[yt-dlp.git] / yt_dlp / extractor / nhk.py

import re

from .common import InfoExtractor
from ..utils import (
    parse_duration,
    traverse_obj,
    unescapeHTML,
    unified_timestamp,
    urljoin,
    url_or_none
)


class NhkBaseIE(InfoExtractor):
    _API_URL_TEMPLATE = 'https://nwapi.nhk.jp/nhkworld/%sod%slist/v7b/%s/%s/%s/all%s.json'
    _BASE_URL_REGEX = r'https?://www3\.nhk\.or\.jp/nhkworld/(?P<lang>[a-z]{2})/ondemand'
    _TYPE_REGEX = r'/(?P<type>video|audio)/'

    def _call_api(self, m_id, lang, is_video, is_episode, is_clip):
        return self._download_json(
            self._API_URL_TEMPLATE % (
                'v' if is_video else 'r',
                'clip' if is_clip else 'esd',
                'episode' if is_episode else 'program',
                m_id, lang, '/all' if is_video else ''),
            m_id, query={'apikey': 'EJfK8jdS57GqlupFgAfAAwr573q01y6k'})['data']['episodes'] or []

    def _extract_episode_info(self, url, episode=None):
        fetch_episode = episode is None
        lang, m_type, episode_id = NhkVodIE._match_valid_url(url).groups()
        if len(episode_id) == 7:
            episode_id = episode_id[:4] + '-' + episode_id[4:]

        is_video = m_type == 'video'
        if fetch_episode:
            episode = self._call_api(
                episode_id, lang, is_video, True, episode_id[:4] == '9999')[0]
        title = episode.get('sub_title_clean') or episode['sub_title']

        def get_clean_field(key):
            return episode.get(key + '_clean') or episode.get(key)

        series = get_clean_field('title')

        thumbnails = []
        for s, w, h in [('', 640, 360), ('_l', 1280, 720)]:
            img_path = episode.get('image' + s)
            if not img_path:
                continue
            thumbnails.append({
                'id': '%dp' % h,
                'height': h,
                'width': w,
                'url': 'https://www3.nhk.or.jp' + img_path,
            })

        info = {
            'id': episode_id + '-' + lang,
            'title': '%s - %s' % (series, title) if series and title else title,
            'description': get_clean_field('description'),
            'thumbnails': thumbnails,
            'series': series,
            'episode': title,
        }
        if is_video:
            vod_id = episode['vod_id']
            info.update({
                '_type': 'url_transparent',
                'ie_key': 'Piksel',
                'url': 'https://player.piksel.com/v/refid/nhkworld/prefid/' + vod_id,
                'id': vod_id,
            })
        else:
            if fetch_episode:
                audio_path = episode['audio']['audio']
                info['formats'] = self._extract_m3u8_formats(
                    'https://nhkworld-vh.akamaihd.net/i%s/master.m3u8' % audio_path,
                    episode_id, 'm4a', entry_protocol='m3u8_native',
                    m3u8_id='hls', fatal=False)
                for f in info['formats']:
                    f['language'] = lang
            else:
                info.update({
                    '_type': 'url_transparent',
                    'ie_key': NhkVodIE.ie_key(),
                    'url': url,
                })
        return info


class NhkVodIE(NhkBaseIE):
    # the 7-character IDs can have alphabetic chars too: assume [a-z] rather than just [a-f], eg
    _VALID_URL = r'%s%s(?P<id>[0-9a-z]{7}|[^/]+?-\d{8}-[0-9a-z]+)' % (NhkBaseIE._BASE_URL_REGEX, NhkBaseIE._TYPE_REGEX)
    # Content available only for a limited period of time. Visit
    # https://www3.nhk.or.jp/nhkworld/en/ondemand/ for working samples.
    _TESTS = [{
        # video clip
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999011/',
        'md5': '7a90abcfe610ec22a6bfe15bd46b30ca',
        'info_dict': {
            'id': 'a95j5iza',
            'ext': 'mp4',
            'title': "Dining with the Chef - Chef Saito's Family recipe: MENCHI-KATSU",
            'description': 'md5:5aee4a9f9d81c26281862382103b0ea5',
            'timestamp': 1565965194,
            'upload_date': '20190816',
        },
    }, {
        # audio clip
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/r_inventions-20201104-1/',
        'info_dict': {
            'id': 'r_inventions-20201104-1-en',
            'ext': 'm4a',
            'title': "Japan's Top Inventions - Miniature Video Cameras",
            'description': 'md5:07ea722bdbbb4936fdd360b6a480c25b',
        },
        'params': {
            # m3u8 download
            'skip_download': True,
        },
    }, {
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2015173/',
        'only_matching': True,
    }, {
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/plugin-20190404-1/',
        'only_matching': True,
    }, {
        'url': 'https://www3.nhk.or.jp/nhkworld/fr/ondemand/audio/plugin-20190404-1/',
        'only_matching': True,
    }, {
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/j_art-20150903-1/',
        'only_matching': True,
    }, {
        # video, alphabetic character in ID #29670
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999a34/',
        'only_matching': True,
        'info_dict': {
            'id': 'qfjay6cg',
            'ext': 'mp4',
            'title': 'DESIGN TALKS plus - Fishermen’s Finery',
            'description': 'md5:8a8f958aaafb0d7cb59d38de53f1e448',
            'thumbnail': r're:^https?:/(/[a-z0-9.-]+)+\.jpg\?w=1920&h=1080$',
            'upload_date': '20210615',
            'timestamp': 1623722008,
        }
    }]

