From 0f6e60bb5722f03c6b64712f70aaf9b0b6915795 Mon Sep 17 00:00:00 2001 From: u-spec-png <54671367+u-spec-png@users.noreply.github.com> Date: Thu, 21 Oct 2021 23:39:50 +0000 Subject: [PATCH] [tagesschau] Fix extractor (#1227) Closes #1124 Authored by: u-spec-png --- yt_dlp/extractor/extractors.py | 5 +- yt_dlp/extractor/tagesschau.py | 277 +++++++-------------------------- 2 files changed, 61 insertions(+), 221 deletions(-) diff --git a/yt_dlp/extractor/extractors.py b/yt_dlp/extractor/extractors.py index f4f817fcb..8ea7d2ed8 100644 --- a/yt_dlp/extractor/extractors.py +++ b/yt_dlp/extractor/extractors.py @@ -1387,10 +1387,7 @@ from .swrmediathek import SWRMediathekIE from .syfy import SyfyIE from .sztvhu import SztvHuIE -from .tagesschau import ( - TagesschauPlayerIE, - TagesschauIE, -) +from .tagesschau import TagesschauIE from .tass import TassIE from .tbs import TBSIE from .tdslifeway import TDSLifewayIE diff --git a/yt_dlp/extractor/tagesschau.py b/yt_dlp/extractor/tagesschau.py index 25c200455..6e03d0a7d 100644 --- a/yt_dlp/extractor/tagesschau.py +++ b/yt_dlp/extractor/tagesschau.py @@ -5,177 +5,63 @@ from .common import InfoExtractor from ..utils import ( - determine_ext, js_to_json, - parse_iso8601, - parse_filesize, + extract_attributes, + try_get, + int_or_none, ) -class TagesschauPlayerIE(InfoExtractor): - IE_NAME = 'tagesschau:player' - _VALID_URL = r'https?://(?:www\.)?tagesschau\.de/multimedia/(?Paudio|video)/(?P=kind)-(?P\d+)~player(?:_[^/?#&]+)?\.html' - - _TESTS = [{ - 'url': 'http://www.tagesschau.de/multimedia/video/video-179517~player.html', - 'md5': '8d09548d5c15debad38bee3a4d15ca21', - 'info_dict': { - 'id': '179517', - 'ext': 'mp4', - 'title': 'Marie Kristin Boese, ARD Berlin, über den zukünftigen Kurs der AfD', - 'thumbnail': r're:^https?:.*\.jpg$', - 'formats': 'mincount:6', - }, - }, { - 'url': 'https://www.tagesschau.de/multimedia/audio/audio-29417~player.html', - 'md5': '76e6eec6ebd40740671cf0a2c88617e5', - 'info_dict': { - 'id': '29417', - 'ext': 'mp3', - 'title': 'Trabi - Bye, bye Rennpappe', - 'thumbnail': r're:^https?:.*\.jpg$', - 'formats': 'mincount:2', - }, - }, { - 'url': 'http://www.tagesschau.de/multimedia/audio/audio-29417~player_autoplay-true.html', - 'only_matching': True, - }] - - _FORMATS = { - 'xs': {'quality': 0}, - 's': {'width': 320, 'height': 180, 'quality': 1}, - 'm': {'width': 512, 'height': 288, 'quality': 2}, - 'l': {'width': 960, 'height': 540, 'quality': 3}, - 'xl': {'width': 1280, 'height': 720, 'quality': 4}, - 'xxl': {'quality': 5}, - } - - def _extract_via_api(self, kind, video_id): - info = self._download_json( - 'https://www.tagesschau.de/api/multimedia/{0}/{0}-{1}.json'.format(kind, video_id), - video_id) - title = info['headline'] - formats = [] - for media in info['mediadata']: - for format_id, format_url in media.items(): - if determine_ext(format_url) == 'm3u8': - formats.extend(self._extract_m3u8_formats( - format_url, video_id, 'mp4', - entry_protocol='m3u8_native', m3u8_id='hls')) - else: - formats.append({ - 'url': format_url, - 'format_id': format_id, - 'vcodec': 'none' if kind == 'audio' else None, - }) - self._sort_formats(formats) - timestamp = parse_iso8601(info.get('date')) - return { - 'id': video_id, - 'title': title, - 'timestamp': timestamp, - 'formats': formats, - } - - def _real_extract(self, url): - mobj = self._match_valid_url(url) - video_id = mobj.group('id') - - # kind = mobj.group('kind').lower() - # if kind == 'video': - # return self._extract_via_api(kind, video_id) - - # JSON api does not provide some audio formats (e.g. ogg) thus - # extracting audio via webpage - - webpage = self._download_webpage(url, video_id) - - title = self._og_search_title(webpage).strip() - formats = [] - - for media_json in re.findall(r'({src\s*:\s*["\']http[^}]+type\s*:[^}]+})', webpage): - media = self._parse_json(js_to_json(media_json), video_id, fatal=False) - if not media: - continue - src = media.get('src') - if not src: - return - quality = media.get('quality') - kind = media.get('type', '').split('/')[0] - ext = determine_ext(src) - f = { - 'url': src, - 'format_id': '%s_%s' % (quality, ext) if quality else ext, - 'ext': ext, - 'vcodec': 'none' if kind == 'audio' else None, - } - f.update(self._FORMATS.get(quality, {})) - formats.append(f) - - self._sort_formats(formats) - - thumbnail = self._og_search_thumbnail(webpage) - - return { - 'id': video_id, - 'title': title, - 'thumbnail': thumbnail, - 'formats': formats, - } - - class TagesschauIE(InfoExtractor): _VALID_URL = r'https?://(?:www\.)?tagesschau\.de/(?P[^/]+/(?:[^/]+/)*?(?P[^/#?]+?(?:-?[0-9]+)?))(?:~_?[^/#?]+?)?\.html' _TESTS = [{ 'url': 'http://www.tagesschau.de/multimedia/video/video-102143.html', - 'md5': 'f7c27a0eff3bfe8c7727e65f8fe1b1e6', + 'md5': '7a7287612fa881a1ae1d087df45c2fd6', 'info_dict': { - 'id': 'video-102143', + 'id': 'video-102143-1', 'ext': 'mp4', 'title': 'Regierungsumbildung in Athen: Neue Minister in Griechenland vereidigt', - 'description': '18.07.2015 20:10 Uhr', - 'thumbnail': r're:^https?:.*\.jpg$', }, }, { 'url': 'http://www.tagesschau.de/multimedia/sendung/ts-5727.html', 'md5': '3c54c1f6243d279b706bde660ceec633', 'info_dict': { - 'id': 'ts-5727', + 'id': 'ts-5727-1', 'ext': 'mp4', - 'title': 'Sendung: tagesschau \t04.12.2014 20:00 Uhr', - 'description': 'md5:695c01bfd98b7e313c501386327aea59', - 'thumbnail': r're:^https?:.*\.jpg$', + 'title': 'Ganze Sendung', }, }, { # exclusive audio 'url': 'http://www.tagesschau.de/multimedia/audio/audio-29417.html', - 'md5': '76e6eec6ebd40740671cf0a2c88617e5', + 'md5': '4cf22023c285f35e99c24d290ba58cc9', 'info_dict': { - 'id': 'audio-29417', + 'id': 'audio-29417-1', 'ext': 'mp3', - 'title': 'Trabi - Bye, bye Rennpappe', - 'description': 'md5:8687dda862cbbe2cfb2df09b56341317', - 'thumbnail': r're:^https?:.*\.jpg$', + 'title': 'Brasilianischer Präsident Bolsonaro unter Druck: Corona-Bericht wird vorgestellt', }, }, { - # audio in article 'url': 'http://www.tagesschau.de/inland/bnd-303.html', - 'md5': 'e0916c623e85fc1d2b26b78f299d3958', + 'md5': '12cfb212d9325b5ba0d52b625f1aa61c', 'info_dict': { - 'id': 'bnd-303', - 'ext': 'mp3', - 'title': 'Viele Baustellen für neuen BND-Chef', - 'description': 'md5:1e69a54be3e1255b2b07cdbce5bcd8b4', - 'thumbnail': r're:^https?:.*\.jpg$', + 'id': 'bnd-303-1', + 'ext': 'mp4', + 'title': 'SPD-Gruppenbild mit Bärbel Bas nach der Fraktionssitzung | dpa', }, }, { 'url': 'http://www.tagesschau.de/inland/afd-parteitag-135.html', 'info_dict': { 'id': 'afd-parteitag-135', - 'title': 'Möchtegern-Underdog mit Machtanspruch', + 'title': 'AfD', + }, + 'playlist_count': 20, + }, { + 'url': 'https://www.tagesschau.de/multimedia/audio/audio-29417~player.html', + 'info_dict': { + 'id': 'audio-29417-1', + 'ext': 'mp3', + 'title': 'Brasilianischer Präsident Bolsonaro unter Druck: Corona-Bericht wird vorgestellt', }, - 'playlist_count': 2, }, { 'url': 'http://www.tagesschau.de/multimedia/sendung/tsg-3771.html', 'only_matching': True, @@ -206,62 +92,6 @@ class TagesschauIE(InfoExtractor): 'only_matching': True, }] - @classmethod - def suitable(cls, url): - return False if TagesschauPlayerIE.suitable(url) else super(TagesschauIE, cls).suitable(url) - - def _extract_formats(self, download_text, media_kind): - links = re.finditer( - r'
(?P.+?)
', - download_text) - formats = [] - for l in links: - link_url = l.group('url') - if not link_url: - continue - format_id = self._search_regex( - r'.*/[^/.]+\.([^/]+)\.[^/.]+$', link_url, 'format ID', - default=determine_ext(link_url)) - format = { - 'format_id': format_id, - 'url': l.group('url'), - 'format_name': l.group('name'), - } - title = l.group('title') - if title: - if media_kind.lower() == 'video': - m = re.match( - r'''(?x) - Video:\s*(?P[a-zA-Z0-9/._-]+)\s*&\#10; - (?P[0-9]+)x(?P[0-9]+)px&\#10; - (?P[0-9]+)kbps&\#10; - Audio:\s*(?P[0-9]+)kbps,\s*(?P[A-Za-z\.0-9]+)&\#10; - Größe:\s*(?P[0-9.,]+\s+[a-zA-Z]*B)''', - title) - if m: - format.update({ - 'format_note': m.group('audio_desc'), - 'vcodec': m.group('vcodec'), - 'width': int(m.group('width')), - 'height': int(m.group('height')), - 'abr': int(m.group('abr')), - 'vbr': int(m.group('vbr')), - 'filesize_approx': parse_filesize(m.group('filesize_approx')), - }) - else: - m = re.match( - r'(?P.+?)-Format\s*:\s*(?P\d+)kbps\s*,\s*(?P.+)', - title) - if m: - format.update({ - 'format_note': '%s, %s' % (m.group('format'), m.group('note')), - 'vcodec': 'none', - 'abr': int(m.group('abr')), - }) - formats.append(format) - self._sort_formats(formats) - return formats - def _real_extract(self, url): mobj = self._match_valid_url(url) video_id = mobj.group('id') or mobj.group('path') @@ -271,34 +101,46 @@ def _real_extract(self, url): title = self._html_search_regex( r']*class="headline"[^>]*>(.+?)', - webpage, 'title', default=None) or self._og_search_title(webpage) + webpage, 'title', default=None) or self._og_search_title(webpage, fatal=False) - DOWNLOAD_REGEX = r'(?s)

