Metrics evaluation using open source

Abstract

Software metrics play an crucial role in the administration of the software projects. Metrics used to track development process, quantify restructuring impact and to calculate code quality. Open source software is a software product whose source code is freely available to all its intended users. Open source software during the last decade has got very remarkable success. This thesis thus been escorted concentrating on gradual directional change especially leading to more advanced form of open source software and software metrics. In this research, we analyze various software metrics at two levels that is at class level and at method level. Metrics analyze at class level are number of methods, Lack of Cohesion of Method, Average Cyclomatic Complexity, Number of Java Statements, Halstead Bug, Halstead Effort, Unweighted class size, Total Instance Variables declared, Total Packages Imported, Response for Class, Coupling between Objects, Maintainability Index, Total Number of Comment Lines in the class and Total Line of code. Metrics analyze at method level are Complexity, Number of Comment Lines, Number of Java Statements, Halstead Length, Halstead Vocabulary, Halstead Effort, Halstead Predicted Bug, Number of Classes Referenced, Number of Methods External to class called by method, Number of Methods local to class called by method and Total Line of Code in the method. Four java-based open source softwares are analyzed by assuming that the number of download indicates the success of these softwares. Tool used to find the value of metrics of these open source software is JHawk which is also a java-based open source framework which can be enclosed in any java application.