Skip to content

A lightweight HTTP framework for Typescript / JS with zero dependencies

License

Notifications You must be signed in to change notification settings

TomShacham/http4js

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

http4js

A lightweight HTTP framework for Typescript / JS, with zero dependencies

Use http4js in your project

npm install --save http4js

Contents

Example

An example server and client

//define our routes
const routing = routes('GET', ".*", async (req: Req) => {
    console.log(req);
    return ResOf(Status.OK, 'OK');
})

//add csrf token header to every request and vary gzip to every response
const headerFilter = (handler: HttpHandler) => {
    return async (req: Req) => {
        const response = await handler(req.withHeader(Headers.X_CSRF_TOKEN, Math.random()))
        return response.withHeader(Headers.VARY, "gzip");
    }
};

// start routing as a NativeHttpServer on port 3000
routing
    .withFilter(headerFilter)
    .asServer(HttpServer(3000))
    .start();

// make an http request to our server and log the response
HttpClient(ReqOf(Method.GET, "http://localhost:3000/any/path"))

Latest features

Full Release Notes here

5.0.0 move to HttpHandler interface

4.2.6 fix HttpsClient: post body

4.2.4 res.fullBodyString() for bodies > 65kb

4.2.3 Adds gzip filter. Eg. Filters.GZIP(HttpClient) or .withFilter(Filters.GZIP)

4.2.2 throw if no server in Routing when calling start, stop, serveE2E.

4.2.1 Query params are now passed to serveE2E

4.2.0: Breaking change: Most precise handler no longer beats first declared match. Fix: Composed routes filter as expected.

To find a matching handler for a Req, we recurse "left to right and deepest first" through nested routes, ie. routes attached to top level routes using withRoutes(routes), ending finally with the top level routes e.g.

get('/', async()=> ResOf())
    .withRoutes(
        routes.withRoutes(furtherNestedRoutes)
    )

furtherNestedRoutes is traversed followed by routes then finally the top level routes. Further docs here

4.1.3: Breaking change: Res Convenience methods for responding

Redirect is now a static method Res.Redirect as we provide a number of convenience methods eg. Res.OK() and Res.GatewayTimeout.

4.1.2: Convenience methods for starting server

We provide HttpServer(3000) and HttpsServer(3000, certs) as quick easy ways to provide a server.

4.1.1: Fix: HttpClient was not streaming out

See streaming docs for more info

4.1.0: streaming by default

NativeHttpServer and HttpClient stream in and out by default. A handle on the stream is provided by req.bodyStream() and a res is streamed out if a Res(200, readable) is provided, i.e. a Readable stream body.

4.0.0: ! Breaking change: drop support for Koa and Express backends

In order to evolve the core library faster support for Express and Koa backends has been dropped. Happy to add back later.

Contributing

I'd be very happy if you'd like to contribute :)

To run:

git clone [email protected]:TomShacham/http4js.git && \ 
cd http4js && \
npm i --save && \
./create-ssl-certs.sh && \
npm test

History and Design

http4js is a port of http4k.

Early ideas and influence from Daniel Bodart's Utterly Idle

To dos

  • example app
    • withOptions on withPost
  • generalise routing to an interface
  • client side httpclient (from stu)

Running HTTPS Server tests

We need our own certs to run an HTTPS server locally.

These Commands get you most of the way, I altered them slightly for this script, that may work for you

./create-ssl-certs.sh

If not, follow these Instructions to create your own certificates in order to run an HTTPS server locally.

Then run

npm test

Sanity Testing Streaming

Create a big file

cat /dev/urandom | base64 >> bigfile.txt
# wait ...
# ^C

Start up a server and stream the file

get('/bigfile', async() => ResOf(200, fs.createReadStream('./bigfile.txt')))
    .asServer()
    .start();

Check the memory of usage of the process.

  • If we are not streaming, then the whole file will be read into memory before the server responds, using lots of memory.
  • If we are streaming then the memory usage should be much lower.