RDF Triple Storage – ordf.handler.rdf

This module implements some triplestores. Most basic but least scalable is the storage that comes with RDFLib. More advanced and scalable but more complicated to set up is 4store.

RDFLib Basic Storage – RDFLib

class ordf.handler.rdf.RDFLib(*av, **kw)

Bases: ordf.handler.HandlerPlugin

The RDFLib handler can use any of the back-ends that are supported by rdflib itself. The constructor takes a store keyword argument that specifies which to use. This defaults to “IOMemory” which is not terribly useful. Initialisation arguments for the store should be passed as positional arguments, for example:

RDFLib("/some/where/data", store="Sleepycat")

or the equivalent in the configuration file,:

rdflib.args = /some/where/data
rdflib.store = Sleepycat

4Store Quad-Store – FourStore

class ordf.handler.rdf.FourStore(config)

Bases: ordf.handler.rdf.RDFLib

Use of this back-end requires the py4s bindings to 4store. It also requires 4store to be built from the multiclient branch to support multiple simultaneous client connections.

Initialisation takes a comma-separated string of arguments, only the first, the name of the kb is required. For example:

FourStore("kbname")
FourStore("kbname,soft_limit=-1")

or equivalently in the configuration file:

fourstore.args = kbname
fourstore.args = kbname,soft_limit=-1