Advances in Software Engineering and Code Analysis

The field of software engineering is witnessing significant advancements in code analysis, generation, and maintenance. Researchers are exploring innovative approaches to improve the accuracy and efficiency of code generation, commit message generation, and bug localization. One notable trend is the integration of Large Language Models (LLMs) with traditional software engineering techniques to enhance code completion, commit untangling, and code retrieval. Additionally, there is a growing interest in developing tools and frameworks that support the analysis and management of software repositories, such as commit history analysis and API breaking change detection. These developments have the potential to improve software development productivity, quality, and maintainability. Noteworthy papers in this area include HistoryFinder, which introduces a new method for accurate and efficient method history generation, and RepoScope, which proposes a call chain-aware multi-view context for repository-level code generation. Roseau, a novel static analysis tool for API breaking change analysis, also demonstrates high accuracy and performance in detecting breaking changes in software libraries.

Sources

HistoryFinder: Advancing Method-Level Source Code History Generation with Accurate Oracles and Enhanced Algorithm

Enhancing Repository-Level Code Generation with Call Chain-Aware Multi-View Context

Observing Fine-Grained Changes in Jupyter Notebooks During Development Time

AI-Powered Commit Explorer (APCE)

LLM-Driven Collaborative Model for Untangling Commits via Explicit and Implicit Dependency Reasoning

VulGuard: An Unified Tool for Evaluating Just-In-Time Vulnerability Prediction Models

Roseau: Fast, Accurate, Source-based API Breaking Change Analysis in Java

Contextual Code Retrieval for Commit Message Generation: A Preliminary Study

SMECS: A Software Metadata Extraction and Curation Software

Gotta catch 'em all! Towards File Localisation from Issues at Large

A Deep Dive into Retrieval-Augmented Generation for Code Completion: Experience on WeChat

Evaluation of a Provenance Management Tool for Immersive Virtual Fieldwork

Built with on top of