tests normalized vs denormalized+gzip data

denormalized.js

const {writeFileSync} = require('fs');
const { normalize, schema } = require('normalizr');
const faker = require('faker');
const murmur = require("murmurhash-js");

const genId = (data) => murmur(JSON.stringify(data));

// Define a users schema
const user = new schema.Entity('users', undefined, {
    idAttribute: genId
});

// Define your comments schema
const comment = new schema.Entity('comments', {
  author: user
});

// Define your article
const article = new schema.Entity('articles', {
  authors: [user],
  comments: [comment]
});

function range(min, max) {
    const count = max - min;
    return Array
      .from({length: (faker.random.number() % count) + min}, (_, i) => i + min)
}

function pickSample(data, min, max) {
  const count = (max||data.length) - min;
  return Array
    .from({length: (faker.random.number() % count) + min}, () => faker.random.number() % data.length)
    .map(i => data[i]);
}


function generateUser() {
  return {
    name: faker.name.findName(), // Rowan Nikolaus
    email: faker.internet.email(), // Kassandra.Haley@erich.biz
    card: faker.helpers.userCard() // random contact card containing many properties
  };
}

function generateComment(users) {
  return {
    id: faker.random.uuid(),
    body: faker.lorem.paragraph(),
    author: users[faker.random.number()%users.length],
  };
}

function generateArticle(users) {
    const comments = range(0, 5)
        .map(() => generateComment(users));

    const authors = pickSample(users, 2, 5);

    return {
      id: faker.random.uuid(),
      slug: faker.lorem.slug(),
      body: faker.lorem.paragraphs(),
      comments,
      authors
    };
}

function generateArticles() {
    const users = Array.from({length: 100}, generateUser);
    return Array.from({length: 100}, () => generateArticle(users))
}


const data = generateArticles();
const normalized = normalize(data, [article]);

writeFileSync('./in.json', JSON.stringify(data, null, 2));
writeFileSync('./normalized.json', JSON.stringify(normalized, null, 2));

// gzip -c normalized.json > normalized.json.gz && gzip -c in.json > in.json.gz

532K input.json (not normalized)
128K input.json.gz (not normalized)

256K normalized.json (50% reusability)
128K normalized.json.gz