A Comprehensive Survey on Code Summarization
Radhika v. Kulkarni, Piyush Kothekar, Hrishikesh Kadival, Rimzim Khinchi, Sangram Gunjal · 2024
A thorough understanding of the program code is required for software development and maintenance projects. Although well-written comments are crucial for supporting this understanding, modern software frequently lacks them or has out-of-date versions. Automatic code summarization has surfaced as a viable solution to this problem. Even though this field has advanced significantly in the past ten years, there is still a noticeable lack of a current survey that compiles the most recent developments. This study classifies the articles based on using Abstract Syntax Trees (AST) in code summarization. It conducts a comprehensive evaluation of several code summarization research articles that employ AST and those that do not. The paper explores each of them based on state-of-the-art techniques. Finally, a brief overview of evaluation metrics for code summarization is provided.