[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
)


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)
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,
	9	urljoin
	10	)
298a120a AN	11
298a120a AN	12
29f7c58a	13	class NhkBaseIE(InfoExtractor):
6d1b3489	14	_API_URL_TEMPLATE = 'https://nwapi.nhk.jp/nhkworld/%sod%slist/v7b/%s/%s/%s/all%s.json'
29f7c58a	15	_BASE_URL_REGEX = r'https?://www3\.nhk\.or\.jp/nhkworld/(?P<lang>[a-z]{2})/ondemand'
29f7c58a	16	_TYPE_REGEX = r'/(?P<type>video\|audio)/'
298a120a	17
29f7c58a	18	def _call_api(self, m_id, lang, is_video, is_episode, is_clip):
	19	return self._download_json(
	20	self._API_URL_TEMPLATE % (
	21	'v' if is_video else 'r',
	22	'clip' if is_clip else 'esd',
	23	'episode' if is_episode else 'program',
	24	m_id, lang, '/all' if is_video else ''),
	25	m_id, query={'apikey': 'EJfK8jdS57GqlupFgAfAAwr573q01y6k'})['data']['episodes'] or []
	26
	27	def _extract_episode_info(self, url, episode=None):
	28	fetch_episode = episode is None
5ad28e7f	29	lang, m_type, episode_id = NhkVodIE._match_valid_url(url).groups()
6d1b3489	30	if len(episode_id) == 7:
061d1cd9	31	episode_id = episode_id[:4] + '-' + episode_id[4:]
f9b373af	32
061d1cd9	33	is_video = m_type == 'video'
29f7c58a	34	if fetch_episode:
	35	episode = self._call_api(
	36	episode_id, lang, is_video, True, episode_id[:4] == '9999')[0]
061d1cd9	37	title = episode.get('sub_title_clean') or episode['sub_title']
45396dd2	38
061d1cd9 RA	39	def get_clean_field(key):
061d1cd9 RA	40	return episode.get(key + '_clean') or episode.get(key)
45396dd2	41
061d1cd9	42	series = get_clean_field('title')
45396dd2	43
061d1cd9 RA	44	thumbnails = []
	45	for s, w, h in [('', 640, 360), ('_l', 1280, 720)]:
	46	img_path = episode.get('image' + s)
	47	if not img_path:
	48	continue
	49	thumbnails.append({
	50	'id': '%dp' % h,
	51	'height': h,
	52	'width': w,
	53	'url': 'https://www3.nhk.or.jp' + img_path,
	54	})
298a120a	55
061d1cd9 RA	56	info = {
061d1cd9 RA	57	'id': episode_id + '-' + lang,
f9b373af	58	'title': '%s - %s' % (series, title) if series and title else title,
061d1cd9 RA	59	'description': get_clean_field('description'),
061d1cd9 RA	60	'thumbnails': thumbnails,
f9b373af S	61	'series': series,
	62	'episode': title,
	63	}
061d1cd9	64	if is_video:
29f7c58a	65	vod_id = episode['vod_id']
061d1cd9 RA	66	info.update({
061d1cd9 RA	67	'_type': 'url_transparent',
a373befa	68	'ie_key': 'Piksel',
29f7c58a	69	'url': 'https://player.piksel.com/v/refid/nhkworld/prefid/' + vod_id,
29f7c58a	70	'id': vod_id,
061d1cd9 RA	71	})
061d1cd9 RA	72	else:
29f7c58a	73	if fetch_episode:
	74	audio_path = episode['audio']['audio']
	75	info['formats'] = self._extract_m3u8_formats(
	76	'https://nhkworld-vh.akamaihd.net/i%s/master.m3u8' % audio_path,
	77	episode_id, 'm4a', entry_protocol='m3u8_native',
	78	m3u8_id='hls', fatal=False)
	79	for f in info['formats']:
	80	f['language'] = lang
	81	else:
	82	info.update({
	83	'_type': 'url_transparent',
	84	'ie_key': NhkVodIE.ie_key(),
	85	'url': url,
	86	})
061d1cd9	87	return info
29f7c58a	88
	89
	90	class NhkVodIE(NhkBaseIE):
6d1b3489	91	# the 7-character IDs can have alphabetic chars too: assume [a-z] rather than just [a-f], eg
6d1b3489	92	_VALID_URL = r'%s%s(?P<id>[0-9a-z]{7}\|[^/]+?-\d{8}-[0-9a-z]+)' % (NhkBaseIE._BASE_URL_REGEX, NhkBaseIE._TYPE_REGEX)
29f7c58a	93	# Content available only for a limited period of time. Visit
	94	# https://www3.nhk.or.jp/nhkworld/en/ondemand/ for working samples.
