南京大学学报(自然科学版)
南京大學學報(自然科學版)
남경대학학보(자연과학판)
JOURNAL OF NANJING UNIVERSITY(NATURAL SCIENCES)
2007年
5期
509-519
,共11页
郭铭铭%窦建华%杨彬%曹璟
郭銘銘%竇建華%楊彬%曹璟
곽명명%두건화%양빈%조경
复杂性度量%AOP%信息熵
複雜性度量%AOP%信息熵
복잡성도량%AOP%신식적
complexity metric%aspect-oriented programming%information entropy
AOP(aspect-oriented programming)是一种新的程序设计范式,它能够有效解决横切关注点分离问题.由于AO系统的行为依赖于aspect与类之间的交互,因而定义一种新的、适合于AO的复杂性度量,以更加有效的理解和维护AO程序就很有必要.提出了一种新的复杂性度量,用以评估AO系统中aspect的复杂性.首先定义了aspect的信息流图,用以描述组成aspect的元素之间的交互和aspect与其他程序单元的交互;在此基础上,提出了一种基于信息熵的aspect复杂性度量,它从aspect负载的平均信息量角度量化aspect的复杂性.分析了它的外部行为特性,并使用多个AO项目进行了验证.实验结果表明,该度量反映了aspect的设计度量,能够有效的预测aspect的复杂性.
AOP(aspect-oriented programming)是一種新的程序設計範式,它能夠有效解決橫切關註點分離問題.由于AO繫統的行為依賴于aspect與類之間的交互,因而定義一種新的、適閤于AO的複雜性度量,以更加有效的理解和維護AO程序就很有必要.提齣瞭一種新的複雜性度量,用以評估AO繫統中aspect的複雜性.首先定義瞭aspect的信息流圖,用以描述組成aspect的元素之間的交互和aspect與其他程序單元的交互;在此基礎上,提齣瞭一種基于信息熵的aspect複雜性度量,它從aspect負載的平均信息量角度量化aspect的複雜性.分析瞭它的外部行為特性,併使用多箇AO項目進行瞭驗證.實驗結果錶明,該度量反映瞭aspect的設計度量,能夠有效的預測aspect的複雜性.
AOP(aspect-oriented programming)시일충신적정서설계범식,타능구유효해결횡절관주점분리문제.유우AO계통적행위의뢰우aspect여류지간적교호,인이정의일충신적、괄합우AO적복잡성도량,이경가유효적리해화유호AO정서취흔유필요.제출료일충신적복잡성도량,용이평고AO계통중aspect적복잡성.수선정의료aspect적신식류도,용이묘술조성aspect적원소지간적교호화aspect여기타정서단원적교호;재차기출상,제출료일충기우신식적적aspect복잡성도량,타종aspect부재적평균신식량각도양화aspect적복잡성.분석료타적외부행위특성,병사용다개AO항목진행료험증.실험결과표명,해도량반영료aspect적설계도량,능구유효적예측aspect적복잡성.
Aspect-oriented programming has been proposed as a new paradigm for improving separation of crosscutting concerns. The behavior of an aspect-oriented system depends on the interaction between aspects and classes. The metrics that adapt to this new paradigm are very favorable for effective understanding and maintaining the aspect-oriented software. This paper presents an entropy-based complexity measure to assess aspects in aspect-oriented programs. The aspect information flow graph is developed to represent the inter-aspect and intra-aspect interaction associated with an aspect. Based on this graph, the complexity measure is defined in terms of the average amount of information conveyed by an aspect. The behavior properties of the measure are discussed and validated by several aspect-oriented projects. The experimental results show that the measure has a positive correlation with the aspect design measures and can be used to predict the complexity of aspects.