[utils] Add a function to sanitize consecutive slashes in URLs

This commit is contained in:
Yen Chi Hsuan 2015-03-06 12:43:49 +08:00
parent 5c7495a194
commit 55969016e9
3 changed files with 34 additions and 5 deletions

View file

@ -1789,3 +1789,18 @@ class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
return None # No Proxy
return compat_urllib_request.ProxyHandler.proxy_open(
self, req, proxy, type)
def url_sanitize_consecutive_slashes(url):
"""Sanitize URLs with consecutive slashes
For example, transform both
http://hostname/foo//bar/filename.html
and
http://hostname//foo/bar/filename.html
into
http://hostname/foo/bar/filename.html
"""
parsed_url = list(compat_urlparse.urlparse(url))
parsed_url[2] = re.sub(r'/{2,}', '/', parsed_url[2])
return compat_urlparse.urlunparse(parsed_url)