Updating unit tests to match code changes

When you make changes to your code, it's important to update any tests to verify the new behavior and catch any bugs that the new code has introduced. Discussion avec Copilot can help you quickly update tests to match your code changes, ensuring your test suite stays in sync with your implementation.

Example scenario

Imagine you have a Python function, calculate_discount, that determines the discount for a given purchase amount. In the original code, you get a 10% discount for amounts over $100. You're making changes to the logic of the function, so that only prices over $150 get a 10% discount, and there's now a 20% discount for amounts over $200.

Original code

In the original code, purchase prices above $100 get a 10% discount.

def calculate_discount(amount: float) -> float:
    if amount > 100:
        return amount * 0.1  # 10% discount
    return 0.0

Updated code

In the changed code, only amounts above $150 get 10% discount, and amounts above $200 now get a 20% discount.

def calculate_discount(amount: float) -> float:
    if amount > 200:
        return amount * 0.2  
    elif amount > 150:
        return amount * 0.1
    return 0.0

Existing tests

This example shows the existing tests that were written for the original code.

import unittest
from discount import calculate_discount

class TestCalculateDiscount(unittest.TestCase):
    def test_discount_above_100(self):
        # $140 should get a $14 discount (10%)
        self.assertAlmostEqual(calculate_discount(140), 14.0)
    
    def test_discount_below_100(self):
        # $80 should get no discount
        self.assertEqual(calculate_discount(80), 0.0)

    def test_discount_exactly_100(self):
        # $100 is not above 100, so no discount
        self.assertEqual(calculate_discount(100), 0.0)

if __name__ == '__main__':
    unittest.main()

Example prompt

You can ask Discussion avec Copilot to identify and update any tests affected by your logic change.

Given the update to the `calculate_discount` function, update the unit tests that may fail or become outdated as a result.

Example response

Discussion avec Copilot analyzes your code and existing tests, and identifies which tests will fail or are now misleading following your code change.

For example, Discussion avec Copilot explains that:

• test_discount_above_100 will now fail
• test_discount_below_100 and test_discount_exactly_100 will pass, but they are now misleading, since the threshold for the 10% discount is now $150, and not $100.

In addition, Discussion avec Copilot identifies any missing test coverage, such as the new 20% discount tier for amounts over $200.

Discussion avec Copilot proposes an updated set of tests.

Updated tests

import unittest
from discount import calculate_discount

class TestCalculateDiscount(unittest.TestCase):
    def test_discount_above_200(self):
        # $250 should get a $50 discount (20%)
        self.assertAlmostEqual(calculate_discount(250), 50.0)

    def test_discount_above_150(self):
        # $175 should get a $17.50 discount (10%)
        self.assertAlmostEqual(calculate_discount(175), 17.5)

    def test_discount_at_or_below_150(self):
        # $150 is not above 150, so no discount
        self.assertEqual(calculate_discount(150), 0.0)
        # $80 should get no discount
        self.assertEqual(calculate_discount(80), 0.0)

if __name__ == '__main__':
    unittest.main()

Further reading

• Ingénierie rapide pour GitHub Copilot Chat
• Meilleures pratiques pour l’utilisation de GitHub Copilot