Adding Stardog as target graph database to write SEPSES kg triples

[Radu3000]

What do we need to do to be able to write triples to Stardog?

I have posted a related question at https://community.stardog.com/t/sepses-kg-engine-to-write-triples-to-stardog/2251

So far I have tried:

================SPARQL Endpoint=========================

SparqlEndpoint=https://host:port/Test2/query

Triplestore=fuseki;virtuoso;stardog;dummy

Triplestore=stardog UseAuth=true ...

[fekaputra]

Hi, we have an interface that needs to be implemented for each triplestore that you are using https://github.com/sepses/cyber-kg-converter/blob/master/src/main/java/ac/at/tuwien/ifs/sepses/storage/Storage.java.

We have the implementation for Jena Fuseki (https://github.com/sepses/cyber-kg-converter/blob/master/src/main/java/ac/at/tuwien/ifs/sepses/storage/impl/FusekiStorage.java) and Virtuoso (https://github.com/sepses/cyber-kg-converter/blob/master/src/main/java/ac/at/tuwien/ifs/sepses/storage/impl/VirtuosoStorage.java), but yet to create the implementation for Stardog. Since this is open-source software, it would be encouraged that you write the implementation for stardog (and make a PR afterward).

[jbalint]

@Radu3000 assuming the SPARQL INSERT statements will work already with the Fuseki implementation, you can just add s-put and s-post scripts which load the data into Stardog instead of Fuseki. no code changes needed

[Radu3000]

@jbalint - fuseki s-post script has over 725 lines of code ... I don't think you are suggesting to re-write it for Stardog use... ;) ... it would be like re-writing "stardog-admin import" command ...

Maybe API calls is a better route?

My 2c

[jbalint]

@Radu3000 no, I am not suggesting you rewrite anything. I am suggesting you create a script called s-post where you just call stardog data add with the endpoint, named graph, and filename... a 1-liner