Fetch Requests in ClojureScript

Asynchronous Channels

Making a GET request for data in Clojure is pretty simple with the clj-http library. Simply call http/get and pass in the URL. The value returned from the function is the response of the get request, and it can be easily saved in a global def or a let binding to be used later in the code.

Get requests in ClojureScript are a little trickier. Since ClojureScript compiles to JavaScript, we are dealing with JavaScript's async/await structure. Get requests using cljs-http have a little more going on:

(defn get-data []
  (go (let [response (<! (http/get "https://api.github.com/users"))]
           (prn (:status response)))))

Notice that everything wrapped in a go block. This is because we don't want the asynchronous act of fetching data to completely block the once thread that JavaScript has. Since we can't delegate to another thread, wrapping something in a go block makes sure that the operation inside is paused if necessary rather than making everything wait for the async call to complete. This is similar to using the async keyword in JavaScript.

Then we have the <! function, which is the asynchronous "take" function. This is how we actually read data from the asynchronous channel and only works when used inside a go block.

Using the Response Data

Great! So now we're getting data back from an API, and we should just be able to return that response variable in the let to use the data, right?

Unfortunately not. Trying to return the data and print it resulted in #object[cljs.core.async.impl.channels.ManyToManyChannel], a string representation of the channel object, but not the actual data. The data can only be read inside the go block. To make the data accessible elsewhere, I created an atom outside the get-data function, and then I stored the response data to the atom inside the go block using reset!.

Further Reading

For a closer look at the core.async library, see this article.

For more information on the cljs-http library, see the GitHub repo.