diff options
Diffstat (limited to 'program/lib/Roundcube')
| -rw-r--r-- | program/lib/Roundcube/rcube_html2text.php | 691 | ||||
| -rw-r--r-- | program/lib/Roundcube/rcube_message.php | 2 | ||||
| -rw-r--r-- | program/lib/Roundcube/rcube_spellchecker.php | 2 | ||||
| -rw-r--r-- | program/lib/Roundcube/rcube_string_replacer.php | 27 | ||||
| -rw-r--r-- | program/lib/Roundcube/rcube_washtml.php | 451 |
5 files changed, 1170 insertions, 3 deletions
diff --git a/program/lib/Roundcube/rcube_html2text.php b/program/lib/Roundcube/rcube_html2text.php new file mode 100644 index 000000000..0b172ebfa --- /dev/null +++ b/program/lib/Roundcube/rcube_html2text.php @@ -0,0 +1,691 @@ +<?php + +/** + +-----------------------------------------------------------------------+ + | This file is part of the Roundcube Webmail client | + | Copyright (C) 2008-2012, The Roundcube Dev Team | + | Copyright (c) 2005-2007, Jon Abernathy <jon@chuggnutt.com> | + | | + | Licensed under the GNU General Public License version 3 or | + | any later version with exceptions for skins & plugins. | + | See the README file for a full license statement. | + | | + | PURPOSE: | + | Converts HTML to formatted plain text (based on html2text class) | + +-----------------------------------------------------------------------+ + | Author: Thomas Bruederli <roundcube@gmail.com> | + | Author: Aleksander Machniak <alec@alec.pl> | + | Author: Jon Abernathy <jon@chuggnutt.com> | + +-----------------------------------------------------------------------+ + */ + +/** + * Takes HTML and converts it to formatted, plain text. + * + * Thanks to Alexander Krug (http://www.krugar.de/) to pointing out and + * correcting an error in the regexp search array. Fixed 7/30/03. + * + * Updated set_html() function's file reading mechanism, 9/25/03. + * + * Thanks to Joss Sanglier (http://www.dancingbear.co.uk/) for adding + * several more HTML entity codes to the $search and $replace arrays. + * Updated 11/7/03. + * + * Thanks to Darius Kasperavicius (http://www.dar.dar.lt/) for + * suggesting the addition of $allowed_tags and its supporting function + * (which I slightly modified). Updated 3/12/04. + * + * Thanks to Justin Dearing for pointing out that a replacement for the + * <TH> tag was missing, and suggesting an appropriate fix. + * Updated 8/25/04. + * + * Thanks to Mathieu Collas (http://www.myefarm.com/) for finding a + * display/formatting bug in the _build_link_list() function: email + * readers would show the left bracket and number ("[1") as part of the + * rendered email address. + * Updated 12/16/04. + * + * Thanks to Wojciech Bajon (http://histeria.pl/) for submitting code + * to handle relative links, which I hadn't considered. I modified his + * code a bit to handle normal HTTP links and MAILTO links. Also for + * suggesting three additional HTML entity codes to search for. + * Updated 03/02/05. + * + * Thanks to Jacob Chandler for pointing out another link condition + * for the _build_link_list() function: "https". + * Updated 04/06/05. + * + * Thanks to Marc Bertrand (http://www.dresdensky.com/) for + * suggesting a revision to the word wrapping functionality; if you + * specify a $width of 0 or less, word wrapping will be ignored. + * Updated 11/02/06. + * + * *** Big housecleaning updates below: + * + * Thanks to Colin Brown (http://www.sparkdriver.co.uk/) for + * suggesting the fix to handle </li> and blank lines (whitespace). + * Christian Basedau (http://www.movetheweb.de/) also suggested the + * blank lines fix. + * + * Special thanks to Marcus Bointon (http://www.synchromedia.co.uk/), + * Christian Basedau, Norbert Laposa (http://ln5.co.uk/), + * Bas van de Weijer, and Marijn van Butselaar + * for pointing out my glaring error in the <th> handling. Marcus also + * supplied a host of fixes. + * + * Thanks to Jeffrey Silverman (http://www.newtnotes.com/) for pointing + * out that extra spaces should be compressed--a problem addressed with + * Marcus Bointon's fixes but that I had not yet incorporated. + * + * Thanks to Daniel Schledermann (http://www.typoconsult.dk/) for + * suggesting a valuable fix with <a> tag handling. + * + * Thanks to Wojciech Bajon (again!) for suggesting fixes and additions, + * including the <a> tag handling that Daniel Schledermann pointed + * out but that I had not yet incorporated. I haven't (yet) + * incorporated all of