MVP for Android: how to organize the presentation layer
MVP (Model View Presenter) pattern is a derivative from the well known MVC (Model View Controller), and one of the most popular patterns to organize the presentation layer in Android Applications.
This article was first published in April 2014, and been the most popular since then. So I’ve decided to update it solving most of the doubts people had, and also convert the code to
There have been breaking changes about architectural patterns since then, such as MVVM with architecture components, but MVP is still valid and an option to take into account.
What is MVP?
The MVP pattern allows separating the presentation layer from the logic so that everything about how the UI works is agnostic from how we represent it on screen. Ideally, the MVP pattern would achieve that the same logic might have completely different and interchangeable views.
The first thing to clarify is that MVP is not an architecture by itself, it’s only responsible for the presentation layer. This has been a controversial assessment, so I want to explain it a bit deeper.
You may find that MVP is defined as an architectural pattern because it can become part of the architecture of your App, but don’t consider that just because you are using MVP, your architecture is complete. MVP only models the presentation layer
An example of a complete architecture could be Clean Architecture, though there are many other options.
In any case, it is always better to use it for your architecture that not using it at all.
Want to learn Kotlin?
Check my free guide to create your first project in 15 minutes!
Why use MVP?
In Android, we have a problem arising from the fact that Android activities are closely coupled to both UI and data access mechanisms. We can find extreme examples such as CursorAdapter, which mixes adapters, which are part of the view, with cursors, something that should be relegated to the depths of data access layer.
For an application to be easily extensible and maintainable, we need to define well-separated layers. What do we do tomorrow if, instead of retrieving the same data from a database, we need to do it from a web service? We would have to redo our entire view.
MVP makes views independent from our data source. We divide the application into at least three different layers, which lets us test them independently. With MVP we take most of the logic out from the activities so that we can test it without using instrumentation tests.
How to implement MVP for Android
Well, this is where it all starts to become more diffuse. There are many variations of MVP and everyone can adjust the pattern to their needs and the way they feel more comfortable. It varies depending basically on the number of responsibilities that we delegate to the presenter.
Is the view responsible to enable or disable a progress bar, or should it be done by the presenter? And who decides which actions should be shown in the Action Bar? That’s where the tough decisions begin. I will show how I usually work, but I want this article to be more a place for discussion rather than strict guidelines on how to apply MVP, because up to there is no “standard” way to implement it.
For this article, I’ve implemented a very simple example that you may find on my Github with a login screen and a main screen. For simplicity purposes, the code in the article is in Kotlin, but you can also check the code in Java 8 in the repository.
The model
In an application with a complete layered architecture, this model would only be the gateway to the domain layer or business logic. If we were using Uncle Bob’s clean architecture, the model would probably be an interactor that implements a use case. But for the purpose of this article, it is enough to see it as the provider of the data we want to display in the view.
If you check the code, you will see that I’ve created two mock interactors with artificial delays to simulate requests to a server. The structure of one of this interactors:
123456789101112131415 | classLoginInteractor{...fun login(username:String,password:String,listener:OnLoginFinishedListener){// Mock login. I'm creating a handler to delay the answer a couple of secondspostDelayed(2000){when{username.isEmpty()->listener.onUsernameError()password.isEmpty()->listener.onPasswordError()else->listener.onSuccess()}}}} |
It’s a simple function that receives the username and the password, and does some validation.
The View
The view, usually implemented by an Activity (it may be a Fragment, a View… depending on how the app is structured), will contain a reference to the presenter. The presenter will be ideally provided by a dependency injector such as Dagger, but in case you don’t use something like this, it will be responsible for creating the presenter object. The only thing that the view will do is calling a presenter method every time there is a user action (a button click for example).
As the presenter must be view agnostic, it uses an interface that needs to be implemented. Here’s the interface that the example uses:
1234567 | interfaceLoginView{fun showProgress()fun hideProgress()fun setUsernameError()fun setPasswordError()fun navigateToHome()} |
It has some utility methods to show and hide progress, show errors, navigate to the next screen… As mentioned above, there are many ways to do this, but I prefer to show the simplest one.
Then, the activity can implement those methods. Here I show you some, so that you get an idea:
123456789101112131415 | classLoginActivity:AppCompatActivity(),LoginView{...override fun showProgress(){progress.visibility=View.VISIBLE}override fun hideProgress(){progress.visibility=View.GONE}override fun setUsernameError(){username.error=getString(R.string.username_error)}} |
But if you remember, I also told you that the view uses the presenter to notify about user interactions. This is how it’s used:
123456789101112131415161718192021 | classLoginActivity:AppCompatActivity(),LoginView{privateval presenter=LoginPresenter(this,LoginInteractor())override fun onCreate(savedInstanceState:Bundle?){super.onCreate(savedInstanceState)setContentView(R.layout.activity_login)button.setOnClickListener{validateCredentials()}}privatefun validateCredentials(){presenter.validateCredentials(username.text.toString(),password.text.toString())}override fun onDestroy(){presenter.onDestroy()super.onDestroy()}...} |
The presenter is defined as a property for the activity, and when the button is clicked, it calls validateCredentials()
, which will notify the presenter.
Same happens with onDestroy()
. We’ll see later why it needs to notify in that case.
The presenter
The presenter is responsible to act as the middleman between view and model. It retrieves data from the model and returns it formatted to the view
Also, unlike the typical MVC, it decides what happens when you interact with the view. So it will have a method for each possible action the user can do. We saw it in the view, but here’s the implementation:
123456789 | classLoginPresenter(varloginView:LoginView?,val loginInteractor:LoginInteractor):LoginInteractor.OnLoginFinishedListener{fun validateCredentials(username:String,password:String){loginView?.showProgress()loginInteractor.login(username,password,this)}...} |
MVP has some risks, and the most important we use to forget is that the presenter is attached to the view forever. And the view is an activity, which means that:
- We can leak the activity with long-running tasks
- We can try to update activities that have already died
For the first point, if you can ensure that your background tasks finish in a reasonable amount of time, I wouldn’t worry much. Leaking an activity 5-10 seconds won’t make your App much worse, and the solutions to this are usually complex.
The second point is more worrying. Imagine you send a request to a server that takes 10 seconds, but the user closes the activity after 5 seconds. By the time the callback is called and the UI is updated, it will crash because the activity is finishing.
To solve this, we call the onDestroy()
method that cleans the view:
123 | fun onDestroy(){loginView=null} |
That way we avoid calling the activity in an inconsistent state.
Conclusion
Separating interface from logic in Android is not easy, but the MVP pattern makes it easier to prevent our activities end up degrading into very coupled classes consisting of hundreds or even thousands of lines. In large Apps, it is essential to organize our code well. Otherwise, it becomes impossible to maintain and extend.
Nowadays, there are other alternatives like MVVM, and I will create new articles comparing them and helping with the migration. So keep tuned!
Remember you have the repository, where you can take a look at the code in both Kotlin and Java
I’m in love with Kotlin. I’ve been learning about it for a couple of years, applying it to Android and digesting all this knowledge so that you can learn it with no effort.
Shares