[Bug] the HTML validator doesn't handle the Content-Encoding header

Hey guys,

I've just found that the HTTP client of the HTML Validator doesn't
acknowledge the Content-Encoding header, in that a document served as
text/html + content-encoding: gzip will be handled in it binary format,
which of course won't work (and it might actually generate an invalid
Invalid page, since the source of the file will be included in the error
message).

This bug is present both in the production and development versions of
the validator. The correct behavior would be:
- either to discard any document served with a content-encoding as non
supported
- or to actually decode the content-encoded document for a set of
well-known encoding

Dom
-- 
Dominique Hazaël-Massieux - http://www.w3.org/People/Dom/
W3C/INRIA
mailto:dom@w3.org