A JS client lib to communicate with a triple store database through SPARQL queries over HTTP.
Experimental: Expect the unexpected. Please provide feedback on api and your use-case.
If you think this is a fully fledged SPARQL client for JS: nope. This is rather a starting point than a complete implementation. Also, I am not exactly an expert in using SPARQL (that's an understatement) so there might be all sorts of bullshit going on in the lib. Feel free to bash me in the issues...
The library is supposed to be modeled after the 1.1 version of the SPARQL spec but is not yet complete.
Node (min 0.11 or iojs): npm install sparql-hollandaise
Browser: bower install sparql-hollandaise or download/clone the repo and use the files in dist/web/.
Node:
var SPH = require('sparql-hollandaise'),
query = new SPH.Query('https://here.goes.the/endpoint');Browser:
<head>
<script type="text/javascript" src="/bower_components/babel-polyfill/browser-polyfill.js"></script>
<script type="text/javascript" src="/bower_components/sparql-hollandaise/dist/web/sparql-hollandaise.js"></script>
</head>Alternatively, if you are using ES6 JavaScript you can also just import the untranspiled code:
import * as SPH from 'sparql-hollandaise/src/index'Make sure to check out the basic Angular example as well.
The .where() function takes either a single GraphPattern object, or a GroupGraphPattern that in turn contains multiple GraphPattern objects.
var graphPattern = new SPH.GraphPattern(
['array of triple strings'] ||
'single triple string' ||
new SPH.Filter('filter string')
);
// additional args for the GraphPattern
var graphPatternOptional = new SPH.GraphPattern('my super triple', true, false); // this pattern is OPTIONAL
var graphPatternAlternative = new SPH.GraphPattern('my super triple', false, true); // this pattern is an alternative (UNION)
var groupGraphPattern = new SPH.GraphPattern([graphPattern, graphPatternAlternative] || graphPattern);
groupGraphPattern.addElement(graphPatternOptional);Once you have created your pattern, you can pass it to any of the queries.
var query = new SPH.Query('https://here.goes.the/endpoint')
.prefix(['foo: <http://bar>', 'pre: <http://fix>'] || 'pre: <http://fix>')
.select('*') // you can add 'DISTINCT' or 'REDUCED' as a modifier in the second parameter
.from('dataset clause', true) // second param indicates a named dataset (only pass this if named set)
// you can add items to the where clause in any order
// and at any time before calling exec()
.where(graphPattern)
.order('order string')
.limit(10)
.offset(5)
.exec().then(function (result) {
console.log(result);
});var query = new SPH.Query('https://here.goes.the/endpoint')
.prefix('foo: <http://bar>')
.describe('?x')
.where(graphPattern)
.exec().then(function (result) {
console.log(result);
});var query = new SPH.Query('https://here.goes.the/endpoint')
.prefix('foo: <http://bar>')
.ask()
.where(graphPattern)
.exec().then(function (result) {
console.log(result);
});var query = new SPH.Query('https://here.goes.the/endpoint')
.prefix('foo: <http://bar>')
.construct(['array of triple strings'] || 'single triple string')
.where(graphPattern)
.order('order string')
.limit(10)
.offset(5)
.exec().then(function (result) {
console.log(result);
});You can initialize a QueryFactory to create queries from the same endpoint (mandatory), base (optional) and prefix(es) (optional).
var queryFactory = new SPH.QueryFactory(
'http://localhost', // endpoint URL
{}, // auth (optional)
'GET', // method (optional)
'<http://example.org/foo/>', // base (optional)
'foo: http://bar' // prefix(es) (optional)
);
// then get a query object based on those settings
var query = queryFactory.make();
// further configure the query as stated above
query.ask()
.where(...)
...These might be useful to you:
// get the query's string representation that will be sent to the server
var queryString = query.toString();
// reset the entire query to reuse the object
query.reset();API docs for the classes in src/sparql/ can be found here
The classes in src/ use most of the ES6 keywords, so transpiling is necessary. The project uses gulp, Babel and Browserify.
npm install
# build browser lib and transpile for NodeJS
gulp
# you can also do:
gulp build-web
gulp build-nodeRun the tests with npm test and generate code coverage with npm run-script cover (view coverage).
YUIDocs can be generated with npm run-script docs.
This client was conceived at a hackathon initiated by the Pina Bausch Foundation at the Mediencampus Hochschule Darmstadt in December 2015.