编译原理词法分析程序C++实现

题目要求：对给定的C语言程序，识别出关键字、整型常量、实型常量（包括1.23E-23这样的科学计数法）、标识符、分隔符、运算符。
语言：C++
集成环境：VS 2017
算法：使用状态转换图实现

程序代码：

/*
关键字：<AUTO, >；标示符：<IDENT,my_name>；整型常量<200,数值>；实型常量<300,数值>；分隔符：<400-408, >；运算符：<500-, >
*/
#include <iostream>
#include <vector>
#include <string>
#include <fstream>
#include <cassert>
#include <sstream>
#include <map>
using namespace std;
struct Tuple{
	string val;
	string content;
};
vector<string> constant = { "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","UNSIGNED","VOID","VOLATILE","WHILE"};
vector<Tuple> result;
vector<char>file_content;
vector<string> keyword = { "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","unsigned","void","volatile","while" };
vector<char>separater = { ';',',','{','}','[',']','(',')' };
vector<char>operators = { '+','-','*','/','>','<','=','!','&','|'};
void GetFile(string file_name)
{
	ifstream in_file;
	in_file.open(file_name.data());
	assert(in_file.is_open());
	char c;
	in_file >> noskipws;
	while (!in_file.eof())
	{
		in_file >> c;
		file_content.push_back(c);
	}
	in_file.close();
}
bool IsLetter(char c)
{
	if ((c >= 'A'&&c <= 'Z') || (c >= 'a'&&c <= 'z'))
		return true;
	return false;
}
bool IsBlank(char c)
{
	if (c == ' ')
		return true;
	return false;
}
bool IsNumber(char c) 
{
	if (c >= '0'&&c <= '9')
		return true;
	return false;
}
int IsKeyword(string str)
{
	for (int i = 0; i < keyword.size(); i++)
		if (str == keyword[i])
			return i+1;
	return 0;
}
bool IsInteger(string str)
{
	for (int i = 0; i < str.size(); i++)
	{
		if (!IsNumber(str[i]))
			return false;
		return true;
	}
}
int IsSeparater(char c)
{
	for (int i = 0; i < separater.size(); i++)
		if (c == separater[i])
			return i + 1;
	return 0;
}
bool IsOperator(char c)
{
	for (int i = 0; i < operators.size(); i++)
		if (c == operators[i])
			return true;
	return false;

}
int MatchOperator(char c)
{
	for (int i = 0; i < operators.size(); i++)
		if (c == operators[i])
			return i + 1;
	return 0;
}
void LexicalAnalyzer()
{
	char temp_ch=' ';
	for (int i = 0; i < file_content.size();)
	{
		string temp_word = "";
		temp_ch = file_content[i];
		if (IsBlank(temp_ch))
		{
			i++;
			continue;
		}
		if (IsLetter(temp_ch))
		{
			while (IsLetter(temp_ch)||IsNumber(temp_ch)||temp_ch=='_')
			{
				temp_word += temp_ch;
				temp_ch = file_content[++i];
			}
			int index = IsKeyword(temp_word);
			Tuple temp_tuple;
			if (index!=0)
			{
				temp_tuple.val = constant[index - 1];
				temp_tuple.content = "";
				result.push_back(temp_tuple);
			}
			else
			{
				temp_tuple.val = "IDENT";
				temp_tuple.content = temp_word;
				result.push_back(temp_tuple);
			}
			continue;
		}
		if (IsNumber(temp_ch))
		{
			while (IsNumber(temp_ch)||temp_ch=='.'||temp_ch=='E'|temp_ch=='e'||
				(temp_ch=='+'&&(file_content[i-1]=='E'||file_content[i-1]=='e'))||
				(temp_ch=='-'&&(file_content[i - 1] == 'E' || file_content[i - 1] == 'e')))
			{
				temp_word += temp_ch;
				temp_ch = file_content[++i];
			}
			Tuple temp_tuple;
			temp_tuple.content = temp_word;
			if (IsInteger(temp_word))
				temp_tuple.val = "200";
			else
				temp_tuple.val = "300";
			result.push_back(temp_tuple);
			continue;
		}
		if (IsSeparater(temp_ch) != 0)
		{
			Tuple temp_tuple;
			temp_tuple.val = to_string(400 + IsSeparater(temp_ch));
			temp_tuple.content = temp_word;
			result.push_back(temp_tuple);
			i++;
			continue;
		}
		if (IsOperator(temp_ch))
		{
			while (IsOperator(temp_ch))
			{
				temp_word += temp_ch;
				temp_ch = file_content[++i];
			}
			Tuple temp_tuple;
			if (temp_word == "++")
			{
				temp_tuple.val = "500";
				temp_tuple.content = " ";
			}
			if (temp_word == "--")
			{
				temp_tuple.val = "501";
				temp_tuple.content = " ";
			}
			if (temp_word == "&&")
			{
				temp_tuple.val = "502";
				temp_tuple.content = " ";
			}
			if (temp_word == "||")
			{
				temp_tuple.val = "503";
				temp_tuple.content = " ";
			}
			else
			{
				if (temp_word.size() == 1)
				{
					char ch = temp_word[0];
					temp_tuple.val = to_string(503 + MatchOperator(ch));
					temp_tuple.content = " ";
				}
				else
				{
					temp_tuple.val = "UNKNOWN";
					temp_tuple.content = " ";
				}
			}
			result.push_back(temp_tuple);
			continue;
		}
		i++;
	}
}
void PrintResult()
{
	for (int i = 0; i < result.size(); i++)
	{
		cout << "<" << result[i].val << "," << result[i].content <<">"<< endl;
	}
}
int main()
{
	GetFile("test_c.txt");
	LexicalAnalyzer();
	PrintResult();
	system("pause");
	return 0;
}

被读取的text_c.txt文件：

int main()
{
 int i=1;
 char c;
 if(i==1)
{
 i++;
 c='a';
}
 return 0;
}

运行结果：

---

完整程序下载链接：https://download.csdn.net/download/NiZjiTouA/12605353