    def _real_extract(self, url):
        return self._extract_episode_info(url)


class NhkVodProgramIE(NhkBaseIE):
    _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)
    _TESTS = [{
        # video program episodes
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway',
        'info_dict': {
            'id': 'japanrailway',
            'title': 'Japan Railway Journal',
        },
        'playlist_mincount': 1,
    }, {
        # video program clips
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway/?type=clip',
        'info_dict': {
            'id': 'japanrailway',
            'title': 'Japan Railway Journal',
        },
        'playlist_mincount': 5,
    }, {
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/10yearshayaomiyazaki/',
        'only_matching': True,
    }, {
        # audio program
        'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/audio/listener/',
        'only_matching': True,
    }]

    def _real_extract(self, url):
        lang, m_type, program_id, episode_type = self._match_valid_url(url).groups()

        episodes = self._call_api(
            program_id, lang, m_type == 'video', False, episode_type == 'clip')

        entries = []
        for episode in episodes:
            episode_path = episode.get('url')
            if not episode_path:
                continue
            entries.append(self._extract_episode_info(
                urljoin(url, episode_path), episode))

        program_title = None
        if entries:
            program_title = entries[0].get('series')

        return self.playlist_result(entries, program_id, program_title)


class NhkForSchoolBangumiIE(InfoExtractor):
    _VALID_URL = r'https?://www2\.nhk\.or\.jp/school/movie/(?P<type>bangumi|clip)\.cgi\?das_id=(?P<id>[a-zA-Z0-9_-]+)'
    _TESTS = [{
        'url': 'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id=D0005150191_00000',
        'info_dict': {
            'id': 'D0005150191_00003',
            'title': 'にている かな',
            'duration': 599.999,
            'timestamp': 1396414800,

            'upload_date': '20140402',
            'ext': 'mp4',

            'chapters': 'count:12'
        },
        'params': {
            # m3u8 download
            'skip_download': True,
        },
    }]

    def _real_extract(self, url):
        program_type, video_id = self._match_valid_url(url).groups()

        webpage = self._download_webpage(
            f'https://www2.nhk.or.jp/school/movie/{program_type}.cgi?das_id={video_id}', video_id)

        # searches all variables
        base_values = {g.group(1): g.group(2) for g in re.finditer(r'var\s+([a-zA-Z_]+)\s*=\s*"([^"]+?)";', webpage)}
        # and programObj values too
        program_values = {g.group(1): g.group(3) for g in re.finditer(r'(?:program|clip)Obj\.([a-zA-Z_]+)\s*=\s*(["\'])([^"]+?)\2;', webpage)}
        # extract all chapters
        chapter_durations = [parse_duration(g.group(1)) for g in re.finditer(r'chapterTime\.push\(\'([0-9:]+?)\'\);', webpage)]
        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)]

        # this is how player_core.js is actually doing (!)
        version = base_values.get('r_version') or program_values.get('version')
        if version:
            video_id = f'{video_id.split("_")[0]}_{version}'

        formats = self._extract_m3u8_formats(
            f'https://nhks-vh.akamaihd.net/i/das/{video_id[0:8]}/{video_id}_V_000.f4v/master.m3u8',
            video_id, ext='mp4', m3u8_id='hls')

        duration = parse_duration(base_values.get('r_duration'))

        chapters = None
        if chapter_durations and chapter_titles and len(chapter_durations) == len(chapter_titles):
            start_time = chapter_durations
            end_time = chapter_durations[1:] + [duration]
            chapters = [{
                'start_time': s,
                'end_time': e,
                'title': t,
            } for s, e, t in zip(start_time, end_time, chapter_titles)]

        return {
            'id': video_id,
            'title': program_values.get('name'),
            'duration': parse_duration(base_values.get('r_duration')),
            'timestamp': unified_timestamp(base_values['r_upload']),
            'formats': formats,
            'chapters': chapters,
        }


class NhkForSchoolSubjectIE(InfoExtractor):
    IE_DESC = 'Portal page for each school subjects, like Japanese (kokugo, 国語) or math (sansuu/suugaku or 算数・数学)'
    KNOWN_SUBJECTS = (
        'rika', 'syakai', 'kokugo',
        'sansuu', 'seikatsu', 'doutoku',
        'ongaku', 'taiiku', 'zukou',
        'gijutsu', 'katei', 'sougou',
        'eigo', 'tokkatsu',
        'tokushi', 'sonota',
    )
    _VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>%s)/?(?:[\?#].*)?$' % '|'.join(re.escape(s) for s in KNOWN_SUBJECTS)

