Awesome
jsdom-testing-mocks
A set of tools for emulating browser behavior in jsdom environment
Mocks
matchMedia
Intersection Observer
Resize Observer
Web Animations API
Installation
npm i -D jsdom-testing-mocks
yarn add -D jsdom-testing-mocks
Setup
With react
To avoid having to wrap everything in act calls, you can pass act to configMocks:
import { configMocks } from 'jsdom-testing-mocks';
import { act } from '...';
configMocks({ act });
It can be done in a setup file, or in a test file, before rendering the component.
With vitest
Some mocks require lifecycle hooks to be defined on the global object. To make it work with vitest, you need to enable globals in your config. If you don't want to do that you can pass it manually using configMocks.
Also, if you're using fake timers, at the time of writing this, vitest doesn't enable faking performance.now, requestAnimationFrame and cancelAnimationFrame by default, so you need to do it manually:
vi.useFakeTimers({
toFake: [
// vitests defaults
'setTimeout',
'clearTimeout',
'setInterval',
'clearInterval',
'setImmediate',
'clearImmediate',
'Date',
// required for mocks
'performance',
'requestAnimationFrame',
'cancelAnimationFrame',
],
});
Testing framework support
We aim to support all major testing frameworks that support jsdom. Internally, there are no dependencies on any of them, so it's likely that it will work out of the box. Currently tested and confirmed to work with jest, @swc/jest and vitest (with some setup). If you encounter any problems with other testing frameworks, please open an issue.
Mock viewport
Mocks matchMedia, allows testing of component's behavior depending on the viewport description (supports all of the Media Features). mockViewport must be called before rendering the component
Example, using React Testing Library:
import { mockViewport } from 'jsdom-testing-mocks';
it('shows the right lines on desktop and mobile', () => {
const viewport = mockViewport({ width: '320px', height: '568px' });
render(<TestComponent />);
expect(
screen.getByText('Content visible only on small screens')
).toBeInTheDocument();
expect(
screen.queryByText('Content visible only on large screens')
).not.toBeInTheDocument();
act(() => {
viewport.set({ width: '1440px', height: '900px' });
});
expect(
screen.queryByText('Content visible only on small screens')
).not.toBeInTheDocument();
expect(
screen.getByText('Content visible only on large screens')
).toBeInTheDocument();
viewport.cleanup();
});
Also, you can mock the viewport for a group of tests, using mockViewportForTestGroup:
import { mockViewportForTestGroup } from 'jsdom-testing-mocks'
describe('Desktop specific tests', () => {
mockViewportForTestGroup({ width: '1440px', height: '900px' })
test('this', () = {
// ...
})
test('that', () = {
// ...
})
})
Mock IntersectionObserver
Provides a way of triggering intersection observer events
Example, using React Testing Library:
import { mockIntersectionObserver } from 'jsdom-testing-mocks';
const io = mockIntersectionObserver();
/*
Assuming html:
<div data-testid="container">
<img src="..." alt="alt text" />
</div>
And an IntersectionObserver, observing the container
*/
it('loads the image when the component is in the viewport', () => {
const { container } = render(<TestComponent />);
expect(screen.queryByAltText('alt text')).not.toBeInTheDocument();
// when the component's observed node is in the viewport - show the image
act(() => {
io.enterNode(screen.getByTestId('container'));
});
expect(screen.getByAltText('alt text')).toBeInTheDocument();
});
API
mockIntersectionObserver returns an object, that has several useful methods:
.enterNode(node, desc)
Triggers all IntersectionObservers observing the node, with isIntersected set to true and intersectionRatio set to 1. Other IntersectionObserverEntry params can be passed as desc argument, you can override any parameter except isIntersected
.leaveNode(node, desc)
Triggers all IntersectionObservers observing the node, with isIntersected set to false and intersectionRatio set to 0. Other IntersectionObserverEntry params can be passed as desc argument, you can override any parameter except isIntersected
.enterNodes(nodeDescriptions)
Triggers all IntersectionObservers observing the nodes in nodeDescriptions with multiple nodes entering at once. Each IntersectionObserver callback will receive only the nodes it's observing:
io.enterNodes([
// you can pass multiple nodes each with its own state
{ node: screen.getByText('First Node'), desc: { intersectionRatio: 0.5 } },
// description is optional:
{ node: screen.getByText('Second Node') },
// or you can use a shorthand:
screen.getByText('Third Node'),
]);
.leaveNodes(nodeDescriptions)
Triggers all IntersectionObservers observing the nodes in nodeDescriptions with multiple nodes leaving at once. Each IntersectionObserver callback will receive only the nodes it's observing.
