From 73f67430f0cdf9dbcfcd34c7bc29cea20b64b22f Mon Sep 17 00:00:00 2001 From: Simon Sawicki Date: Fri, 15 Sep 2023 22:39:25 +0200 Subject: [PATCH 1/2] Extend bpb extractor --- yt_dlp/extractor/bpb.py | 172 +++++++++++++++++++++++++++++++++------- yt_dlp/utils/_utils.py | 1 + 2 files changed, 143 insertions(+), 30 deletions(-) diff --git a/yt_dlp/extractor/bpb.py b/yt_dlp/extractor/bpb.py index f28e581b87b..77e6dd9c26c 100644 --- a/yt_dlp/extractor/bpb.py +++ b/yt_dlp/extractor/bpb.py @@ -1,56 +1,168 @@ +import functools import re from .common import InfoExtractor from ..utils import ( + clean_html, + extract_attributes, + get_element_text_and_html_by_tag, + get_elements_by_class, + join_nonempty, js_to_json, - determine_ext, + mimetype2ext, + unified_strdate, + url_or_none, + urljoin, + variadic, ) +from ..utils.traversal import traverse_obj + + +def html_get_element(tag=None, cls=None): + assert tag or cls, 'One of tag or class is required' + + if cls: + func = functools.partial(get_elements_by_class, cls, tag=tag) + else: + func = functools.partial(get_element_text_and_html_by_tag, tag) + + def html_get_element_wrapper(html): + return variadic(func(html))[0] + + return html_get_element_wrapper class BpbIE(InfoExtractor): IE_DESC = 'Bundeszentrale für politische Bildung' - _VALID_URL = r'https?://(?:www\.)?bpb\.de/mediathek/(?P[0-9]+)/' + _VALID_URL = r'https?://(?:www\.|m\.)?bpb\.de/(?:[^/?#]+/)*(?P\d+)(?:[/?#]|$)' - _TEST = { + _TESTS = [{ 'url': 'http://www.bpb.de/mediathek/297/joachim-gauck-zu-1989-und-die-erinnerung-an-die-ddr', - 'md5': 'c4f84c8a8044ca9ff68bb8441d300b3f', 'info_dict': { 'id': '297', 'ext': 'mp4', + 'creator': 'Kooperative Berlin', + 'description': 'md5:f4f75885ba009d3e2b156247a8941ce6', + 'release_date': '20160115', + 'series': 'Interview auf dem Geschichtsforum 1989 | 2009', + 'tags': ['Friedliche Revolution', 'Erinnerungskultur', 'Vergangenheitspolitik', 'DDR 1949 - 1990', 'Freiheitsrecht', 'BStU', 'Deutschland'], + 'thumbnail': 'https://www.bpb.de/cache/images/7/297_teaser_16x9_1240.jpg?8839D', 'title': 'Joachim Gauck zu 1989 und die Erinnerung an die DDR', - 'description': 'Joachim Gauck, erster Beauftragter für die Stasi-Unterlagen, spricht auf dem Geschichtsforum über die friedliche Revolution 1989 und eine "gewisse Traurigkeit" im Umgang mit der DDR-Vergangenheit.' + 'uploader': 'Bundeszentrale für politische Bildung', + }, + }, { + 'url': 'https://www.bpb.de/mediathek/video/522184/krieg-flucht-und-falschmeldungen-wirstattdesinformation-2/', + 'info_dict': { + 'id': '522184', + 'ext': 'mp4', + 'creator': 'Institute for Strategic Dialogue Germany gGmbH (ISD)', + 'description': 'md5:f83c795ff8f825a69456a9e51fc15903', + 'release_date': '20230621', + 'tags': ['Desinformation', 'Ukraine', 'Russland', 'Geflüchtete'], + 'thumbnail': 'https://www.bpb.de/cache/images/4/522184_teaser_16x9_1240.png?EABFB', + 'title': 'md5:9b01ccdbf58dbf9e5c9f6e771a803b1c', + 'uploader': 'Bundeszentrale für politische Bildung', + }, + }, { + 'url': 'https://www.bpb.de/lernen/bewegtbild-und-politische-bildung/webvideo/518789/krieg-flucht-und-falschmeldungen-wirstattdesinformation-1/', + 'info_dict': { + 'id': '518789', + 'ext': 'mp4', + 'creator': 'Institute for Strategic Dialogue Germany gGmbH (ISD)', + 'description': 'md5:85228aed433e84ff0ff9bc582abd4ea8', + 'release_date': '20230302', + 'tags': ['Desinformation', 'Ukraine', 'Russland', 'Geflüchtete'], + 'thumbnail': 