通过自学习生成库函数识别模板库  

GENERATING PATTERN BASE FOR LIBRARY FUNCTIONS RECOGNITION BY SELF-LEARNING

作　　者：刘宗田[1] 

机构地区：[1]合肥工业大学微机所

出　　处：《小型微型计算机系统》1996年第9期57-59,共3页Journal of Chinese Computer Systems

摘　　要：库函数识别是反编译中的重要的环节，也是难点之一。成功地识别库函数，不仅有利于反编译后继阶段的实现，而且对用户阅读理解反编译生成的高级语言程序大有助益。实现库函数识别，最困难的是为各种编译系统各个版本各种模式的库函数分别构造库函数识别模板库，这是工作量很大的任务。本文介绍了ＤＥＣＬＥＲ中所采用的通过自学习生成模板库的方法。该方法减轻了构造库函数识别模板库的工作量并降低了技术难度。该技术和工具将向用户开放，使用户能参与改进软件的功能。It is an importent and a difficult phase in decompiling process to recognize library functions. Succesful recognition not only is beneficial for its next phases to be implemented,bElt is useful for users to understand the programs generated by decompiling system. The reason of difficulty is that each pattern base for recognizing the library functions of each mode of each version of each compiling system has to be built respectively. The paper intruduces a method for generating pattern bases by self--learning. 16 pattern bases of decompiling system, DECLER, has been generated by the method.

关 键 词：自学习 反编译系统 库函数识别 程序语言 

分 类 号：TP314[自动化与计算机技术—计算机软件与理论]