Huge News!Announcing our $40M Series B led by Abstract Ventures.Learn More
Socket
Sign inDemoInstall
Socket

magic_lamp

Package Overview
Dependencies
Maintainers
1
Alerts
File Explorer

Advanced tools

Socket logo

Install Socket

Detect and block malicious and high-risk dependencies

Install

magic_lamp

  • 1.9.0
  • Rubygems
  • Socket score

Version published
Maintainers
1
Created
Source

Magic Lamp

Gem Version Build Status

Magic Lamp helps you get your Rails templates into your JavaScript tests. This means that way your JavaScript tests break if you change your templates and you don't have to create so many fixtures. Plus, it lets you test your views in JavaScript. All you have to do is set up your data just like you would in a controller.

Table of Contents

  1. Installation
  2. Basic Usage
  3. Where the files go
  4. Tasks
  5. Ruby API
  6. JavaScript API
  7. Errors
  8. Sweet Aliases
  9. Contributing

Installation

Add this line to your application's Gemfile:

  gem "magic_lamp"

And then execute:

$ bundle install

Then paste mount MagicLamp::Genie, at: "/magic_lamp" if defined?(MagicLamp) into your config/routes.rb like so:

Rails.application.routes.draw do
  # ...
  mount MagicLamp::Genie, at: "/magic_lamp" if defined?(MagicLamp)
  # ...
end

This mounts the Magic Lamp engine in your app.

Then drop this:

//= require magic_lamp

at the top of your spec_helper.js (assuming you're using JavaScript spec runner for Rails that allows the use of Sprockets directives).

If your JavaScript test runner doesn't support XHR requests, you can drop this in instead:

//= require magic_lamp
//= require magic_lamp/all_fixtures

//= require magic_lamp/all_fixtures will load all of your fixtures without making any XHR requests.

Note: when loading your fixtures via Sprockets directives you will likely want to disable Sprockets caching in whatever environment your JavaScript tests are running. You can do so by putting the following in the appropriate environment config file:

config.assets.configure do |env|
  env.cache = ActiveSupport::Cache.lookup_store(:null_store)
end

Now you've got the basic setup.

Debugging

Visit /magic_lamp/lint in your browser to lint your fixtures. You can also run rake magic_lamp:lint (or rake mll for short) to lint your fixtures from the command line.

Loading Helpers

Simply load your helpers in the magic_lamp_config.rb file like so:

# in magic_lamp_config.rb
Dir[Rails.root.join("spec", "support", "magic_lamp_helpers/**/*.rb")].each { |f| load f }

Supported test runners

Magic Lamp has been shown to work with:

It will likely work with any test runner that supports XHR requests or Sprockets directives.

With Database Cleaner

You don't need Database Cleaner to use this gem, but this is probably the setup most people want.

First make sure you have Database Cleaner installed, then you'll want to do something like this:

require "database_cleaner"

MagicLamp.configure do |config|

  DatabaseCleaner.strategy = :transaction

  config.before_each do
    DatabaseCleaner.start
  end

  config.after_each do
    DatabaseCleaner.clean
  end
end

in a file called magic_lamp_config.rb which you can place anywhere in your spec or test directories.

This way you can take advantage of after_create callbacks for your fixture setup without polluting your database every time you run your JavaScript specs.

With FactoryGirl

If you're using FactoryGirl, Magic Lamp will call FactoryGirl.reload for you to save you some autoloader pain.

Basic Usage

Magic Lamp will load all files in your spec or test directory that end with _lamp.rb (your app's "lamp files). I'd recommend starting with a single magic_lamp.rb file and breaking it into smaller files once it gets unwieldy (one for each controller would be a good approach).

In your lamp files you just call MagicLamp.fixture like so:

MagicLamp.fixture do
  @order = Order.new
  render partial: "orders/form"
end

Inside the block you pass to fixture you're in the scope of a controller so you can set up any instance variables your templates depend on. In this case we're using the default controller which is your ApplicationController. We're also using the default name for the fixture which is whatever render receives to identify the template (ie the symbol or string argument to render or whatever is at the :template or :partial key in the argument hash).

render here also works normally except that it won't render the layout by default.

Then in your JavaScript you can call load:

beforeEach(function() {
  MagicLamp.load("orders/form");
});

which will put the orders/form partial in a div with a class of magic-lamp (this all happens synchronously). Then you can go nuts testing your JavaScript against your actual template. If you'd like to only make one request for your templates, simply call MagicLamp.preload(); in your spec_helper.js to populate Magic Lamp's cache.