Wojciech's changes, though I may at some + * future time. + * + * *** End of the housecleaning updates. Updated 08/08/07. + */ + +/** + * Converts HTML to formatted plain text + * + * @package Framework + * @subpackage Utils + */ +class rcube_html2text +{ + /** + * Contains the HTML content to convert. + * + * @var string $html + */ + protected $html; + + /** + * Contains the converted, formatted text. + * + * @var string $text + */ + protected $text; + + /** + * Maximum width of the formatted text, in columns. + * + * Set this value to 0 (or less) to ignore word wrapping + * and not constrain text to a fixed-width column. + * + * @var integer $width + */ + protected $width = 70; + + /** + * Target character encoding for output text + * + * @var string $charset + */ + protected $charset = 'UTF-8'; + + /** + * List of preg* regular expression patterns to search for, + * used in conjunction with $replace. + * + * @var array $search + * @see $replace + */ + protected $search = array( + "/\r/", // Non-legal carriage return + "/[\n\t]+/", // Newlines and tabs + '/<head[^>]*>.*?<\/head>/i', // <head> + '/<script[^>]*>.*?<\/script>/i', // <script>s -- which strip_tags supposedly has problems with + '/<style[^>]*>.*?<\/style>/i', // <style>s -- which strip_tags supposedly has problems with + '/<p[^>]*>/i', // <P> + '/<br[^>]*>/i', // <br> + '/<i[^>]*>(.*?)<\/i>/i', // <i> + '/<em[^>]*>(.*?)<\/em>/i', // <em> + '/(<ul[^>]*>|<\/ul>)/i', // <ul> and </ul> + '/(<ol[^>]*>|<\/ol>)/i', // <ol> and </ol> + '/<li[^>]*>(.*?)<\/li>/i', // <li> and </li> + '/<li[^>]*>/i', // <li> + '/<hr[^>]*>/i', // <hr> + '/<div[^>]*>/i', // <div> + '/(<table[^>]*>|<\/table>)/i', // <table> and </table> + '/(<tr[^>]*>|<\/tr>)/i', // <tr> and </tr> + '/<td[^>]*>(.*?)<\/td>/i', // <td> and </td> + ); + + /** + * List of pattern replacements corresponding to patterns searched. + * + * @var array $replace + * @see $search + */ + protected $replace = array( + '', // Non-legal carriage return + ' ', // Newlines and tabs + '', // <head> + '', // <script>s -- which strip_tags supposedly has problems with + '', // <style>s -- which strip_tags supposedly has problems with + "\n\n", // <P> + "\n", // <br> + '_\\1_', // <i> + '_\\1_', // <em> + "\n\n", // <ul> and </ul> + "\n\n", // <ol> and </ol> + "\t* \\1\n", // <li> and </li> + "\n\t* ", // <li> + "\n-------------------------\n", // <hr> + "<div>\n", // <div> + "\n\n", // <table> and </table> + "\n", // <tr> and </tr> + "\t\t\\1\n", // <td> and </td> + ); + + /** + * List of preg* regular expression patterns to search for, + * used in conjunction with $ent_replace. + * + * @var array $ent_search + * @see $ent_replace + */ + protected $ent_search = array( + '/&(nbsp|#160);/i', // Non-breaking space + '/&(quot|rdquo|ldquo|#8220|#8221|#147|#148);/i', + // Double quotes + '/&(apos|rsquo|lsquo|#8216|#8217);/i', // Single quotes + '/>/i', // Greater-than + '/</i', // Less-than + '/&(copy|#169);/i', // Copyright + '/&(trade|#8482|#153);/i', // Trademark + '/&(reg|#174);/i', // Registered + '/&(mdash|#151|#8212);/i', // mdash + '/&(ndash|minus|#8211|#8722);/i', // ndash + '/&(bull|#149|#8226);/i', // Bullet + '/&(pound|#163);/i', // Pound sign + '/&(euro|#8364);/i', // Euro sign + '/&(amp|#38);/i', // Ampersand: see _converter() + '/[ ]{2,}/', // Runs of spaces, post-handling + ); + + /** + * List of pattern replacements corresponding to patterns searched. + * + * @var array $ent_replace + * @see $ent_search + */ + protected $ent_replace = array( + ' ', // Non-breaking space + '"', // Double quotes + "'", // Single quotes + '>', + '<', + '(c)', + '(tm)', + '(R)', + '--', + '-', + '*', + '£', + 'EUR', // Euro sign. € ? + '|+|amp|+|', // Ampersand: see _converter() + ' ', // Runs of spaces, post-handling + ); + + /** + * List of preg* regular expression patterns to search for + * and replace using callback function. + * + * @var array $callback_search + */ + protected $callback_search = array( + '/<(a) [^>]*href=("|\')([^"\']+)\2[^>]*>(.*?)<\/a>/i', // <a href=""> + '/<(h)[123456]( [^>]*)?>(.*?)<\/h[123456]>/i', // h1 - h6 + '/<(b)( [^>]*)?>(.*?)<\/b>/i', // <b> + '/<(strong)( [^>]*)?>(.*?)<\/strong>/i', // <strong> + '/<(th)( [^>]*)?>(.*?)