Comunica — rete in the RDF/JS ecosystem

Comunica is the modular JavaScript SPARQL framework behind LDflex, GraphQL-LD, and much of the Solid ecosystem. rete plugs into it at two levels — pick by where the joins should run.

Level 1 — zero code: rete datasets as SPARQL endpoints

Every published dataset — and any .rete URL on the web — is a standard SPARQL 1.1 endpoint via the gateway, and Comunica speaks to endpoints natively (verified, this exact command):

$ npx -y -p @comunica/query-sparql comunica-sparql \
    "sparql@https://katospiegel-rete.hf.space/sparql/boe" \
    "SELECT ?title WHERE { <https://www.boe.es/eli/es/c/1978/12/27/(1)> <http://data.europa.eu/eli/ontology#title> ?title }"
[{"title":"\"Constitución Española.\""}]

import { QueryEngine } from "@comunica/query-sparql";

const engine = new QueryEngine();
const bindings = await (await engine.queryBindings(sparql, {
  sources: [
    { type: "sparql", value: "https://katospiegel-rete.hf.space/sparql/boe" },
    // …mix freely with TPF, RDF files, other endpoints — Comunica federates.
  ],
})).toArray();

The whole query is pushed down to the rete engine server-side: rete runs its own optimized joins over the file's indexes and Comunica receives finished bindings. Use this level for heavy multi-join queries over big remote files, and to let Comunica federate rete data with everything else it speaks. Unregistered files work too: …/sparql/https://example.org/any.rete.

Level 2 — native: ReteSource (npm rete-graph ≥ 0.2.0)

An RDF/JS Source over an open graph — local bytes or a lazy URL — pluggable into any Comunica pipeline with no server anywhere:

import { QueryEngine } from "@comunica/query-sparql";
import { open, ReteSource } from "rete-graph";

const source = new ReteSource(await open("https://data.graphplaza.com/boe/boe.rete"));
const bindings = await (await new QueryEngine().queryBindings(
  `SELECT ?who ?label WHERE {
     ?s <urn:x:knows> ?who .
     ?who <http://www.w3.org/2000/01/rdf-schema#label> ?label .
   }`,
  { sources: [source] },
)).toArray();

What the source does, precisely:

• match(s, p, o, g) is one pattern lookup against the file's permutation indexes (a fully-bound pattern becomes an ASK). Comunica then executes the joins itself over the returned quad streams.
• countQuads(…) is implemented, so Comunica's planner can order joins by cardinality.
• RDF/JS semantics are honored: a null graph argument matches the default graph ∪ every named graph, DefaultGraph/NamedNode narrow it, blank-node arguments are matched by their stable labels, datatypes and language tags survive round-trips, and RDF-star quoted triples come back as nested RDF/JS Quads.
• Zero dependencies — the package ships its own minimal RDF/JS terms and stream (nothing else enters your bundle).

Which level, when

Level 3 — native Comunica actors that auto-recognize .rete URLs by content type — is deliberately not built yet; the two levels above cover the use cases without tracking Comunica's actor API across majors. Ask for it in the issues if your pipeline needs it.

Tests behind this page: the client's suite runs the real @comunica/query-sparql engine over a ReteSource (multi-pattern joins, datatype/language fidelity, named graphs), and the published package is smoke-tested from a clean npm install. See also the JavaScript client and triple-store interop.