An Integrated Visualization Approach Combining Dynamic Data-Flow Analysis with Symbolic Execution

Troost, L.; Winkelmann, H.; Kuchen, H.

Abstract

Although studies have emphasized that generating test cases with respect to data-flow coverage is a highly effective approach to ensure software quality, there is still a lack of appropriate tooling. We contribute to this by extending the open source dynamic data-flow analysis and visualization tool Dacite with symbolic execution using the open source tool Mulib. Thereby, given a Java program and JUnit test cases, the covered data flow cannot only be identified but the user is able to receive feedback about the data flow not covered by existing test cases and can automatically generate test cases for those. This is especially suited for unit testing and early integration testing. Furthermore, to enhance the comprehensibility the identified data flow is visualized for the user with an integrated visualization using the Language Server Protocol.