check.java~362~

java写的词法分析器

package lexicalanalyzer2;
import java.io.*;
/**
 * Title: Check
 * Description: Check and return the four attributes of the token.java
 * Copyright: Copyright (c) 2004 15th,Oct
 * Company: B02251132
 * @author: ZengBo Huang
 * @version 1.0
 */
public class check {

  //These are all the variable in this part

  private int line,position;//Token's line and position

  private int state;//All the case's state

  private int tknln;//Length of one token

  private int chrnum;//Length of the file

  private int num_before_point_or_E;//For float number and exponent number,for example,a float number like XXX.XXXXX ,that is three

  private int forward;//The point of character in the file

  private int oneortwo=0;//Check the token's like //,==,>=,<= and etc.

  private FileReader filereader;//The file input

  private char[] chr=null;//This is the array that put the character of the input file

  private boolean is_int=true,is_float,is_exponent;//Check if a number is int,float or exponent

  private StringBuffer stringbuffer = new StringBuffer();//Create a buffer for token

  public check()throws IOException {
        chr=new char[10000];
          filereader = new FileReader("text.txt");//Create a FileReader to put the file
          chrnum = filereader.read(chr);//Put the file into the array,chrnum
  }
  public token nexttoken(){//This is a class from token
  char c;
    while (true) {//Create a roop that never down
      switch (state) {//Begin the switch include 16 cases

case 0:              //Check the if it is a letter or digit
          c = chr[forward];
          ++forward;
          position++;
          if (Character.isDigit(c)) {//If the char is a digit,turn to case 16
            state=16;
            stringbuffer.append(c);//Put the char into the buffer
            tknln++;
          }
          else
          if (Character.isLetter(c)) {//If the char is a letter,turn to case 1
            state=1;
            stringbuffer.append(c);//Put the char into the buffer
            tknln++;
          }
          else {//If not a letter or digit,turn to case 2 to check if it is other case
              forward--;
              position--;
              state=2;
          }
          break;
case 1:              //This case is fall from case 0,check if it isletter or digit
          c=chr[forward];
          ++forward;
          position++;
          if (Character.isLetterOrDigit(c)){//If it is letter or digit
            if(Character.isDigit(c)){//If it is a number
              num_before_point_or_E++;//Check the number before point
            }
            state=1;//Back to the begin of case 1 to check the next char
            stringbuffer.append(c);
            tknln++;
          }
          else {
            //Check if it is a situation like "asd12.8"
            //It should be taken as a identify "asd" and a float number 12.8
            if((c=='.'&&Character.isDigit(chr[forward+1]))||(c=='E'&&Character.isDigit(chr[forward+1]))){//If it is a float number
              state=16;//Turn to the state 16 to check the number
              forward=forward-num_before_point_or_E;//The forward point should be put before the number begin
            }
            else{//Not a float number
              tknln--;
              forward--;
              position--;
              state = 2;
              //If it is a keywords
              if (stringbuffer.toString().equals("int") ||
                  stringbuffer.toString().equals("real") ||
                  stringbuffer.toString().equals("if") ||
                  stringbuffer.toString().equals("then") ||
                  stringbuffer.toString().equals("else") ||
                  stringbuffer.toString().equals("while")) {
                token tkn = new token(stringbuffer.toString(), "Keywords",
                                      line + 1, position - tknln);//Output
                for(int count=tknln;count>=0;count--){//Cleanout the buffer
                  stringbuffer.deleteCharAt(count);
                }
                tknln = 0;
                state=0;
                return tkn;
              }
              else {//Not the keywords,then it is a identifier
                token tkn = new token(stringbuffer.toString(), "Identifier",
                                      line + 1, position - tknln);
                for(int count=tknln;count>=0;count--){
                  stringbuffer.deleteCharAt(count);
                }
                tknln = 0;
                state=0;
                return tkn;
              }
            }
          }
          break;

