摘要: Regression testing is an important part of software development, but it can be very time consuming. test selection (RTS) aims to speed up regression by running only impacted tests—the subset tests that change behavior due code changes. We present STARTS, a tool for STAtic Test Selection. Unlike dynamic RTS, STARTS requires no instrumentation or runtime information find tests; instead, uses compile-time information. Specifically, builds dependency graph program types and finds, as impacted, reach some changed type in the transitive closure graph. Maven plugin easily integrated into any Maven- based Java project. selects on average 35.2% tests, leading end-to-end 81.0% all tests. A video demo found at https://youtu.be/PCNtk8jphrM.