Yes! Its actually great that you mention OpenCC because it is actually currently used as a non-Latin transliteration example! Basically how this works is that the current algorithm goes through all the languages the user knows and checks if there exists a way to transliterate to that language. Again, for all latin-based languages this is straightforward, but for all non-latin based languages it does this by checking by function. Currently an example of this using Simplified and Traditional Chinese is that if the user knows Traditional Chinese and the script is Simplified, the algorithm jumps to the function that knows how to transliterate Simplified to Traditional (using OpenCC)!

This can be found here starting at line 182 if you are interested!