I want to find every url in text and wrap them with a tag (<a href="...">...</a>).
var src = "bla blaaa blaaaaaa 1 http://yahoo.com text something about and. http://www.google.com";
var match = /http:\/\/([a-z0-9.-]+)/.exec(src); //this only can one matched
// result: ["http://yahoo.com", "yahoo.com"]
But i need to wrap every links.
You can use /g (global) to match all occurences and the backreference like this:
var src = "bla blaaa blaaaaaa 1 http://yahoo.com text something about and. http://www.google.com";
var match = src.replace(/http:\/\/([a-z0-9.-]+)/g, '<a href="$1">$1</a>');
/g explanation.var res = src.replace(/(http:\/\/([a-z0-9.-]+))/g, '<a href="$1">$2</a>');
Outputs:
bla blaaa blaaaaaa 1 <a href="http://yahoo.com">yahoo.com</a> text something about and. <a href="http://www.google.com">www.google.com</a>
Not sure that was the intention, but what I could think of. Use <a href="$1">$1</a> as replacement if you want to preserve the http:// prefix in the link text, too.
(In the meantime Nick Craver provided the answer and introduced the g modifier.)
\0 isn't a backreference, you can test it here: jsfiddle.net/nick_craver/9beM2
<a href="$1">$2</a> i think it will use first link on second address like <a href="yahoo.com">google.com</a> and all these links don't open the real page because of they don't have http:// do they ?!?$1) is the first parenthesis (I added those to the pattern), the second ($2) is without the http:// prefix. BTW <a href="www.google.com"> is not a link to google.com but to the relative resource www.google.com!