<\/th>/i', // <th> and </th> + ); + + /** + * List of preg* regular expression patterns to search for in PRE body, + * used in conjunction with $pre_replace. + * + * @var array $pre_search + * @see $pre_replace + */ + protected $pre_search = array( + "/\n/", + "/\t/", + '/ /', + '/<pre[^>]*>/', + '/<\/pre>/' + ); + + /** + * List of pattern replacements corresponding to patterns searched for PRE body. + * + * @var array $pre_replace + * @see $pre_search + */ + protected $pre_replace = array( + '<br>', + ' ', + ' ', + '', + '' + ); + + /** + * Contains a list of HTML tags to allow in the resulting text. + * + * @var string $allowed_tags + * @see set_allowed_tags() + */ + protected $allowed_tags = ''; + + /** + * Contains the base URL that relative links should resolve to. + * + * @var string $url + */ + protected $url; + + /** + * Indicates whether content in the $html variable has been converted yet. + * + * @var boolean $_converted + * @see $html, $text + */ + protected $_converted = false; + + /** + * Contains URL addresses from links to be rendered in plain text. + * + * @var array $_link_list + * @see _build_link_list() + */ + protected $_link_list = array(); + + /** + * Boolean flag, true if a table of link URLs should be listed after the text. + * + * @var boolean $_do_links + * @see __construct() + */ + protected $_do_links = true; + + /** + * Constructor. + * + * If the HTML source string (or file) is supplied, the class + * will instantiate with that source propagated, all that has + * to be done it to call get_text(). + * + * @param string $source HTML content + * @param boolean $from_file Indicates $source is a file to pull content from + * @param boolean $do_links Indicate whether a table of link URLs is desired + * @param integer $width Maximum width of the formatted text, 0 for no limit + */ + function __construct($source = '', $from_file = false, $do_links = true, $width = 75, $charset = 'UTF-8') + { + if (!empty($source)) { + $this->set_html($source, $from_file); + } + + $this->set_base_url(); + + $this->_do_links = $do_links; + $this->width = $width; + $this->charset = $charset; + } + + /** + * Loads source HTML into memory, either from $source string or a file. + * + * @param string $source HTML content + * @param boolean $from_file Indicates $source is a file to pull content from + */ + function set_html($source, $from_file = false) + { + if ($from_file && file_exists($source)) { + $this->html = file_get_contents($source); + } + else { + $this->html = $source; + } + + $this->_converted = false; + } + + /** + * Returns the text, converted from HTML. + * + * @return string Plain text + */ + function get_text() + { + if (!$this->_converted) { + $this->_convert(); + } + + return $this->text; + } + + /** + * Prints the text, converted from HTML. + */ + function print_text() + { + print $this->get_text(); + } + + /** + * Sets the allowed HTML tags to pass through to the resulting text. + * + * Tags should be in the form "<p>", with no corresponding closing tag. + */ + function set_allowed_tags($allowed_tags = '') + { + if (!empty($allowed_tags)) { + $this->allowed_tags = $allowed_tags; + } + } + + /** + * Sets a base URL to handle relative links. + */ + function set_base_url($url = '') + { + if (empty($url)) { + if (!empty($_SERVER['HTTP_HOST'])) { + $this->url = 'http://' . $_SERVER['HTTP_HOST']; + } + else { + $this->url = ''; + } + } + else { + // Strip any trailing slashes for consistency (relative + // URLs may already start with a slash like "/file.html") + if (substr($url, -1) == '/') { + $url = substr($url, 0, -1); + } + $this->url = $url; + } + } + + /** + * Workhorse function that does actual conversion (calls _converter() method). + */ + protected function _convert() + { + // Variables used for building the link list + $this->_link_list = array(); + + $text = trim(stripslashes($this->html)); + + // Convert HTML to TXT + $this->_converter($text); + + // Add link list + if (!empty($this->_link_list)) { + $text .= "\n\nLinks:\n------\n"; + foreach ($this->_link_list as $idx => $url) { + $text .