Publication Date
Fall 2024
Degree Type
Master's Project
Degree Name
Master of Science in Computer Science (MSCS)
Department
Computer Science
First Advisor
Thomas Austin
Second Advisor
Fabio Di Troia
Third Advisor
Faranak Abri
Keywords
Java, Comment, GPT, Sublime Text, Plugin
Abstract
Developers are notoriously disinterested in writing and maintaining comments for their code. The J-CAG plugin automates code analysis and comment generation to save developer time and improve code quality. The project utilizes the Generative Pretrained Transformer model to analyze and provide users with constructive feedback on JavaDoc comments and functions. Embedded within Sublime Text, J-CAG is designed to provide developers with useful advice. It can also generate JavaDoc comments based on the function itself, reducing the time developers spend on writing documentation. The plugin integrates seamlessly with Sublime Text, offering an intuitive interface. With positive results in comment analysis and near-perfect accuracy in comment generation, our analysis demonstrates J-CAG’s effectiveness and potential for widespread application, despite its limitations in handling function creation due to insufficient context in function logic.
Recommended Citation
Le, Linh, "J-CAG: JAVA COMMENT ANALYSIS & GENERATION - A SUBLIME TEXT PLUGIN POWERED BY GPT" (2024). Master's Projects. 1424.
https://scholarworks.sjsu.edu/etd_projects/1424
