C++ 代码解析（词法分析、语法分析）

转载请注明：一雨田的专栏（http://blog.csdn.net/dylgsy）

 

     很久以前就想写个C++ 的类解析器了，终于花了一晚上时间，写了一个C++ 头文件和 CPP文件的解析。其实跟我上篇“编译原理”的文章说的一样，词法分析和语法分析最基本就是一堆 if-else 或者 switch-case。不要被什么有穷自动机吓倒了。。。

    下面的代码是一个基本的词法和语法分析，能工作，但现在这个实验品，肯定不会很完善，请各位看官见谅。

    能解析出来所包含的头文件、类名（输入 .h 文件可以解析出类名）、函数名等，需要的功能自己再加吧。

    程序我直接一个main函数写下来的，完全C风格，不用跟我讲什么分函数、分类之类的话，我也知道，后期我在代码上会分好，不过不一定贴上来了，要分你可以自己分一下，如果能共享出来给我学习一下，那更感谢了。

    我这个程序就是让你看看一个词法分析和一个语法分析的最简单的写法，同时也实现一些有点用的功能。

 

下面就是这个程序代码，能直接拷贝编译运行的。代码也有注释，逻辑比较清楚了。

 

// CodeParser.cpp : Defines the entry point for the console application.//#include "stdafx.h"#include <stdio.h>#include <tchar.h>#include <stdlib.h>#include <conio.h>#include <vector>#include <string>#include <map>using namespace std;// 符号表enum eType{Symbol};map<string, eType> g_mapSymbol;enum eCharFlag{noFlag,explain};eCharFlag g_charFlag = noFlag;int _tmain(int argc, _TCHAR* argv[]){// 输入常用标识符g_mapSymbol["if"] = Symbol;// 获取输入int i=0,j=0,k=0;char oneChar;char fname[200];memset(fname, 0, 200);printf("请输入要打开的文件名:\n");scanf("%s",fname);// for 测试//strcpy(fname, "d:\\1.cpp");// 打开文件FILE *fp;if ((fp= fopen(fname,"r"))== NULL){printf("Cannot open infile.\n");exit(0); }//int 0i;// 分析文件oneChar=fgetc(fp);char oneWord[256]={0};int nBuf = 0;vector<string> vecWords;while (oneChar!=EOF){if(nBuf == 0){if(oneChar == ' ' || oneChar == '\n' || oneChar == '\t'){}else if(oneChar == '(' || oneChar == '{'){oneWord[nBuf++]= oneChar;string strWord = oneWord;vecWords.push_back(strWord);nBuf = 0;g_charFlag = noFlag;memset(oneWord, 0, 256);//continue;}else if(oneChar == ')' || oneChar == '}'){oneWord[nBuf++]= oneChar;string strWord = oneWord;vecWords.push_back(strWord);nBuf = 0;g_charFlag = noFlag;memset(oneWord, 0, 256);//continue;}else if(oneChar >= '0' && oneChar <= '9'){/*printf("error, 不可以数字打头");return -1;*/oneWord[nBuf++]= oneChar;}else if(oneChar >= 'a' && oneChar <= 'z'){oneWord[nBuf++]= oneChar;}else if(oneChar >= 'A' && oneChar <= 'Z'){oneWord[nBuf++]= oneChar;}else if(oneChar == '#'){oneWord[nBuf++]= oneChar;}else if(oneChar == ':'){// 界限符号oneWord[nBuf++]= oneChar;}else if(oneChar == '/'){// 注释的开始，找第二根横线// 或者是路径描述oneWord[nBuf++]= oneChar;g_charFlag = explain;}else if(oneChar == '('){// 函数的开始，找第二个括号oneWord[nBuf++]= oneChar;}else if(oneChar == ')'){// 函数的结束oneWord[nBuf++]= oneChar;}else if(oneChar == '{'){oneWord[nBuf++]= oneChar;}else if(oneChar == '}'){oneWord[nBuf++]= oneChar;}else if(oneChar == ';'){// 语句结束符oneWord[nBuf++]= oneChar;}else if(oneChar == '~'){// 析构函数符oneWord[nBuf++]= oneChar;}else if(oneChar == '\"'){oneWord[nBuf++]= oneChar;}else if(oneChar == '='){oneWord[nBuf++]= oneChar;}else if(oneChar == '_'){oneWord[nBuf++]= oneChar;}else if(oneChar == '?'){oneWord[nBuf++]= oneChar;}else{//printf("未处理情况：%c", oneChar);oneWord[nBuf++]= oneChar;}}else{if(oneChar == ' ' || oneChar == '\t' || oneChar == '\n'){if(g_charFlag != explain){string strWord = oneWord;vecWords.push_back(strWord);nBuf = 0;g_charFlag = noFlag;memset(oneWord, 0, 256);}else if(g_charFlag == explain && oneChar == '\n'){// 如果是注释状态,则碰到回车才截断string strWord = oneWord;vecWords.push_back(strWord);nBuf = 0;g_charFlag = noFlag;memset(oneWord, 0, 256);}}else if(oneChar == '(' || oneChar == '{'){//oneWord[nBuf++]= oneChar;string strWord = oneWord;vecWords.push_back(strWord);nBuf = 0;g_charFlag = noFlag;memset(oneWord, 0, 256);continue;}else if(oneChar == ')' || oneChar == '}'){//oneWord[nBuf++]= oneChar;string strWord = oneWord;vecWords.push_back(strWord);nBuf = 0;g_charFlag = noFlag;memset(oneWord, 0, 256);continue;}//else if(oneChar == '\n')//{//// 行末, 处理一些情况, 例如注释等//}else if(oneChar >= '0' && oneChar <= '9'){oneWord[nBuf++]= oneChar;}else