Ignition

Basic configuration and tooling shared across applications

Install

npm install ghost-ignition --save

or

yarn add ghost-ignition

Usage

Ignition offers the following features:

• Logging
• Errors
• Config using nconf
• HTTP Server
• Debug

Logging

Configuration

Example:

const ignition = require('ghost-ignition');

const logging = ignition.logging({
    domain: 'example.com',
    env: 'production',
    mode: 'long',
    level: 'info',
    transports: ['file'],
    rotation: {enabled: true, period: '1d', count: 10},
    path: '/var/log'
});

Examples

logging.info({req: req, res: res});
logging.info({req: req, res: res, err: err});
logging.info('Info');
logging.error(new Error());
logging.warn('this', 'is', 'a', 'warning');
logging.debug('this is a debug mode');
logging.warn(err, 'Caught an error from service X.');
logging.warn('A friendly message.', err);
logging.warn('A friendly message.', {err: err});

Transports

File

Ignition creates two log files by default:

• An errors log file, which only contains logs from logging.error
• A general log file, which contains all logs from logging.info, logging.warn and logging.error

If you would like to open a log file on disk, we highly recommend to install bunyan with NPM (npm i -g bunyan). You can then open your log file with bunyan your.log in the shell, which makes it possible to read the content.

Loggly

The loggly transport makes it possible to send your logs to loggly. The stream will only send errors to loggly at the moment.

Example:

const ignition = require('ghost-ignition');

const logging = ignition.logging({
    transports: ['file', 'loggly'],
    loggly: {
      token: 'token',
      subdomain: 'subdomain',
      // The "match" property is helpful if you only want to send specific errors to loggly. It's a regex string.
      match: 'level:critical' // or 'statusCode:500|statusCode:403'
    },
    ...
});

GELF

The transport makes it possible to send logs to the GELF UDP collector.

Example:

const ignition = require('ghost-ignition');

const logging = ignition.logging({
    transports: ['gelf'],
    gelf: {
      host: 'gelf.example.com', // Default: '127.0.0.1'
      post: 12345               // Default: 12201
    },
    ...
});

Shell

ENV Variables

Ignition accepts some env variables to modify the log output.

LEVEL=error - Only print errors. MODE=long - Show full & long log output. LOIN=true - Set's the level to "info" and the mode to "long".

Errors

Ignition errors contains a set of useful & common error classes. Each Ignition error inherits from Node's native error and keeps the structure!

Extra properties

On top of the native error properties (message, code, stack), Ignition errors support the following properties:

List of errors

Examples

new logging.errors.InternalServerError({
    message: 'Something went very wrong',
    context: {
        user: 1
    }
})

// Ignition supports nested errors. It will try to inherit properties and extend the stack trace.
// This is super useful if you receive an error from a calling unit, but you would like to wrap it into a custom error.
new logging.errors.InternalServerError({
    err: err
})

Error utils

const ignition = require('ghost-ignition');

// you can pass any error and ignition will tell you if this is a custom ignition error
ignition.errors.utils.isIgnitionError(err);

// serialize an error to a specific format
ignition.errors.utils.serialize(err, {format: 'jsonapi|oauth'});

// deserialize specific format to error instance
ignition.errors.utils.deserialize(err);

Config

Ignition config uses nconf to create a configuration object based on your environment.

Requirements

• Create config files based on your available environments.
• Instantiate Ignition config.
• Read nconf documentation to understand how to use the config object.

Examples

config.example.json (defaults)

{
  "port": 9999
}

config.production.json

{
  "host": "blog.com"
}

config.development.json

{
  "host": "localhost"
}

// As soon as you call the config object, Ignition will read your config files from disk and returns a config object.
// The config object is then cached. You can operate on the config object using `set` and `get` (see [nconf](https://github.com/indexzero/nconf#readme))
const config = require('ghost-ignition').config();

// -> {port: 9999, host: localhost}

Debug

Ignition debug offers an easy way to add debugging to your application. It wraps the debug NPM module to simplify how to add debug information to your files. Ignition debug will try to read your package.json to get the name/alias of your application. You can enable the debug log by passing the "DEBUG" environment variable.

Requirements

• Read debug documentation.

Examples

package.json

  "name": "myproject",
  "alias": "proj" 

const debug = require('ghost-ignition').debug('api-controller');

debug('Calling the model layer.');

// DEBUG=proj:api-controller yarn start

Server

The HTTP server bundles common logic in one place.

• error handling for the HTTP server
• port normalisation

Requirements

• Express
• Ignition config

Examples

const ignition = require('ghost-ignition');
ignition.server.start(app);
ignition.server.stop(app);

Test

• yarn lint run just eslint
• yarn test run eslint && then tests

Publish

• yarn ship

Copyright & License

Copyright (c) 2013-2022 Ghost Foundation - Released under the MIT license.