Awesome
CoEdPilot: Recommending Code Edits with Learned Prior Edit Relevance, Project-wise Awareness, and Interactive Nature
Description
This repository contains the source code and supporting materials for the paper "CoEdPilot: Recommending Code Edits with Learned Prior Edit Relevance, Project-wise Awareness, and Interactive Nature" by Chenyan Liu, Yufan Cai, Yun Lin, Yuhuan Huang, Yunrui Pei, Bo Jiang, Ping Yang, Jin Song Dong, and Hong Mei. Presented at ISSTA'24.
Demo
<div align="center"> <a href="https://youtu.be/6G2-7Gf0Fhc"> <img src="./demo_cover.png" width="600" /> </a> </div>[!NOTE] Please click the image to watch the demo video on YouTube.
Contents
More detailed READMEs are available in each subdirectory.
/dependency_analyzer: The inference script and pre-trained model for the dependency analyzer./file_locator: The training script for semantic embedding model and the inference script to combine the score of dependency and semantic similarity./line_locator: The training and inference script for line-locator./generator: The training and inference script for edit-generator.
Getting Started
Our model scripts require Python 3.10 and Pytorch with CUDA.
Using pip (with Python 3.10):
pip install -r requirements.txt
Or using conda :
conda create -n coedpilot python=3.10.13
conda activate coedpilot
python -m pip install -r requirements.txt