基于神经网络模型的类C算法向C语言程序的自动转换  

Automatic Conversion of C-Like Algorithms to C Language Programs Based on Neural Network Models

作　　者：张恒熙 陈灏然 蒋开心 李嘉卿 姚建业 ZHANG Hengxi;CHEN Haoran;JIANG Kaixin;LI Jiaqing;YAO Jianye(School of Computer and Artificial Intelligence,Zhengzhou University,Zhengzhou Henan 453003,China)

机构地区：[1]郑州大学计算机与人工智能学院,河南郑州453003

出　　处：《信息与电脑》2022年第14期76-78,共3页Information & Computer

摘　　要：为强化类C算法设计中对C语言程序的理解,基于神经网络模型研究类C算法向C语言程序的自动转换。利用编译器编写类C算法高级语言代码,通过伪码与C语言描述类C算法的文字和符号;通过类C算法代码片段语法和语义分析自动生成相应的语法分析表;基于神经网络模型实现向C语言程序的自动转换,生成语法树和目标代码,完成类C算法向C语言程序的自动转换。实验结果表明,转换后的代码与实际C语言程序函数执行结果完全一致,且自动转换得到的程序性能与直接用C语言编写的程序性能相差不大,证明本文方法具有可行性。In order to strengthen the understanding of C language program in the design of C-like algorithm, this study studies the automatic conversion of C-like algorithm to C language program based on neural network model. Use the compiler to write the high-level language code of the C-like algorithm, and describe the text and symbols of the C-like algorithm through pseudocode and C language.Through the syntax and semantic analysis of the C-like algorithm code fragment, the corresponding syntax analysis table is automatically generated. Based on the neural network model, the automatic conversion of C language programs, generating syntax trees and object codes,and completing the automatic conversion of C-like algorithms to C language programs. The experimental results show that the code converted in this paper is completely consistent with the actual C language program function execution results, and the performance of the program automatically converted in this paper is not much different from that of the program written directly in C language, which proves that the method in this paper is feasible.

关 键 词：神经网络模型 类C算法 C语言程序 代码生成技术 编译器 数据类型 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]