Pagination

The last topic that we’ll cover in this tutorial is pagination. You’ll implement a simple pagination approach so that users are able to view the links in smaller chunks rather than having an extremely long list of Link elements.

Preparing the React Components

Once more, you first need to prepare the React components for this new functionality. In fact, we’ll slightly adjust the current routing setup. Here’s the idea: The LinkList component will be used for two different use cases (and routes). The first one is to display the 10 top voted links. Its second use case is to display new links in a list separated into multiple pages that the user can navigate through.

Make sure to import the Redirect component, so you don’t get any errors.

You now added two new routes: /top and /new/:page. The latter reads the value for page from the url so that this information is available inside the component that’s rendered, here that’s LinkList.

The root route / now redirects to the first page of the route where new posts are displayed.

Before moving on, quickly add a new navigation item to the Header component that brings the user to the /top route.

You also need to add quite some logic to the LinkList component to account for the two different responsibilities it now has.

The query now accepts arguments that we’ll use to implement pagination and ordering. skip defines the offset where the query will start. If you passed a value of e.g. 10 for this argument, it means that the first 10 items of the list will not be included in the response. first then defines the limit, or how many elements, you want to load from that list. Say, you’re passing the 10 for skip and 5 for first, you’ll receive items 10 to 15 from the list.

But how can we pass the variables when using the graphql container which is fetching the data under the hood? You need to provide the arguments right where you’re wrapping your component with the query.

You’re now passing a function to graphql that takes in the props of the component (ownProps) before the query is executed. This allows you to retrieve the information about the current page from the router (ownProps.match.params.page) and use it to calculate the chunk of links that you retrieve with first and skip.

Also note that you’re including the ordering attribute createdAt_DESC for the new page to make sure the newest links are displayed first. The ordering for the /top route will be calculated manually based on the number of votes for each link.

You also need to define the LINKS_PER_PAGE constant and then import it into the LinkList component.

Implementing Navigation

Next, you need functionality for the user to switch between the pages. First add two button elements to the bottom of the LinkList component that can be used to navigate back and forth.

Since the setup is slightly more complicated now, you are going to calculate the list of links to be rendered in a separate method.

For the newPage, you’ll simply return all the links returned by the query. That’s logical since here you don’t have to make any manual modifications to the list that is to be rendered. If the user loaded the component from the /top route, you’ll sort the list according to the number of votes and return the top 10 links.

Next, you’ll implement the functionality for the Previous- and Next-buttons.

The implementation of these is very simple. You’re retrieving the current page from the url and implement a sanity check to make sure that it makes sense to paginate back or forth. Then you simply calculate the next page and tell the router where to navigate next. The router will then reload the component with a new page in the url that will be used to calculate the right chunk of links to load. Run the app by typing yarn start in a Terminal and use the new buttons to paginate through your list of links!

Final Adjustments

Through the changes that we made to the FEED_QUERY, you’ll notice that the update functions of your mutations don’t work any more. That’s because readQuery now also expects to get passed the same variables that we defined before.

Note: readQuery essentially works in the same way as the query method on the ApolloClient that you used to implement the search. However, instead of making a call to the server, it will simply resolve the query against the local store! If a query was fetched from the server with variables, readQuery also needs to know the variables to make sure it can deliver the right information from the cache.

All that's happening here is the computation of the variables depending on whether the user currently is on the `/top` or `/new` route.

Finally, you also need to adjust the implementation of update when new links are created.

You have now added a simple pagination system to the app, allowing users to load links in small chunks instead of loading them all up front.

Unlock the next chapter
What's the difference between the 'query' and 'readQuery' methods on the 'ApolloClient'?
'readQuery' always fetches data over the network while 'query' can retrieve data either from the cache or remotely
'readQuery' can only be used to reading data while 'query' can also be used to write data
'readQuery' was formerly called 'query' and the functionality of both is identical
'readQuery' always reads data from the local cache while 'query' might retrieve data either from the cache or remotely