As an alternative for  qeremy [atta] gmail [dotta] com
There is much shorter way for binarysafe chunking of multibyte string:

<?php
function word_chunk($str, $len = 76, $end = "\n") {
    $pattern = '~.{1,' . $len . '}~u'; // like "~.{1,76}~u"
    $str = preg_replace($pattern, '$0' . $end, $str);
    return rtrim($str, $end);
}

$str = 'русский';
echo chunk_split($str, 3) ."\n";
echo word_chunk($str, 3) . "\n";
?>

р�
�с
с�
�и
й

рус
ски
й

An alternative for unicode strings;

<?php
function chunk_split_unicode($str, $l = 76, $e = "\r\n") {
    $tmp = array_chunk(
        preg_split("//u", $str, -1, PREG_SPLIT_NO_EMPTY), $l);
    $str = "";
    foreach ($tmp as $t) {
        $str .= join("", $t) . $e;
    }
    return $str;
}

$str = "Yarım kilo çay, yarım kilo şeker";
echo chunk_split($str, 4) ."\n";
echo chunk_split_unicode($str, 4);
?>

Yar�
�m k
ilo 
çay
, ya
rım
 kil
o ş
eker

Yarı
m ki
lo ç
ay, 
yarı
m ki
lo ş
eker

I'm not sure what versions this also occurs in but the output of chunk_split() in PHP 5.0.4 does not match the output in other versions of PHP.

In all versions of PHP I have used, apart from 5.0.4 chunk_split() adds the separator (\r\n) to the end of the string.  But in PHP 5.0.4 this does not happen.  This had a fairly serious impact on a library I maintain so it may also affect others who are not aware of this.

The description of the function is slightly inaccurate. A trailing $end is also added.

the best way to solve the problem with the last string added by chunk_split() is:

<?php
$string = '1234';
substr(chunk_split($string, 2, ':'), 0, -1);
// will return 12:34
?>

Important note is the maximum line length and the recommended one. The standard says:
"Lines in a message MUST be a maximum of 998 characters excluding the CRLF, but it is RECOMMENDED that lines be limited to 78 characters excluding the CRLF. "

See PHP manual for chunk_split() Which is set to 76 characters long chunk and "\r\n" at the end of line by default.

This function is very simple and many other functions make this on PHP 5 and even some ones in 4 the good think about this one is that work on php 3.0.6 and 4 

function split_hjms_chars($xstr, $xlenint, $xlaststr)
{
    $texttoshow = chunk_split($xstr,$xlenint,"\r\n"); 
    $texttoshow  = split("\r\n",$texttoshow);
    $texttoshow = $texttoshow[0].$xlaststr;
    return $texttoshow;
}

// For use 

echo split_hjms_chars("This is your text",6,"...");

// Will return

This i...

It is useful to cut long text on preview lists and if the server it's old.

Hope it helps some one. Hans Svane

"version" of chunk_split for cyrillic characters in UTF-8

public function UTFChunk($Text,$Len = 10,$End = "\r\n")
{
    if(mb_detect_encoding($Text) == "UTF-8")
    {
        return mb_convert_encoding(
                chunk_split(
                    mb_convert_encoding($Text, "KOI8-R","UTF-8"), $Len,$End
                ),
                "UTF-8", "KOI8-R"
            );
    } else
    {
        return chunk_split($Text,$Len,$End);
    }
}

this is example for russian language

Here's a version of Chunk Split I wrote that will not split html entities. Useful if you need to inject something in html (in my case, <wbr/> tags to allow for long text wrapping).

<?php
function HtmlEntitySafeSplit($html,$size,$delim)
{
  $pos=0;
  for($i=0;$i<strlen($html);$i++)
  {
    if($pos >= $size && !$unsafe)
    {
      $out.=$delim;
      $unsafe=0;
      $pos=0;
    }
    $c=substr($html,$i,1);
    if($c == "&")
      $unsafe=1;
    elseif($c == ";")
      $unsafe=0;
    $out.=$c;
    $pos++;
  }
  return $out;
}
?>

In reply to "adrian at zhp dot inet dot pl" digit grouping function:
<?php
    $number = strrev(chunk_split (strrev($number), 3,' '));
    //If $number is '1234567', result is '1 234 567'.
?>

There is a much more simple way of doing this, by using the built-in number_format() function. 

