MarcoBuster/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-04 03:53:49 +01:00
Code Issues Projects Releases Wiki Activity

[xfileshare] Improve removed videos detection

Browse Source
This commit is contained in:
Sergey M․ 2016-06-13 01:19:54 +07:00
parent cf2bf840ba
commit 33b72ce64e
No known key found for this signature in database
GPG Key ID: 2C393E0F18A9236D
1 changed files with 9 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
youtube_dl/extractor/xfileshare.py
View File

@ -32,7 +32,10 @@ class XFileShareIE(InfoExtractor):
_VALID_URL = (r'https?://(?P<host>(?:www\.)?(?:%s))/(?:embed-)?(?P<id>[0-9a-zA-Z]+)' _VALID_URL = (r'https?://(?P<host>(?:www\.)?(?:%s))/(?:embed-)?(?P<id>[0-9a-zA-Z]+)'
% '|'.join(re.escape(site) for site in list(zip(*_SITES))[0])) % '|'.join(re.escape(site) for site in list(zip(*_SITES))[0]))
_FILE_NOT_FOUND_REGEX = r'>(?:404 - )?File Not Found<' _FILE_NOT_FOUND_REGEXES = (
r'>(?:404 - )?File Not Found<',
r'>The file was removed by administrator<',
)
_TESTS = [{ _TESTS = [{
'url': 'http://gorillavid.in/06y9juieqpmi', 'url': 'http://gorillavid.in/06y9juieqpmi',
@ -88,6 +91,10 @@ class XFileShareIE(InfoExtractor):
'ext': 'mp4', 'ext': 'mp4',
'title': 'Chucky Prank 2015.mp4', 'title': 'Chucky Prank 2015.mp4',
}, },
}, {
# removed by administrator
'url': 'http://xvidstage.com/amfy7atlkx25',
'only_matching': True,
}] }]
def _real_extract(self, url): def _real_extract(self, url):
@ -97,7 +104,7 @@ def _real_extract(self, url):
url = 'http://%s/%s' % (mobj.group('host'), video_id) url = 'http://%s/%s' % (mobj.group('host'), video_id)
webpage = self._download_webpage(url, video_id) webpage = self._download_webpage(url, video_id)
if re.search(self._FILE_NOT_FOUND_REGEX, webpage) is not None: if any(re.search(p, webpage) for p in self._FILE_NOT_FOUND_REGEXES):
raise ExtractorError('Video %s does not exist' % video_id, expected=True) raise ExtractorError('Video %s does not exist' % video_id, expected=True)
fields = self._hidden_inputs(webpage) fields = self._hidden_inputs(webpage)