1
1
mirror of https://github.com/ytdl-org/youtube-dl synced 2024-12-21 07:07:47 +01:00
youtube-dl/youtube_dl/extractor/cbssports.py

39 lines
1.5 KiB
Python
Raw Normal View History

from __future__ import unicode_literals
from .cbs import CBSBaseIE
class CBSSportsIE(CBSBaseIE):
_VALID_URL = r'https?://(?:www\.)?cbssports\.com/[^/]+/(?:video|news)/(?P<id>[^/?#&]+)'
_TESTS = [{
'url': 'https://www.cbssports.com/nba/video/donovan-mitchell-flashes-star-potential-in-game-2-victory-over-thunder/',
'info_dict': {
'id': '1214315075735',
'ext': 'mp4',
'title': 'Donovan Mitchell flashes star potential in Game 2 victory over Thunder',
'description': 'md5:df6f48622612c2d6bd2e295ddef58def',
'timestamp': 1524111457,
'upload_date': '20180419',
'uploader': 'CBSI-NEW',
},
'params': {
# m3u8 download
'skip_download': True,
}
}, {
'url': 'https://www.cbssports.com/nba/news/nba-playoffs-2018-watch-76ers-vs-heat-game-3-series-schedule-tv-channel-online-stream/',
'only_matching': True,
}]
def _extract_video_info(self, filter_query, video_id):
return self._extract_feed_info('dJ5BDC', 'VxxJg8Ymh8sE', filter_query, video_id)
def _real_extract(self, url):
display_id = self._match_id(url)
webpage = self._download_webpage(url, display_id)
2018-04-21 23:57:22 +02:00
video_id = self._search_regex(
[r'(?:=|%26)pcid%3D(\d+)', r'embedVideo(?:Container)?_(\d+)'],
webpage, 'video id')
return self._extract_video_info('byId=%s' % video_id, video_id)