词法分析器.cpp

课程设计时作的词法分析器

#include<stdio.h>
#include<stdlib.h>
#include<string.h>
#include<malloc.h>
#define length 42
#define MAX 2048
char *keytable[]={"include","static","auto","double","int","struct", "break","else",
                "long","switch","case","return","union","const","float","short",
				"unsigned","printf","stdio","typedef","char","extern","long",
		        "continue","for","signed","void","default","goto","sizeof","define",
				"volatile","do","if","static","while","main","malloc","\\n","\\r","\\t","\\0"};
struct linklist {    //建立链表
	char data[MAX];
	struct linklist *next;
	};
struct hashlink{    //创建哈希表
		int id;
	struct linklist *link;
}hashlist[MAX];
struct charstring   //创建字符串结构
{
	char s[MAX];
}w[MAX];
struct record{     //记录字符的机内码
	int ip;
	char string[MAX];
}hashrecord[MAX];
int l=0;
//***************************************************************************************************
void initrecord()//record结构初始化
{ int i;
	for(i=0;i<MAX;i++)
	{  hashrecord[MAX].ip=-1; }
}
//***************************************************************************************************
charstring chartable(FILE *fp)  /*字符串表*/
{
 char c;
 struct charstring  q,p;
 int i=0,k=1;
 if(!feof(fp))
 {
   q.s[i]=fgetc(fp);
   if(q.s[i]==' '||q.s[i]=='\n')
   {
	 while((c=fgetc(fp))=='	'||c=='\n'||c=='\r'||c=='\0');
	 q.s[i]=c;
   } 
   switch(q.s[i])
  {		
   case '#':case ';':case '?':case '"':case ':':case'\'':
   case '(':case ')':case '[': case ']':case '{':case '}': 
	   break;
   case '\\': 
		if((c=fgetc(fp))=='n'||c=='t'||c=='r'||c=='0')
		{  q.s[++i]=c; break;  }
		else{
        fseek(fp,-1,1);
		break;
		}
   case'.': 
	   if((c=fgetc(fp))=='0'||c=='1'||c=='2'||c=='3'||c=='4'||c=='5'||c=='6'||c=='7'||c=='8'||c=='9')
		   { q.s[++i]=c;
	         fseek(fp,-1,1);
	         c=fgetc(fp);
              goto l;
	   }	       
	   else{ fseek(fp,-1,1); break; }
    case '!':if((c=fgetc(fp))=='=')
		   {      
			   q.s[++i]=c;
			   break;
		   }
	       else {fseek(fp,-1,1);break;}
    case '*':if((c=fgetc(fp))=='='||c=='/')
		   {      
			   q.s[++i]=c;
			   break;
		   }
		   else 
		   { fseek(fp,-1,1);break;}
    case '&':if((c=fgetc(fp))=='&')
		   {      
			   q.s[++i]=c;
			   break;
		   }
	       else {fseek(fp,-1,1);break;}
			  
    case '/':    if((c=fgetc(fp))=='*')//1
				{   
			      q.s[++i]=c;
			      while(c=fgetc(fp))
				  {
				   if(c=='*')//2
				     if((c=fgetc(fp))=='/')//3
					 {
						 fseek(fp,-2,1);
						 break;
					 }
				     else fseek(fp,-1,1);
				  }//while
			      break;
				}//if1
	       else 
			   if(c=='/')
			   {
			    q.s[++i]=c;
			    fgets(p.s,MAX-1,fp);//读入一行字符；
		 	    break;
			   }
			   else 
			   { 
				     fseek(fp,-1,1);
				     break;
			   }
  case '+': if((c=fgetc(fp))=='='||c=='+')
		   {     
			   q.s[++i]=c;
			   break;
		   }
	  else
	  {fseek(fp,-1,1); 
		  if((c=fgetc(fp))=='0'||c=='1'||c=='2'||c=='3'||c=='4'||c=='5'||c=='6'||c=='7'||c=='8'||c=='9')
		   { q.s[++i]=c;
	       fseek(fp,-1,1);
	       c=fgetc(fp);
            goto l;
		   }//正数.......
	         else
			 {   fseek(fp,-1,1);break;}
	  }
  case '-':if((c=fgetc(fp))=='='||c=='-')
		   {      
			   q.s[++i]=c;
			   break;
		   }
	       else
		   {  /* fseek(fp,-1,1);
			   if((c=fgetc(fp))=='>')
			   {   q.s[i++];
		           break;
			   }
		      else
			  {  */fseek(fp,-1,1);
			    if((c=fgetc(fp))=='0'||c=='1'||c=='2'||c=='3'||c=='4'||c=='5'||c=='6'||c=='7'||c=='8'||c=='9')
				{ q.s[++i]=c;
	             fseek(fp,-1,1);
	             c=fgetc(fp);
                 goto l;
				}//负数......
	           else {fseek(fp,-1,1);break;}
		   }

