Svelte.js + Flask — Combining Svelte with a Backend Server

Alex Cabrera
3 min readSep 25, 2019

TL;DR — You can use Flask to add a simple backend server to your Svelte app. Check out the example code here.

I recently discovered Svelte.js, a web application framework similar to React.js or Vue.js. While it has the same powerful features, like components and props, as the more established frameworks, I find it much more intuitive and straight-forward. A super simple Svelte component looks like this:

Interacting with the component updates the state of a JS variable.

A Svelte component looks like a regular HTML file, but with the inclusion of a <script>tag and a couple special syntax commands it becomes a fully functional reactive framework. The functionality ends up being similar to React.js, allowing you to pass props, import NPM libraries and components, and reactively update the HTML body. There is really a ton Svelte.js can do, and I highly recommend learning more with their fantastic interactive tutorial.

Combining Svelte and Flask

Svelte.js is powerful by itself, but I recently ran into a problem where I needed some sort of backend or API— I had a dataset of about 200,000 images I wanted to be able to filter and query.

While I could have created a standalone server to act as an API, it becomes a bit of a hassle to deal with CORS and eventually hosting two different applications. I found it much simpler to combine the two and serve the Svelte app from Flask.

In a new folder, we can create a tiny Flask server,server.py, that returns a random number (you may have to run pip install flaskfirst):

To test it out you can run python server.py and open localhost:5000/rand in your browser. You should see a random number between 0 and 100 each time you refresh.

Adding Svelte

We can now add Svelte into our Flask application. Inside your folder, run the following commands to create a new Svelte app:

npx degit sveltejs/template client
cd client
npm install

This will create a new Svelte template and install the required dependencies. You can run npm run dev to check that the Svelte app works, and you should see a purple Hello World at localhost:5000 in your browser.

Now we want Flask to serve our Svelte project. We can do that by adding the following to our server:

This tells our Flask server to serve all the static files that Svelte compiles into the public folder. Now if you run python server.py you should see our Svelte app!

Sending Requests to Flask

Lastly, we can modify our Svelte file App.svelte to send a request and get a random number from the Flask server:

This component now queries the /rand path of our Flask server and shows the returned number.

Final Setup

This simple framework ends up being a simple and convenient way to quickly prototype ideas that need a server. In two terminal windows you can run the following:

  • python server.py to run the flask server
  • npm run autobuild in the Svelte app folder to automatically build when you save a file

You can find the full code for this simple app here.

--

--