Re: How to get WIN-1255 encoded string

On Sun, 2004-11-07 at 13:57 +0200, Marina Markus wrote:
> Hello,

Hi!
 
> I need a PHP script to create an email message in a way that a subject line
> in Hebrew will be readable in all mail clients. Some mail clients cannot
> cope with Hebrew if they don't have character set explicitly denoted. 

I believe that should be _all_ e-mail clients. Since the default MIME
charset is US-ASCII (as per RFC 2047 and 2822), only faulty e-mail
clients would be able to cope with non-ASCII characters without wrapping
the header in MIME.
 
> Is there a possibility in PHP allowing to encode a string
> with WIN-1255 character set encoding? The result should look like:
>  
> =?windows-1255?B?Rlc6IOz26eHl+CDk8uXj6e0=?=
>  
> which the mail client then is able to decode back into Hebrew. 

If you're running PHP5, you might want to look at iconv_mime_encode. If
not, you can take these functions, which are from a GPL:d webmail I'm
writing, and adapt them to your purpose:

function mimifyhdr($header)
{
    if(strpos($header, " ") !== FALSE)
    {
	$temp = "";
	$cp = 0;
	while(($np = strpos($header, " ", $cp)) !== FALSE)
	{
	    $temp .= mimifyhdr(substr($header, $cp, $np - $cp)) . " ";
	    $cp = $np + 1;
	}
	$temp .= mimifyhdr(substr($header, $cp));
	return($temp);
    }
    $nh = "";
    $num = 0;
    for($i = 0; $i < strlen($header); $i++)
    {
	$c = substr($header, $i, 1);
	if(ord($c) >= 128)
	    $num++;
    }
    if($num == 0)
	return($header);
    if($num > (strlen($header) / 4))
	return("=?UTF-8?B?" . encodemime($header, "base64") . "?=");
    $nt = "";
    for($i = 0; $i < strlen($header); $i++)
    {
	$c = substr($header, $i, 1);
	if(($c == "=") || (ord($c) >= 128) || ($c == "_"))
	{
	    $nt .= "=" . strtoupper(dechex(ord($c)));
	} else if($c == " ") {
	    $nt .= "_";
	} else {
	    $nt .= $c;
	}
    }
    return("=?UTF-8?Q?" . $nt . "?=");
}

function addhdr($name, $val)
{
    global $headers;
    
    if(trim($val) != "")
    {
	$temp .= $name . ": " . mimifyhdr($val);
	$maxlen = strlen($temp);
	$ls = 0;
	while($maxlen > 70)
	{
	    $cp = $ls + 70;
	    while(strpos("\t ", substr($temp, $cp, 1)) === FALSE)
	    {
		if(--$cp < $ls)
		    break;
	    }
	    if($cp < $ls)
		break;
	    $temp = substr($temp, 0, $cp) . "\r\n\t" .
		    substr($temp, $cp+ 1);
	    $ls = $cp + 3;
	    $maxlen = strlen($temp) - $ls;
	}
	$headers .= $temp . "\r\n";
    }
}

Note that they take values which are already in the UTF-8 encoding. If
your values aren't in UTF-8, take a look at the iconv function.

> If anyone has a solution for another character set, I suppose it can also
> help.

I'd recommend that you use UTF-8 instead. It's more general than
Windows-1255 (copes with characters outside the Hebrew range), and is
probably supported by more MUAs.

Hope it helps!

Fredrik Tolf

-- 
PHP Internationalization Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php