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

Author:

Yao Wenjun1,Jiang Ying1,Yang Yang1

Affiliation:

1. Kunming University of Science and Technology, China

Abstract

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

Publisher

IGI Global

Subject

Software

Reference18 articles.

1. A Survey of Smart Code Completion Research.;Y.Bo;Journal of Software,2020

2. Bruch, M., Monperrus, M., & Mezini, M. Learning from Examples to Improve Code Completion Systems. in Joint Meeting of the European Software Engineering Conference & the Acm Sigsoft Symposium on the Foundations of Software Engineering. 2009.

3. Ding, Y. (2022). CoCoMIC: Code Completion By Jointly Modeling In-file and Cross-file Context. arXiv preprint arXiv:2212.10007.

4. Software defect prediction via convolutional neural network.;L.Jian;2017 IEEE International Conference on Software Quality, Reliability and Security (QRS),2017

5. A unified multi-task learning model for AST-level and token-level code completion

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3