The pattern that you need to follow depends on your use case.
First: You might have a situation where you need to add event listener during the initial mount and clean them up at unmount and another case where a particular listener needs to be cleaned up and re-added on a prop change.
In such a case, using two different useEffect
is better to keep the relevant logic together as well as having performance benefits
useEffect(() => {
// adding event listeners on mount here
return () => {
// cleaning up the listeners here
}
}, []);
useEffect(() => {
// adding listeners everytime props.x changes
return () => {
// removing the listener when props.x changes
}
}, [props.x])
Second: You need to trigger an API call or some other side-effect when any of the state or props change from a defined set. In such a case a single useEffect
with the relevant dependencies to monitor would be better
useEffect(() => {
// side effect here on change of any of props.x or stateY
}, [props.x, stateY])
Third: You need separate side-effect for different sets of changes. In such a case, separate out relevant side-effects into different useEffect
s
useEffect(() => {
// some side-effect on change of props.x
}, [props.x])
useEffect(() => {
// another side-effect on change of stateX or stateY
}, [stateX, stateY])