On Wed, Dec 07, 2005 at 05:34:11AM -0500, Joel E. Denny wrote: > Should it be possible to reference a key in the match pattern of another? > I did not find this topic addressed in the XSLT 1.0 recommendation or > elsewhere on the web. Did I miss it? > > libxslt's treatment of this case seems odd. I've tested with 1.1.15. For > an example, please find the XML and XSLT documents at the end of this > message. Inside the match pattern of key `keyAnotB', the key `keyB' is > referenced. If `keyAnotB' is declared before `keyB', then the key() > invocation for `keyB' returns the expected set of nodes. Otherwise, it > returns the empty set. > > Reading xsltproc -v, I see that key hashes are built in the reverse order > of the declarations. That explains this behavior. However, the opposite > build order seems more intuitive. What's correct?
I think you're reaching the limits of what the XSLT-1.0 spec specifies and are then in "implementation specific beaviour" area. A risky place to be in... Daniel -- Daniel Veillard | Red Hat http://redhat.com/ [EMAIL PROTECTED] | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/ _______________________________________________ xslt mailing list, project page http://xmlsoft.org/XSLT/ [email protected] http://mail.gnome.org/mailman/listinfo/xslt