'https://www.bpb.de/cache/images/9/518789_teaser_16x9_1240.jpeg?56D0D', + 'title': 'md5:3e956f264bb501f6383f10495a401da4', + 'uploader': 'Bundeszentrale für politische Bildung', + }, + }, { + 'url': 'https://www.bpb.de/mediathek/podcasts/apuz-podcast/539727/apuz-20-china/', + 'only_matching': True, + }, { + 'url': 'https://www.bpb.de/mediathek/audio/315813/folge-1-eine-einfuehrung/', + 'info_dict': { + 'id': '315813', + 'ext': 'mp3', + 'creator': 'Axel Schröder', + 'description': 'md5:eda9d1af34e5912efef5baf54fba4427', + 'release_date': '20200921', + 'series': 'Auf Endlagersuche. Der deutsche Weg zu einem sicheren Atommülllager', + 'tags': ['Atomenergie', 'Endlager', 'hoch-radioaktiver Abfall', 'Endlagersuche', 'Atommüll', 'Atomendlager', 'Gorleben', 'Deutschland'], + 'thumbnail': 'https://www.bpb.de/cache/images/3/315813_teaser_16x9_1240.png?92A94', + 'title': 'Folge 1: Eine Einführung', + 'uploader': 'Bundeszentrale für politische Bildung', + }, + }, { + 'url': 'https://www.bpb.de/517806/die-weltanschauung-der-neuen-rechten/', + 'info_dict': { + 'id': '517806', + 'ext': 'mp3', + 'creator': 'Bundeszentrale für politische Bildung', + 'description': 'md5:594689600e919912aade0b2871cc3fed', + 'release_date': '20230127', + 'series': 'Vorträge des Fachtags "Modernisierer. Grenzgänger. Anstifter. Sechs Jahrzehnte \'Neue Rechte\'"', + 'tags': ['Rechtsextremismus', 'Konservatismus', 'Konservativismus', 'neue Rechte', 'Rechtspopulismus', 'Schnellroda', 'Deutschland'], + 'thumbnail': 'https://www.bpb.de/cache/images/6/517806_teaser_16x9_1240.png?7A7A0', + 'title': 'Die Weltanschauung der "Neuen Rechten"', + 'uploader': 'Bundeszentrale für politische Bildung', + }, + }, { + 'url': 'https://www.bpb.de/mediathek/reihen/zahlen-und-fakten-soziale-situation-filme/520153/zahlen-und-fakten-die-soziale-situation-in-deutschland-migration/', + 'only_matching': True, + }] + + BASE_URL = 'https://www.bpb.de' + _TITLE_RE = re.compile('(?P.*?)<[^>]+>(?P<series>[^<]*)') + + def _parse_vue_attributes(self, name, string, video_id): + attributes = extract_attributes(self._search_regex(rf'(<{name}(?:"[^"]*?"|[^>])*>)', string, name)) + + for key, value in attributes.items(): + if key.startswith(':'): + attributes[key] = self._parse_json(value, video_id, transform_source=js_to_json, fatal=False) + + return attributes + + @staticmethod + def _process_source(source): + url = url_or_none(source['src']) + if not url: + return None + + source_type = source.get('type', '') + extension = mimetype2ext(source_type) + note = url.rpartition('.')[0].rpartition('_')[2] if source_type.startswith('video') else None + return { + 'url': url, + 'ext': extension, + 'quality': 10 if note == 'high' else 0, + 'format_note': note, + 'format_id': join_nonempty(extension, note), } - } def _real_extract(self, url): video_id = self._match_id(url) webpage = self._download_webpage(url, video_id) - title = self._html_search_regex( - r'<h2 class="white">(.*?)