beyondlimits/processHTML.php

## processHTML.php
<?php

return function($html)
{
  $allowedTags = array(
    'body',
    'p', 'div', 'span',
    'b', 'i', 'u',
    'h1', 'h2', 'h3', 'h4', 'h5', 'h6',
    'font',
    'li', 'ul', 'ol',
    'hr', 'br',
    'a', 'blockquote',
    'img',
  );

  $mimetypes = array(
    'image/jpeg' => 'jpeg',
    'image/gif' => 'gif',
    'image/png' => 'png',
  );

  $html = "<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8"></head><body>$html</body></html>";

  $dom = new DOMDocument;

  set_error_handler(function($number, $error) {
    if (preg_match('/^DOMDocument::loadHTML\(\): (.+)$/', $error, $m) == 1) {
      throw new Error($m[1]);
    }
  });

  try {
    $dom->loadHTML($html);
  } catch (Exception $e) {
    $dom = null;
  }

  restore_error_handler();

  if (!$dom) {
    return;
  }

  $stack = array();
  $current = $dom->getElementsByTagName('body')->item(0);

  # usuń niedozwolone tagi oraz atrybuty

  do {
    $passed = true;

    if ($current instanceof DOMElement) {
      if (!in_array($current->tagName, $allowedTags)) {
        $passed = false;

        if ($current->childNodes instanceof Traversable) {
          foreach (iterator_to_array($current->childNodes) as $child) {
            $stack[] = $current->parentNode->insertBefore($child, $current);
          }
        }

        $current->parentNode->removeChild($current);
      }
    }

    if ($passed) {
      $attributes = $current->attributes;

      if ($attributes) {
        foreach ($attributes as $name => $value) {
          if (substr($name, 0, 2) == 'on') {
            $current->removeAttribute($name);
          }
        }
      }

      if ($current->childNodes instanceof Traversable) {
        foreach ($current->childNodes as $child) {
          $stack[] = $child;
        }
      }
    }

    $current = array_pop($stack);
  } while ($current);

  # przekształć IMG

  $host = getservbyport($_SERVER['SERVER_PORT'], 'tcp') . '://' . $_SERVER['SERVER_NAME'];

  foreach ($dom->getElementsByTagName('img') as $img) {
    $src = $img->getAttribute('src');

    if ($src) {
      $url = parse_url($src);

      if ($url['scheme'] == 'data') {
        $md5 = md5_file($src);
        $mime = mime_content_type($src);
        $ext = $mimetypes[$mime];

        if (!$ext) {
          return;
        }

        $dest = MAINDIR . '/' . FILEDIR . "/newsletter/$md5.$ext";

        if (!file_exists($dest)) {
          copy($src, $dest);
        }

        $img->setAttribute('src', "$host/pub/newsletter/$md5.$ext");
      }
    }
  }

  $result = '';

  foreach ($dom->getElementsByTagName('body')->item(0)->childNodes as $node) {
    $result .= $dom->saveHTML($node);
  }

  return $result;
};
	<?php

	return function($html)
	{
	$allowedTags = array(
	'body',
	'p', 'div', 'span',
	'b', 'i', 'u',
	'h1', 'h2', 'h3', 'h4', 'h5', 'h6',
	'font',
	'li', 'ul', 'ol',
	'hr', 'br',
	'a', 'blockquote',
	'img',
	);

	$mimetypes = array(
	'image/jpeg' => 'jpeg',
	'image/gif' => 'gif',
	'image/png' => 'png',
	);

	$html = "<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8"></head><body>$html</body></html>";

	$dom = new DOMDocument;

	set_error_handler(function($number, $error) {
	if (preg_match('/^DOMDocument::loadHTML\(\): (.+)$/', $error, $m) == 1) {
	throw new Error($m[1]);
	}
	});

	try {
	$dom->loadHTML($html);
	} catch (Exception $e) {
	$dom = null;
	}

	restore_error_handler();

	if (!$dom) {
	return;
	}

	$stack = array();
	$current = $dom->getElementsByTagName('body')->item(0);

	# usuń niedozwolone tagi oraz atrybuty

	do {
	$passed = true;

	if ($current instanceof DOMElement) {
	if (!in_array($current->tagName, $allowedTags)) {
	$passed = false;

	if ($current->childNodes instanceof Traversable) {
	foreach (iterator_to_array($current->childNodes) as $child) {
	$stack[] = $current->parentNode->insertBefore($child, $current);
	}
	}

	$current->parentNode->removeChild($current);
	}
	}

	if ($passed) {
	$attributes = $current->attributes;

	if ($attributes) {
	foreach ($attributes as $name => $value) {
	if (substr($name, 0, 2) == 'on') {
	$current->removeAttribute($name);
	}
	}
	}

	if ($current->childNodes instanceof Traversable) {
	foreach ($current->childNodes as $child) {
	$stack[] = $child;
	}
	}
	}

	$current = array_pop($stack);
	} while ($current);

	# przekształć IMG

	$host = getservbyport($_SERVER['SERVER_PORT'], 'tcp') . '://' . $_SERVER['SERVER_NAME'];

	foreach ($dom->getElementsByTagName('img') as $img) {
	$src = $img->getAttribute('src');

	if ($src) {
	$url = parse_url($src);

	if ($url['scheme'] == 'data') {
	$md5 = md5_file($src);
	$mime = mime_content_type($src);
	$ext = $mimetypes[$mime];

	if (!$ext) {
	return;
	}

	$dest = MAINDIR . '/' . FILEDIR . "/newsletter/$md5.$ext";

	if (!file_exists($dest)) {
	copy($src, $dest);
	}

	$img->setAttribute('src', "$host/pub/newsletter/$md5.$ext");
	}
	}
	}

	$result = '';

	foreach ($dom->getElementsByTagName('body')->item(0)->childNodes as $node) {
	$result .= $dom->saveHTML($node);
	}

	return $result;
	};