- From: Sandro Hawke <sandro@w3.org>
- Date: Fri, 13 Jul 2007 11:53:55 -0400
- To: Dave Reynolds <der@hplb.hpl.hp.com>
- Cc: public-rif-wg@w3.org
Dave Reynolds <der@hplb.hpl.hp.com> writes: > Consider: > > <Var><name>?x</name></Var> > > In the case of simple string names we know that two occurrences of the > same variable name within the same scope refer to the same semantic > object but we don't need to turn the parse tree into some parse lattice > in order to represent that. The lattice is constructed when you turn the > parse tree into your internal rule representation using symbol tables or > whatever. > > Presumably you don't want to use nodeIDs to tie up the Var bNodes? Right -- this is issue #1 in my other e-mail [1]. I don't feel strongly about this, but, yes I do lean towards using nodeIDs to connect occurances of the same variable, instead of making RIF consumers implement per-scope nested symbol tables. I think it better fits the notion that it is the same variable being used in different places. (In the larger syntax discussion, the question of repeated-serialization is confusing when the only property a thing has is its identifier.) > Consts labeled by IRIs seem like the same thing to me. Yes -- it's clearer on vars. -- Sandro [1] http://lists.w3.org/Archives/Public/public-rif-wg/2007Jul/0069.html
Received on Friday, 13 July 2007 15:54:56 UTC