返回Google圖書搜尋

Holistic Use of Analysis Models in Model-based System Testing

出版	2011
URL	http://books.google.com.hk/books?id=MbNNMwEACAAJ&hl=&source=gbs_api

註釋Nowadays software testing techniques have to fulfill the requirements of growing complexity of evolving software systems. To handle the requirements current research is strongly interested in the field of model-based testing (MBT). While MBT is becom- ing the next generation of testing, using it in practice at the system test level two main problems arise: missing use of analysis models for testing purposes and low internal quality (as completeness, understandability, analysability or traceability) of automatically generated test artefacts. Those problems arise in the context of the model-driven development, when several inter-related modelling viewpoints of an analysis model (e.g. structure, behaviour and interaction) are not used while generating a test model. Such a holistic view on all viewpoints is needed to ensure the testability of the analysis model and to generate high-quality test artefacts from it. Therefore, a holistic usage of analysis models in functional model-based system testing is missing. In order to tackle the mentioned problems, we introduce a novel model-based test specification process. It consist of four steps, which result in automatically generated high quality test cases. In the first two steps a test model is automatically generated from a manually annotated analysis model. Afterwards, the test model has to be manually reviewed and extended with test data. In the last step concrete test cases are automatically generated. In our approach we use the meta-models of a representative analysis model from an industry research project and the customized meta-model of the UML Testing Profile. The whole approach is prototypically implemented and an experiment providing empirical evidence for the improvement of the internal test quality and improvement of the modelling effort is conducted. ; eng