Remove part of a string with regex

Question

I'm trying to strip part of a string (which happens to be a url) with Regex. I'm getting better out regex but can't figure out how to tell it that content before or after the string is optional. Here is what I have

$string='http://www.example.com/username?refid=22';
$new_string= preg_replace('/[/?refid=0-9]+/', '', $string);
echo $new_string;

I'm trying to remove the ?refid=22 part to get http://www.example.com/username

Ideas?

EDIT I think I need to use Regex instead of explode becuase sometimes the url looks like http://example.com/profile.php?id=9999&refid=22 In this case I also want to remove the refid but not get id=9999

Answer 1

parse_url() is good for parsing URLs :)

$string = 'http://www.example.com/username?refid=22';

$url = parse_url($string);

// Ditch the query.
unset($url['query']);

echo array_shift($url) . '://' . implode($url);

CodePad.

Output

http://www.example.com/username

If you only wanted to remove that specific GET param, do this...

parse_str($url['query'], $get);

unset($get['refid']);

$url['query'] = http_build_query($get);

CodePad.

Output

http://example.com/profile.php?id=9999

If you have the extension, you can rebuild the URL with http_build_url().

Otherwise you can make assumptions about username/password/port and build it yourself.

Update

Just for fun, here is the correction for your regular expression.

preg_replace('/\?refid=\d+\z/', '', $string);

• [] is a character class. You were trying to put a specific order of characters in there.
• \ is the escape character, not /.
• \d is a short version of the character class [0-9].
• I put the last character anchor (\z) there because it appears it will always be at the end of your string. If not, remove it.

Answer 2

Dont use regexs if you dont have to

echo current( explode( '?', $string ) );