Evaluating the use of dominance trees for C and COBOL

This paper evaluates the suitability of dominance trees as a means of re-modularising legacy systems. It uses two factors to assess the suitability of the dominance tree approach as a restructuring mechanism. Firstly, it considers if dominance tree analysis identifies suitable reuse candidates and secondly, if the candidates generated represent an improvement to the structure of the code. Applications from two domains and software written in two development languages are used to evaluate the dominance tree's suitability factors. Sequential versions of the software are used to identify specific changes that are made to the software over time, to ensure that the reuse candidates identified are supportive of software change. The results show that overall the dominance trees are useful for software re-modularisation. Furthermore, the results of this study have revealed some interesting observations regarding the general characteristics of software and the process of software change.