'extractor_args': {'nhkradirulive': {'area': ['fukuoka']}}},
 877         'info_dict': {
 878             'id': 'r2-fukuoka',
 879             'title': 're:^ＮＨＫネットラジオ第2 福岡.+$',
 880             'ext': 'm4a',
 881             'thumbnail': 'https://www.nhk.or.jp/common/img/media/r2-200x200.png',
 882             'live_status': 'is_live',
 883         },
 884     }, {
 885         # fm, area specified
 886         'url': 'https://www.nhk.or.jp/radio/player/?ch=fm',
 887         'params': {'extractor_args': {'nhkradirulive': {'area': ['sapporo']}}},
 888         'info_dict': {
 889             'id': 'fm-sapporo',
 890             'title': 're:^ＮＨＫネットラジオＦＭ 札幌.+$',
 891             'ext': 'm4a',
 892             'thumbnail': 'https://www.nhk.or.jp/common/img/media/fm-200x200.png',
 893             'live_status': 'is_live',
 894         },
 895     }]
 896
 897     _NOA_STATION_IDS = {'r1': 'n1', 'r2': 'n2', 'fm': 'n3'}
 898
 899     def _real_extract(self, url):
 900         station = self._match_id(url)
 901         area = self._configuration_arg('area', ['tokyo'])[0]
 902
 903         config = self._download_xml(
 904             'https://www.nhk.or.jp/radio/config/config_web.xml', station, 'Downloading area information')
 905         data = config.find(f'.//data//area[.="{area}"]/..')
 906
 907         if not data:
 908             raise ExtractorError('Invalid area. Valid areas are: {}'.format(', '.join(
 909                 [i.text for i in config.findall('.//data//area')])), expected=True)
 910
 911         noa_info = self._download_json(
 912             f'https:{config.find(".//url_program_noa").text}'.format(area=data.find('areakey').text),
 913             station, note=f'Downloading {area} station metadata', fatal=False)
 914         present_info = traverse_obj(noa_info, ('nowonair_list', self._NOA_STATION_IDS.get(station), 'present'))
 915
 916         return {
 917             'title': ' '.join(traverse_obj(present_info, (('service', 'area'), 'name', {str}))),
 918             'id': join_nonempty(station, area),
 919             'thumbnails': traverse_obj(present_info, ('service', 'images', ..., {
 920                 'url': 'url',
 921                 'width': ('width', {int_or_none}),
 922                 'height': ('height', {int_or_none}),
 923             })),
 924             'formats': self._extract_m3u8_formats(data.find(f'{station}hls').text, station),
 925             'is_live': True,
 926         }