Characterizing Reverse Engineering Practices on Decayed Software Applications

Abstract

Architecture reverse engineering is an approach to reproduce architectural contents once an application has deviated from its planned architecture. Unassisted understandings of an application by an individual, interviewing a person knowing the subject system, and computer-aided tools are few approaches that produce architectural contents from a decayed software application. The former two approaches are helpful when there is an individual in the organization who can understand the software application. Worst comes when computer-aided tools remain the only way to produce architectural contents from an application's source code. This research aims to identify architectural contents that industrial practitioners reproduce through reverse engineering, finding out the users of identified architectural contents and how the existing tools help in meeting industrial practitioners' needs. To achieve the research objectives, a qualitative study was performed by choosing a homogenous sampling approach from the organizations where software applications were under gradual development for many years. Semi-structured interviews were conducted, and a coding approach was used to find out themes from transcribed data. We identified different architectural contents that practitioners produce from source code. Our findings show that practitioners use reverse engineering tools to produce architectural content from an application's source code. However, there are some architectural contents that practitioners need to reverse engineer, but no available tool produces those contents. The reverse engineering tools produce a wide range of architectural contents from source code but, contents visualization as required by practitioners is a challenge that needs to be addressed.