Re: GRAPH details and test cases

On 17 Aug 2006, at 10:27, Seaborne, Andy wrote:

>
>
>
> Lee Feigenbaum wrote:
> . . .
>> The test is source-named/nested-graph-001 (for lack of a better  
>> place to put it). Data:
>> (no default graph)
>> <ng1.n3>
>> @prefix : <http://example.org/ns#> .
>> :s :p :o, :o1.
>> <ng2.n3>
>> @prefix : <http://example.org/ns#> .
>> :s :p :o, :o2.
>> Query:
>> <nested-graph-query.rq>
>> PREFIX : <http://example.org/ns#>
>> SELECT ?o
>> FROM NAMED <ng1.n3>
>> FROM NAMED <ng2.n3>
>> {
>>   GRAPH <ng1.n3> {
>>   GRAPH <ng2.n3> {
>>       :s :p ?o .
>>   }
>>   }
>> }
>> Results:
>> <nested-graph-query-result.n3>
>> @prefix rs:      <http://www.w3.org/2001/sw/DataAccess/tests/ 
>> result-set#> .
>> []  <http://www.w3.org/1999/02/22-rdf-syntax-ns#type>
>>                 rs:ResultSet ;
>>     rs:resultVariable "o";
>>     rs:solution [ rs:binding  [ rs:value    <http://example.org/ 
>> ns#o> ;
>>                                 rs:variable "o"
>>                               ]                 ] ,
>>                 [ rs:binding  [ rs:value    <http://example.org/ 
>> ns#o2> ;
>>                                 rs:variable "o"
>>                               ]                 ]                 .
>> The test runs successfully in my engine.
>
> ARQ passes this test.

FWIW, JXT fails it, rasqal 0.9.12 gets the GRAPHs upside down, though  
I'm not chasing CVS. Also I get a warning, "WHERE omitted".

I only just noticed that WHERE is optional in the grammar. For the  
record I suspect this is not a good idea, and we shouldn't do it in  
testcases.

- Steve

Received on Thursday, 17 August 2006 10:05:40 UTC