= '[' . ($idx+1) . '] ' . $url . "\n"; + } + } + + $this->text = $text; + $this->_converted = true; + } + + /** + * Workhorse function that does actual conversion. + * + * First performs custom tag replacement specified by $search and + * $replace arrays. Then strips any remaining HTML tags, reduces whitespace + * and newlines to a readable format, and word wraps the text to + * $width characters. + * + * @param string Reference to HTML content string + */ + protected function _converter(&$text) + { + // Convert <BLOCKQUOTE> (before PRE!) + $this->_convert_blockquotes($text); + + // Convert <PRE> + $this->_convert_pre($text); + + // Run our defined tags search-and-replace + $text = preg_replace($this->search, $this->replace, $text); + + // Run our defined tags search-and-replace with callback + $text = preg_replace_callback($this->callback_search, array($this, 'tags_preg_callback'), $text); + + // Strip any other HTML tags + $text = strip_tags($text, $this->allowed_tags); + + // Run our defined entities/characters search-and-replace + $text = preg_replace($this->ent_search, $this->ent_replace, $text); + + // Replace known html entities + $text = html_entity_decode($text, ENT_QUOTES, $this->charset); + + // Remove unknown/unhandled entities (this cannot be done in search-and-replace block) + $text = preg_replace('/&([a-zA-Z0-9]{2,6}|#[0-9]{2,4});/', '', $text); + + // Convert "|+|amp|+|" into "&", need to be done after handling of unknown entities + // This properly handles situation of "&quot;" in input string + $text = str_replace('|+|amp|+|', '&', $text); + + // Bring down number of empty lines to 2 max + $text = preg_replace("/\n\s+\n/", "\n\n", $text); + $text = preg_replace("/[\n]{3,}/", "\n\n", $text); + + // remove leading empty lines (can be produced by eg. P tag on the beginning) + $text = ltrim($text, "\n"); + + // Wrap the text to a readable format + // for PHP versions >= 4.0.2. Default width is 75 + // If width is 0 or less, don't wrap the text. + if ( $this->width > 0 ) { + $text = wordwrap($text, $this->width); + } + } + + /** + * Helper function called by preg_replace() on link replacement. + * + * Maintains an internal list of links to be displayed at the end of the + * text, with numeric indices to the original point in the text they + * appeared. Also makes an effort at identifying and handling absolute + * and relative links. + * + * @param string $link URL of the link + * @param string $display Part of the text to associate number with + */ + protected function _build_link_list( $link, $display ) + { + if (!$this->_do_links || empty($link)) { + return $display; + } + + // Ignored link types + if (preg_match('!^(javascript:|mailto:|#)!i', $link)) { + return $display; + } + + if (preg_match('!^([a-z][a-z0-9.+-]+:)!i', $link)) { + $url = $link; + } + else { + $url = $this->url; + if (substr($link, 0, 1) != '/') { + $url .= '/'; + } + $url .= "$link"; + } + + if (($index = array_search($url, $this->_link_list)) === false) { + $index = count($this->_link_list); + $this->_link_list[] = $url; + } + + return $display . ' [' . ($index+1) . ']'; + } + + /** + * Helper function for PRE body conversion. + * + * @param string HTML content + */ + protected function _convert_pre(&$text) + { + // get the content of PRE element + while (preg_match('/<pre[^>]*>(.*)<\/pre>/ismU', $text, $matches)) { + $this->pre_content = $matches[1]; + + // Run our defined tags search-and-replace with callback + $this->pre_content = preg_replace_callback($this->callback_search, + array($this, 'tags_preg_callback'), $this->pre_content); + + // convert the content + $this->pre_content = sprintf('<div><br>%s<br></div>', + preg_replace($this->pre_search, $this->pre_replace, $this->pre_content)); + + // replace the content (use callback because content can contain $0 variable) + $text = preg_replace_callback('/<pre[^>]*>.*<\/pre>/ismU', + array($this, 'pre_preg_callback'), $text, 1); + + // free memory + $this->pre_content = ''; + } + } + + /** + * Helper function for BLOCKQUOTE body conversion. + * + * @param string HTML content + */ + protected function _convert_blockquotes(&$text) + { + if (preg_match_all('/<\/*blockquote[^>]*>/i', $text, $matches, PREG_OFFSET_CAPTURE)) { + $level = 0; + $diff = 0; + foreach ($matches[0] as $m) { + if ($m[0][0] == '<' && $m[0][1] == '/') { + $level--; + if ($level < 0) { + $level = 0; // malformed HTML: go to next blockquote + } + else if ($level > 0) { + // skip inner blockquote + } + else { + $end = $m[1]; + $len = $end - $taglen - $start; + // Get blockquote content + $body = substr($text, $start + $taglen - $diff, $len); + + // Set text width + $p_width = $this->width; + if ($this->width > 0) $this->width -= 2; + // Convert blockquote content + $body = trim($body); + $this->_converter($body); + // Add citation markers and create PRE