Pycord18n

Ported by YoungTrep for the new Discord.py fork - Pycord

This is a open sourced version of the internal internationalization engine used for Kolumbao.

Installation

To install the Pycord18n, you can just run the following command:

# Windows
py -m pip install pycord18n

# Linux/MacOS
python3 -m pip install pycord18n

You can now use the library!

Usage

Setting up languages

A language can be initialized like this:

french = Language("French", "fr", {
    "hello": "Bonjour",
    "goodbye": "Au revoir",
    "francais": "Français"
})

But you may want to store languages seperately and create them as follows:

import json
french = Language("French", "fr", json.load(open("fr.json")))

Base I18n class

When setting up the i18n class, we need to setup our languages and declare a fallback language:

i18n = I18n([
    Language("English", "en", {
        "hello": "Hello",
        "goodbye": "Goodbye",
        "english": "English"
    }),
    Language("French", "fr", {
        "hello": "Bonjour",
        "goodbye": "Au revoir",
        "francais": "Français"
    }),
], fallback="en")

i18n will now fallback to english if it can't find a translation for other languages.

>>> i18n.get_text("hello", "en")
'Hello'
>>> i18n.get_text("hello", "fr")
'Bonjour'
>>> # "english" is not a listed translation in the French locale, so we revert to english
>>> i18n.get_text("english", "fr")
'English'
>>> # However we can make it not fallback, but this will throw an error if the translation isn't found
>>> i18n.get_text("english", "fr", should_fallback=False) 
Traceback (most recent call last):
  ...      
py18n.i18n.InvalidTranslationKeyError: 'Translation foo not found for en!'

Discord

For Pycord, we can use the extension py18n.extension.I18nExtension. Setup your bot as you would usually, and then run i18n.init_bot as follows.

from discord.ext import commands
from py18n.extension import I18nExtension

# Make our bot
bot = commands.Bot("prefix")

# Setup similarly to the usual class
i18n = I18nExtension([
    Language("English", "en", {
        "hello": "Hello",
        "goodbye": "Goodbye",
        "english": "English"
    }),
    Language("French", "fr", {
        "hello": "Bonjour",
        "goodbye": "Au revoir",
        "francais": "Français"
    }),
], fallback="en")

# Setup the bot by giving it a function to get the user's locale.
# This could potentially refer to a database or other file.
# Anything you want!
# Otherwise, it will always be the fallback locale.
async def get_locale(ctx: commands.Context):
    preferences = {
       301736945610915852: "en"
    }
    return preferences[ctx.author.id]

# Set it up!
i18n.init_bot(bot, get_locale)

@bot.command(pass_context=True)
async def hello(ctx):
    await ctx.send(i18n.contextual_get_text("hello"))

This is all good, but because of our line i18n.init_bot(bot, get_locale), we can shorten things.

This function adds a pre-invoke hook that sets the language based on the result of get_locale. The contextually_get_text function is also exposed as py18n.extension._, and it is a classmethod.

We can change it by adding the following import and change our function:

from py18n.extension import I18nExtension, _

# ...

@bot.command(pass_context=True)
async def hello(ctx):
    await ctx.send(_("hello"))

There, much tidier!

• The _ function considers the current context and uses the correct locale by default.
• When initializing any I18nExtension, as we did earlier, it becomes the default i18n instance. The default instance is used by _ and contextually_get_text.

Issues

If you encounter any problems, check out current issues or make a new issue.

Notes

• Feel free to contribute! This is released under the GLP-3 license. (If you suggest another license, make an issue suggesting).