更多pythonic测试方法

Question

Ok I'm writing tests for my Django app. I'm trying to figure out a way to iterate over the testing for a ajax register call for the appropriate responses. It's working but I know it can be done in a more efficient way.

def test_ajax_register(self):
    c = Client()
    # Check register success
    response = c.post('/register/', {
        'register-username': 'testuser',
        'register-email': 'testuser@email.com',
        'register-password': 'password'
    })
    self.assertEqual(json.loads(response.content)['status'], 'success')
    self.assertEqual(response.status_code, 200)

    # Check register failed  username taken
    response = c.post('/register/', {
        'register-username': 'testuser',
        'register-email': 'testuser@email.com',
        'register-password': 'password'
    })
    self.assertEqual(json.loads(response.content)['status'], 'fail')
    self.assertEqual(json.loads(response.content)['error_msg'], 'username already in use')

    # Check register failed email in use
    response = c.post('/register/', {
        'register-username': 'testuser1',
        'register-email': 'testuser@email.com',
        'register-password': 'password'
    })
    self.assertEqual(json.loads(response.content)['status'], 'fail')
    self.assertEqual(json.loads(response.content)['error_msg'], 'email already in use')

    # Check register failed  password length
    response = c.post('/register/', {
        'register-username': 'testuser2',
        'register-email': 'testuser2@email.com',
        'register-password': 'pass'
    })
    self.assertEqual(json.loads(response.content)['status'], 'fail')
    self.assertEqual(json.loads(response.content)['error_msg'], 'password must be atleast 8 characters long')

Answer 1

Just separate your tests into separate methods. Each of them will be triggered as long as they have prefix test_ , and each test can also run as a stand alone test. Use setUp and tearDown methods, if need to preset anything.

def test_ajax_register(self):
    # Test registration 
    c = Client()
    # Check register success
    response = c.post('/register/', {
        'register-username': 'testuser',
        'register-email': 'testuser@email.com',
        'register-password': 'password'
    })
    self.assertEqual(json.loads(response.content)['status'], 'success')
    self.assertEqual(response.status_code, 200)


def test_failed_registration(self):
    # Check register failed  username taken
    response = c.post('/register/', {
        'register-username': 'testuser',
        'register-email': 'testuser@email.com',
        'register-password': 'password'
    })
    self.assertEqual(json.loads(response.content)['status'], 'fail')
    self.assertEqual(json.loads(response.content)['error_msg'], 'username already in use')


def test_email_in_use(self):
    # Check register failed email in use
    response = c.post('/register/', {
        'register-username': 'testuser1',
        'register-email': 'testuser@email.com',
        'register-password': 'password'
    })
    self.assertEqual(json.loads(response.content)['status'], 'fail')
    self.assertEqual(json.loads(response.content)['error_msg'], 'email already in use')


def test_password_length(self):
    # Check register failed  password length
    response = c.post('/register/', {
        'register-username': 'testuser2',
        'register-email': 'testuser2@email.com',
        'register-password': 'pass'
    })
    self.assertEqual(json.loads(response.content)['status'], 'fail')
    self.assertEqual(json.loads(response.content)['error_msg'], 'password must be atleast 8 characters long')

Answer 2

In general, you should split the big test logically and follow the UnitOfWork_StateUnderTest_ExpectedBehavior test method naming practice .

The concept of data-driven tests is also a relevant idea, I've personally used ddt package in the past and then switched to pytest parameterized tests - that is a great way to test the same specific "symptom" of a function under test using different input data - this would help to prevent code duplication ( DRY principle ) and make your tests more organized and readable.