Testing and Debugging Notes

Review from Last Class

• The previous class involved writing a Python program to find unique combinations of 3 numbers from a list that sum to a target number.

• The task included providing doctests for specific test cases.

• Example:

  • List: [1, 2, 3, 4, 5, 6, 7, 8, 9]

  • Target: 12

  • Combinations: [(1, 3, 8), (1, 4, 7), (1, 2, 9), (1, 5, 6), (3, 4, 5), (2, 3, 7), (2, 4, 6)]

• Example:

  • List: [1, 2, 3, 4, 5, 6, 7, 8, 9]

  • Target: 17

  • Combinations: [(4, 5, 8), (2, 6, 9), (3, 5, 9), (2, 7, 8), (4, 6, 7), (3, 6, 8), (1, 7, 9)]

Unit Testing

• Unit testing involves testing individual units or components of code to ensure proper functioning.

• The correct approach to unit testing is to test all code parts individually.

Key Aspects of Unit Testing:

• All tests should be within a subclass TestCase from the unittest library.

• Each test is a function named starting with test_.

• Assert methods are used to check for expected results.

• The setUp() method is called before each test, used to define elements needed for the tests.

• The tearDown() method is called after each test, used to reset definitions after each test.

Assert Methods:

• assertTrue: Succeeds if the expression is true.

• assertFalse: Succeeds if the expression is false.

• assertEqual, assertNotEqual: Checks equality and inequality.

• assertAlmostEqual: Used for comparing floating-point numbers.

• assertNotAlmostEqual

Running Tests in VSCode:

1. Create a workspace where all tests are located.

2. Configure tests in the testing tab.

3. Select unittest -> Root directory -> test_*.py.

Troubleshooting:

• If test files are not appearing, reload the window:

  • Control + Shift + p (Windows)

  • Command + Shift + p (Mac)

  • Options: "Reload window", "Configure python tests"

• If the explorer is not working, run tests in the terminal:

  • python -m unittest discover -v -s . -p "test_*.py"

• For test discovery errors, reload the window.

Tips:

• Create a new folder for each exercise.

• Configure Python tests in each exercise by changing the folder in the configuration.

• Each exercise should have at least two files: class_name.py and test_class_name.py.

Exercises

• Exercise 1 - Calculator Class:

  • Step 1: Create tests for all methods (add, subtract, multiply, divide) using pytest and unittest.

  • Step 2: Implement the functions to make tests pass.

  • Test cases provided for each method with expected results.

  • For example:

    • add(1, 3) should return 4.

    • subtract(8, 7) should return 1.

    • multiply(8, 3) should return 24.

    • divide(10, 5) should return 2.

• Exercise 2 - Temperature Converter Class:

  • Step 1: Create tests for all methods (celsius_to_fahrenheit, fahrenheit_to_celsius, celsius_to_kelvin, kelvin_to_celsius) using pytest and unittest.

  • Step 2: Implement the functions to make tests pass.

  • Test cases provided for each method with expected results.

    • celsius_to_fahrenheit(0) should return 32.

    • fahrenheit_to_celsius(32) should return 0.

    • celsius_to_kelvin(25) should return 298.15.

    • kelvin_to_celsius(273.15) should return 0.

PyTest

• Supports unittests.

• Allows selecting specific tests using the -k flag.

• Provides mark tests:

  • @pytest.mark.skip: Skips the test.

  • @pytest.mark.xfail: Marks the test as expected to fail.

• Filter by marked tests using the -m flag.

• Run from the last failed test using the --lf flag.

Mocking

• Mock objects imitate real objects but with fake data.

• Commonly used for API calls.

• Decouples tested code from other calls.

• Example:

  • Mocking an expensive API call to speed up tests.

    import time
    def compute(x):
        response = expensive_api_call()
        return response + x
    
    def expensive_api_call():
        time.sleep(1000) # takes 1,000 seconds
        return 123
    
    import unittest
    import compute_expensive
    
    class TestComputeExpensive(unittest.TestCase):
        def test_compute(self):
            self.assertEqual(compute_expensive.compute(1), 124)
    
    import unittest
    from unittest.mock import patch
    import compute_expensive
    
    class TestComputeExpensive(unittest.TestCase):
        @patch('compute_expensive.expensive_api_call', return_value=123)
        def test_compute(self, mock_expensive_api_call):
            # calls the compute function mocking the expensive_api_call
            self.assertEqual(124, compute_expensive.compute(1))
            # asserts that expensive_api_call was called once
            mock_expensive_api_call.assert_called_once()
    

Code Coverage

• Tracks which lines of code are executed during tests.

• Provides a report describing test coverage.

• Tool to gain insight into what the tests are doing.

• Not the sole definition of a good test suite.

How to Run Test Coverage:

• Right-click in the folder where test folders are located and select "Run Tests with Coverage".

• If a No module named 'coverage' error occurs, install the coverage package:

  • pip install coverage or pip3 install coverage if pip is not working.

• The percentage of code covered appears next to the file name once configured correctly.

Exercises

• Exercise 3 - Bank Account Class:

  • Step 1: Create test cases for BankAccount class methods (__init__, deposit, withdraw, get_balance).

  • Step 2: Implement failing tests.

  • Step 3: Implement the code to make the tests pass.

• Exercise 4 - Counter Class:

  • Step 1: Create test cases for Counter class methods (__init__, increment, decrement, reset, get_value).

  • Step 2: Implement failing tests.

  • Step 3: Implement the code to make the tests pass.

• Exercise 5 - To-Do List Class:

  • Step 1: Create test cases for ToDoList class methods (__init__, add_task, remove_task, get_tasks).

  • Step 2: Implement failing tests.

  • Step 3: Implement the code to make the tests pass.

• Exercise 6 - Shapes:

  • Create tests using inheritance and abstract methods for shapes (rectangle, circle, square).

  • Each shape should have perimeter and area calculations.