MIST: MODELING INPUT SPACE FOR TESTING TOOL

Dependencies among software input parameters make test case generation difficult. One possible solution is using an input space model, which allows the generation of all valid (satisfied all dependencies) and only valid test combinations. For large number of dependencies, automatic creation of the model is necessary. For this purpose, the paper presents MIST – a new tool for Modeling Input Space for Testing. MIST implements a new algorithm based on subgraph splitting and creates a software input space graphical model, where paths through the graph represent sets of valid test cases. The paper considers MIST design, implementation, and usage issues. MIST inputs and outputs are illustrated for a case study of a fragment of a nuclear power plant control system. Integration of MIST with other tools, in particular, the JUMBL tool, is addressed. This allows automatic test case generation for statistical testing. KEY WORDS Software testing, software tools, input space, dependencies among parameters