Skip to content

alexlafroscia/qunit-wait-for

master
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 
 
 
src
 
 
 
 
 
 
 
 
 
 
 
 
 
 

qunit-wait-for

Verify

Wait for a QUnit Assertion

Installation

Install the dependency:

yarn add -D qunit-wait-for

and then add the following in your JavaScript code:

import QUnit from "qunit";
import { installWaitFor } from "qunit-wait-for";

installWaitFor(QUnit);

If you're using Ember, the right place for that snippet is your tests/test-helper.js.

What does it do?

A common pattern in UI testing is the idea of needing to wait for some condition to be met before moving on in your tests. Normally we do some setup, interact with our application, wait for some event to take place, and then perform our assertions. For example, you might see something like this in an Ember integration test:

// Start with some modal rendered
await render(hbs`<ModalDialog />`);

// Close the modal, which might need time to animate off-screen
await click("[data-test-close-button]");

// Wait for the modal to actually be gone
await waitUntil(() => findAll("[data-test-modal-element]").length === 0);

// Actually _assert_ that the modal is gone
assert.dom("[data-test-modal-element").doesNotExist();

Many testing libraries provide helper functions to wait for your tests to catch up to the desired state:

  • In Ember, there are many test helpers that serve this purpose; waitUntil, settled, and await-ing the promise returned from other test helpers all serve this purpose
  • In React, the async utilities from @testing-library/dom provide this functionality

However, needing to know when -- and how -- to correctly wait for the UI under test to reach the right state adds complexity to your tests and can couple them tightly to the underlying implementation of the code being tested.

There's another approach that you can take that's much cleaner; rather than waiting for your tests to reach some state and then asserting against it, you can let your assertions run immediately and gracefully handle an initial failure. Then, try the assertion again, over and over until it passes (or a timeout is reached). I call this pattern "convergence testing" based on work from The Frontside on BigTest.js. The result is a test that both correctly waits for asynchronous operations to complete and is decoupled from specific logic around how to wait for the right state.

With qunit-wait-for, the example above can be simplified to this:

await render(hbs`<ModalDialog />`);

await click("[data-test-close-button]");

await assert.waitFor(() => {
  assert.dom("[data-test-modal-element").doesNotExist();
});

Usage

To use it, pass a callback to assert.waitFor and within it place your normal assertion:

await assert.waitFor(() => {
  assert.dom("[data-test-my-element]").exists();
});

The resulting promise resolves when either the condition is met or the timeout is reached. This promise should always be await-ed so ensure one of those two things has happened before moving on.

You can also provide an override for the amount of time to wait for the assertion to pass, if needed. By default it will wait for up to 1000ms (1 second):

// Wait for up to 2 seconds instead of 1
await assert.waitFor(
  () => {
    assert.dom("[data-test-my-element]").exists();
  },
  { timeout: 2000 }
);

Prior Art