What is the difference between this and pycel?

[ckp95]

I'm looking for a Python library to help with reverse-engineering many complicated spreadsheets in my organization. I found this library, and also pycel. I'm not sure what the major differences are between the two. In the documentation for koala, it says that pycel

needs an Excel instance running at least once per file

But this is no longer true, afaik it now uses openpyxl to read the file and implements the Excel functions natively (or a subset of them, at least).

In the end, we have a program that is able to:

read the core information of an .xlsx file
build a graph that represents the calculation structure of the file
modify the graph to add/change cells or formulas
reduce the graph size to speed up the calculations, when possible
make calculations
draw a graphic representation of your graph
save the graph into a .gzip file
load a graph from a .gzip file And all that, without having Excel installed.

again I think pycel does all of this too.

So what is the benefit of this library over pycel (or vice versa?) Which one is more mature, feature-complete?

[bradbase]

This project leveraged from PyCel to begin. The projects satisfied different needs.

PyCel was born to assist modelling in the aviation industry with a focus on visualization of the model. It is also the "engine" behind PyXll.
https://dirkgorissen.com/2011/10/19/pycel-compiling-excel-spreadsheets-to-python-and-making-pretty-pictures/

Koala was born as an experiment to see if using Python could be used to assist in speeding up calculation of business cases for a speculative petroleum company.
https://github.com/vallettea/koala/blob/master/doc/presentation.md

Support has diverged since the beginning. They support different featuresets. PyCel appears to have been modernized where koala hasn't (yet.. I'm working on that -- I have an interest as I have a project FlyingKoala).

There is a third, Formulas (and associated Schedula). That's been born from modelling complex systems in the automotive industry.
https://github.com/vinci1it2000/formulas