使用flex和bison实现的sql引擎解析(一)

2014-11-24 17:10:20 · 作者: · 浏览: 0

由于老师要求,最近在做oceanbase存储过程的实现,在oceanbase 0.4以前是不支持存储过程的。实现的主要步骤主要包括

1、语法解析

2、词法解析

3、具体执行语法树的步骤

现在先来说说语法解析吧,在这一块主要是使用的flex( 词法分析器生成工具) 和bison(语法分析器生成器) 这两个是对用户输入的存储过程语句进行解析的

\

来具体说说该怎么实现对sql语句的分析吧

1、首先建立一个lex的文件

%option noyywrap nodefault yylineno case-insensitive

%{

#include "prosql.tab.hpp"
#include 
#include 
#include 
#include 
#include 
//YYSTYPE yylval;
int oldstate;
extern "C" int yylex();
//extern "C" int yyparse();
extern "C" void yyerror(const char *s, ...);
extern char globalInputText[10000];
extern int readInputForLexer( char *buffer, int *numBytesRead, int maxBytesToRead );
#undef YY_INPUT
#define YY_INPUT(b,r,s) readInputForLexer(b,&r,s)
%}

%x COMMENT

%%

CREATE		{ return CREATE; }
PROCEDURE	{ return PROCEDURE; }
SQL		{ return SQL; }

DECLARE		{ return DECLARE; }
SET		{ return SET; }
BEGIN		{ return BEGINT; }
END		{ return END; }

INT		{ return INT; }
VARCHAR		{ return VARCHAR; }
DATE		{ return DATE; }
TIME		{ return TIME; }
DOUBLE		{ return DOUBLE; }

IF		{ return IF; }
THEN		{ return THEN; }
ELSE		{ return ELSE; }
ENDIF		{ return ENDIF; }
FOR		{ return FOR; }
WHEN		{ return WHEN; }
WHILE		{ return WHILE; }


[0-9]+	{ yylval.strval = strdup(yytext);/*printf("number=%s\n",yylval.strval);*/ return INTNUM; }/*number*/

[0-9]+"."[0-9]* |
"."[0-9]+	|
[0-9]+E[-+] [0-9]+	|
[0-9]+"."[0-9]*E[-+] [0-9]+ |
"."[0-9]*E[-+] [0-9]+	{ yylval.strval = strdup(yytext);/*printf("float=%s\n",yylval.strval);*/ return APPROXNUM; }/*double*/

TRUE	{ yylval.strval = "1";/*printf("bool=%s\n",yylval.strval);*/ return BOOL; }/*bool*/

FALSE	{ yylval.strval = "0";/*printf("bool=%s\n",yylval.strval);*/ return BOOL; }/*bool*/

'(\\.|''|[^'\n])*'	|
\"(\\.|\"\"|[^"\n])*\"  {
				char *temp = strdup(yytext); 
				yylval.strval = strdup(yytext);

				//GetCorrectString(yylval.strval, temp);
				
				/*printf("string=%s\n",yylval.strval);*/
				return STRING;
			}/*string*/
'(\\.|[^'\n])*$		{ yyerror("Unterminated string %s", yytext); }
\"(\\.|[^"\n])*$		{ yyerror("Unterminated string %s", yytext); }


X'[0-9A-F]+' |  
0X[0-9A-F]+  	{ yylval.strval = strdup(yytext); return STRING; }


0B[01]+      |
B'[01]+'     { yylval.strval = strdup(yytext); return STRING; }


[-+&~|^/%*(),.;!]   { return yytext[0]; }

"&&"	{ return ANDOP; }
"||"	{ return OR; }

"<"	{ yylval.subtok = 1; return COMPARISON; }
">"	{ yylval.subtok = 2; return COMPARISON; }
"!="	|
"<>"	{ yylval.subtok = 3; return COMPARISON; }
"="	{ yylval.subtok = 4; return COMPARISON; }
"<="	{ yylval.subtok = 5; return COMPARISON; }
">="	{ yylval.subtok = 6; return COMPARISON; }
"<=>"	{ yylval.subtok = 12; return COMPARISON; }

"<<"	{ yylval.subtok = 1; return SHIFT; }
">>"	{ yylval.subtok = 2; return SHIFT; }


[A-Za-z][A-Za-z0-9_]*	{ yylval.strval = strdup(yytext);
			  /*printf("name 1=%s\n",yylval.strval);*/
                          return NAME; }
`[^`/\\.\n]+`           { yylval.strval = strdup(yytext+1);
			  /*printf("name 2=%s\n",yylval.strval);*/
                          yylval.strval[yyleng-2] = 0;
                          return NAME; }

`[^`\n]*$               { yyerror("unterminated quoted name %s", yytext); }


@[0-9a-z_.$]+ |
@\"[^"\n]+\" |
@`[^`\n]+` |
@'[^'\n]+' { yylval.strval = strdup(yytext+1);  return USERVAR; }

@\"[^"\n]*$ { yyerror("unterminated quoted user variable %s", yytext); }
@`[^`\n]*$ { yyerror("unterminated quoted user variable %s", yytext); }
@'[^'\n]*$ { yyerror("unterminat