- From: Florent Georges <fgeorges@fgeorges.org>
- Date: Tue, 3 Feb 2009 15:59:15 +0100
- To: Toman_Vojtech@emc.com
- Cc: public-xml-processing-model-comments@w3.org
2009/2/3 ? wrote:
Hi,
> <c:multipart boundary="-=-=-=-" content-type="multipart/mixed;
> boundary='-=-=-=-'">
> ...
> </c:multipart>
> Or should the result be:
> <c:multipart boundary="-=-=-=-" content-type="multipart/mixed">
> ...
> </c:multipart>
That's an interesting question. I've been facing the same
problem with an extension function for XSLT that I wrote a few
years ago. IMHO I think it is convenient to provide the user
with common useful values in a strict format (for instance
@content-type is the content type string, and only this, without
extra param) on the one hand, and anyway, we provide the original
headers on the other hand (in c:header elements.)
So I think that if the HTTP server does return the following
header:
Content-Type: multipart/mixed; boundary='-=-=-=-'
then the response should look like the following:
<c:response status="...">
<c:header name="Content-Type"
value="multipart/mixed; boundary='-=-=-=-'"/>
<c:header .../>
<c:header .../>
<c:multipart boundary="-=-=-=-"
content-type="multipart/mixed">
...
</c:multipart>
</c:response>
For the headers, I think the actual draft is clear enough:
§7.1.10.3 says "2. Each response header is translated into a
c:header element." But I think that the way an attribute value
(as @content-type) is got from the original header is maybe
under-specified.
Just my 2 cents...
Regards,
--
Florent Georges
http://www.fgeorges.org/
Received on Tuesday, 3 February 2009 14:59:55 UTC