This paper presents a novel approach for generation of test cases from UML design diagrams. In this new generation scheme, we have considered use case diagram, activity diagram and sequence diagram. Our approach consists of converting the use case diagram into use case diagram graph (UDG), activity diagram into activity diagram graph (ADG) and sequence diagram into sequence diagram graph (SDG). After that three graphs UDG, ADG and SDG are integrated to form System Graph (SYTG). The System Graph is then traversed to generate test cases also optimized using Genetic Algorithm. The generated test cases are suitable to detect maximum number of faults like use case dependency, interaction, scenario, pre-post condition faults and error handling.

Abstract-During last two decades, object-oriented paradigm has arisen as a prevailing software engineering practice for solving software problems. A large number of object-oriented metrics are proposed during that period to measure the properties like abstraction, encapsulation, inheritance, polymorphism, coupling, cohesion, information hiding and reusability. Most of the researchers mainly depict attributes like class, object, and method to measure the properties in their research. In this paper metrics are defined to depict Member Access Control mechanism and then employed in the class hierarchy. By successfully implemented the proposed metrics in object-oriented programing, we quantify the derived classes which directly results to reduce code, time and complexity of the object-oriented software systems. The proposed metrics provides a new way to understand and imply these concepts in research and development of the software using object-oriented approach.

In Object-oriented programming languages like Java; it is the basic need to define a class within another class. These classes are known as nested classes or inner classes. The scope of a nested class is limited to its outer class. All the variables and methods of outer class are accessible inside inner class enhances encapsulation. Nested classes also help in packaging of the classes. In this paper, we propose a new metric, namely, Nested Class Complexity Metric (NCCM) to measure the complexity of nested classes and the results are compared with existing metrics, which are quite encouraging. 