Finding Used Characters


When dealing with multiple languages having different Character Sets (Latin, Chinese, Korean, Arabic, etc) the size and number of texture atlases used to store all the different characters for that language can become prohibitively large.



To help solving this issue I2 Localization provides a tool that allows you buildomh minimal size fonts containing only the needed information.

The Charset Tool can be opened by clicking the Tools tab in the Language Source and then selecting Charset.



This tools, allows you to select a set of languages and it displays ALL characters used in the translations for that language. Then those characters can be copy/pasted into the editor you use to build the font (e.g. TextMeshPro Font Asset Creator, etc) and generate a font that contains only the characters you need in your game.


For languages like Chinese with a large number of individual characters, this tools is particularly useful as it allows you to avoid baking ALL characters given that most of them are not used in the game. Keep in mind, that chinese have around 50000 characters, while most persons only use around 2000-8000 of them, and most games only use an average of 200!


Furthermore, you can combine smaller/similar languages together (English, Spanish, Russian). Just select them, and the tool will show you all (and only those) characters your font needs to be able of correctly render those languages.



Once you generate several fonts to cover all of your languages, then you can make I2 Localization automatically pick the font to use based on the active language.

See this tutorial on how to it.

Created with the Personal Edition of HelpNDoc: Easy EPub and documentation editor