word.cpp

一个简单的类C语言的编译器的实现 可以参考学习编译原理的基本实现

#include<iostream>
#include<vector>
#include<stdio.h>
using namespace std;

//单词编码表
struct symbolrecord{
char symbol[10];
int id;
};
//符号表
struct entrytype{
char idname[10];
int address;
char type[10]; 
};
//常数表
struct consttype{
char idname[10];
int address;
char type[10]; 
};
//token字
struct tokentype{
	int id;
	int entry;
};
vector <symbolrecord> symbol;
vector <entrytype> entype;
vector <tokentype> token;
vector <consttype> cntype;

int ident1=0,ident2=0,counti=1,token_num=0;
void isword(char c,FILE *pf,int n);
void isnumber(char c,FILE *pf,int n);
int iscode(char c,FILE *pf,int n);

int word(char *source)
{

    FILE *pt,*pf,*pot,*poe,*poc;//单词编码表，分析文件，token，符号文件指针,const表 
    symbolrecord node;
    int flag0=0,i;
    int n;
    char c,c1;
	pt=fopen("table.txt","r");
	fscanf(pt,"%d",& n);
	while(fscanf(pt,"%s%d",& node.symbol, &node.id)==2)//输入单词编码表
	{
        symbol.push_back(node);
	}
    pf=fopen(source,"r");
	if(pt == NULL)printf("文件打开错误！");
NN:	while(flag0==0&&(c=fgetc(pf))!=EOF)
	{
		if(c=='/')//去掉注释（/**/）
		{
			if((c1=fgetc(pf))=='*')
			{
MM:				while((c1=fgetc(pf))!='*'&&c1!=EOF)
					;
				if(c1=='*')
				{
					if(fgetc(pf)=='/')
						goto NN;
					else
						goto MM;
				}
				else
				{
				    ungetc(c1,pf);
					goto NN;
				}
			}
			else
				ungetc(c1,pf);
		}
		   if(!isspace(c))//判断不是空格
		   {
              if(isalpha(c))
			  {
                isword(c,pf,n);//若是标示符调用
			  }
              else if(c>='0'&&c<='9')
			  {
				isnumber(c,pf,n);//若是数字调用
			  }
			 else
			 {
				flag0=iscode(c,pf,n);//若是关键字，算符，界符调用
			 }
		   }
		   else if(c=='\n')
			   counti++;//计数行数
	}
    pot=fopen("token.txt","w");
    poe=fopen("id.txt","w");
    poc=fopen("const.txt","w");
	//printf("符号表为：\n");
    for(i=0;i<entype.size();i++)//输出符号表
	{
		fprintf(poe,"%d\t%s\t%s\n",entype[i].address,entype[i].idname,entype[i].type);
		//printf("%d\t%s\t%s\n",entype[i].address,entype[i].idname,entype[i].type);
	}
	//printf("常数表为：\n");
    for(i=0;i<cntype.size();i++)//输出符号表
	{
		fprintf(poc,"%d\t%s\t%s\n",cntype[i].address,cntype[i].idname);
		//printf("%d\t%s\t%s\n",cntype[i].address,cntype[i].idname,cntype[i].type);
	}
	//printf("token字为：\n");
    for(i=0;i<token.size();i++)//输出token字
	{
		fprintf(pot,"%d\t%d\n",token[i].id,token[i].entry);
		//printf("%d\t%d\n",token[i].id,token[i].entry);
	}
	
	fclose(pot);
	fclose(poe);
	fclose(poc);
	fclose(pf);
	fclose(pt);
	
	return 0;
}
void isword(char c,FILE *pf,int n)
{
   char ch[10];
   ch[0]=c;
   int i=1;
   int flag=0,flag1;
   tokentype node1;
   while(isalnum((c=fgetc(pf))))//若是字符或数字继续读   
   {
	   ch[i]=c;
	   i++;
   }
   ungetc(c,pf);
   ch[i]='\0';
   for(i=0;i<(n-1)&&flag==0;i++)
   {
       if(!strcmp(ch,symbol[i].symbol))//搜索是否为关键字
	   {
            node1.entry=-1;
			node1.id=symbol[i].id;
			token.push_back(node1);
			flag=1;
	   }
   }
   if(flag==0)//不是关键字搜索符号表
   {
	   flag1=0;
	   for(i=0;i<entype.size()&&flag1==0;i++)
	   {
           if(!strcmp(ch,entype[i].idname))
		   {
		      flag1=1;
		   }
	   }
	   if(flag1==0)
	   {
           entrytype node2;
	       node2.address=ident1;
	       strcpy(node2.idname,ch);
	       strcpy(node2.type,"\0");
		   entype.push_back(node2);
		   node1.entry=ident1;
	       ident1++;
	   }
	   else
		   node1.entry=entype[i-1].address;
    node1.id=symbol[n].id;
	token.push_back(node1);
	token_num++;

   }
}
void isnumber(char c,FILE *pf,int n)
{
   char ch[10];
   ch[0]=c;
   int i=1;
   int flag=0,flag1;
   tokentype node1;
   while((c=fgetc(pf))>='0'&&c<='9')//是数字继续输入
   {
	   ch[i]=c;
	   i++;
   }
   ungetc(c,pf);
   ch[i]='\0'; 

	   flag1=0;
	   for(i=0;i<cntype.size()&&flag1==0;i++)
	   {
           if(!strcmp(ch,cntype[i].idname))
		   {
		      flag1=1;
		   }
	   }
	   if(flag1==0)
	   {
           consttype node2;
	       node2.address=ident2;
	       strcpy(node2.idname,ch);
           strcpy(node2.type,"int");
           cntype.push_back(node2);
		   node1.entry=ident2;
	       ident2++;
	   }
	   else
		   node1.entry=entype[i-1].address;
    node1.id=symbol[n-1].id;
	token.push_back(node1); 
    token_num++;  
}
int iscode(char c,FILE *pf,int n)
{
   char ch[10];
   ch[0]=c;
   int i=1;
   int flag=0;
   tokentype node1;
   if(c=='>'||c=='<'||c=='='||c=='!')//是这些符号判断是否下一个为等号
   {
	     if((c=fgetc(pf))=='=') 
		 {
			 ch[1]=c;
			 ch[2]='\0';
		 }
		 else
		 {
               ungetc(c,pf);
			   ch[1]='\0';
		 }
   }
   else
     ch[1]='\0';
   for(i=n;i<symbol.size()&&flag==0;i++)
   {
       if(!strcmp(ch,symbol[i].symbol))
	   {
            node1.entry=-1;
			node1.id=symbol[i].id;
			token.push_back(node1);
			token_num++;
			flag=1;
	   }
   }
   if(flag==1)
   {
	   return 0;
   }
   else
   {
	   printf("ERROR: WRONG CODE!!! in the line %d\n",counti);//遇到非法字符输出行数
       return 1;
   }
}