block + $body = preg_replace('/((^|\n)>*)/', '\\1> ', trim($body)); + $body = '<pre>' . htmlspecialchars($body) . '</pre>'; + // Re-set text width + $this->width = $p_width; + // Replace content + $text = substr($text, 0, $start - $diff) + . $body . substr($text, $end + strlen($m[0]) - $diff); + + $diff = $len + $taglen + strlen($m[0]) - strlen($body); + unset($body); + } + } + else { + if ($level == 0) { + $start = $m[1]; + $taglen = strlen($m[0]); + } + $level ++; + } + } + } + } + + /** + * Callback function for preg_replace_callback use. + * + * @param array PREG matches + * @return string + */ + public function tags_preg_callback($matches) + { + switch (strtolower($matches[1])) { + case 'b': + case 'strong': + return $this->_toupper($matches[3]); + case 'th': + return $this->_toupper("\t\t". $matches[3] ."\n"); + case 'h': + return $this->_toupper("\n\n". $matches[3] ."\n\n"); + case 'a': + // Remove spaces in URL (#1487805) + $url = str_replace(' ', '', $matches[3]); + return $this->_build_link_list($url, $matches[4]); + } + } + + /** + * Callback function for preg_replace_callback use in PRE content handler. + * + * @param array PREG matches + * @return string + */ + public function pre_preg_callback($matches) + { + return $this->pre_content; + } + + /** + * Strtoupper function with HTML tags and entities handling. + * + * @param string $str Text to convert + * @return string Converted text + */ + private function _toupper($str) + { + // string can containg HTML tags + $chunks = preg_split('/(<[^>]*>)/', $str, null, PREG_SPLIT_NO_EMPTY | PREG_SPLIT_DELIM_CAPTURE); + + // convert toupper only the text between HTML tags + foreach ($chunks as $idx => $chunk) { + if ($chunk[0] != '<') { + $chunks[$idx] = $this->_strtoupper($chunk); + } + } + + return implode($chunks); + } + + /** + * Strtoupper multibyte wrapper function with HTML entities handling. + * + * @param string $str Text to convert + * @return string Converted text + */ + private function _strtoupper($str) + { + $str = html_entity_decode($str, ENT_COMPAT, $this->charset); + $str = mb_strtoupper($str); + $str = htmlspecialchars($str, ENT_COMPAT, $this->charset); + + return $str; + } +} diff --git a/program/lib/Roundcube/rcube_message.php b/program/lib/Roundcube/rcube_message.php index c45dbfcd0..9fea8382a 100644 --- a/program/lib/Roundcube/rcube_message.php +++ b/program/lib/Roundcube/rcube_message.php @@ -272,7 +272,7 @@ class rcube_message $out = $this->get_part_content($mime_id); // create instance of html2text class - $txt = new html2text($out); + $txt = new rcube_html2text($out); return $txt->get_text(); } } diff --git a/program/lib/Roundcube/rcube_spellchecker.php b/program/lib/Roundcube/rcube_spellchecker.php index e9e1ceb0f..3d4d3a3d6 100644 --- a/program/lib/Roundcube/rcube_spellchecker.php +++ b/program/lib/Roundcube/rcube_spellchecker.php @@ -443,7 +443,7 @@ class rcube_spellchecker private function html2text($text) { - $h2t = new html2text($text, false, true, 0); + $h2t = new rcube_html2text($text, false, true, 0); return $h2t->get_text(); } diff --git a/program/lib/Roundcube/rcube_string_replacer.php b/program/lib/Roundcube/rcube_string_replacer.php index 0fe982b26..6b289886b 100644 --- a/program/lib/Roundcube/rcube_string_replacer.php +++ b/program/lib/Roundcube/rcube_string_replacer.php @@ -36,7 +36,7 @@ class rcube_string_replacer // Support unicode/punycode in top-level domain part $utf_domain = '[^?&@"\'\\/()\s\r\t\n]+\\.?([^\\x00-\\x2f\\x3b-\\x40\\x5b-\\x60\\x7b-\\x7f]{2,}|xn--[a-zA-Z0-9]{2,})'; $url1 = '.:;,'; - $url2 = 'a-zA-Z0-9%=#$@+?!&\\/_~\\[\\]{}\*-'; + $url2 = 'a-zA-Z0-9%=#$@+?!&\\/_~\\[\\]\\(\\){}\*-'; $this->link_pattern = "/([\w]+:\/\/|\W[Ww][Ww][Ww]\.|^[Ww][Ww][Ww]\.)($utf_domain([$url1]?[$url2]+)*)/"; $this->mailto_pattern = "/(" @@ -161,6 +161,9 @@ class rcube_string_replacer // "http://example.com/?a[b]=c". However we need to handle // properly situation when a bracket is placed at the end // of the link e.g. "[http://example.com]" + // Yes, this is not perfect handles correctly only paired characters + // but it should work for common cases + if (preg_match('/(\\[|\\])/', $url)) { $in = false; for ($i=0, $len=strlen($url); $i<$len; $i++) { @@ -182,6 +185,28 @@ class rcube_string_replacer } } + // Do the same for parentheses + if (preg_match('/(\\(|\\))/', $url)) { + $in = false; + for ($i=0, $len=strlen($url); $i<$len; $i++) { + if ($url[$i] == '(') { + if ($in) + break; + $in = true; + } + else if ($url[$i] == ')') { + if (!