Since micropython presumably already has a built-in lexer, how difficult / size-costly would it be to expose this in a way suitable to allow the implementation of the tokenize module from the python standard library for micropython?

The reason for my request is as follows:
I am still somewhat disappointed that I cannot use help(my_func) to be reminded of the correct way to use a function. And while clearly the CPython approach to the function help functionality is unsuitable for micropython, the tokenize module would allow a pure-python implementation that mimics this behaviour by parsing the plain-text code.