yt-dlp/yt_dlp/extractor/mdr.py

141 lines
5.0 KiB
Python
Raw Permalink Normal View History

import urllib.parse
2013-12-10 18:40:50 +01:00
from .common import InfoExtractor
2015-10-31 17:17:09 +01:00
from ..utils import (
determine_ext,
int_or_none,
2021-11-06 02:05:24 +01:00
join_nonempty,
2015-10-31 17:17:09 +01:00
parse_duration,
parse_iso8601,
2021-01-01 13:26:37 +01:00
url_or_none,
2015-10-31 17:17:09 +01:00
xpath_text,
)
2013-12-10 18:40:50 +01:00
2013-12-16 05:44:34 +01:00
2013-12-10 18:40:50 +01:00
class MDRIE(InfoExtractor):
IE_DESC = 'MDR.DE'
_VALID_URL = r'https?://(?:www\.)?mdr\.de/(?:.*)/[a-z-]+-?(?P<id>\d+)(?:_.+?)?\.html'
2014-11-23 20:41:03 +01:00
2021-01-01 13:26:37 +01:00
_GEO_COUNTRIES = ['DE']
2015-10-31 17:17:09 +01:00
_TESTS = [{
2016-01-10 16:17:47 +01:00
# MDR regularly deletes its videos
'url': 'http://www.mdr.de/fakt/video189002.html',
'only_matching': True,
2015-10-31 18:00:36 +01:00
}, {
2015-10-31 17:24:21 +01:00
# audio
'url': 'http://www.mdr.de/kultur/audio1312272_zc-15948bad_zs-86171fdd.html',
'md5': '64c4ee50f0a791deb9479cd7bbe9d2fa',
'info_dict': {
'id': '1312272',
'ext': 'mp3',
'title': 'Feuilleton vom 30. Oktober 2015',
'duration': 250,
'uploader': 'MITTELDEUTSCHER RUNDFUNK',
},
'skip': '404 not found',
}, {
# audio with alternative playerURL pattern
'url': 'http://www.mdr.de/kultur/videos-und-audios/audio-radio/operation-mindfuck-robert-wilson100.html',
'info_dict': {
'id': '100',
'ext': 'mp4',
'title': 'Feature: Operation Mindfuck - Robert Anton Wilson',
'duration': 3239,
'uploader': 'MITTELDEUTSCHER RUNDFUNK',
},
'skip': '404 not found',
2016-02-29 20:24:26 +01:00
}, {
'url': 'http://www.mdr.de/mediathek/mdr-videos/a/video-1334.html',
'only_matching': True,
2015-10-31 17:17:09 +01:00
}]
2013-12-10 18:40:50 +01:00
def _real_extract(self, url):
2015-10-31 17:17:09 +01:00
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
2013-12-10 18:40:50 +01:00
2015-10-31 17:17:09 +01:00
data_url = self._search_regex(
r'(?:dataURL|playerXml(?:["\'])?)\s*:\s*(["\'])(?P<url>.+?-avCustom\.xml)\1',
webpage, 'data url', group='url').replace(r'\/', '/')
2013-12-10 18:40:50 +01:00
2015-10-31 17:17:09 +01:00
doc = self._download_xml(
urllib.parse.urljoin(url, data_url), video_id)
2015-10-31 17:17:09 +01:00
2015-10-31 17:45:45 +01:00
title = xpath_text(doc, ['./title', './broadcast/broadcastName'], 'title', fatal=True)
2013-12-16 05:44:34 +01:00
2021-01-01 13:26:37 +01:00
type_ = xpath_text(doc, './type', default=None)
2013-12-16 05:44:34 +01:00
formats = []
2015-10-31 17:17:09 +01:00
processed_urls = []
for asset in doc.findall('./assets/asset'):
for source in (
2021-01-01 13:26:37 +01:00
'download',
2015-10-31 17:17:09 +01:00
'progressiveDownload',
'dynamicHttpStreamingRedirector',
'adaptiveHttpStreamingRedirector'):
url_el = asset.find(f'./{source}Url')
2015-10-31 17:17:09 +01:00
if url_el is None:
continue
2021-01-01 13:26:37 +01:00
video_url = url_or_none(url_el.text)
if not video_url or video_url in processed_urls:
2015-10-31 17:17:09 +01:00
continue
processed_urls.append(video_url)
2021-01-01 13:26:37 +01:00
ext = determine_ext(video_url)
2015-10-31 17:17:09 +01:00
if ext == 'm3u8':
2021-01-01 13:26:37 +01:00
formats.extend(self._extract_m3u8_formats(
2015-10-31 17:17:09 +01:00
video_url, video_id, 'mp4', entry_protocol='m3u8_native',
quality=1, m3u8_id='HLS', fatal=False))
2015-10-31 17:17:09 +01:00
elif ext == 'f4m':
2021-01-01 13:26:37 +01:00
formats.extend(self._extract_f4m_formats(
2015-10-31 17:17:09 +01:00
video_url + '?hdcore=3.7.0&plugin=aasp-3.7.0.39.44', video_id,
quality=1, f4m_id='HDS', fatal=False))
2015-10-31 17:17:09 +01:00
else:
media_type = xpath_text(asset, './mediaType', 'media type', default='MP4')
vbr = int_or_none(xpath_text(asset, './bitrateVideo', 'vbr'), 1000)
abr = int_or_none(xpath_text(asset, './bitrateAudio', 'abr'), 1000)
filesize = int_or_none(xpath_text(asset, './fileSize', 'file size'))
f = {
'url': video_url,
2021-11-06 02:05:24 +01:00
'format_id': join_nonempty(media_type, vbr or abr),
2015-10-31 17:17:09 +01:00
'filesize': filesize,
'abr': abr,
2021-01-01 13:26:37 +01:00
'vbr': vbr,
2015-10-31 17:17:09 +01:00
}
if vbr:
f.update({
2021-01-01 13:26:37 +01:00
'width': int_or_none(xpath_text(asset, './frameWidth', 'width')),
'height': int_or_none(xpath_text(asset, './frameHeight', 'height')),
2015-10-31 17:17:09 +01:00
})
2021-01-01 13:26:37 +01:00
if type_ == 'audio':
f['vcodec'] = 'none'
2015-10-31 17:17:09 +01:00
2021-01-01 13:26:37 +01:00
formats.append(f)
2015-10-31 19:01:08 +01:00
2015-10-31 17:17:09 +01:00
description = xpath_text(doc, './broadcast/broadcastDescription', 'description')
timestamp = parse_iso8601(
2015-10-31 17:45:45 +01:00
xpath_text(
doc, [
'./broadcast/broadcastDate',
'./broadcast/broadcastStartDate',
'./broadcast/broadcastEndDate'],
'timestamp', default=None))
2015-10-31 17:17:09 +01:00
duration = parse_duration(xpath_text(doc, './duration', 'duration'))
uploader = xpath_text(doc, './rights', 'uploader')
2013-12-16 05:44:34 +01:00
return {
'id': video_id,
'title': title,
2015-10-31 17:17:09 +01:00
'description': description,
'timestamp': timestamp,
'duration': duration,
'uploader': uploader,
2013-12-16 05:44:34 +01:00
'formats': formats,
}