@inrupt/solid-common-vocab

The Solid Common Vocab library for JavaScript

A library providing JavaScript objects to represent the individual terms (i.e., the classes and properties) defined in RDF vocabularies, both existing vocabularies (like http://schema.org, FOAF, vCard, LDP, ActivityStreams, etc.), and your own custom RDF vocabularies.

A major feature of this library is that it provides easy access to any rdfs:label and rdfs:comment values provided for these vocabulary terms, and provides very easy-to-use support for multilingual values for these labels and comments. We also support other term metadata (such as rdfs:seeAlso and rdfs:isDefinedBy), and also generic message strings (using SKOS-XL) that can be used for error message strings, or labels, or tooltip text for user interfaces, etc.

Setup

The demo directory provides a working example that you can run with the following commands (starting from the root of this project, which we need to build first):

npm ci
npm run build

cd demo
npm ci
node index.js

The solid-common-vocab library is distributed as a GitHub npm package: @inrupt/solid-common-vocab. For more information about GitHub npm packages, please visit the dedicated documentation.

Note: This library is used extensively by the Artifact Generator project that can automatically generate source-code (in multiple programming languages, including JavaScript and TypeScript) that can provide Vocab Term instances for every term defined within any RDF vocabulary. Due to the ease of simply pointing the Artifact Generator at any RDF vocabulary and having it automatically generate all the Vocab Term instances for you, we don't expect manual instantiation of Vocab Terms to be very common. However, this documentation describes the Vocab library without any dependency or requirement to use the Artifact Generator at all.

Advanced usage by demonstration

For more detailed examples that go beyond the common uses featured here, please see the demonstration test suite.

That test code can be run using the following (assuming the base project has been built first!):

jest demonstrateByUsage.test.js

RDF library support

The Vocab Term object from this library is intended to be a simple wrapper around the 'NamedNode' object conforming to the RDF/JS interface.

This means that Vocab Term instances can be used natively with libraries that are RDF/JS-compliant, such as rdf-ext, rdflib.js, rdf-data-factory, graphy, etc. An instance of a VocabTerm may be built by passing an RDF/JS DataFactory implemented by any library, but it also includes its own very basic DataFactory implementation, just for convenience if you don't wish to include an existing implementation.

Introductory example

For example, if we have the following simple RDF vocabulary defining a single Person term (in this case an RDF Class):

prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
prefix ex:   <https://example.com#>

ex:Person a rdfs:Class ;
  rdfs:label "My Person class"@en ;
  rdfs:comment "Full description of my Person class..."@en .

We could represent this as a Vocab Term in JavaScript like so:

const {VocabTerm, buildStore} = require('@inrupt/solid-common-vocab')
// Any other implementation of the RDF/JS interfaces would also be appropriate.
const rdf = require('rdfFactory-ext')

// The third argument provides as a context - it will commonly store things like
// the current language preference of the user, which can be used to lookup term
// labels or comments in that language. It's always there for browsers, but in
// NodeJS we expose a local implementation accessible through the method
// `buildStore`, which returns either said local implementation or the browser
// store depending on the environment.
// The last parameter indicates whether we want a 'strict' behaviour or not
// (see below for an explanation).  
const person = new VocabTerm('https://example.com#Person', rdf, buildStore(), true)
  .addLabel('My Person class','en')
  .addComment('Full description of my Person class...','en')

We can use this Vocab Term in various ways:

// To access the term's full IRI value:
const personIri = person.value

// The label and the comment are available as RDF/JS RDFLiteral instances:
// - get the RDFLiteral object (which contains not just the text value, but also
// the language tag of that text (e.g. 'en' for English, or 'es' for Spanish).
// Solid Common can potentially offer further meta-data - such as a description
// of how the text was determined. For example if a user's current language
// preference (as stored in localStorage) was 'French', but our original RDF
// vocabulary didn't provide a French label (in which case the vocab term will
// fallback to using an English label by default), then we can describe that
// behaviour in another field saying: "Current language is French, but only
// German, Spanish and English labels are available: using English", which can
// be extremely useful in a User Interface tooltip, for instance):
const personLabel = person.label
const personComment = person.comment

