Generalizing Solutions with TDD

Welcome back to our course on Test-Driven Development (TDD) using Scala and ScalaTest. In our previous lesson, we introduced the fundamentals of TDD and the Red-Green-Refactor workflow. Now, we will advance our TDD skills by focusing on generalizing solutions and enhancing the complexity of our testing scenarios.

As a brief reminder, TDD involves a repetitive cycle known as Red-Green-Refactor:

• Red: Write a failing test to clarify the new functionality you aim to implement.
• Green: Develop the smallest amount of code needed to make that test pass.
• Refactor: Clean up the code, enhancing its quality while maintaining its functionality and ensuring all tests remain passing.

In this lesson, we will expand upon the sum method, demonstrating how to generalize it while following these TDD principles.

Before we dive into coding, let's review our current setup. You are already familiar with the sum method and its corresponding test in the MathFunSuite:

import org.scalatest.funsuite.AnyFunSuite

class MathFunSuite extends AnyFunSuite:

  test("sum adds two numbers correctly"):
    val result = sum(2, 3)
    assert(result == 5)

def sum(a: Int, b: Int): Int = 5

This existing setup serves as a foundation. Now, we'll focus on expanding your understanding by generalizing the approach using TDD principles. Understanding where you've come from will help ensure future changes enhance our function without straying too far from the core logic.

To embrace the Red phase, let's introduce a new test case that will fail.

Update MathFunSuite to include more input scenarios:

import org.scalatest.funsuite.AnyFunSuite

class MathFunSuite extends AnyFunSuite:

  test("sum adds two numbers correctly"):
    val result = sum(2, 3)
    assert(result == 5)

  test("sum should add two more numbers correctly"):
    val result = sum(5, 6)
    assert(result == 11)

By including a new scenario with new inputs, this step is intentionally set to fail to define our target goal clearly.

Running this test will show that our implementation does not account for varying inputs, prompting us to generalize our solution.