Re: [xsl] top level params and xsl:attribute magic?

[S Woodside <sbwoodside@xxxxxxxxx>]

On Sunday, January 19, 2003, at 02:29  AM, Dimitre Novatchev wrote:

> "S Woodside" <sbwoodside@xxxxxxxxx> wrote in message
> news:7FCB3007-2B28-11D7-8385-000393414368@xxxxxxxxxxxx
> > Thanks, I considered this, but I think that passing the full XPath
> > makes debugging easier and will be more robust if I change the source
> > XML (it has a chance of still working, instead of being guaranteed not
> > to work.)
<snip/>
> As for easier debugging when using XPath expressions, this is right,
> but only in the case we know well the source xml document. In case the
> transformation is designed to be generic and process *any* source xml
> document, then in the general case when the structure of the source xml
> is not known to the programmer, using XPath expressions is exactly as
> helpful as using another uid system.

In this case, the source XML is guaranteed to be a Relax NG schema.

> The big advantage of the coordinate uid method is the fact that nodes
> can be located very fast by using a key. Future XSLT processors could
> be optimised to automatically maintain the coordinate uid of every
> node, so that a programmer would not even have to declare an xsl:key
> for node() on its coordinates.

I see. I'll keep it in mind. I like your proposal of a fixed ID 
function, by the way.

simon
---
www.simonwoodside.com


XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list