    _TESTS = [{
        'url': 'https://www.nhk.or.jp/school/sougou/',
        'info_dict': {
            'id': 'sougou',
            'title': '総合的な学習の時間',
        },
        'playlist_mincount': 16,
    }, {
        'url': 'https://www.nhk.or.jp/school/rika/',
        'info_dict': {
            'id': 'rika',
            'title': '理科',
        },
        'playlist_mincount': 15,
    }]

    def _real_extract(self, url):
        subject_id = self._match_id(url)
        webpage = self._download_webpage(url, subject_id)

        return self.playlist_from_matches(
            re.finditer(rf'href="((?:https?://www\.nhk\.or\.jp)?/school/{re.escape(subject_id)}/[^/]+/)"', webpage),
            subject_id,
            self._html_search_regex(r'(?s)<span\s+class="subjectName">\s*<img\s*[^<]+>\s*([^<]+?)</span>', webpage, 'title', fatal=False),
            lambda g: urljoin(url, g.group(1)))


class NhkForSchoolProgramListIE(InfoExtractor):
    _VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>(?:%s)/[a-zA-Z0-9_-]+)' % (
        '|'.join(re.escape(s) for s in NhkForSchoolSubjectIE.KNOWN_SUBJECTS)
    )
    _TESTS = [{
        'url': 'https://www.nhk.or.jp/school/sougou/q/',
        'info_dict': {
            'id': 'sougou/q',
            'title': 'Ｑ～こどものための哲学',
        },
        'playlist_mincount': 20,
    }]

    def _real_extract(self, url):
        program_id = self._match_id(url)

        webpage = self._download_webpage(f'https://www.nhk.or.jp/school/{program_id}/', program_id)

        title = (self._generic_title('', webpage)
                 or self._html_search_regex(r'<h3>([^<]+?)とは？\s*</h3>', webpage, 'title', fatal=False))
        title = re.sub(r'\s*\|\s*NHK\s+for\s+School\s*$', '', title) if title else None
        description = self._html_search_regex(
            r'(?s)<div\s+class="programDetail\s*">\s*<p>[^<]+</p>',
            webpage, 'description', fatal=False, group=0)

        bangumi_list = self._download_json(
            f'https://www.nhk.or.jp/school/{program_id}/meta/program.json', program_id)
        # they're always bangumi
        bangumis = [
            self.url_result(f'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id={x}')
            for x in traverse_obj(bangumi_list, ('part', ..., 'part-video-dasid')) or []]

        return self.playlist_result(bangumis, program_id, title, description)


class NhkRadiruIE(InfoExtractor):
    _GEO_COUNTRIES = ['JP']
    IE_DESC = 'NHK らじる (Radiru/Rajiru)'
    _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]+))?'
    _TESTS = [{
        'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=0449_01_3853544',
        'skip': 'Episode expired on 2023-04-16',
        'info_dict': {
            'channel': 'NHK-FM',
            'description': 'md5:94b08bdeadde81a97df4ec882acce3e9',
            'ext': 'm4a',
            'id': '0449_01_3853544',
            'series': 'ジャズ・トゥナイト',
            'thumbnail': 'https://www.nhk.or.jp/prog/img/449/g449.jpg',
            'timestamp': 1680969600,
            'title': 'ジャズ・トゥナイト　ＮＥＷジャズ特集',
            'upload_date': '20230408',
            'release_timestamp': 1680962400,
            'release_date': '20230408',
            'was_live': True,
        },
    }, {
        # playlist, airs every weekday so it should _hopefully_ be okay forever
        'url': 'https://www.nhk.or.jp/radio/ondemand/detail.html?p=0458_01',
        'info_dict': {
            'id': '0458_01',
            'title': 'ベストオブクラシック',
            '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。',
            'channel': 'NHK-FM',
            'thumbnail': 'https://www.nhk.or.jp/prog/img/458/g458.jpg',
        },
        'playlist_mincount': 3,
    }, {
        # one with letters in the id
        'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F300_06_3738470',
        'note': 'Expires on 2024-03-31',
        'info_dict': {
            'id': 'F300_06_3738470',
            'ext': 'm4a',
            'title': '有島武郎「一房のぶどう」',
            '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）',
            'channel': 'NHKラジオ第1、NHK-FM',
            'timestamp': 1635757200,
            'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F300/img/corner/box_109_thumbnail.jpg',
            'release_date': '20161207',
            'series': 'らじる文庫 by ラジオ深夜便 ',
            'release_timestamp': 1481126700,
            'upload_date': '20211101',
        }
    }, {
        # news
        'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F261_01_3855109',
        'skip': 'Expires on 2023-04-17',
        'info_dict': {
            'id': 'F261_01_3855109',
            'ext': 'm4a',
            'channel': 'NHKラジオ第1',
            'timestamp': 1681635900,
            'release_date': '20230416',
            'series': 'NHKラジオニュース',
            'title': '午後６時のNHKニュース',
            'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
            'upload_date': '20230416',
            'release_timestamp': 1681635600,
        },
    }]

