mirror of
https://github.com/ytdl-org/youtube-dl.git
synced 2024-11-22 02:01:50 +00:00
[youjizz] Modernize (#4131)
This commit is contained in:
parent
ef89dba58f
commit
28465df1ff
1 changed files with 11 additions and 21 deletions
|
@ -9,40 +9,30 @@ from ..utils import (
|
|||
|
||||
|
||||
class YouJizzIE(InfoExtractor):
|
||||
_VALID_URL = r'^https?://(?:\w+\.)?youjizz\.com/videos/(?P<videoid>[^.]+)\.html$'
|
||||
_VALID_URL = r'https?://(?:\w+\.)?youjizz\.com/videos/[^/#?]+-(?P<id>[0-9]+)\.html(?:$|[?#])'
|
||||
_TEST = {
|
||||
'url': 'http://www.youjizz.com/videos/zeichentrick-1-2189178.html',
|
||||
'file': '2189178.flv',
|
||||
'md5': '07e15fa469ba384c7693fd246905547c',
|
||||
'info_dict': {
|
||||
'id': '2189178',
|
||||
'ext': 'flv',
|
||||
"title": "Zeichentrick 1",
|
||||
"age_limit": 18,
|
||||
}
|
||||
}
|
||||
|
||||
def _real_extract(self, url):
|
||||
mobj = re.match(self._VALID_URL, url)
|
||||
|
||||
video_id = mobj.group('videoid')
|
||||
|
||||
# Get webpage content
|
||||
video_id = self._match_id(url)
|
||||
webpage = self._download_webpage(url, video_id)
|
||||
|
||||
age_limit = self._rta_search(webpage)
|
||||
video_title = self._html_search_regex(
|
||||
r'<title>\s*(.*)\s*</title>', webpage, 'title')
|
||||
|
||||
# Get the video title
|
||||
video_title = self._html_search_regex(r'<title>(?P<title>.*)</title>',
|
||||
webpage, 'title').strip()
|
||||
|
||||
# Get the embed page
|
||||
result = re.search(r'https?://www.youjizz.com/videos/embed/(?P<videoid>[0-9]+)', webpage)
|
||||
if result is None:
|
||||
raise ExtractorError('ERROR: unable to extract embed page')
|
||||
|
||||
embed_page_url = result.group(0).strip()
|
||||
video_id = result.group('videoid')
|
||||
|
||||
webpage = self._download_webpage(embed_page_url, video_id)
|
||||
embed_page_url = self._search_regex(
|
||||
r'(https?://www.youjizz.com/videos/embed/[0-9]+)',
|
||||
webpage, 'embed page')
|
||||
webpage = self._download_webpage(
|
||||
embed_page_url, video_id, note='downloading embed page')
|
||||
|
||||
# Get the video URL
|
||||
m_playlist = re.search(r'so.addVariable\("playlist", ?"(?P<playlist>.+?)"\);', webpage)
|
||||
|
|
Loading…
Reference in a new issue