Loading multiple templates

Just pass more fixture names to MagicLamp.load and it will load them all. For example:

  beforeEach(function() {
    MagicLamp.load("orders/sidebar", "orders/form");
  });

with a sidebar template/partial that looks like this:

  <div class="sidebar content">Links!</div>

and a form template/partial that looks like this:

  <div class="form content">Inputs!</div>

will yield:

  <div class="magic-lamp">
    <div class="sidebar content">Links!</div>
    <div class="form content">Inputs!</div>
  </div>

Loading JSON fixtures and arbitrary strings

If you pass a block to fixture that does not invoke render, Magic Lamp will assume that you want the to_json representation of the return value of the block as your fixture (Magic Lamp will call this for you). If the return value is already a string, then Magic Lamp will return that string as is. In your JavaScript MagicLamp.loadJSON will return the JSON.parsed string while MagicLamp.loadRaw will return the string as is (you can also use loadRaw get the string version of a template rendered with render without Magic Lamp appending it to the DOM. loadJSON also won't interact with the DOM).

When sending down JSON or arbitrary strings, you must provide the fixture with a name since inferring one is impossible.

It's also good to remember that in the fixture block that even though the controller isn't rendering anything for us, the block is still scoped to the given controller which gives us access to any controller methods we might want to use to massage our data structures.

For example:

MagicLamp.define(controller: OrdersController) do
  fixture(name: "some_json") do
    OrderSerializer.new(Order.new(price: 55))
  end

  fixture(name: "some_arbitrary_string") do
    some_method_on_the_controller_that_returns_a_string("Just some string")
  end
end

Then in your JavaScript:

beforeEach(function() {
  var jsonObject = MagicLamp.loadJSON("orders/some_json");
  var someString = MagicLamp.loadRaw("orders/some_arbitrary_string");
});

A few more examples

Here we're specifying which controller should render the template via the arguments hash to fixture. This gives us access to helper methods in the fixture block and in the template. It also means we don't need a fully qualified path to the rendered template or partial.

Since we didn't specify the name of our fixture and we're not using the ApplicationController to render the template the fixture will be named "orders/order".

We're also taking advantage of render's :collection option.

MagicLamp.fixture(controller: OrdersController) do
  orders = 3.times.map { Order.new }
  render partial: "order", collection: orders
end

Here we're specifying a name with the :name option that's passed to fixture. This way we can load the fixture in our JavaScript with MagicLamp.load("custom/name") instead of the default MagicLamp.load("orders/foo"). Custom names for fixtures must be url safe strings. We're also extending the controller and its view with the AuthStub module to stub some methods that we don't want executing in our fixtures.

MagicLamp.fixture(name: "custom/name", extend: AuthStub) do
  render "orders/foo"
end

Here we're specifying both a controller and custom name. We're also setting the params[:foo] mostly to demonstrate that we have access to all of the usual controller methods.

MagicLamp.fixture(controller: OrdersController, name: "other_custom_name") do
  params[:foo] = "test"
  render :foo
end

Drying up your fixtures

If you have several fixtures that depend on the same setup (same controller, extensions, etc), you can use the define method to dry things up:

MagicLamp.define(controller: OrdersController, extend: AuthStub) do
  fixture do # orders/new
    @order = Order.new
    render :new
  end

  fixture(name: "customized_new") do # orders/customized_new
    session[:custom_user_info] = "likes movies"
    @order = Order.new
    render :new
  end

  fixture do # orders/form
    @order = Order.new
    render partial: "form"
  end

  define(namespace: "errors", extend: SomeErrorHelpers) do
    fixture(name: "form_without_price") do # orders/errors/form_without_price
      @order = Order.new
      @order.errors.add(:price, "can't be blank")
      render partial: "form"
    end

    fixture do # orders/errors/form
      @order = Order.new
      @order.errors.add(:address, "can't be blank")
      render partial: "form"
    end
  end
end

Where the files go

Config File

Magic Lamp first loads the magic_lamp_config.rb file. It can be anywhere in your spec or test directory but it's not required.

Lamp files

Magic Lamp will load any files in your spec or test directory that end with _lamp.rb.

Tasks

fixture_names

Call rake magic_lamp:fixture_names (or rake mlfn) to see a list of all of your app's fixture names.

lint

Call rake magic_lamp:lint (or rake mll) to see if there are any errors when registering or rendering your fixtures.

Ruby API

fixture

(also aliased to register_fixture and register)

It requires a block that is invoked in the context of a controller. If render is called, it renders the specified template or partial the way the controller normally would minus the layout. If you'd like a particular layout to be rendered simply specify it when you call render:

render :index, layout: "admin"

Note: Rendering the layout could be useful under some circumstances, but it is important to be aware that things will get weird if you use MagicLamp to load and entire HTML document. Especially if it contains script tags to your JavaScript.

