scan.h

<compiler construction principles and practice>书中定义的tiny语言编译器。

struct
{ 
	char* str;
    TokenType tok;
} reservedWords[8] = {  {"if",IF},{"then",THEN},{"else",ELSE},{"end",END},{"repeat",REPEAT},
                                  {"until",UNTIL},{"read",READ},{"write",WRITE}  };

const int MAXTOKENLEN = 40;

const int BUFLEN = 256;

typedef enum {	START,INASSIGN,INCOMMENT,INNUM,INID,DONE  } StateType;

class Scan
{
private:
	
	char lineBuf [ BUFLEN ];
	int linepos;
	int bufsize;
	int EOF_flag;
	FILE *source;
	
protected:
	int getNextChar();
	void ungetNextChar ();
	TokenType reservedLookup (char * s);

public:
	Scan (const char* filename);
	char tokenString [ MAXTOKENLEN+1 ] ;
	TokenType getToken ();
	void getEachToken ();
};



Scan::Scan (const char* filename)
{
	source = fopen (filename,"r");
	linepos = 0;
	bufsize = 0;
	EOF_flag = 0;
	
};

int Scan::getNextChar()
{ 
	if (!(linepos < bufsize))
	{  
		lineno++;
		if (fgets(lineBuf,BUFLEN-1,source))
		{ 
			bufsize = strlen(lineBuf);
			linepos = 0;
			return lineBuf[linepos++];
		}
		else
		{ 
			EOF_flag = 1;
			return EOF;
		}
	}
	else return lineBuf[linepos++];
};

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

TokenType Scan:: reservedLookup (char * s)
{ int i;
  for (i=0;i<8;i++)
    if (!strcmp(s,reservedWords[i].str))
      return reservedWords[i].tok;
  return ID;
}

TokenType Scan::getToken()
{  
   int tokenStringIndex = 0;
   TokenType currentToken;
   StateType state = START;
   int save;

   while (state != DONE)
   { int c = getNextChar();
     save = 1;
	 switch (state)
     { case START:
         if (isdigit(c))
           state = INNUM;
         else if (isalpha(c))
           state = INID;
         else if (c == ':')
           state = INASSIGN;
         else if ((c == ' ') || (c == '\t') || (c == '\n'))
           save = 0;
         else if (c == '{')
         { save = 0;
           state = INCOMMENT;
         }
         else
         { state = DONE;
           switch (c)
           { case EOF:
               save = 0;
               currentToken = ENDFILE;
               break;
             case '=':
               currentToken = EQ;
               break;
             case '<':
               currentToken = LT;
               break;
             case '+':
               currentToken = PLUS;
               break;
             case '-':
               currentToken = MINUS;
               break;
             case '*':
               currentToken = TIMES;
               break;
             case '/':
               currentToken = OVER;
               break;
             case '(':
               currentToken = LPAREN;
               break;
             case ')':
               currentToken = RPAREN;
               break;
             case ';':
               currentToken = SEMI;
               break;
             default:
               currentToken = ERROR;
               break;
           }
         }
         break;
       case INCOMMENT:
         save = 0;
         if (c == EOF)
         { state = DONE;
           currentToken = ENDFILE;
         }
         else if (c == '}') state = START;
         break;
       case INASSIGN:
         state = DONE;
         if (c == '=')
           currentToken = ASSIGN;
         else
         { 
           ungetNextChar();
           save = 0;
           currentToken = ERROR;
         }
         break;
       case INNUM:
         if (!isdigit(c))
         { 
           ungetNextChar();
           save = 0;
           state = DONE;
           currentToken = NUM;
         }
         break;
       case INID:
         if (!isalpha(c))
         { 
           ungetNextChar();
           save = 0;
           state = DONE;
           currentToken = ID;
         }
         break;
       case DONE:
       default: 
     	 cout<<"Scanner Bug: state= "<<state<<endl;
         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);
     }
   }
     printToken(currentToken,tokenString);
     return currentToken;
} 

void Scan::getEachToken ()
{
	while ( getToken() != ENDFILE );
}