Cleaning User Input With Regex

[kvarnerexpress 0]

As anyone who works with user input knows, not everyone who submits information makes it look proper.

One one of my web forms, I parse all the needed fields that I wish to be title cased; their name, address, city, etc.

I use this to perform this action, which works nicely:

PHP Code:

$string = ucwords(strtolower($string));

This fixes the input if the user types in all caps or all lowercase. There is a problem I have been noticing about those who enter a generation after their name, such as "Bill Warner III".

The above function changes it to "Bill Warner Iii", which is incorrect. I have been toying with a regular expression to catch these I's (or sometimes V's) and turn them all uppercase, but to no avail. Here are versions of my non-working code:

PHP Code:

$string = preg_replace("/\b([IiVv])+\b/e", strtoupper("$1"), $string);$string = preg_replace("/\b([IiVv]+)\b/e", strtoupper("$1"), $string);$string = preg_replace("/\b([IiVv]+)$/e", strtoupper("$1"), $string)

Thanks,kivarnerexpress

[Spectre 0]

This pattern should match almost any Roman numeral below L/50 (it is unlikely you will ever encounter anything beyond, say, 10th/X generation names).

 

/\b(i+|i+[vx]*?|v[i]*?|x+[vi]*?)\b/ei

Here's a quick example (the names are, of course, randomly chosen). Note that the 'strtoupper()' function is being evaluated as PHP code and is enclosed in quotes.

 

<?php$names = array(   'john sMith iv',   'Richard jones Iiv',   'JOE NOBODY II',   'Viva vivi ixen ivan III',   'Person XI',   'person xv');$pattern = '/\b(i+|i+[vx]*?|v[i]*?|x+[vi]*?)\b/ei';for( $i=0;$i<count($names);$i++ ) {   $name = ucwords(strtolower($names[$i]));   $string = preg_replace($pattern, 'strtoupper("$1")', $name);   $names_p[$names[$i]] = $string;}print_r($names_p);?>

Outputs:

Array

(

    [john sMith iv] => John Smith IV

    [Richard jones Iiv] => Richard Jones IIV

    [JOE NOBODY II] => Joe Nobody II

    [Viva vivi ixen ivan III] => Viva Vivi Ixen Ivan III

    [Person XI] => Person XI

    [person xv] => Person XV

)