A question regarding sparql endpoint

[Maxx]

Hi,

First of all I am really sorry to be asking this question on your forum. I understand your focus is on Gremlin. However, I have been struggling with this for quite sometime now and did not know where I could ask this question to get a helpful response. I am doing a project for my master thesis that requires me to benchmark some graph data stores for rdf storage and sparql querying. I was able to load the data using tinkerpop blueprints. My next task is to query the stored data using a sparql endpoint for my test driver, I have been strugglinf with this for quite sometime and I am hoping someone can shed some light in this area. So, here is the question:

I have created a SAIL object using tinkerpop blueprints to load RDF data into a Graph database. Ive successfully managed to load the data. The Graph database in question does not have a sparql endpoint to run my query mix test driver. Does anyone know if it possible and if so then how to make a sparql endpoint accessible to my test driver java code over an HTTP server? the sparql endpoint needs to be of the form http:// (somehost) /sparql (I will be sending it as a commandline argument to my benchmark test driver). I am aware 4store, BigData, Neo4j and some other stores have their own built in sparql servers for querying. I am wondering if it is possible to make a sparql server like that available for my Sail object store if the underlying graph database does not provide its own implementation of the server.

Thanks in advance.

[Stephen Mallette]

it's perhaps a bit of an unknown feature, but Rexster does have a sparql extension:

https://github.com/tinkerpop/rexster/tree/master/rexster-kibbles/sparql-kibble

I haven't tried it out in years, but I would assume it is still good to go.

HTH,

Stephen

--
You received this message because you are subscribed to the Google Groups "Gremlin-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gremlin-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gremlin-users/63719656-bbdb-41d1-b439-ff48c66ca935%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Maxx]

Thanks Stephen!

I learnt a lot in the process of configuring!

I was able to configure my graph (orientgraph) in rexster, but kibbles is unable to recognize it as a SailGraph (as discussed http://stackoverflow.com/questions/25874497/sparql-endpoint-for-sail-object). I got the following message on running a sparql query:

{"message":"the graph to which this extension is applied is not a SailGraph implementation","api":{"description":"execute SPARQL queries against a SAIL graph.","parameters":{"rexster.returnKeys":"an array of element property keys to return (default is to return all element properties)","query":"the SPARQL query to be evaluated","rexster.showTypes":"displays the properties of the elements with their native data type (default is false)"}},"success":false}

I cannot do anything more here using kibbles or?

Best,

Maxxie

[Stephen Mallette]

As I stated in the comments of that SO question, I can't offer much more except to say that you can communicate via RexPro or the REST Gremlin Extension and in those requests, you simply wrap your OrientDB instance in a GraphSail Ouplementation (https://github.com/tinkerpop/blueprints/wiki/Sail-Ouplementation) then issue your sparql to that.  You will have to wrap your OrientDB instance in GraphSail on every single request with the REST Gremlin Extension.  If you used RexPro with a session i guess you could do that once and then reference the wrapped SailGraph on future requests.

To view this discussion on the web visit https://groups.google.com/d/msgid/gremlin-users/208e187f-f42b-4abf-aae3-f92e08ba604f%40googlegroups.com.