Archive for the 'Behaviour Driven Development' Category

Internal And External Collaborators

The Single Responsibility Principal (SRP) states that every object should have a single responsibility, and that all its services should be aligned with that responsibility. By separating object responsibilities we are able to achieve a clear Separation of Concerns. Objects need to collaborate with each other in order to perform a behaviour. I find I use two distinct styles of collaboration with other objects which I have called internal and external collaborators.

The principals of object collaboration are nothing new, but I have found that defining these roles has helped me to better understand how to design and test the behaviour of objects.

External Collaborators

External collaborators are objects that provide a service or resource to an object but are not directly controlled by the object. They are passed to an object by dependency injection or through a service locator. An object makes calls to its external collaborators to perform actions or retrieve data. When testing an object, any external collaborators are stubbed-out. We can then write tests that perform an action, then determine if the right calls were made to the external object.

Examples of external collaborator objects include: services, repositories, presenters, framework classes, email senders, loggers and file system wrappers.

We are interested in testing how we interact with the external collaborator and not how it affects the behaviour of our object. For example, if we are testing a controller that retrieves a list of customers from a repository, we want to know that we have asked the repository for a list of customers, but we are not concerned that the repository returns the correct customers (this is a test for the repository itself). Of course, we might need some particular customer objects returned by the stub for the purpose of testing the behaviour of the controller. These customer objects then become internal collaborators, which we’ll come to next.

External collaborators can be registered with an IoC container to manage the creation and lifecycle of the object, and to provide an instance to dependent objects.

Internal Collaborators

Internal collaborators have a smaller scope than external collaborators. They are used in the context of the local object to provide functions or hold state. An object and its internal collaborators work together closely and should be treated as a single unit of behaviour.

Examples of internal collaborators include: DTOs, domain entities, view-models, utilities, system types and extension methods.

When testing an object with internal collaborators, we are interested in the effect on behaviour, not the interaction with the object. Therefore we shouldn’t stub-out internal collaborators. We don’t care how we interact with them, just that the correct behaviour occurs.

These objects are not affected by external influences, such as a database, email server, or file system. They are also not volatile or susceptible to environmental changes, such as a web request context. Therefore, they should not require any special context setup before testing.

We don’t get passed an instance of a internal collaborator through dependency injection, instead they may be passed to us by an external collaborator (e.g. a repository returning an entity), or we create an instance within our own object when we need it (such as a DTO).

By understanding the roles and responsibilities of collaboration between objects, our design becomes clearer and tests are more focused and easier to maintain.

object_collaborators

Advertisements

Features and Specs

I’m going to take a short diversion from my series on BDD with ASP.NET MVC to talk about the difference between Features and Specs in Behaviour-Driven Development.

I am currently reading the beta version of The RSpec Book. This is an excellent book that covers doing BDD in Ruby using RSpec and Cucumber. From reading this book, I have come to realise there is a distinct difference between higher-level user acceptance criteria (features) and and lower-level context-specifications (specs).

Features

Features are the things that the customer is likely to be most concerned with. They are high-level behaviours that are derived from the acceptance criteria of a user story. They are purely concerned with how the user interacts with the application and don’t provide any details of implementation.

Features can be used to facilitate communication with the customer on what business objectives the application should achieve.

A feature consists of a user story and one or more example scenarios, which serve as our acceptance criteria. An example of a feature would be:

Story: Account Holder withdraws cash

  As an Account Holder 
  I want to withdraw cash from an ATM
  So that I can get money when the bank is closed

  Scenario 1: Account has sufficient funds 
    Given the account balance is $100
    And the card is valid
    And the machine contains enough money
    When the Account Holder requests $20
    Then the ATM should dispense $20
    And the account balance should be $80
    And the card should be returned

Features can be turned into automated acceptance criteria. In Ruby, we can use Cucumber to parse a plain-text file and run each scenario against the application. These are usually slow-running, end-to-end integration tests. In .NET we would use a BDD framework such as NBehave, or a standard unit-testing framework to implement the features as automated acceptance criteria.

Recently, I have been using Cucumber for automating the acceptance criteria for a .NET application, using Waitr to drive the browser. With IronRuby, we will be able to access our .NET objects directly from Cucumber. As of this post date, IronRuby does not work with Cucumber, but a fix should be out very soon.

Specs

When we have defined a new feature, we must create the objects required to implement it. To do this we need to “drop-down” to the object-level and use a context-specification style to define how these objects should behave.

Specs are implemented as fast-running, unit-level tests with stubbed dependencies. In Ruby, RSpec is used to define the specs. In .NET we would use a standard unit-testing tool, such as NUnit.

