[LocalNews8] add extractor (Closes #9200)
parent
6c114b1210
commit
6756602be6
2 changed files with 30 additions and 0 deletions
@ -0,0 +1,29 @@ |
||||
# coding: utf-8 |
||||
from __future__ import unicode_literals |
||||
|
||||
from .common import InfoExtractor |
||||
|
||||
|
||||
class LocalNews8IE(InfoExtractor): |
||||
_VALID_URL = r'https?://(?:www\.)?localnews8\.com/.+?/(?P<id>[0-9]+)' |
||||
_TEST = { |
||||
'url': 'http://www.localnews8.com/news/rexburg-business-turns-carbon-fiber-scraps-into-wedding-rings/35183304', |
||||
'md5': '477bdb188f177788c65db27ecb56649b', |
||||
'info_dict': { |
||||
'id': '35183304', |
||||
'ext': 'mp4', |
||||
'title': 'Rexburg business turns carbon fiber scraps into wedding ring', |
||||
'description': 'The process was first invented by Lamborghini and less than a dozen companies around the world use it.', |
||||
'duration': '153', |
||||
'timestamp': '1441844822', |
||||
'uploader_id': 'api', |
||||
}} |
||||
|
||||
def _real_extract(self, url): |
||||
video_id = self._match_id(url) |
||||
webpage = self._download_webpage(url, video_id) |
||||
|
||||
partner_id = self._search_regex(r'partnerId\s*:\s*"(\d+)"', webpage, video_id) |
||||
kaltura_id = self._search_regex(r'var\s+videoIdString\s*=\s*"kaltura:(.+)";', webpage, video_id) |
||||
|
||||
return self.url_result('kaltura:%s:%s' % (partner_id, kaltura_id), 'Kaltura') |
Loading…
Reference in new issue