The Metric for Automatic Code Generation Based on Dynamic Abstract Syntax Tree

In order to improve the efficiency and quality of software development, automatic code generation technology is current focus. The generated by also an important issue. However, existing metrics for ignore that programming process a continuous dynamic changeable process. So metric This article proposes method based on abstract syntax tree (DAST). More specifically, first builds DAST through interaction in behavior information between tool programmer. Then measurement contents are extracted DAST. Finally, completed with extracted. experiment results show can effectively realize generation. Compared MAST method, this convergence speed 80% when training model, shorten time-consuming average 46% doing prediction.