React

Introduction To React Testing

new

React Course

Introduction

We’ve learned how to test our vanilla JavaScript applications in a previous section. Testing is a powerful tool that allows us to write maintainable and flexible code. If you’ve followed along with our lessons, so far we’ve been using the Jest framework. However, since we’re using Vite we’ll switch over to Vitest as our test runner since it integrates nicely with Vite. We’ll add more capabilities to our tests using the React Testing library (RTL).

Lesson overview

This section contains a general overview of topics that you will learn in this lesson.

  • How to set up a React testing environment.
  • How to test UI elements.
  • Understanding snapshot tests.

Setting up a React testing environment

Follow along Robin Wieruch’s guide on setting up Vitest with RTL. Once you’ve completed the setup, let’s meet back here.

Phew, that was a lot of setup. But there’s one more tiny package to install before we can begin:

npm install @testing-library/user-event --save-dev

Now that we have everything we need, let’s briefly go over what some of those packages do. We’ll primarily focus on the @testing-library packages.

  • @testing-library/react will give us access to useful functions like render which we’ll demonstrate later on.

  • @testing-library/jest-dom includes some handy custom matchers (assertive functions) like toBeInTheDocument and more. (complete list on jest-dom’s github). Jest already has a lot of matchers so this package is not compulsory to use.

  • @testing-library/user-event provides the userEvent API that simulates user interactions with the webpage.

Our first query

First, we’ll render the component using render. The API will return an object and we’ll use destructuring syntax to obtain a subset of the methods required. You can read all about what render can do in the React Testing Library API docs about render.

// App.jsx

const App = () => <h1>Our First Test</h1>;

export default App;
// App.test.jsx

import { describe, it, expect } from "vitest";
import { render, screen } from "@testing-library/react";
import App from "./App";

describe("App component", () => {
  it("renders correct heading", () => {
    render(<App />);
    expect(screen.getByRole("heading").textContent).toMatch(/our first test/i);
  });
});

Vitest globals and ESLint

Even if you set globals: true in vite.config.js like in the setup tutorial, ESLint will still yell at you, as it will not recognize these globals without some extra configuration in your .eslintrc.cjs file. The most straightforward resolution would be to explicitly import the globals you’d need. You can omit globals: true from vite.config.js in this case.

Execute npm test App.test.jsx on the terminal and see the test pass. getByRole is just one of the dozen query methods that we could’ve used. Essentially, queries are classified into three types: getBy, queryBy and findBy. Go through the React Testing Library docs page about queries. Pay extra attention to the “Types of Queries” and “Priority” sections.

As stated by the React Testing Library docs, ByRole methods are favored methods for querying, especially when paired with the name option. For example, we could improve the specificity of the above query like so: getByRole("heading", { name: "Our First Test" }). Queries that are done through ByRole ensure that our UI is accessible to everyone no matter what mode they use to navigate the webpage (i.e. mouse or assistive technologies).

Simulating user events

There are numerous ways a user can interact with a webpage. Even though live user feedback and interaction is irreplaceable, we can still build some confidence in our components through tests. Here’s a button which changes the heading of the App:

// App.jsx

import { useState } from "react";

const App = () => {
  const [heading, setHeading] = useState("Magnificent Monkeys");

  const clickHandler = () => {
    setHeading("Radical Rhinos");
  };

  return (
    <>
      <button type="button" onClick={clickHandler}>
        Click Me
      </button>
      <h1>{heading}</h1>
    </>
  );
};

export default App;

Let’s test if the button works as intended. In this test suite, we’ll use a separate utility to query our UI elements. React Testing Library provides the screen object which has all the methods for querying. With screen, we don’t have to worry about keeping render’s destructuring up-to-date. Hence, it’s better to use screen to access queries rather than to destructure render.

