An application is composed of many objects that collaborate with each other. Objects usually depend on other objects to perform some task. When an object is responsible for referencing its own dependencies it leads to a highly coupled, hard-to-test and hard-to-change code.
Dependency injection is a software design pattern that implements inversion of control for resolving dependencies. An injection is passing of dependency to a dependent object that would use it. This allows a separation of client’s dependencies from the client’s behaviour, which allows the application to be loosely coupled.
Not to be confused with the above definition - a dependency injection simply means giving an object its instance variables.
It’s that simple, but it provides a lot of benefits:
There are three most commonly used ways Dependency Injection (DI) can be implemented in an application:
There is an interesting article with links to more articles about Dependency Injection so check it out if you want to dig deeper into DI and Inversion of Control principle.
Let’s show how to use DI with View Controllers - an every day task for an average iOS developer.
We’ll have two View Controllers: LoginViewController and TimelineViewController. LoginViewController is used to login and upon successful loign, it will switch to the TimelineViewController. Both view controllers are dependent on the FirebaseNetworkService.
LoginViewController
class LoginViewController: UIViewController {
var networkService = FirebaseNetworkService()
override func viewDidLoad() {
super.viewDidLoad()
}
}
TimelineViewController
class TimelineViewController: UIViewController {
var networkService = FirebaseNetworkService()
override func viewDidLoad() {
super.viewDidLoad()
}
@IBAction func logoutButtonPressed(_ sender: UIButton) {
networkService.logutCurrentUser()
}
}