{"title":"ConfigFile++: Automatic comment enhancement for misconfiguration prevention","authors":"Yuanliang Zhang, Shanshan Li, Xiangyang Xu, Xiangke Liao, Shazhou Yang, Yun Xiong","doi":"10.1109/MALTESQUE.2018.8368457","DOIUrl":null,"url":null,"abstract":"Nowadays, misconfiguration has become one of the key factors leading to system problems. Most current research on the topic explores misconfiguration diagnosis, but is less concerned with educating users about how to configure correctly in order to prevent misconfiguration before it happens. In this paper, we manually study 22 open source software projects and summarize several observations on the comments of their configuration files, most of which lack sufficient information and are poorly formatted. Based on these observations and the general process of misconfiguration diagnosis, we design and implement a tool called ConfigFile++ that automatically enhances the comment in configuration files. By using name-based analysis and machine learning, ConfigFile++ extracts guiding information about the configuration option from the user manual and source code, and inserts it into the configuration files. The format of insert comment is also designed to make enhanced comments concise and clear. We use real-world examples of misconfigurations to evaluate our tool. The results show that ConfigFile++ can prevent 33 out of 50 misconfigurations.","PeriodicalId":345739,"journal":{"name":"2018 IEEE Workshop on Machine Learning Techniques for Software Quality Evaluation (MaLTeSQuE)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 IEEE Workshop on Machine Learning Techniques for Software Quality Evaluation (MaLTeSQuE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/MALTESQUE.2018.8368457","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Nowadays, misconfiguration has become one of the key factors leading to system problems. Most current research on the topic explores misconfiguration diagnosis, but is less concerned with educating users about how to configure correctly in order to prevent misconfiguration before it happens. In this paper, we manually study 22 open source software projects and summarize several observations on the comments of their configuration files, most of which lack sufficient information and are poorly formatted. Based on these observations and the general process of misconfiguration diagnosis, we design and implement a tool called ConfigFile++ that automatically enhances the comment in configuration files. By using name-based analysis and machine learning, ConfigFile++ extracts guiding information about the configuration option from the user manual and source code, and inserts it into the configuration files. The format of insert comment is also designed to make enhanced comments concise and clear. We use real-world examples of misconfigurations to evaluate our tool. The results show that ConfigFile++ can prevent 33 out of 50 misconfigurations.