编译原理丨第七周 ——1000. 词法分析程序设计 **

Description

设一语言的关键词、运算符、分界符的个数与单词如下： 
struct { int number; string str[10]; } keywords={3,"int","main","return"} ; //关键词
struct { int number; string str[10]; } operators ={5,"+","*","=","+=","*="}; //运算符
struct { int number; string str[10]; } boundaries ={6,"(",")","{","}",",",";"} ; //分界符
struct { int number; string str[100];} identifieres={0}; //标识符
struct { int number; string str[100];} Unsigned_integer={0}; //无符号整数
以上类号分别为1~5，序号从0开始；
标识符是字母开头的字母数字串；常量为无符号整数；
用C++设计一程序实现词法分析。

Input

输入一程序，结束符用”#”；

Output

输出单词数对：<类号，序号>。 输出标识符表，用空格分隔； 输出无符号整数表，用空格分隔；

Sample Input

 Copy sample input to clipboard

main()
{ int a=2,b=3;
  return 2*b+a;
}#

Sample Output

<1,1><3,0><3,1><3,2><1,0><4,0><2,2><5,0><3,4><4,1><2,2><5,1><3,5><1,2><5,0><2,1>
<4,1><2,0><4,0><3,5><3,3>
identifieres:a b
Unsigned_integer:2 3

Problem Source: 词法分析

编译原理西西里的题目，还是比较简单的，需要注意的是要用cin.get（ch），不然读不进空格

// Problem#: 20907
// Submission#: 5186361
// The source code is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License
// URI: http://creativecommons.org/licenses/by-nc-sa/3.0/
// All Copyright reserved by Informatic Lab of Sun Yat-sen University
//
//  main.cpp
//  cifa
//
//  Created by xie on 2017/10/18.
//  Copyright © 2017年 xie. All rights reserved.
//

#include <iostream>


using namespace std;

struct { int number; string str[10]; } keywords={3,"int","main","return"} ; //关键词
struct { int number; string str[10]; } operators ={5,"+","*","=","+=","*="}; //运算符
struct { int number; string str[10]; } boundaries ={6,"(",")","{","}",",",";"} ; //分界符
struct { int number; string str[100];} identifieres={0}; //标识符
struct { int number; string str[100];} Unsigned_integer={0}; //无符号整数


void handle(string my_str)
{
    string temp_int = "";//用来暂时存储无符号整数的一部分
    string temp_word = "";//用来暂时存储字符串


    for(int i = 0;i<my_str.size();++i)
    {
        if(my_str[i] == ' ' || my_str[i] == '\n')//空格与换行不用处理
            continue;

        //先处理分界符的情况
        if(my_str[i] == '(') 
        {
            cout<<"<3,0>";
            continue;
        }   
        if(my_str[i] == ')') 
        {
            cout<<"<3,1>";
            continue;
        }
        if(my_str[i] == '{') 
        {
            cout<<"<3,2>";
            continue;
        }
        if(my_str[i] == '}') 
        {
            cout<<"<3,3>";
            continue;
        }
        if(my_str[i] == ',') 
        {
            cout<<"<3,4>";
            continue;
        }
        if(my_str[i] == ';') 
        {
            cout<<"<3,5>";
            continue;
        }



        //再到运算符
        if(my_str[i] == '=')
        {
            cout<<"<2,2>";
            continue;
        }
        if(my_str[i] == '+' )
        {
            if(my_str[i+1] == '=')
            {
                cout<<"<2,3>";
                continue;
            }
            else
            {
                cout<<"<2,0>";
                continue;
            }
        }
        if(my_str[i] == '*')
        {
            if(my_str[i+1] == '=')
            {
                cout<<"2,4";
                continue;
            }
            else
            {
                cout<<"<2,1>";
                continue;
            }
        }

        //再到无符号整数
        if(isdigit(my_str[i]))
        {
            temp_int += my_str[i];
            if(isdigit(my_str[i+1]))//如果下一位还是数字的话，留到下一次再处理
                continue;
            else
            {
                bool flag = false;//判断数字在Unsigned_integer中是否已经存在
                for(int j = 0;j < Unsigned_integer.number;++j)
                {
                    if(temp_int == Unsigned_integer.str[j])
                    {
                        cout<<"<5,"<<j<<">";
                        temp_int = "";//记得复原temp_int
                        flag = true;
                        break;
                    }
                }
                if(flag == false)
                {
                    cout<<"<5,"<<Unsigned_integer.number<<">";
                    Unsigned_integer.str[Unsigned_integer.number++] = temp_int;
                    temp_int = "";
                    
                }
            }
        }

        if(isalpha(my_str[i]))
        {
            temp_word += my_str[i];
            if(('a'<=my_str[i+1]&&my_str[i+1]<='z')||('A'<=my_str[i+1]&&my_str[i+1]<='Z'))
            {
                //cout<<endl<<my_str[i+1];
                continue;
            }


            bool flag1 = false;
            for(int j = 0;j<3;++j)//处理关键字
            {
                if(temp_word == keywords.str[j])
                {
                    cout<<"<1,"<<j<<">";
                    flag1 = true;
                    temp_word = "";
                    break;
                }
            }
            if(flag1) continue;
            
            bool flag = false;//处理标识符，过程与上面类似
            for(int j = 0;j<identifieres.number;++j)
            {
                if(temp_word == identifieres.str[j])
                {
                    cout<<"<4,"<<j<<">";
                    temp_word = "";
                    flag = true;
                    break;
                }
            }

            if(!flag)
            {
                cout<<"<4,"<<identifieres.number<<">";
                identifieres.str[identifieres.number++] = temp_word;
                temp_word = "";
                
            }



        }



    }

    cout<<endl;
    cout<<"identifieres:";

    if(identifieres.number == 0)
        ;
    else    
        for(int i = 0;i<identifieres.number;++i)
            cout<<identifieres.str[i]<<" ";
    cout<<endl;

    cout<<"Unsigned_integer:";

    if(Unsigned_integer.number == 0)
        ;
    else
        for(int i = 0;i<Unsigned_integer.number;++i)
            cout<<Unsigned_integer.str[i]<<" ";

    cout<<endl;





}





int main(int argc, const char * argv[]) {
    
    char temp_char;
    string temp_string;
    
    while(cin.get(temp_char) && temp_char != '#')
        temp_string += temp_char;

    //cout<<temp_string;
    handle(temp_string);

    return 0;
}