摘要
软件维护是软件开发生命周期中的关键环节，其中代码理解与重构技术在提升软件质量、降低维护成本方面具有重要作用。随着软件规模和复杂度的持续增长，传统代码理解方法已难以满足实际需求，亟需引入更高效的技术手段以应对挑战。本研究旨在探索软件维护过程中代码理解与重构的核心问题，提出一种基于静态分析与动态跟踪相结合的混合代码理解框架，并设计了一套自动化重构支持工具。通过结合程序依赖图与语义分析技术，该框架能够准确识别代码逻辑结构并生成直观的可视化表示，从而显著降低开发人员的认知负担。同时，所提出的自动化重构工具集支持多种常见重构模式，能够在保证代码功能一致性的同时优化其可维护性。实验结果表明，相较于现有方法，本研究提出的框架在代码理解效率上提升了约35%，并在重构过程中减少了约40%的人工干预。此外，本研究还验证了混合分析方法在处理大规模复杂系统时的优势，为未来相关领域的研究提供了新的思路。主要创新点在于提出了融合静态与动态信息的代码理解模型，并实现了高效的自动化重构支持机制，为软件维护实践提供了有力的技术支撑。
关键词：软件维护；代码理解；自动化重构；静态分析；动态跟踪
Abstract
Software maintenance is a critical phase in the software development lifecycle, where code comprehension and refactoring techniques play significant roles in enhancing software quality and reducing maintenance costs. As the size and complexity of software continue to grow, traditional code comprehension methods are increasingly insufficient to meet practical demands, necessitating the introduction of more efficient technical approaches to address these challenges. This study focuses on exploring the core issues of code comprehension and refactoring during software maintenance, proposing a hybrid code comprehension fr amework that integrates static analysis with dynamic tracing, as well as designing an automated refactoring support toolset. By combining program dependence graphs with semantic analysis technologies, the fr amework can accurately identify the logical structure of code and generate intuitive visual representations, thereby significantly alleviating the cognitive burden on developers. Moreover, the proposed automated refactoring toolset supports various common refactoring patterns, enabling the optimization of maintainability while ensuring functional consistency of the code. Experimental results demonstrate that the proposed fr amework improves code comprehension efficiency by approximately 35% compared to existing methods and reduces manual intervention during the refactoring process by about 40%. Additionally, this research validates the advantages of the hybrid analysis approach in handling large-scale complex systems, providing new insights for future studies in related fields. The primary innovations lie in the proposal of a code comprehension model that integrates static and dynamic information and the realization of an efficient automated refactoring support mechanism, offering robust technical support for software maintenance practices.
Keywords：Software Maintenance; Code Comprehension; Automation Refactoring; Static Analysis; Dynamic Tracking
目 录
摘要 I
Abstract II
一、绪论 1
(一) 软件维护与代码理解的研究背景 1
(二) 国内外研究现状与技术发展 1
(三) 研究方法与技术路线 2
二、代码理解的技术与方法 2
(一) 静态分析在代码理解中的应用 2
(二) 动态分析对代码行为的解析 3
(三) 可视化技术提升代码理解效率 3
三、重构技术及其实践路径 4
(一) 重构的基本原则与分类 4
(二) 自动化工具支持下的重构实现 5
(三) 重构过程中的风险评估与管理 5
四、代码理解与重构的综合案例分析 6
(一) 案例背景与问题描述 6
(二) 代码理解的具体实施步骤 6
(三) 重构效果评估与优化建议 7
结 论 9
参考文献 10
