A little off topic, The author here describes the use case for this and mentions for those situations where "if you do not need probabilities attached to your parse trees, you are probably better off using nearley instead"
This lead me to wonder (and unsuccessfully answer via google-fu): What is the difference between [jison](https://zaa.ch/jison/docs/) and [nearley](https://nearley.js.org) and when would I use one over the other?