异常传播网络静态建模方法及特性分析

Static modeling method and characteristic analysis of exception propagation network

异常传播现象已经成为软件开发者在异常处理过程中必须要考虑的主要问题之一.以往研究者更多关注于异常传播路径以及异常抛出的根源,而忽略异常传播结构的网络化特征.通过定义异常传播过程中方法之间的异常传递关系对异常传播结构进行网络化建模,提出了异常传播网络模型.通过对5个开源软件包的异常传播网络进行度分布、传播异常类型层次分布、聚集系数和平均最短路径等网络特征的分析发现:异常传播网络的出度和入度基本服从幂律分布;异常传播过程几乎都与"不要传播抽象异常"原则相违背;由于异常传播过程的"弱边界性",异常传播网络并没有传统软件的高内聚特征.上述分析成果可以为软件开发者在设计异常处理过程时提供指导和建议.

In area of the exception handling,exception propagation is of catholic concern.Previous researchers focused more on exception propagation paths and the source of a thrown exception but ignored the networked characteristic of exception propagation.An exception propagation networked model was proposed based on the definition of exception transitivity relationship between two methods.A case study was presented for the analysis of main characteristics of complex network,which includes the degree distribution,the exception hierarchy distribution,the clustering coefficient and the average shortest path.The results show that the in-degree and out-degree distribution are basically subject to a power law distribution;Most exception propagation processes violate the rule of ＂Do not spread the abstract exception＂;The exception propagation network appears the low cohesion feature due to the ＂Weak Border＂ characteristic.These empirical findings can provide the guidance for software developers in the design process for exception handling.