// App.test.jsx

import { describe, it, expect } from "vitest";
import { render, screen } from "@testing-library/react";
import userEvent from "@testing-library/user-event";
import App from "./App";

describe("App component", () => {
  it("renders magnificent monkeys", () => {
    // since screen does not have the container property, we'll destructure render to obtain a container for this test
    const { container } = render(<App />);
    expect(container).toMatchSnapshot();
  });

  it("renders radical rhinos after button click", async () => {
    const user = userEvent.setup();

    render(<App />);
    const button = screen.getByRole("button", { name: "Click Me" });

    await user.click(button);

    expect(screen.getByRole("heading").textContent).toMatch(/radical rhinos/i);
  });
});

The tests speak for themselves. In the first test, we utilize snapshots to check whether all the nodes render as we expect them to. In the second test, we simulate a click event. Then we check if the heading changed. toMatch is one of the various assertions we could have made. Notice that the callback function for the second test is an async one, as we need this in order to await user.click().

It’s also important to note that after every test, React Testing Library unmounts the rendered components. That’s why we render for each test. For a lot of tests for a component, the beforeEach Vitest function could prove handy.

What are snapshots?

Snapshot testing is just comparing our rendered component with an associated snapshot file. For example, the snapshot file which was automatically generated after we ran the “magnificent monkeys renders” test was:

// Vitest Snapshot v1, https://vitest.dev/guide/snapshot.html

exports[`App component > renders magnificent monkeys 1`] = `
<div>
  <button
    type="button"
  >
    Click Me
  </button>
  <h1>
    Magnificent Monkeys
  </h1>
</div>
`;

It’s an HTML representation of the App component. And it will be compared against the App in future snapshot assertions. If the App changes even slightly, the test fails.

Snapshot tests are fast and easy to write. One assertion saves us from writing multiple lines of code. For example, with a toMatchSnapshot, we’re spared of asserting the existence of the button and the heading. They also don’t let unexpected changes creep into our code. Read all about what can be achieved with snapshots in the Vitest snapshot docs.

Snapshots might seem the best thing that has happened to us while testing thus far. But we are forced to wonder, what exactly are we testing? What’s being validated? If a snapshot passes, what does it convey about the correctness of the component?

Snapshot tests may cause false positives. Since we cannot ascertain the validity of the component from a snapshot test, a bug might go undetected. Over-reliance on snapshots can make developers more confident about their code than they should be.

The other issue with snapshots is false negatives. Even the most insignificant of changes compel the test to fail. Fixing punctuation? Snapshot will fail. Replacing an HTML tag to a more semantic one? Snapshot will fail. This might cause us to lose our confidence in the test suite altogether. Snapshots aren’t inherently bad; they do serve a purpose. But it’s beneficial to understand when to snapshot, and when not to snapshot.

Assignment

Focus on the concepts

Even though some articles use Jest and the Enzyme testing library, the concepts should be transferrable.

  1. Testing Implementation Details by Kent C. Dodds shows us how we can reduce false test results and inflexible tests by avoiding testing the implementation of things.
  2. Take a glance at all of the available query methods on the React Testing Library’s cheatsheet page. There’s no need to use them all, but it’s optimal to employ a specific method for a specific query. If none of the query methods suffice, there’s an option to use test ids. Learn about test ids on the React Testing Library’s test id docs.
  3. Read the userEvent API docs to get a feel of how to achieve user simulation.
  4. This article on the Pros and Cons of Snapshot Tests goes in depth regarding the advantages and disadvantages of snapshot testing. And this one, Snapshot Testing: Benefits and Drawbacks, does an excellent job of explaining what snapshot testing is for programming in general.

Knowledge check

The following questions are an opportunity to reflect on key topics in this lesson. If you can’t answer a question, click on it to review the material, but keep in mind you are not expected to memorize or master this knowledge.

Additional resources

This section contains helpful links to related content. It isn’t required, so consider it supplemental.

Support us!

The Odin Project is funded by the community. Join us in empowering learners around the globe by supporting The Odin Project!