[XSL-LIST Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message] RE: keys and idrefs - XSLT2 request?
At 03:59 AM 10/10/01, DaveP wrote:
Wendell wrote: > I think it'd be a tall order to get much consensus on IDREFS being so > special that the definition of key() should be stretched that > far (even > broken, since it makes its functionality very different > depending on a > DTD's being parsed). This would be a source of much > confusion, I imagine, > and also complaints since many users simply wouldn't > understand or accept > the rationale for the design. 1. No need to take something that users already find confusing as a precedent, is there? 2. The distinction between attributes named 'id' and attributes of type ID has an impact on the id() function, which users unfamiliar with the ID mechanism find confusing. But id() has no other use besides resolving cross-references. key() has many other uses (significantly, grouping). Especially given the powerful but subtle feature of a key declared with a node set as its 'use' attribute, or a key() function called with a node set as its second argument, keys are already a bit steep on the learning curve (though in this case the tradeoff is well worth it). So I'd be very reluctant to mess with the key mechanism as it stands, if there are other approaches. 3. I think the WG -- and Jeni -- with their alternatives, have already made this debate moot! :-) I ought to start an faq section entitled come here when you're desperate. I could include Mike Kays subtleties on namespaces, missing/mistyped namespaces in a stylesheet, lack of DTD with screwed up cross references etc. Yes, there are always more! Anyway, that's why IDREFS make a special case, despite our many users of well formed not valid XML. I agree that IDREFS make a special case, but consider a need to change the way key() works to be a non sequitur. Also, as a general principle I think it's good to keep DTD dependencies to an absolute minimum. (This is coming from a user in whose world DTDs are very useful: unlike Mike K. I *like* DTDs.) We'll have plenty enough to worry about with post-schema-validation-infosets! Cheers, Wendell
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
|
PURCHASE STYLUS STUDIO ONLINE TODAY!Purchasing Stylus Studio from our online shop is Easy, Secure and Value Priced! Download The World's Best XML IDE!Accelerate XML development with our award-winning XML IDE - Download a free trial today! Subscribe in XML format
|