Link to home
Start Free TrialLog in
Avatar of ChristoferDutz
ChristoferDutzFlag for Germany

asked on

ANTLR Grammar Problem

Hi,

I am currently trying to create a formula-evaluator using an ANTLR generated Java Parser. In oder to get a kick-start, I used a grammar from this site http://arcanecoder.blogspot.com/2008/04/using-antlr-to-create-excel-like.html
At first it seemed to work like a charm. Until one of my test-users created a false expression. In the Grammar-File I am using, the Or operation is represented by "||" a user entered a forumual using "OR" instead. Unfortunately my parser doesn't fire an exception but simply quits after parsing the first part of the formula. As an example

TRUE OR FALSE evaluates to TRUE
FALSE OR TRUE evaluates to FALSE

How can I change the grammar not to quit but to identify the formula as invalid and to fire an exception?


grammar SecurityRules;

options {
        backtrack = true;
        output=AST;
	ASTLabelType=CommonTree;
        k=2;
}

tokens {
	POS;
	NEG;
	CALL;
}

@header {package de.cware.cweb.security.service.parser;}
@lexer::header{package de.cware.cweb.security.service.parser;}

formula
	: (EQ!)? expression
	;

//The highest precedence expression is the most deeply nested
//Precedence ties are parsed left to right
//Expression starts with the lowest precedece rule
expression		
	: boolExpr
	;
boolExpr
	: concatExpr ((AND | OR | LT | LTEQ | GT | GTEQ | EQ | NOTEQ)^ concatExpr)*
	;
concatExpr
	: sumExpr (CONCAT^ sumExpr)*
	;
sumExpr
	: productExpr ((SUB | ADD)^ productExpr)*
	;
productExpr
	: expExpr ((DIV | MULT)^ expExpr)*
	;
expExpr
	: unaryOperation (EXP^ unaryOperation)*
	;
unaryOperation
	: NOT^ operand
	| ADD o=operand -> ^(POS $o)
	| SUB o=operand -> ^(NEG $o)
	| operand
	;
// the highest precedence rule uses operand
operand
	: literal 
	| functionExpr -> ^(CALL functionExpr)
	| percent
	| VARIABLE
	| LPAREN expression RPAREN -> ^(expression)
	;
functionExpr
	: FUNCNAME LPAREN! (expression (COMMA! expression)*)? RPAREN!
	;
literal
	: NUMBER 
	| STRING 
	| TRUE
	| FALSE
	;
percent
	: NUMBER PERCENT^
	;

STRING
	:
	'\"'
		( options {greedy=false;}
		: ESCAPE_SEQUENCE
		| ~'\\'
		)*
	'\"'
	|
	'\''
		( options {greedy=false;}
		: ESCAPE_SEQUENCE
		| ~'\\'
		)*
	'\''
	;
WHITESPACE
	: (' ' | '\n' | '\t' | '\r')+ {skip();};
TRUE
	: ('t'|'T')('r'|'R')('u'|'U')('e'|'E')
	;
FALSE
	: ('f'|'F')('a'|'A')('l'|'L')('s'|'S')('e'|'E')
	;
	
NOTEQ           : '!=';
LTEQ            : '<=';
GTEQ            : '>=';
AND		: '&&';
OR		: '||';
NOT		: '!';
EQ              : '=';
LT              : '<';
GT              : '>';

EXP             : '^';
MULT            : '*';
DIV             : '/';
ADD             : '+';
SUB             : '-';

CONCAT          : '&';

LPAREN          : '(';
RPAREN          : ')';
COMMA           : ',';
PERCENT         : '%';

VARIABLE
	: '[' ~('[' | ']')+ ']'
	;
FUNCNAME
	: (LETTER)+
	;
NUMBER
	: (DIGIT)+ ('.' (DIGIT)+)?
	;

fragment
LETTER 
	: ('a'..'z') | ('A'..'Z')
	;
fragment
DIGIT
	: ('0'..'9')
	;
fragment
ESCAPE_SEQUENCE
	: '\\' 't'
	| '\\' 'n'
	| '\\' '\"'
	| '\\' '\''
	| '\\' '\\'
	;

Open in new window

Avatar of arch-itect
arch-itect

Have you noticed that both the TRUE and FALSE constructs contain ('e'|'E')

If that is not the only problem, could you post their test formula?
ROFL, never mind
ASKER CERTIFIED SOLUTION
Avatar of ChristoferDutz
ChristoferDutz
Flag of Germany image

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial