Media targets and "sensors" are not toys - they define the state of your Application. Like a Finite State Machine state
.
Handle it holistically. Do not use react media query - use media match.
- 📦 all required matchers are built in
- 🐍 mobile-first "gap-less", and (!)bug-less approach.
- 💻 SSR friendly. Customize the target rendering mode and
SSR
for any device. - 💡 Provides
Media Matchers
to render Components andMedia Pickers
to pick a value depending on the current media. - 🎣 Provide
hooks
interface forpickers
- 🧠 Good typing out of the box - written in TypeScript
- 🚀 more performant than usual - there is only one top level query
- 🧨 Controllable matchers
https://codesandbox.io/s/react-media-match-example-g28y3
Use prebuild matchers or define your own
// custom
import { createMediaMatcher } from 'react-media-match';
const customMatcher = createMediaMatcher({
portrait: '(orientation: portrait)',
landscape: '(orientation: landscape)',
});
// prebuild
import {
breakpoints,
orientation,
darkMode,
hover,
reducedMotion,
// what else?
} from 'react-media-match/targets';
// ...
orientation.useMedia({
portrait: '📱',
landscape: '💻', // (well, actually not, but yes)
});
- Rule 1: Don't mix concerns
You shall never mix size
and orientation
, hover
and reduced-motion
- they are different slices of a one big state.
💡 If you need to respond to screen size
and orientation
- create 2 separate matchers, and use them separately!
- Rule 2: Don't match explicit target - think in states
For the every case you might have two or more
states
, only one of which can be active in a single point of view- mobile/tablet/desktop - who you are,
- portrait/landscape - how you are holding it
- hover/no-hover - there is no way they both can be true
- and visa versa.
👉 Each Media Query should be responsible only for a single dimension
- width, height, hover or orientation.
- Rule 3: Intervals Started with desktop and mobile? Then added tablet? Then added small mobile, and then large desktop? You shall be ready for a change. All API in react-media-matcher follow the pick value to the left pattern, making impossible situations when you might miss a target.
👉 Pick value to the left is the core concept. It protects you from mistakes, and allows to skip intermediate resolutions, if they should inherit styles from "lesser" query.
- Rule 4: Match all rules at once Every matcher should match only one consern, and every matcher should handle all possible variations simultaneously - it's not about what do to in case of mobile, it's also what to do in any other case.
👉 The core idea is to use object hashes to define how something should look on all targets, protecting from wide bug variations and making everything more declarative and readable.
npm install react-media-match
yarn add react-media-match
- render "forking"
<MediaMatcher
mobile={'render for mobile'}
// tablet={"tablet"} // mobile will be rendered for a "skipped" tablet - "pick value to the left"
desktop={'render desktop'}
/>
- hook interface
const title = useMedia({
mobile: shortName,
tablet: name,
// desktop: tablet will be used
});
- custom media
const Orientation = createMediaMatcher({
portrait: '(orientation: portrait)',
landscape: '(orientation: landscape)',
});
const height = Orientation.useMedia({
portrait: "50vw",
landscape: "50vw"
});
<Orientation.Matcher portrait="One" landscape="Second" />;
<Orientation.ServerRender portrait="rendering on portrait" landscape="rendering on landscape" />;
import { MediaMatcher, ProvideMediaMatchers } from 'react-media-match';
// this component will calculate all Media's and put data into the React Context
// if you will not provide it - values would be caclucaed only once, on the application start
// keep in mind - some values (like hoverability) could not change, and it's legal to skip some providers.
<ProvideMediaMatchers>
<MediaMatcher
mobile={'render for mobile'}
// tablet={"tablet"} // mobile will be rendered for "skipped" tablet
desktop={'render desktop'}
/>
<MediaMatcher
mobile={'render for mobile'}
tablet={null} // nothing will be rendered for tablet, as long you clearly "defined" it
desktop={'render desktop'}
/>
// there are also "Range" Components
<Above mobile>will be rendered on tablet and desktop</Above>
<Below desktop>will be rendered on mobile and tablet</Above>
<Below including desktop>
will be rendered on mobile, tablet and desktop
</Below>
<MediaMatches>
{' '}
// will provide matches information via render-props
{(matches) => (
<span>
{' '}
testing {
// pick matching values
pickMatch(matches, {
mobile: 'mobile',
// tablet: "tablet", // the same rules are applied here
desktop: 'desktop',
})
}
</span>
)}
</MediaMatches>
<MediaMatches>
{' '}
// will provide matches information via render-props
{(
_,
pickThisMatch // you can get pickMatch from MediaMatches
) => (
<span>
{' '}
testing {
// pick matching values, there is no need to provide "matches"
pickThisMatch({
mobile: 'mobile',
// tablet: "tablet", // the same rules are applied here
desktop: 'desktop',
})
}
</span>
)}
</MediaMatches>
// there is also "hooks" API for pickMatch
</ProvideMediaMatchers>;
PS: Don’t forget to wrap all this with ProvideMediaMatchers - without it MediaMatches will always picks the "last" branch.
