Ah right, thank you for the detailed explanation.
Currently, we are processing the text using a Python backend. In order to process it on the user's side, I guess we'll have to use Javascript.
I will try to fix that in the next version.
Thank you very much.