di.ngdoc

@ngdoc overview
@name Developer Guide: Dependency Injection
@description

# Dependency Injection

Dependency Injection (DI) is a software design pattern that deals with how code gets hold of its
dependencies.

For in-depth discussion about DI, see {@link http://en.wikipedia.org/wiki/Dependency_injection
Dependency Injection} at Wikipedia, {@link http://martinfowler.com/articles/injection.html
Inversion of Control} by Martin Fowler, or read about DI in your favorite software design pattern
book.

## DI in a nutshell

There are only three ways an object or a function can get a hold of its dependencies:

  1. The dependency can be created, typically using the `new` operator.

  2. The dependency can be looked up by referring to a global variable.

  3. The dependency can be passed in to where it is needed.


The first two options of creating or looking up dependencies are not optimal because they hard
code the dependency. This makes it difficult, if not impossible, to modify the dependencies.
This is especially problematic in tests, where it is often desirable to provide mock dependencies
for test isolation.

The third option is the most viable, since it removes the responsibility of locating the
dependency from the component. The dependency is simply handed to the component.

<pre>
  function SomeClass(greeter) {
    this.greeter = greeter;
  }
  
  SomeClass.prototype.doSomething = function(name) {
    this.greeter.greet(name);
  }
</pre>

In the above example `SomeClass` is not concerned with locating the `greeter` dependency, it
is simply handed the `greeter` at runtime.

This is desirable, but it puts the responsibility of getting hold of the dependency on the
code that constructs `SomeClass`.

To manage the responsibility of dependency creation, each Angular application has an {@link
api/angular.injector injector}. The injector is a service locator that is responsible for
construction and lookup of dependencies.

Here is an example of using the injector service:

<pre>
  // Provide the wiring information in a module
  angular.module('myModule', []).
  
    // Teach the injector how to build a 'greeter'
    // Notice that greeter itself is dependent on '$window'
    factory('greeter', function($window) {
      // This is a factory function, and is responsible for 
      // creating the 'greet' service.
      return {
        greet: function(text) {
          $window.alert(text);
        }
      };
    });

  // New injector is created from the module. 
  // (This is usually done automatically by angular bootstrap)
  var injector = angular.injector(['myModule', 'ng']);
  
  // Request any dependency from the injector
  var greeter = injector.get('greeter');
</pre>

Asking for dependencies solves the issue of hard coding, but it also means that the injector needs
to be passed throughout the application. Passing the injector breaks the {@link
http://en.wikipedia.org/wiki/Law_of_Demeter Law of Demeter}. To remedy this, we turn the
dependency lookup responsibility to the injector by declaring the dependencies as in this example:

<pre>
  <!-- Given this HTML -->
  <div ng-controller="MyController">
    <button ng-click="sayHello()">Hello</button>
  </div>
</pre>  
<pre>
  // And this controller definition
  function MyController($scope, greeter) {
    $scope.sayHello = function() {
      greeter.greet('Hello World');
    };
  }
  
  // The 'ng-controller' directive does this behind the scenes
  injector.instantiate(MyController);
</pre>

Notice that by having the `ng-controller` instantiate the class, it can satisfy all of the
dependencies of `MyController` without the controller ever knowing about the injector. This is
the best outcome. The application code simply asks for the dependencies it needs, without having to
deal with the injector. This setup does not break the Law of Demeter.

## Dependency Annotation

How does the injector know what service needs to be injected?

The application developer needs to provide annotation information that the injector uses in order
to resolve the dependencies. Throughout Angular, certain API functions are invoked using the
injector, as per the API documentation. The injector needs to know what services to inject into
the function. Below are three equivalent ways of annotating your code with service name
information. These can be used interchangeably as you see fit and are equivalent.

### Inferring Dependencies

The simplest way to get hold of the dependencies, is to assume that the function parameter names
are the names of the dependencies.

<pre>
  function MyController($scope, greeter) {
    ...
  }
</pre>

Given a function the injector can infer the names of the service to inject by examining the
function declaration and extracting the parameter names. In the above example `$scope`, and
`greeter` are two services which need to be injected into the function.

While straightforward, this method will not work with JavaScript minifiers/obfuscators as they
rename the method parameter names. This makes this way of annotating only useful for {@link
http://www.pretotyping.org/ pretotyping}, and demo applications.

### `$inject` Annotation

To allow the minifers to rename the function parameters and still be able to inject right services
the function needs to be annotated with the `$inject` property. The `$inject` property is an array
of service names to inject.

<pre>
  var MyController = function(renamed$scope, renamedGreeter) {
    ...
  }
  MyController.$inject = ['$scope', 'greeter'];
</pre>

In this scenario the ordering of the values in the '$inject' array must match the ordering of the arguments to inject.
Using above code snippet as an example, '$scope' will be injected into 'renamed$scope' and 'greeter' into 'renamedGreeter'.
Care must be taken that the `$inject` annotation is kept in sync with the actual arguments in the
function declaration.

This method of annotation is useful for controller declarations since it assigns the annotation
information with the function.

### Inline Annotation

Sometimes using the `$inject` annotation style is not convenient such as when annotating
directives.

For example:

<pre>
  someModule.factory('greeter', function($window) {
    ...
  });
</pre>

Results in code bloat due to needing a temporary variable:

<pre>
  var greeterFactory = function(renamed$window) {
    ...
  };
  
  greeterFactory.$inject = ['$window'];
  
  someModule.factory('greeter', greeterFactory);
</pre>

For this reason the third annotation style is provided as well.

<pre>
  someModule.factory('greeter', ['$window', function(renamed$window) {
    ...
  }]);
</pre>

Keep in mind that all of the annotation styles are equivalent and can be used anywhere in Angular
where injection is supported.

## Where can I use DI?

DI is pervasive throughout Angular. It is typically used in controllers and factory methods. 

### DI in controllers

Controllers are classes which are responsible for application behavior. The recommended way of 
declaring controllers is using the array notation:

<pre>
  someModule.controller('MyController', ['$scope', 'dep1', 'dep2', function($scope, dep1, dep2) {
    ...
    $scope.aMethod = function() {
      ...
    }
    ...
  }]);
</pre>

This avoids the creation of global functions for controllers and also protects against minification.


### Factory methods

Factory methods are responsible for creating most objects in Angular. Examples are directives,
services, and filters. The factory methods are registered with the module, and the recommended way
of declaring factories is:

<pre>
  angular.module('myModule', []).
    config(['depProvider', function(depProvider){
      ...
    }]).
    factory('serviceId', ['depService', function(depService) {
      ...
    }]).
    directive('directiveName', ['depService', function(depService) {
      ...
    }]).
    filter('filterName', ['depService', function(depService) {
      ...
    }]).
    run(['depService', function(depService) {
      ...
    }]);
</pre>