Golang Testing - flowup/knowledge GitHub Wiki

• Testify - testing and mocking framework
• Goconvey - automatic test runner with UI

🖼️ These concepts are all working with the Testify framework.

We will define a simple fibonacci function for the purpose of direct testing and table testing.

func Fib(n int) int {...}

Direct testing relays on the ability of code to be directly executed and asserted. These are the most simple tests, however, they are not as flexible as Table tests.

🤖 Use direct testing when there is no way of generalization of the given problem

func TestFib(t *testing.T) {
  // let's assert that third number of fibonacci sequence is 3
  assert.Equal(t, 2, Fib(3))
}

Table tests are used whenever a generalization of inputs and outputs is available. E.g. Fibonacci sequence is well known and can be calculated. However, to test that it calculates the results correctly, we need to create a lookup table with known results.

func TestFib(t *testing.T) {
  candidates := map[int]int{
    1: 1,
    2: 1,
    3: 2,
    4: 3,
    5: 5,
  }

  for seq, res := range candidates {
    assert.Equal(t, res, Fib(seq))
  }
}

Any cases which caused bugs can be rapidly tested thanks to the table tests. All it involves is adding a new pair to the table.

Testing with mocks and HTTP testing require more advanced scenario, e.g. User registration process. We will cover this scenario with our Auth service.

// MailService is a service that is able to send mails to users
type MailService interface {
  Send(to string, text string)
}

type AuthService struct {
  mails MailService
}

// Register registers given user to the database and sends him
// confirmation email via s.mails.Send()
func (s *AuthService) Register(user User) {...}

🤖 This code is missing implementations, as the tester should always be able to determine what is the function doing at the first place