1
0
mirror of https://github.com/l1ving/youtube-dl synced 2024-11-26 07:12:59 +08:00
youtube-dl/youtube_dl/extractor/clipfish.py
2013-11-24 07:51:44 +01:00

54 lines
1.7 KiB
Python

import re
import time
import xml.etree.ElementTree
from .common import InfoExtractor
class ClipfishIE(InfoExtractor):
IE_NAME = u'clipfish'
_VALID_URL = r'^https?://(?:www\.)?clipfish\.de/.*?/video/(?P<id>[0-9]+)/'
_TEST = {
u'url': u'http://www.clipfish.de/special/supertalent/video/4028320/supertalent-2013-ivana-opacak-singt-nobodys-perfect/',
u'file': u'4028320.f4v',
u'md5': u'5e38bda8c329fbfb42be0386a3f5a382',
u'info_dict': {
u'title': u'Supertalent 2013: Ivana Opacak singt Nobody\'s Perfect',
u'duration': 399,
}
}
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
video_id = mobj.group(1)
info_url = ('http://www.clipfish.de/devxml/videoinfo/%s?ts=%d' %
(video_id, int(time.time())))
info_xml = self._download_webpage(
info_url, video_id, note=u'Downloading info page')
doc = xml.etree.ElementTree.fromstring(info_xml)
title = doc.find('title').text
video_url = doc.find('filename').text
thumbnail = doc.find('imageurl').text
duration_str = doc.find('duration').text
m = re.match(
r'^(?P<hours>[0-9]+):(?P<minutes>[0-9]{2}):(?P<seconds>[0-9]{2}):(?P<ms>[0-9]*)$',
duration_str)
if m:
duration = (
(int(m.group('hours')) * 60 * 60) +
(int(m.group('minutes')) * 60) +
(int(m.group('seconds')))
)
else:
duration = None
return {
'id': video_id,
'title': title,
'url': video_url,
'thumbnail': thumbnail,
'duration': duration,
}