Welcome to java-rdfa

The cruftiest RDFa parser in the world, I'll bet. Apologies that there isn't much documentation. Things may explode: you have been warned.

Currently passing all conformance tests for XHTML, and the HTML 4 and 5 tests with one exception.

This was written by Damian Steer. It is an offshoot of the Stars Project which was funded by JISC

Useful Links

• Maven repository (snapshots)
• Java api documentation
• Online parser

Basic Use

$ ls
htmlparser-1.2.0.jar    java-rdfa-0.3.jar

$ java -jar java-rdfa-0.3.jar http://examples.tobyinkster.co.uk/hcard
<http://examples.tobyinkster.co.uk/hcard> <http://xmlns.com/foaf/0.1/primaryTopic> <http://examples.tobyinkster.co.uk/hcard#jack> .
...

or (equivalent):

$ java -cp '*' rdfa.simpleparse http://examples.tobyinkster.co.uk/hcard
<http://examples.tobyinkster.co.uk/hcard> <http://xmlns.com/foaf/0.1/primaryTopic> <http://examples.tobyinkster.co.uk/hcard#jack> .
...

For HTML sources add the format argument, and you will need the validator.nu parser:

$ java -cp '*' rdfa.simpleparse --format HTML http://www.slideshare.net/intdiabetesfed/world-diabetes-day-2009
<http://www.slideshare.net/intdiabetesfed/world-diabetes-day-2009> <http://www.w3.org/1999/xhtml/vocab#stylesheet> <http://public.slidesharecdn.com/v3/styles/combined.css?1265372095> .
...

The output of simpleparse is n-triples, and hard to read. If you have jena try adding it to you classpath and using rdfa.parse instead:

$ java -cp '*:/path/to/jena/lib/*' rdfa.parse --format HTML http://www.slideshare.net/intdiabetesfed/world-diabetes-day-2009
@prefix dc:      <http://purl.org/dc/terms/> .
@prefix hx:      <http://purl.org/NET/hinclude> .
... nice turtle output ...

Java Use

To use the parser directly, without the assistance of an RDF toolkit (a bold choice) implement a StatementSink to collect the triples, then use a parser from the Factory to make a reader:

XMLReader reader = ParserFactory.createReaderForFormat(sink, Format.XHTML); // or HTML, still an XMLReader
reader.parse(source); // Your sink will be sent triples

java-rdfa can be used from jena. Simply invoke:

Class.forName("net.rootdev.javardfa.RDFaReader");

Which will hook the two readers in to jena, then you will be able to:

model.read(url, "XHTML"); // xml parsing
model.read(other, "HTML"); // html parsing

java-rdfa is available in the maven central repositories. Note that it does not depend on jena.

A sesame reader provided by Henry Story is also available.

Form Mode

There is a secret form mode (that prompted the development of this parser). In this mode you can generate basic graph patterns by including ?variables where curies are allowed, and INPUT tags generate @name variables.

Simple example (from the tests) and the query that results.

Changes

0.3

• Updated to current conformance tests
• Switched validator.nu to streaming mode (may live to regret this).
• Created very simple n-triple and rdf/xml streaming serialisers.
• Usual bug fixes etc.
• Jena is now a provided maven dependency. Using java-rdfa won't pull in jena.
• Sesame reader create by Henry Story added. Can't be added to central maven repository since Sesame isn't available, so spun out in small module.
• Tests for query, and some utilities.