    def _extract_episode_info(self, headline, programme_id, series_meta):
        episode_id = f'{programme_id}_{headline["headline_id"]}'
        episode = traverse_obj(headline, ('file_list', 0, {dict}))

        return {
            **series_meta,
            'id': episode_id,
            'formats': self._extract_m3u8_formats(episode.get('file_name'), episode_id, fatal=False),
            'container': 'm4a_dash',  # force fixup, AAC-only HLS
            'was_live': True,
            'series': series_meta.get('title'),
            'thumbnail': url_or_none(headline.get('headline_image')) or series_meta.get('thumbnail'),
            **traverse_obj(episode, {
                'title': 'file_title',
                'description': 'file_title_sub',
                'timestamp': ('open_time', {unified_timestamp}),
                'release_timestamp': ('aa_vinfo4', {lambda x: x.split('_')[0]}, {unified_timestamp}),
            }),
        }

    def _real_extract(self, url):
        site_id, corner_id, headline_id = self._match_valid_url(url).group('site', 'corner', 'headline')
        programme_id = f'{site_id}_{corner_id}'

        if site_id == 'F261':
            json_url = 'https://www.nhk.or.jp/s-media/news/news-site/list/v1/all.json'
        else:
            json_url = f'https://www.nhk.or.jp/radioondemand/json/{site_id}/bangumi_{programme_id}.json'

        meta = self._download_json(json_url, programme_id)['main']

        series_meta = traverse_obj(meta, {
            'title': 'program_name',
            'channel': 'media_name',
            'thumbnail': (('thumbnail_c', 'thumbnail_p'), {url_or_none}),
        }, get_all=False)

        if headline_id:
            return self._extract_episode_info(
                traverse_obj(meta, (
                    'detail_list', lambda _, v: v['headline_id'] == headline_id), get_all=False),
                programme_id, series_meta)

        def entries():
            for headline in traverse_obj(meta, ('detail_list', ..., {dict})):
                yield self._extract_episode_info(headline, programme_id, series_meta)

        return self.playlist_result(
            entries(), programme_id, playlist_description=meta.get('site_detail'), **series_meta)


class NhkRadioNewsPageIE(InfoExtractor):
    _VALID_URL = r'https?://www\.nhk\.or\.jp/radionews/?(?:$|[?#])'
    _TESTS = [{
        # airs daily, on-the-hour most hours
        'url': 'https://www.nhk.or.jp/radionews/',
        'playlist_mincount': 5,
        'info_dict': {
            'id': 'F261_01',
            'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
            'description': 'md5:bf2c5b397e44bc7eb26de98d8f15d79d',
            'channel': 'NHKラジオ第1',
            'title': 'NHKラジオニュース',
        }
    }]

