Reusability metrics for software components

Summary form only given. Assessing the reusability, adaptability, compose-ability and flexibility of software components is more and more of a necessity due to the growing popularity of component based software development (CBSD). Even if there are some metrics defined for the reusability of object-oriented software (OOS), they cannot be used for CBSD because these metrics require analysis of source code. The aim of this paper is to study the adaptability and compose-ability of software components, both qualitatively and quantitatively. We propose metrics and a mathematical model for the above-mentioned characteristics of software components. The interface characterization is the starting point of our evaluation. The adaptability of a component is discussed in conjunction with the complexity of its interface. The compose-ability metric defined for database components is extended for general software components. We also propose a metric for the complexity and adaptability of the problem solved by a component, based on its use cases. The number of alternate flows from the use case narrative is considered as a measurement for the complexity of the problem solved by a component. This was our starting point in developing a set of metrics for evaluating components functionality-wise. The main advantage of defining these metrics is the possibility to measure adaptability, reusability and quality of software components, and therefore to identify the most effective reuse strategy.