Minecraft

Downloading the dataset:

The Minecraft dataset can be downloaded from Zenodo.

Instructions to generate dataset:

First, clone the repository using

git clone --recursive https://github.com/SET-IITGN/Minecraft.git

Now download the required libraries using:

pip3 install -r requirements.txt

Once downloaded all the dependencies, the dataset can be generated by running the following command:

python3 main.py projects.csv

projects.csv is composed of all open source repositories of multiple languages like C, C++, Java, and Python.

Extending the dataset

Dataset can be further extended by including different languages. To add languges refer to tree-sitter library and make the necessary changes in AST.py.

Citation

If you use this dataset + tool in your research, please cite our ASE 2023 paper in the format specified below:

  author={Avula, Sai Krishna and Vobbilisetti, Venkatesh and Mondal, Shouvick},
  booktitle={2023 38th IEEE/ACM International Conference on Automated Software Engineering (ASE)}, 
  title={Minecraft: Automated Mining of Software Bug Fixes with Precise Code Context}, 
  year={2023},
  volume={},
  number={},
  pages={1969-1979},
  doi={10.1109/ASE56229.2023.00116}}