Join egghead, unlock knowledge.

Want more egghead? It's 45% off for a limited time only!

This lesson is for members. Join us? Get access to all 3,000+ tutorials + a community with expert developers around the world.

Unlock All Content for 45% Off

Already subscribed? Sign In

Save 45% for a limited time.

Get access to all courses and lessons on egghead today.

Autoplay

    Support a state change handler for all control props

    Kent C. DoddsKent C. Dodds

    In our simple situation, having an onToggleChange handler is sufficient because the on state is the only state we have, but in a more complex component, there could be many items of state. We could make an prop callback for each of those items of state, but that could drastically increase the complexity and repetition of our component. Instead, let's make an onStateChange that's called when any state changes.

    reactReact
    javascriptJavaScript
    Code

    Code

    Become a Member to view code

    You must be a Member to view code

    Access all courses and lessons, track your progress, gain confidence and expertise.

    Become a Member
    and unlock code for this lesson
    Transcript

    Transcript

    Instructor: 00:00 Currently, the way that users of the toggle component know that state has changed is because we call this onToggle prop, but this onToggle prop is just notifying consumers of a single state change.

    00:11 Often, your components will have more than just a single boolean that they're managing. Writing code like this for every single time you're going to be calling this.setState might get a little bit overwhelming. Instead, let's package this up to be a little bit more generic and allow for any number of items of state.

    00:27 Rather than an onToggle prop, our consumers now are expecting an onStateChange prop that will be called anytime any state changes. We'll expect it to be called with an object with only the state that changed. Here, we're destructuring that to update our controlled state.

    00:42 What we're going to do is we'll create a new method called internalSetState. This will resemble the same API that setState has, that is, changes -- that could be an object or an updater function -- and callback. Here we'll call this.setState with changes and callback. Then we'll update our setState call to be internalSetState. Nothing has changed so far.

    01:04 Where the magic is going to happen is inside of this setState call. We're going to create an updater function. First, we'll get the combined state. That'll be this.getState. We can't get it from this.state like we are here because if this is running as part of a batch, this.state is unreliable. We need to use the state that we're given from our updater function.

    01:24 What we'll do is, in our getState, we'll adapt that to accept a state and default to this.state. Then we'll pass our state here. Next, we're going to get a changes object. That will be type of changes is a function. We'll call changes with our combined state. Otherwise, it will simply be changes.

    01:46 Next, we'll create an object called non-controlled changes. That will be based on object entries of our changes object. We'll reduce that into an object that represents only the changes that are not controlled. We can call our accumulator newChanges. We'll destructure the key and value out of our entries. We'll return the newChanges. We'll default newChanges to an object.

    02:12 Next, we'll say if this is not controlled, then newChanges at the key will equal the value. Then we'll return the non-controlled changes. In our callback, that also needs to be updated, so we call the onStateChange. We'll invoke the callback as we were before, but we'll also call this.props.onStateChange with this.getState.

    02:32 Let's go ahead and try that out. Looks like onToggle is not a function, so let's go ahead and make a default. In our case, we no longer need onToggle, because we have onStateChange. You really just need one or the other to use the toggle component properly.

    02:48 We'll have a static defaultProps. We'll have an onToggle that defaults to an empty function. OnStateChange will also default to an empty function. Now let's try this. Uh-oh. It's not working.

    03:03 The next thing we need to do is update our toggle function so that it's going to call internalSetState regardless of whether it's controlled, because we've handled that case in the internalSetState function. We'll get rid of this.

    03:14 Let's go ahead and try that out. That's not working. Here's the problem. When we call this.props.onStateChange, we're calling it with this.getState. This.getState is going to get us the combined state between this.state and this.props. Our onState is actually controlled, and so it is going to come from this.props.

    03:35 If we look at the props we're being passed, we're getting passed this both on. Both on gets updated when handle state change happens. That happens when we call onStateChange. Our problem is that we're calling onStateChange with this.getState at the point where our props have not yet been updated. GetState is going to return us an object that has on as false. That's what's happening here.

    03:59 How do we fix this? We need to call onStateChange with not only our internal state but also our recommendations for what we believe the state change should be. That resides in the changes object. What we're going to do here is we'll make an allChanges variable. Then we'll assign allChanges to the changes object.

    04:19 Instead of calling onStateChange with this.getState, which contains only the current state derived from internal state and props, we'll pass allChanges, which includes our recommendations for the changes. In this case, we're recommending that they toggle the onState. With those changes, we can now use our component properly.

    04:38 In review, the problem we're trying to solve is to make our controlled component more generic so we can simplify every time we call this.setState. It does add a fair amount of complexity, but in a more real-world component that manages multiple items of state, this will save you a lot of complexity every place you call this.setState.

    04:55 To make this work, we used this.setState with a custom updater function. We used the state that we get from that custom updater function to get the combined state between our props and our internal managed state. Then we used that combined state to get the changes object from the changes function that were passed when we call internalSetState.

    05:15 With that changes object, we maintain a reference to that, so we can use it later. We filter that object down to just the changes that are not controlled so that we avoid unnecessary re-renders. Then we return that non-controlled changes object.

    05:27 After the setState has happened, we call onStateChange with all the changes that we recommend and any callback that is provided to our internalSetState call. With the way that this is set up, this could actually be an empty object if all the changes are controlled, which is the case with our toggle component, because we're returning an object that could actually cause a re-render.

    05:47 To make sure we avoid that, we're going to say Object.keys.length. If it has length, then we'll return the non-controlled changes. Otherwise, we'll return null. That will ensure React does not re-render our component if there are no internal state changes.

    06:03 Let's make one last simple change. When we call onStateChange, as a convenience, we'll also provide this.getState so that users of the onStateChange API get the changes as well as the current state.

    Discuss

    Discuss