A process for identifying changes when source code is not available

Abstract

Various regression test selection techniques have been developed and shown to improve fault detection effectiveness. The majority of these test selection techniques rely on access to source code for change identification. However, when new releases of COTS components are made available for integration and testing, source code is often not available to guide regression test selection. This paper describes a process for identifying changed functions when code is not available. This change information is beneficial for selecting white-box regression tests of customer/glue code. This process is applicable when COTS licensing agreements do not preclude decompilation. A feasibility study of the process was conducted with four releases of a medium-scale internal ABB product. The results of the feasibility study indicate that this process can be effective in identifying changed functions.