基于代码变更与历史信息的测试用例优先级排序OA

The existing methods for prioritizing test cases often rely on a single data source or analysis technique,which cannot fully capture the potential factors that affect the importance of test cases and limits the effectiveness of the ranking.To this end,a test case prioritization framework SSH-TCP based on code changes and historical information is proposed.Firstly,in the coarse-grained filtering stage,a candidate test case set is obtained through an information retrieval model;Secondly,in the fine-grained sorting stage,pre trained language models and centrality analysis based on abstract syntax trees are used to calculate the semantic and syntactic similarity of code changes and test cases,re-spectively;Finally,the final ranking result is determined based on the average historical failure rate of the test cases.The experiment shows that SSH-TCP outperforms all baseline models in terms of average fault detection rate considering cost,verifying the effectiveness of the pro-posed framework in improving the prioritization of test cases.