Improving Program Comprehension through Dynamic Code Analysis

Abstract

Abstract. Most of the software that is currently being developed in the industry tends to be very complex and it is safe to assume that a lot of future software projects will keep escalating in their complexity. This means that developers need to keep track of numerous aspects of their system at all time and that new additions to the team will have difficulty adjusting to projects. In this paper, we will propose a set of techniques that combine already existing dynamic code analysis concepts to resolve the aforementioned problems, which can be summarized as being program comprehension difficulties. To do this we introduce a few novel software analysis and visualization techniques that facilitate program comprehension. The approach proposed within this paper allows for easy identification of semantic information, data available at execution time, which might be difficult or even impossible to portray in an easy to understand representation using existing software visualization techniques. During this paper we will be considering traditional object oriented programming languages, however these ideas should be useful in the context of other programming paradigms as well.