Re: The suitability of XSL
Kip Hampton <[email protected]>
| Newsgroups | gmane.comp.lang.perl.xml |
|---|---|
| Message-ID | <[email protected]> |
Howdy, Lot's of worthwhile discussion about using the right tool for the job (XSL vs Perlishness) but not a lot of detail about when to use which and what criteria to use to decide. Really, that's the underlying question, isn't it? Broadly speaking, XSLT is best for transforming deeply nested document-centric XML (docbook, XHTML) while Perlish, programmatic solutions (SAX, Twig, XML::Simple, etc) are best for shallower data-centric XML. Why? Because XSLT is at its best (and simplest and most powerful) when it is used for *recursive* transformations. That is, "find this element, transform it in this way, apply these other rules to that element's children", rinse and repeat. Most of the heartache (and crappy XSL) I've seen over the years has come from cases where people-- myself included-- tried to use XSL as a functional, iterative programing language. They try to write big rules that transform large chunks of the document at once instead of writing smaller template rules that use selective recursion. Just my .02 (not adjusted for inflation) -kip _______________________________________________ Perl-XML mailing list [email protected] To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs