- From: Holger Knublauch <holger@topquadrant.com>
- Date: Wed, 15 Apr 2015 05:38:56 +1000
- To: public-data-shapes-wg <public-data-shapes-wg@w3.org>
On 4/15/15 12:34 AM, Peter F. Patel-Schneider wrote: > You are arguing that it is a good thing to be able to query SHACL > syntax when performing SHACL checks. If that is a good thing, then > using SPARQL syntax in SHACL is a *bad* thing because it prevents > effective querying of SHACL syntax. (Yes, you could pull the SPARQL > syntax apart and try to analyze it, but that's not going to be > effective.) Like in the real world, "good" and "bad" are not boolean terms, but rather fuzzy. Trade-offs need to be made. In SPIN we had a complete RDF syntax for all of SPARQL, allowing RDF-based querying. But this was deemed too cryptic to edit by hand, and people need tools for that. In practice few people wanted to query the SPARQL structures. If there is interest in, for example, finding out which URIs are used within a SPARQL query (string) then we could define appropriate built-in functions. But at this stage, I guess the most reasonable thing we could do is to draw the line between high-level terms and low-level SPARQL queries, and assume that it will usually suffice to analyze the former. Holger
Received on Tuesday, 14 April 2015 19:39:30 UTC