$in) + break; + $in = false; + } + } + + if ($i < $len) { + $suffix = substr($url, $i); + $url = substr($url, 0, $i); + } + } + return $suffix; } } diff --git a/program/lib/Roundcube/rcube_washtml.php b/program/lib/Roundcube/rcube_washtml.php new file mode 100644 index 000000000..715c46047 --- /dev/null +++ b/program/lib/Roundcube/rcube_washtml.php @@ -0,0 +1,451 @@ +<?php + +/** + +-----------------------------------------------------------------------+ + | This file is part of the Roundcube Webmail client | + | Copyright (C) 2008-2012, The Roundcube Dev Team | + | | + | Licensed under the GNU General Public License version 3 or | + | any later version with exceptions for skins & plugins. | + | See the README file for a full license statement. | + | | + | PURPOSE: | + | Utility class providing HTML sanityzer (based on Washtml class) | + +-----------------------------------------------------------------------+ + | Author: Thomas Bruederli <roundcube@gmail.com> | + | Author: Aleksander Machniak <alec@alec.pl> | + | Author: Frederic Motte <fmotte@ubixis.com> | + +-----------------------------------------------------------------------+ + */ + +/** + * Washtml, a HTML sanityzer. + * + * Copyright (c) 2007 Frederic Motte <fmotte@ubixis.com> + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR + * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES + * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. + * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, + * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT + * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF + * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * OVERVIEW: + * + * Wahstml take an untrusted HTML and return a safe html string. + * + * SYNOPSIS: + * + * $washer = new washtml($config); + * $washer->wash($html); + * It return a sanityzed string of the $html parameter without html and head tags. + * $html is a string containing the html code to wash. + * $config is an array containing options: + * $config['allow_remote'] is a boolean to allow link to remote images. + * $config['blocked_src'] string with image-src to be used for blocked remote images + * $config['show_washed'] is a boolean to include washed out attributes as x-washed + * $config['cid_map'] is an array where cid urls index urls to replace them. + * $config['charset'] is a string containing the charset of the HTML document if it is not defined in it. + * $washer->extlinks is a reference to a boolean that is set to true if remote images were removed. (FE: show remote images link) + * + * INTERNALS: + * + * Only tags and attributes in the static lists $html_elements and $html_attributes + * are kept, inline styles are also filtered: all style identifiers matching + * /[a-z\-]/i are allowed. Values matching colors, sizes, /[a-z\-]/i and safe + * urls if allowed and cid urls if mapped are kept. + * + * Roundcube Changes: + * - added $block_elements + * - changed $ignore_elements behaviour + * - added RFC2397 support + * - base URL support + * - invalid HTML comments removal before parsing + * - "fixing" unitless CSS values for XHTML output + * - base url resolving + */ + +/** + * Utility class providing HTML sanityzer + * + * @package Framework + * @subpackage Utils + */ +class rcube_washtml +{ + /* Allowed HTML elements (default) */ + static $html_elements = array('a', 'abbr', 'acronym', 'address', 'area', 'b', + 'basefont', 'bdo', 'big', 'blockquote', 'br', 'caption', 'center', + 'cite', 'code', 'col', 'colgroup', 'dd', 'del', 'dfn', 'dir', 'div', 'dl', + 'dt', 'em', 'fieldset', 'font', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'hr', 'i', + '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', + // form elements + 'button', 'input', 'textarea', 'select', 'option', 'optgroup' + ); + + /* Ignore these HTML tags and their content */ + static $ignore_elements = array('script', 'applet', 'embed', 'object', 'style'); + + /* Allowed HTML attributes */ + static $html_attribs = array('name', 'class', 'title', 'alt', 'width', 'height', + 'align', 'nowrap', 'col', 'row', 'id', 'rowspan', 'colspan', 'cellspacing', + 'cellpadding', 'valign', 'bgcolor', 'color', 'border', 'bordercolorlight', + 'bordercolordark', 'face', 'marginwidth', 