A simple, lightweight module to send queries to SPARQL endpoints and retrieve their results in a streaming fashion.

All results are compatible with the RDFJS specification.

Currently, SPARQL queries such as SELECT, ASK, CONSTRUCT and DESCRIBE are supported. SPARQL UPDATE to insert, delete and patch data is not supported yet.

Internally, this library supports SPARQL results in SPARQL JSON, SPARQL XML, and Turtle.

This package can be installed via npm.

$ npm install fetch-sparql-endpoint

This package also works out-of-the-box in browsers via tools such as webpack and browserify.

import {SparqlEndpointFetcher} from "fetch-sparql-endpoint";

const myFetcher = new SparqlEndpointFetcher();

Optionally, you can pass an options object with the following optional entries:

const myFetcher = new SparqlEndpointFetcher({
  fetch: fetch,                     // A custom fetch-API-supporting function
  dataFactory: DataFactory,         // A custom RDFJS data factory
  prefixVariableQuestionMark: false // If variable names in bindings should be prefixed with '?', defaults to false
});

SPARQL SELECT queries returns a (promise to a) stream of bindings.

const bindingsStream = await fetcher.fetchBindings('https://dbpedia.org/sparql', 'SELECT * WHERE { ?s ?p ?o } LIMIT 100');
bindingsStream.on('data', (bindings) => console.log(bindings));

This will output bindings in the following form, where keys correspond to variables in the queries, and values and RDFJS terms:

{ s: namedNode('s1'), p: namedNode('p1'), o: namedNode('o1') }
{ s: namedNode('s2'), p: namedNode('p2'), o: namedNode('o2') }
{ s: namedNode('s3'), p: namedNode('p3'), o: namedNode('o3') }
...

SPARQL ASK queries answer with a (promise to a) boolean value.

const answer = await fetcher.fetchAsk('https://dbpedia.org/sparql', 'ASK WHERE { ?s ?p ?o }');

This will output true or false.

SPARQL CONSTRUCT and SPARQL DESCRIBE queries returns a (promise to a) stream of triples.

const tripleStream = await fetcher.fetchTriples('https://dbpedia.org/sparql', 'CONSTRUCT { ?s ?p ?o } LIMIT 100');
tripleStream.on('data', (triple) => console.log(triple));

This will output RDFJS triples as follows:

triple(namedNode('s1'), namedNode('p1'), namedNode('o1'))
triple(namedNode('s2'), namedNode('p2'), namedNode('o2'))
triple(namedNode('s3'), namedNode('p3'), namedNode('o3'))
...

If you want to know the query type in order to determine which of the above fetch methods to call, then you can use the getQueryType method as follows:

fetcher.getQueryType('SELECT * WHERE { ?s ?p ?o } LIMIT 100'); // Outputs 'SELECT'
fetcher.getQueryType('ASK WHERE { ?s ?p ?o }');                // Outputs 'ASK'
fetcher.getQueryType('CONSTRUCT { ?s ?p ?o } LIMIT 100');      // Outputs 'CONSTRUCT'

This method will also throw an error if the query contains a syntax error.

A command-line tool is provided to quickly query any SPARQL endpoint:

Usage:

$ fetch-sparql-endpoint https://dbpedia.org/sparql [-q] 'SELECT * WHERE { ?s ?p ?o } LIMIT 100'
$ fetch-sparql-endpoint https://dbpedia.org/sparql -f query.sparql
$ cat query.sparql | fetch-sparql-endpoint https://dbpedia.org/sparql

This software is written by Ruben Taelman.

This code is released under the MIT license.