Reputation: 651
I am new to Angular 2 testing. I am trying to figure out what is the difference in using testsbed.get()
and just using inject
at the test level.
eg:
beforeEach(() => {
TestBed.configureTestingModule({
providers: [SomeService]
});
const testbed = getTestBed();
someService= testbed.get(SomeService);
});
});
vs
it('test service', inject([SomeService], (someService: SomeService) => {
Upvotes: 60
Views: 25197
Reputation: 4592
These used to be equivalent, but with Angular 9 the preferred method became inject().
TestBed.get()
is deprecated in Angular 9+, and TestBed.inject()
is now the preferred type-safe way to inject a dependency.
Read the documentation for clarity: TestBed.get() and TestBed.inject(). The change is one of deprecation.
Upvotes: 2
Reputation: 2054
Just to add to the existing answer and if like me you found this question because you are wondering what the difference is between TestBed.get()
and TestBed.inject()
which I know was not quite what the OP originally asked but it is relevant and is very much related.
I thought it was worth posting that according to the latest Angular documentation that TestBed.inject()
is the type safe replacement of TestBed.get()
.
From the Angular documentation on TestBed
that can be found here.
Upvotes: 47
Reputation: 222651
inject
helper function was historically used since AngularJS as an alternative to direct injector calls. In Angular 1, it was necessary to bootstrap a test with ngMock
. It is entirely optional in Angular 2 and higher and is just a suggested way for DI in TestBed tests.
It a convenience wrapper for testBed.get
that allows to avoid multiple testBed.get
calls, similarly to:
const [foo, bar] = [Foo, Bar].map(TestBed.get);
Other helper functions can be optionally used in conjunction with inject
, namely async
and fakeAsync
.
Upvotes: 29