Configuring Query Processors

Query Processors are classes that can process SPARQL Queries are return a SparqlResultSet or IGraph as appropriate. Query Processors implement the ISparqlQueryProcessor interface and the library provides 5 concrete implementations of this all of which can be configured using the Configuration API.

Basic Configuration

Basic Configuration for a Query Processor looks like the following:

@prefix dnr: <http://www.dotnetrdf.org/configuration#> .

_:proc a dnr:SparqlQueryProcessor ;
  dnr:type "VDS.RDF.Query.LeviathanQueryProcessor" .

Leviathan Query Processor

The Leviathan Query Processor is used to process queries on in-memory stores using the library's Leviathan SPARQL Engine. It is configured quite simply by adding a dnr:usingStore property to the basic configuration, the object pointed to by this property must be a Triple Store which implements the IInMemoryQueryableStore interface e.g.

@prefix dnr: <http://www.dotnetrdf.org/configuration#> .

_:proc a dnr:SparqlQueryProcessor ;
  dnr:type "VDS.RDF.Query.LeviathanQueryProcessor" ;
  dnr:usingStore _:store .

_:store a dnr:TripleStore ;
  dnr:type "VDS.RDF.TripleStore" .

For information on how to configure Triple Stores see Configuration API - Triple Stores.

Alternatively you may use the dnr:usingDataset property to connect it to a Dataset instead. See Configuration API - Datasets for details. If both dnr:usingDataset and dnr:usingStore are present then dnr:usingDataset has priority and dnr:usingStore will be ignored.

Generic Query Processor

The Generic Query Processor is used to process queries against some arbitrary store's SPARQL engine where the store you wish to connect to has an implementation of IQueryableStorage. To configure these handlers simply add a dnr:storageProvider property to the basic configuration like so:

@prefix dnr: <http://www.dotnetrdf.org/configuration#> .

_:proc a dnr:SparqlQueryProcessor ;
  dnr:type "VDS.RDF.Query.GenericQueryProcessor" ;
  dnr:storageProvider _:manager .

_:manager a dnr:StorageProvider ;
  dnr:type "VDS.RDF.Storage.FusekiConnector" ;
  dnr:server <http://localhost:3030/dataset> .

The above specifies a Query Processor which passes the queries to the SPARQL engine of the specified Fuseki store. See Configuration API - Storage Providers for more detail on configuring storage providers.

Simple Query Processor

Similar to the Generic Query Processor the Simple Query Processor passes queries to the ExecuteQuery() method of a Triple Store that implements the INativelyQueryablerStore interface. To configure this add a using Store property that points to a Triple Store that implements the relevant interface e.g.

@prefix dnr: <http://www.dotnetrdf.org/configuration#> .

_:proc a dnr:SparqlQueryProcessor ;
  dnr:type "VDS.RDF.Query.SimpleQueryProcessor" ;
  dnr:usingStore _:store .

_:store a dnr:TripleStore ;
  dnr:type "VDS.RDF.NativeTripleStore" ;
  dnr:storageProvider _:manager .

_:manager a dnr:StorageProvider ;
  dnr:type "VDS.RDF.Storage.TalisPlatformConnector" ;
  dnr:storeID "space" .

The above configuration is effectively identical to the previous example, generally using Generic Query Processors is better and easier to configure.

Remote Query Processor

The Remote Query Processor is used to farm out queries to a remote SPARQL endpoint, it can be configured to use any endpoint configurable as described on Configuration API - SPARQL Clients.

@prefix dnr: <http://www.dotnetrdf.org/configuration#> .

_:proc a dnr:SparqlQueryProcessor ;
  dnr:type "VDS.RDF.Query.RemoteQueryProcessor" ;
  dnr:queryEndpoint _:endpoint .

_:endpoint a dnr:SparqlQueryClient ;
  dnr:type "VDS.RDF.Query.SparqlQueryClient" ;
  dnr:queryEndpointUri <http://example.org/sparql> .

The above configures a Remote Query Processor which passes queries to the endpoint at http://example.org/sparql

Pull Query Processor

The Pull Query Processor is an alternate in-memory SPARQL query engine implementation which aims to minimize the size of intermediate results sets during processing by using a streaming approach to query evaluation.

The Pull Query Processor must be configured with a Triple Store instance to query against and accepts options for setting the default query execution timeout and whether the default graph is the union of all graphs.

@prefix dnr: <http://www.dotnetrdf.org/configuration#> .

_:proc a dnr:SparqlQueryProcessor
  dnr:type "VDS.RDF.Query.Pull.PullQueryProcessor" ;
  dnr:usingStore _:store ;
  dnr:timeout 60000 ;
  dnr:unionDefaultGraph true ;
.

_:store a dnr:TripleStore ;
  dnr:type "VDS.RDF.TripleStore" .

The example above configures a Pull Query Processor to query over an (initially empty) in-memory triple store using the union of all graphs as the default graph and with a query execution timeout set to 1 minute (60 000ms).