摘要： 为恢复变量的数据类型信息并提高反编译的质量，提出一种基于规则的数据类型重构方法。采用静态和动态相结合的分析方式，将可执行代码转化为中间语言，在优化中间语言的基础上为每条指令制定约束规则，并给出相应的求解算法，从而推导出变量的数据类型信息。实验结果表明，该方法能对简单数据变量类型和复杂数据变量类型进行恢复，具有较高的准确性。

关键词: 反编译, 类型重构, 类型约束, 中间语言, 类型格

Abstract: In order to recover types of variables and improve the quality of decompilation, this paper proposes a rule-based data type reconstruction technology method. It translates the executable into intermediate language with dynamic and static analysis, generates corresponding constrain for each instruction, gives out solving algorithm, and infers the real type of corresponding variable. Experimental results show that this method can recover simple data variable type and complex data variable type, and it has a high practicability.

Key words: decompilation, type reconstruction, type constraint, intermediate language, type lattice

中图分类号: 

• TP311