Fetch, extract, parse, expand, frame and compact JSON-LD

json-ld.js

const { JSDOM } = require('jsdom')
const { compact, expand, frame } = require('jsonld')

const url = 'https://www.bbc.co.uk/schedules/p00fzl6p/2020/06/14'

// fetch and parse HTML
const { window: { document } } = await JSDOM.fromURL(url)

// select the script elements containing JSON-LD
const elements = document.querySelectorAll('script[type="application/ld+json"]')

for (const element of elements) {
    // parse the JSON
    const doc = JSON.parse(element.textContent)
    console.log({ doc })

    // expand the doc to use full URIs for property names
    // const expanded = await expand(doc)
    // console.log({ expanded })

    // frame the doc to filter and pick certain properties
    const framed = await frame(doc, {
      '@context': 'https://schema.org/',
      '@type': 'TVEpisode',
      '@explicit': true,
      identifier: {},
      description: {},
      name: {},
      published: {},
      image: {},
      partOfSeries: {
       '@type': 'TVSeries',
       '@explicit': true,
       '@default': [],
        identifier: {},
        name: {},
      },
      partOfSeason: {
       '@type': 'TVSeason',
       '@explicit': true,
       '@default': [],
        identifier: {},
        name: {},
        position: {},
      },
    })
    console.log({ framed })

    // compact the framed doc to use own property names
    const compacted = await compact(framed, {
      '@context': {
        identifier: 'http://schema.org/identifier',
        description: 'http://schema.org/description',
        name: 'http://schema.org/name',
        series: 'http://schema.org/partOfSeries',
        season: 'http://schema.org/partOfSeason',
        position: 'http://schema.org/position',
        image: {
          '@type': '@id',
          '@id': 'http://schema.org/image',
        },
        published: {
          '@type': 'http://schema.org/Date',
          '@id': 'http://schema.org/datePublished',
        },
      },
    })
    console.log({ compacted })
}