⚠️ This lesson is retired and might contain outdated information.

Handle Error Conditions in a Reducer using redux-actions

Andy Van Slaars
InstructorAndy Van Slaars
Share this video with your friends

Social Share Links

Send Tweet

In this lesson, we’ll see how we can handle success and error conditions independently using next and throw properties in a reducerMap with handleActions. We’ll add calls to catch for our fetch calls and display error messages when we encounter an error.

Instructor: [00:01] Currently, the reducer functions that are in our reducer map only handle the happy path. Nothing in here accounts for an error condition.

[00:08] Let's simulate an error and see what happens with the load to do's reducer when our service throws an exception. In that to do services JS file, I'm going to come into this get to do's method where we're calling fetch and returning the JSON for the result.

[00:22] I'm going to attack on an extra then. Inside that function, I'm just going to throw a new exception. I'll throw a new error and will give it the message. Since I only load the page, our call to get to do's should throw an exception. Let's save this and see what happens in the browser.

[00:46] We'll see that our loading indicator ran for a little bit longer than expected, and now we get this unhandled rejection with our boom error. Now, let's go back to our register file and handle this exception. In reducer JS, I'm going to find the fetch to do function, and this is where a calling gets to do's which is throwing exception. You see here we have a call that gets to do's with it then, but no catch.

[01:11] Let's fix that. We'll drop on here and will throw a catch on here, and we're going to get an error. We need to do something with that. Since we're controlling our state with Redux, let's dispatch an action. At this point, you might consider creating a new action creator and adding another step to our reducer that basically allows us to dispatch an error action.

[01:35] Instead, what we're going to do is we're going to dispatch load to do's just like we did before, but this time we're going to pass our error through. We're also going to dispatch hide loader, so that it hides that better.

[01:54] Now, let's go back to our reducer function and see how we can make this work. Down here in load to do's, you'll see that we're expecting our payload to be an array of to do's. That works in our happy path, but with the way we just dispatched this action, now we're sending an error in.

[02:12] Obviously, the error isn't going to work in place of our array of to do objects. The flux standard action format that we're using for these action objects has an error property that is a Boolean flag. If our payload object is an error, that error flag is going to be set to true by our action creator, and because we're using Redux actions for our action creators, that's handled automatically for us and we don't have to do anything to create that.

[02:38] What we do need to do is differentiate how we want to handle our reduction based on whether this is an error object or not. Handle actions gives us the capability right through a reducer map. What we're going to do is instead of just using reducer function as our value for this load to do's key, we're going to pass it an object.

[02:58] That object is going to have two keys of its own, and next key, and next is going to be our reducer function. I'll just cut that and paste that right there.

[03:11] Our second property here is going to be throw. Throw is also reducer function, but this is the one that we called if our payload is an error and the error key on that action is set to true. If we throw the final reducer function the same signature is going to take state and our action. We're going to return a new state.

[03:36] Just like before, I spread out the existing state. Now, what I want to do is I want to set the message property, and this is going to display a message at the top of our page.

[03:48] Because, I know I'm inside this throw reducer. I know that the payload is going to be an error object. I'm going to use the action.payload which is my error object, and I'm going to use that error object's message property to define what that message text should be.

[04:07] I'll save this. Go back to the browser and this time when it reloads, we'll get our error, but our indicator is hidden and our message is shown with the text that we gave in that error.

[04:22] Of course, I might not want to use the error message as my display message, so I can just come in here and I can change this. This error handling is still tied to my original action, I can make this handling very specific instead of having a new action that's just for showing error messages that I'd then have to pass the message into.

[04:48] I can do this all in place and I can handle errors for different action types completely independently. I'll save this. Reload the browser one more time, and we'll see that we get our new custom error message.

[05:01] I'm going to go back into to do services. I'm going to take this, then that there is an error out. Just make sure my application still works as expected. When we don't get an error, you'll see that it's still loading the to do's and it's not showing our message.