Plugin 1: SPARQL Query to Triple Pattern Query Transformer
The plugin can take a SPARQL Query and extract a set of Triple Patterns from the WHERE clause. The data layer API needs to be able to build a single Triple Pattern Query with multiple patterns inside. This is not currently possible as the current Triple Pattern Query in the API contains only a single pattern and also requries a graph name to be specified
Plugin 2: Sindice Triple Pattern Query Identifier
This plugin takes a Triple Pattern Query containing multiple triple patterns and sends it Sindice. It receives a list of rdf graph locations from which it should build ?RdfGraphs. This appears to work correctly.
Plugin 3: Pipeline Filter
This plugin is needed because the output of identify and transform are Collections of ?InformationSets, while the input to select is a Dataset. Essentially the pipeline filter takes only the ?RdfGraphs from the pipeline (discarding any untransformed Natural Language Documents or other information sets) and puts them together in a Dataset. It is currently unclear how to build a Dataset from a set of ?RemoteRdf objects. This needs clarification.
Plugin 3: Grab Everything Selecter
In the current Baby LarKC we use the Grab Everything Selecter. Where the tripleset that comes out from the selecter is equal in size to the input Dataset. Therefore I need a means to take all the ?RdfGraphs from the Dataset and place them in the ?TripleSet. I cannot find a way to do this with ?DataFactory
Plugin 4: Reason
I have not managed to get this far yet as I cannot get the Pipeline Filter or Grab Everything Selecter to work properly. Of course I should be able to pass in the ?TripleSet from the Selecter and execute the users input SparQL Query against it.