JSLint reports “Insecure ^” for my regex — what does that mean?

Question

I\'m trying to get my Javascript code 100% JSLint clean.

I\'ve got a regular expression:

 linkRgx = /https?:\\/\\/[^\\s;|\\\\*\'\"!,()<>]+/g;
<         


        

Answer 1

[^\s;|\\*'"!,()<>] matches any ASCII character other than the ones listed, and any non-ASCII character. Since JavaScript strings are Unicode-aware, that means every character known to Unicode. I can see a lot of potential for mischief there.

Rather than disable the warning, I would rewrite the character class to match the characters you do want to allow, as this regex from the Regular Expressions Cookbook does:

/\bhttps?:\/\/[-\w+&@#/%?=~|$!:,.;]*[\w+&@#/%=~|$]/g

Answer 2

(answering my own question) I did some digging... JSLint documentation says:

Disallow insecure . and [^...]. in /RegExp/ regexp: true if . and [^...] should not be allowed in RegExp literals. These forms should not be used when validating in secure applications.

What I have done is disable the JSLint error for the offending line (as I'm not dealing with needing to be secure from potentially malicious user input:

/*jslint regexp: false*/
.... Javascript statement(s) ....
/*jslint regexp: true*/

Answer 3

You should use:

/*jslint regexp: true*/
linkRgx = /https?:\/\/[^\s;|\\*'"!,()<>]+/g;
/*jslint regexp: false*/