mirror of
https://github.com/ytdl-org/youtube-dl.git
synced 2024-12-15 21:57:53 +01:00
[ku6] Improve (#3015)
This commit is contained in:
parent
05741e05d9
commit
90e0fd4bad
@ -1,30 +1,29 @@
|
||||
from __future__ import unicode_literals
|
||||
|
||||
import re
|
||||
|
||||
from .common import InfoExtractor
|
||||
|
||||
|
||||
class Ku6IE(InfoExtractor):
|
||||
_VALID_URL = r'http://v\.ku6\.com/show/(?P<id>[a-zA-Z0-9\-\_]+)(?:\.)*html'
|
||||
_TEST = {
|
||||
'url': 'http://v.ku6.com/show/JG-8yS14xzBr4bCn1pu0xw...html',
|
||||
'md5': '01203549b9efbb45f4b87d55bdea1ed1',
|
||||
'info_dict': {
|
||||
'id': 'JG-8yS14xzBr4bCn1pu0xw',
|
||||
'ext': 'f4v',
|
||||
u'title': u'techniques test',
|
||||
'title': 'techniques test',
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
|
||||
def _real_extract(self, url):
|
||||
mobj = re.match(self._VALID_URL, url)
|
||||
video_id = mobj.group('id')
|
||||
|
||||
webpage = self._download_webpage(url, video_id)
|
||||
#title = self._html_search_meta('title', webpage, 'title')
|
||||
title = self._search_regex(r'<h1 title=.*>(.*?)</h1>', webpage, 'title')
|
||||
self.to_screen('title: '+title)
|
||||
|
||||
dataUrl = 'http://v.ku6.com/fetchVideo4Player/'+video_id+'.html'
|
||||
dataUrl = 'http://v.ku6.com/fetchVideo4Player/%s.html' % video_id
|
||||
jsonData = self._download_json(dataUrl, video_id)
|
||||
downloadUrl = jsonData['data']['f']
|
||||
|
||||
@ -32,7 +31,5 @@ class Ku6IE(InfoExtractor):
|
||||
'id': video_id,
|
||||
'title': title,
|
||||
'url': downloadUrl
|
||||
# TODO more properties (see youtube_dl/extractor/common.py)
|
||||
|
||||
}
|
||||
|
||||
|
Loading…
Reference in New Issue
Block a user