/**
* This function runs various checks against the attributes.
*
* @param $tagname String with the name of the tag.
* @param $attary Array with all tag attributes.
* @param $rm_attnames See description for tln_sanitize
* @param $bad_attvals See description for tln_sanitize
* @param $add_attr_to_tag See description for tln_sanitize
* @return Array with modified attributes.
*/
function tln_fixatts($tagname, $attary, $rm_attnames, $bad_attvals, $add_attr_to_tag)
{
$me = 'tln_fixatts';
while (list($attname, $attvalue) = each($attary)) {
/**
* See if this attribute should be removed.
*/
foreach ($rm_attnames as $matchtag => $matchattrs) {
if (preg_match($matchtag, $tagname)) {
foreach ($matchattrs as $matchattr) {
if (preg_match($matchattr, $attname)) {
unset($attary[$attname]);
continue;
}
}
}
}
/**
* Remove any backslashes, entities, or extraneous whitespace.
*/
tln_defang($attvalue);
tln_unspace($attvalue);
/**
* Now let's run checks on the attvalues.
* I don't expect anyone to comprehend this. If you do,
* get in touch with me so I can drive to where you live and
* shake your hand personally. :)
*/
foreach ($bad_attvals as $matchtag => $matchattrs) {
if (preg_match($matchtag, $tagname)) {
foreach ($matchattrs as $matchattr => $valary) {
if (preg_match($matchattr, $attname)) {
/**
* There are two arrays in valary.
* First is matches.
* Second one is replacements
*/
list($valmatch, $valrepl) = $valary;
$newvalue = preg_replace($valmatch, $valrepl, $attvalue);
if ($newvalue != $attvalue) {
$attary[$attname] = $newvalue;
}
}
}
}
}
}
/**
* See if we need to append any attributes to this tag.
*/
foreach ($add_attr_to_tag as $matchtag => $addattary) {
if (preg_match($matchtag, $tagname)) {
$attary = array_merge($attary, $addattary);
}
}
return $attary;
}
function tln_fixstyle($body, $pos, $trans_image_path, $block_external_images)
{
$me = 'tln_fixstyle';
// workaround for </style> in between comments
$iCurrentPos = $pos;
$content = '';
$sToken = '';
$bSucces = false;
$bEndTag = false;
for ($i = $pos, $iCount = strlen($body); $i < $iCount; ++$i) {
$char = $body[$i];
switch ($char) {
case '<':
$sToken = $char;
break;
case '/':
if ($sToken == '<') {
$sToken .= $char;
$bEndTag = true;
} else {
$content .= $char;
}
break;
case '>':
if ($bEndTag) {
$sToken .= $char;
if (preg_match('/\\<\\/\\s*style\\s*\\>/i', $sToken, $aMatch)) {
$newpos = $i + 1;
$bSucces = true;
break 2;
} else {
$content .= $sToken;
}
$bEndTag = false;
} else {
$content .= $char;
}
break;
case '!':
if ($sToken == '<') {
// possible comment
if (isset($body[$i + 2]) && substr($body, $i, 3) == '!--') {
$i = strpos($body, '-->', $i + 3);
if ($i === false) {
// no end comment
$i = strlen($body);
}
$sToken = '';
}
} else {
$content .= $char;
}
break;
default:
if ($bEndTag) {
$sToken .= $char;
} else {
$content .= $char;
}
break;
}
}
if ($bSucces == FALSE) {
return array(FALSE, strlen($body));
}
/**
* First look for general BODY style declaration, which would be
* like so:
* body {background: blah-blah}
* and change it to .bodyclass so we can just assign it to a <div>
*/
$content = preg_replace("|body(\\s*\\{.*?\\})|si", ".bodyclass\\1", $content);
$trans_image_path = $trans_image_path;
/**
* Fix url('blah') declarations.
*/
// $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
// "url(\\1$trans_image_path\\2)", $content);
// first check for 8bit sequences and disallowed control characters
if (preg_match('/[\\16-\\37\\200-\\377]+/', $content)) {
$content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
return array($content, $newpos);
}
// remove @import line
$content = preg_replace("/^\\s*(@import.*)\$/mi", "\n<!-- @import rules forbidden -->\n", $content);
$content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
preg_match_all("/url\\s*\\((.+)\\)/si", $content, $aMatch);
if (count($aMatch)) {
$aValue = $aReplace = array();
foreach ($aMatch[1] as $sMatch) {
// url value
$urlvalue = $sMatch;
tln_fixurl('style', $urlvalue, $trans_image_path, $block_external_images);
$aValue[] = $sMatch;
$aReplace[] = $urlvalue;
}
$content = str_replace($aValue, $aReplace, $content);
}
/**
* Remove any backslashes, entities, and extraneous whitespace.
*/
$contentTemp = $content;
tln_defang($contentTemp);
tln_unspace($contentTemp);
$match = array('/\\/\\*.*\\*\\//', '/expression/i', '/behaviou*r/i', '/binding/i', '/include-source/i', '/javascript/i', '/script/i', '/position/i');
$replace = array('', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', '');
$contentNew = preg_replace($match, $replace, $contentTemp);
if ($contentNew !== $contentTemp) {
$content = $contentNew;
}
return array($content, $newpos);
}
