scan.cpp

词法分析

#include "globals.h"
#include "util.h"
#include "scan.h"

#define BUFLEN 256
 

//static int bufsize = 0; /* current size of buffer string */
//static int EOF_flag = FALSE; /* corrects ungetNextChar behavior on EOF */

char tokenString[MAXTOKENLEN+1];



 /* states in scanner DFA */
typedef enum
   { START,INASSIGN,INCOMMENT,INNUM,INID,DONE,INEQ,INLE,INGE,INCOMMENT_S1,INCOMMENT_S2,INNOEQ}
   StateType;

 int getNextChar(void)
{ if (!(linepos < bufsize))
  { lineno++;
    if (fgets(lineBuf,BUFLEN-1,source))
    { if (EchoSource) fprintf(scan,"%4d: %s",lineno,lineBuf);
      bufsize = strlen(lineBuf);
      linepos = 0;
      return lineBuf[linepos++];
    }
    else
    { EOF_flag = TRUE;
      return EOF;
    }
  }
  else return lineBuf[linepos++];
}

 void ungetNextChar(void)
{ if (!EOF_flag) linepos-- ;}

/* lookup table of reserved words */
static struct
    { char* str;
      TokenType tok;
    } reservedWords[MAXRESERVED]
   = {{"if",IF},{"else",ELSE},
      {"int",INT},{"return",RETURN},{"void",VOID},
      {"while",WHILE}};

/* lookup an identifier to see if it is a reserved word */
/* uses linear search */
static TokenType reservedLookup (char * s)
{ int i;
  for (i=0;i<MAXRESERVED;i++)
    if (!strcmp(s,reservedWords[i].str))
      return reservedWords[i].tok;
  return ID;
}
TokenType getToken(void)
{  /* index for storing into tokenString */
   int tokenStringIndex = 0;
   /* holds current token to be returned */
   TokenType currentToken;
   /* current state - always begins at START */
   StateType state = START;
   /* flag to indicate save to tokenString */
   int save;
   while (state != DONE)
   { char c = getNextChar();
     save = TRUE;
     switch (state)
     { case START:
         if (isdigit(c))
           state = INNUM;
         else if (isalpha(c))
           state = INID;
         else if (c == '=')
           state = INEQ;
		 else if (c == '<')
           state = INLE;
		 else if (c == '>')
           state = INGE;
		 else if (c=='!')
			 state=INNOEQ;
		 
         else if ((c == ' ') || (c == '\t') || (c == '\n'))
           save = FALSE;
         else if (c == '/')
         { save=false;
           state = INCOMMENT_S1;
         }
         else
         { state = DONE;
           switch (c)
           { case EOF:
               save = FALSE;
               currentToken = ENDFILE;
               break;
             case '+':
               currentToken = PLUS;
               break;
             case '-':
               currentToken = MINUS;
               break;
            case '*':
               currentToken = TIMES;
               break;
			 case',':
			   currentToken = COMMA;
               break;

             case '(':
               currentToken = LPAREN;
               break;
             case ')':
               currentToken = RPAREN;
               break;
			  case '[':
               currentToken = LINDEX;
               break;
             case ']':
               currentToken = RINDEX;
               break;
			 case '{':
               currentToken = LBRACE;
               break;
             case '}':
               currentToken = RBRACE;
               break;
             case ';':
               currentToken = SEMI;
               break;
             default:
               currentToken = ERROR;
			   Error=TRUE;
               break;
           }
         }
         break;
	 case INCOMMENT_S1:
		 save = FALSE;
         if (c == '*')
         { 
			 state = INCOMMENT;
         }
         else 
		 { ungetNextChar();
           save = FALSE;
		   state = DONE;
           currentToken = OVER;}
         break;

	

       case INCOMMENT:
         save = FALSE;
         if (c == EOF)
         { state = DONE;
           currentToken = ENDFILE;
         }
         else if (c == '*') state = INCOMMENT_S2;
         break;
	   case INCOMMENT_S2:
		 save = FALSE;
         if (c == '/')
         { 
			 state = START;
         }
         else 
		 {
		   state = INCOMMENT;
           currentToken = OVER;}
         break;

       case INEQ:
         state = DONE;
         if (c == '=')
           currentToken = EQ;
         else
         { /* backup in the input */
           ungetNextChar();
           save = FALSE;
		   state = DONE;
           currentToken = ASSIGN;
         }
         break;
		 case INNOEQ:
         state = DONE;
         if (c == '=')
           currentToken = NOEQ;
         else
         { /* backup in the input */
           ungetNextChar();
           save = FALSE;
		   state = DONE;
           currentToken = ERROR;
         }
         break;
		 case INLE:
         state = DONE;
         if (c == '=')
           currentToken = LE;
         else
         { /* backup in the input */
           ungetNextChar();
           save = FALSE;
		   state = DONE;
           currentToken = LT;
         }
         break;
		  case INGE:
         state = DONE;
         if (c == '=')
           currentToken = GE;
         else
         { /* backup in the input */
           ungetNextChar();
           save = FALSE;
		   state = DONE;
           currentToken = GT;
         }
         break;
       case INNUM:
         if (!isdigit(c))
         { /* backup in the input */
           ungetNextChar();
           save = FALSE;
           state = DONE;
           currentToken = NUM;
         }
         break;
       case INID:
         if (!isalpha(c))
         { /* backup in the input */
           ungetNextChar();
           save = FALSE;
           state = DONE;
           currentToken = ID;
         }
         break;
       case DONE:
       default: /* should never happen */
         fprintf(scan,"Scanner Bug: state= %d\n",state);
         state = DONE;
         currentToken = ERROR;
         break;
     }
     if ((save) && (tokenStringIndex <= MAXTOKENLEN))
       tokenString[tokenStringIndex++] = (char) c;
     if (state == DONE)
     { tokenString[tokenStringIndex] = '\0';
       if (currentToken == ID)
         currentToken = reservedLookup(tokenString);
     }
   }
 if (TraceScan) 
     {
     fprintf(scan,"\t%d: ",lineno);
	// printf("%d ",tokenno++);
     printToken(currentToken,tokenString);
   }
   return currentToken;
} /* end getToken */