Wir bieten dieses (?PVideo|Audio) in folgenden Formaten zum Download an:

\s*
(?P.*?)
\s*

' - - webpage_type = self._og_search_property('type', webpage, default=None) - if webpage_type == 'website': # Article - entries = [] - for num, (entry_title, media_kind, download_text) in enumerate(re.findall( - r'(?s)]+class="infotext"[^>]*>\s*(?:]+>)?\s*(.+?).*?

.*?%s' % DOWNLOAD_REGEX, - webpage), 1): + entries = [] + videos = re.findall(r']+>', webpage) + num = 0 + for video in videos: + video = extract_attributes(video).get('data-config') + if not video: + continue + video = self._parse_json(video, video_id, transform_source=js_to_json, fatal=False) + video_formats = try_get(video, lambda x: x['mc']['_mediaArray'][0]['_mediaStreamArray']) + if not video_formats: + continue + num += 1 + for video_format in video_formats: + media_url = video_format.get('_stream') or '' + formats = [] + if media_url.endswith('master.m3u8'): + formats = self._extract_m3u8_formats(media_url, video_id, 'mp4', m3u8_id='hls') + elif media_url.endswith('.hi.mp3') and media_url.startswith('https://download'): + formats = [{ + 'url': media_url, + 'vcodec': 'none', + }] + if not formats: + continue entries.append({ 'id': '%s-%d' % (display_id, num), - 'title': '%s' % entry_title, - 'formats': self._extract_formats(download_text, media_kind), + 'title': try_get(video, lambda x: x['mc']['_title']), + 'duration': int_or_none(try_get(video, lambda x: x['mc']['_duration'])), + 'formats': formats }) - if len(entries) > 1: - return self.playlist_result(entries, display_id, title) - formats = entries[0]['formats'] - else: # Assume single video - download_text = self._search_regex( - DOWNLOAD_REGEX, webpage, 'download links', group='links') - media_kind = self._search_regex( - DOWNLOAD_REGEX, webpage, 'media kind', default='Video', group='kind') - formats = self._extract_formats(download_text, media_kind) - thumbnail = self._og_search_thumbnail(webpage) - description = self._html_search_regex( - r'(?s)

(.*?)

', - webpage, 'description', default=None) + if len(entries) > 1: + return self.playlist_result(entries, display_id, title) + formats = entries[0]['formats'] + video_info = self._search_json_ld(webpage, video_id) + description = video_info.get('description') + thumbnail = self._og_search_thumbnail(webpage) or video_info.get('thumbnail') + timestamp = video_info.get('timestamp') + title = title or video_info.get('description') self._sort_formats(formats) @@ -307,5 +149,6 @@ def _real_extract(self, url): 'title': title, 'thumbnail': thumbnail, 'formats': formats, + 'timestamp': timestamp, 'description': description, }