W3C home > Mailing lists > Public > public-rdf-dawg@w3.org > April to June 2010

Re: issue round-up, part 1

From: Olivier Corby <Olivier.Corby@sophia.inria.fr>
Date: Mon, 17 May 2010 14:03:58 +0200
Message-ID: <4BF130AE.2070706@sophia.inria.fr>
To: SPARQL Working Group <public-rdf-dawg@w3.org>
 > PROPOSED: Close ISSUE-8 with the consensus that subqueries share the 
same RDF
 > dataset as their parent query, and that FROM and FROM NAMED clauses 
are not
 > permitted in subqueries.

As I said in a previous post, I think there is an interaction between 
graph  and subquery patterns. Does the graph pattern apply to subquery:


graph ?g {
   {select * where { ... }}
}


I think this is not obvious because we can consider that the subquery 
creates a fresh new evaluation environment without graph ?g as current 
graph.

This interpretation is based on this case:

graph ?g1 {
   graph ?g2 { }
}

In this example, graph ?g2 creates a new evaluation environment in which 
graph ?g2 overloads/hides graph ?g1. We could have the same behavior 
with subquery.

I think that the recommendation should explicit the behavior of graph 
pattern vs subquery.


Olivier
Received on Monday, 17 May 2010 12:04:33 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:15:42 GMT