/react-three-a11y

♿️ Accessibility tools for React Three Fiber

Primary LanguageTypeScriptMIT LicenseMIT

@react-three/a11y

Version Downloads Discord Shield

Imgur

npm install @react-three/a11y

@react-three/a11y brings accessibility to webGL with easy-to-use react-three-fiber components:

  • Focus and focus indication
  • Tab index and keyboard navigation
  • Screen reader support and alt-text
  • Roles and cursor shapes
  • Descriptive links

Live demo: https://n4rzi.csb.app

How to use

First, place the A11yAnnouncer component next to the R3F Canvas component. This is critical, because it will manage the screen-reader and help emulate focus!

import { Canvas } from 'react-three-fiber'
import { A11yAnnouncer } from '@react-three/a11y'

function App() {
  return (
    <>
      <Canvas />
      <A11yAnnouncer />
    </>
  )
}

To add accessibility features to your scene you'll have to wrap components you want to make focusable with the A11y component:

import { A11y } from '@react-three/a11y'

<A11y>
  <MyComponent />
</A11y>

MyComponent can now receive focus. More accurately, the emulated "focus" will handled at the A11y components which acts as a provider for children to access its state. But even if objects are focusable, nothing will be displayed or shown by default.

Accessing the hover, focused & pressed state

For each child wrapped in the A11y component, you can access the focus, hover and pressed state like so:

import { useA11y } from '@react-three/a11y'

function Box(props) {
  const a11y = useA11y()
  return (
    <mesh {...props}>
      <boxBufferGeometry />
      <meshStandardMaterial color={a11y.hover || a11y.focus ? 'hotpink' : 'orange'} />
    </mesh>
  )
}

Call function on focus

The focusCall prop of A11y will be called each time this component receives focus (usually through tab navigation).

<A11y role="content" focusCall={()=> console.log("in focus")} ... />

Call function on click / keyboard Click

The actionCall prop of A11y will be called each time this component gets clicked, focused, keyboard activated etc.

<A11y role="button" actionCall={()=> console.log("clicked")} ... />

Provide a description of the currently focused / hovered element

When using the description prop, the A11y component will provide a description to the screen reader users on focus/hover. Optionally, you can also show the description to the user on hover by setting showAltText={true}.

// Reads "A rotating red square" to screen readers on focus / hover
<A11y role="content" description="A rotating red square" ... />
// Reads "A bouncing blue sphere" to screen readers on focus / hover while also showing it on mouseover
<A11y role="content" description="A bouncing blue sphere" showAltText ... />

If your A11y component has the role="button", you can use three more props:

  • activationMsg: When the user will click/activate the "button" the screen reader will read the passed message
  • deactivationMsg: When set, it turns your button in a togglable button. Which means it now has a on/off state. Screen readers will read the state of the button as well as the activation/disactivation messages passsed.
  • pressedDescription: When set, it turns your button in a togglable button. Which means it now has a on/off state. This description will replace the one passed via description when the toggle is active.
// Reads the description on hover/focus then will read activationMsg if clicked/pressed
<A11y role="button" description="Sends a thank you email to the team" activationMsg="Email is sending" ... />
// Reads the description on hover/focus then will read activationMsg if turned on or deactivationMsg if tuned off
<A11y
  role="button"
  description="This button can enable dark theme. Dark theme is off"
  pressedDescription="This button can disable dark theme. Dark theme is on"
  activationMsg="Dark theme enabled"
  deactivationMsg="Dark theme disabled" ... />

The three roles of the A11y component

Like in HTML, you can focus different kind of elements and expect different things depending on what you're focusing.

Content

<A11y role="content" ... />

Uses the default cursor. This role is meant to provide information to screen readers or to serve as a step for a user to navigate your site using Tab for instance. It's not meant to trigger anything on click or to be activable with the Keyboard. Therefore it won't show a pointer cursor on hover.

Buttons

<A11y
  role="button"
  activationMsg="Button activated"
  deactivationMsg="Button deactivated"
  pressedDescription="Button pressed" ... />

Uses the pointer cursor. Special attributes: activationMsg, deactivationMsg and pressedDescription.

This role is meant to emulate the behaviour of a button or a togglable button. It will display a cursor pointer when your cursor is over the linked 3D object. It will call a function on click but also on any kind of action that would trigger a focused button (Enter, Double-Tap, ...). It is also actionnable by user using a screen reader.

You can turn it into a button with aria-pressed by providing the following properties deactivationMsg, pressedDescription in addition to the usual description and activationMsg properties.

Links

<A11y role="link" href="https://url.com" ... />

Uses the pointer cursor. Special attributes: href.

This role is meant to emulate the behaviour of a regular html link. It should be used in combination with something that will trigger navigation on click.

- Don't forget to provide the href attribute as he is required for screen readers to read it correctly !
- It will have no effect on the navigation, it's just used as information

Screen Reader Support

In order to provide informations to screen reader users and use this package at its full potential, fill the description prop of all your A11y components and use the appropriate role prop on each of them.

Additionals Features

Use a custom tabindex with for your A11y components by providing a number to the tabIndex attribute

<A11y tabIndex={2} ... />

More about the use of tabIndex on developer.mozilla.org

Next Steps

  • Provide a documentation inside the IDE
  • Provide more examples showing how to use it

Author: