-
Notifications
You must be signed in to change notification settings - Fork 10
module 번역
@ngdoc overview @name Developer Guide: Modules @description
Most applications have a main method which instantiates, wires, and bootstraps the application. Angular apps don't have a main method. Instead modules declaratively specify how an application should be bootstrapped. There are several advantages to this approach:
- The process is more declarative which is easier to understand
- In unit-testing there is no need to load all modules, which may aid in writing unit-tests.
- Additional modules can be loaded in scenario tests, which can override some of the configuration and help end-to-end test the application
- Third party code can be packaged as reusable modules.
- The modules can be loaded in any/parallel order (due to delayed nature of module execution).
Ok, I'm in a hurry. How do I get a Hello World module working?
Important things to notice:
- {@link api/angular.Module Module} API
- Notice the reference to the
myAppmodule in the<html ng-app="myApp">, it is what bootstraps the app using your module.
<doc:example module='myApp'> doc:source <script> // declare a module var myAppModule = angular.module('myApp', []);
// configure the module.
// in this example we will create a greeting filter
myAppModule.filter('greet', function() {
return function(name) {
return 'Hello, ' + name + '!';
};
});
</script>
<div>
{{ 'World' | greet }}
</div>
</doc:source> </doc:example>
While the example above is simple, it will not scale to large applications. Instead we recommend that you break your application to multiple modules like this:
- A service module, for service declaration
- A directive module, for directive declaration
- A filter module, for filter declaration
- And an application level module which depends on the above modules, and which has initialization code.
The reason for this breakup is that in your tests, it is often necessary to ignore the initialization code, which tends to be difficult to test. By putting it into a separate module it can be easily ignored in tests. The tests can also be more focused by only loading the modules that are relevant to tests.
The above is only a suggestion, so feel free to tailor it to your needs.
<doc:example module='xmpl'> doc:source <script> angular.module('xmpl.service', []). value('greeter', { salutation: 'Hello', localize: function(localization) { this.salutation = localization.salutation; }, greet: function(name) { return this.salutation + ' ' + name + '!'; } }). value('user', { load: function(name) { this.name = name; } });
angular.module('xmpl.directive', []);
angular.module('xmpl.filter', []);
angular.module('xmpl', ['xmpl.service', 'xmpl.directive', 'xmpl.filter']).
run(function(greeter, user) {
// This is effectively part of the main method initialization code
greeter.localize({
salutation: 'Bonjour'
});
user.load('World');
})
// A Controller for your app
var XmplController = function($scope, greeter, user) {
$scope.greeting = greeter.greet(user.name);
}
</script>
<div ng-controller="XmplController">
{{ greeting }}!
</div>
</doc:source> </doc:example>
A module is a collection of configuration and run blocks which get applied to the application during the bootstrap process. In its simplest form the module consist of collection of two kinds of blocks:
- Configuration blocks - get executed during the provider registrations and configuration phase. Only providers and constants can be injected into configuration blocks. This is to prevent accidental instantiation of services before they have been fully configured.
- Run blocks - get executed after the injector is created and are used to kickstart the application. Only instances and constants can be injected into run blocks. This is to prevent further system configuration during application run time.
angular.module('myModule', []).
config(function(injectables) { // provider-injector
// This is an example of config block.
// You can have as many of these as you want.
// You can only inject Providers (not instances)
// into the config blocks.
}).
run(function(injectables) { // instance-injector
// This is an example of a run block.
// You can have as many of these as you want.
// You can only inject instances (not Providers)
// into the run blocks
});
There are some convenience methods on the module which are equivalent to the config block. For example:
angular.module('myModule', []).
value('a', 123).
factory('a', function() { return 123; }).
directive('directiveName', ...).
filter('filterName', ...);
// is same as
angular.module('myModule', []).
config(function($provide, $compileProvider, $filterProvider) {
$provide.value('a', 123);
$provide.factory('a', function() { return 123; });
$compileProvider.directive('directiveName', ...);
$filterProvider.register('filterName', ...);
});
The configuration blocks get applied in the order in which they are registered. The only exception to it are constant definitions, which are placed at the beginning of all configuration blocks.
Run blocks are the closest thing in Angular to the main method. A run block is the code which needs to run to kickstart the application. It is executed after all of the service have been configured and the injector has been created. Run blocks typically contain code which is hard to unit-test, and for this reason should be declared in isolated modules, so that they can be ignored in the unit-tests.
Modules can list other modules as their dependencies. Depending on a module implies that required module needs to be loaded before the requiring module is loaded. In other words the configuration blocks of the required modules execute before the configuration blocks or the requiring module. The same is true for the run blocks. Each module can only be loaded once, even if multiple other modules require it.
Modules are a way of managing $injector configuration, and have nothing to do with loading of scripts into a VM. There are existing projects which deal with script loading, which may be used with Angular. Because modules do nothing at load time they can be loaded into the VM in any order and thus script loaders can take advantage of this property and parallelize the loading process.
In its simplest form a unit test is a way of instantiating a subset of the application in test and then applying a stimulus to it. It is important to realize that each module can only be loaded once per injector. Typically an app has only one injector. But in tests, each test has its own injector, which means that the modules are loaded multiple times per VM. Properly structured modules can help with unit testing, as in this example:
In all of these examples we are going to assume this module definition:
angular.module('greetMod', []).
factory('alert', function($window) {
return function(text) {
$window.alert(text);
}
}).
value('salutation', 'Hello').
factory('greet', function(alert, salutation) {
return function(name) {
alert(salutation + ' ' + name + '!');
}
});
Let's write some tests:
describe('myApp', function() {
// load the relevant application modules then load a special
// test module which overrides the $window with a mock version,
// so that calling window.alert() will not block the test
// runner with a real alert box. This is an example of overriding
// configuration information in tests.
beforeEach(module('greetMod', function($provide) {
$provide.value('$window', {
alert: jasmine.createSpy('alert')
});
}));
// The inject() will create the injector and inject the greet and
// $window into the tests. The test need not concern itself with
// wiring of the application, only with testing it.
it('should alert on $window', inject(function(greet, $window) {
greet('World');
expect($window.alert).toHaveBeenCalledWith('Hello World!');
}));
// this is another way of overriding configuration in the
// tests using an inline module and inject methods.
it('should alert using the alert service', function() {
var alertSpy = jasmine.createSpy('alert');
module(function($provide) {
$provide.value('alert', alertSpy);
});
inject(function(greet) {
greet('World');
expect(alertSpy).toHaveBeenCalledWith('Hello World!');
});
});
});