Message Preprocessor

Hey,

Can you use a custom component in NLU for a message preprocessing like spell correction for a message send from core, or do you need to do this with the message preprocessor? How would I do that?

I would assume both is possible since processor.py from core calls NLU parsing where each component is called?

You’d have to customise the message preprocessor in rasa_core rasa_core/processor.py at 9d0272e9e7e9e5c9e9eb68f8e80607680ccdaa96 · RasaHQ/rasa_core · GitHub

May I know how can we customize the message processor and make sure our custom class gets hit?