Programming the Semantic Web

Chapter 9. Introspecting Objects from Data

A common problem you face when working with graph databases is figuring out how to access the data easily and efficiently in your code. By using an OWL ontology to describe the common shape of our data, we can map OWL classes in our graph database to Python objects directly and easily, making it easy to read and write data from the database. This approach to database programming is commonly called object-relational mapping and has been popularized in libraries like Python’s SQLObject, Java’s Hibernate, and Ruby’s ActiveRecord. Performing this kind of mapping is very easy with a graph database and an OWL ontology, as we’ll demonstrate in this chapter. The RDFObject framework described here demonstrates how easy it is to tightly integrate code with graph data.

RDFObject Examples

Let’s get a quick overview of how the RDFObject framework works by looking at some examples using our film data. We’ve started by loading the OWL ontology into a Sesame repository called semprog running on localhost. First we’ll initialize the connection to the Sesame repository and create an RDFObjectFactory, the base object for creating RDFObjectGraphs, which we will use to access our graph data. The RDFObjectFactory queries the Sesame graph for ontology information on initialization and then pre-computes and caches mappings from Python attributes to RDF URIs:

>>> from rdfobject import * >>> sc = SesameConnection("localhost:8080", "semprog") >>> factory = RDFObjectGraphFactory(sc) ...