From 77cc7c6e6093c7b16a96870098edec70f943c62c Mon Sep 17 00:00:00 2001 From: "Lesmiscore (Naoya Ozaki)" Date: Wed, 23 Feb 2022 01:15:08 +0900 Subject: [PATCH] [nhk] Add support for NHK for School (#2850) Authored by: Lesmiscore --- yt_dlp/extractor/common.py | 4 + yt_dlp/extractor/extractors.py | 3 + yt_dlp/extractor/nhk.py | 149 ++++++++++++++++++++++++++++++++- 3 files changed, 155 insertions(+), 1 deletion(-) diff --git a/yt_dlp/extractor/common.py b/yt_dlp/extractor/common.py index 04d4c0733a..d8bb211370 100644 --- a/yt_dlp/extractor/common.py +++ b/yt_dlp/extractor/common.py @@ -1307,6 +1307,10 @@ def _og_search_video_url(self, html, name='video url', secure=True, **kargs): def _og_search_url(self, html, **kargs): return self._og_search_property('url', html, **kargs) + def _html_extract_title(self, html, name, **kwargs): + return self._html_search_regex( + r'(?s)(.*?)', html, name, **kwargs) + def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs): name = variadic(name) if display_name is None: diff --git a/yt_dlp/extractor/extractors.py b/yt_dlp/extractor/extractors.py index 15bc74915c..7d4262acf1 100644 --- a/yt_dlp/extractor/extractors.py +++ b/yt_dlp/extractor/extractors.py @@ -994,6 +994,9 @@ from .nhk import ( NhkVodIE, NhkVodProgramIE, + NhkForSchoolBangumiIE, + NhkForSchoolSubjectIE, + NhkForSchoolProgramListIE, ) from .nhl import NHLIE from .nick import ( diff --git a/yt_dlp/extractor/nhk.py b/yt_dlp/extractor/nhk.py index 4998fed831..626c6379b3 100644 --- a/yt_dlp/extractor/nhk.py +++ b/yt_dlp/extractor/nhk.py @@ -1,8 +1,15 @@ from __future__ import unicode_literals +import re from .common import InfoExtractor -from ..utils import urljoin +from ..utils import ( + parse_duration, + traverse_obj, + unescapeHTML, + unified_timestamp, + urljoin +) class NhkBaseIE(InfoExtractor): @@ -176,3 +183,143 @@ def _real_extract(self, url): 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/(?Pbangumi|clip)\.cgi\?das_id=(?P[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'
(scene\s*\d+)?([^<]+?)
', 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') + self._sort_formats(formats) + + 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%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)\s*\s*([^<]+?)', webpage, 'title', fatal=False), + lambda g: urljoin(url, g.group(1))) + + +class NhkForSchoolProgramListIE(InfoExtractor): + _VALID_URL = r'https?://www\.nhk\.or\.jp/school/(?P(?:%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._og_search_title(webpage, fatal=False) or self._html_extract_title(webpage, fatal=False) or self._html_search_regex(r'

([^<]+?)とは？\s*

', 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)\s*

[^<]+

', + 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)