Hopiu/linkchecker
1
0
Fork 0
mirror of https://github.com/Hopiu/linkchecker.git synced 2026-04-04 23:20:34 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

content encoding getter method

Browse source 
git-svn-id: https://linkchecker.svn.sourceforge.net/svnroot/linkchecker/trunk/linkchecker@2380 e7d03fd6-7b0d-0410-9947-9c21f3af8025
This commit is contained in:
calvin 2005-03-09 13:39:20 +00:00
parent 12b726b4df
commit ca49e04730
1 changed files with 13 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

16
linkcheck/checker/httpurl.py
View file

@ -535,7 +535,7 @@ class HttpUrl (urlbase.UrlBase, proxysupport.ProxySupport):
response = self._get_http_response()
self.headers = response.msg
self.data = response.read()
encoding = self.headers.get("Content-Encoding")
encoding = self.get_content_encoding()
if encoding in _supported_encodings:
try:
if encoding == 'deflate':
@ -563,7 +563,7 @@ class HttpUrl (urlbase.UrlBase, proxysupport.ProxySupport):
return False
if self.headers.gettype()[:9] != "text/html":
return False
encoding = self.headers.get("Content-Encoding")
encoding = self.get_content_encoding()
if encoding and encoding not in _supported_encodings and \
encoding != 'identity':
self.add_warning(_('Unsupported content encoding %r.') % encoding)
@ -593,6 +593,16 @@ class HttpUrl (urlbase.UrlBase, proxysupport.ProxySupport):
ptype = ptype.split(';')[0]
return ptype.strip()
def get_content_encoding (self):
"""
Get the content encoding from the Content-Encoding header value, or
an empty string if not found.
@return: encoding string
@rtype: string
"""
return self.headers.get("Content-Encoding", "").strip()
def is_parseable (self):
"""
Check if content is parseable for recursion.
@ -604,7 +614,7 @@ class HttpUrl (urlbase.UrlBase, proxysupport.ProxySupport):
return False
if self.get_content_type() not in ("text/html", "text/css"):
return False
encoding = self.headers.get("Content-Encoding")
encoding = self.get_content_encoding()
if encoding and encoding not in _supported_encodings and \
encoding != 'identity':
self.add_warning(_('Unsupported content encoding %r.') % encoding)