@vocabulary/builder

Helper to create vocabulary packages, exactly like those in @zazuko/rdf-vocabularies

Usage

Fetch vocabulary

npx vocab-build datasets

Execute this in the directory where you have your vocabulary package.

In the package.json, add a configuration key which tells the CLI where to get the contents of your vocabularies. At the minimum, prefix and namespace are required.

{
  "name": "@my-vocab/foobar",
  "vocabulary": {
    "prefix": "foobar",
    "namespace": "https://example.com/foobar"
  }
}

When executed, the vocabulary will be dereferenced from the namespace URL and save it as {prefix}.nt. Additionally, a meta.nt document is created which is used by prefix.zazuko.com.

In many cases, vocabularies do not dereference nicely. In those circumstances, it is possible to add overrides:

Force a specific media type

When the server does not respond with the matching content-type header.

{
  "name": "@my-vocab/foobar",
  "vocabulary": {
    "prefix": "foobar",
    "namespace": "https://example.com/foobar",
    "mediaType": "application/rdf+xml"
  }
}

Override the URL

{
  "name": "@my-vocab/foobar",
  "vocabulary": {
    "prefix": "foobar",
    "namespace": "https://purl.org/ns/foobar",
    "file": "https://example.com/ontology.ttl"
  }
}

Set base URI for RDF/XML

RDF/XML sources frequently do not have the correct base

{
  "name": "@my-vocab/foobar",
  "vocabulary": {
    "prefix": "foobar",
    "namespace": "https://example.com/foobar",
    "xmlParserOptions": {
      "baseIRI": "https://example.com/foobar"
    }
  }
}

Multiple sources

{
  "name": "@my-vocab/foobar",
  "vocabulary": {
    "prefix": "foobar",
    "namespace": "https://example.com/foobar",
    "files": [
      {
        "mediaType": "application/rdf+xml",
        "file": "https://example.com/foobar-ext.rdf"
      },
      {
        "mediaType": "text/turtle",
        "file": "https://example.com/foobar.ttl"
      }
    ]
  }
}

Generate vocabulary module

npx vocab-build module

This will generate an index.ts with a single export of the entire vocabulary as an array of RDF/JS quads. Use it to load the vocabulary without the need for parsers.

import foobarQuads from '@my-vocab/foobar'
import $rdf from '@zazuko/env'

const dataset = $rdf.dataset(foobarQuads({ factory: $rdf }))