Kibana Localisations

Kibana relies on several UI frameworks (ReactJS and AngularJS) and Internationalization engine is framework agnostic and consumable in all parts of Kibana (ReactJS, AngularJS and NodeJS).

Localization files are JSON files.

Using comments can help to understand which section of the application the localization key is used for. Also namespaces are used in order to simplify message location search. For example, if we are going to translate the title of /management/sections/objects/_objects.html file, we should use message path like this: 'management.objects.objectsTitle'.

Each Kibana plugin has a separate folder with translation files located at

where locale is ISO 639 language code.

For example:

The engine scans x-pack/legacy/plugins/*/translations, src/core_plugins/*/translations, plugins/*/translations and src/legacy/ui/translations folders on initialization, so there is no need to register translation files.

The engine uses a config/kibana.yml file for locale resolution process. If locale is defined via i18n.locale option in config/kibana.yml then it will be used as a base locale, otherwise i18n engine will fall back to en. The en locale will also be used if translation can't be found for the base non-English locale.

One of our technical requirements is to have default messages in the templates themselves, and those messages will always be in English, so we don't have to keep en.json file in repository. We can generate that file from defaultMessages defined inline.

Note: locale defined in i18n.locale and the one used for translation files should match exactly, e.g. i18n.locale: zh and .../translations/zh-CN.json won't match and default English translations will be used, but i18n.locale: zh-CN and.../translations/zh-CN.json or i18n.locale: zh and .../translations/zh.json will work as expected.

Note: locale should look like zh-CN where zh - lowercase two-letter or three-letter ISO-639 code and CN - uppercase two-letter ISO-3166 code (optional). ISO-639 and ISO-3166 codes should be separated with - character.