  case '|':if((c=fgetc(fp))=='|')
		   {      
			   q.s[++i]=c;
			   break;
		   }
	       else {fseek(fp,-1,1);break;}
  
  case '%':if((c=fgetc(fp))=='d'||c=='f'||c=='l'||c=='s'||c=='c')
		   {    
			   q.s[++i]=c;
			   break;
		   }
	       else {fseek(fp,-1,1);break;}
  case '<':
	  if((c=fgetc(fp))=='='||c=='<')
		   {      
			   q.s[++i]=c;
			   break;
		   }
       	    else {fseek(fp,-1,1);break;}
  case '>':if((c=fgetc(fp))=='='||c=='>')
		   {      
			   q.s[++i]=c;
			   break;
		   }
	       else {fseek(fp,-1,1);break;}

  case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g': case 'h':
  case 'i': case 'j': case 'k': case 'l': case 'm': case 'n': case 'o': case 'p': 
  case 'q': case 'r': case 's': case 't': case 'u': case 'v': case 'w': case 'x':
  case 'y': case 'z': case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
  case 'G': case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N': 
  case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U': case 'V': 
  case 'W': case 'X': case 'Y': case 'Z': case '_':
    while(!feof(fp))
	{
       c=fgetc(fp);
	   switch(c)
		{
		 case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g': case 'h':
         case 'i': case 'j': case 'k': case 'l': case 'm': case 'n': case 'o': case 'p': 
         case 'q': case 'r': case 's': case 't': case 'u': case 'v': case 'w': case 'x':
         case 'y': case 'z': case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
         case 'G': case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N': 
         case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U': case 'V': 
         case 'W': case 'X': case 'Y': case 'Z': case '_': q.s[++i]=c; break;
		 default : fseek(fp,-1,1); k=0; break;	 
		}
		if(k==0) break;
	}//???????????????????????!!!!!!!!!!!!!!!!!!!!!!!!!!
  
l: case '0': case '1': case '2': case '3': case '4':
   case '5': case '6': case '7': case '8': case '9':
	 while(!feof(fp))
	 {
		  c=fgetc(fp);
		switch(c)
		{
		case '0': case '1': case '2': case '3': case '4': case '5': case '6':case '7': case '8': case '9':   
				q.s[++i]=c;   break;
			
			case'.': q.s[++i]=c;  break;
		    case'E':case'e': 
				if(((c=fgetc(fp))=='+'||c=='-')&&((c=fgetc(fp))=='0'||c=='1'||c=='2'||c=='3'||c=='4'||c=='5'||c=='6'||c=='7'||c=='8'||c=='9')) 
					 {   fseek(fp,-3,1);
				         c=fgetc(fp);
						 q.s[++i]=c; 
                         c=fgetc(fp);
                         q.s[++i]=c;
                         c=fgetc(fp);
						 q.s[++i]=c;
                         goto l;
                         fseek(fp,1,1);
					 }//科学计数法................
				else {  fseek(fp,-2,1);break; }
				default : fseek(fp,-1,1);  k=0;break;
        }
		if(k==0) break;
	  }
    default :  break;
   }
  q.s[++i]='\0';//其他字符的处理
 }
    return q;
}

//****************************************************************************************************
int chartype(char *p)/*找字符串的类型*/
{
 int q,i;  char c;
 for(i=0;i<length;i++)
  if((strcmp(p,keytable[i]))==0)
  {  q=1; break; }
  
  if(i>=length)
 {
   c=*p;
   switch(c)
   {

    case '>':case '<':case '=':case '#':case '(':
    case ')':case '{':case '}':case '[':case ']':case '$':
    case '%':case '&':case '|':case '?':case '!':case '~':
		                              q=2;        break;
    case '-':   c=*(p+1);
	            if(c=='>')  q=3;/*->的类型的判断*/
				if(c=='0'||c=='1'||c=='2'||c=='3'||c=='4'||c=='5'||c=='6'||c=='7'||c=='8'||c=='9'||c=='.') q=5;
	            else  q=2; //科学计数法
				break;   
	case'+':  c=*(p+1);
		if(c=='0'||c=='1'||c=='2'||c=='3'||c=='4'||c=='5'||c=='6'||c=='7'||c=='8'||c=='9'||c=='.') q=5;
		 else  q=2; //科学计数法
				break;
    case '/':   c=*(p+1);
	     if(c=='*')q=3;
	     else   q=2; break;
    case '*':c=*(p+1);
	     if(c=='/')  q=3;
	     else q=2; break;  /*注释的类型*/
    case ';':case '"':case ',':case ':':case '.':case '\'': 
		q=4;  break;
	case '0':case '1':case '2':case '3':case '4':case '5':case '6': case '7':case '8':case '9': 
         q=5;    break;
    default : q=6;
   }
 }
 return q;
}

