Hi all,

 I'm trying to develop an expression that can determine if a statement is in a valid syntax.  The statement is contained within nested parens, and I need to see if the expressions not only is balanced (solved that), but valid.

The following is my expression to see if the statement is balanced, but it lacks seeing if it is valid. 

/^(\((?:[^()]+|(?1))*\))$/ 

Statements can be of the following 2 examples:

(1 = 1)

or

(1 => (3 = 3) || ((3 < (3 = 4)) && ((1 < 3) = 2)))

or any breakdown there in where it follows the pattern of 

([number | nested boolean expression] [= | => | <= | < | > | && | ||]  [number | nested boolean BLOCKED EXPRESSION

So far I've come up with the following, but they are not working right.

 15 define('RULE_OPERATOR', '=|==|>|>=|<|<=|!=');

 16 //define('RULE_VALID', '/^(\((?:[^(?:'.RULE_OPERATOR.')]*(?:'.RULE_OPERATOR.')[^(?:'.RULE_OPERATOR.')]*\))|(?1))*$/');

 17 //define('RULE_VALID', '/^(\((?:[^\((?:'.RULE_OPERATOR.')]*\s[(?:'.RULE_OPERATOR.')]\s[^)(?:'.RULE_OPERATOR.')]*|(?1))\))$/');

I'm writing in PHP, so it's a PCRE based expression engine.

Thanks for any advice or help.

--

Dave 

You might consider just taking the easy way out and using PHP's eval function to test this:

<pre>
<?php
function check($str){
        if(!preg_match('/^(\((?:[^()]+|(?1))*\))$/',$str)) return ("unbalanced.");
        if(!preg_match('/^[0-9|\s=<>()!&]+$/',$str)) return ("illegal characters found.");
        $return=eval("return true; if $str { }"); // doesn't execute the if $str code yet, just checks for syntax errors
        if(!$return) return ("syntax error.");
        return eval("return $str ? true : false;"); // no syntax error found, executes the code to evaluate result
}
$str='($test==1)'; // illegal characters found
var_dump(check($str));
$str='(1==1 || (2==)'; // unbalanced
var_dump(check($str));
$str='(1 => (3 = 3) || ((3 < (3 = 4)) && ((1 < 3) = 2)))'; // syntax errors, => should be >= and = should be ==
var_dump(check($str));
$str='(1 >= (3 == 3) || ((3 < (3 == 4)) && ((1 < 3) == 2)))'; // no syntax errors, evaluates true since 1 >= 1
var_dump(check($str));
$str='(1 >= 2)'; // no syntax errors and result is false
var_dump(check($str));
?>

Output:

string(25) "illegal characters found."
string(11) "unbalanced."
string(13) "syntax error."
bool(true)
bool(false)

dup post removed.

Great, that is a step in the right direction!

Sadly it doesn't account for the following 2 strings:

$str1 = "(1==1==1)";

$str2 = "(3 >= (4 == exit()))";