Worker Middleware
A suite of standalone HTTP server middlewares for Worker Runtimes.
Work In Progress
It is meant to be used with Worker Router, but can also be used with simple request handlers.
What's Included?
Cookies
Supports singed, unsigned and encrypted cookies.
Signed and encrypted cookies use the Web Cryptography API internally to en/decrypt sign and verify cookies.
router.get('/', withSignedCookies({ secret: 'password123' }), (request, { cookies, cookieStore }) => {
cookieStore.set('foo', 'bar') // no await necessary
return ok(cookie.foo === 'bar' ? 'Welcome back!' : 'Welcome!')
})
The cookieStore
property implements the web's Cookie Store API for maximum standard compatibility.
For better DX, the middleware also provides a read-optimized cookies
property, which are the request's cookies parsed into a plain JS object.
Modifying cookies is done via the cookie store. While the cookie store API is async, there is no need to await every result, as the cookie store keeps track of all operations and awaits them internally before sending the headers.
Session
There are two session middlewares. The cookie session encodes the entire session object into a cookie and is meant for prototyping and small use cases. The storage session uses a KV Storage API-compatible storage object to persist the session object between requests. Worker Tools provides storage adapters for Cloudflare's KV storage and SQLite/Postgres for Deno.
The session is a plain JavaScript object that is serialized/deserialized via the Structured Clone Algorithm, i.e. it behaves largely the same as storing an object in IndexedDB. In other words, you can add Maps, Sets, and ArrayBuffers to it.
router.get('/', combine(
signedCookies({
secret: 'password123'
}),
cookieSession({
// Type is inferred from default session.
defaultSession: { id: '', iv: new Uint8Array([]) }
})
), (request, { session }) => {
if (!session.id) {
session.id = crypto.randomUUID();
session.iv = crypto.getRandomValues(new Uint8Array(32))
}
return ok()
})
The session object is persisted at the end of the request.
Body Parser
Because Worker Runtimes already provide helpers like .json()
and .formData()
in the Request type, the need for a body parser is less pronounced. The value of Middleware's body parser mainly comes from content negotiation:
router.any('/form', bodyParser(), (request, { accepted, ...ctx }) => {
switch (accepted) {
case 'application/x-www-form-urlencoded': {
ctx.form // instanceof URLSearchParams
ctx.formParams // { [name: string]: string }
return ok()
}
case 'multipart/form-data': {
ctx.formData // instanceof FormData
ctx.formParams // { [name: string]: string }
ctx.files // { [name: string]: File }
return ok()
}
case 'application/octet-stream':
case 'application/binary': { // commonly used non-standard mime type
ctx.blob // instanceof Blob
ctx.buffer // instanceof ArrayBuffer
return ok()
}
default: {
// application/json, text/html, text/plain
return ok()
}
}
return ok()
})
You can also limit what is acceptable to the endpoint by combining the content negotiation middleware (see below) and bodyParserStandalone
.
Content Negotiation
Provides generic content negotiation for HTTP endpoints.
Another use case is to combine with the body parser to limit what the body parser accepts:
router.any('/form', combine(
accepts({ accepts: ['application/x-www-form-urlencoded', 'multipart/form-data'] }),
bodyParserExCn()
), (request, { accepted, ...ctx }) => {
// Accepted property are now limited to forms.
// Endpoint will not accept other types
}
NOTE: It's currently only possible to limit what the body parser accepts.
The contentTypes
middleware lets you specify what content types the endpoint can provide. For example, we can build a mini deno.land that either serves raw JavaScript or a HTML page documenting the function:
router.get('/form', combine(
contentTypes({ types: ['text/html', 'application/javascript'] })
), (request, { type }) => {
// `type` is either 'text/html' or 'application/json',
// depending on the client's `Accept` header (best match)
switch (type) {
case 'application/javascript':
return ok('function foo() {}', {
headers: { 'content-type': type }
})
case 'text/html':
return ok('HTML documentation for JS', {
headers: { 'content-type': type }
})
}
}
Caching
Basics
Use with Netlify Functions
import { Context } from "netlify:edge"
import {
withMiddleware,
combine,
signedCookies,
cookieSession,
} from 'https://ghuc.cc/worker-tools/middleware/index.ts';
export default withMiddleware(
combine(
signedCookies({ secret: 'password123' }),
cookieSession(),
),
async (req, { cookies, body, args: [, _context] }) => {
const context = _context as Context
return await context.next()
}
)
This module is part of the Worker Tools collection
β
Worker Tools are a collection of TypeScript libraries for writing web servers in Worker Runtimes such as Cloudflare Workers, Deno Deploy and Service Workers in the browser.
If you liked this module, you might also like:
- π§ Worker Router --- Complete routing solution that works across CF Workers, Deno and Service Workers
- π Worker Middleware --- A suite of standalone HTTP server-side middleware with TypeScript support
- π Worker HTML --- HTML templating and streaming response library
- π¦ Storage Area --- Key-value store abstraction across Cloudflare KV, Deno and browsers.
- π Response Creators --- Factory functions for responses with pre-filled status and status text
- π Stream Response --- Use async generators to build streaming responses for SSE, etc...
- π₯ JSON Fetch --- Drop-in replacements for Fetch API classes with first class support for JSON.
- π¦ JSON Stream --- Streaming JSON parser/stingifier with first class support for web streams.
Worker Tools also includes a number of polyfills that help bridge the gap between Worker Runtimes:
- βοΈ HTML Rewriter --- Cloudflare's HTML Rewriter for use in Deno, browsers, etc...
- π Location Polyfill --- A
Location
polyfill for Cloudflare Workers. - π¦ Deno Fetch Event Adapter --- Dispatches global
fetch
events using Denoβs native HTTP server.
Fore more visit workers.tools.