	95	_TESTS = [{
	96	# video clip
	97	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999011/',
	98	'md5': '7a90abcfe610ec22a6bfe15bd46b30ca',
	99	'info_dict': {
	100	'id': 'a95j5iza',
	101	'ext': 'mp4',
	102	'title': "Dining with the Chef - Chef Saito's Family recipe: MENCHI-KATSU",
	103	'description': 'md5:5aee4a9f9d81c26281862382103b0ea5',
	104	'timestamp': 1565965194,
	105	'upload_date': '20190816',
	106	},
	107	}, {
	108	# audio clip
	109	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/r_inventions-20201104-1/',
	110	'info_dict': {
	111	'id': 'r_inventions-20201104-1-en',
	112	'ext': 'm4a',
	113	'title': "Japan's Top Inventions - Miniature Video Cameras",
	114	'description': 'md5:07ea722bdbbb4936fdd360b6a480c25b',
	115	},
	116	'params': {
	117	# m3u8 download
	118	'skip_download': True,
	119	},
	120	}, {
	121	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2015173/',
	122	'only_matching': True,
	123	}, {
	124	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/plugin-20190404-1/',
	125	'only_matching': True,
	126	}, {
	127	'url': 'https://www3.nhk.or.jp/nhkworld/fr/ondemand/audio/plugin-20190404-1/',
	128	'only_matching': True,
	129	}, {
	130	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/j_art-20150903-1/',
	131	'only_matching': True,
6d1b3489	132	}, {
	133	# video, alphabetic character in ID #29670
	134	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999a34/',
	135	'only_matching': True,
	136	'info_dict': {
	137	'id': 'qfjay6cg',
	138	'ext': 'mp4',
	139	'title': 'DESIGN TALKS plus - Fishermen’s Finery',
	140	'description': 'md5:8a8f958aaafb0d7cb59d38de53f1e448',
	141	'thumbnail': r're:^https?:/(/[a-z0-9.-]+)+\.jpg\?w=1920&h=1080$',
	142	'upload_date': '20210615',
	143	'timestamp': 1623722008,
	144	}
29f7c58a	145	}]
	146
	147	def _real_extract(self, url):
	148	return self._extract_episode_info(url)
	149
	150
	151	class NhkVodProgramIE(NhkBaseIE):
	152	_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)
	153	_TESTS = [{
	154	# video program episodes
	155	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway',
	156	'info_dict': {
	157	'id': 'japanrailway',
	158	'title': 'Japan Railway Journal',
	159	},
	160	'playlist_mincount': 1,
	161	}, {
	162	# video program clips
	163	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway/?type=clip',
	164	'info_dict': {
	165	'id': 'japanrailway',
	166	'title': 'Japan Railway Journal',
	167	},
	168	'playlist_mincount': 5,
	169	}, {
	170	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/10yearshayaomiyazaki/',
	171	'only_matching': True,
	172	}, {
	173	# audio program
	174	'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/audio/listener/',
	175	'only_matching': True,
	176	}]
	177
	178	def _real_extract(self, url):
5ad28e7f	179	lang, m_type, program_id, episode_type = self._match_valid_url(url).groups()
29f7c58a	180
	181	episodes = self._call_api(
	182	program_id, lang, m_type == 'video', False, episode_type == 'clip')
	183
	184	entries = []
	185	for episode in episodes:
	186	episode_path = episode.get('url')
	187	if not episode_path:
	188	continue
	189	entries.append(self._extract_episode_info(
	190	urljoin(url, episode_path), episode))
	191
	192	program_title = None
	193	if entries:
	194	program_title = entries[0].get('series')
	195
	196	return self.playlist_result(entries, program_id, program_title)
77cc7c6e LNO	197
	198
	199	class NhkForSchoolBangumiIE(InfoExtractor):
	200	_VALID_URL = r'https?://www2\.nhk\.or\.jp/school/movie/(?P<type>bangumi\|clip)\.cgi\?das_id=(?P<id>[a-zA-Z0-9_-]+)'
	201	_TESTS = [{
	202	'url': 'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id=D0005150191_00000',
	203	'info_dict': {
	204	'id': 'D0005150191_00003',
	205	'title': 'にているかな',
	206	'duration': 599.999,
	207	'timestamp': 1396414800,
	208
	209	'upload_date': '20140402',
	210	'ext': 'mp4',
	211
	212	'chapters': 'count:12'
	213	},
	214	'params': {
	215	# m3u8 download
	216	'skip_download': True,
	217	},
	218	}]
	219
	220	def _real_extract(self, url):
	221	program_type, video_id = self._match_valid_url(url).groups()
	222
	223	webpage = self._download_webpage(
	224	f'https://www2.nhk.or.jp/school/movie/{program_type}.cgi?das_id={video_id}', video_id)
	225
	226	# searches all variables
	227	base_values = {g.group(1): g.group(2) for g in re.finditer(r'var\s+([a-zA-Z_]+)\s=\s"([^"]+?)";', webpage)}
	228	# and programObj values too
	229	program_values = {g.group(1): g.group(3) for g in re.finditer(r'(?:program\|clip)Obj\.([a-zA-Z_]+)\s=\s(["\'])([^"]+?)\2;', webpage)}
	230	# extract all chapters
	231	chapter_durations = [parse_duration(g.group(1)) for g in re.finditer(r'chapterTime\.push\(\'([0-9:]+?)\'\);', webpage)]
	232	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)]
	233
	234	# this is how player_core.js is actually doing (!)
