My PDF
EOF;
$doc = new DOMDocument();
libxml_use_internal_errors(true);
$doc->loadHTML($html); // loads your html
$nodeList = $doc->getElementsByTagName('a');
for($i=0; $i < $nodeList->length; $i++) {
$node = $nodeList->item($i);
$children = $node->childNodes;
$hasImage = false;
foreach ($children as $child) {
if ($child->nodeName == 'img') {
$hasImage = true;
break;
}
}
if (!$hasImage)
continue;
if ($node->hasAttributes())
foreach ($node->attributes as $attr) {
$name = $attr->nodeName;
$value = $attr->nodeValue;
if ( $attr->nodeName == 'href' && preg_match('/\.(doc|pdf)$/i', $attr->nodeValue) )
echo $attr->nodeValue . " - Image is wrapped in a link to a PDF or DOC file\n";
}
}
?>