mokacoding

unit and acceptance testing, automation, productivity

Specta global before and after each hooks (Updated)

Specta is one of the best xSpec style testing library for Objective-C.

One of the reason I like it is the modular approach it takes, you can plug in Specta your favourite expectation library, matcher library, and mocking library. It doesn't make any assumptions and the only things it focuses on is providing a nice DSL to write expressive tests, and to run them leveraging XCTest.

One little feature that Specta has is the ability to set global beforeEach and afterEach hooks.

Say that we have a Banana class, and that we want to test it's looks yellow behaviour. We get a banana from the fruit bowl, and wherever we put it it should still look yellow.

Unfortunately our Banana is a stateful fruit, so to make the test accurate we have to put it back in the fruit bowl every time.

A Specta spec to describe this behaviour might look like this:

describe(@"Banana", ^{
  it(@"should look yellow when put on the table", ^{
    [banana removeFromTheFruitBowl];
    [banana putOnTheTable];
    expect([banana looksYellow]).to.beTruthy();
    [banana putInTheFruitBowl];
  });

  it(@"should look yellow when put on the snack bag", ^{
    [banana removeFromTheFruitBowl];
    [banana putInTheSnackBag];
    expect([banana looksYellow]).to.beTruthy();
    [banana putInTheFruitBowl];
  });
});

This spec looks a bit dense, and wet. Thanks to the beforeEach and afterEach hooks we can write it in a clearer and drier way:

describe(@"Banana", ^{
  beforeEach(^{
    [banana removeFromTheFruitBowl];
  });

  it(@"should look yellow when put on the table", ^{
    [banana putOnTheTable];
    expect([banana looksYellow]).to.beTruthy();
  });

  it(@"should look yellow when put on the snack bag", ^{
    [banana putInTheSnackBag];
    expect([banana looksYellow]).to.beTruthy();
  });

  afterEach(^{
    [banana putInTheFruitBowl];
  });
});

Configuring a global hook in Specta

Now, not only our Banana is stateful, but also is an hidden dependency in several other pieces of the system. Let's for a moment forget that the best thing to do in this case would be to do some good old refactoring to extract that dependency, and imagine that the only thing we can do is call the removeFromTheFruitBowl and putInTheFruitBowl methods in every test.

This option is quite annoying isn't it? Our specs shouldn't have to care about setup and tear down operations that are at the system level. Plus it's easy to forget about this step.

Specta takes care of this problem for us. Before running the suite it will inspect the run time for classes that conform to the SPTGlobalBeforeAfterEach protocol, class methods, and run their beforeEach and afterEach methods, before and after each example in the suite.

@interface BeforeAfterEachHelper : NSObject <SPTGlobalBeforeAfterEach>
@end

@implementation BeforeAfterEachHelper

+ (void)beforeEach {
  // this code will run beforeEach example
  //...
}

+ (void)afterEach {
  // this code will run afterEach example
  //...
}

@end

Pretty handy, isn't it?

The advantage of using the protocol is that if any of your classes has either a beforeEach or an afterEach method that will not be pick up by the test. This approach is called whitelisting.


One could argue that a suite that needs to run code before and/or after each test has some intrinsic problem, and that time should be spent to improve the architecture of the system under test. Regardless of that I think this option offered by Specta is interesting and powerful.

Have a look at the example project to see global hooks in action, or check out the video below, and feel free to leave a comment or tweet me at @mokagio.

Happy coding, and leave the codebase better than you found it.

Vote on Hacker News