Research And Implementation Of Path Selection Method For Loop Structure In Covreage Testing

Software testing is an important assurance of software quality,and the basic method of unit coverage testing is to cover the set of specific target elements by generating test cases.Because the iterations of the loop structure are difficult to figure out by the method of static analysis,it is always a difficult problem to select path for the loop structures in the test.Common 0-1 expansion strategy cannot meet the need to expand a loop structure several times,fixed K times expansion strategy also has the limitations of the number of times.This paper puts forward a concolic path selection method for loop structure which is based on static selecting and dynamic execution.This paper introduces two common strategies for the path selection of the loop structure,analyzes their limitations,and gives the improved method for the two limitations of them.Then,this paper proposes a concolic path selection method for loop structure.The method selects the prefix path for target element by static path selection,and then generates a feasible path which contains the target element by dynamic execution.However,this method is random in the selection of the iterations of the loop structure,and gives the improved dynamic and static method.The method classifies target coverage elements into free target elements and non-free target elements,and selects path for all free target elements preferentially.The records of actual iteration of loop structure are saved,which can be used in path selection for non-free target elements.Then the method selects path for non-free target elements with the reference of the records.