    def _real_extract(self, url):
        return self.url_result('https://www.nhk.or.jp/radio/ondemand/detail.html?p=F261_01', NhkRadiruIE)
Commit	Line	Data
77cc7c6e	1	import re
061d1cd9	2
298a120a	3	from .common import InfoExtractor
77cc7c6e LNO	4	from ..utils import (
	5	parse_duration,
	6	traverse_obj,
	7	unescapeHTML,
	8	unified_timestamp,
8f0be90e	9	urljoin,
8f0be90e	10	url_or_none
77cc7c6e	11	)
298a120a AN	12
298a120a AN	13
29f7c58a	14	class NhkBaseIE(InfoExtractor):
6d1b3489	15	_API_URL_TEMPLATE = 'https://nwapi.nhk.jp/nhkworld/%sod%slist/v7b/%s/%s/%s/all%s.json'
29f7c58a	16	_BASE_URL_REGEX = r'https?://www3\.nhk\.or\.jp/nhkworld/(?P<lang>[a-z]{2})/ondemand'
29f7c58a	17	_TYPE_REGEX = r'/(?P<type>video\|audio)/'
298a120a	18
29f7c58a	19	def _call_api(self, m_id, lang, is_video, is_episode, is_clip):
	20	return self._download_json(
	21	self._API_URL_TEMPLATE % (
	22	'v' if is_video else 'r',
	23	'clip' if is_clip else 'esd',
	24	'episode' if is_episode else 'program',
	25	m_id, lang, '/all' if is_video else ''),
	26	m_id, query={'apikey': 'EJfK8jdS57GqlupFgAfAAwr573q01y6k'})['data']['episodes'] or []
	27
	28	def _extract_episode_info(self, url, episode=None):
	29	fetch_episode = episode is None
5ad28e7f	30	lang, m_type, episode_id = NhkVodIE._match_valid_url(url).groups()
6d1b3489	31	if len(episode_id) == 7:
061d1cd9	32	episode_id = episode_id[:4] + '-' + episode_id[4:]
f9b373af	33
061d1cd9	34	is_video = m_type == 'video'
29f7c58a	35	if fetch_episode:
	36	episode = self._call_api(
	37	episode_id, lang, is_video, True, episode_id[:4] == '9999')[0]
061d1cd9	38	title = episode.get('sub_title_clean') or episode['sub_title']
45396dd2	39
061d1cd9 RA	40	def get_clean_field(key):
061d1cd9 RA	41	return episode.get(key + '_clean') or episode.get(key)
45396dd2	42
061d1cd9	43	series = get_clean_field('title')
45396dd2	44
061d1cd9 RA	45	thumbnails = []
	46	for s, w, h in [('', 640, 360), ('_l', 1280, 720)]:
	47	img_path = episode.get('image' + s)
	48	if not img_path:
	49	continue
	50	thumbnails.append({
	51	'id': '%dp' % h,
	52	'height': h,
	53	'width': w,
	54	'url': 'https://www3.nhk.or.jp' + img_path,
	55	})
298a120a	56
061d1cd9 RA	57	info = {
061d1cd9 RA	58	'id': episode_id + '-' + lang,
f9b373af	59	'title': '%s - %s' % (series, title) if series and title else title,
061d1cd9 RA	60	'description': get_clean_field('description'),
061d1cd9 RA	61	'thumbnails': thumbnails,
f9b373af S	62	'series': series,
	63	'episode': title,
	64	}
061d1cd9	65	if is_video:
29f7c58a	66	vod_id = episode['vod_id']
061d1cd9 RA	67	info.update({
061d1cd9 RA	68	'_type': 'url_transparent',
a373befa	69	'ie_key': 'Piksel',
29f7c58a	70	'url': 'https://player.piksel.com/v/refid/nhkworld/prefid/' + vod_id,
29f7c58a	71	'id': vod_id,
061d1cd9 RA	72	})
061d1cd9 RA	73	else:
29f7c58a	74	if fetch_episode:
	75	audio_path = episode['audio']['audio']
	76	info['formats'] = self._extract_m3u8_formats(
	77	'https://nhkworld-vh.akamaihd.net/i%s/master.m3u8' % audio_path,
	78	episode_id, 'm4a', entry_protocol='m3u8_native',
	79	m3u8_id='hls', fatal=False)
	80	for f in info['formats']:
	81	f['language'] = lang
	82	else:
	83	info.update({
	84	'_type': 'url_transparent',
	85	'ie_key': NhkVodIE.ie_key(),
	86	'url': url,
	87	})
061d1cd9	88	return info
29f7c58a	89
	90
	91	class NhkVodIE(NhkBaseIE):
6d1b3489	92	# the 7-character IDs can have alphabetic chars too: assume [a-z] rather than just [a-f], eg
6d1b3489	93	_VALID_URL = r'%s%s(?P<id>[0-9a-z]{7}\|[^/]+?-\d{8}-[0-9a-z]+)' % (NhkBaseIE._BASE_URL_REGEX, NhkBaseIE._TYPE_REGEX)
29f7c58a	94	# Content available only for a limited period of time. Visit
	95	# https://www3.nhk.or.jp/nhkworld/en/ondemand/ for working samples.
	96	_TESTS = [{
	97	# video clip
	98	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999011/',
	99	'md5': '7a90abcfe610ec22a6bfe15bd46b30ca',
	100	'info_dict': {
	101	'id': 'a95j5iza',
	102	'ext': 'mp4',
	103	'title': "Dining with the Chef - Chef Saito's Family recipe: MENCHI-KATSU",
	104	'description': 'md5:5aee4a9f9d81c26281862382103b0ea5',
	105	'timestamp': 1565965194,
	106	'upload_date': '20190816',
	107	},
	108	}, {
	109	# audio clip
	110	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/r_inventions-20201104-1/',
	111	'info_dict': {
	112	'id': 'r_inventions-20201104-1-en',
	113	'ext': 'm4a',
