Econfig is a gem which allows you to easily configure your Ruby applications in a multitude of ways.

Add this to your Gemfile:

gem "econfig"

If you're using Rails, you'll want to require the Rails extension:

gem "econfig", require: "econfig/rails"

Extend your main application module with the Econfig shortcut.

In Rails, you'll want to add this in config/application.rb:

module MyApp
  extend Econfig::Shortcut

  class Application < Rails::Application
    # …
  end
end

In a modular Sinatra application, extend your controller class and copy its settings to Econfig in app.rb:

require 'sinatra'
require 'econfig'

class MyApp < Sinatra::Base
  extend Econfig::Shortcut

  Econfig.env = settings.environment.to_s
  Econfig.root = settings.root

  # …
end

In either case, you can now you can access configuration like this:

MyApp.config.aws_access_key_id

If the key you accessed is not configured, Econfig will raise an error. To access optional configuration, which can be nil, use brackets:

MyApp.config[:aws_access_key_id]

Sometimes you might want to bypass the strictness requirement in econfig, for example if you're running the application as part of a build process. In that case you can set the environment variable ECONFIG_PERMISSIVE, and econfig will not raise errors on missing keys, instead returning nil.

You can specify configuration through:

1. ENV variables
2. Redis
3. Relational database
4. YAML files
5. OSX Keychain

This allows you to set up Econfig on most kinds of hosting solutions (EngineYard, Heroku, etc) without any additional effort, and to switch between them easily.

Just set an environment variable whose name is the name of the option being accessed uppercased.

For example:

AWS_ACCESS_KEY_ID=xyz rails server

You can now read it like this:

MyApp.config.aws_access_key_id # => "xyz"

This is especially convenient for Heroku.

Add a yaml file to config/app.yml. This should have a similar layout to config/database.yml:

development:
  aws_access_key_id: "xyz"
  aws_secret_access_key: "xyz"
production:
  aws_access_key_id: "xyz"
  aws_secret_access_key: "xyz"

Econfig also reads configuration from config/secret.yml which is the new standard for secret configuration parameters in Rails 4.1.

This needs to be explicitly enabled. In config/application.rb add this code:

require "econfig/active_record"
Econfig.backends.insert_after :env, :db, Econfig::ActiveRecord.new

You probably want environment variables to take precendence over configuration from ActiveRecord, hence the insert_after. If you'd rather have ActiveRecord configuration take precendence you can use this instead:

require "econfig/active_record"
Econfig.backends.unshift :db, Econfig::ActiveRecord.new

You will also need to create a migration to create the necessary database tables:

rails generate econfig:migration
rake db:migrate

This needs to be explicitly enabled. In config/application.rb add this code:

require "econfig/redis"
redis = Redis.new(:host => "myredis.com")
Econfig.backends.insert_after :env, :redis, Econfig::Redis.new(redis)

If you wish to namespace your keys in Redis, you can use redis namespace.

For the OSX keychain backend, see econfig-keychain.

You can set options by simply assigning them:

MyApp.config[:aws_access_key_id] = "xyz"

This will set the value in the default write backend, which by default is :memory. This means that by default, configuration which is set like this is not persisted in any way.

If you always want to assign values to a different backend, for example the database backend, you can set the default write backend like this:

Econfig.default_write_backend = :db

You can also explicitly supply the backend when setting a configuration value:

MyApp.config[:db, :aws_access_key_id] = "xyz"

MIT, see separate LICENSE.txt file