TinkerPop over SPARQL-star
Do you want Gremlins in your RDF store? Then you have come to the right place!
This is an implementation of Gremlin/TinkerPop over SPARQL-star. You can do two things with it:
1) I have no interest in RDF, I happen to have a triple store and just want to use it as a storage back-end.
If your triple store supports SPARQL-star then you can just point SparqlStarGraph at it, and Gremlin away as normal.
We support the majority of graph features so maybe we can do something that your current implementation can't (or do it better).
Most graph analytic tools assume a standard definition of a graph, but RDF isn't (it is a directed 3-uniform hypergraph, and that is before we throw in named graphs and RDF-star). TinkerPopStar lets you expose a slice of your RDF data as a standard graph and use Gremlin to process it. Real-world data often does not have the structure of a standard graph, so RDF can provide a good fit. At the other end, analytics need a standard graph. TinkerPopStar can provide the bridge.
Essentially:
An edge (a)-[e]-(b) maps to the RDF triple a e b.
A vertex property a.prop = value maps to the RDF triple a prop value.
An edge property e.prop = value maps to the RDF* triple <<a e b>> prop value.
A vertex meta-property (a.prop = value).metaprop = metavalue maps to the RDF* triple <<a prop value>> metaprop metavalue.
These are accompanied by some RDFS statements to make the data understandable by RDF tools and allow mixed data - other triples can freely co-exist with the TinkerPopStar triples.
Exceptions:
Multiedges (enabled by default) (a)-[e]-(b) map to the RDF triples a e uuid. uuid rdfs:seeAlso b.
List vertex properties a.prop = value maps to the RDF triples a prop uuid. uuid rdf:value value.