无效的正则表达式Javascript

Question

I'm trying to parse CSS file with this expression:

rulePatt = new RegExp("(?<=\\s)(?<!\\:\\s)#.*?\\{.*?\\}","gm");

It works in two different regex testers I used, but when I put it in the wild it generates javascript error:

Uncaught SyntaxError: Invalid regular expression: /(?<=\s)(?<!\:\s)#.*?\{.*?\}/: Invalid group

Fiddle here to form a impression on what I'm trying to achieve: http://jsfiddle.net/LoomyBear/44XtU/1/

I guess I miss something really important here. Please help ... Thanks!

Answer 1

You can use this pattern

var regex = /(?:^|[^:])\s(#[^{]*\{[^}]*\})/gm;

Your result is in the first capturing group.

explanations:

(?:^|[^:])   # begining of the line or a character that is not a :
\s           # a white space
(#
[^{]*        # all characters expect { zero or more times
\{
[^}]*        # all characters expect } zero or more times
\})

The first lookbehind (?<=\\s) imposes a white space before the sharp, it s the reason why i wrote a \\s before. The second lookbehind (?<!:\\s) forbids a : before this space, thus, at this position is only allowed nothing (ie: the begining of the line) or a character that is not a :

Answer 2

You can mimic look-behinds in JavaScript by using memory groups, eg:

var patt1 = /(?:\}|$)\s*(#[^\{]+)/g,
    m;

while ((m = patt1.exec(parseStr)) !== null) {
    output += m[1] + '<br>';
}

Demo

First you match either the start of the subject or the closing brace of the previous definition, followed by optional white space. The rest is more or less your original expression.

To iterate over each match you must use RegExp.exec() instead of $.each() so that you can choose which memory group to add to output .