lamcfg

Bereft and easy configuration management for NodeJS projects on AWS Lambda via Serverless.

Summary

The idea is to provide a quick wrapper around configurations to make them easily consumed in AWS Lambda functions, while also providing an easy way to make code testable. If the configuration is provided as an argument to the class that needs it then values can be easily fed into your unit tests via configuration defaults.

The secondary objective is to keep the code footprint small and have no reliance on outside libraries, since these quickly blow up the code footprint.

config.child('name')

This provides you the ability to create a configuration object from an object within your configuration source, that can be passed around and used with its children continuing to be overridable by environment variables.

const Config = require('lamcfg');
const config = new Config({defaults:{parent:{child:{with:'some',values:{value1:'DefaultValue'}}}}});
const childConfig = config.child('parent.child');

function MyObject(childConfig){
  childConfig.get('with'); //some
  childConfig.get('values.value'); //DefaultValue
}

import * as Config from 'lamcfg';
const config = new Config({defaults:{parent:{child:{with:'some',values:{value1:'DefaultValue'}}}}});
const childConfig = config.child('parent.child');

function MyObject(childConfig){
  childConfig.get('with'); //some
  childConfig.get('values.value'); //DefaultValue
}

config.get('name','inlineDefault')

There are 3 points you can provide a configuration, inline as a second argument to getting the value, as 'defaults' to the configuration instance and as process.env values. The value is selected based on the first value found.

1. process.env, if there is a value here it is used.
2. configuration defaults, if there is a value found here it is used.
3. inline defaults, this is the last place looked and the value is used if it is found.

NOTE: You do not have to use the full dot notation to a string/bool/number value. If you want to return a full object, you can do that. It will be appropriately overriden with all environment variable overrides.

process.env

You can specify an envPrefix value if you want to keep your configurations 'separate' from other process.env values. Its a primitive method to help avoid name collissions but should work in most scenarios.

Serverless

Within serverless, to create process.env values all you need to do is add environment variables to your configuration. This can make for a pretty convenient usage scenario. This is the intended use case for this module, but other simple scenarios will work just as well.

service: superspecial

provider:
  name: aws
  runtime: nodejs4.3
  region: us-west-2
  environment:
    variable_name: ABC777
    name: XYZ123

const Config = require('lamcfg');
const config = new Config({envPrefix:'MYCFG_',defaults:{variable:{name:'x'},name:'y'}});

function MyObject(config){
  config.get('variable.name'); //ABC777
  config.get('name'); //XYZ123
}

Config Defaults with process.env

This assumes you have a central file or configuration source, as is usually needed in larger applications where the value is used in multiple places. This module does not provide you a method for storing and retreiving the configuration. If you would like this feature please request it at https://github.com/matt-filion/lamcfg/issues.

process.env['MYCFG_dot_notation_to_value'] = "Override Inline Default";

const Config = require('lamcfg');
const config = new Config({envPrefix:'MYCFG_',defaults:{dot:{notation:{to:{value:'DefaultValue'}}}}});

function MyObject(config){
  /*
   * Assuming the value is not overriden by a process.env value, then the value specified
   *  in defaults will be used.
   */
  config.get("dot.notation.to.value");
}

Default Overrides, config.update({override:'values'})

Its possible to have a second source of configuration. There is a convenience method for updating an existing default configuration, with new values, using config.update({}).

const Config = require('lamcfg');

function MyObject(additionalOrOverwritingConfigs){
  const config = new Config({envPrefix:'MYCFG_',defaults:{dot:{notation:{to:{value:'DefaultValue'}}}}});
  config.update(additionalOrOverwritingConfigs);
  /*
   * The default values stored will be overwritten with the values provided in additionalOrOverwrittenConfigs, passed in.
   */
  config.get("dot.notation.to.value");
}

Inline and process.env

This strategy aimes at providing you two points of configuration, and is appropriate for small and quick applications. Inline at the point of needing the configuration and then overriding using process.env variables. On AWS Lambda this will be in the form of environment variables, see http://docs.aws.amazon.com/lambda/latest/dg/env_variables.html.

process.env['dot_notation_to_value'] = "Override Inline Default";

const Config = require('lamcfg');
const config = new Config();

function MyObject(config){
  /*
   * Assumin the value is not overriden by a process.env value, then the inline value
   *  will be used.
   */
  config.get("dot.notation.to.value","inlineDefault");
}