RDF Hop
An RDF library that treats graphs like graphs.
Intallation
npm i rdfhopUsage
Graphs can represent anything in the world, and for this usage example, it will represent a story. Specifically the story of "Avatar the Last Airbender"
Create a Graph
Let's start with some worldbuilding. Through JavaScript we can easily define a node representing each of the 4 elements. "Water", "Earth", "Fire", and "Air".
import {
Store,
SuperStore,
NamedNode,
NodeSet
} from 'rdfhop'
import rdf from 'rdfhop/vocab/rdf'
import foaf from 'rdfhop/vocab/foaf'
const ELEMENT_PREFIX = 'https://avatar.example/elements#'
const VOCAB_PREFIX = 'https://avatar.example/vocab#'
const elementStore = new Store()
const elements: NodeSet = elementStore.add([
new NamedNode(ELEMENT_PREFIX + 'water').addOut(rdf.label, new Literal('water')),
new NamedNode(ELEMENT_PREFIX + 'earth').addOut(rdf.label, new Literal('earth')),
new NamedNode(ELEMENT_PREFIX + 'fire').addOut(rdf.label, new Literal('fire')),
new NamedNode(ELEMENT_PREFIX + 'air').addOut(rdf.label, new Literal('air'))
])
elements.forEach((element: NamedNode) => {
element.addOut(rdf.type, new NamedNode(VOCAB_PREFIX + 'Element'))
})
console.log(elementStore.toString())
import hop, {
NamedNode,
RDFNode
} from './index'
import rdf from 'rdfhop/vocab/rdf'
import foaf from 'rdfhop/vocab/foaf'
const ELEMENT_PREFIX = 'https://avatar.example/elements#'
const VOCAB_PREFIX = 'https://avatar.example/vocab#'
const elementGraph = hop.graph('https://avatar.example/elements')
const elements = [
hop.namedNode(ELEMENT_PREFIX + 'water').addOut(rdf.label, hop.literal('water'), [elementGraph]),
hop.namedNode(ELEMENT_PREFIX + 'earth').addOut(rdf.label, hop.literal('earth'), [elementGraph]),
hop.namedNode(ELEMENT_PREFIX + 'fire').addOut(rdf.label, hop.literal('fire'), [elementGraph]),
hop.namedNode(ELEMENT_PREFIX + 'air').addOut(rdf.label, hop.literal('air'), [elementGraph])
]
elements.forEach((element) => {
element.addOut(rdf.type, hop.namedNode(VOCAB_PREFIX + 'Element'), [elementGraph])
})
console.log(elements.map((element) => element.toString()))This will log:
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>.
@prefix elements: <https://avatar.example/element#>.
@prefix vocab: https://avatar.example/vocab#>.
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
elements:water a vocab:Element ;
rdf:label "water"^^xsd:string .
elements:earth a vocab:Element ;
rdf:label "earth"^^xsd:string ,
elements:fire a vocab:Element ;
rdf:label "fire"^^xsd:string .
elements:air a vocab:Element ;
rdf:label "air"^^xsd:string .Already we can see how intuitive it is to set up an RDF graph. RDF Hop is designed to work with all the JavaScript paridigms you're used to.
Travserse a the Graph
Now that we've defined a graph and put it into a store, we can traverse the graph.
const water = hop.namedNode(ELEMENT_PREFIX + "water")
const earth = hop.namedNode(ELEMENT_PREFIX + "earth")
const fire = hop.namedNode(ELEMENT_PREFIX + "fire")
const air = hop.namedNode(ELEMENT_PREFIX + 'air')
console.log(fire.out(rdf.label).one().value()) // "fire^^xsd:string"Notice the out command. This will yeild all nodes that are the object of the provided predicate. Because this returns a collection of nodes, we call one() which picks a singular node from the collection. value yeilds the real string value of the literal.
const air = elementStore.getLiteral("air").in(rdf.label).one()Similarly, you can use the in method to trace the graph backwards.
Multiple Stores
You might want to keep graphs with overlapping data separate in the case you have multiple RDF files. To do so create a new store:
const CHARACTER_PREFIX = 'https://avatar.example/characters#'
const characterGraph = hop.graph('https://avatar.example/characters')
// Define Characters
const rawCharacterData: { name: string, elems: RDFNode[] }[] = [
{ name: 'Katara', elems: [water] },
{ name: 'Sokka', elems: [] },
{ name: 'Aang', elems: [water, earth, fire, air]},
{ name: 'Zuko', elems: [fire]},
{ name: 'Jet', elems: []},
{ name: 'Toph', elems: [earth]}
]
const characters = rawCharacterData.map(({ name, elems }) => {
const character = hop.namedNode(CHARACTER_PREFIX + name)
.addOut(rdf.type, foaf.Person, [characterGraph])
.addOut(foaf.name, hop.literal(name), [characterGraph])
elems.forEach((elem: RDFNode) => character.addOut(hop.namedNode(VOCAB_PREFIX + 'bends'), elem, [characterGraph]))
return character
})
// Ensure all characters know each other
characters.forEach((personA: NamedNode) => {
characters.forEach((personB: NamedNode) => {
if (personA !== personB) {
personA.addOut(foaf.knows, personB)
}
})
})
console.log(characterGraph.nodes().map(node => node.toString()))Notice that despite the elements being added as an object to the characters, the knows predicate will not be applied to them becasue these elements are not in the characterStore, and would not be included in the set yeilded by getAllNamedNodes()
Removing Data
Let's say it's mid season 2 and Toph doesn't know Zuko or Jet. We can remove her know connections:
const toph = hop.namedNode(CHARACTER_PREFIX + 'Toph')
const zuko = hop.namedNode(CHARACTER_PREFIX + 'Zuko')
const jet = hop.namedNode(CHARACTER_PREFIX + 'Jet')
toph.deleteOut(foaf.knows, zuko)
toph.deleteIn(foaf.knows, zuko)
toph.deleteOut(foaf.knows, jet)
toph.deleteIn(foaf.knows, jet)
console.log(characterGraph.nodes().map(node => node.toString()))
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent