Getting started with state management using useReducer and Context
Choosing a state management library for your React app can be tricky. Some of your options include:
- Using React’s
useReducer
hook in combination with React Context - Going for a longstanding and popular library like Redux or MobX
- Trying something new like react-sweet-state or Recoil (if you’re feeling adventurous!)
To help you make a more informed decision, this series aims to give a quick overview of creating a to-do list app using a variety of state management solutions.
In this post we will be using a combination of the useReducer
hook and React Context to build our example app, as well as a quick detour to take a look at a library called React Tracked.
If you want to follow along, I have created a repository for the example app created in this guide at react-state-comparison.
This post assumes knowledge of how to render functional components in React, as well as a general understanding of how hooks work.
App functionality and structure
The functionality we will be implementing in this app will include the following:
- Editing the name of the to-do list
- Creating, deleting and editing a task
The structure of the app will look something like this:
Creating our common components
First we’ll be creating some components in our common
folder. These “view” components won’t have any knowledge of what state management library we are using. Their sole purpose will be to render a component, and to use callbacks that we pass in as props. We’re putting them in a common folder so that we can re-use them in future posts in this series.
We’ll need four components:
NameView
- a field to let us edit the to-do list’s nameCreateTaskView
- a field with a “create” button so we can create a new taskTaskView
- a checkbox, name of the task, and a “delete” button for the taskTasksView
- loops through and renders all the tasks
As an example, the code for the Name
component will look like this:
Each time we edit the name, we’ll be calling the onSetName
callback with the current value of the input (accessed through the event
object).
In a real-life app, you might think about holding off on making this call until the user has saved the task’s name. You could either have a “save” button for this, or listen for the user to leaving the input field by clicking away or pressing enter.
The code for the other three components follow a similar sort of pattern, which you can check out in the common/components folder.
Defining the shape of our store
Next we should think about how our store should look. With local state, your state lives inside of individual React components. In contrast to this, a store is a central place where you can put all the state for your app.
We’ll be storing the name of our to-do list, as well as a tasks map that contains all our tasks mapped against their IDs:
Creating our reducer and actions
A reducer and actions is what we use to modify the data in our store.
An action’s job is to ask for the store to be modified. It will say:
“Hey, I want to change the to-do list’s name to be ‘Fancy new name’”.
The reducer’s job is to modify the store. The reducer will receive that request, and go:
“Okay, I will change the to-do list’s name to be ‘Fancy new name‘“
Actions
Each action will have two values:
- An action’s
type
- to update the list’s name you could define the type asupdateListName
- An action’s
payload
- to update the list’s name, the payload would contain “Fancy new name”
Dispatching our updateListName
action would look something like this:
Reducers
A reducer is where we define how we will modify the state using the action’s payload. It’s a function that takes in the current state of the store as its first argument, and the action as its second:
With a switch statement, the reducer will attempt to find a matching case for the action. If the action isn’t defined in the reducer, we would enter the default
case and return the state
object unchanged.
If it is defined, we will go ahead and return a modified version of the state
object. In our case, we would change the listName
value.
A super-important thing to note here is that we never directly modify the state object that we receive. e.g. Don’t do this:
We need our app to re-render when values in our store are changed, but if we directly modify the state object this won’t happen. We need to make sure that we return new objects. If you don’t want to do this manually, there are libraries like immer that will do this safely for you.
Creating and initialising our store
Now that we’ve defined our reducer and actions, we need to create our store using React Context and useReducer
:
The useReducer
hook allows us to create a reducer using the reducer function we defined earlier. We also pass in an initial state object, which might look something like this:
When we wrap the Provider around our app, any component will be able to access the state
object to render what it needs, as well as the dispatch
function to dispatch actions as the user interacts with the UI.
Wrapping our app with the Provider
We need to create our React app in our src/react/components
folder, and wrap it in our new provider:
You can see all the state-aware components we are using here and I’ll be covering the Name
component below.
Accessing data and dispatching actions
Using our NameView
component that we created earlier, we’ll be re-using it to create our Name
component. It can access values from Context using the useContext
hook:
We can use the state
value to render our list’s name, and the dispatch
function to dispatch an action when the name is edited. And then our reducer will update the store. And it’s as simple as that!
The problem with React Context
Unfortunately, with this simplicity comes a catch. Using React Context will cause re-renders for any components that are using the useContext
hook. In our example, we’ll have a useContext
hook in both the Name
and Tasks
components. If we modify the list’s name, it causes the Tasks
component to re-render, and vice versa.
This won’t pose any performance issues for our small to-do list app, but lots of re-renders isn’t very good for performance as your app gets bigger. If you want the ease of use of React Context and useReducer without the re-render issues, there is a workaround library that you can use instead.
Replacing React Context with React Tracked
React Tracked is a super small (1.6kB) library that acts as a wrapper on top of React Context.
Your reducer and actions file can stay the same, but you’ll need to replace your store
file with this:
There are three hooks you can use to access your state and dispatch values:
And that’s the only difference! Now if you edit the name of your list, it won’t cause the tasks to re-render.
Conclusion
Using useReducer
in conjunction with React Context is a great way to quickly get started with managing your state. However re-rendering can become a problem when using Context. If you’re looking for a quick fix, React Tracked is a neat little library that you can use instead.
To check out any of the code that we’ve covered today, you can head to react-state-comparison to see the full examples. You can also take a sneak peek at the Redux example app we’ll be going through next week! If you have any questions, or a suggestion for a state management library that I should look into, please let me know.
Thanks for reading!