Subject: [xsl] Re: Assignment no, dynamic scoping si From: Mark Nahabedian <naha@xxxxxxxxxx> Date: Sun, 6 Jan 2002 14:42:08 -0500 (EST) |
I've not caught up with all of my holiday email (still working around some computer trouble) yet so I don't know if these points have already been made, but they seem novel enough so here goes ... Assuming the use of the XSLT "push" model, the dynamic state of the processing is dictated by the static content of the input document. Since the language makes no promises about the order in which sibling (co-selected) nodes are processed, to achieve consistent results, dynamic state could only be communicated downward, not across siblings, which are processed "in parallel". Thus, any data that could be communicated through dynamic scoping could also be extracted via the ancestor axis. Dynamic scoping would thus only provide a benefit if there was some complicated, expensive or redundant calculation to be economised on either in execution time, clarity or code space. Rather than dynamic scoping, One could instead imagine a mechanism that would allow one to look at the ancestors of the already computed higher up levels of the result tree to achieve a similar affect. One would want a means to supress serialization of such data that was only used for further computation and was not required in the output. XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
[xsl] Re: Assignment no, dynamic sc, Terje Norderhaug | Thread | Re: [xsl] Re: Assignment no, dynami, Terje Norderhaug |
[xsl] RE: Higher-order function sup, Dimitre Novatchev | Date | RE: [xsl] Re: . in for, Michael Kay |
Month |