// Get the term's label or comment as a simple string value:
const personLabelAsString = person.label.value
const personCommentAsString = person.comment.value

To use the included DataFactory implementation to build a VocabTerm, the previous example would become:

const {buildBasicTerm, buildStore} = require('@inrupt/solid-common-vocab')

const person = buildBasicTerm('https://example.com#Person', buildStore(), true)
  .addLabel('My Person class','en')
  .addComment('Full description of my Person class...','en')

Note: The solid-common-vocab library is implemented in TypeScript, and embeds its types. The following code snippet demonstrates basic TypeScript usage:

import {buildBasicTerm, buildStore, VocabTerm} from '@inrupt/solid-common-vocab'

const person: VocabTerm = buildBasicTerm(
  'https://example.com#Person',
  buildStore(),
  true
).addLabel('My Person class','en')
.addComment('Full description of my Person class...','en')

Messages

An important feature of solid-common-vocab is support for parameterized messages. This can be extremely useful when creating your own RDF vocabularies and including message strings (thereby providing those messages with globally unique IRI identifiers, and allowing for easy translations of those message strings). For instance, to report errors to the user with contextual information (and in multiple languages):

const term = new VocabTerm("https://test.com/vocab#Unauthorized", rdf, buildStore(), true)
    .addMessage('Your account ({{0}}), does not have sufficient credentials for this operation', 'en')
    .addMessage('Votre compte ({{0}}) ne dispose pas des informations d'identification suffisantes pour cette opération', 'fr')
    
term.messageParams('Current Account').value // Evaluates to "Your account (Current Account)..."

Multilingual support

Unless we explicitly mandate a specific language, English will be used by default. Best practice for RDF vocabularies in general is to provide labels (short human-readable descriptions) and comments (longer, more detailed descriptions), and to also provide these descriptions in multiple languages if appropriate and possible.

(Technical note: the language tag defaults to an empty string in the case of falling back to the local part of the term's IRI (see the next section about strictness)).

const storage = buildStore()
const person = new VocabTerm('https://example.com#Person', rdf, storage, true)
  .addLabel('Person','en')
  .addLabel('Personne', 'fr')
  .addLabel('Persona', 'es')

// Default to the English label (if there is one).
var personLabel = person.label

// Request an explicit language for the label - but if there isn't one, we'll 
// fallback to the English one, if there is one).
personLabel = person.asLanguage('fr').label

// Change the default language in our context (i.e. localStorage).
storage.setItem(VocabContext.CONTEXT_KEY_LOCALE, 'es')

personLabel = person.label // personLabel now contains the Spanish literal.

Strictness

The last parameter to the Vocab Term constructor indicates if the behaviour of the term should be strict or loose. In the case of 'loose' behaviour, in the absence of any label, term.label will default to the local part of the term's IRI (i.e., the last segment of the full path component). With 'strict' behaviour it will return undefined. When the local part of the IRI is returned as a label, the language tag will be empty (i.e., "").

// Here we specify 'loose' behaviour(i.e. 'false' parameter to constructor)...
var person = new VocabTerm('https://example.com#Person', rdf, buildStore(), false)

// 'personLabel' will default to an RDF literal with the value "Person", and an
// empty language tag (i.e. "").
var personLabel = person.labelLiteral 
 
// Now strict behaviour...
person = new VocabTerm('https://example.com#Person', rdf, buildStore(), true)
// personLabel will default to 'undefined'.
personLabel = person.labelLiteral

This behaviour (i.e. returning the local part of the IRI, or undefined) may be overridden to instead throw an error when no label is found by using the .mandatory accessor.

// Here 'strictness' has no impact...
const person = new VocabTerm('https://example.com#Person', rdf, buildStore(), true)

// An exception will be thrown here, because our term has no label.
const personLabel = person.mandatory.label