Test Case Generation from Statecharts

1. Test Case Generation from Statecharts Thomas Gutzeitgutzeit@upb.de ContentIntroductionBasic PrincipleAdditional Criteria for Statecharts Test Results and Conclusion

2. Statecharts as behavior model for testing graphical user interfaces • Increasing influence and use of statecharts in software engineering (UML) • Here: Statecharts defined by D. Harel • Applying the approach of F. Belli to statecharts • Are the additional properties of statecharts covered by the approach? Introduction Basic Principle Additional Criteria for Statecharts Test Results and Conclusion 

3. Basic principle of test case generation to cover all Event Pairs Introduction Basic Principle Additional Criteria for Statecharts Test Results and Conclusion 

4. Hierarchical State Coverage: By means of the event pairs Introduction Basic Principle Additional Criteria for Statecharts Test Results and Conclusion  Condition Coverage: Events will be tested by covering the outcomes TRUE and FALSE

5. History Event Coverage: Unique test sequence which uses history functionality EP: FEP: Introduction Basic Principle Additional Criteria for Statecharts Test Results and Conclusion  • Concurrent Event Coverage is reduced to testing: • parallel states separately and • [IN] conditions by setting TRUE and FALSE

6. Designed Statecharts for Real Jukebox (RJB) Introduction Basic Principle Additional Criteria for Statecharts Test Results and Conclusion 

7. 0 5 10 15 20 25 30 35 15 Number of detected Faults 10 Detected faults by Belli, but not detected by Statecharts ES number of faults Faults detected by both methods FES 5 Faults detected by Gutzeit, but not detected by Belli 0 • Results from RJB Case-Study: • Revealed faults: #15 • Conclusion: • Generated test cases from statecharts are efficient in detecting faults • But designing statecharts dedicatedly for testing is very expensive Introduction Basic Principle Additional Criteria for Statecharts Test Results and Conclusion