[1]刘雨,姜瑛.面向程序员编程过程中重点代码修改识别研究[J].计算机技术与发展,2025,(01):81-87.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0263]
　LIU Yu,JIANG Ying.Research of Focused Code Modification Identification in Programmer-oriented Programming Processes[J].,2025,(01):81-87.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0263]
点击复制

面向程序员编程过程中重点代码修改识别研究()

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2025年01期
页码:
81-87
栏目:
软件技术与工程
出版日期:
2025-01-10

文章信息/Info

Title:
Research of Focused Code Modification Identification in Programmer-oriented Programming Processes
文章编号:
1673-629X(2025)01-0081-07
作者:
刘雨1; 2; 姜瑛1; 2
1. 云南省计算机技术应用重点实验室,云南 昆明 650500;2. 昆明理工大学 信息工程与自动化学院,云南 昆明 650500
Author(s):
LIU Yu1; 2; JIANG Ying1; 2
1. Computer Technology Application Key Laboratory of Yunnan Province,Kunming 650500,China;2. School of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650500,China
关键词:
软件维护; 语法结构; 语义信息; 注意力机制; 深度学习; 重点代码修改
Keywords:
software maintenance; syntactic structure; semantic information; attention mechanisms; deep learning; focused code modifica-tion
分类号:
TP311.5
DOI:
10.20165/j.cnki.ISSN1673-629X.2024.0263
摘要:
软件的持续变化增加了维护的复杂性。 为提高维护效率,该文提出了一种基于动态抽象语法树(Dynamic Abstract Syntax Tree,DAST)的重点代码修改识别方法。传统的修改识别方法主要依赖语法分析,忽略了程序语义的重要信息,导致识别的精度和粒度受到限制。 通过 DAST 提取出代码的重点信息,结合语法结构和语义信息,以提高程序员在编程过程中重点代码修改识别的准确性。 同时,引入了注意力机制,进一步突出重点代码的频繁修改区域。 最后在版本变化和三个不同程序员的数据集上进行实验,结果表明,与传统的重点代码修改识别方法相比,该方法在准确性和稳定性方面均有提升,验证了在重点代码修改识别任务上的有效性,从而提升了软件维护效率。
Abstract:
Continuous changes in software increase the complexity of maintenance. To improve the maintenance efficiency,we propose a focused code modification identification method based on Dynamic Abstract Syntax Tree ( DAST ). The traditional modification identification method mainly relies on syntactic analysis and ignores the important information of program semantics,resulting in limited accuracy and granularity of identification. The key information of the code is extracted by DAST and the syntactic structure and semantic information is combined to improve the programmer ’s key code modification identification accuracy in the programming process. Meanwhile,we introduce an attention mechanism to further highlight the frequently modified regions of focused code. Finally,experiments are conducted on version change and three different programmers’ datasets. It is showed that compared with the traditional focused code modification identification method,the proposed method improves both accuracy and stability,and verifies the effectiveness in the task of focused code modification identification,thus improving the software maintenance efficiency.

相似文献/References:

[1]严秀 李龙澍.软件逆向工程技术研究[J].计算机技术与发展,2009,(04):20.
　YAN Xiu,LI Long-shu.Research of Technology in Software Reverse Engineering[J].,2009,(01):20.
[2]阚红星 马溪骏 桂宏新.基于COCOMOⅡ的自动测试维护代价实例研究[J].计算机技术与发展,2008,(11):47.
　KAN Hong-xing,ma Xi-jun,GUI Hong-xin.A Case Study on Maintenance Cost for Regression Test Automation Based on COCOMO Ⅱ[J].,2008,(01):47.
[3]陈永郑 李龙澍.基于程序切片技术的回归测试方法研究[J].计算机技术与发展,2007,(12):113.
　CHEN Yong-zheng,LI Long-shu.Regression Testing Based on Program Slicing[J].,2007,(01):113.
[4]丁剑洁 鱼滨 侯红.软件维护中程序理解的应用与研究[J].计算机技术与发展,2007,(04):218.
　DING Jian-Jie,YU Bin,HOU Hong.Research and Application of Program Understanding in the Software Maintenance[J].,2007,(01):218.
[5]姜文,刘立康. 现代应用软件的维护与技术支持[J].计算机技术与发展,2015,25(04):116.
　JIANG Wen,LIU Li-kang. Maintenance and Technical Support of Modern Application Software[J].,2015,25(01):116.
[6]侯 敏,张丽萍.克隆代码检测技术研究[J].计算机技术与发展,2019,29(08):86.[doi:10. 3969 / j. issn. 1673-629X. 2019. 08. 017]
　HOU Min,ZHANG Li-ping.Research on Software Clone Detection Technology[J].,2019,29(01):86.[doi:10. 3969 / j. issn. 1673-629X. 2019. 08. 017]
[7]贾 清,杨 抒.基于 Word2vec 的克隆代码检测方法研究[J].计算机技术与发展,2020,30(08):124.[doi:10. 3969 / j. issn. 1673-629X. 2020. 08. 021]
　JIA Qing,YANG Shu.Research on Clone Code Detection Method Based on Word2vec[J].,2020,30(01):124.[doi:10. 3969 / j. issn. 1673-629X. 2020. 08. 021]

更新日期/Last Update: 2025-01-10