The output is correct in all 4 cases. The problem is the improper use of the second parameter of rtrim :

When you say rtrim( $string, "<hr>" ) , you are saying "remove all occurrences of the characters < , h , r and > from the end of the string, and leave the rest."

Note that r is in the list, so it will be removed as requested.

To remove any other string from the end, replace it like this:

$string = '<hr>teste<hr><hr>';

while( substr( $string, -4) == '<hr>' ) {
   $string = substr( $string, 0, -4 );
}

echo $string;

Output:

<hr>teste

  

This was just a simple example as a starting point. Using arrays can do a function that removes several different groups of strings at a time.

There's even an example in the PHP site comments that does something of the sort , but that can be very simplified in my view.

You are including 4 items for deletion at the end of your string: .

Why not make use of the good old regex?

<?= preg_replace('/<hr( \/)?>$/i', '', $string); ?>

This will replace <hr> or <hr /> of the end of the string with nothing.

  

Obs. : Avoid double quotes, as php will process the content. Instead, the use of single quotation marks is preferable, unless   that you want the content to be processed. Ex : <?= "Meu nome é $nome"; ?>

Hugs!

You can in a loop check the position of the characters to be removed from the string by using strpos() " and "cut" the string with the substr() ".

You can implement this as follows:

function rtrim2($str, $charlist=null) 
{ 
    $str      = (string)$str; 
    $charlist = (string)$charlist;    

    if(empty($charlist)) 
       return rtrim($str); 

    $len = strlen($charlist); 
    $offset = strlen($str) - $len; 
    while($offset > 0 && $offset == strpos($str, $charlist, $offset)) 
    { 
        $str = substr($str, 0, $offset); 
        $offset = strlen($str) - $len; 
    } 

    return rtrim($str);    
}

Example at IdeOne

Try to use str_replace("<br>", "". $palavra);