If render is not called in the block then MagicLamp will render the to_json representation of the return value of the block unless the return value is already a string. In that case, the string is rendered as is.

It also takes an optional hash of arguments. The arguments hash recognizes:

  • :controller
    • specifies any controller class that you'd like to have render the template or partial or have the block scoped to.
    • if specified it removes the need to pass fully qualified paths to templates to render
    • the controller's name becomes the default namespace, ie OrdersController provides a default namespace of orders resulting in a template named orders/foo
  • :name
    • whatever you'd like name the fixture.
    • Specifying this option also prevents the block from being executed twice which could be a performance win. See configure for more.
    • this is required when you want to send down JSON or arbitrary strings.
  • :extend
    • takes a module or an array of modules
    • extends the controller and view context (via Ruby's extend) Also note that only symbol keys are recognized.

fixture will also execute any callbacks you've specified. See configure for more.

Example:

MagicLamp.fixture(name: "foo", controller: OrdersController) do
  @order = Order.new
  render partial: "form"
end

define

Allows you scope groups of fixtures with defaults and can be nested arbitrarily. It takes an optional hash and a required block. The hash accepts the following options:

  • :controller
    • specifies any controller class that you'd like to have render the template or partial.
    • if specified it removes the need to pass fully qualified paths to templates to render
    • the controller's name becomes the default namespace if no namespace is provided, ie OrdersController provides a default namespace of orders resulting in a template named orders/foo
  • :name
    • whatever you'd like name the fixture.
    • Specifying this option also prevents the block from being executed twice which could be a performance win. See configure for more.
  • :extend
    • takes a module or an array of modules
    • extends the controller and view context (via Ruby's extend)
  • :namespace
    • namespaces all fixtures defined within it
    • overrides the default controller namespace if passed Also note that only symbol keys are recognized.

Inside of the block you can nest more calls to define and create fixtures via the fixture method or one of its aliases.

Example:

module DefinesFoo
  def foo
    "foo!"
  end
end

module AlsoDefinesFoo
  def foo
    "also foo!"
  end
end

MagicLamp.define(controller: OrdersController, extend: DefinesFoo) do

  fixture do # orders/edit
    foo #=> "foo!"
    @order = Order.create!
    render :edit
  end

  fixture do # orders/new
    foo #=> "foo!"
    @order = Order.new
    render :new
  end

  define(namespace: "errors", extend: AlsoDefinesFoo) do

    fixture do # orders/errors/edit
      foo #=> "also foo!"
      @order = Order.create!
      @order.errors.add(:price, "Can't be negative")
      render :edit
    end

    fixture(extend: DefinesFoo) do # orders/errors/new
      foo #=> "foo!"
      @order = Order.new
      @order.errors.add(:price, "Can't be negative")
      render :new
    end
  end
end

configure

It requires a block to which it yields the configuration object. Here you can set:

  • before_each
    • takes a block
    • defaults to nil
    • called before each block you pass to fixture is executed
    • note: if you call it a second time with a second block, only the second block will be executed
  • after_each
    • takes a block
    • defaults to nil
    • called after each block you pass to fixture is executed
    • note: if you call it a second time with a second block, only the second block will be executed
  • global_defaults
    • can be set to a hash of default options that every fixture generated will inherit from. Options passed to define and fixture take precedence.
    • accepts all of the keys define accepts
  • infer_names
    • defaults to true
    • if set to true, Magic Lamp will try to infer the name of the fixture when not provided with a name parameter.
    • if set to false, the name parameter becomes required for MagicLamp.fixture (this can be done to improve performance or force your team to be more explicit)

Example:

# spec/support/magic_lamp/helpers/auth_stub.rb
module AuthStub
  def current_user
    @current_user ||= User.create!(
      email: "foo@example.com",
      password: "password"
    )
  end
end

# spec/magic_lamp_config.rb (can be anywhere in /spec)
MagicLamp.configure do |config|

  Dir[Rails.root.join("spec", "support", "magic_lamp_helpers/**/*.rb")].each { |f| load f }

  # if you want to require the name parameter for the `fixture` method
  config.infer_names = false

  config.global_defaults = { extend: AuthStub }

  config.before_each do
    puts "I appear before the block passed to `fixture` executes!"
  end

  config.after_each do
    puts "I appear after the block passed to `fixture` executes!"
  end
end

JavaScript API

clean

Calling MagicLamp.clean() will remove the Magic Lamp fixture container from the page.

If you don't want any dom elements from a fixture hanging around between specs, throw it in a global afterEach block. Calling it with nothing to clean won't result in an error.

