Recovery Of Use Case Structure In Source Code Using Static Analysis Technique

Use case was proposed by Ivar Jacobson, an important contributor of UML and rational unified process. It is the core of the whole software development process which uses UML. It can drive the whole life cycle of software. All stages can be driven by use case such as requirements analysis, software design, coding and testing, and it can maintain consistency of these stages. But in the coding phase, use cases are difficult to maintain its own structure. Recovering use cases from source code is beneficial to recover traceability of use case structure in the whole life of the software and understanding and maintenance of software.At present, the most research of recovering use cases from code focus on normal application, and further,the methods which focus on web application lack theory basis and have a poor effect, which can’t accurately recovering uses cases and its internal structure, there is existing method that recovering uses cases and its internal structure based on purpose success statement and with dominator as core, which has a better theory basis but shortages in the related experiments. Based on this method, according to the features of web application, this paper extends the method both in theory and experiment. We take JSP application as our research object.Firstly, we introduced the method of web interface modeling, so that we can describe the interactions between web interface and user accurately, and the UIML was used to model the interactions. Then we apply the existing method which based on purpose success statement to separate web page, so that we can obtain the control flow of every page which stands for use case. And then, we extract the event flow from UIML behavior model, combine with the former control flow of every web page. At last, we did the identification still from the purpose success statements.We did some research about how to identify the relationship between use cases and in this paper, we take advantage of the relationship between use case and the database which is common in web application.Experiments on several open-source systems show that the methods above can effectively recovering use case structurein web application.