react-media-match provides an API for "default" queries, and a factory method to create custom media queries.
createMediaMatcher(breakPoints: { key: string })
- factory for a new API for provided breakpoints. The object with following keys will be returned:pickMatch
useMedia
Matches
Matcher
Provider
Mock
ServerRender
Consumer
There is also pre-exported API for default breakpoints - mobile
, tablet
, desktop
-
pickMatch(mediaMatches, matchers)
- function, returns value from matchers matchingmatchers
. -
useMatch(matchers)
- hook, returns value from matchers matching matches. This call is equal topickMatch
with autowired context. -
ProvideMediaMatchers
- component, calculates media queries and stores them in context. -
MediaMatches
- component, returns current matchers as a render prop -
MediaMatcher
- component, renders path for active match -
Above
- component, renders children above specified point. Or including specified point ifincluding
prop is set. -
Below
- component, renders children below specified point. Or including specified point ifincluding
prop is set. -
MediaServerRender
- component, helps render server-size -
MediaConsumer
- React Context Consumer
- Define secondary Query for orientation
import { createMediaMatcher } from 'react-media-match';
const Orientation = createMediaMatcher({
portrait: '(orientation: portrait)',
landscape: '(orientation: landscape)',
});
<Orientation.Match portrait="One" landscape="Second" />;
Keep in mind - only value picker should be used as a hook, the render selection should
be declarative and use MediaMatcher
.
const MyComponent = ({ shortName, name }) => {
const title = useMedia({
mobile: shortName,
tablet: name,
});
return <span>Hello {title}</span>;
};
Requires React16.6+
import { MediaConsumer, pickMatch } from 'react-media-match';
// use createMediaMatcher to create your own matches
class App extends React.Component {
// provide Consumer as a contextType
static contextType = MediaConsumer;
componentDidMount() {
// use `pickMatch` matching the consumer
pickMatch(this.context, {
mobile: 'a',
tablet: 'b',
});
}
}
If you want to react to a media change you have to wrap your application with ProvideMediaMatchers
.
But if you don't - you might skip this moment.
Mobile phones(touch devices) don't have "hover" effects, while the onces with mouse
- do support it.
More of it - this could not be changed in runtime - device type is constant.
This information might be quite important - for example you might control autoFocus, as long as
auto-focusing input on a touch device would open a virtual keyboard
(consuming 50% of the screen), which may be
not desired.
In this case you might omit ProvideMediaMatchers
and use default values, which would be computed on start time.
const HoverMedia = createMediaMatcher({
touchDevice: '(hover: none)',
mouseDevice: '(hover: hover)',
});
const MyComponent = () => {
const autoFocus = HoverMedia.useMedia({
touchDevice: false,
mouseDevice: true,
});
return <input autoFocus={autoFocus} />;
};
There is no way to support MediaQuery on the Server Side, so the only way to generate the expected result is to mock a predicted device.
We are providing a special component which will
- render data in predicted device on server side,
- hydrate into it on the client side
- switch to the real values after initial hydration
import { MediaMatcher, MediaServerRender } from 'react-media-match';
<MediaServerRender predicted="desktop" hydrated={optionallyTrue}>
<MediaMatcher
mobile={'render for mobile'}
// tablet={"tablet"} // mobile will be rendered for "skipped" tablet
desktop={'render desktop'}
/>
</MediaServerRender>;
- set
hydrated
totrue
if your application is already hydrated by any reason. Omit the field to letMediaServerRender
handle hydration process automatically.
You may use ua-parser-js, to detect device type, and pick desired screen resolution, or use react-ugent to make it a bit more declarative.
Define query based on user settings
import { MediaMock, ProvideMediaMatchers } from "react-media-match";
// override all the data
<ProvideMediaMatchers state={{mobile:true, tablet:false, desktop:false}}>
....
</ProvideMediaMatchers>
<MediaMock mobile>
....
</MediaMock>
<Orientation.Mock portrait>
....
</Orientation.Mock>
Just provide state
for ProvideMediaMatchers, and it will control all the nested matchers. Could be used to provide not media-based rules.
ProvideMediaMatchers
has astate
parameter, and you might specify it override any information, and control all the nested matchers.MediaMock
will completely mock media settings.
Both mocks are not working for Inline
component.
Testing and mocking are related to SSR rendering, and you may use MediaServerRender for tests and Mocks for SSR as well.
- react-media-query-hoc implements the same idea of SSR friendly and Multiple-breakpoints approach.
- Dev.to article - Take the Responsivebility
- Medium article - Adaptive?! Responsive? Reactive!
- Spectrum chat - React-Media-Match
MIT