正则表达式上缺少奇怪的空白

Question

I have this kind of regex on PHP :

$str="first word https://www.helloz.it last word";
$str=preg_replace(
    '#[^"](((http|https|ftp)://)[^\s
]+)#',
    '<a class="lforum" href="$1">$1</a>',
    $str);
echo nl2br($str);

And the output that I'll expect is :

first word <a class="lforum" href="https://www.helloz.it">https://www.helloz.it</a> last word

but in fact the output is :

first word<a class="lforum" href="https://www.helloz.it">https://www.helloz.it</a> last word

(notice the missing whitespace between first word and <a class...

Where is it vanished that whitespace? :) Thanks

Answer 1

[^"] says "match a character that isn't "". A space character is matched by this, so it is replaced by your regex.

Use a negative lookbehind instead:

'#(?<!")(((http|https|ftp)://)[^\s
]+)#',

This says "match the string if it doesn't follow a quotation mark". Preceding characters are not therefore included in your matched content.

See regular-expressions.info for information about lookbehinds.