How to Build a REST API in NodeJS

Building a REST API in NodeJS is simple with a framework specifically made for the task. Eran Hammer and the team at Walmart Labs created HapiJS to connect their mobile platforms with a legacy Java backend. Sounds like complex integrations coupled with with large scale maintainability requirements to me!

I use Hapi to build my platform APIs for that very reason. I’ll demonstrate how simple it is to construct a good foundaton for your REST API.

Today let’s focus on the first step of creating a server. We’ll go over the basics and touch on some handy features.

Installing Hapi

First make sure you have a recent version of NodeJS installed. To get started let’s create a new Node project with npm init`. Just hit enter to accept all the defaults if you want.

npm init
npm install --save hapi nodemon
touch server.js

The --save flag will write the module dependencies to your package.json file. For local development I use nodemon to run my API. It watches the project folder for changes and restarts automatically.

Creating a server

Open the server.js file and type in the following code:

 1 var port = process.env.PORT || 9000;
 2 
 3 var Hapi = require('hapi');
 4 
 5 var server = new Hapi.Server();
 6 
 7 server.connection({ port: port });
 8 
 9 server.start(function () {
10     console.log('Server started at port %d', port);
11     console.log('Resistance is futile');
12 });

No ceremony here! If you want you can provide a port number to start your API at. Otherwise it will default to port 9000.

PORT=9001 nodemon server.js

Application settings

Real world projects usually have some kind of configuration file with environment specific values.

module.exports = {
    "development": {
        "someProvider": "http://dev.some.provider.com"
    },
    "qa": {
        "someProvider": "http://qa.some.provider.com"
    }
};

Instead of having require('./config.js') wherever you need configuration values, you can DRY up your code by giving your configuration values to the server when you call the constructor function.

 1 var port = process.env.PORT || 9000;
 2 var env = process.env.NODE_ENV || 'development';
 3 
 4 var Hapi = require('hapi');
 5 
 6 var server = new Hapi.Server({ app: require('./config')[env] });
 7 
 8 server.connection({ port: port });
 9 
10 server.start(function () {
11     console.log('app setings', server.app);
12 });

Multiple servers

Sometimes it’s useful to partition your app into different servers. Maybe you want the admin part of your API to be completely separate from your core services. A lot of times I need to run a proxy to legacy APIs too.

 1 var port = process.env.PORT || 9000;
 2 var env = process.env.NODE_ENV || 'development';
 3 
 4 var Hapi = require('hapi');
 5 
 6 var server = new Hapi.Server({ app: require('./config')[env] });
 7 
 8 server.connection({ port: 9000, labels: ['core'] });
 9 
10 server.connection({ port: 9001, labels: ['admin'] });
11 
12 server.connection({ port: 9002, labels: ['proxy'] });
13 
14 server.start(function () {
15     console.log('core', server.select('core').info);
16     console.log('admin', server.select('admin').info);
17     console.log('proxy', server.select('proxy').info);
18 });

That’s a wrap on some Hapi server methods I find useful. Next time we’ll learn how to configure routes and add request handlers.