A spec consists of an action on an object, within a certain context, and the observations made from that action. An example of a spec would be:

When ATM requests withdrawal and account has sufficient funds
  Should check account balance
  Should send “Please wait” message to screen
  Should begin transaction
  Should withdraw funds from account
  Should dispense cash
  Should end transaction
  Should return card
  Should send “please take your card” message to screen

These steps describe the interaction between our object and its dependencies. They do not necessarily reflect the user’s observations (e.g the user doesn’t observe a transaction taking place).

We are not as concerned about testing the state of an object as we are about defining how it interacts with other objects. For instance, using the Model-View-Controller pattern, if we’re specifying a controller, we want to make sure an action sends a certain “message” to the view, rather than testing how it affects the state of the view. By specifying interaction over state, we are better able to define the roles and responsibilities of our objects.

I will try to incorporate some of this into the PlaylistShare sample application to demonstrate the distinction between Features and Specs.

Using Behaviour-Driven Development with ASP.NET MVC – Part 2

In part one, I introduced the PlaylistShare project, created some initial user stories, defined the acceptance criteria and created a template for writing executable specifications. In this part, I am going to write some executable specs for viewing playlists and implement the behaviour to make them pass. You can download the source code at the end of this post.

There a couple of approaches you can take when writing specs for an ASP.NET MVC user interface. Either you can run them directly against the UI from inside a browser, using a tool such as WaitiN. Or, you can drop in just below the UI and run the specs against the controllers. I prefer running them against the controllers as they run much faster and are not susceptible to changes to the HTML elements. However, you may find some aspects are better specified using a UI testing tool, such as behaviour that relies heavily on client-side scripting, which the controllers know nothing about.

I will be starting with the executable specification template I created in part one:

namespace view_playlist_specs
{
    namespace scenario_of
    {
        public abstract class listener_views_playlists : Specification<PlaylistController>
        {
            protected override PlaylistController create_subject()
            {
                // Create the system under test
                return new PlaylistController();
            }
 
            protected void given_a_list_of_playlists()
            {
                // Establish a list of playlists
            }
        }
    }
 
    namespace viewing_a_list_of_playlists
    {
        [TestFixture]
        public class when_the_listener_views_the_playlists 
            : scenario_of.listener_views_playlists
        {
            protected override void setup_scenario()
            {
                // Arrange
                given_a_list_of_playlists();
            }
 
            protected override void execute_scenario()
            {
                // Act
            }
 
            [Test]
            public void should_display_a_list_of_playlists()
            {
                // Assert
            }
 
            [Test]
            public void playlists_should_be_ordered_by_date_submitted_from_newest_to_oldest()
            {
                // Assert
            }
        }
    }
}

The first scenario we are going to implement is:

Story: Listener views playlists

As a playlist listener, I can view a list of playlists that have been submitted, so that I can find a playlist I might enjoy.

Scenario 1: Viewing a list of playlists

Given a list of playlists

When the listener views the playlists

Then should display a list of playlists

  And playlists should be ordered by date submitted from newest to oldest

First, lets create a new ASP.NET MVC application and add a PlaylistController to give us a subject to work with.

[HandleError]
public class PlaylistController : Controller
{
    public ActionResult Index()
    {
        return View();
    }
}

Now that we have a controller, let’s specify the first outcome of the scenario.

[Test]
public void should_display_a_list_of_playlists()
{
    the_view_model.should_be_equivalent_to(playlists);
}

The outcome states that a list of playlists should be viewed on the page.

the_view_model is a property on the spec base class that wraps calls to ProductsController.ViewData.Model. This makes the code slightly more readable. We can use ViewData.Model on the controller to pass data back to the view. In this case we want to pass back a list of Playlist objects.

You will notice I have used a method called should_be_equivalent_to to verify that the_view_model contains the same elements as playlists. This is an extension method defined in a SpecificationExtensions class I have added to the project. This class contains some helpful extension methods that wrap-up calls to NUnit assertions. This enables us to assert conditions directly on the objects, producing slightly more readable outcomes. Again, this is a personal preference and you can use ordinary Assert calls if you wish.

Running this spec should result in a failure, as we are not yet returning any playlists from the Index() controller method. We need to provide the controller with our list of playlists. To do this, we need to somehow pass a list of Playlist objects to the controller. The best way to do this is to use Dependency Injection and Mocking to pass a Repository containing our Playlist objects into the controller. Woah! Sounds complicated. But, it’s actually quite simple…

Let’s modify the PlaylistController to take an IPlaylistRepository and initialise a field in the constructor.

