C语言词法分析器-优快云博客

保留字表：

关键字	种别码	关键字	种别码
main	0	return	32
auto	1	变量	33
short	2	int常量	34
int	3	实型常量	35
long	4	char常量	36
float	5	=	37
double	6	+	38
char	7	-	39
struct	8	*	40
union	9	/	41
enum	10	++	42
typedef	11	--	43
const	12	+=	44
usigned	13	-=	45
signed	14	*=	46
extern	15	/=	47
register	16	==	48
static	17	!=	49
volatile	18	>	50
void	19	<	51
if	20	>=	52
else	21	<=	53
switch	22	(	54
case	23	)	55
for	24	[	56
do	25	]	57
while	26	{	58
goto	27	}	59
continue	28	,	60
break	29	:	61
default	30	;	62
sizeof	31

源代码：

#include <iostream>

#include<math.h>

#include<string>

#include<stdlib.h>

#include<stdio.h>

#include<fstream>

using namespace std;

string key[34]={"main",

"auto","short","int","long","float","double","char","struct"

,"union","enum","typedef","const","unsigned","signed","extern","register"

,"static","volatile","void","if","else","switch","case","for"

,"do","while","goto","continue","break","default","sizeof","return"};

string symbol[26]={

"=","+","-","*","/","++","--","+=","-=",

"*=","/=","==","!=",">","<",">=","<=","(",

")","[","]","{","}",",",":",";" };

char ch; ///存放最新读进的源程序字符

string strToken; ///存放构成单词的字符串

string text="int main()\n{\n int a,b;\n a=10.3211;\n a+=20; if (b==15) break; \n float fa=3.141592653; \n char s='z';\n}"; ///要读进的文本

//string text="main(){ int a,b; a=10; b=a+20;}"; ///要读进的文本

int pText=0; ///搜索指示器

int line=1;

int FindInSymbol(string str)

{

for(int i=0;i<26;i++)

{

if(str==symbol[i]) return i+37;

}

return -1;

}

void GetChar() ///---------1

{

ch=text[pText];

pText++;

}

void GetBC() ///-------2

{

while(1)

{

if(ch==' ') GetChar();

else break;

}

void Concat() ///----------3

{

strToken+=ch;

}

bool IsLetter() ///---------4

{

if( ( ch>='a' &&ch<='z') ||( ch>='A' &&ch<='Z') )

return true;

return false;

}

bool IsDigit() ///---------5

{

if( ch>='0' &&ch<='9')

return true;

return false;

}

int Reserve() ///---------6

{

for(int i=0;i<=33;i++)

{

if(strToken==key[i]) return i;

}

return -1;

}

void Retract() ///----------7

{

ch=' ';

pText--;

}

void Analyse()

{

int code;

strToken="";

GetChar();

GetBC();

if(IsLetter()) ///识别标示符和关键字

{

while(IsLetter() || IsDigit() ||ch=='_')

{

Concat();

GetChar();

}

Retract();

code=Reserve();

if(code!=-1) ///识别出来的是常量

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

else ///识别出来的是标示符

cout<<"("<<33<<","<<strToken<<",line="<<line<<")"<<endl;

}

else if(IsDigit()) ///识别整形常量、浮点型常量

{

while(IsDigit())

{

Concat();

GetChar();

}

if(ch!='.') ///识别整形常量

{

Retract();

int temp=atoi(strToken.c_str());

///---------------------二进制显示----------------

int a[128],i;

for(i=0; ;i++)

{

if(temp<=0) break;

a[i]=temp%2;

temp/=2;

}

cout<<"("<<34<<",";

for(int j=i-1;j>=0;j--)

{

cout<<a[j];

}

cout<<",line"<<line<<")"<<endl;

}

else if(ch=='.') ///识别实型常量

{

Concat();

GetChar();

while(IsDigit())

{

Concat();

GetChar();

}

Retract();

cout<<"("<<35<<","<<strToken<<",line="<<line<<")"<<endl;

}

else if(ch==39) ///识别字符常量

{

Concat();

GetChar();

while(ch!=39)

{

Concat();

GetChar();

}

cout<<"("<<36<<","<<strToken<<"',line="<<line<<")"<<endl;

}

else if(ch=='=') ///识别=、==

{

Concat();

GetChar();

int code;

if(ch=='=')

{

Concat();

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

}

else

{

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

Retract();

}

else if(ch=='+') ///识别+、+=、++

{

Concat();

GetChar();

int code;

if(ch=='=')

{

Concat();

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

}

else if(ch=='+')

{

Concat();

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

}

else

{

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

Retract();

}

else if(ch=='-') ///识别-、-=、--

{

Concat();

GetChar();

int code;

if(ch=='=')

{

Concat();

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

}

else if(ch=='-')

{

Concat();

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

}

else

{

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

Retract();

}

else if(ch=='*') ///识别*、*=

{

Concat();

GetChar();

int code;

if(ch=='=')

{

Concat();

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

}

else

{

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

Retract();

}

else if(ch=='/') ///识别/、/=

{

Concat();

GetChar();

int code;

if(ch=='=')

{

Concat();

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

}

else

{

code=FindInSymbol(strToken);

cout<<"("<<code<<","<<strToken<<",line="<<line<<")"<<endl;

Retract();

}

else if(ch==';') cout<<"("<<34<<",;"<<",line="<<line<<")"<<endl;

else if(ch=='(') cout<<"("<<26<<",("<<",line="<<line<<")"<<endl;

else if(ch==')') cout<<"("<<27<<",)"<<",line="<<line<<")"<<endl;

else if(ch=='{') cout<<"("<<30<<",{"<<",line="<<line<<")"<<endl;

else if(ch=='}') cout<<"("<<31<<",}"<<",line="<<line<<")"<<endl;

else if(ch==',') cout<<"("<<32<<",,"<<",line="<<line<<")"<<endl;

else if(ch=='\n') line++;

}

void choice()

{

cout<<"\t\t---------从文件中读------>1----"<<endl;

cout<<"\t\t---------默认输入------>2----"<<endl;

cout<<"\t\t---------退出------>0----"<<endl;

int choice;

while(1)

{

cin>>choice;

if(choice==1)

{

fstream data;

data.open("D:\data.txt");

text.clear();

char temp;

while(1)

{

if(data.eof()) break;

data.get(temp);

text+=temp;

}

cout<<"要分析的C程序内容："<<endl;

cout<<text<<endl;

break;

}

else if(choice==2)

{

cout<<"要分析的C程序内容："<<endl;

cout<<text<<endl;

break;

}

else if(choice==0)

exit(0);

}

int main()

{

choice();

while(pText<text.length())

{

Analyse();

}

cout<<strToken;

return 0;

}