W3C home > Mailing lists > Public > public-rdb2rdf-wg@w3.org > August 2010

Re: Comments on the SQL-based proposal

From: Souri Das <Souripriya.Das@oracle.com>
Date: Mon, 16 Aug 2010 18:04:55 -0400
Message-ID: <4C69B607.1080507@Oracle.com>
To: public-rdb2rdf-wg@w3.org
I have created new versions (Revision 1) of the schema and example based 
upon the Richard's comments and last week's meeting:

http://www.w3.org/2001/sw/rdb2rdf/wiki/An_XML_Schema_for_SQL-based_RDB2RDF_Mapping_%28Revision_1%29
http://www.w3.org/2001/sw/rdb2rdf/wiki/Example_of_SQL-based_RDB2RDF_Mapping:_Revision_1

I am still not sure about a few things, most importantly

- Regarding Comment 2.5 (Lookup Table): How can we translate a SPARQL 
query with FILTER involving transformed values?

   Example: Consider the following SPARQL query involving a property, 
:transfProp, whose values are transformed, via a lookup table, into (-1) 
times the original value
        (e.g., 10 in an INTEGER column, i.e., "10"^^xsd:integer in RDF, 
is transformed into "-10"^^xsd:integer):

        SELECT ?x ?y WHERE {?x  :transfProp ?v1 . ?y   :transfProp  ?v2 
. FILTER (?v1 > ?v2)}
  
         How can SQL know, without having the lookup table somehow 
specified (in a CASE statement or something equivalent), how to evaluate 
the FILTER clause?

Thanks,
- Souri.
Received on Monday, 16 August 2010 22:06:14 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 16 August 2010 22:06:15 GMT