27
Enzyme is dead. Now what?
I'm the maintainer of @wojtekmaj/enzyme-adapter-react-17. I'm here to warn you. Enzyme, a popular utility to test React components, is dead. It's time to move on. Here's why I think so.
For a long time, Enzyme has been the 1st choice when it comes to testing React applications, despite of how bumpy the road to Enzyme React 16 support was. Even today, 1/3 of React apps are still being tested using Enzyme!
Back in August 2020, React 17 Release Candidate came out. Shortly after, an issue has been raised in Enzyme repository to add support for React 17. Immediately after, @layershifter has opened a PR adding an official enzyme-adapter-react-17.
Unfortunately, there were (and, spoiler alert, still are) some issues with testing that prevented this PR from being merged.
Now, here's the thing. I'm the maintainer of many popular React packages, React-PDF, React-Calendar, and React-Date-Picker just to name a few. Professionally, I maintain several large projects, which collectively have more than 30,000 Enzyme-based unit tests.
I was in no position to wait for Enzyme to add support for React 17 to support React 17 in my packages. Rewriting all unit tests also looked like no fun to me. So, I've decided to act and publish a temporary adapter based on Oleksandr's work.
It wasn't only the adapter though: certain changes were also necessary in Enzyme adapter utils, which all Enzyme adapters depend on, so a fork of them was also needed. A couple of tireless evenings later, @wojtekmaj/enzyme-adapter-react-17 was born.
The issue for adding React 17 support remains open and has become nothing but a source of bitter comments mixed with unproductive "any updates?". Oleksandr's PR remains unmerged. enzyme-adapter-react-17 is nowhere in sight. The unofficial adapter I published with an intention of being just a stepping stone before everyone eventually migrates to enzyme-adapter-react-17, has become de facto the default adapter for React 17, with 16 million downloads so far.
Retrospectively, I'm not sure if publishing the package was a good decision.
On one hand, this move has helped thousands of developers worldwide to upgrade to React 17, relatively hassle-free.
On the other, I helped prolonging Enzyme's inevitable death and thus gave many developers hope they shouldn't get to feel. And I'm sorry for that.
Before I came to this conclusion though, React 18 beta was announced, and of course, I jumped to see if @wojtekmaj/enzyme-adapter-react-18 could be released.
The big thing in React 18 are concurrent features. To opt-in, after upgrading to React 18, you also need to switch to the new createRoot API:
// before
const container = document.getElementById('root');
ReactDOM.render(<App />, container);
// after
const container = document.getElementById('root');
const root = ReactDOM.createRoot(container);
root.render(<App/>);Until you switch to the new API, your app will behave as if it's running React 17. So if you don't plan to do this, there's no point of upgrading!
I quickly realized that the API changes in React 18 meant that releasing a React 18 Enzyme adapter will not be possible without a huge rework of not only the adapter itself, but also Enzyme, enzyme-adapter-utils and enzyme-adapter-react-helper.
Trust me, it's not gonna happen. It's over. No more "stepping stones". And certainly no more official adapters. Whether or not you plan to upgrade to React 18 in the near future, you should consider looking for Enzyme alternative right now.
The answer is, as always, it depends. You don't have to upgrade React, after all.
Here's what I would do:
- Start familiarizing yourself with React Testing Library, an officially recommended library for React components.
- Make a rule to write new tests using RTL only.
- Consider making a rule to rewrite tests to RTL whenever you need to touch them and/or the component they are testing.
- In your designated time for repaying technical debt (you have designated time for repaying technical debt, right? …right?), rewrite your remaining Enzyme-based tests to RTL.
- Clean up your repo from Enzyme-specific bits
- When you're ready, upgrade to React 18.
While Migrate from Enzyme support article is available, I suggest you to just start fresh, forgetting that Enzyme has ever existed. RTL is by no means an Enzyme drop-in replacement, so having a completely fresh mindset will help you getting the most of it.
27