.triggerNodes(nodeDescriptions)
Triggers all IntersectionObservers observing the nodes in nodeDescriptions with multiple nodes at once with custom descriptions (isIntersected is not enforced). Each IntersectionObserver callback will receive only the nodes it's observing
.enterAll(desc) and .leaveAll(desc)
Triggers all IntersectionObservers for each of the observed nodes
Mock ResizeObserver
Mocks ResizeObserver class. Resize callbacks are triggered manually using resize method returned by the mock. Elements' size must not be 0 (at least on one axis) for the element to appear in the list of callback entries (you can mock the size using mockElementSize or mockElementBoundingClientRect)
Example, using React Testing Library:
import { mockResizeObserver } from 'jsdom-testing-mocks';
const DivWithSize = () => {
const [size, setSize] = useState({ width: 0, height: 0 });
const ref = useRef(null);
useEffect(() => {
const observer = new ResizeObserver((entries) => {
setSize({
width: entries[0].contentRect.width,
height: entries[0].contentRect.height,
});
});
observer.observe(ref.current);
return () => {
observer.disconnect();
};
}, []);
return (
<div data-testid="theDiv" ref={ref}>
{size.width} x {size.height}
</div>
);
};
const resizeObserver = mockResizeObserver();
it('prints the size of the div', () => {
render(<DivWithSize />);
const theDiv = screen.getByTestId('theDiv');
expect(screen.getByText('0 x 0')).toBeInTheDocument();
resizeObserver.mockElementSize(theDiv, {
contentBoxSize: { inlineSize: 300, blockSize: 200 },
});
act(() => {
// on the first run you don't have to pass the element,
// it will be included in the list of entries automatically
// because of the call to .observe
resizeObserver.resize();
});
expect(screen.getByText('300 x 200')).toBeInTheDocument();
resizeObserver.mockElementSize(theDiv, {
contentBoxSize: { inlineSize: 200, blockSize: 500 },
});
act(() => {
// on subsequent calls to `resize` you have to include it
// explicitly, unless observe has been called on it again
resizeObserver.resize(theDiv);
});
expect(screen.getByText('200 x 500')).toBeInTheDocument();
});
Caveats
Triggering the callback on observe
Although the mock doesn't call the resize callback on its own, it keeps track of all the cases when it should be implicitly called (like when the element first begins being observed), and it auto-adds them to the list of elements when resize is called. You can disable this in ResizeOptions
Mocking element's size
The mock uses the size provided by mockElementSize if present and fallbacks to getBoundingClientRect (that you can mock using mockElementBoundingClientRect). The issue with getBoundingClientRect however is that in the real world the value it returns takes CSS Transforms into account, while the values returned in the observer callback don't. It doesn't really matter because it is you who mocks sizes, but for consistency it is preferred that you use mockElementSize
API
mockResizeObserver returns an object, that has several methods:
.resize(elements?: HTMLElement | HTMLElement[], options: ResizeOptions)
Triggers all resize observer callbacks for all observers that observe the passed elements. Some elements are implicitly resized by the Resize Observer itself, for example when they first attached using observe. This mock doesn't call the callback by itself. Instead, it adds them to the list of entries when the next resize is called (it happens only once per observe per element).
