Create and deploy a minimal Python version of this package

[jeancochrane]

As we continue writing more production Python code, particularly in the context of our dbt project, it's become clear that having a Python version of this package would be helpful for reducing duplication of code across our Python scripts. This would follow the same pattern set by assesspy and assessr, although we should colocate the R and Python versions of the ccao package in this repo so that they can share data files in data-raw.

We should refactor this repo to add a Python ccao package that shares data with the existing R package but is deployed separately to PyPI. Since there's a lot of code to be ported, we should start small with the highest-leverage parts of the package, get the package deployed to prod, and then migrate more features over incrementally.

The scope of this issue includes:

• Create a Python package in a folder called python
• Copy vars_funs.R to Python
• Get data pipeline working for data in data-raw so that it works with Python (see the issue template for context on how this pipeline works for R)
• Set up CI checks
• Set up PyPI deployment EDIT: This package does not get deployed to any package authorities

Once these steps are complete and deployed, we can open up a follow-up issue to copy the rest of the package functionality to Python.