针对目前已有的可观测性覆盖评估方法计算复杂度高、时间开销大的问题,建立了一种可观测性信息模型,并在此基础上提出一种可观测性语句覆盖评估算法.该算法采用变量的赋值与引用作为分析基础,计算效率大大提高,同时可以很容易地与其他考虑可控制性的覆盖准则相结合.实验结果表明,可观测性语句覆盖准则增强了语句覆盖准则的评估能力,同时与已有的可观测性语句覆盖评估方法相比,文中算法时间开销大大降低.
For the observability assessment algorithms proposed up to now, the overhead of computing the metrics is very large, so they could not be integrated into simulation tools easily. A precise and concise abstract representations is proposed to model observability information. Based on this, a novel observability evaluation method is presented. This method is more computatlonally efficient than prior efforts to assess observability, and can be combined with controllability metrics easily. Experimental results show that observability evaluation enhances the capability of traditional statement coverage metric and the proposed method is efficient.