	114	'title': "Japan's Top Inventions - Miniature Video Cameras",
	115	'description': 'md5:07ea722bdbbb4936fdd360b6a480c25b',
	116	},
	117	'params': {
	118	# m3u8 download
	119	'skip_download': True,
	120	},
	121	}, {
	122	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2015173/',
	123	'only_matching': True,
	124	}, {
	125	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/plugin-20190404-1/',
	126	'only_matching': True,
	127	}, {
	128	'url': 'https://www3.nhk.or.jp/nhkworld/fr/ondemand/audio/plugin-20190404-1/',
	129	'only_matching': True,
	130	}, {
	131	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/j_art-20150903-1/',
	132	'only_matching': True,
6d1b3489	133	}, {
	134	# video, alphabetic character in ID #29670
	135	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999a34/',
	136	'only_matching': True,
	137	'info_dict': {
	138	'id': 'qfjay6cg',
	139	'ext': 'mp4',
	140	'title': 'DESIGN TALKS plus - Fishermen’s Finery',
	141	'description': 'md5:8a8f958aaafb0d7cb59d38de53f1e448',
	142	'thumbnail': r're:^https?:/(/[a-z0-9.-]+)+\.jpg\?w=1920&h=1080$',
	143	'upload_date': '20210615',
	144	'timestamp': 1623722008,
	145	}
29f7c58a	146	}]
	147
	148	def _real_extract(self, url):
	149	return self._extract_episode_info(url)
	150
	151
	152	class NhkVodProgramIE(NhkBaseIE):
	153	_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)
	154	_TESTS = [{
	155	# video program episodes
	156	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway',
	157	'info_dict': {
	158	'id': 'japanrailway',
	159	'title': 'Japan Railway Journal',
	160	},
	161	'playlist_mincount': 1,
	162	}, {
	163	# video program clips
	164	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway/?type=clip',
	165	'info_dict': {
	166	'id': 'japanrailway',
	167	'title': 'Japan Railway Journal',
	168	},
	169	'playlist_mincount': 5,
	170	}, {
	171	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/10yearshayaomiyazaki/',
	172	'only_matching': True,
	173	}, {
	174	# audio program
	175	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/audio/listener/',
	176	'only_matching': True,
	177	}]
	178
	179	def _real_extract(self, url):
5ad28e7f	180	lang, m_type, program_id, episode_type = self._match_valid_url(url).groups()
29f7c58a	181
	182	episodes = self._call_api(
	183	program_id, lang, m_type == 'video', False, episode_type == 'clip')
	184
	185	entries = []
	186	for episode in episodes:
	187	episode_path = episode.get('url')
	188	if not episode_path:
	189	continue
	190	entries.append(self._extract_episode_info(
	191	urljoin(url, episode_path), episode))
	192
	193	program_title = None
	194	if entries:
	195	program_title = entries[0].get('series')
	196
	197	return self.playlist_result(entries, program_id, program_title)
77cc7c6e LNO	198
	199
	200	class NhkForSchoolBangumiIE(InfoExtractor):
	201	_VALID_URL = r'https?://www2\.nhk\.or\.jp/school/movie/(?P<type>bangumi\|clip)\.cgi\?das_id=(?P<id>[a-zA-Z0-9_-]+)'
	202	_TESTS = [{
	203	'url': 'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id=D0005150191_00000',
	204	'info_dict': {
	205	'id': 'D0005150191_00003',
	206	'title': 'にているかな',
	207	'duration': 599.999,
	208	'timestamp': 1396414800,
	209
	210	'upload_date': '20140402',
	211	'ext': 'mp4',
	212
	213	'chapters': 'count:12'
	214	},
	215	'params': {
	216	# m3u8 download
	217	'skip_download': True,
	218	},
	219	}]
	220
	221	def _real_extract(self, url):
	222	program_type, video_id = self._match_valid_url(url).groups()
	223
	224	webpage = self._download_webpage(
	225	f'https://www2.nhk.or.jp/school/movie/{program_type}.cgi?das_id={video_id}', video_id)
	226
	227	# searches all variables
	228	base_values = {g.group(1): g.group(2) for g in re.finditer(r'var\s+([a-zA-Z_]+)\s=\s"([^"]+?)";', webpage)}
	229	# and programObj values too
	230	program_values = {g.group(1): g.group(3) for g in re.finditer(r'(?:program\|clip)Obj\.([a-zA-Z_]+)\s=\s(["\'])([^"]+?)\2;', webpage)}
	231	# extract all chapters
	232	chapter_durations = [parse_duration(g.group(1)) for g in re.finditer(r'chapterTime\.push\(\'([0-9:]+?)\'\);', webpage)]
	233	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)]
	234
	235	# this is how player_core.js is actually doing (!)
	236	version = base_values.get('r_version') or program_values.get('version')
	237	if version:
	238	video_id = f'{video_id.split("_")[0]}_{version}'
