Abstract
Recently, software, especially CPS and Internet of Things (IoT), increasingly have high requirements for quality, while program defects exist inevitably duo to the high complexity. Program defect repair faces serious challenges in that such repairs require considerable manpower, and the existing automatic repair approaches have difficulty generating correct patches efficiently. This paper proposes an automatic method for repairing semantic defects in Java programs based on restricted sets which refer to the interval domains of related variables that can trigger program semantic defects. Our work introduces a repair mechanism symmetrically combining defect patterns and repair templates. First, the program semantic defects are summarized into defect patterns according to their grammar and semantic features. A repair template for each type of defect pattern is predefined based on a restricted-set. Then, for each specific defect, a patch statement is automatically synthesized according to the repair template, and the detected defect information is reported by the static detection tool (DTSJava). Next, the patch location is determined by the def-use chain of defect-related variables. Finally, we evaluate the patches generated by our method using DTSJava. We implemented the method in the defect automatic repair prototype tool DTSFix to verify the effect of repairing the semantic defects detected by DTSJava in 6 Java open-source projects. The experimental results showed that 109 of 129 program semantic defects were repaired.
Highlights
With the development of information technology, the Internet of Things (IoT) has made breakthrough progress in smart transportation, smart home, public safety, and so forth, and extended it to satellites, airplanes, submarines, and other areas
By comparing the repair effect of different repair methods, we found that repair methods for specific defect patterns are more targeted and the success rate of repairs is higher
At present, related researches on the automatic repair of program semantic defects are mainly divided into three main categories—program semantic defect repair based on constraint solving, program semantic defect repair based on program specifications, and template-based program semantic defect repair
Summary
With the development of information technology, the Internet of Things (IoT) has made breakthrough progress in smart transportation, smart home, public safety, and so forth, and extended it to satellites, airplanes, submarines, and other areas. DTSJava is a static analysis tool based on defect patterns that can detect potential defects in programs, such as null pointer dereferences. The generated patch is retested by DTSJava to ensure that the patch is both correct (i.e., the defect was repaired) and safe (i.e., no new defect was induced) This approach fully utilizes the defect information reported by DTSJava to synthesize the precise condition and determine the patch location, avoiding blind automatic program repairs and improving the repair precision and efficiency. An automatic program repair method that utilizes defect information detected by the static analysis tool DTSJava. Our approach achieves an optimal balance in scalability (repairing the large-scale real-world projects), repairability (repairing more types of semantic defects detected by DTSJava), and repair quality (obtaining a functionally equivalent patch). The approach we proposed can repair the 84.5% semantic defects, and achieve the functional-equivalence repair
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