'marginheight', 'axis', 'border', + 'abbr', 'char', 'charoff', 'clear', 'compact', 'coords', 'vspace', 'hspace', + 'cellborder', 'size', 'lang', 'dir', 'usemap', 'shape', 'media', + // attributes of form elements + 'type', 'rows', 'cols', 'disabled', 'readonly', 'checked', 'multiple', 'value' + ); + + /* Block elements which could be empty but cannot be returned in short form (<tag />) */ + static $block_elements = array('div', 'p', 'pre', 'blockquote', 'a', 'font', 'center', + 'table', 'ul', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'ol', 'dl', 'strong', + 'i', 'b', 'u', 'span', + ); + + /* State for linked objects in HTML */ + public $extlinks = false; + + /* Current settings */ + private $config = array(); + + /* Registered callback functions for tags */ + private $handlers = array(); + + /* Allowed HTML elements */ + private $_html_elements = array(); + + /* Ignore these HTML tags but process their content */ + private $_ignore_elements = array(); + + /* Block elements which could be empty but cannot be returned in short form (<tag />) */ + private $_block_elements = array(); + + /* Allowed HTML attributes */ + private $_html_attribs = array(); + + + /** + * Class constructor + */ + public function __construct($p = array()) + { + $this->_html_elements = array_flip((array)$p['html_elements']) + array_flip(self::$html_elements) ; + $this->_html_attribs = array_flip((array)$p['html_attribs']) + array_flip(self::$html_attribs); + $this->_ignore_elements = array_flip((array)$p['ignore_elements']) + array_flip(self::$ignore_elements); + $this->_block_elements = array_flip((array)$p['block_elements']) + array_flip(self::$block_elements); + + unset($p['html_elements'], $p['html_attribs'], $p['ignore_elements'], $p['block_elements']); + + $this->config = $p + array('show_washed' => true, 'allow_remote' => false, 'cid_map' => array()); + } + + /** + * Register a callback function for a certain tag + */ + public function add_callback($tagName, $callback) + { + $this->handlers[$tagName] = $callback; + } + + /** + * Check CSS style + */ + private function wash_style($style) + { + $s = ''; + + foreach (explode(';', $style) as $declaration) { + if (preg_match('/^\s*([a-z\-]+)\s*:\s*(.*)\s*$/i', $declaration, $match)) { + $cssid = $match[1]; + $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).')'; + } + else { + $this->extlinks = true; + } + } + else if (preg_match('/^data:.+/i', $match[2])) { // RFC2397 + $value .= ' url('.htmlspecialchars($match[2], ENT_QUOTES).')'; + } + } + else { + // whitelist ? + $value .= ' ' . $match[0]; + + // #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]) + ) { + $value .= 'px'; + } + } + + $str = substr($str, strlen($match[0])); + } + + if (isset($value[0])) { + $s .= ($s?' ':'') . $cssid . ':' . $value . ';'; + } + } + } + + return $s; + } + + /** + * Take a node and return allowed attributes and check values + */ + private function wash_attribs($node) + { + $t = ''; + $washed = ''; + + foreach ($node->attributes as $key => $plop) { + $key = strtolower($key); + $value = $node->getAttribute($key); + + if (isset($this->_html_attribs[$key]) || + ($key == 'href' && !preg_match('!^(javascript|vbscript|data:text)!i', $value) + && preg_match('!^([a-z][a-z0-9.+-]+:|//|#).+!i', $value)) + ) { + $t .= ' ' . $key . '="' . htmlspecialchars($value, ENT_QUOTES) . '"'; + } + else if ($key == 'style' && ($style = $this->wash_style($value))) { + $quot = strpos($style, '"') !== false ? "'" : '"'; + $t .= ' style=' . $quot . $style . $quot; + } + else if ($key == 'background' || ($key == 'src' && strtolower($node->tagName) == 'img')) { //check tagName anyway + if (($src = $this->config['cid_map'][$value]) + || ($src = $this->config['cid_map'][$this->config['base_url'].$value]) + ) { + $t .= ' ' . $key . '="' . htmlspecialchars($src, ENT_QUOTES) . '"'; + } + else if (preg_match('/^(http|https|ftp):.+/i', $value)) { + if ($this->config['allow_remote']) { + $t .= ' ' . $key . '="' . htmlspecialchars($value, ENT_QUOTES) . '"'; + } + else { + $this->extlinks = true; + if ($this->config['blocked_src']) { + $t .= ' ' . $key . '="' . htmlspecialchars($this->config['blocked_src'], ENT_QUOTES) . '"'; + } + } + } + else if (preg_match('/^data:.+/i', $value)) { // RFC2397 + $t .= ' ' . $key . '="' . htmlspecialchars($value, ENT_QUOTES) . '"'; + } + } + else { + $washed .