Static Testing For Bad Behavior Of Resource Based On Fault Pattern Detection

With modern society becoming increasingly dependent on software, high confidence software testing is needed in a wide range. The sudden emergence of software testing based on defect pattern detection, as an important guarantee for high assurance software, can greatly reduce the defect density of software and improve its reliability. In this paper, we will use the theory of resource-related static defects based on defection pattern to detect bad behavior of resource.This paper introduces the basic conception of software testing firstly, and then analyzes the theory of software defection pattern and static testing. Using basic theory of static testing, it will do interprocedural analysis with function summary, reduce the limitations to interval arithmetic with constraint solving and perform defect detection with defection pattern state machine based on the framework of DTSJava. Finally, we put emphasis on analyzing bad behavior detection of resource about resource leaked and expired resources recalled and give the corresponding solution to improve DTSJava detection based on DTSJava framework.In this paper, we change the DTSJava resource detection from existing resource detection in order to improve detection accuracy of DTSJava. We verify the detection effectiveness and detection accuracy of bad behavior through more than two hundred typical test cases and ten open-source projects, which indicates that static testing system has more contribution to high confidence software testing.