github/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-21 20:46:36 -05:00
Code Issues Projects Releases Packages Wiki Activity

[youporn] Improve uploader extraction

Browse source
This commit is contained in:
Sergey M․ 2015-10-25 23:12:12 +06:00
parent feb7711cf5
commit 4f13f8f798
1 changed files with 27 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

32
youtube_dl/extractor/youporn.py
View file

@ -15,8 +15,9 @@
class YouPornIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?youporn\.com/watch/(?P<id>\d+)/(?P<display_id>[^/?#&]+)'
_TEST = {
_TESTS = [{
'url': 'http://www.youporn.com/watch/505835/sex-ed-is-it-safe-to-masturbate-daily/',
'md5': '71ec5fcfddacf80f495efa8b6a8d9a89',
'info_dict': {
'id': '505835',
'display_id': 'sex-ed-is-it-safe-to-masturbate-daily',
@ -31,8 +32,29 @@ class YouPornIE(InfoExtractor):
'categories': list,
'tags': list,
'age_limit': 18,
}
}
},
}, {
# Anonymous User uploader
'url': 'http://www.youporn.com/watch/561726/big-tits-awesome-brunette-on-amazing-webcam-show/?from=related3&al=2&from_id=561726&pos=4',
'info_dict': {
'id': '561726',
'display_id': 'big-tits-awesome-brunette-on-amazing-webcam-show',
'ext': 'mp4',
'title': 'Big Tits Awesome Brunette On amazing webcam show',
'description': 'http://sweetlivegirls.com Big Tits Awesome Brunette On amazing webcam show.mp4',
'thumbnail': 're:^https?://.*\.jpg$',
'uploader': 'Anonymous User',
'upload_date': '20111125',
'average_rating': int,
'view_count': int,
'categories': list,
'tags': list,
'age_limit': 18,
},
'params': {
'skip_download': True,
},
}]
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
@ -97,8 +119,8 @@ def _real_extract(self, url):
r'(?:imageurl\s*=|poster\s*:)\s*(["\'])(?P<thumbnail>.+?)\1',
webpage, 'thumbnail', fatal=False, group='thumbnail')
uploader = self._search_regex(
r'<div[^>]+class=["\']videoInfoBy["\'][^>]*>\s*By:\s*</div>\s*<a[^>]+href="[^"]*">([^<]+)</a>',
uploader = self._html_search_regex(
r'(?s)<div[^>]+class=["\']videoInfoBy["\'][^>]*>\s*By:\s*</div>(.+?)</(?:a|div)>',
webpage, 'uploader', fatal=False)
upload_date = unified_strdate(self._html_search_regex(
r'(?s)<div[^>]+class=["\']videoInfoTime["\'][^>]*>(.+?)</div>',