<?php 
   $number = number_format($number,2,"."," ");

   //This will round $number to 2 decimals, use the dot (".")
   //as decimal point, and the space (" ") as thousand sepparator.

?>

Not quite completely obvious, but... 

you can un_chunk_split() by:

$long_str = str_replace( "\r\n", "", $chunked_str );

another way to group thousands in a number, which is much simpler, is built into PHP :)

www.php.net/number_format

When using ssmtp for simple command line mailing:

$mail_to = "[email protected]";
$msg = "this would be an actual base64_encoded gzip msg";
$date = date(r);
$mail  = "X-FROM: [email protected] \n";
$mail .= "X-TO: ".$mail_to. " \n";
$mail .= "To: ".$mail_to. " \n";
$mail .= "Date: $date \n";
$mail .= "From: [email protected] \n";
$mail .= "Subject: lifecheck \n";
$mail .= $msg." \n";
exec("echo '$mail' | /usr/sbin/ssmtp ".$mail_to);

be sure to invoke chunk_split() on your message body - ssmtp becomes unhappy with long lines and will subsequently trash  your message.

I've found this quite useful for simulating various kinds of shuffles with cards.  It is humorous but can imitate multiple deck cuts and other (imperfectly) random events.

<?php
function truffle_shuffle($body, $chunklen = 76, $end = "\r\n")
{
     $chunk = chunk_split($body, $chunklen, "-=blender=-");
     $truffle = explode("-=blender=-",$chunk);
     $shuffle = shuffle($truffle);
     $huknc = implode($end,$shuffle);
     return $huknc;
}
?>

>> chunk_split will also add the break _after_ the last occurence.

this should be not the problem 

substr(chunk_split('FF99FF', 2, ':'),0,8);
will return FF:99:FF

If you are using UTF-8 charset you will face a problem with Arabic language
to solve this problem i used this function

<?php
function chunk_split_($text,$length,$string_end)
{
    $text = iconv("UTF-8","windows-1256",$text);
    $text = str_split($text);
    foreach($text as $val)
    {
        if($a !== $val)
        {
            $a = $val;
            $x = 0;
        }else{
            $a = $val;
            $x++;
        }
        if($x > $length)
        {
            $new_text .= $val.$string_end;
            $x = 0;
        }else
        {
             $new_text .= $val;
        }
        
    }
    $new_text = iconv("windows-1256","UTF-8",$new_text);
    return $new_text;
}
?>

To phpkid:

This is a much simpler solution.

<?php
function longWordWrap($string) {
    $string = str_replace("\n", "\n ", $string); // add a space after newline characters, so that 2 words only seperated by \n are not considered as 1 word
    $words = explode(" ", $string); // now split by space
    foreach ($words as $word) {
        $outstring .= chunk_split($word, 12, " ") . " ";
    }
    return $outstring;
}
?>

@Royce

I think this is better, since you can still use the ampersand in your text:

<?php
function HtmlEntitySafeSplit($html,$size,$delim)
{
  $pos=0;
  for($i=0;$i<strlen($html);$i++)
  {
    if($pos >= $size && !$unsafe)
    {
      $out.=$delim;
      $unsafe=0;
      $pos=0;
    }
    $c=substr($html,$i,1);
    if($c == "&")
      $unsafe=1;
    elseif($c == ";")
      $unsafe=0;
    elseif($c == " ")
      $unsafe=0;
    $out.=$c;
    $pos++;
  }
  return $out;
}
?>

Well I have been having issues with a shoutbox I am coding it would keep expanding the <TD> if there were large words in it but I fixed it with this:

function PadString($String){
    $Exploded = explode(" ", $String);
    $Max_Parts = count($Exploded);
    
    $CurArray = 0;
    $OutString = '';
    while($CurArray<=$Max_Parts)
    {
        $Peice_Size = strlen($Exploded[$CurArray]);
        if($Peice_Size>15)
        {
            $OutString .= chunk_split($Exploded[$CurArray], 12, " ");
            $CurArray++;
        } else {
            $OutString .= " ".$Exploded[$CurArray];
            $CurArray++;
        }
    }
    
    return $OutString;
}