</h2>', webpage, 'title') - video_info_dicts = re.findall( - r"({\s*src\s*:\s*'https?://film\.bpb\.de/[^}]+})", webpage) - - formats = [] - for video_info in video_info_dicts: - video_info = self._parse_json( - video_info, video_id, transform_source=js_to_json, fatal=False) - if not video_info: - continue - video_url = video_info.get('src') - if not video_url: - continue - quality = 'high' if '_high' in video_url else 'low' - formats.append({ - 'url': video_url, - 'quality': 10 if quality == 'high' else 0, - 'format_note': quality, - 'format_id': '%s-%s' % (quality, determine_ext(video_url)), - }) + title_result = traverse_obj(webpage, ({html_get_element(cls='opening-header__title')}, {self._TITLE_RE.match})) + json_lds = list(self._yield_json_ld(webpage, video_id, fatal=False)) return { 'id': video_id, - 'formats': formats, - 'title': title, - 'description': self._og_search_description(webpage), + 'title': traverse_obj(title_result, ('title', {str.strip})) or None, + # This metadata could be interpreted otherwise, but it fits "series" the most + 'series': traverse_obj(title_result, ('series', {str.strip})) or None, + 'description': join_nonempty(*traverse_obj(webpage, [( + {html_get_element(cls='opening-intro')}, + [{html_get_element(tag='bpb-accordion-item')}, {html_get_element(cls='text-content')}], + ), {clean_html}]), delim='\n\n') or None, + 'creator': self._html_search_meta('author', webpage), + 'uploader': self._html_search_meta('publisher', webpage), + 'release_date': unified_strdate(self._html_search_meta('date', webpage)), + 'tags': traverse_obj(json_lds, (..., 'keywords', {lambda x: x.split(',')}, ...)), + **traverse_obj(self._parse_vue_attributes('bpb-player', webpage, video_id), { + 'formats': (':sources', ..., {self._process_source}), + 'thumbnail': ('poster', {lambda x: urljoin(self.BASE_URL, x)}), + }), } diff --git a/yt_dlp/utils/_utils.py b/yt_dlp/utils/_utils.py index f5552ce8021..180bec245a3 100644 --- a/yt_dlp/utils/_utils.py +++ b/yt_dlp/utils/_utils.py @@ -2847,6 +2847,7 @@ def mimetype2ext(mt, default=NO_DEFAULT): 'quicktime': 'mov', 'webm': 'webm', 'vp9': 'vp9', + 'video/ogg': 'ogv', 'x-flv': 'flv', 'x-m4v': 'm4v', 'x-matroska': 'mkv', From 3b55ef42946af7c85935f3a7a6cd9087f0881010 Mon Sep 17 00:00:00 2001 From: Simon Sawicki <contact@grub4k.xyz> Date: Sat, 16 Sep 2023 00:18:59 +0200 Subject: [PATCH 2/2] Apply suggestions from code review --- yt_dlp/extractor/bpb.py | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/yt_dlp/extractor/bpb.py b/yt_dlp/extractor/bpb.py index 77e6dd9c26c..7fe08994492 100644 --- a/yt_dlp/extractor/bpb.py +++ b/yt_dlp/extractor/bpb.py @@ -112,8 +112,7 @@ class BpbIE(InfoExtractor): 'only_matching': True, }] - BASE_URL = 'https://www.bpb.de' - _TITLE_RE = re.compile('(?P<title>.*?)<[^>]+>(?P<series>[^<]*)') + _TITLE_RE = re.compile('(?P<title>[^<]*)<[^>]+>(?P<series>[^<]*)') def _parse_vue_attributes(self, name, string, video_id): attributes = extract_attributes(self._search_regex(rf'(<{name}(?:"[^"]*?"|[^>])*>)', string, name)) @@ -132,10 +131,13 @@ def _process_source(source): source_type = source.get('type', '') extension = mimetype2ext(source_type) - note = url.rpartition('.')[0].rpartition('_')[2] if source_type.startswith('video') else None + is_video = source_type.startswith('video') + note = url.rpartition('.')[0].rpartition('_')[2] if is_video else None + return { 'url': url, 'ext': extension, + 'vcodec': None if is_video else 'none', 'quality': 10 if note == 'high' else 0, 'format_note': note, 'format_id': join_nonempty(extension, note), @@ -163,6 +165,6 @@ def _real_extract(self, url): 'tags': traverse_obj(json_lds, (..., 'keywords', {lambda x: x.split(',')}, ...)), **traverse_obj(self._parse_vue_attributes('bpb-player', webpage, video_id), { 'formats': (':sources', ..., {self._process_source}), - 'thumbnail': ('poster', {lambda x: urljoin(self.BASE_URL, x)}), + 'thumbnail': ('poster', {lambda x: urljoin(url, x)}), }), }