Software Reengineering

• Target project : Deep-River

• Group Project Initial Setup

  • Single shared repository

• Task

  • 1. Introduction
  • 2. Understanding the system
    • System Evolution and Driving Forces
      • System Evolution and Driving Forces Draft
  • 3. Restructuring the system
  • 4. Conclusion

Way Point

• System Evolution and Driving Forces Draft
• Repository Mining for Deep-River
  • Detect when big changes are made and separate core contributors
• Deep-River Repository Evolution

Approach

• Navigating the GUI Repository Mining tool

  • No such thing
  • Model Context Protocol and Cursor, I thought that if I used the lab files, I could analyze them using the skills I learned in class.

• Challenges

  • The author used multiple accounts, which made it difficult to analyze commits due to the spread of contributors.
  • Sometimes too many changes were committed at once, or the commit message didn’t make sense.

• Filtering

  • Python file only
    • Extract only the date when the python file was changed the most from commits that changed Python.
    • Core module is written in python, but in the repository, ipynb or docs accounted for a large amount of changes.
  • Track changes based on key contributor and bulk changeI
    • Usually bulk happens when there is a key change.