RE: [ubl-cmsc] Transformation Language

[Matthew Gertner <matthew.gertner@schemantix.com>]

I would support this. We will design a specific context rule syntax that
makes it easier to specify context rules than if XSLT were used. At the same
time, we should try to ensure that a direct mapping to XSLT (preferably an
automated one) is possible. Personally I don't see that this mapping has to
be implementable using XSLT, but I'd be interested to hear if someone has a
strong feeling about this too.

Matt

> -----Original Message-----
> From: Eduardo Gutentag [mailto:eduardo.gutentag@sun.com]
> Sent: Saturday, November 17, 2001 12:34 AM
> To: Matthew Gertner
> Cc: 'ubl-cmsc@lists.oasis-open.org'
> Subject: Re: [ubl-cmsc] Transformation Language
> 
> 
> Matthew Gertner wrote:
> > 
> > 4) Should the transformation be represented as an adhoc XML 
> document (like
> > the Vienna approach) or using some standard transformation 
> language (such as
> > XSLT)?
> 
> I'm having a hard time answering this one shortly and 
> coherently, but in any
> event please remember that I am a strong supporter of XSLT 
> for just about 
> anything short of brushing my teeth.
> 
> However, I think the question should be presented as "Should 
> the expression of
> context rules be represented as...", because there is no doubt in
> my mind that the transformation itself should be done with 
> XSLT (even though
> some could do it through Java or through some proprietary method).
> 
> The problem I see with representing the rules with XSLT is 
> that XSLT appears
> very complicated and long to the untrained eye; the rules 
> presented in v1.04
> have a variety of shortcuts that would have to be expressed 
> in very long and
> incomprehensible stylesheets. Or consider the issue of rule 
> application order;
> it is not XSLT's priority (which selects only one rule out of 
> many); in order
> to express order with XSLT one would probably have to write a 
> very complicated
> XSLT stylesheet indeed. The axis of action in the rules, 
> also, is different
> than the one in stylesheets. While the former concentrate on 
> what happens
> if a given context is in play, the latter concentrate on what 
> happens when
> a given element is encountered. It's a different way of looking at the
> problem.
> 
> That being said, I believe it is possible to write an XSLT 
> stylesheet that, when
> applied to a given context rules document, will result in a 
> second XSLT stylesheet
> that, when applied to a schema, will execute the actions 
> indicated in the
> context rules document. So XSLT is not totally out of the picture...
> 
> 
> > 
> > ----------------------------------------------------------------
> > To subscribe or unsubscribe from this elist use the subscription
> > manager: <http://lists.oasis-open.org/ob/adm.pl>
> 
> -- 
> Eduardo Gutentag               |         e-mail: 
> eduardo.gutentag@Sun.COM
> XML Technology Center          |         Phone:  (510) 986-3651 x73651
> Sun Microsystems Inc.          |
>