Example:

  afterEach(function() {
    MagicLamp.clean();
  });

load

Call MagicLamp.load to load a fixture. It requires the name of the fixture and which will be loaded into a div with a class of magic-lamp. It will destroy the previous fixture container if present so you never end up with duplicate fixture containers or end up with dom elements from previous loads. It will hit the network only on the first request for a given fixture. If you never want load to hit the network, call MagicLamp.preload() or use //= require magic_lamp/all_fixtures before your specs.

You can load multiple fixtures into the dom at the same time by simply passing more arguments to load.

The call to get your template is completely synchronous.

Example:

  beforeEach(function() {
    MagicLamp.load("orders/foo");
  });

  // or if you wanted multiple fixtures...

  beforeEach(function() {
    MagicLamp.load("orders/foo", "orders/bar", "orders/foo", "orders/baz");
  });

loadJSON

Returns the JSON.parsed version of the fixture. It's a convenience method for JSON.parse(MagicLamp.loadRaw('some_json_fixture'));. Look here for more.

loadRaw

Returns the template, partial, JSON, or string as a raw string. Look here for more.

preload

Call MagicLamp.preload to load all of your templates into MagicLamp's cache. This means you'll only hit the network once, so the rest of your specs will be quicker and you can go wild stubbing the network.

The call to get your templates is completely synchronous.

Note: this is completely unnecessary if you use //= require magic_lamp/all_fixtures

Example:

// Probably should be in your `spec_helper.js`
MagicLamp.preload();

fixtureNames

MagicLamp.fixtureNames() will return an array of all of the fixture names available in the cache (which is all of them if you've called preload). It will also console.log them out.

Example

MagicLamp.preload();
MagicLamp.fixtureNames(); // => ['orders/foo', 'orders/bar', 'orders/baz']
// logs 'orders/foo'
// logs 'orders/bar'
// logs 'orders/baz'

globalize

MagicLamp.globalize() will put MagicLamp.clean and MagicLamp.load onto window for convenience.

Example:

MagicLamp.globalize();

describe("Foo", function() {
  beforeEach(function() {
    load("orders/foo");
  });

  afterEach(function() {
    clean();
  });

  // ...
});

Errors

If there are errors rendering any of your templates, Magic Lamp will often throw a JavaScript error. Errors will also appear in your server log (if you're running the in-browser specs).

To see errors outside of the server log (which may be noisy), you can run rake magic_lamp:lint (or rake mll) or visit /magic_lamp/lint in your browser and display any errors in your fixtures.

You can debug into your fixture generation code or even the templates themselves just as you normally would in any other part of your Rails app. If you start seeing a bunch of RenderCatcher objects instead of the code you expected, just continue through and wait for the debugger to catch again and things go back to normal. If you want to skip continuing through, simply provide an explicit name for the fixture.

If you get an ActionView::MissingTemplate error, try specifying the controller. This error is caused by a template that renders a partial without using the fully qualified path to the partial. Specifying the controller should help Rails find the template.

When rails is inferring routes for form_for, link_to, and a few other methods you'll sometimes need to change request.env["action_dispatch.request.path_parameters"] by setting action and controller. When the route is nested with with a parameter in the url like /orders/:order_id/foos/new you'll need to add the :order_id to it like so:

fixture do
  request.env["action_dispatch.request.path_parameters"] = {
    action: "new",
    controller: "foos",
    order_id: 3
  }
  render :new
end

Sweet aliases

Ruby

MagicLamp.register_fixture => fixture
MagicLamp.register => fixture
MagicLamp.rub => fixture
MagicLamp.wish => fixture

JavaScript

MagicLamp.rub => load
MagicLamp.wish => load
MagicLamp.massage => preload
MagicLamp.wishForMoreWishes => preload
MagicLamp.polish => clean

Rake Tasks

rake mlfn => rake magic_lamp:fixture_names
rake mll => rake magic_lamp:lint

Contributing

  1. Fork it
  2. Clone it locally
  3. cd into the project root
  4. ruby bootstrap
  5. appraisal install
  6. Run the specs with appraisal rake
  7. Update the CHANGELOG.md
  8. Create your feature branch (git checkout -b my-new-feature)
  9. Commit your changes (git commit -am 'Add some feature')
  10. Push to the branch (git push origin my-new-feature)
  11. Create new Pull Request

FAQs

Package last updated on 26 Feb 2019

Did you know?

Socket

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Install

Related posts

SocketSocket SOC 2 Logo

Product

  • Package Alerts
  • Integrations
  • Docs
  • Pricing
  • FAQ
  • Roadmap
  • Changelog

Packages

npm

Stay in touch

Get open source security insights delivered straight into your inbox.


  • Terms
  • Privacy
  • Security

Made with ⚡️ by Socket Inc