Server : Apache/2.4.6 (CentOS) OpenSSL/1.0.2k-fips PHP/7.4.20 System : Linux st2.domain.com 3.10.0-1127.10.1.el7.x86_64 #1 SMP Wed Jun 3 14:28:03 UTC 2020 x86_64 User : apache ( 48) PHP Version : 7.4.20 Disable Function : NONE Directory : /usr/lib/python2.7/site-packages/youtube_dl/extractor/ |
from __future__ import unicode_literals
import re
from .common import InfoExtractor
from ..utils import extract_attributes
class TheSunIE(InfoExtractor):
_VALID_URL = r'https://(?:www\.)?thesun\.co\.uk/[^/]+/(?P<id>\d+)'
_TEST = {
'url': 'https://www.thesun.co.uk/tvandshowbiz/2261604/orlando-bloom-and-katy-perry-post-adorable-instagram-video-together-celebrating-thanksgiving-after-split-rumours/',
'info_dict': {
'id': '2261604',
'title': 'md5:cba22f48bad9218b64d5bbe0e16afddf',
},
'playlist_count': 2,
}
BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/%s/default_default/index.html?videoId=%s'
def _real_extract(self, url):
article_id = self._match_id(url)
webpage = self._download_webpage(url, article_id)
entries = []
for video in re.findall(
r'<video[^>]+data-video-id-pending=[^>]+>',
webpage):
attrs = extract_attributes(video)
video_id = attrs['data-video-id-pending']
account_id = attrs.get('data-account', '5067014667001')
entries.append(self.url_result(
self.BRIGHTCOVE_URL_TEMPLATE % (account_id, video_id),
'BrightcoveNew', video_id))
return self.playlist_result(
entries, article_id, self._og_search_title(webpage, fatal=False))