A greedy approach to object identification in imperative code

Abstract

The benefits of such recent innovations as object-oriented programming are not realized in most systems currently in use because they are, on average, over 10 years old. Additionally, they suffer from extensive maintenance. The paper addresses these concerns in the context of reverse engineering. It discusses the development of a method to identify objects in imperative code, specifically FORTRAN-77. An algorithm that uses a greedy approach to object extraction is presented. The imperative code is analyzed at the subroutine level and, using the concepts of graph theory, a set of objects is generated.<>