Skip to content
A handy set of tools for creating a Cloudflare Worker app.
Branch: master
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
template pre-commit eslint and prettier Mar 21, 2019
.gitignore init Mar 5, 2019
CODE_OF_CONDUCT.md add CoC and contribution guide Mar 18, 2019
CONTRIBUTING.md add CoC and contribution guide Mar 18, 2019
LICENSE-APACHE license Mar 15, 2019
LICENSE-MIT license Mar 15, 2019
README.md docs: Add links to README May 19, 2019
index.js pre-commit eslint and prettier Mar 21, 2019
package-lock.json 1.1.0 Mar 21, 2019
package.json 1.1.0 Mar 21, 2019
readme-template.txt add readme to generated project Mar 14, 2019

README.md

Cloudflare Worker App Kit

Postlight's Cloudflare Worker App Kit is a handy set of tools for creating, developing, testing, and deploying a Cloudflare Worker app. No configuration, just build it and ship it. Read all about it in this handy introduction.

Demo

A real app built with these tools — secretmsg.app (repo).

Get started

npx @postlight/cloudflare-worker-app-kit create <AppName>

This create command scaffolds your project with the following:

How the app works

Like any server-side rendered app this runs in two environments — client and worker. The client entry point is typical of most single-page, JavaScript applications. The root component is rendered into a container on the page, and in this case, nothing happens since the page was delivered fully-rendered.

The worker entry point is where things are a little different. Instead of sending the request all the way back to a Node.js server running in a central location, a quick-to-start Worker sitting in front of Cloudflare's cache handles the request. If it's a request for a static asset, like an image or JavaScript file, we fetch it from the cache or S3 like we normally would. If it's for a page, we render the app, delivering all the html. With all this happening at the edge, your user gets great response times no matter where they're located.

That's it, no origin server to worry about. Just some files on S3 and a little JS distributed around the world. The result is a fast, server-rendered web app that can handle 10 million requests per month for $5 — not bad!

Project structure

# Static assets will be copied to S3 and served by the worker
assets/
 └─ images/

# Scripts for build, deploy, and the dev server
scripts/
 ├─ build.js
 ├─ deploy.js
 └─ start.js

# App files
src/
 ├─ components/
 ├─ lib/
 ├─ client.ts
 └─ worker.ts

After you build the app there will also be dist directory where you'll find JS and CSS bundles along with their associated source maps. During deploys those files get copied to S3 along with your other assets.

Environment variables

These environment variables are required to deploy the app. If you're not sure where to find these values, there's more info below.

BUCKET=bucket-name
AWS_KEY=XXXACCESSKEYXXX
AWS_SECRET=XXXXXXXXXSECRETXXXXXXXXX
AWS_REGION=us-east-1
CF_ZONE_ID=XXXXXXXXXWORKERZONEIDXXXXXXXXX
CF_KEY=XXXXCLOUDFLAREAUTHKEYXXXX
CF_EMAIL=account@email.com

If you want to use Workers KV you'll need the CF_KV_NAMESPACES environment variable during development and when you deploy.

# single KV namespace
CF_KV_NAMESPACES="NAME XXXXXXXXXNAMESPACEIDXXXXXXXXX"

# multiple namespace are supported, separate with a comma
CF_KV_NAMESPACES="NS_1 XXXXXXXNAMESPACEIDXXXXXXX, NS_2 XXXXXXXNAMESPACEIDXXXXXXX"

Similarly, you can bind other strings with CF_WORKER_BINDINGS

CF_WORKER_BINDINGS="KEY_1 value1, KEY_2 value2"

Dev tools

The local dev environment consists of two servers. One delivers static assets using webpack middleware, basically standing in for S3. The other wraps Cloudworker — a mock implementation of Cloudflare Workers. It loads the worker.js bundle and handles requests just like a worker would.

# Start a dev server at http://localhost:3333
npm start

Colocate test files with your source files with a .test.ts extension, and Jest will do the rest.

# Run jest tests
npm test

You know the drill, ESlint is here to keep you honest.

# Check source files for common errors
npm run lint

Build and deploy

Using webpack, this script creates production-ready bundles for the client, worker and any css you imported into you components. If you're going to use the deploy script below, don't worry about this. It will run its own fresh build anyway.

# Output production-ready JS & CSS bundles in dist folder
npm run build

First, dependencies are installed and a fresh build is made. Then all the static files are copied to your S3 bucket Finally, the new worker script, along with some metadata, is deployed to Cloudflare.

# Build files, copy static assets to S3, and deploy worker to Cloudflare
npm run deploy

Setup Cloudflare and S3

Two things need to happen before you can deploy. Get Cloudflare in front of your domain and setup an S3 bucket to serve static files to the public.

Cloudflare

  1. If you don't have a domain setup with Cloudflare, step 2 and step 3 of their Getting Started guide is a good place to start.
  2. Visit your Cloudflare dashboard to turn on workers.
  3. In the Workers tab, launch the workers editor then go to the routes section on the left. These routes filter what requests will be handled by the worker, something like example.net/* is a good place to start. All requests to your domain will go through the worker. If you're familiar with curl, you can do all of this from the command-line.
  4. If you want to play with the worker editor, it's a nice way to see some immediate results. There are also some handy debugging tools that might be helpful later on.
  5. Before finishing, now's a good time to gather some environment variable values — CF_ZONE_ID and CF_KEY. Zone ID is on the dashboard overview page in the right column. For the API key, visit your user profile page, scroll to the bottom and copy the Global API Key.

AWS S3

One way to handle this is to treat your S3 bucket like a static web server. You give it a name that matches your domain, like example.net and change your Cloudflare DNS settings to point at that bucket. If you haven't done this before, here's a good step-by-step explanation.

You can also name the bucket anything you want, leaving the DNS settings alone. You would then handle all the routing in the worker, like this.

While setting up your bucket, you should also gather some environment variable values for BUCKET, AWS_KEY, AWS_SECRET, and AWS_REGION. For the AWS credentials you should setup a user with read and write access to the new bucket.


🔬 A Labs project from your friends at Postlight. Happy coding!

You can’t perform that action at this time.