Unit Testing

While most of our testing workflow is automated (CI/CD), understanding how to manually test your code locally helps you:

• Validate changes before pushing.
• Debug failing tests.
• Run only the tests relevant to your feature.

1. 🧠 What Is Jest?

Jest is a JavaScript testing framework built for simplicity.

It works perfectly with NestJS and supports:

• Unit tests (testing a single service, controller or other component)
• Integration tests (testing multiple components together)
• Mocks (fake services, data, etc.)

2. 🧪 Using Jest with NestJS

Before trying to run tests, make sure you have installed the dependencies:

1. Check your package.json for the @nestjs/testing package.

2. If the package is not installed, run npm install --save-dev @nestjs/testing. (--save-dev adds the dependency to devDependencies only.)

3. Check your package.json -> scripts for the test script.

4. If the script is not defined, add it:

   "scripts": {
     "test": "jest",
   }

5. You good to go! Run npm run test to execute the tests (the project already contains simple tests inside the tests folder).

6. You should see something like this:

 PASS  src/app.controller.spec.ts
  AppController
    root
      ✓ should return "Hello World!" (5 ms)

Test Suites: 1 passed, 1 total
Tests:       1 passed, 1 total

3. 🔬 Testing the Tasks module

Now its time to write some tests!

Create a new folder inside the tests called tasks.

3.1. Controller tests

Let's start with the controller tests. Create a new file called tasks.controller.spec.ts inside the tests/tasks folder.

In this file we should think about possible use cases and edge-cases where the controller might fail.

For example, we have the findAll method that returns all tasks. We should test the following cases:

• The database is empty.
• The database contains one task.
• The database contains multiple tasks.

We will not go deep into the tests here, as it is just a high-level overview, but this gives you an idea of how to write tests.

3.2 Service tests

Now let's move on to the service tests. Create a new file called tasks.service.spec.ts inside the tests/tasks folder.

This one has something special:

constructor(
    @InjectModel(Task.name)
    private readonly taskModel: Model<TaskDocument>,
  ) {}

The InjectModel decorator is a NestJS decorator that allows you to inject a Mongoose model into a service by its name.

We have 2 options here:

1. Create "in-memory" database and use it for testing (the same MongoDB database but created on a fly just for testing).
2. "Mock" the database/model using Jest or any other mocking library.

For this tutorial, we will use the second option as it is easier to set up.

By creating and injecting our own mock object

const mockModel = {
  find: jest.fn(),
};

we can still use the service as before, but instead of using the real database, it will use the mockModel object. This way, we can check if the methods are called correctly and with the correct arguments.

describe('TasksService', () => {
  let service: TasksService;

  beforeEach(async () => {
    const module: TestingModule = await Test.createTestingModule({
      providers: [
        TasksService,
        {
          provide: getModelToken(Task.name), // <- this matches @InjectModel(Task.name)
          useValue: mockTaskModel,          // <- your Jest mock
        },
      ],
    }).compile();

3.3 DTO tests

Finally, let's write some tests for the DTO. Create 1 test file for each dto object inside the test/tasks/dto folder.

We should verify that the DTO is correctly constructed and that it has the correct properties.

Example:

// create-task.dto.test.ts

export const CreateTaskSchema = z.object({
  title: z
    .string()
    .min(1, "Title is required")
    .max(80, "Keep the title under 80 characters"),
  description: z.string().max(200, "Description should be short").optional(),
});

// given the schema above we can write a test like this:

describe("CreateTaskDto", () => {
  it("should throw an error if the title length is above 80 characters", () => {
    const data = {
      title: "a".repeat(81),
      description: "test description",
    };

    const result = CreateTaskSchema.safeParse(data);

    expect(result.success).toBeFalsy();

    // other expects to be more specific (what error was thrown, etc.)
  });
});

In the next section we will cover how to write integration tests.