Fangliding 谈 AI 编写测试的用途与长期主线收录
在 Xray-core PR #7201 关于未来是否从 PR 中排除测试的讨论中，Fangliding 表示，AI 编写的测试可能用上游不可能出现的调用触发问题，并导致测试用例无限膨胀。他有时会阅读这些测试以理解提交的问题，但总体认为它们不适合并入长期主线测试；他同时强调，不能说这些测试完全没有意义，需要进行清理。
作者原文@FanglidingAI written tests always have the possibility of using impossible upstream calls to trigger bugs, and can lead to infinite expansion of test cases
Sometimes I may read them to understand submitted bugs, but overall they are not suitable for being merged into long-term mainline testing. It will be incorrect if say they are completely meaningless, just do some clean