基于标识符的代码影响域自动化分析方法

An Automatic Analysis Method of Code Influence Domain Based on Identifier

摘要:

装备软件质量评测自动化需求日益提升。装备软件在质量评测过程要经过多轮迭代修改,修改后代码的影响域分析是软件质量评测中重要的一环。目前修改后代码的影响域分析完全依赖人工。由于装备软件代码规模大,人工分析耗时较长,拉长了装备交付周期。为了提高装备软件质量评测的效率,代码影响域的自动化分析的研究势在必行。针对装备软件代码影响域分析问题,提出一种基于标识符的代码修改特征提取方法,该方法可以快速提取大规模代码的修改特征；提出一种基于标识符检测队列的影响域分析方法,可以实现装备软件影响域的快速自动化分析,准确率达85.7%,该方法可以满足现有装备软件影响域自动化分析的需求。

Abstract:

The demand for automation of equipment software quality evaluation is increasing. Multiple rounds of iterative modification are necessary in the quality evaluation process of equipment software, the impact domain analysis of the modified code is focused on the software quality evaluation process. At present, the influence domain analysis of the modified code completely relies on manual labor. As the large scale of equipment software code, manual analysis takes a long time, and the equipment delivery cycle is lengthened. In order to improve the efficiency of equipment software quality evaluation, it is imperative to study the automatic analysis of code influence domain. To solve the problem of analyzing the influence domain of equipment software code, an identifier-based code modification feature extraction method is proposed, which can quickly extract the modification characteristics of large-scale codes; an analysis method of influence domain based on identifier detection queue is proposed, which can realize rapid automatic analysis of influence domain of equipment software, with an accuracy rate of 85.7%. This method can satisfy automatic analysis of influence domain of existing equipment software.