Research On Full-Path Test Case Generation Technology

Full-Path in this paper is defined as the path travelling from the entry of program to some specified function point of the program.Full-Path has characteristics of more complex constraints and larger scale when compared to traditional path in software testing.Test case generation is a particularly important problem in software testing,including black-box testing and white-box testing.And path-oriented test case generation is an important technology in test case generation,which plays an important role in coverage testing of software testing.Path-oriented test case generation obtains constraints on the selected path by symbolic execution technology,and generates test cases covering the path by solving path-constraint.At present,symbolic execution is faced with the problem that the scale of path-constraint will increase sharply with the increase in program volume,which leads to the low efficiency of test case generation.As a result,automatic test case generation for large program is difficult to be accomplished in tolerable time and resource consumption.Concerning the above problems,this paper proposes several related technologies to improve the efficiency of test case generation.This paper mainly includes the following parts:In the first part,the simplification technology of path-constraint is studied for reducing the complexity of path-constraint and improving the efficiency of test case generation.While solving path-constraint,the number of expressions in path-constraint is an important factor affecting the efficiency of constraint solving.In the process of test case generation,the expressions in path-constraint extracted by symbolic execution are often not independent with each other and with implicit restrictive relations existing,so there may be redundant expressions in the path-constraint.Several methods in this paper is researched to detect and eliminate redundant expressions in path-constraint,achieving the goal of reducing the scale of path-constraint and improving the efficiency of constraint solving.Experiments verify the validity of the path-constraint simplification technology by comparing the constraint solving time of the path-constraint before and after simplification.Experiments show that the simplification technology for path-constraint can eliminate redundant expressions in path-constraint,and performing more significantly when the path-constraint is more complex.The second part is the segmentation technology for path-constraint and the constraint-segment sorting technology based on the complexity of constraint-segment.These two technologies are combined to improve efficiency by partitioning and sorting path-constraint.At present,while solving path-constraint,the solution time increases exponentially with the increase in the scale of path-constraint.In this paper,a segmentation strategy for path-constraint is proposed facing the above problem.Path-constraint is segmented according to a certain strategy,and is divided into several segments with smaller scale.And each segment is solved iteratively,optimizing the trend of time increasing with the scale of path-constraint.Meanwhile,in the process of constraint solving,the author finds that the order of segment solving has a great influence on the efficiency.So the paper defines the complexity of constraint-segment and designs the sort method of constraint-segment based on the complexity of constraint-segment.By introducing the segmentation technology and constraint-segment sorting technology for path-constraint into the Function Test System,we can prove that these technologies have significant effects on improving the efficiency of path-constraint solving and test case automatic generation.