= ($washed ? ' ' : '') . $key; + } + } + + return $t . ($washed && $this->config['show_washed'] ? ' x-washed="'.$washed.'"' : ''); + } + + /** + * The main loop that recurse on a node tree. + * It output only allowed tags with allowed attributes + * and allowed inline styles + */ + private function dumpHtml($node) + { + if (!$node->hasChildNodes()) { + return ''; + } + + $node = $node->firstChild; + $dump = ''; + + do { + switch($node->nodeType) { + case XML_ELEMENT_NODE: //Check element + $tagName = strtolower($node->tagName); + if ($callback = $this->handlers[$tagName]) { + $dump .= call_user_func($callback, $tagName, + $this->wash_attribs($node), $this->dumpHtml($node), $this); + } + else if (isset($this->_html_elements[$tagName])) { + $content = $this->dumpHtml($node); + $dump .= '<' . $tagName . $this->wash_attribs($node) . + ($content != '' || isset($this->_block_elements[$tagName]) ? ">$content</$tagName>" : ' />'); + } + else if (isset($this->_ignore_elements[$tagName])) { + $dump .= '<!-- ' . htmlspecialchars($tagName, ENT_QUOTES) . ' not allowed -->'; + } + else { + $dump .= '<!-- ' . htmlspecialchars($tagName, ENT_QUOTES) . ' ignored -->'; + $dump .= $this->dumpHtml($node); // ignore tags not its content + } + break; + + case XML_CDATA_SECTION_NODE: + $dump .= $node->nodeValue; + break; + + case XML_TEXT_NODE: + $dump .= htmlspecialchars($node->nodeValue); + break; + + case XML_HTML_DOCUMENT_NODE: + $dump .= $this->dumpHtml($node); + break; + + case XML_DOCUMENT_TYPE_NODE: + break; + + default: + $dump . '<!-- node type ' . $node->nodeType . ' -->'; + } + } while($node = $node->nextSibling); + + return $dump; + } + + /** + * Main function, give it untrusted HTML, tell it if you allow loading + * remote images and give it a map to convert "cid:" urls. + */ + public function wash($html) + { + // Charset seems to be ignored (probably if defined in the HTML document) + $node = new DOMDocument('1.0', $this->config['charset']); + $this->extlinks = false; + + $html = $this->cleanup($html); + + // Find base URL for images + if (preg_match('/<base\s+href=[\'"]*([^\'"]+)/is', $html, $matches)) { + $this->config['base_url'] = $matches[1]; + } + else { + $this->config['base_url'] = ''; + } + + @$node->loadHTML($html); + return $this->dumpHtml($node); + } + + /** + * Getter for config parameters + */ + public function get_config($prop) + { + return $this->config[$prop]; + } + + /** + * Clean HTML input + */ + private function cleanup($html) + { + // 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 + ); + + $html_replace = array( + '\\1'.' '.'\\3', + '', + '', + '<html>', + ); + $html = preg_replace($html_search, $html_replace, trim($html)); + + // PCRE errors handling (#1486856), should we use something like for every preg_* use? + if ($html === null && ($preg_error = preg_last_error()) != PREG_NO_ERROR) { + $errstr = "Could not clean up HTML message! PCRE Error: $preg_error."; + + if ($preg_error == PREG_BACKTRACK_LIMIT_ERROR) { + $errstr .= " Consider raising pcre.backtrack_limit!"; + } + if ($preg_error == PREG_RECURSION_LIMIT_ERROR) { + $errstr .= " Consider raising pcre.recursion_limit!"; + } + + rcube::raise_error(array('code' => 620, 'type' => 'php', + 'line' => __LINE__, 'file' => __FILE__, + 'message' => $errstr), true, false); + return ''; + } + + // fix (unknown/malformed) HTML tags before "wash" + $html = preg_replace_callback('/(<[\/]*)([^\s>]+)/', array($this, 'html_tag_callback'), $html); + + // Remove invalid HTML comments (#1487759) + // Don't remove valid conditional comments + $html = preg_replace('/<!--[^->[\n]*>/', '', $html); + + // turn relative into absolute urls + $html = self::resolve_base($html); + + return $html; + } + + /** + * Callback function for HTML tags fixing + */ + public static function html_tag_callback($matches) + { + $tagname = $matches[2]; + $tagname = preg_replace(array( + '/:.*$/', // Microsoft's Smart Tags <st1:xxxx> + '/[^a-z0-9_\[\]\!-]/i', // forbidden characters + ), '', $tagname); + + return $matches[1] . $tagname; + } + + /** + * Convert all relative URLs according to a <base> in HTML + */ + public static function resolve_base($body) + { + // check for <base href=...> + if (preg_match('!(<base.*href=["\']?)([hftps]{3,5}://[a-z0-9/.%-]+)!i', $body, $regs)) { + $replacer = new rcube_base_replacer($regs[2]); + $body = $replacer->replace($body); + } + + return $body; + } +} + |