	239
	240	formats = self._extract_m3u8_formats(
	241	f'https://nhks-vh.akamaihd.net/i/das/{video_id[0:8]}/{video_id}_V_000.f4v/master.m3u8',
	242	video_id, ext='mp4', m3u8_id='hls')
77cc7c6e LNO	243
	244	duration = parse_duration(base_values.get('r_duration'))
	245
	246	chapters = None
	247	if chapter_durations and chapter_titles and len(chapter_durations) == len(chapter_titles):
	248	start_time = chapter_durations
	249	end_time = chapter_durations[1:] + [duration]
	250	chapters = [{
	251	'start_time': s,
	252	'end_time': e,
	253	'title': t,
	254	} for s, e, t in zip(start_time, end_time, chapter_titles)]
	255
	256	return {
	257	'id': video_id,
	258	'title': program_values.get('name'),
	259	'duration': parse_duration(base_values.get('r_duration')),
	260	'timestamp': unified_timestamp(base_values['r_upload']),
	261	'formats': formats,
	262	'chapters': chapters,
	263	}
	264
	265
	266	class NhkForSchoolSubjectIE(InfoExtractor):
	267	IE_DESC = 'Portal page for each school subjects, like Japanese (kokugo, 国語) or math (sansuu/suugaku or 算数・数学)'
	268	KNOWN_SUBJECTS = (
	269	'rika', 'syakai', 'kokugo',
	270	'sansuu', 'seikatsu', 'doutoku',
	271	'ongaku', 'taiiku', 'zukou',
	272	'gijutsu', 'katei', 'sougou',
	273	'eigo', 'tokkatsu',
	274	'tokushi', 'sonota',
	275	)
	276	_VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>%s)/?(?:[\?#].*)?$' % '\|'.join(re.escape(s) for s in KNOWN_SUBJECTS)
	277
	278	_TESTS = [{
	279	'url': 'https://www.nhk.or.jp/school/sougou/',
	280	'info_dict': {
	281	'id': 'sougou',
	282	'title': '総合的な学習の時間',
	283	},
	284	'playlist_mincount': 16,
	285	}, {
	286	'url': 'https://www.nhk.or.jp/school/rika/',
	287	'info_dict': {
	288	'id': 'rika',
	289	'title': '理科',
	290	},
	291	'playlist_mincount': 15,
	292	}]
	293
	294	def _real_extract(self, url):
	295	subject_id = self._match_id(url)
	296	webpage = self._download_webpage(url, subject_id)
	297
	298	return self.playlist_from_matches(
	299	re.finditer(rf'href="((?:https?://www\.nhk\.or\.jp)?/school/{re.escape(subject_id)}/[^/]+/)"', webpage),
	300	subject_id,
	301	self._html_search_regex(r'(?s)<span\s+class="subjectName">\s<img\s[^<]+>\s*([^<]+?)</span>', webpage, 'title', fatal=False),
	302	lambda g: urljoin(url, g.group(1)))
	303
	304
	305	class NhkForSchoolProgramListIE(InfoExtractor):
	306	_VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>(?:%s)/[a-zA-Z0-9_-]+)' % (
307	'\|'.join(re.escape(s) for s in NhkForSchoolSubjectIE.KNOWN_SUBJECTS)
308	)
309	_TESTS = [{
310	'url': 'https://www.nhk.or.jp/school/sougou/q/',
311	'info_dict': {
312	'id': 'sougou/q',
313	'title': 'Ｑ～こどものための哲学',
314	},
315	'playlist_mincount': 20,
316	}]
317
318	def _real_extract(self, url):
319	program_id = self._match_id(url)
320
321	webpage = self._download_webpage(f'https://www.nhk.or.jp/school/{program_id}/', program_id)
322
62b8dac4	323	title = (self._generic_title('', webpage)
04f3fd2c	324	or self._html_search_regex(r'<h3>([^<]+?)とは？\s*</h3>', webpage, 'title', fatal=False))
77cc7c6e LNO	325	title = re.sub(r'\s\\|\sNHK\s+for\s+School\s*$', '', title) if title else None
	326	description = self._html_search_regex(
	327	r'(?s)<div\s+class="programDetail\s">\s<p>[^<]+</p>',
	328	webpage, 'description', fatal=False, group=0)
	329
	330	bangumi_list = self._download_json(
	331	f'https://www.nhk.or.jp/school/{program_id}/meta/program.json', program_id)
	332	# they're always bangumi
	333	bangumis = [
	334	self.url_result(f'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id={x}')
	335	for x in traverse_obj(bangumi_list, ('part', ..., 'part-video-dasid')) or []]
	336
	337	return self.playlist_result(bangumis, program_id, title, description)
8f0be90e	338
	339
	340	class NhkRadiruIE(InfoExtractor):
	341	_GEO_COUNTRIES = ['JP']
	342	IE_DESC = 'NHK らじる (Radiru/Rajiru)'
	343	_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]+))?'
	344	_TESTS = [{
	345	'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=0449_01_3853544',
	346	'skip': 'Episode expired on 2023-04-16',
	347	'info_dict': {
	348	'channel': 'NHK-FM',
	349	'description': 'md5:94b08bdeadde81a97df4ec882acce3e9',
	350	'ext': 'm4a',
	351	'id': '0449_01_3853544',
	352	'series': 'ジャズ・トゥナイト',
	353	'thumbnail': 'https://www.nhk.or.jp/prog/img/449/g449.jpg',
	354	'timestamp': 1680969600,
	355	'title': 'ジャズ・トゥナイト　ＮＥＷジャズ特集',
