[SOLVED] How to strip our Accent Characters from text string?

[fatmikey]

Hi, I have a website where people signup using a form and sometimes they have a name that contains a French accented character in their name.  Unfortunately, when the PHP script tries to process the record the name gets all screwed up and doesn't display properly.

 

Is there an easy script or function to strip our French accented characters from a string?

 

Thanks for your help,

Mikey

[jonsjava]

<?php
function transcribe($string) {
    $string = strtr($string,
       "\xA1\xAA\xBA\xBF\xC0\xC1\xC2\xC3\xC5\xC7
        \xC8\xC9\xCA\xCB\xCC\xCD\xCE\xCF\xD0\xD1
        \xD2\xD3\xD4\xD5\xD8\xD9\xDA\xDB\xDD\xE0
        \xE1\xE2\xE3\xE5\xE7\xE8\xE9\xEA\xEB\xEC
        \xED\xEE\xEF\xF0\xF1\xF2\xF3\xF4\xF5\xF8
        \xF9\xFA\xFB\xFD\xFF",
        "!ao?AAAAAC
        EEEEIIIIDN
        OOOOOUUUYa
        aaaaceeeei
        iiidnooooo
        uuuyy");  
    $string = strtr($string, array("\xC4"=>"Ae", "\xC6"=>"AE", "\xD6"=>"Oe", "\xDC"=>"Ue", "\xDE"=>"TH", "\xDF"=>"ss", "\xE4"=>"ae", "\xE6"=>"ae", "\xF6"=>"oe", "\xFC"=>"ue", "\xFE"=>"th"));
    return($string);
}
//usage:
$data = "Àorvar";
print transcribe($data);

[JonnoTheDev]

You shouldn't try to strip them out. You should be using the correct character encoding for your DB storage, php, HTML document. UTF-8 Unicode should work. Check your HTML headers:

 

<meta http-equiv="Content-Type" content="text/html; charset=utf-8">

 

If that doesnt work use php (place at the top of each script preferably in a common include prior to any screen output).

 

header('Content-Type: text/html; charset=UTF-8');

[.josh]

I agree with the "don't just strip them out".  If you are wanting to stick with the "get rid of them" instead of "make sure they display" route, I would give the user the option to enter in something of the "correct" format, rather than you "stripping" them out.  Check if they are there, tell the user he can't use them, re-enter name, sort of thing.

 

if(!preg_match('~^[a-z ]+$~i',$name)) {
  // name has more than a-z, A-Z or space, give error
}

 

 

[Daniel0]

*snip*

 

Ø and Å aren't O and A, but OE and AA. Same goes for their lower case variants.

 

There is another problem with your function. If you for instance have the word "Æble" (Apple in Danish) it would be transliterated into Aeble. However, if it's in all caps, ÆBLE, then it would be AEBLE. Your function doesn't take that into account. You would have to figure the case of the other characters out as well. Otherwise you could end up with something like AeBLE or AEble, but of which look stupid. The same goes for all the other letters that stand for more than one letter.

 

Niel and CV, it could be the case that he needs it in a URL (e.g. /user/Daniel). In that case he might only want the letters A-Z without any sort of diacritics.

[.josh]

well that's why I opt for the opinion to tell the user that non <insert good stuff here>'s are not allowed, and let the user pick an alternative.

[Daniel0]

Well, that's easy to say for someone whose primary language is English. English doesn't really use other characters than a through z. Many other languages use various diacritics to give different meanings. Compare these words in Spanish for instance: año (year) vs. ano (anus), papá (dad) vs. papa (potato (or pope)).

[fatmikey]

Thanks for your help everyone!  Really appreciate it.