RE: [xsl] XSL template "namespace" problem

[Ian Bonnycastle <ibonny@xxxxxxxxxxxx>]

On Wed, 29 Mar 2006, Michael Kay wrote:

> > One
> > snippet of XSLT code I'm using as an example is:
> >
> > <A NAME="linkone"><xsl:attribute
> > name="HREF">javascript:process('linkone',<xsl:value-of
> > select='/rootnode/value'/>)</xsl:attribute>Click me, value
> > is: <xsl:value-of select="/rootnode/value" />.</A>
> >
> > (Notice the <xsl:value-of> tag inside the Javascript call
> > inside the A
> > HREF, and inside the A HREF data field.)
>
> This is now well-formed (though not very pretty) XML, and it's correct XSLT.
> The HTML it generates is slightly dubious, because javascript:process(.....)
> is not a legal URL, so some HTML validators may get upset about it (and the
> XSLT serializer will replace spaces by %20, which would probably stop it
> working); you may be better off generating attributes like onClick that are
> designed to hold Javascript rather than URLs.

   This is all well and good, but then that means that there is no way to 
create javascript-linked links in XSLT. The reason I'm doing it this way 
is because I'm trying to create a menu system that is based on an XML file 
to define its content. Unfourtunately, using a structure like:

    <A OnClick="funccall()">Some text</A>

   doesn't create a clickable link, even though it might be more "well 
formed" than:

    <A HREF="javascript:funccall()">Some text</A>

   This latter code actually creates a link that one can click on. From 
what I'm understanding, this type of structure, because its not well 
formed, is completely incompatible with XSLT, even thought xsltproc 
processes it properly and IE interprets it correctly. I guess I'm a little 
confused why these browsers and processors are interpreting it if its 
technically illegal.

Ian