word.cpp

词法分析器

#include<iostream>
#include<string>
#include<vector>
#include<fstream>
using namespace std;
class CIV               //存储的输出结构
{
public:
	int code;           //种别码
	int index;          //变量或常量在变量表或常量表中的位置索引,其他则为-1;
	string value;       //单词值
public:
	CIV(int c,int in,string&v)
	{
		code=c;
		value=v;
		index=in;
	}
	CIV& operator=(const CIV &s)
	{
		code=s.code;
		value=s.value;
		index=s.index;
		return *this;
	}
	CIV(const CIV&s)
	{
		code=s.code;
		value=s.value;
		index=s.index;
	}
		
};
class  word
{
    string* keywords;    
private:
	vector<string> symbols;    //符号表
	vector<CIV> out;           //输出
	vector<string>  consts;    //常量表
	string source;             //读入的源字符串
	int len;                   //长度
	int current;               //当前位置
	char ch;                  //当前字符
	string token;
public:
	word(string&s,int l,string* key_list)
	{
		source=s;
		len=l;
		current=0;
		ch=' ';
		keywords=key_list;
	}
	bool GetChar()
	{
		if(current>=len)
		{
			ch=0;
			return false;
		}
	    ch=source[current++];
		return true;
	}
	bool GetBC()
	{
		while(ch==' ')
		{
			if(!GetChar())
				return false;
		}
		return true;
	}
	void Concat()
	{
    	token+=ch;
	}
	int Reserve()
	{
		for(int i=0;i<8;i++)
			if(token==keywords[i])
				return 1;
	     return 2; 
	}
	bool Retract()
	{
		if(current==0)
		{
			ch=0;
			return false;
		}
		else
		{
			current--;
			ch=0;
		}
	}
    int InsertId()
	{
		symbols.push_back(token); 
		return symbols.size()-1;
	}
	bool IsLetter(char c)
	{
		if((c>='a'&&c<='z')||(c>='A'&&c<='Z'))
			return true;
		return false;
	}
	bool IsDigit(char c)
	{
		if(c>='0'&&c<='9')
			return true;
		return false;
	}
	bool parse();
	void print();
	 
};
bool word::parse()
{
	int code,index;
	string value;
	while(GetChar())
	{
		if(!GetBC()) continue;
		if(IsLetter(ch))
		{
			while(IsLetter(ch)||IsDigit(ch))
			{
				Concat();
				if(!GetChar()) break;
			}
			Retract();
			code=Reserve();
			if(code==2)
			{
				index=InsertId();
				value=token;
				token.erase();
			}
			else
			{
				index=-1;
				value=token;
				token.erase();
			}
			CIV cv(code,index,value);
            out.push_back(cv);  
		}
		else if(IsDigit(ch))
		{
			while(IsDigit(ch))
			{
				Concat();
				if(!GetChar()) break;
			}
     		if(IsLetter(ch))
			{
			  while(IsLetter(ch))
			  {
				Concat();
				if(!GetChar()) break;
			  }
			  cout<<"单词:"<<token<<" error!"<<endl;
			  Retract();
			  token.erase();
			  continue;
			}
 
			Retract();
			consts.push_back(token);
			index=consts.size()-1;
			CIV cv(3,index,token);
			out.push_back(cv);
			token.erase();
		}
		else if(ch=='+'||ch=='-'||ch=='*'||ch=='/')
		{
			char ss[2]={'\0'};
			ss[0]=ch;			 
			CIV cv(4,-1,string(ss));
			out.push_back(cv);
		}else if(ch=='>'||ch=='<'||ch=='='||ch=='!')
		{
			char ss[3]={'\0'};
			ss[0]=ch;
			GetChar();
			if(ch=='=')
		 		ss[1]='=';
			else
				Retract();
			string s(ss);
			CIV cv(4,-1,s);
			out.push_back(cv);
		}
		else if(ch==','||ch==';'||ch=='{'||ch=='}'||ch=='('||ch==')')
		{
			char ss[2]={'\0'};
			ss[0]=ch;
			string s(ss);
			CIV cv(5,-1,s);
			out.push_back(cv);
		}
	
	}
	return true;
}
void word::print()
{
	vector<CIV>::iterator it;
	for(it=out.begin();it!=out.end();it++)
		 cout<<"     ("<<(*it).code<<",\""<<(*it).value<<"\")"<<endl;
}
int main()
{
	string x[8]={"if","int","for","while","do","return","break","continue"};
	word* pw=NULL;
    FILE*fp;
	string filename;
	cout<<"请输入源文件名(绝对路径):";
	cin>>filename;
    

 	string source;
	fp=fopen(filename.c_str(),"r");
    if(fp==NULL)
	{
		cout<<"不能打开文件 "<<filename<<endl;
		exit(0);
	}
	char buff[1024];
	while(!feof(fp))
	{
		fgets(buff,1024,fp);
		string ss(buff);
		source+=ss;
	}
	cout<<"源文件 "<<filename<<"内容为:"<<endl;
	cout<<source<<endl;
    pw=new word(source,source.size(),x);
	pw->parse();
	cout<<"输出结果为:"<<endl;
	pw->print();
 	
	return 1;
}