case 2:             //Whitespace or change the line
          c=chr[forward];
          ++forward;
          position++;
          if(c==' '){//If it is a whitespace
            state=0;//Turn to state 0 and begin to check next token
            tknln=0;
          }
          //else
          else
          if(chr[forward]=='\n'){//Change the line
          forward++;//Forward should also change the line
          state=0;
          tknln=0;
          line++;
          position=0;
          //token tkn=new token("Newline","Other"+"\t",line+1,position);
          //return tkn;
          }
          else {//Not a whitespace or line change,next case
            forward--;
            position--;
            state=3;
          }
          break;
case 3:             //Operators(+)
          c=chr[forward];
          ++forward;
          position++;
          if(c=='+'){
            token tkn=new token("+","Operator",line+1,position);//Output
            state=0;
            return tkn;
          }
          else {
            forward--;
            position--;
            state=4;
          }
          break;
case 4:             //Operators(-)
          c=chr[forward];
          ++forward;
          position++;
          if(c=='-'){
            token tkn=new token("-","Operator",line+1,position+1);//Output
            state=0;
            return tkn;
          }
          else {
            forward--;
            position--;
            state=5;
          }
          break;
case 5:             //Operators(*)
        c=chr[forward];
        ++forward;
        position++;
        if(c=='*'){
          token tkn=new token("*","Operator",line+1,position+1);//Output
          state=0;
          return tkn;
        }
        else {
          forward--;
          position--;
          state=6;
        }
        break;
case 6:             //Operators(/) and cut the comment
        c=chr[forward];
        ++forward;
        position++;
        if(c=='/'){//if it is a '/' turn back to case 6 to check if there is comment
        state=6;
        oneortwo++;
        if(oneortwo>=2){     //There are two '/'s,then cut the comment
          while(chr[forward-1]!='\n'){
            forward++;
          }
          oneortwo=0;
          state=0;
          tknln=0;
          line++;
          position=0;
        }else
        if(oneortwo==1&&chr[forward]!='/'){//One '/' and then output the '/'
          token tkn = new token("/", "Operator", line + 1, position + 1);
          oneortwo=0;
          state = 0;
          return tkn;
        }
        }
        else {
        forward--;
        position--;
        state=7;
        }
        break;
case 7:             //Operators( =  or  ==  or  <=  or  >=  or  <  or  > )
              c=chr[forward];
              ++forward;
              position++;
              if(c=='='){//Check the operator '=' and '=='
              stringbuffer.append(c);
              tknln++;
              oneortwo++;
              state=7;
              }
              else {
                state=8;
                forward--;
                position--;
                if(oneortwo==1){//There is one operator, output'<' or '>' or '=' or '!'
                  oneortwo=0;
                  if(chr[forward-1]=='<'){
                    token tkn=new token("<","Operator",line+1,position);//Output '<'
                    for(int count=tknln-1;count>=0;count--){
                        stringbuffer.deleteCharAt(count);
                      }
                    state = 0;
                    tknln = 0;
                    return tkn;
                  }
                  else
                  if(chr[forward-1]=='>'){
                    token tkn=new token(">","Operator",line+1,position);//Output '>'
                    for(int count=tknln-1;count>=0;count--){
                        stringbuffer.deleteCharAt(count);
                      }
                    state = 0;
                    tknln = 0;
                    return tkn;
                  }
                  else
                  if(chr[forward-1]=='='){
                    token tkn=new token("=","Operator",line+1,position);//Output '='
                    for(int count=tknln-1;count>=0;count--){
                        stringbuffer.deleteCharAt(count);
                      }
                    state = 0;
                    tknln = 0;
                    return tkn;
                  }
                  else
                  if(chr[forward-1]=='!'){
                    token tkn=new token("!","Operator",line+1,position);//Output '!'
                    for(int count=tknln-1;count>=0;count--){
                        stringbuffer.deleteCharAt(count);
                      }
                    state = 0;
                    tknln = 0;