<?php


//callback function for the regex
function utf8_entity_decode($entity){
        $convmap =  array(0x0, 0x10000, 0, 0xfffff);
        return mb_decode_numericentity($entity, $convmap, 'UTF-8');
}

$xmlDoc = file_get_contents("php://stdin");

$xmlDoc = str_replace("&#34;",'"', $xmlDoc);
$xmlDoc = str_replace("&#38;","&amp;", $xmlDoc);
$xmlDoc = str_replace("&#39;","'", $xmlDoc);
$xmlDoc = str_replace("&#60;","&lt;", $xmlDoc);
$xmlDoc = str_replace("&#133;","...", $xmlDoc);
$xmlDoc = str_replace("&#145;","'", $xmlDoc);
$xmlDoc = str_replace("&#146;","'", $xmlDoc);
$xmlDoc = str_replace("&#147;","'", $xmlDoc);
$xmlDoc = str_replace("&#149;","-", $xmlDoc);
//$xmlDoc = str_replace("&#149;","·", $xmlDoc);
$xmlDoc = str_replace("&#151;","-", $xmlDoc);
$xmlDoc = str_replace("&#8212;","-", $xmlDoc);
$xmlDoc = str_replace("&#8217;","'", $xmlDoc);
$xmlDoc = str_replace("&#8220;","'", $xmlDoc);
$xmlDoc = str_replace("&#8221;","'", $xmlDoc);

//decode decimal html entities added by web browser
$xmlDoc = preg_replace_callback('/&#\d{2,5};/u',
        function($match){
                return utf8_entity_decode($match[0]);
        },
        $xmlDoc
);

//$xmlDoc = preg_replace_callback('/&#\d{2,5};/ue', utf8_entity_decode('$0'), $xmlDoc);

//decode hex html entities added by web browser
$xmlDoc = preg_replace_callback('/&#x([a-fA-F0-7]{2,8});/u',
                function($match){
                return utf8_entity_decode('&#'.hexdec($match[1]).';');
        },
        $xmlDoc
);

//$xmlDoc = preg_replace_callback('/&#x([a-fA-F0-7]{2,8});/ue', utf8_entity_decode('&#'.hexdec('$1').';'), $xmlDoc);

//Replace "&amp;#???; encoded characters with equivalents

$xmlDoc = str_replace("&amp;#160;"," ", $xmlDoc);
$xmlDoc = str_replace("&amp;#172;","", $xmlDoc);
$xmlDoc = str_replace("&amp;#173;","", $xmlDoc);