	235	version = base_values.get('r_version') or program_values.get('version')
	236	if version:
	237	video_id = f'{video_id.split("_")[0]}_{version}'
	238
	239	formats = self._extract_m3u8_formats(
	240	f'https://nhks-vh.akamaihd.net/i/das/{video_id[0:8]}/{video_id}_V_000.f4v/master.m3u8',
	241	video_id, ext='mp4', m3u8_id='hls')
77cc7c6e LNO	242
	243	duration = parse_duration(base_values.get('r_duration'))
	244
	245	chapters = None
	246	if chapter_durations and chapter_titles and len(chapter_durations) == len(chapter_titles):
	247	start_time = chapter_durations
	248	end_time = chapter_durations[1:] + [duration]
	249	chapters = [{
	250	'start_time': s,
	251	'end_time': e,
	252	'title': t,
	253	} for s, e, t in zip(start_time, end_time, chapter_titles)]
	254
	255	return {
	256	'id': video_id,
	257	'title': program_values.get('name'),
	258	'duration': parse_duration(base_values.get('r_duration')),
	259	'timestamp': unified_timestamp(base_values['r_upload']),
	260	'formats': formats,
	261	'chapters': chapters,
	262	}
	263
	264
	265	class NhkForSchoolSubjectIE(InfoExtractor):
	266	IE_DESC = 'Portal page for each school subjects, like Japanese (kokugo, 国語) or math (sansuu/suugaku or 算数・数学)'
	267	KNOWN_SUBJECTS = (
	268	'rika', 'syakai', 'kokugo',
	269	'sansuu', 'seikatsu', 'doutoku',
	270	'ongaku', 'taiiku', 'zukou',
	271	'gijutsu', 'katei', 'sougou',
	272	'eigo', 'tokkatsu',
	273	'tokushi', 'sonota',
	274	)
	275	_VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>%s)/?(?:[\?#].*)?$' % '\|'.join(re.escape(s) for s in KNOWN_SUBJECTS)
	276
	277	_TESTS = [{
	278	'url': 'https://www.nhk.or.jp/school/sougou/',
	279	'info_dict': {
	280	'id': 'sougou',
	281	'title': '総合的な学習の時間',
	282	},
	283	'playlist_mincount': 16,
	284	}, {
	285	'url': 'https://www.nhk.or.jp/school/rika/',
	286	'info_dict': {
	287	'id': 'rika',
	288	'title': '理科',
	289	},
	290	'playlist_mincount': 15,
	291	}]
	292
	293	def _real_extract(self, url):
	294	subject_id = self._match_id(url)
	295	webpage = self._download_webpage(url, subject_id)
	296
	297	return self.playlist_from_matches(
	298	re.finditer(rf'href="((?:https?://www\.nhk\.or\.jp)?/school/{re.escape(subject_id)}/[^/]+/)"', webpage),
	299	subject_id,
	300	self._html_search_regex(r'(?s)<span\s+class="subjectName">\s<img\s[^<]+>\s*([^<]+?)</span>', webpage, 'title', fatal=False),
	301	lambda g: urljoin(url, g.group(1)))
	302
	303
	304	class NhkForSchoolProgramListIE(InfoExtractor):
	305	_VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P<id>(?:%s)/[a-zA-Z0-9_-]+)' % (
306	'\|'.join(re.escape(s) for s in NhkForSchoolSubjectIE.KNOWN_SUBJECTS)
307	)
308	_TESTS = [{
309	'url': 'https://www.nhk.or.jp/school/sougou/q/',
310	'info_dict': {
311	'id': 'sougou/q',
312	'title': 'Ｑ～こどものための哲学',
313	},
314	'playlist_mincount': 20,
315	}]
316
317	def _real_extract(self, url):
318	program_id = self._match_id(url)
319
320	webpage = self._download_webpage(f'https://www.nhk.or.jp/school/{program_id}/', program_id)
321
62b8dac4	322	title = (self._generic_title('', webpage)
04f3fd2c	323	or self._html_search_regex(r'<h3>([^<]+?)とは？\s*</h3>', webpage, 'title', fatal=False))
77cc7c6e LNO	324	title = re.sub(r'\s\\|\sNHK\s+for\s+School\s*$', '', title) if title else None
	325	description = self._html_search_regex(
	326	r'(?s)<div\s+class="programDetail\s">\s<p>[^<]+</p>',
	327	webpage, 'description', fatal=False, group=0)
	328
	329	bangumi_list = self._download_json(
	330	f'https://www.nhk.or.jp/school/{program_id}/meta/program.json', program_id)
	331	# they're always bangumi
	332	bangumis = [
	333	self.url_result(f'https://www2.nhk.or.jp/school/movie/bangumi.cgi?das_id={x}')
	334	for x in traverse_obj(bangumi_list, ('part', ..., 'part-video-dasid')) or []]
	335
	336	return self.playlist_result(bangumis, program_id, title, description)