Identifying reusable abstract data types in code

Abstract

The basic aim of this thesis is to analyse the state of the art in the field of the software reuse Software reuse is widely regarded as offering the opportunity for improving the software production process. It is expected that a massive increase in software reuse is the most promising way of overcoming the software crisis. It can lead to substantial increases in productivity and also to software systems which are more robust and more reliable. A Reuse Re-engineering process together with techniques from reverse engineering represent a method to achieve software reuse. A reference paradigm is established to implement the Reuse Re-engineering process. The reference paradigm is composed of five sequential phases, each characterised by the object it produces. This thesis deals mainly with the first phase of the reference paradigm. This phase is called Candidature and it is concerned with the analysis of the source code for the identification of sets of software components that can be candidate to make up a reusable component. Various methods involved in this phase exist in the literature. Each of them has different characteristic and different qualities. One of these approaches is analysed and it is extended in the new method to give a more precise set of reusable abstract data types. In this thesis the new method is presented. A formalisation followed by implementation of it and an evaluation of its quality through quality attributes is given.