Ember CLI Rails

Ember CLI Rails is an integration story between (surprise suprise) Ember CLI and Rails 3.1 and up. It is designed to provide an easy way to organize your Rails backed Ember CLI application with a specific focus on upgradeability. Rails and Ember [slash Ember CLI] are maintained by different teams with different goals. As such, we believe that it is important to ensure smooth upgrading of both aspects of your application.

A large contingent of Ember developers use Rails. And Rails is awesome. With the upcoming changes to Ember 2.0 and the Ember community's desire to unify around Ember CLI it is now more important than ever to ensure that Rails and Ember CLI can coexist and development still be fun!

To this end we have created a minimum set of features (which we will outline below) to allow you keep your Rails workflow while minimizing the risk of upgrade pain with your Ember build tools.

For example, end-to-end tests with frameworks like Cucumber should just work. You should still be able leverage the asset pipeline, and all the conveniences that Rails offers. And you should get all the new goodies like ES6 modules and Ember CLI addons too! Without further ado, let's get in there!

Installation

Firstly, you'll have to include the gem in your Gemfile and bundle install

gem "ember-cli-rails"

Then you'll want to configure your installation by adding an ember.rb initializer. There is a generator to guide you, run:

rails generate ember-cli:init

This will generate an initializer that looks like the following:

EmberCLI.configure do |c|
  c.app :frontend
end

options

app - this represents the name of the Ember CLI application.
path - the path where your Ember CLI application is located. The default value is the name of your app in the Rails root.
enable - a lambda that accepts each request's path. The default value is a lambda that returns true.

EmberCLI.configure do |c|
  c.app :adminpanel # path is "<your-rails-root>/adminpanel"
  c.app :frontend,
    path: "/path/to/your/ember-cli-app/on/disk",
    enable: -> path { path.starts_with?("/app/") }
end

Once you've updated your initializer to taste, install Ember CLI if it is not already installed, and use it to generate your Ember CLI app in the location/s specified in the initializer. For example:

cd frontend
ember init

You will also need to install the ember-cli-rails-addon. For each of your Ember CLI applications, run:

npm install --save-dev ember-cli-rails-addon@0.0.12

And that's it! You should now be able to start up your Rails server and see your Ember CLI app.

Multiple Ember CLI apps

In the initializer you may specify multiple Ember CLI apps, each of which can be referenced with the view helper independently. You'd accomplish this like so:

EmberCLI.configure do |c|
  c.app :frontend
  c.app :admin_panel, path: "/somewhere/else"
end

Usage

You render your Ember CLI app by including the corresponding JS/CSS tags in whichever Rails view you'd like the Ember app to appear.

For example, if you had the following Rails app

# /config/routes.rb
Rails.application.routes.draw do
  root 'application#index'
end

# /app/controllers/application_controller.rb
class ApplicationController < ActionController::Base
  def index
    render :index
  end
end

and if you had created an Ember app :frontend in your initializer, then you could render your app at the / route with the following view:

<!-- /app/views/application/index.html.erb -->
<%= include_ember_script_tags :frontend %>
<%= include_ember_stylesheet_tags :frontend %>

Your Ember application will now be served at the / route.

Other routes

Rendering Ember applications at routes other than / requires additional setup to avoid an Ember UnrecognizedURLError.

For instance, if you had Ember applications named :frontend and :admin_panel and you wanted to serve them at /frontend and /admin_panel, you would set up the following Rails routes:

# /config/routes.rb
Rails.application.routes.draw do
  root 'application#index'
  get  'frontend'    => 'frontend#index'
  get  'admin_panel' => 'admin_panel#index'
end

# /app/controllers/frontend_controller.rb
class FrontendController < ActionController::Base
  def index
    render :index
  end
end

# /app/controllers/admin_panel_controller.rb
class AdminPanelController < ActionController::Base
  def index
    render :index
  end
end

Additionally, you would have to modify each Ember app's baseURL to point to the correct route:

/* /app/frontend/config/environment.js */
module.exports = function(environment) {
  var ENV = {
    modulePrefix: 'frontend',
    environment: environment,
    baseURL: '/frontend', // originally '/'
    ...
  }
}

/* /app/admin_panel/config/environment.js */
module.exports = function(environment) {
  var ENV = {
    modulePrefix: 'admin_panel',
    environment: environment,
    baseURL: '/admin_panel',  // originally '/'
    ...
  }
}

Lastly, you would configure each app's router.js file so that rootURL points to the baseURL you just created:

/* app/frontend/app/router.js */
var Router = Ember.Router.extend({
  rootURL:  config.baseURL, // add this line
  location: config.locationType
});

Repeat for app/admin_panel/app/router.js. Now your Ember apps will render properly at the alternative routes.

CSRF Tokens

Your Rails controllers, by default, are expecting a valid authenticity token to be submitted with non-GET requests. Without it you'll receive a 422 Unprocessable Entity error, specifically: ActionController::InvalidAuthenticityToken.

In order to add that token to your requests, you need to add into your template:

<!-- /app/views/application/index.html.erb -->
# ... your ember script and stylesheet includes ...
<%= csrf_meta_tags %>

This will add the tokens to your page.

