Question

I have this regex to filter out urls but its also filtering out some invalid urls

    $regexUrl = "((https?|ftp)\:\/\/)?"; // SCHEME 
    $regexUrl .= "([a-zA-Z0-9+!*(),;?&=\$_.-]+(\:[a-zA-Z0-9+!*(),;?&=\$_.-]+)?@)?"; // User and Pass 
    $regexUrl .= "([a-zA-Z0-9-.]*)\.([a-zA-Z]{2,3})"; // Host or IP 
    $regexUrl .= "(\:[0-9]{2,5})?"; // Port 
    $regexUrl .= "(\/([a-zA-Z0-9+\$_-]\.?)+)*\/?"; // Path 
    $regexUrl .= "(\?[a-zA-Z+&\$_.-][a-zA-Z0-9;:@&%=+\/\$_.-]*)?"; // GET Query 
    $regexUrl .= "(#[a-zA-Z_.-][a-zA-Z0-9+\$_.-]*)?"; // Anchor

for instance “http://…XYZ” is also filtered by the above regex but this is invalid url.

Any help would be appreciated

Answer 1

In your Host or IP line, change * to + and remove the . from the first []

$regexUrl .= "([a-zA-Z0-9-]+)\.([a-zA-Z]{2,3})"; // Host or IP 

The effect of this is to require (with +) some characters from the first [] and not permit a . among them since the . is handled (and required) by the \. which follows the first group.