ahm-config

Config management module. It uses nconf under the bonnet to manage config files, environment variables as a single system. It aims to simplify system configuration and config management. It also offers means of config validation and normalization using json schema.

Installation

$ npm i @ahmdigital/config

Usage

const config = require('@ahmdigital/config');

module.exports = config.make({ path: `${__dirname}/..` });

Options

There're several options that config.make takes:

• default string - Path to default config file (default dir/default.config.json).
• local string - Path to local config file (default dir/local.config.json).
• normalise boolean - Attempt to normalise string literals that come from env (default true).
• onError function - Error callback function, e.g. when config doesn't match schema (default console.error).
• path string - Directory to look for json configs.
• schema string - Path to json schema file (default dir/schema.json).

See Example.

Load order

Configuration is loaded with the following priority:

1. Environment variables (process.env) with __ as the separator.
2. Local config (usually not checked in to version control).
3. Environment specific config <env>.config.json.
4. Default config.
5. System defaults.

Environment configuration control

• NODE_ENV current environment name, e.g. development or production
• NODE_CONFIG_DIR directory where to look for config files, project's base directory by default
• NODE_CONFIG_FILE custom config file with overrides
• foo__bar__baz to override foo: { bar: { baz: value } }

Considerations

• Use hierarchical configs where possible
• Use json types where it makes sense, avoid type casting later on in your consumer code
• Keep devops specific config values (like elasticache hostname) in infrastructure config
• Keep sensitive values in infrastructure config or/and consider config encryption - nconf can do it