Re: [xsl] Re: Output validation with XSLT 2.0

Subject: Re: [xsl] Re: Output validation with XSLT 2.0
From: "Andrew Welch" <andrew.j.welch@xxxxxxxxx>
Date: Mon, 5 May 2008 20:44:43 +0100
On 05/05/2008, Michael Kay <mike@xxxxxxxxxxxx> wrote:
> > The "xs" namespace is a bit odd in that it's known to
>  > processors, like the "xsl" namespace,  but unlike the xsl
>  > namespace it makes it into the result tree and requires pruning.
>
>
> The reason for this: if the "xs" namespace had been excluded by default, we
>  would have had do something special for users who want to generate something
>  like <d xsi:type="xs:date"/>.

If it was excluded by default, could you not override that by
explicitly declaring it - for example on the output root element?  But
then you'd need to declare it twice if you really wanted it in the
output...

Or perhaps use namespace-alias, just like when you really need the xsl
namespace in the output.

Either way not great, but maybe preferable to have to work a bit
harder in the special case, rather then every time for every
transform.

-- 
Andrew Welch
http://andrewjwelch.com
Kernow: http://kernowforsaxon.sf.net/

Current Thread