[HandleError]
public class PlaylistController : Controller
{
    private readonly IPlaylistRepository _playlistRepository;
 
    public PlaylistController(IPlaylistRepository playlistRepository)
    {
        _playlistRepository = playlistRepository;
    }
 
    public ActionResult Index()
    {
        return View();
    }
}

Now we need to pass an instance of this repository when we create the PlaylistController in our specs. To do this I am going to generate a stub using RhinoMocks. This creates a runtime implementation of the IPlaylistRespository that we can pass into the controller.

protected override void setup_scenario()
{
    playlist_repository = MockRepository.GenerateStub<IPlaylistRepository>();
}
 
protected override PlaylistController create_subject()
{
    return new PlaylistController(playlist_repository);
}

The setup_scenario method gets called right before create_subject, so we can use this to create any dependencies before the PlaylistController gets instantiated. Any scenarios can create additional context by overriding the setup_scenario method.

Now we need to add a GetAll() method to IPlaylistRepository that allows the PlaylistController to retrieve a list of Playlist objects.

public ActionResult Index()
{
    var playlists = _playlistRepository.GetAll();
    return View(playlists);
}

Back in our specs, we can now tell the stubbed playlist_repository to return some Playlist objects whenever the GetAll() method is called.

protected void given_a_list_of_playlists()
{
    playlists = new List<Playlist>
                    {
                        new Playlist(),
                        new Playlist(),
                        new Playlist()
                    };
    playlist_repository.Stub(r => r.GetAll()).Return(playlists.ToArray());
}

Our spec should now be passing! We have now confirmed the controller is passing a list of playlists to the view to display on the page.

So, let’s write an assertion for our next outcome: “playlists should be ordered by date submitted from newest to oldest”. Here we verify that the Playlist objects sent to the view are ordered by date-submitted in a descending order.

[Test]
public void playlists_should_be_ordered_by_date_submitted_from_newest_to_oldest()
{
    the_view_model.should_be_ordered_by_descending(p => p.DateSubmitted);
}

For this to compile we need to add a DateSubmitted property to the Playlist object.

I have created a helper method called should_be_ordered_by_descending that will verify that the playlists are ordered correctly. It’s really useful to create extension methods for complicated assertions to better explain what is being verified.

We can now make this spec pass by using LINQ to order the list of Playlist objects in the controller action before they are returned to the view.

public ActionResult Index()
{
    var playlists = _playlistRepository.GetAll();
    return View(playlists.OrderByDescending(p => p.DateSubmitted));
}

Here is the output from the Resharper test runner:

test-run-resharper-1-small

Using these naming conventions means that if a spec fails, it’s easy to identify what behaviour is causing the problem. You also get clear, concise, executable documentation that describes how your application behaves.

You might have noticed that I have been implementing the playlist controller without even running the web site. By implementing the controller first, we can focus on the core behaviour of the application without being concerned about how it is presented. In the next part, we will create the view that displays the playlists on the page.

Download the source code for Part 2

Using Behaviour-Driven Development with ASP.NET MVC – Part 1

Introduction

Welcome to the first post in a series where I hope to demonstrate creating an ASP.NET MVC application using Behaviour-Driven Development (BDD).

I agonised over how to introduce this blog post. In the end, I figured it would be best just to get straight into it! There are plenty of excellent articles that introduce the concepts behind both BDD and ASP.NET MVC, so I won’t bother repeating it here. I will explain core concepts along the way and provide links to further information. If you feel that I haven’t explained something properly, or have missed something, then please post a comment and I’ll try to elaborate further.

Right, let’s get started. The application I am going to build is a playlist sharing web site aptly called PlaylistShare (slightly unimaginative, but if you have a better name, let me know! :-). Music streaming services like Spotify are becoming increasingly popular and a number of sites have sprung up that allow you to share playlists with other people. I don’t intend this example to be anything unique or new, but I thought it would be an interesting subject and not yet-another-storefront example.

I will provide a link to download the source code at the end of each post. I have also created a Google Code project to host the source code.

It All Starts With a Story

BDD is an outside-in development process. We start by identifying goals and motivations, then drill down to the features that will achieve those goals. So, I’ve come up with a few user stories to get us started.

Story: Listener views playlists

As a playlist listener, I can view a list of playlists that have been submitted, so that I can find a playlist I might enjoy.

Story: Listener views playlist details

As a playlist listener, I can view the playlist details, so that I can get more information about the playlist.

Story: Contributor submits a playlist

As a playlist contributor, I can submit a playlist, so that others can listen to it.

There are a lot of other features I can add, but as a “business”, these are my top priorities and will allow me to get something functional delivered first.

Defining Acceptance Criteria

In order to determine what behaviour to implement, we need to create a list of acceptance criteria for each story. The acceptance criteria allows us to define when a story is “done”.

Note: this is a very important stage. It allows all members of the team to share a common understanding of what the system is meant to do. It is tempting to jump straight to writing the executable specifications, but this stage needs to be carefully considered and have input from the customer and testers.

