| | |
| | | <?php |
| | | |
| | | /** |
| | | /* |
| | | +-----------------------------------------------------------------------+ |
| | | | This file is part of the Roundcube Webmail client | |
| | | | Copyright (C) 2008-2012, The Roundcube Dev Team | |
| | |
| | | +-----------------------------------------------------------------------+ |
| | | */ |
| | | |
| | | /** |
| | | /* |
| | | * Washtml, a HTML sanityzer. |
| | | * |
| | | * Copyright (c) 2007 Frederic Motte <fmotte@ubixis.com> |
| | |
| | | 'ins', 'label', 'legend', 'li', 'map', 'menu', 'nobr', 'ol', 'p', 'pre', 'q', |
| | | 's', 'samp', 'small', 'span', 'strike', 'strong', 'sub', 'sup', 'table', |
| | | 'tbody', 'td', 'tfoot', 'th', 'thead', 'tr', 'tt', 'u', 'ul', 'var', 'wbr', 'img', |
| | | 'video', 'source', |
| | | // form elements |
| | | 'button', 'input', 'textarea', 'select', 'option', 'optgroup' |
| | | ); |
| | |
| | | */ |
| | | private function wash_style($style) |
| | | { |
| | | $s = ''; |
| | | $result = array(); |
| | | |
| | | foreach (explode(';', $style) as $declaration) { |
| | | if (preg_match('/^\s*([a-z\-]+)\s*:\s*(.*)\s*$/i', $declaration, $match)) { |
| | |
| | | $str = $match[2]; |
| | | $value = ''; |
| | | |
| | | while (sizeof($str) > 0 && |
| | | preg_match('/^(url\(\s*[\'"]?([^\'"\)]*)[\'"]?\s*\)'./*1,2*/ |
| | | '|rgb\(\s*[0-9]+\s*,\s*[0-9]+\s*,\s*[0-9]+\s*\)'. |
| | | '|-?[0-9.]+\s*(em|ex|px|cm|mm|in|pt|pc|deg|rad|grad|ms|s|hz|khz|%)?'. |
| | | '|#[0-9a-f]{3,6}'. |
| | | '|[a-z0-9", -]+'. |
| | | ')\s*/i', $str, $match) |
| | | ) { |
| | | if ($match[2]) { |
| | | if (($src = $this->config['cid_map'][$match[2]]) |
| | | || ($src = $this->config['cid_map'][$this->config['base_url'].$match[2]]) |
| | | ) { |
| | | $value .= ' url('.htmlspecialchars($src, ENT_QUOTES) . ')'; |
| | | } |
| | | else if (preg_match('!^(https?:)?//[a-z0-9/._+-]+$!i', $match[2], $url)) { |
| | | if ($this->config['allow_remote']) { |
| | | $value .= ' url('.htmlspecialchars($url[0], ENT_QUOTES).')'; |
| | | foreach ($this->explode_style($str) as $val) { |
| | | if (preg_match('/^url\(/i', $val)) { |
| | | if (preg_match('/^url\(\s*[\'"]?([^\'"\)]*)[\'"]?\s*\)/iu', $val, $match)) { |
| | | $url = $match[1]; |
| | | if (($src = $this->config['cid_map'][$url]) |
| | | || ($src = $this->config['cid_map'][$this->config['base_url'].$url]) |
| | | ) { |
| | | $value .= ' url('.htmlspecialchars($src, ENT_QUOTES) . ')'; |
| | | } |
| | | else { |
| | | $this->extlinks = true; |
| | | else if (preg_match('!^(https?:)?//[a-z0-9/._+-]+$!i', $url, $m)) { |
| | | if ($this->config['allow_remote']) { |
| | | $value .= ' url('.htmlspecialchars($m[0], ENT_QUOTES).')'; |
| | | } |
| | | else { |
| | | $this->extlinks = true; |
| | | } |
| | | } |
| | | } |
| | | else if (preg_match('/^data:.+/i', $match[2])) { // RFC2397 |
| | | $value .= ' url('.htmlspecialchars($match[2], ENT_QUOTES).')'; |
| | | else if (preg_match('/^data:.+/i', $url)) { // RFC2397 |
| | | $value .= ' url('.htmlspecialchars($url, ENT_QUOTES).')'; |
| | | } |
| | | } |
| | | } |
| | | else { |
| | | else if (!preg_match('/^(behavior|expression)/i', $val)) { |
| | | // whitelist ? |
| | | $value .= ' ' . $match[0]; |
| | | $value .= ' ' . $val; |
| | | |
| | | // #1488535: Fix size units, so width:800 would be changed to width:800px |
| | | if (preg_match('/(left|right|top|bottom|width|height)/i', $cssid) |
| | | && preg_match('/^[0-9]+$/', $match[0]) |
| | | if (preg_match('/^(left|right|top|bottom|width|height)/i', $cssid) |
| | | && preg_match('/^[0-9]+$/', $val) |
| | | ) { |
| | | $value .= 'px'; |
| | | } |
| | | } |
| | | |
| | | $str = substr($str, strlen($match[0])); |
| | | } |
| | | |
| | | if (isset($value[0])) { |
| | | $s .= ($s?' ':'') . $cssid . ':' . $value . ';'; |
| | | $result[] = $cssid . ':' . $value; |
| | | } |
| | | } |
| | | } |
| | | |
| | | return $s; |
| | | return implode('; ', $result); |
| | | } |
| | | |
| | | /** |
| | |
| | | $t .= ' ' . $key . '="' . htmlspecialchars($value, ENT_QUOTES) . '"'; |
| | | } |
| | | else if ($key == 'style' && ($style = $this->wash_style($value))) { |
| | | $quot = strpos($style, '"') !== false ? "'" : '"'; |
| | | $t .= ' style=' . $quot . $style . $quot; |
| | | // replace double quotes to prevent syntax error and XSS issues (#1490227) |
| | | $t .= ' style="' . str_replace('"', '"', $style) . '"'; |
| | | } |
| | | else if ($key == 'background' || ($key == 'src' && strtolower($node->tagName) == 'img')) { //check tagName anyway |
| | | else if ($key == 'background' |
| | | || ($key == 'src' && preg_match('/^(img|source)$/i', $node->tagName)) |
| | | || ($key == 'poster' && strtolower($node->tagName) == 'video') |
| | | ) { |
| | | if (($src = $this->config['cid_map'][$value]) |
| | | || ($src = $this->config['cid_map'][$this->config['base_url'].$value]) |
| | | ) { |
| | |
| | | |
| | | /** |
| | | * The main loop that recurse on a node tree. |
| | | * It output only allowed tags with allowed attributes |
| | | * and allowed inline styles |
| | | * It output only allowed tags with allowed attributes and allowed inline styles |
| | | * |
| | | * @param DOMNode $node HTML element |
| | | * @param int $level Recurrence level (safe initial value found empirically) |
| | | */ |
| | | private function dumpHtml($node, $level = 0) |
| | | private function dumpHtml($node, $level = 20) |
| | | { |
| | | if (!$node->hasChildNodes()) { |
| | | return ''; |
| | |
| | | $this->max_nesting_level = (int) @ini_get('xdebug.max_nesting_level'); |
| | | |
| | | // Use optimizations if supported |
| | | if (version_compare(PHP_VERSION, '5.4.0', '>=')) { |
| | | if (PHP_VERSION_ID >= 50400) { |
| | | @$node->loadHTML($html, LIBXML_PARSEHUGE | LIBXML_COMPACT); |
| | | } |
| | | else { |
| | |
| | | { |
| | | // special replacements (not properly handled by washtml class) |
| | | $html_search = array( |
| | | '/(<\/nobr>)(\s+)(<nobr>)/i', // space(s) between <NOBR> |
| | | '/<title[^>]*>[^<]*<\/title>/i', // PHP bug #32547 workaround: remove title tag |
| | | '/^(\0\0\xFE\xFF|\xFF\xFE\0\0|\xFE\xFF|\xFF\xFE|\xEF\xBB\xBF)/', // byte-order mark (only outlook?) |
| | | '/<html\s[^>]+>/i', // washtml/DOMDocument cannot handle xml namespaces |
| | | // space(s) between <NOBR> |
| | | '/(<\/nobr>)(\s+)(<nobr>)/i', |
| | | // PHP bug #32547 workaround: remove title tag |
| | | '/<title[^>]*>[^<]*<\/title>/i', |
| | | // remove <!doctype> before BOM (#1490291) |
| | | '/<\!doctype[^>]+>[^<]*/im', |
| | | // byte-order mark (only outlook?) |
| | | '/^(\0\0\xFE\xFF|\xFF\xFE\0\0|\xFE\xFF|\xFF\xFE|\xEF\xBB\xBF)/', |
| | | // washtml/DOMDocument cannot handle xml namespaces |
| | | '/<html\s[^>]+>/i', |
| | | ); |
| | | |
| | | $html_replace = array( |
| | | '\\1'.' '.'\\3', |
| | | '', |
| | | '', |
| | | '', |
| | | '<html>', |
| | |
| | | // Remove invalid HTML comments (#1487759) |
| | | // Don't remove valid conditional comments |
| | | // Don't remove MSOutlook (<!-->) conditional comments (#1489004) |
| | | $html = preg_replace('/<!--[^->\[\n]+>/', '', $html); |
| | | $html = preg_replace('/<!--[^-<>\[\n]+>/', '', $html); |
| | | |
| | | // fix broken nested lists |
| | | self::fix_broken_lists($html); |
| | |
| | | } |
| | | } |
| | | } |
| | | |
| | | /** |
| | | * Explode css style value |
| | | */ |
| | | protected function explode_style($style) |
| | | { |
| | | $style = trim($style); |
| | | |
| | | // first remove comments |
| | | $pos = 0; |
| | | while (($pos = strpos($style, '/*', $pos)) !== false) { |
| | | $end = strpos($style, '*/', $pos+2); |
| | | |
| | | if ($end === false) { |
| | | $style = substr($style, 0, $pos); |
| | | } |
| | | else { |
| | | $style = substr_replace($style, '', $pos, $end - $pos + 2); |
| | | } |
| | | } |
| | | |
| | | $strlen = strlen($style); |
| | | $result = array(); |
| | | |
| | | // explode value |
| | | for ($p=$i=0; $i < $strlen; $i++) { |
| | | if (($style[$i] == "\"" || $style[$i] == "'") && $style[$i-1] != "\\") { |
| | | if ($q == $style[$i]) { |
| | | $q = false; |
| | | } |
| | | else if (!$q) { |
| | | $q = $style[$i]; |
| | | } |
| | | } |
| | | |
| | | if (!$q && $style[$i] == ' ' && !preg_match('/[,\(]/', $style[$i-1])) { |
| | | $result[] = substr($style, $p, $i - $p); |
| | | $p = $i + 1; |
| | | } |
| | | } |
| | | |
| | | $result[] = (string) substr($style, $p); |
| | | |
| | | return $result; |
| | | } |
| | | } |