Internationalization (i18n)

Updating the translation files

wger uses Django’s translation infrastructure, but there are a couple of things that need to be considered. First, you need to extract some translatable strings from the database such as exercise categories and muscle names:

python manage.py extract-i18n

Then, update your po files with the usual Django command (run this in the wger sub folder, not the root one):

django-admin makemessages --all --extension py,html,tpl

and finally, you can compile them with:

django-admin compilemessages

Adding new languages

Besides adding the new translations to the locale folder, they have to be activated in the Django settings file and in the application itself.

• django: add an entry to LANGUAGES in wger/settings_global.py

• wger: add the new language in the language admin page and set the visibility of exercises and ingredients. For the short name, use the language code such as ‘fr’, for the long name the native name, in this example ‘français’.

• compile: to use the new language files, the translation files have to be compiled. Do this by changing to the wger folder (so you see a locale folder there) and invoking django-admin compilemessages. You will also need to restart the webserver.

• flag icon: add an appropriate flag icon in SVG format in images/icons/flag-CODE.svg in the static folder of the core application.

• fixtures: after having added the language in the admin module, the data has to be exported so the current language configuration can be reproduced. This is done with the filter-fixtures.py script:

  • while in extras/scripts, export the whole database to a JSON file with:

    python ../../manage.py dumpdata core.language --indent 4 --natural-foreign > data.json
    

  • filter the database dump, this will generate a json file for each “important” module:

    python filter-fixtures.py
    

  • languages.json to the fixtures folder in core and optionally delete the json files:

    cp languages.json ../../wger/core/fixtures/
    rm *.json
    

Getting new languages

If you have a local installation and new languages arrive from upstream, you need to load the necessary data to the language tables in the database (note that you’ll need to reload/restart the webserver so the new po files are picked up):

python manage.py loaddata languages
python manage.py loaddata language_config

Please note that this will overwrite any changes you might have done from the language administration module.

Server translations

There are a handful of translations that are not part of the Django translation files, but are saved in the db instead. These are things like categories and muscles. These can be extracted to be used in the Django, react and flutter applications with the following command:

python manage.py extract-i18n

This command will generate the following files:

• wger/i18n.tpl: this allows the django makemessages command to extract the strings so they can be translated.

• wger/i18n.tsx: this should be copied to the react repository.

• wger/app_en.arb and wger/i18n.dart: the content of this file should be copied to the end of the intl_en.arb file in the flutter repository.

Yes, this process is by far not optimal and should definitely be changed one of these days, however these translations are very rarely updated so this is kinda ok. The downside is that local instances can’t easily add new categories or muscles.