	356	'upload_date': '20230408',
	357	'release_timestamp': 1680962400,
	358	'release_date': '20230408',
	359	'was_live': True,
	360	},
	361	}, {
	362	# playlist, airs every weekday so it should _hopefully_ be okay forever
	363	'url': 'https://www.nhk.or.jp/radio/ondemand/detail.html?p=0458_01',
	364	'info_dict': {
	365	'id': '0458_01',
	366	'title': 'ベストオブクラシック',
	367	'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。',
	368	'channel': 'NHK-FM',
	369	'thumbnail': 'https://www.nhk.or.jp/prog/img/458/g458.jpg',
	370	},
	371	'playlist_mincount': 3,
	372	}, {
	373	# one with letters in the id
	374	'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F300_06_3738470',
	375	'note': 'Expires on 2024-03-31',
	376	'info_dict': {
	377	'id': 'F300_06_3738470',
	378	'ext': 'm4a',
	379	'title': '有島武郎「一房のぶどう」',
	380	'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）',
	381	'channel': 'NHKラジオ第1、NHK-FM',
	382	'timestamp': 1635757200,
	383	'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F300/img/corner/box_109_thumbnail.jpg',
	384	'release_date': '20161207',
	385	'series': 'らじる文庫 by ラジオ深夜便 ',
	386	'release_timestamp': 1481126700,
	387	'upload_date': '20211101',
	388	}
	389	}, {
	390	# news
	391	'url': 'https://www.nhk.or.jp/radio/player/ondemand.html?p=F261_01_3855109',
	392	'skip': 'Expires on 2023-04-17',
	393	'info_dict': {
	394	'id': 'F261_01_3855109',
	395	'ext': 'm4a',
	396	'channel': 'NHKラジオ第1',
	397	'timestamp': 1681635900,
	398	'release_date': '20230416',
	399	'series': 'NHKラジオニュース',
	400	'title': '午後６時のNHKニュース',
	401	'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
402	'upload_date': '20230416',
403	'release_timestamp': 1681635600,
404	},
405	}]
406
407	def _extract_episode_info(self, headline, programme_id, series_meta):
408	episode_id = f'{programme_id}_{headline["headline_id"]}'
409	episode = traverse_obj(headline, ('file_list', 0, {dict}))
410
411	return {
412	**series_meta,
413	'id': episode_id,
414	'formats': self._extract_m3u8_formats(episode.get('file_name'), episode_id, fatal=False),
415	'container': 'm4a_dash', # force fixup, AAC-only HLS
416	'was_live': True,
417	'series': series_meta.get('title'),
418	'thumbnail': url_or_none(headline.get('headline_image')) or series_meta.get('thumbnail'),
419	**traverse_obj(episode, {
420	'title': 'file_title',
421	'description': 'file_title_sub',
422	'timestamp': ('open_time', {unified_timestamp}),
423	'release_timestamp': ('aa_vinfo4', {lambda x: x.split('_')[0]}, {unified_timestamp}),
424	}),
425	}
426
427	def _real_extract(self, url):
428	site_id, corner_id, headline_id = self._match_valid_url(url).group('site', 'corner', 'headline')
429	programme_id = f'{site_id}_{corner_id}'
430
431	if site_id == 'F261':
432	json_url = 'https://www.nhk.or.jp/s-media/news/news-site/list/v1/all.json'
433	else:
434	json_url = f'https://www.nhk.or.jp/radioondemand/json/{site_id}/bangumi_{programme_id}.json'
435
436	meta = self._download_json(json_url, programme_id)['main']
437
438	series_meta = traverse_obj(meta, {
439	'title': 'program_name',
440	'channel': 'media_name',
441	'thumbnail': (('thumbnail_c', 'thumbnail_p'), {url_or_none}),
442	}, get_all=False)
443
444	if headline_id:
445	return self._extract_episode_info(
446	traverse_obj(meta, (
447	'detail_list', lambda _, v: v['headline_id'] == headline_id), get_all=False),
448	programme_id, series_meta)
449
450	def entries():
451	for headline in traverse_obj(meta, ('detail_list', ..., {dict})):
452	yield self._extract_episode_info(headline, programme_id, series_meta)
453
454	return self.playlist_result(
455	entries(), programme_id, playlist_description=meta.get('site_detail'), **series_meta)
456
457
458	class NhkRadioNewsPageIE(InfoExtractor):
459	_VALID_URL = r'https?://www\.nhk\.or\.jp/radionews/?(?:$\|[?#])'
460	_TESTS = [{
461	# airs daily, on-the-hour most hours
462	'url': 'https://www.nhk.or.jp/radionews/',
463	'playlist_mincount': 5,
464	'info_dict': {
465	'id': 'F261_01',
466	'thumbnail': 'https://www.nhk.or.jp/radioondemand/json/F261/img/RADIONEWS_640.jpg',
467	'description': 'md5:bf2c5b397e44bc7eb26de98d8f15d79d',
468	'channel': 'NHKラジオ第1',
469	'title': 'NHKラジオニュース',
470	}
471	}]
472
473	def _real_extract(self, url):
474	return self.url_result('https://www.nhk.or.jp/radio/ondemand/detail.html?p=F261_01', NhkRadiruIE)