remi/architect

Your web workers’ supervisor

Architect is a JavaScript library built on top of Web Workers.
He will manage and polyfill them workers so you don’t have to.

Short-lived workers

These will be automatically terminated as soon as the work is done. It will spawn a new worker every time.

proxy

Returns anything it receives in a background process. Useful when dealing with heavy DOM manipulation (i.e. Infinite scroll). It greatly improves initial page load speed, especially on mobiles.

var images = ['foo.png', 'bar.png', 'twiz.png', 'foozle.png', 'barzle.png', 'twizle.png']
Architect.proxy(images, function(data) {
  console.log(data)
  // => ['foo.png', 'bar.png', 'twiz.png', 'foozle.png', 'barzle.png', 'twizle.png']

  data.forEach(function(image) {
    img = document.createElement('img')
    img.src = image

    document.body.appendChild(img)
  })
})

Alias for Architect.work(data, 'proxy', callback).

ajax

Makes an Ajax request in a background process.

Architect.ajax('/users/1', function(data) {
  console.log(data);
  // => { id: 1, name: 'Foo', email: '[email protected]' }
})

Alias for Architect.work(url, 'ajax', callback).

jsonp

Makes a JSONP request in a background process. Do not add ?callback=foo to your URL, Architect will handle JSONP callbacks himself. See the request Architect makes.

Architect.jsonp('https://api.github.com/users/etiennelem', function(data) {
  console.log(data);
  // => { meta: { status: 200, … }, data: { login: 'EtienneLem', company: 'Heliom', … } }
})

Alias for Architect.work(url, 'jsonp', callback).

Long-lived workers

These need to be manually terminated when the work is done. The same worker can therefore be reused many times with different data.

proxyOn

var images, jobName, imagesCount

images = ['foo.png', 'bar.png', 'twiz.png', 'foozle.png', 'barzle.png', 'twizle.png']
jobName = 'appendImages'
imagesCount = 0

images.forEach(function(image) {
  Architect.proxyOn(jobName, image, function(data) {
    imagesCount++

    img = document.createElement('img')
    img.src = data
    document.body.appendChild(img)

    if (imagesCount == images.length) { Architect.endJob(jobName) }
  })
})

Alias for Architect.workOn(jobName, data, 'proxy', callback).

ajaxOn

var totalPages, jobName, queryApi

totalPages = 10
jobName = 'getUsers'

queryApi = function(page) {
  Architect.ajaxOn(jobName, '/users?page=' + page, function(data) {
    // [Add DOM elements, do your thing ;)]

    if (page == totalPages) {
      // Manually terminate the 'getUsers' ajax worker
      Architect.endJob(jobName)
      console.log('Done')
    } else {
      // Reuse the same worker
      queryApi(page + 1)
    }
  })
}

queryApi(1)

Alias for Architect.workOn(jobName, url, 'ajax', callback).

jsonpOn

Architect.jsonpOn('profile', 'https://api.github.com/users/etiennelem', function(data) {
  console.log(data);
  // => { meta: { status: 200, … }, data: { login: 'EtienneLem', company: 'Heliom', … } }

  Architect.endJob('profile')
})

Alias for Architect.workOn(jobName, url, 'jsonp', callback).

Custom workers

You can use Architect with your own workers. Just remember that if you want to be compatible with all the old browsers you need to optionally provide a fallback function that replicates your worker’s work.

workFrom

// workers/foozle.js
addEventListener('message', function(e) {
  data = (e.data + 'zle').toUpperCase()
  postMessage(data)
})

// application.js

// Replicates workers/foozle.js, but in the main thread
var foozleFallback = function(data) {
  return (data + 'zle').toUpperCase()
}

Architect.workFrom('workers/foozle.js', 'foo', foozleFallback, function(data) {
  console.log(data)
  // => FOOZLE
})

Setup

Rails

1. Add gem 'architect' to your Gemfile.
2. Add //= require architect to your JavaScript manifest file (usually found at app/assets/javascripts/application.js).
3. Restart your server and you'll have access to your very own Architect!

Other

You’ll need to serve the worker files at /architect (i.e. http://foo.com/architect/proxy_worker.min.js) and manually include architect.min.js to your HTML pages.

Custom path

You can also specify any path you want to serve the workers from.

Architect.setupWorkersPath('fake/path')
Architect.proxy('Foo', function(data) {
  // => Uses http://yourdomain.com/fake/path/proxy_worker.min.js
})

Tests

Run the rake test task.