A Framework for Guided Test Case Generation in Constraint Logic Programming

Performing test case generation by symbolic execution on large programs becomes quickly impracticable due to the path explosion problem. A common limitation that this problem poses is the generation of unnecessarily large number of possibly irrelevant or redundant test cases even for medium-size programs. Tackling the path explosion problem and selecting high quality test cases are considered major challenges in the software testing community. In this paper we propose a constraint logic programming-based framework to guide symbolic execution and thus test case generation towards a more relevant and potentially smaller subset of paths in the program under test. The framework is realized as a tool and empirical results demonstrate its applicability and effectiveness. We show how the framework can help to obtain high quality test cases and to alleviate the scalability issues that limit most symbolic execution-based test generation approaches.