[XSL-LIST Mailing List Archive Home]
[By Thread]
[By Date]
Re: The XSL-List Digest V1 #149
Subject: Re: The XSL-List Digest V1 #149 From: keshlam@xxxxxxxxxx Date: Mon, 14 Sep 1998 10:46:12 -0400 |
> The various rules in a stylesheet should be able to be in any order > without having the order affect which rule is used in what case >(except perhaps as a final arbiter in an "error recovery" situation). This requires a most-specific-matching-rule system, as opposed to the "lazy evaluation" technique of first-matching-rule. Which in turn is going to require a very tightly nailed down explanation of how that distance is calculated. A problem arises when two rules wind up being the same distance from the node being evaluated (eg, matching on different attributes). This requires some disambiguation mechanism. In other rulesystems I've used, this has usually been handled by some form of first-best-match or last-best-match. If you're going to declare that position will not be meaningful, I think you're going to need something else to break those ties consistantly. This might be expressed as a "rule priority number". Workable. But I'm not sure whether it improves convenience more than it obscures intent or vice versa. Ordering is arguably easier for humans to visualize, and keep track of, than priority numbers would be. ______________________________________ Joe Kesselman / IBM Research Unless stated otherwise, all opinions are solely those of the author. XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
Re: multiple Formatting namespaces., James Tauber | Thread | ?indent-result, Frank Boumphrey |
Re: XSL Tools, Tony Graham | Date | ?indent-result, Frank Boumphrey |
Month |