Backbone.conduit NPM

Backbone.Conduit

Moving Data Through Backbone

Conduit is a Backbone plugin that improves the ability of Backbone to handle large scale data sets.

Install it with npm or bower:

$ bower install backbone.conduit
... or ...
$ npm install backbone.conduit

Option #1: Faster Model Creation

Conduit.QuickCollection is optimized to create Backbone.Model instances faster:

var collection = new Backbone.Conduit.QuickCollection();

// If you have a large amount of data injected onto the page, instead of 'reset(...)' do ...
var aBigArray = [ {}, {}, {} ]; // let's assume this was 100K objects
collection.refill(aBigArray);

// Or, if you need to get it asynchronously, instead of 'fetch()' do ...
collection.haul();

Performance varies, but typically loading data into a Conduit.QuickCollection is ~ 40% faster than a Backbone.Collection.

Option #2: Async Data Management + Deferred Model Creation

Conduit.SparseCollection drastically changes how data is managed in a Collection. Raw data is managed in a web worker, and model creation only happens when data is prepared for use in a View. Sort, Filter, Map and Reduce, and other methods are asynchronous, using Promises to manage the flow:

var MyCollection = Backbone.Conduit.SparseCollection.extend({

    // Assume this endpoint returns 100K items:
    url: '/some/enormous/data',
    
    // ...
});

Backbone.Conduit.enableWorker({
    paths: '/your/path/to/backbone.conduit'
}).then(function() {
    var collection = new MyCollection();
    return collection.haul();
}).then(function() {
   console.log('Length: ' + collection.length); // <== "Length: 100000"

   // Prepare the first 10 models for use
   return collection.prepare({
       indexes: { min: 0, max: 9}
   });
}).then(function(models) {
  console.log('Prepared: ' + models.length); // <== "Prepared: 10"
  // Note the prepared models are also available via
  // 'collection.get(...)' or 'collection.at(...)';
});

Since model creation happens at the last possible moment, and the worker thread handles the data management, a SparseCollection can handle hundreds of thousands of items easily.