CODEWITHSUNDEEP
javascriptregex
CdB
CdB

Reputation: 4908

javascript regex - global replace issue

I'm getting content from a textarea (which besides simple text may include html markup), and try to parse it and replace all occurences of "[File:#xx#]" with a string contained in an array. So, lets say the contents of the textarea is in var html.
I do the following:

html = html.replace(/\[File:#(.*)#\]/g, 
                    function($0, $1){ return furls[$1]; });

everything works fine when the contents the text area are like this: 

<img src="[File:#111#]" alt="image1" />
<img src="[File:#222#]" />

but when there is no line break between 2 elements witch have attribute with [File:#xx#] value, then the problem appears.
So, given this as the textarea's value: 

<img src="[File:#111#]" alt="image1" /><img src="[File:#222#]" />

seems like it matces the first img's [File:#111# but closes it not with the first bracket, rathen than the second one's. so, what gets replaced is all this:
#]" alt="image1" /><img src="[File:#222

What is wrong with my regular expression? How can i prevent this look-ahead from happening and stop at the first closing bracket?

Thanks in advance.

Upvotes: 1

Views: 278

Answers (2)

Russell G
Russell G

Reputation: 478

The problem is that it's grabbing everything from the first # sign to the last one, because you're using (.*), which matches all characters. Try this instead, which limits the matched part to just numeric digits:

html = html.replace(/\[File:#([0-9]*)#\]/g, 
                function($0, $1){ return furls[$1]; });

Upvotes: 1

Bogdan Emil Mariesan
Bogdan Emil Mariesan

Reputation: 5647

Well the correct regex for your case would be:

/[File:#[\w]+#]/g

Why is this the case?

Because in your regex:

The . Matches any character, except for line breaks if dotall is false.

The * Matches 0 or more of the preceeding token. This is a greedy match, and will match as many characters as possible before satisfying the next token.

And in the regex i've provided:

The \w Matches any word character (alphanumeric & underscore).

Upvotes: 1

Related Questions