Python Readability Enhancer with Test Validation

This Python-based tool is part of a Master Thesis and designed to automate the enhancement of readability in Python-Files, optionally incorporating unit tests to ensure code integrity post-enhancement. By applying readability improvements with the help of a LLM using Autogen as an Orchestrator, it refines Python scripts to make them cleaner and more maintainable. The tool (optionally) not only processes individual Python files but also pairs them with their respective unit tests to verify that readability optimizations do not introduce errors or alter the intended functionality of the code.

For a quick overview please check out the simplified Flowchart at the end of this File.

Cloning Repository and setting up Environment:

• clone this repository to your local drive
• create a new .venv with python (Version: >= 3.8, <3.13) as the base interpreter. Here's a short Video Tutorial
• run the following command in the terminal within your IDE: python -m pip install -r requirements.txt

Project Set-Up

Chose Project Name and set verbosity of the script (amount of output in the console) - config_user.py

Inside this configuration-file only edit the variables that are listed in 1. Project Set-Up (Settings for the User)

• set the use_gui variable to True in order to have the GUI displayed (if you decide to use the GUI, all the mandatory variables listed below can be set there)
• select a project_name (it will later be used to create a directory inside your_projects_edit - folder to store the edited files as well as some temporary ones)
• set the suppress_autogen_conversation and suppress_miscellaneous_info to True in order to have less output in the console
• the seed is used by AutoGen in order to have a reference to store previous conversations in its .cache folder
• set the temperature for the LLM
• set the oai_api_key variable either directly in code or leave it empty and instead set it up as described in Chapter 2.2
• select a LLM-Model with the user_specified_llm variable by setting it to either 0, 1 or 2
• provide the directories to the .py-files you want to be processed by this script inside the directories_that_contain_files_to_be_modified list
• (optionally) provide the directories to the test-.py-files you want to be used as unit tests for the verification of the changes made inside the 'directories_that_contain_test_files' list (Note: the order has to be the same as in the previous list)

Optional Step: Set 'oai_api_key' as environment variable - Windows Operating System

This step is only necessary if the API Key has not been set directly in code (left blank). The benefit of setting the API Key as an environment variable is that you will not accidentally commit to GitHub with the API Key still inside the Code for others to see.

• Right-Click on This PC or My Computer on your desktop or inside File Explorer
• Select Properties
• Click on Advanced system settings to open the System Properties window
• In the System Properties window, go to the Advanced tab
• Click on the Environment Variables button
• Under System variables (for all users) or User variables (for your user only), click New
• Create a new environment variable named OPENAI_API_KEY
• Enter your actual OpenAI API key as the variable value
• Click OK

Some Notes and Troubleshooting

How to get an OpenAI API Key?

• Log into your OpenAI account and visit: OpenAI API Keys
• As of 2024, the usage of API Services is not free. Make sure to deposit funds (at least $1) into your OpenAI account

Important Security Note

• Keep your API keys and other sensitive information secure.
• Avoid exposing them in places where unauthorized users might access them.
• When distributing your code, ensure you're not including these keys directly in the source-code or in any files that might be shared publicly.

Unit tests are not executing

• Currently only the pytest environment is supported.
• Make sure that the unit test is executable in its set up environment (it is possible to run the unit test file and get proper results)

Simplified Flowchart of the main.py-file