Automated Testing For Anti-Patterns For Software Quality Analysing Using AOP

Abstract

This dissertation is based on the automatic detection of anti-patterns in the given code. Anti-patterns are wrong design patterns that deteriorate the quality of the software. Detection and removal of anti-patterns can enhance the software quality and will decrease the chances of the bug in the software. If we may automate the process of anti-patterns detection in the code it will reduce the time and cost of anti-pattern detection. So our goal is to automate the process of anti-pattern detection. In the dissertation we are using an Aspect oriented based approach for detection of the anti-patterns in the code. We will use aspect which is a java enhancement of implement AOP concept. AOP separate cross cutting concerns from the main goal. That is why we are using AOP this will separate testing module from main program to be tested. Program to be tested will be in java. First we will create UML diagram for the program, UML diagram will be created using ArgoUML. Generated UML will be converted into XMI, this XMI will be used as input for aspectJ testing module. XMI will have important information about the code to be tested and this will be used by asepctJ to get input that will be used for testing of anti-patterns. After getting the required information from the XMI, our aspectJ testing module will analyze the main code and find the anti-pattern in it. There are different type of anti-patterns, we will try to find some most common anti-patterns in the code. In the end results i.e. anti-patterns in the code will be displayed. These results can be used as reference to refactor the main code to remove anti-patterns.