Mock
canvas
when run unit test cases with jest. For more browser environment.
- jest-random-mock Mock
Math.random
in jest, with deterministic random generator.- jest-date-mock Mock
Date
when run unit test with jest, testDate
easily.
This should only be installed as a development dependency (devDependencies
) as it is only designed for testing.
npm i --save-dev jest-canvas-mock
In your package.json
under the jest
, create a setupFiles
array and add jest-canvas-mock
to the array.
{
"jest": {
"setupFiles": ["jest-canvas-mock"]
}
}
If you already have a setupFiles
attribute you can also append jest-canvas-mock
to the array.
{
"jest": {
"setupFiles": ["./__setups__/other.js", "jest-canvas-mock"]
}
}
More about in configuration section.
Alternatively you can create a new setup file which then requires this module or
add the require
statement to an existing setup file.
__setups__/canvas.js
import 'jest-canvas-mock';
// or
require('jest-canvas-mock');
Add that file to your setupFiles
array:
"jest": {
"setupFiles": [
"./__setups__/canvas.js"
]
}
If you reset the jest mocks (for example, with jest.resetAllMocks()
), you can
call setupJestCanvasMock()
to re-create it.
import { setupJestCanvasMock } from 'jest-canvas-mock';
beforeEach(() => {
jest.resetAllMocks();
setupJestCanvasMock();
});
This mock strategy implements all the canvas functions and actually verifies the parameters. If a
known condition would cause the browser to throw a TypeError
or a DOMException
, it emulates the
error. For instance, the CanvasRenderingContext2D#arc
function will throw a TypeError
if the
radius is negative, or if it was not provided with enough parameters.
// arc throws a TypeError when the argument length is less than 5
expect(() => ctx.arc(1, 2, 3, 4)).toThrow(TypeError);
// when radius is negative, arc throws a dom exception when all parameters are finite
expect(() => ctx.arc(0, 0, -10, 0, Math.PI * 2)).toThrow(DOMException);
The function will do Number
type coercion and verify the inputs exactly like the browser does. So
this is valid input.
expect(() => ctx.arc('10', '10', '20', '0', '6.14')).not.toThrow();
Another part of the strategy is to validate input types. When using the
CanvasRenderingContext2D#fill
function, if you pass it an invalid fillRule
it will throw a
TypeError
just like the browser does.
expect(() => ctx.fill('invalid!')).toThrow(TypeError);
expect(() => ctx.fill(new Path2D(), 'invalid!')).toThrow(TypeError);
We try to follow the ECMAScript specification as closely as possible.
There are multiple ways to validate canvas state using snapshots. There are currently three methods
attached to the CanvasRenderingContext2D
class. The first way to use this feature is by using the
__getEvents
method.
/**
* In order to see which functions and properties were used for the test, you can use `__getEvents`
* to gather this information.
*/
const events = ctx.__getEvents();
expect(events).toMatchSnapshot(); // jest will assert the events match the snapshot
The second way is to inspect the current path associated with the context.
ctx.beginPath();
ctx.arc(1, 2, 3, 4, 5);
ctx.moveTo(6, 7);
ctx.rect(6, 7, 8, 9);
ctx.closePath();
/**
* Any method that modifies the current path (and subpath) will be pushed to an event array. When
* using the `__getPath` method, that array will sliced and usable for snapshots.
*/
const path = ctx.__getPath();
expect(path).toMatchSnapshot();
The third way is to inspect all of the successful draw calls submitted to the context.
ctx.drawImage(img, 0, 0);
/**
* Every drawImage, fill, stroke, fillText, or strokeText function call will be logged in an event
* array. This method will return those events here for inspection.
*/
const calls = ctx.__getDrawCalls();
expect(calls).toMatchSnapshot();
In some cases it may be useful to clear the events or draw calls that have already been logged.
// Clear events
ctx.__clearEvents();
// Clear draw calls
ctx.__clearDrawCalls();
Finally, it's possible to inspect the clipping region calls by using the __getClippingRegion
function.
const clippingRegion = ctx.__getClippingRegion();
expect(clippingRegion).toMatchSnapshot();
The clipping region cannot be cleared because it's based on the stack values and when the .clip()
function is called.
You can override the default mock return value in your test to suit your need. For example, to override return value of toDataURL
:
canvas.toDataURL.mockReturnValueOnce(
'data:image/png;base64, iVBORw0KGgoAAAANSUhEUgAAAAUAAAAFCAYAAACNbyblAAAAHElEQVQI12P4//8/w38GIAXDIBKE0DHxgljNBAAO9TXL0Y4OHwAAAABJRU5ErkJggg=='
);
MIT@hustcc.