Data

This dataset contains a list of 102 code smells detected from Copilot-generated Python code, along with Python code files generated by Copilot from the Repositories and Code label, respectively. This dataset also includes Copilot Chat’s responses to fixing the 102 detected code smells. A brief description of each document and folder in the dataset is provided below: 1. files folder contains 311 Python code files generated by Copilot. In the 311 Python files, 171 are retrieved under the Repositories label, indicating Python code files entirely generated by Copilot, and 140 are retrieved under the Code label, indicating Python code snippets generated by Copilot. 2. results of RQ1.xlsx contains a list of 102 code smells detected from Copilot-generated Python code. 3. results of RQ2.xlsx contains Copilot Chat’s responses to fixing the detected 102 code smells instructed by three prompts of varying detail levels.本数据集包含从GitHub Copilot生成的Python代码中检测得到的102组代码坏味（code smells），同时收录了分别取自「Repositories」与「Code」标签下、由Copilot生成的Python代码文件，此外还附带了Copilot Chat针对这102种检测出的代码坏味所给出的修复响应。下文将对数据集中的各文档与文件夹进行简要说明： 1. files文件夹：内含311份由Copilot生成的Python代码文件。其中171份取自「Repositories」标签，代表完全由Copilot生成的完整Python代码文件；剩余140份取自「Code」标签，代表Copilot生成的Python代码片段。 2. RQ1结果.xlsx：收录了从Copilot生成的Python代码中检测得到的102组代码坏味清单。 3. RQ2结果.xlsx：收录了Copilot Chat基于三种不同详细程度的提示词，针对上述102种检测出的代码坏味所生成的修复响应。