In this example the resize callback will be triggered with all observed elements from within TestedComponent:
// a component that begins to observe elements in a useEffect
render(<TestedComponent />);
// ...don't forget to mock sizes
act(() => {
// triggers the `resize` callback with the elements for which `observe` has been called
resizeObserver.resize();
});
ResizeOptions.ignoreImplicit (false by default)
If true, do not include imlicit elements in the resize callback entries array
.mockElementSize(element: HTMLElement, size: Size)
Mocks element's size only for the ResizeObserver. size accepts 2 properties: contentBoxSize and borderBoxSize they're both similar to what you see in the ResizeObserver's callback entry. At least one of them must be present (if the other isn't it is set to be equal to the one present), and the other entry properties are derived from these two (and window.devicePixelRatio).
Example:
mockElementSize(myDiv, {
// both contentBoxSize and borderBoxSize accept plain objects instead of arrays
contentBoxSize: { inlineSize: 400, blockSize: 200 },
});
mockElementSize(myOtherDiv, {
// only one dimension is required, the other one will be assumed to be 0
borderBoxSize: { inlineSize: 200 },
});
.getObservers(element?: HTMLElement)
Returns all observers (observing the element if passed)
.getObservedElements(observer?: ResizeObserver)
Returns all observed elements (of the observer if passed)
Mock Web Animations API
Warning: experimental, bug reports, tests and feedback are greatly appreciated
Mocks WAAPI functionality using requestAnimationFrame. With one important limitation — there are no style interpolations. Each frame applies the closest keyframe from list of passed keyframes or a generated "initial keyframe" if only one keyframe is passed (initial keyframe removes/restores all the properties set by the one keyframe passed). As the implementation is based on the official spec it should support the majority of cases, but the test suite is far from complete, so here be dragons
Example, using React Testing Library:
import { mockAnimationsApi } from 'jsdom-testing-mocks';
const TestComponent = () => {
const [isShown, setIsShown] = useState(false);
return (
<div>
{/* AnimatePresence is a component that adds its children in the dom
and fades it in using WAAPI, with 2 keyframes: [{ opacity: 0 }, { opacity: 1 }],
also adding a div with the word "Done!" after the animation has finished
You can find implementation in examples
*/}
<AnimatePresence>{isShown && <div>Hehey!</div>}</AnimatePresence>
<button
onClick={() => {
setIsShown(true);
}}
>
Show
</button>
</div>
);
};
mockAnimationsApi();
it('adds an element into the dom and fades it in', async () => {
render(<TestComponent />);
expect(screen.queryByText('Hehey!')).not.toBeInTheDocument();
await userEvent.click(screen.getByText('Show'));
// assume there's only one animation present in the document at this point
// in practice it's better to get the running animation from the element itself
const element = screen.getByText('Hehey!');
const animation = document.getAnimations()[0];
// our AnimatePresence implementation has 2 keyframes: opacity: 0 and opacity: 1
// which allows us to test the visibility of the element, the first keyframe
// is applied right after the animation is ready
await animation.ready;
expect(element).not.toBeVisible();
// this test will pass right after 50% of the animation is complete
// because this mock doesn't interpolate keyframes values,
// but chooses the closest one at each frame
await waitFor(() => {
expect(element).toBeVisible();
});
// AnimatePresence will also add a div with the text 'Done!' after animation is complete
await waitFor(() => {
expect(screen.getByText('Done!')).toBeInTheDocument();
});
});
Using with fake timers
It's perfectly usable with fake timers, except for the issue with promises. Also note that you would need to manually advance timers by the duration of the animation taking frame duration (which currently is set to 16ms in jest/sinon.js) into account. So if you, say, have an animation with a duration of 300ms, you will need to advance your timers by the value that is at least the closest multiple of the frame duration, which in this case is 304ms (19 frames * 16ms). Otherwise the last frame may not fire and the animation won't finish.
Current issues
- Needs more tests