//****************************************************************************************************

void creatlinklist()//链表初始化
{   	int i;
	for(i=0;i<=MAX;i++)
	{   
		hashlist[i].id=-1;
		hashlist[i].link=NULL;
	}
	
}

//****************************************************************************************************


int  hash(char *p)
{ 
	int i,k,g;
  k=0;
 char a[MAX];
 i=0;
 strcpy(a,p); 
while(a[i]!=NULL)
  {  k+=a[i];
	i++;
  }  
 g=k%2039;
 return g; 
}

//****************************************************************************************************
int searchhash(char *p)/*hash表的查找*/
{
 int w,t;
 w=hash(p);
 t=chartype(p);
 if(hashlist[w].link==NULL)
 {   
     return -1;
 }

 else
 {
	 struct linklist *x;
	 x=hashlist[w].link;
	 while(x->next!=NULL&&(strcmp(x->data,p)!=0))
		 x=x->next;
	 if(strcmp(x->data,p)==0)
	 {
		return 0;
	 }
	 else
	 {
          return -1;
	 }
 }
}
//****************************************************************************************************

void creathash(char *p)//哈希插入
{    int w,v;
     w=hash(p);
     v=chartype(p);
    struct linklist *q;
    q=(struct linklist *)malloc(sizeof(struct linklist));
    strcpy(q->data,p);
     q->next=NULL;
 	 if(hashlist[w].link==NULL)
	 { 	   hashlist[w].id=v;
            hashlist[w].link=q;
	 }
     else
	 {
      struct linklist *p2;  
      p2=hashlist[w].link;
      while(p2->next!=NULL) p2=p2->next;
      p2->next=q;
	 }
	 strcpy(hashrecord[l].string,p);
	 hashrecord[l].ip=w;
	 l++;
}
//*****************************************************************************************************
/*void tdisplay(charstring c)
{   
	int i,j;
    struct charstring *p;
	j=1;
	p=&c;
	i=0;
	while(p->s[i]!=NULL)
	{
		   printf("%c",p->s[i]);
			j++;
			++i;
	}
	printf("\n");
}*/

//*****************************************************************************************************
void display()
{
	int i/*,j*/;
    struct linklist *p;
	//j=0;
	for(i=0;i<MAX;i++)
	{
		p=hashlist[i].link;
		while(p!=NULL)
		{  // if(j%5==0){  printf("\n"); }
			printf(" (%s,%d)",p->data,hashlist[i].id);
			if(p->next!=NULL)
			{	printf("->");}  			
			else 
			{ printf("\n");  }
			   p=p->next;
		//	j++;
		}
		
	}
}
//*****************************************************************************************************
void qdisplay()
{  int i,j;
  for(i=0;i<l;i++)
	{
		if(hashrecord[i].ip!=-1)
		{   if(j%5==0){  printf("\n"); }
			printf(" (%s,%d)   ",hashrecord[i].string,hashrecord[i].ip);
			j++;
		}
	}
}
//*****************************************************************************************************
void print()
{ 
	printf("1:关键字  2:运算符   3:指针   4:界限符   5:数字  6:字符串");
}
//*****************************************************************************************************
void print();
void display();
void creathash(char *p);
int searchhash(char *p);
int chartype(char *p);
charstring chartable(FILE *fp);
void creatlinklist();
//*****************************************************************************************************
void main ()
{
 FILE *fp;
 char ch;
 
 int i,k,r;
 i=0;
 printf("**********************************词法分析器************************************");
 if ((fp=fopen("2.txt","r"))==NULL)
  printf("源程序无法打开！\n"); 
 else
 {  
	 printf("\n************************************源程序**************************************");
	 while((ch=fgetc(fp))!=EOF) 
	 {
		 printf("%c",ch);
	 }
	 fclose(fp);
	 printf("\n");
   printf("********************************************************************************");
   fp=fopen("2.txt","r");
	creatlinklist();
	initrecord();
   while(!feof(fp))
	{
     w[i]=chartable(fp);
     r=searchhash(w[i].s);
	  if(r=-1)
	  creathash(w[i].s);
        ++i;
	}
    printf("分析结果:\n");
    display();
	
	/*int j=0; 
	for(j=0;j<=i;j++)
	tdisplay(w[j]);
	*/
	printf("********************************************************************************");
    printf("是否查看帮助文件？：**是—键入'1'**否键入—'0'**:");
    scanf("%d",&k);
    if(k==1)  print();
	printf("\n********************************************************************************");
	printf("(符号串,机内码):");
	qdisplay();
 }
  printf("\n***********************************词法分析器结束*******************************");
}