Javascript RDFParser from Tabulator
David Sheets introduces himself and the new RDF/XML parser he added to the Tabulator all-javascript RDF browser. I've been using the RDF parser from Jim Ley in a generic JavaScript RDF Editor (to work with any OWL/schema files; no public release yet, but let me know if you're interested, especially in helping out). So recently I worked on using Tabulator's rdf parser to see if it was better. It seems to be designed a bit better (separate data store from parser from XML loading) and David's post says it is more accurate and often faster.
I had been hoping to find some additional functionality to use in my JsRdfEditor. Namely, modelling the RDF document instead of just getting all the triples. For example, knowing where a nodeID was used, and the XML arrangement of the triples. That way the JsRdfEditor could parse a file, add some triples, and output a new document that is arranged much like the one that was loaded. But no luck there; sounds like I'll have to add that capability to the parser myself, if I want to have elegant RDF/XML output.
So here's how you can use Tabulator's RDF Parser in your own projects:
// depends on:
// dig.csail.mit.edu/2005/ajar/ajaw/rdf/term.js
// dig.csail.mit.edu/2005/ajar/ajaw/uri.js
// dig.csail.mit.edu/2005/ajar/ajaw/rdf/rdfparser.js
// TestStore implementation from dig.csail.mit.edu/2005/ajar/ajaw/test/rdf/rdfparser.test.html
// see also RDFIndexedFormula from dig.csail.mit.edu/2005/ajar/ajaw/rdf/identity.js
// (extends RDFFormula from dig.csail.mit.edu/2005/ajar/ajaw/rdf/term.js no indexing and smushing)
// for the real implementation used by Tabulator which uses indexing and smushing
var store = new TestStore()
var parser = new RDFParser(store);
parser.reify = parser.forceRDF = true;
// forceRDF isn't used??
var url = 'http://something.or/other';
// get the XML
var xhr = XMLHTTPFactory(); // returns a new XMLHttpRequest, or ActiveX XMLHTTP object
if (xhr.overrideMimeType) {
xhr.overrideMimeType("text/xml");
}
// the "data/" path and encoding is just how I store files locally
xhr.open("GET", "data/" + encodeURIComponent(encodeURIComponent(url)), false);
xhr.send("");
var nodeTree = xhr.responseXML;
if (nodeTree === null && xhr.responseText !== null) {
// Only if the server fails to set Content-Type: text/xml AND xmlhttprequest doesn't have the overrideMimeType method
console.debug("no responseXML, parsing responseText");
nodeTree = (new DOMParser()).parseFromString(xhr.responseText, 'text/xml');
}
// must be an XML document node tree
parser.parse(nodeTree,url);
// use FireBug extension to inspect console.debug'd objects
// Using TestStore you can access store.triples
console.debug('store',store);
Syndicated 2006-11-25 23:43:01 (Updated 2006-11-25 23:43:43) from Dave Brondsema's Blog - Programming
And there is a challenge to get 500 CodeMash blog entries on technorati (blog post aggregrator) before the end of the conference. So here's my contribution to help get one of the coordinators to shave his head.