JavaScript RegExp objects

Question

I try to write a simple Markdown parser in JavaScript. Therefore I want to check for the [link content][link id] syntax. I use the following code:

data = data.replace( /\[(.*?)\][ ]*\[([0-9]+)\]/g, '<a href="$2">$1</a>' );

It works well, but now I want to do this with a RegExp object. So I set up the following bit of code:

var r = new RegExp( '\[(.*?)\][ ]*\[([0-9]+)\]', 'g' );
data = data.replace( r, '<a href="$2">$1</a>' );

But it doesn't work. It even says that my regular expression (which works since the first example does a good job) is invalid:

unmatched ) in regular expression

I think it must have to do with some RegExp-object peculiarities I am not aware of. What am I doing wrong and how can the problem be solved?

On a side note you might find it useful to run your JavaScript regular expressions through an online tester (pagecolumn.com/tool/regtest.htm). — Dscoduc
– Dscoduc, Commented Feb 5, 2009 at 17:27
Thanks, seems to be a great tool. I used this (regexp-evaluator.de/evaluator) PHP-ish evaluator. Maybe a JS-ish is better when scripting in JS :) — okoman
– okoman, Commented Feb 5, 2009 at 17:31
data.replace(/foo/g, 'bar') and data.replace(new RegExp('/foo', 'g'), 'bar') are the same thing — Tivie
– Tivie, Commented Jul 10, 2015 at 20:29

score 13 · Accepted Answer · 2013-11-28 14:45:23Z

13

Because the first argument of the RegExp constructor is a string, not a pattern literal, you have to escape the backslashes, since you want literal backslashes in the pattern:

var r = new RegExp( '\\[(.*?)\\][ ]*\\[([0-9]+)\\]', 'g' );

edited Nov 28, 2013 at 14:45

answered Feb 5, 2009 at 17:25

user21926

Sign up to request clarification or add additional context in comments.

1 Comment

okoman Over a year ago

Uhm.. didn't think about those string issues. Now it's obvious. Thanks :-)

bobince · Accepted Answer · 2009-02-05 18:51:44Z

1

In addition to the pattern's backslash problem, this:

data = data.replace( r, '<a href="$2">$1</a>' );

could be dangerous. I'll assume you've already taken care of the HTML-escaping, so I won't be able to do this:

[<script>stealCookies()</script>][http://oops.example.com/]
[hover me][http://hello" onmouseover="stealCookies()]

but you'll still need to check the URL is a known-good scheme so I can't do this:

[click me][javascript:stealCookies()]

You'll probably want to use the String.replace(r, func) variant of the method, and include validation in your replacement-making 'func'.

answered Feb 5, 2009 at 18:51

bobince

538k111 gold badges675 silver badges846 bronze badges

Comments

chills42 · Accepted Answer · 2009-02-05 17:25:48Z

0

var r = /\[(.*?)\][ ]*\[([0-9]+)\]/g;
data = data.replace( r, '<a href="$2">$1</a>' );

answered Feb 5, 2009 at 17:25

chills42

14.6k4 gold badges45 silver badges78 bronze badges

Comments

Quassnoi · Accepted Answer · 2009-02-05 17:26:01Z

0

Double escape you backslashes:

var r = new RegExp( '\\[(.*?)\\][ ]*\[([0-9]+)\\]', 'g' );

answered Feb 5, 2009 at 17:26

Quassnoi

427k94 gold badges628 silver badges623 bronze badges

Comments

meouw · Accepted Answer · 2009-02-05 17:30:24Z

0

You need to double escape:

var r = new RegExp( '\\[(.*?)\\][ ]*\\[([0-9]+)\\]', 'g' )

answered Feb 5, 2009 at 17:30

meouw

42.1k11 gold badges55 silver badges69 bronze badges

Collectives™ on Stack Overflow

JavaScript RegExp objects

5 Answers 5

1 Comment

Comments

Comments

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

5 Answers 5

1 Comment

Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related