Tests, general and extendedType-literal-ne in particular

Hi,

I've already sent some comments on the SPARQL tests regarding syntax.
I'm now using the tests to debug the semantic part of the SWI-Prolog
SPARQL engine.  I fear I'll find a few issues :-)  Some questions:

  * I'm now using the XML test-data.  Part of this is actually Turtle
    data, so I generalised my test framework to use both.  I have the
    impression the test-data is maintained in Turtle.  If this is the case 
I'll
    switch and I'd appreciate if the website would state (near the
    downloads) that the Turtle data is the one maintained.

    If the RDF/XML data is provided, it would be great if it was formatted
    a bit better.  I'm now load and save many of the files using the
    SWI-Prolog RDF parser to get something readable.

  * I'm struggling with the extendedType-literal-ne test.  The result
    data says it should return no results.  Query is this

======================= extendedType-literal-ne ========================
# Test FILTER negative match on extended type.
# $Id: extendedType-literal-ne.rq,v 1.1 2005/08/02 03:30:19 eric Exp $

PREFIX loc: 
<http://www.w3.org/2001/sw/DataAccess/tests/data/ValueTesting/extendedType-0#>
PREFIX air: <http://www.daml.org/2001/10/html/airport-ont#>
SELECT ?a1 ?a2
 WHERE {?a1 air:name ?name1 ;
            loc:location ?l1 .
        ?l1 loc:rotational ?r .
        FILTER (?r != "-7420507.1"^^loc:ECEF_X) }
------------------------------------------------------------------------

Data is this (after reformatting):

================================================================
<?xml version='1.0' encoding='UTF-8'?>
<!DOCTYPE rdf:RDF [
    <!ENTITY air 'http://www.daml.org/2001/10/html/airport-ont#'>
    <!ENTITY rdf 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'>
    <!ENTITY loc 
'http://www.w3.org/2001/sw/DataAccess/tests/data/ValueTesting/extendedType-0#'>
]>

<rdf:RDF
    xmlns:air="&air;"
    xmlns:rdf="&rdf;"
    xmlns:loc="&loc;"
>
<rdf:Description rdf:about="&loc;a"
    air:name="Austin Bergstrom International">
  <loc:location>
    <loc:geodetic>
      <loc:orthogonal 
rdf:datatype="&loc;longitude">-97.700000</loc:orthogonal>
      <loc:rotational rdf:datatype="&loc;latitude">30.300000</loc:rotational>
    </loc:geodetic>
  </loc:location>
</rdf:Description>

<rdf:Description rdf:about="&loc;b"
    air:name="Austin Bergstrom International">
  <loc:location>
    <loc:EarthCenteredEarthFixed>
      <loc:orthogonal rdf:datatype="&loc;ECEF_Y">-5462738.5</loc:orthogonal>
      <loc:rotational rdf:datatype="&loc;ECEF_X">-7420507.1</loc:rotational>
    </loc:EarthCenteredEarthFixed>
  </loc:location>
</rdf:Description>

</rdf:RDF>
================================================================

First of all, the query uses the variable a2 in the projection which is not 
used
in the query.  

My system now returns the row(loc:a, NULL).  This looks like the correct 
result to
me as I do not see where the logic resides that makes these two coordinates 
the
same to the SPARQL engine.  What am I missing?

Comments?

	Cheers --- Jan

Received on Wednesday, 21 June 2006 12:46:16 UTC