We will define the acceptance criteria as example scenarios using a given/when/then vocabulary.

Story: Listener views playlists

Scenario 1: Viewing a list of playlists
Given a list of playlists
When the listener views the playlists
Then should display a list of playlists
  And playlists should be ordered by date submitted from newest to oldest

Story: Listener views playlist details

Scenario 1: Viewing playlist details
Given a playlist
When the listener views the playlist details
Then should display the playlist details
  And should display a link to listen to the playlist

Story: Contributor submits a playlist

Scenario 1: Submitting a playlist
Given a playlist
When the contributor submits the playlist
Then should submit the playlist
  And should display a message confirming the submission was successful

Scenario 2: Submitting a playlist without a name
Given a playlist without a name
When the contributor submits the playlist
Then should display a message saying the name is required
  And should not submit the playlist

Now that we have defined some acceptance criteria, we can create the executable specifications used to drive the implementation (a fancy way of saying, “write some tests!”). As we go forward we will probably identify other features we would like. We will make a note of these then add them to our next “iteration”.

Writing Executable Specifications

In BDD, tests are called specifications. This is mainly to get away from the notion that we are “testing”. Rather, we are defining behaviours that will drive the development process. We will still end up with a full suite of automated tests, but this is merely a very handy side-effect, not the goal, of BDD.

There are a number of BDD frameworks available for .NET, the most well-known of these being NBehave. However, these frameworks can be difficult to get started with and can distract us from the main purpose of BDD, which is to effectively describe the application behavior. This is not because these frameworks are bad, but is more to do with the constraints of the C# language. BDD frameworks are much more effective in languages like Ruby, where the syntax makes it very easy to write readable specifications using frameworks such as RSpec and Cucumber. For this example, I am going to stick with using NUnit and structure the test classes and member names to enable us to write BDD specifications. It will be a glorious day when when eventually see RSpec and Cucumber working on IronRuby and integrated seamlessly with our .NET applications. Until then, this is my current preference for writing BDD-style specifications in real-world applications.

I use a common Specification base-class to provide a framework for writing the specifications. This simple class contains the NUnit setup/teardown methods, handles exceptions and coordinates the various stages of executing a scenario.

These are a few conventions I use when writing BDD-style specifications:

  • Specification names are lowercase with underscores. This_is_to_make_the_specifications_more_readable. It is a personal preference, but I find PascalCaseNamingBecomesHardToReadForLongSentences (see what I did there… heh).
  • One test-fixture per scenario. Each test fixture represents a scenario of the story.
  • Each test-fixture is wrapped in a namespace that describes the scenario. This allows us to easily navigate the scenarios and prevents naming conflicts with other actions. This becomes important when we have the same action in different scenarios.
  • Each test-fixture name defines the action. An action is the “when” part of a scenario. It is the event that causes the behaviour.
  • Each test-fixture derives from a base class that describes the story. By inheriting from a base class, we can provide common set up data that is shared across all scenarios in the story. It also provides us with a handy reference back to the original story.
  • The context is set before executing the action. The context is the “given” part of a scenario. The context sets the state of the system-under-test and its dependencies before the action is invoked. Wrapping setup code in helper-methods makes it easy to see what context we are creating and to reuse common setup steps.
  • One test per outcome. An outcome is the result of an action, the “then” part of a scenario. We may have many outcomes from an action, but we should only be testing for one thing per outcome. So try to only make one assertion per test method. Test for other outcomes in separate test methods.

The outline for the first scenario looks like this:

namespace view_playlist_specs
{
    namespace scenario_of
    {
        public abstract class listener_views_playlists : Specification<PlaylistController>
        {
            protected override PlaylistController create_subject()
            {
                // Create the system under test
                return new PlaylistController();
            }
 
            protected void given_a_list_of_playlists()
            {
                // Establish a list of playlists
            }
        }
    }
 
    namespace viewing_a_list_of_playlists
    {
        [TestFixture]
        public class when_the_listener_views_the_playlists 
            : scenario_of.listener_views_playlists
        {
            protected override void setup_scenario()
            {
                // Arrange
                given_a_list_of_playlists();
            }
 
            protected override void execute_scenario()
            {
                // Act
            }
 
            [Test]
            public void should_display_a_list_of_playlists()
            {
                // Assert
            }
 
            [Test]
            public void playlists_should_be_ordered_by_date_submitted_from_newest_to_oldest()
            {
                // Assert
            }
        }
    }
}

Download the source code containing the outline for the first scenario.

I hope this has been a helpful way to get started. I will explain a lot more about writing specifications in upcoming posts. Please feel free to post any questions or comments. In the next part of this series, I will begin implementing a playlist controller and passing the scenarios.