This is an imaginatively named, configurable web server using Hapi.js atop Node.js.
The aim is to provide a standardized node web server that can be used to serve your web application without the need for duplicating from another example, or starting from scratch.
The intention is that you will extend via configuration, such that this provides the baseline functionality of a Hapi web server, and within your own application you will add on the features, logic, etc unique to your situation.
This module requires Node v8.x.x+.
- Installing
- Usage
- Configuration
- Configuration Options
- electrode-confippet
- Adding a Hapi plugin
- API
- Contributions
- License
npm i --save electrode-server
Electrode Server comes with enough defaults such that you can spin up a Hapi server at http://localhost:3000
with one call:
require("electrode-server")();
Of course that doesn't do much but getting a 404
response from http://localhost:3000
.
To handle your routes, you should create a Hapi plugin to install your handlers.
See below for configuration options on how to register your plugin through electrode-server.
You can pass in a config object that controls every aspect of the Hapi server.
For example, if you want to spin up a server with HTTP compression off at port 9000:
const config = {
connection: {
port: 9000,
compression: false
}
};
require("electrode-server")(config);
However, for a more complex application, it's recommended that you use a config composer such as electrode-confippet to manage your app configuration.
Here's what you can configure:
All properties are optional (if not present, the default values shown below will be used).
server.app.config
is set to a object that's the combination of your config with electrode-server's
defaults applied.
- Server options to pass to Hapi's
Hapi.Server
Default:
{
server: {
app: {
electrode: true;
}
}
}
- Connection to setup for the Hapi server. Contains connection details for the server.
- If you want multiple connections, you can start multiple instances of
electrode-server
Default:
{
connection: {
host: process.env.HOST,
address: process.env.HOST_IP || "0.0.0.0",
port: parseInt(process.env.PORT, 10) || 3000,
routes: {
cors: true
}
}
}
connections
Object in previous Electrode no longer supports multiple connections.
Only the default
is allowed.
{
connections: {
default: {
host: process.env.HOST,
address: process.env.HOST_IP || "0.0.0.0",
port: parseInt(process.env.PORT, 10) || 3000,
routes: {
cors: true
}
}
}
}
- plugin registration objects, converted to an array of its values and passed to Hapi's
server.register
Default is just empty object:
{
plugins: {
}
}
-
A function to install event listeners for the electrode server startup lifecycle.
-
The following events are supported:
config-composed
- All configurations have been composed into a single oneserver-created
- Hapi server createdplugins-sorted
- Plugins processed and sorted by priorityplugins-registered
- Plugins registered with Hapiserver-started
- Server startedcomplete
- Final step before returning
To receive events you must set config.listener
before calling electrodeServer
.
For example:
myConfig.listener = (emitter) => {
emitter.on("server-created", (data, next) => {
// do something
next();
});
});
-
The data object will contain these:
emitter
,server
,config
, andplugins
. -
Depending on the stage some may not be present. For example,
server
is not available untilserver-created
event andplugins
is not available untilplugins-sorted
event. -
These are async events so you have to take and call a
next
callback.
You can control how much output the Electrode Server logs to the console by setting the logLevel
.
- Levels are
"info"
,"warn"
,"error"
,"none"
. - A level of
"warn"
means only warnning and error messages will be printed. - Default is
"info"
For example, to suppress the banner that is shown when the server starts up:
Hapi.js server running at http://mypc:4000
set the logLevel to "warn" or "error":
{
electrode: {
logLevel: "warn";
}
}
To keep your environment specific configurations manageable, you can use electrode-confippet.
Once you have your config files setup according to the configuration files setup, you can simply pass the config object to electrode server.
const config = require("electrode-confippet").config;
require("electrode-server")(config);
You can have electrode-server
register any Hapi plugin that you want
through your configuration file.
{
plugins: {
"<plugin-id>": {
enable: true,
options: {},
priority: 210,
register: function () {}, // mutual exclusive with module
module: "<plugin-module-name>",
requireFromPath: process.cwd()
}
}
}
<plugin-id>
- ID for the plugin. Generally the module name for the plugin, which is used to load it for registration.register
- optional The register function to pass to Hapi. Overridesmodule
.module
- optional name of the module to load for the plugin instead of the<plugin-id>
requireFromPath
- optional The path from which to callrequire
to load the plugin moduleenable
- optional if set tofalse
then this plugin won't be registered. If it's not set then it's considered to betrue
.options
- optional Object that's passed to the plugin's register function.priority
- optional integer value to indicate the plugin's registration order- Lower value ones are register first
- Default to
Infinity
if this field is missing or has no valid integer value (NaN
) (string of number accepted)
Priority allows you to arrange plugins to be registered in an order you prefer. The plugins with lower priority values are registered first.
If you don't want to use <plugin-id>
to load the module, then you can optionally specify one of the following:
register
- if specified, then treat as the plugin'sregister
function to pass to Hapi, overides modulemodule
- Only used ifregister
is not specified- If it's a string the used as the name module to
require
for registration. - It it's
false
then electrode server will not load any module. - You can specify a require-from-path for the module using an object.
- If it's a string the used as the name module to
{
plugins: {
myPlugin: {
module: {
requireFromPath: process.cwd(),
name: "my-plugin-module"
}
}
}
}
Electrode server will try to find your Hapi Plugin from your module by looking through these fields:
mod.hapiPlugin
mod.default.hapiPlugin
mod.default
mod
itself
Examples:
- Exporting the plugin directly as the module:
CommonJS example:
module.exports = myHapiPlugin;
ES6 example:
export default myHapiPlugin;
- Exporting the plugin as a field named
hapiPlugin
:
CommonJS example:
module.exports.hapiPlugin = myHapiPlugin;
ES6 example:
const hapiPlugin = myHapiPlugin;
export hapiPlugin;
ES6 default:
export default {
hapiPlugin: myHapiPlugin
};
There are three places you can specify a path to call require
from when loading your plugin modules.
config.plugins.requireFromPath
- The top one used for all pluginsconfig.plugins.<plugin-id>.requireFromPath
- Used for the specific plugin of<plugin-id>
, overrides the one aboveconfig.plugins.<plugin-id>.module.requireFromPath
- Used for the specific plugin of<plugin-id>
, overrides the two above
For more information: check out require-from-path
Here's an example using the crumb
plugin:
First, install the plugin as you normally would from npm
:
npm i --save crumb
Then, add your plugin to the config plugins
section.
{
plugins: {
"crumb": {
enable: true,
options: {},
priority: 210,
requireFromPath: process.cwd()
}
}
}
Above config tells electrode-server
to require
from CWD
the module by its <plugin-id>
"crumb"
and register it as a plugin with Hapi.
The electrode server exports a single API.
electrodeServer(config, [decors], [callback])
-
config
is the electrode server config -
decors
- Optional extraconfig
or array ofconfig
. In case you have common config you want to put inside a dedicated module, you can pass them in here.- If it's an array like
[ decor1, decor2, decor3 ]
then each one is composed into the main config. ie: something similar to_.merge(mainConfig, decor1, decor2, decor3)
.
- If it's an array like
-
callback
is an optional errback with the signaturefunction (err, server)
- where
server
is the Hapi server
- where
-
Returns: a promise resolving to the Hapi server if callback is not provided
Make sure you sign the CLA. Checkout the contribution guide
To run tests
% npm i
% clap test
To run tests and coverage
% clap check
To run sample server
% npm run sample
Hit http://localhost:9000
Copyright 2016-present WalmartLabs
Licensed under the Apache License, Version 2.0.