How can I display a list of characters that fail to match a regular expression?

Question

For example, if I'm doing some form input validation and I'm using the following code for the name field.

    preg_match("/^[a-zA-Z .-]$/", $firstname);

If someone types in Mr. (Awkward) Double-Barrelled I want to be able to display a message saying Invalid character(s): (, )

Answer 1

You could search your input for ([^a-zA-Z .-]) to get all illegal characters.

Answer 2

You can fetch all occurences of characters that are not within your character-class.
Negate the class [...] -> [^...] and then fetch all matches.

$firstname = 'Mr. (Awkward) Double-Barrelled';

if ( 0 < preg_match_all("/[^a-zA-Z .-]+/", $firstname, $cap) ) {
  foreach( $cap[0] as $e ) {
    echo 'invalid character(s): ', htmlspecialchars($e), "\n";
  }
}

using the PREG_OFFSET_CAPTURE flag described at http://docs.php.net/preg_match_all you can even tell the user where that character is in the input.

edit: Or you can use preg_replace_callback() to visually mark the invalid characters somehow. e.g (using an anonymous function/closure, php 5.3+)

$firstname = 'Mr. (Awkward) Double-Barrelled';
$invalid = array();
$result = preg_replace_callback("/[^a-zA-Z .-]+/", function($c) use(&$invalid) { $invalid[] = $c[0]; return '['.$c[0].']'; }, $firstname);
if ( $firstname!==$result ) {
  echo 'invalid characters: "', join(', ', $invalid), '" in your input: ', $result;
}

prints invalid characters: "(, )" in your input: Mr. [(]Awkward[)] Double-Barrelled

Answer 3

You could split along the allowed characters:

$result = preg_split('/[a-zA-Z .-]+/s', $subject);

...and get a list of all the characters that remain.

Answer 4

preg_match("/[^a-zA-Z0-9\s\.\-]/", $text) should do the trick. You're really supposed to escape the ' ', '.', and '-' characters. I personally wouldn't bother wasting space to figure out which characters are invalid. If the person can't figure it out based on a statement saying 'Allowed Characters: (whatever)' then there is no hope for them.

Here's a list of regex characters which also includes a list of characters you're supposed to escape.

Answer 5

you can also simply replace valid characters with "nothing" - the rest, if any, will be invalid.

 $badchars = preg_replace(/[a-z0-9 .-]/, "", $input);
 if(strlen($badchars))
       error