Re: DTD Files !

Eileen> Some existing tools will provide feedback to the author
Eileen> *automatically* if they create an invalid structure. 

Yes of course existing tools will use DTDs that is not surprising.
The question is whether tools need to be that way.

Eileen>	[<ek>]  Typically *outside* of their authoring environment - much
Eileen> less efficient.

again, authoring environments can and do change.

Selim> What about larger company example where everyone needs to use the same
Selim> structure with various XSL files. The XSL validation code should then be
Selim> duplicated in each of them?

Not necessarily, you can have a validating stylesheet that produces
nothing other than error messages (eg those produced by schematron)
and then separately have stylesheets doing transforms.

But note I did not say that you should not use DTD, I just queried the
statement that you _had_ to use DTD to ensure conformance.

Selim> which way do we gain performance,
Selim> By using validating parser and processing it through a light stylesheet or
Selim> by using nonvalidating parser and processing it through a heavier
Selim> (validating) stylesheet?

It is not clear to me that (say) a schematron schema is much `heavier
weight' than a DTD. Also an XSL stylesheet can check a lot more things
than are possible to check in a DTD, so if you want to enforce those
things you need some application layer to do additional checks over and
above the validating parse with a DTD, in which case an XSL stylesheet
might be quite a `lightweight' solution.

David


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