RE: SPAM-HIGH: stream closed

[Rob Vesse]

You are closing the result set indirectly (by closing the QueryExecution) 
before you iterate it. 

Result set objects in ARQ are almost always streaming and the associated 
QueryExecution should not be closed until you've finished iterating over the 
results.

Rob

-----Original Message-----
From: "David Moss" <admo...@gmail.com>
Sent: ‎04/‎01/‎2014 03:04
To: "users@jena.apache.org" <users@jena.apache.org>
Subject: SPAM-HIGH:  stream closed

XMLStreamException: ParseError at [row,col]:[338,7] Message: Stream closed

 

I thought I'd share a curious result. I've been experimenting with Jena,
Fuseki and JavaServer Faces.
I use Netbeans as an IDE and a Fuseki instance on my home server, connected
by WiFi to my laptop.
Jena and JSF are running on Tomcat on my laptop, but I have the same result
when I run them directly on the server.

I first populated a JSF select box with a list of resource name and title
using SPARQL. Everything was good, it loaded all the data from the query
into a list and used the list to populate the select list in JSF. This was a
single query and the results was parsed into selectitems in a list. There
are around 100 results returned by the query. All Good.

Next I tried to build a list of objects using the resources returned by the
first query. As each object instantiated itself it sent off a SPARQL query
to Fuseki and populated its attributes. That meant about 100 separate SPARL
queries. It falls over with the error at the top of this message.



If I limit the results in the SPARQL to, say 30, it all works fine. If I try
to instantiate 40 or more objects it falls over.

 

I've discovered sending 100 separate SPARQL queries isn't very efficient and
takes a lot of time, but I'm stumped over why sending more than 40 in
succession causes it to fall over.

 

 

Here is the code that instantiates the objects:



//constructor

 

  public organisation(String org_resource) {

        org_type = new ArrayList();

        setService("http://10.0.0.1:3030/ds/query";);

        this.setOrg_resource(org_resource);

        this.retrieveorg();

    }

 

 

public final void retrieveorg() {

        

 

        String query = "select ?name ?desc ?website where { " +
this.getOrg_resource() + " <http://purl.org/dc/elements/1.1/title> ?name. "

                + this.getOrg_resource() + "
<http://purl.org/dc/elements/1.1/description> ?desc. "

                + this.getOrg_resource() + "
<http://xmlns.com/foaf/0.1/homepage> ?website. "

                + "} ";

 

        Query q = QueryFactory.create(query);

        QueryExecution qe =
QueryExecutionFactory.sparqlService(getService(), q);

 

        ResultSet r = qe.execSelect();

        qe.close();

        if (r.hasNext()) {

 

            QuerySolution s = r.next();

            this.setOrg_name(s.getLiteral("?name").toString());

            this.setOrg_desc(s.getLiteral("?desc").toString());

            this.setOrg_website(s.getLiteral("?website").toString());

        }

    }
//
**************************************************************************

 

 

And the code that calls the constructor and builds the list of objects:


public List getAllOrgs() {

        

        String query = "select ?s where { ?s
<http://www.w3.org/1999/02/22-rdf-syntax-ns#type>
<http://admoss.info/apr/PoliticalOrganisation>. } limit 90";

 

        ArrayList<organisation> s = new ArrayList();

        Query q = QueryFactory.create(query);

        QueryExecution qe =
QueryExecutionFactory.sparqlService(getService(), q);

        ResultSet r = qe.execSelect();

        qe.close();

        

 

        while (r.hasNext()) {

           

            QuerySolution qs = r.next();

            s.add(new organisation(qs.getResource("?s").toString()));

        }

       

        return s; 

    }