A study on the use of implementation information for selecting specification-based test cases

Abstract

In software development, the requirements have to be established before implementation, and the specification should be in place prior to coding. Generating test cases from the specification can be done in practice at an early stage to facilitate test planning. However, as software systems grow in complexity, many important aspects can often be identified from the specification to be relevant to testing. In such cases, thorough testing may not be practical due to the limitations of project cost and time constraints. It is important to have systematic methods for selecting adequate test cases from the test suite while retaining the effectiveness of the fault detection as much as possible. In this thesis, we first propose a refined approach on the process of test case preparation. The aim is to improve the overall efficiency and effectiveness of the test case preparation process by means of automation. In consideration of the existing specification-based test case selection techniques, there is a need for testing tools to support the test case preparation process. A prototype test tool is developed and presented within which test case generation, selection and prioritization can be applied systematically. It forms a framework for the tester to prepare and select a smaller final test set from the original test suite for testing. The major objective for this thesis is to explore a more efficient and effective way for selecting test cases. In view of the potential benefit of the white-on-black approach, which makes use of implementation information to select test cases derived from the specification, we develop and examine several methods for improving the effectiveness of such an approach. Besides, we study the testing of unusual, exception or error scenarios from the specification. A common approach to reduce testing costs is to test such special scenarios only once. We propose several alternatives to assess the validity of the tester’s hypothesis that such a special scenario can be tested adequately with only one test case. Our proposed approach can also improve the fault-detecting ability of the testing of such a special scenario. The thesis demonstrates, with empirical evidence, that the integration of the white-on-black approach with our new proposed way of testing the special scenarios from the specification not only has the potential to reduce the costs of the testing process, but can also largely maintain the comprehensiveness of the original test suite.

Date of Award	3 Oct 2005
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Y.T. Yu (Supervisor)