You can then override the application DS.RESTAdapter (or whatever flavor of adapter you're using) to send that token with the requests:

// path/to/your/ember-cli-app/app/adapters/application.js
import DS from 'ember-data';
import $ from 'jquery';

export default DS.RESTAdapter.extend({
  headers: {
    "X-CSRF-Token": $('meta[name="csrf-token"]').attr('content')
  }
});

Ember Test Suite

To run an Ember app's tests in a browser, mount the EmberCLI::Engine:

# config/routes.rb

Rails.application.routes.draw do
  mount EmberCLI::Engine => "ember-tests" if Rails.env.development?

  root "application#index"
end

Ember tests are served based on the route you mount the Engine on (in this example, /ember-tests) and the name of the Ember app.

For example, to view tests of the frontend app, visit http://localhost:3000/ember-tests/frontend.

Enabling LiveReload

In order to get LiveReload up and running with Ember CLI Rails, you can install guard and guard-livereload gems, run guard init and then add the following to your Guardfile.

guard "livereload" do
  # ...
  watch %r{your-appname/app/\w+/.+\.(js|hbs|html|css|<other-extensions>)}
  # ...
end

This tells Guard to watch your Ember CLI app for any changes to the JavaScript, Handlebars, HTML, or CSS files within app path. Take note that other extensions can be added to the line (such as coffee for CoffeeScript) to watch them for changes as well.

NOTE: Ember CLI creates symlinks in your-appname/tmp directory, which cannot be handled properly by Guard. This might lead to performance issues on some platforms (most notably on OSX), as well as warnings being printed by latest versions of Guard. As a work-around, one might use directories option, explicitly specifying directories to watch, e.g. adding the following to the Guardfile.

# also add directories that need to be watched by other guard plugins
directories %w[app config lib spec your-appname/app]

Heroku

In order to deploy Ember CLI Rails app to Heroku:

First, enable Heroku Multi Buildpack by running the following command:

heroku buildpacks:set https://github.com/heroku/heroku-buildpack-multi

Next, specify which buildpacks to use by creating a .buildpacks file in the project root containing:

https://github.com/heroku/heroku-buildpack-nodejs
https://github.com/heroku/heroku-buildpack-ruby

Add rails_12factor gem to your production group in Gemfile, then run bundle install:

gem "rails_12factor", group: :production

Add a package.json file containing {} to the root of your Rails project. This is to make sure it'll be detected by the NodeJS buildpack.

Make sure you have bower as a npm dependency of your ember-cli app.

Add a postinstall task to your Ember CLI app's package.json. This will ensure that during the deployment process, Heroku will install all dependencies found in both node_modules and bower_components.

{
  # ...
  "scripts": {
    # ...
    "postinstall": "node_modules/bower/bin/bower install"¬
  }
}

ember-cli-rails adds your ember apps' build process to the rails asset compilation process.

Now you should be ready to deploy.

Experiencing Slow Build/Deploy Times?

Remove ember-cli-uglify from your package.json file, and run npm remove ember-cli-uglify. This will improve your build/deploy time by about 10 minutes.

The reason build/deploy times were slow is because ember uglified the JS and then added the files to the asset pipeline. Rails would then try and uglify the JS again, and this would be considerably slower than normal.

Additional Information

When running in the development environment, Ember CLI Rails runs ember build with the --output-path and --watch flags on. The --watch flag tells Ember CLI to watch for file system events and rebuild when an Ember CLI file is changed. The --output-path flag specifies where the distribution files will be put. Ember CLI Rails does some fancy stuff to get it into your asset path without polluting your git history. Note that for this to work, you must have config.consider_all_requests_local = true set in config/environments/development.rb, otherwise the middleware responsible for building Ember CLI will not be enabled.

Alternatively, if you want to override the default behavior in any given Rails environment, you can manually set the config.use_ember_middleware and config.use_ember_live_recompilation flags in the environment-specific config file.

`RAILS_ENV`

While being managed by EmberCLI Rails, EmberCLI process will have access to the RAILS_ENV environment variable. This can be helpful to detect the Rails environment from within the EmberCLI process.

This can be useful to determine whether or not EmberCLI is running in its own standalone process or being managed by Rails.

For example, to enable ember-cli-mirage API responses in development while being run outside of Rails (while run by ember serve), check for the absence of the RAILS_ENV environment variable:

// config/environment.js
if (environment === 'development') {
  ENV['ember-cli-mirage'] = {
    enabled: typeof process.env.RAILS_ENV === 'undefined',
  }
}

RAILS_ENV will be absent in production builds.

`SKIP_EMBER`

To disable asset compilation entirely, set an environment variable SKIP_EMBER=1.

This can be useful when an application's frontend is developed locally with EmberCLI-Rails, but deployed separately (for example, with ember-cli-deploy).

Ember Dependencies

Ember has several dependencies. Some of these dependencies might already be present in your asset list. For example jQuery is bundled in jquery-rails gem. If you have the jQuery assets included on your page you may want to exclude them from the Ember distribution. You can do so by setting the exclude_ember_deps option like so:

EmberCLI.configure do |c|
  c.app :frontend, exclude_ember_deps: "jquery"
  c.app :admin_panel, exclude_ember_deps: ["jquery", "handlebars"]
end

jQuery and Handlebars are the main use cases for this flag.

Contributing

Fork it (https://github.com/rwz/ember-cli-rails/fork)
Create your feature branch (git checkout -b my-new-feature)
Commit your changes (git commit -am 'Add some feature')
Push to the branch (git push origin my-new-feature)
Create a new Pull Request

benclum11 / ember-cli-rails