/pyfactoriodata

Primary LanguagePythonMIT LicenseMIT

pyfactoriodata

This is a set of scripts that can take the raw data export from Factorio, as found on the Factorio Wiki, and produce a set of Python libraries that can be used for further scripting.

Prerequisites

  1. make
  2. awk
  3. lua
  4. python3
  5. pip3
  6. virtualenv (on my system this was a system package called python3.10-venv)

Then you can:

$ virtualenv -p python3 venv
$ source venv/bin/activate
(venv) $ pip3 install -r requirements.txt .

Instructions

If you want to start all the way from the beginning, download the raw data export from the link at the top of this file, and call this data.raw. Then, run make. This will:

  1. Convert data.raw into data.lua, a Lua script that can export the raw data as JSON. There are various cleanup activities that must be done to accomplish this successfully, most of which are in util.lua, including:

    • Substituting inf or -inf into numbers that are valid JSON.
    • Turning sparse arrays into dictionaries/tables keyed by the string representations of the original integer array indices.
    • Converting mixed tables (i.e. ones that have keys that suggest they can be treated as Lua arrays as well as keys (usually strings) for other properties.

  2. Run data.lua to output data.json.

  3. Run generate.py on data.json to create the hardcoded instances/constants of various Python classes as the __init__.py for the Python module we'll package.

  4. Run python3 -m build to produce the distributable Python packages.