词法分析.cpp

对简单的C程序或者Pascal程序进行词法分析以TOKEN字的形式放入结果文件

#include <stdio.h>
#include<string.h>
#include<ctype.h>
#define BUFFMAXLEN 100
FILE *out;
char token[20],s[20],filename[25],outfile[25],ch;
char blank[33][10]={"auto","break","case","char","const","continue","default",
   "do","double","else","enum","extern","float","for","goto",
   "if","int","long","register","return","short","signed","sizeof",
   "static","struct","switch","typedef","union","unsigned",
   "void","volatile","while"};
void getbe()
{
	int readchar(char *ch);
    while(isspace(ch))
    ch=readchar(filename);
}
void concatenation(char ch)
{
	strcat(token,&ch);
}
void error()
{
	if(ch!=-1)
    fputs("\nyour input a data ERROR!",out);
}
int reserve()
{
	int i,m=0;
    for(i=0;i<=31;i++)
	{ 
		if(strcmp(token,blank[i])==0)m=1;
	}return(m);

}
int readchar(char * filename)
{
	static FILE *fp = NULL;
    static char buff[BUFFMAXLEN];
    static int bufflen = 0;
    static int buffpos = -1;
    if (fp == NULL)
	{
		if ((fp = fopen(filename,"r")) == NULL)
		{
			printf("File open error!\n");
            return 0;
		}
	}
	if (buffpos == -1)
	{
		if (feof(fp))
		{
			fclose(fp);
            return -1;
		}
		bufflen = fread(buff,sizeof(char),BUFFMAXLEN,fp);
        buffpos = 0;
	}
	if (buffpos < bufflen)
    ch = buff[buffpos++];
    if (buffpos >= bufflen)
		buffpos = -1;
	return ch;
}
void main()  //主函数
{
	int c,k=0;
    char b=' ';
    printf("\nPlease input filename:");
    scanf("%s",filename);
    printf("\nPlease input outfile name:");
    scanf("%s",outfile);
    if((out=fopen(outfile,"w+"))==NULL)
	{
		printf("\ncannot open outfile file!");
	}
	ch=b;
	while(ch!=-1&&ch!=0)
	{
		strcpy(token,s);
        if(k==0)
			ch=readchar(filename);
        else
		{
			ch=b;
			k=0;
		}
		getbe();
		if(isalpha(ch))
		{
			while(isalnum(ch))
			{
				concatenation(ch);
				b=ch=readchar(filename);k=1;
			}
			c=reserve();
			if(c==1)
				fprintf(out,"\n< %s ,_ >",token);
			else
				fprintf(out,"\n< ident ,\"%s\" >",token);
		}
		else
		{
			if(isdigit(ch))
			{
				while(isdigit(ch))
				{
					concatenation(ch);
					b=ch=readchar(filename);
					k=1;
				}
				fprintf(out,"\n< const ,%s >",token); 
			}
			else
				switch(ch)
			{
     case '+':
		 b=ch=readchar(filename);
		 if(ch=='+')fputs("\n< ++ , _>",out);
		 else 
		 {
			 fputs("\n< + ,_ >",out);
			 k=1;
		 }
     break;
     case '-':
		 b=ch=readchar(filename);
		 if(ch=='-')fputs("\n< -- , EQ>",out);
		 else 
		 {
			 fputs("\n< - ,_ >",out);
			 k=1;
		 }
     break;
     case '*': 
		 fputs("\n< * ,null >",out);
	 break;
     case '<': 
		 b=ch=readchar(filename);
		 if(ch=='=')fputs("\n< <= , LE>",out);
     else 
	 {
		 fputs("\n< < ,LT >",out);
		 k=1;
	 }
     break;
     case '=': 
		 b=ch=readchar(filename);
		 if(ch=='=')
			 fputs("\n< == , EQ>",out);
     else 
	 {
		 fputs("\n< = ,_ >",out);
		 k=1;
	 }
     break;
     case '>': 
		 b=ch=readchar(filename);
		 if(ch=='=')
			 fputs("\n< >= ,RE >",out);
     else 
	 {
		 fputs("\n< > ,RT >",out);
		 k=1;
	 }
     case '&': 
		 b=ch=readchar(filename);
     if(ch=='&')
		 fputs("\n< && , _>",out);
     else 
	 {
		 fputs("\n< & ,_ >",out);
		 k=1;
	 }
     break;
     case '|': 
		 b=ch=readchar(filename);
		 if(ch=='|')
			 fputs("\n< ||, _>",out);
	 else 
	 {
		 fputs("\n< | ,_ >",out);
		 k=1;
	 }
     break;
     case '!': b=ch=readchar(filename);
     if(ch=='=')fputs("\n< != , _>",out);
     else {fputs("\n< ! ,_ >",out);k=1;}
     break;
     case ':': fputs("\n< : ,_ >",out);break;
     case ',': fputs("\n< , ,_ >",out);break;
     case ';': fputs("\n< ; ,_ >",out);break;
     case '(': fputs("\n< ( ,_ >",out);break;
     case ')': fputs("\n< ) ,_ >",out);break;
     case '{': fputs("\n< { ,_ >",out);break;
     case '}': fputs("\n< } ,_ >",out);break;
     case '[': fputs("\n< [ ,_ >",out);break;
     case ']': fputs("\n< ] ,_ >",out);break;
     case '"': fputs("\n< \",_ >",out);break;
     case '#': fputs("\n< # ,_ >",out);break;
     case '\\': fputs("\n< \\ ,_ >",out);break;
     case '%': fputs("\n< % ,_ >",out);break;
     case '.': fputs("\n< . ,_ >",out);break;
     case '\'': fputs("\n< ' ,_ >",out);break;
     case '?': fputs("\n< ? ,_ >",out);break;
     case '^': fputs("\n< ^ ,_ >",out);break;
     default:
     error();
			}
		}
	}
	fclose(out);
}