{"title":"Towards API-specific automatic program repair","authors":"Sebastian Nielebock","doi":"10.1109/ASE.2017.8115721","DOIUrl":null,"url":null,"abstract":"The domain of Automatic Program Repair (APR) had many research contributions in recent years. So far, most approaches target fixing generic bugs in programs (e.g., off-by-one errors). Nevertheless, recent studies reveal that about 50% of real bugs require API-specific fixes (e.g., adding missing API method calls or correcting method ordering), for which existing APR approaches are not designed. In this paper, we address this problem and introduce the notion of an API-specific program repair mechanism. This mechanism detects erroneous code in a similar way to existing APR approaches. However, to fix such bugs, it uses API-specific information from the erroneous code to search for API usage patterns in other software, with which we could fix the bug. We provide first insights on the applicability of this mechanism and discuss upcoming research challenges.","PeriodicalId":382876,"journal":{"name":"2017 32nd IEEE/ACM International Conference on Automated Software Engineering (ASE)","volume":"140 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 32nd IEEE/ACM International Conference on Automated Software Engineering (ASE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ASE.2017.8115721","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
The domain of Automatic Program Repair (APR) had many research contributions in recent years. So far, most approaches target fixing generic bugs in programs (e.g., off-by-one errors). Nevertheless, recent studies reveal that about 50% of real bugs require API-specific fixes (e.g., adding missing API method calls or correcting method ordering), for which existing APR approaches are not designed. In this paper, we address this problem and introduce the notion of an API-specific program repair mechanism. This mechanism detects erroneous code in a similar way to existing APR approaches. However, to fix such bugs, it uses API-specific information from the erroneous code to search for API usage patterns in other software, with which we could fix the bug. We provide first insights on the applicability of this mechanism and discuss upcoming research challenges.