[
https://issues.apache.org/jira/browse/JENA-2342?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17576285#comment-17576285
]
Andy Seaborne commented on JENA-2342:
-------------------------------------
Thanks for the detailed examples.
It looks like the optimizer is at fault.
A workaround is
{noformat}
ARQ.getContext().set(ARQ.optFilterImplicitJoin, false);
{noformat}
to switch off the transformation at fault.
> SPARQL query comparing "count(distinct ?v)" values is evaluated incorrectly
> ---------------------------------------------------------------------------
>
> Key: JENA-2342
> URL: https://issues.apache.org/jira/browse/JENA-2342
> Project: Apache Jena
> Issue Type: Bug
> Components: SPARQL
> Affects Versions: Jena 4.5.0
> Reporter: Jakub Galgonek
> Priority: Minor
>
> I encountered the following issue while testing the neXtProt database in Jena.
> If the query –
> {code:sql}
> select ?id where {
> values (?id ?v1 ?v2) { ('A' 'B' 'C') }
> } group by ?id having (count(distinct ?v1) = count(distinct ?v2))
> {code}
> – is submitted, then the empty result is returned. However, the result –
> {code:java}
> ( ?id = "A" ){code}
> – should by returned.
> On the other hand, the query –
> {code:sql}
> select ?id where {
> {select ?id (count(distinct ?v1) as ?c1) (count(distinct ?v2) as ?c2) {
> values (?id ?v1 ?v2) { ('A' 'B' 'C') }
> } group by ?id}
> filter (?c1 = ?c2)
> }
> {code}
> – returns the result as expected.
> The following code can be used to reproduce the issue:
> {code:java}
> import org.apache.jena.query.QueryExecution;
> import org.apache.jena.query.QueryExecutionFactory;
> import org.apache.jena.query.ResultSet;
> import org.apache.jena.rdf.model.Model;
> import org.apache.jena.rdf.model.ModelFactory;
> public class Test {
> static String query1 = "select ?id where { values (?id ?v1 ?v2) { ('A' 'B'
> 'C') } } group by ?id having (count(distinct ?v1) = count(distinct ?v2))";
> static String query2 = "select ?id where { {select ?id (count(distinct ?v1)
> as ?c1) (count(distinct ?v2) as ?c2) { values (?id ?v1 ?v2) { ('A' 'B' 'C') }
> } group by ?id} filter (?c1 = ?c2) }";
> public static void main(String[] args) {
> evaluate("Query1:", query1);
> evaluate("Query2:", query2);
> }
> public static void evaluate(String name, String query) {
> Model model = ModelFactory.createDefaultModel();
> QueryExecution qexec = QueryExecutionFactory.create(query, model);
> ResultSet results = qexec.execSelect();
> System.out.println(name);
> while(results.hasNext())
> System.out.println(results.nextSolution());
> }
> }
> {code}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
