private static function _pack_analyze($script, $regexp, $encode)
{
$wall = array();
// all words in the script
// get all words
preg_match_all($regexp, $script, $wall);
// simulate the javascript comportement of global match
$wall = $wall[0];
if (empty($wall)) {
return false;
}
$wsort = array();
// list of words sorted by frequency
$wusrt = array();
// list of words without sorting
$wprot = array();
// instances of "protected" words
$wenco = array();
// dictionary of word->encoding
$dprot = array();
// dictionary of word->"word"
$value = array();
// dictionary of charCode->encoding (eg. 256->ff)
self::$_PWC = array();
// word->count
$i = count($wall);
$j = 0;
//$word = null;
// count the occurrences - used for sorting later
do {
--$i;
$word = '$' . $wall[$i];
if (!isset(self::$_PWC[$word])) {
self::$_PWC[$word] = 0;
$wusrt[$j] = $word;
// make a dictionary of all of the protected words in this script
// these are words that might be mistaken for encoding
$value[$j] = call_user_func(array('self', $encode), $j);
$dprot['$' . $value[$j]] = $j++;
}
// increment the word counter
self::$_PWC[$word]++;
} while ($i > 0);
// prepare to sort the word list, first we must protect
// words that are also used as codes. we assign them a code
// equivalent to the word itself.
// e.g. if "do" falls within our encoding range
// then we store keywords["do"] = "do";
// this avoids problems when decoding
$i = count($wusrt);
do {
$word = $wusrt[--$i];
if (isset($dprot[$word])) {
$wsort[$dprot[$word]] = substr($word, 1);
$iprot[$dprot[$word]] = true;
self::$_PWC[$word] = 0;
}
} while ($i);
// sort the words by frequency
// Note: the javascript and php version of sort can be different :
// in php manual, usort :
// " If two members compare as equal,
// their order in the sorted array is undefined."
// so the final packed script is different of the Dean's javascript version
// but equivalent.
// the ECMAscript standard does not guarantee this behaviour,
// and thus not all browsers (e.g. Mozilla versions dating back to at
// least 2003) respect this.
usort($wusrt, array('self', '_pack_sortwords'));
$j = 0;
// because there are "protected" words in the list
// we must add the sorted words around them
do {
if (!isset($wsort[$i])) {
$wsort[$i] = substr($wusrt[$j++], 1);
}
$wenco[$wsort[$i]] = $value[$i];
} while (++$i < count($wusrt));
return array('sorted' => $wsort, 'encoded' => $wenco, 'protected' => $wprot);
}
