AutoMetric: Towards Measuring Open-Source Software Quality Metrics Automatically

被引：1

作者：

Lee, Taejun ^{[1
]}

Park, Heewon ^{[1
]}

Lee, Heejo ^{[1
]}

机构：

[1] Korea Univ, Seoul, South Korea

来源：

2023 IEEE/ACM INTERNATIONAL CONFERENCE ON AUTOMATION OF SOFTWARE TEST, AST | 2023年

关键词：

Open source; Software test automation; Software metrics;

D O I：

10.1109/AST58925.2023.00009

中图分类号：

TP31 [计算机软件];

学科分类号：

081202 ; 0835 ;

摘要：

In modern software development, open-source software (OSS) plays a crucial role. Although some methods exist to verify the safety of OSS, the current automation technologies fall short. To address this problem, we propose AutoMetric, an automatic technique for measuring security metrics for OSS in repository level. Using AutoMetric which only collects repository addresses of the projects, it is possible to inspect many projects simultaneously regardless of its size and scope. AutoMetric contains five metrics: Mean Time to Update (MU), Mean Time to Commit (MC), Number of Contributors (NC), Inactive Period (IP), and Branch Protection (BP). These metrics can be calculated quickly even if the source code changes. By comparing metrics in AutoMetric with 2,675 reported vulnerabilities in GitHub Advisory Database (GAD), the result shows that the more frequent updates and commits and the shorter the inactivity period, the more vulnerabilities were found.

引用

页码：47 / 55

页数：9