code2doc

A toolkit for automatic code documentation generation. This repository takes un-annotated code from the examples directory, trains a custom model, and generates readable documentation for each code file.

Table of Contents

• code2doc
  • Table of Contents
  • About
  • Research Paper
  • Getting Started
    • Clone the repository
    • Create Virtual Environment
    • Install Requirements
  • Usage
    • 1. Train the Model
    • 2. Generate Documentation
  • Project Structure

About

code2doc provides a pipeline to automatically generate documentation for Python code using a model trained on example scripts. The research paper included in this repository explains the motivation, methodology, and results.

Research Paper

Please refer to research_paper.pdf in this repository for a detailed explanation of the approach, experiments, and findings behind code2doc.

Getting Started

Clone the repository

git clone https://github.com/YARE0909/code2doc.git
cd code2doc

Create Virtual Environment

It is recommended to use a virtual environment to avoid package conflicts:

python -m venv venv
# On Windows:
venv\Scripts\activate
# On macOS/Linux:
source venv/bin/activate

Install Requirements

Make sure you have pip up to date. Then install dependencies:

pip install --upgrade pip
pip install -r requirements.txt

Usage

1. Train the Model

Run main.py to train the code documentation model.

python main.py

2. Generate Documentation

After training, generate documentation for the code files within the examples directory by running:

python generate.py

The generated documentation will be saved in the output directory.

Project Structure

code2doc/
├── examples/             # Input example codes (without docstrings)
├── images/               # Supporting images for this project
├── output/               # Generated documentation will appear here along with training outputs
├── eda.py                # Exploratory Data Analysis scripts
├── generate.py           # Script to generate documentation
├── main.py               # Script to train the model
├── requirements.txt      # Python dependencies
├── research_paper.pdf    # Research paper describing the project
├── .gitignore
└── ...

For full details on methodology and experiments, see research_paper.pdf.

[1] https://github.com/YARE0909/code2doc