Jena SPARQL connection not closed

[datalchemist]

I may have miss something but it seems that using banana-rdf to do SPARQL queries will cause connections to remain open without any easy way to close them.

More specifically, if we do something like:
val result=endpoint.executeSelect(parsedQuery)

Then, we got a Jena ResultSet on which we can iterate to get the solutions but we lost the underlying QueryExecution that needs to be closed when result handling is finished.
As a results, if we do many queries, it ends up with an exception such as "Too much files open".

When I look at the JenaSparqlEngine code in banana-rdf, I see that indeed, no reference is kept on the QueryExecution object... and actually, I wonder how this could be handled in a generic way to fit banana-rdf framework.