I18n-compatible backend for the Cortex localization engine.

Install

gem install i18n-backend-cortex

Usage

Initialize the backend with the Cortex host, and the UUID of the localization you are requesting.

require 'i18n/backend/cortex'

I18n.backend = I18n::Backend::Cortex.new({host: 'http://localhost:3000/', uuid: '61267710-4286-4db9-a074-3dd5ae9993c1'})

Polling

Tries to update all used translations every 10 minutes (using ETag and :cache), can be stopped via I18n.backend.stop_polling.

If a :cache is given, all backends pick one master to do the polling, all others refresh from :cache

require 'i18n/backend/cortex'

I18n.backend = I18n::Backend::Cortex.new({host: 'http://localhost:3000/', uuid: '61267710-4286-4db9-a074-3dd5ae9993c1', cache: Rails.cache})

I18n.t('some.key') == "Old value"
# change in backend + wait 30 minutes
I18n.t('some.key') == "New value"

:cache

If you pass :cache => Rails.cache, translations will be loaded from cache and updated in the cache.

The cache MUST support :unless_exist, so gem_of_thrones can do its job. MemCacheStore + LibmemcachedStore + ActiveSupport::Cache::MemoryStore (edge) should work.

Exceptions

To handle http exceptions provide e.g. :exception_handler => lambda{|e| puts e } (prints to stderr by default).

Fallback

If the http backend is down, it does not translate, but also does not constantly try to poll for new translations. Your app will be untranslated, but not down.

You should either use :default for all I18n.t or use a Chain, so when Cortex is down, an app-defined English translation is used, for example.

I18n.backend = I18n::Backend::Chain.new(
  I18n::Backend::Cortex.new({host: 'http://localhost:3000/', uuid: '61267710-4286-4db9-a074-3dd5ae9993c1'}),
  I18n::Backend::Simple.new
)

Authors

Colin Ewen, CB Content Enablement

Forked from i18n-backend-http, created by Michael Grosser