if(oneChar >= 'a' && oneChar <= 'z'){oneWord[nBuf++]= oneChar;}else if(oneChar >= 'A' && oneChar <= 'Z'){oneWord[nBuf++]= oneChar;}else if(oneChar == '_'){oneWord[nBuf++]= oneChar;}else if(oneChar == ':'){oneWord[nBuf++]= oneChar;}else if(oneChar == '('){// 函数的开始，找第二个括号oneWord[nBuf++]= oneChar;}else if(oneChar == ')'){// 函数的结束oneWord[nBuf++]= oneChar;}else if(oneChar == '{'){oneWord[nBuf++]= oneChar;}else if(oneChar == '}'){oneWord[nBuf++]= oneChar;}else if(oneChar == ';'){// 语句结束符oneWord[nBuf++]= oneChar;}else if(oneChar == '~'){// 析构函数符oneWord[nBuf++]= oneChar;}else if(oneChar == '\"'){oneWord[nBuf++]= oneChar;}else if(oneChar == '*'){// 定义指针oneWord[nBuf++]= oneChar;}else if(oneChar == '.'){// .honeWord[nBuf++]= oneChar;}else if(oneChar == '='){oneWord[nBuf++]= oneChar;}else if(oneChar == '/'){// 如果是注释状态, 等待一行结束//if(g_charFlag == explain){//oneWord[nBuf++]= oneChar;//string strWord = oneWord;//vecWords.push_back(strWord);//nBuf = 0;////g_charFlag = noFlag;//memset(oneWord, 0, 256);//oneWord[nBuf++]= oneChar;}//else{oneWord[nBuf++]= oneChar;}}else if(oneChar == ','){oneWord[nBuf++]= oneChar;}else if(oneChar == '?'){oneWord[nBuf++]= oneChar;}else{oneWord[nBuf++]= oneChar;//printf("未处理情况：%c", oneChar);//string strWord = "未处理情况：%s";//printf("error, 未处理情况.");//return -1;}}oneChar=fgetc(fp);}fclose(fp);// 写入文件FILE *fpIn;if ((fpIn= fopen("d:\\words.txt","w+"))== NULL){printf("Cannot open infile.\n");exit(0); }for(int i = 0; i < vecWords.size(); i++){char szBuf[300];memset(szBuf, 0, 300);sprintf(szBuf, "%d. %s\n", i, vecWords[i].c_str());fwrite(szBuf, strlen(szBuf), 1, fpIn);//printf("%d. %s\n", i, vecWords[i].c_str());}fclose(fpIn);// 语法解析// 首先要对括号进行匹配,验证语法等vector<string> vecClass;// 解析出类vector<string> vecHeader;// 解析出头文件typedef struct{string funcName;string paramList;}stFunc;vector<stFunc> vecFunc;// 解析出函数for(int i = 0; i < vecWords.size(); i++){if(vecWords[i] == "#include" ){// 语法分析会随着词法分析结果而变化//if(i > 0)//{//// 没被注释掉的头文件//if(vecWords[i-1] != "//")//{//vecHeader.push_back(vecWords[i+1]);//}//}vecHeader.push_back(vecWords[i+1]);}else if(vecWords[i] == "("){// 前面是函数名, 后面是参数if(i > 0){if(g_mapSymbol.find(vecWords[i-1]) != g_mapSymbol.end()){//  说明是标准标示词,不是函数continue;}stFunc f;f.funcName = vecWords[i-1];string strParamList = "(";while(vecWords[i++] != ")"){if(vecWords[i] == "("){strParamList += "(";while(vecWords[i++] != ")"){strParamList += " ";strParamList += vecWords[i];}}strParamList += " ";strParamList += vecWords[i];}f.paramList = strParamList;vecFunc.push_back(f);}}else if(vecWords[i] == "()"){// 前面是函数名, 中间无参数if(i > 0){stFunc f;f.funcName = vecWords[i-1];f.paramList = "";vecFunc.push_back(f);}}else if(vecWords[i] == "class"){// 后一个是类名string strClass = vecWords[i+1];vecClass.push_back(strClass);}}// Grammar 写入文件if ((fpIn= fopen("d:\\grammar.txt","w+"))== NULL){printf("Cannot open infile.\n");exit(0); }// 头文件for(int i = 0; i < vecHeader.size(); i++){char szBuf[500];memset(szBuf, 0, 500);sprintf(szBuf, "%d. %s \n", i, vecHeader[i].c_str());fwrite(szBuf, strlen(szBuf), 1, fpIn);}fwrite("\n\n\n", 2, 1, fpIn);// 类名for(int i = 0; i < vecClass.size(); i++){char szBuf[500];memset(szBuf, 0, 500);sprintf(szBuf, "%d. %s \n", i, vecClass[i].c_str());fwrite(szBuf, strlen(szBuf), 1, fpIn);}fwrite("\n\n\n", 2, 1, fpIn);// 函数for(int i = 0; i < vecFunc.size(); i++){char szBuf[500];memset(szBuf, 0, 500);sprintf(szBuf, "%d. %s %s \n", i, vecFunc[i].funcName.c_str(), vecFunc[i].paramList.c_str());fwrite(szBuf, strlen(szBuf), 1, fpIn);//printf("%d. %s\n", i, vecWords[i].c_str());}fclose(fpIn);// 解析完成printf("解析完成...\n");printf("词法分析结果: d:\\words.txt  语法分析结果: d:\\grammar.txt, 请输入任何键结束...\n");getch();return 0;}

自己执行就可以看到了。。。