A Simple and Effective Measure for Complex Low-Level Dependencies
A Simple and Effective Measure for Complex Low-Level Dependencies
复制标题
针对复杂低级依赖关系的简单而有效的测量
DOI：
--
复制
发表时间：
2010
期刊：
影响因子：
--
通讯作者：
Ashgan Fararooy
中科院分区：
文献类型：
--
作者：
Dirk Beyer;Ashgan Fararooy
度量深度是结构问题和代码级复杂依赖的一个简单指标。我们将程序操作之间的低级依赖建模为use-def图，该图是从到达变量的定义中生成的。程序操作的依赖性越多，必须考虑的不同程序状态就越多，理解操作就越困难。依赖度计算和解释简单，应用灵活，可扩展，并独立补充其他指标。初步的实验表明，dep-degree的测量，它简单地计算使用定义图中的依赖边的数量，是可读性和可理解性的一个很好的指标。
The measure dep-degree is a simple indicator for structural problems and complex dependencies on code-level. We model low-level dependencies between program operations as use-def graph, which is generated from reaching definitions of variables. The more dependencies a program operation has, the more different program states have to be considered and the more difficult it is to understand the operation. Dep-degree is simple to compute and interpret, flexible and scalable in its application, and independently complementing other indicators. Preliminary experiments suggest that the measure dep-degree, which simply counts the number of dependency edges in the use-def graph, is a good indicator for readability and understandablity.