Measuring Complexity of Object Oriented Programs

Abstract

In this paper, a metric for object oriented language is formulated and validated. On the contrary of the other metrics used for object oriented programming (OOPS), the proposed metric calculates the complexity of a class at method level and hence considers the internal architecture of the classes, subclasses and member functions. The proposed metric is evaluated against Weyuker's proposed set of measurement principles through examples and validated through experimentation, case study and comparative study with similar measures. The practical usefulness of the metric is evaluated by a practical framework.