On Mon, 12 Oct 2009 17:51:45 +0200, Boris Zbarsky <[email protected]> wrote:
On 10/12/09 5:06 AM, Anne van Kesteren wrote:
FWIW, this is what I've done now. It gives authors a reasonable level of
control over the Content-Type header, it does make sure an incorrect
charset parameter is fixed, and includes a charset parameter and MIME
type if the author did not include a Content-Type header.
Note that it also makes it a PITA for authors to correctly send their
XHTML document as application/xhtml+xml with the right charset set. But
if that's ok....
If you do something as simple as
application/xhtml+xml;charset=utf-8
it will be corrected so it should be fine.
However, it is not clear to me why you'd want a charset parameter for XML
resources to begin with. The metadata the XML parser can extract is much
more likely to be correct.
--
Anne van Kesteren
http://annevankesteren.nl/
