Update InfoExtractors.py

pull/8/head
kkalpakloglou 12 years ago committed by Filippo Valsorda
parent 95506f1235
commit 43113d92cc

@ -4160,6 +4160,46 @@ class SpiegelIE(InfoExtractor):
}
return [info]
class liveleakIE(InfoExtractor):
_VALID_URL = r'^(?:http?://)?(?:\w+\.)?liveleak\.com/view\?(?:.*?)i=(?P<video_id>\d+)(?:.*)'
IE_NAME = u'liveleak'
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
if mobj is None:
self._downloader.trouble(u'ERROR: invalid URL: %s' % url)
return
video_id = mobj.group(1)
if video_id.endswith('/index.html'):
video_id = video_id[:-len('/index.html')]
webpage = self._download_webpage(url, video_id)
video_url = u'http://edge.liveleak.com/80281E/u/u/ll2_player_files/mp55/player.swf?config=http://www.liveleak.com/player?a=config%26item_token=' + video_id
m = re.search(r'<meta property="og:title" content="(?P<title>.*?)"', webpage)
if not m:
self._downloader.trouble(u'Cannot find video title')
title = unescapeHTML(m.group('title'))
m = re.search(r'<meta property="og:description" content="(?P<desc>.*?)"', webpage)
if m:
desc = unescapeHTML(m.group('desc'))
else:
desc = None
info = {
'id': video_id,
'url': video_url,
'ext': 'mp4',
'title': title,
'description': desc
}
return [info]
def gen_extractors():
""" Return a list of an instance of every supported extractor.
@ -4210,7 +4250,6 @@ def gen_extractors():
TEDIE(),
MySpassIE(),
SpiegelIE(),
liveleakIE(),
GenericIE()
]

Loading…
Cancel
Save