filter/urltolink/filter.php

   1 <?php
   2
   3 // This file is part of Moodle - http://moodle.org/
   4 //
   5 // Moodle is free software: you can redistribute it and/or modify
   6 // it under the terms of the GNU General Public License as published by
   7 // the Free Software Foundation, either version 3 of the License, or
   8 // (at your option) any later version.
   9 //
  10 // Moodle is distributed in the hope that it will be useful,
  11 // but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13 // GNU General Public License for more details.
  14 //
  15 // You should have received a copy of the GNU General Public License
  16 // along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
  17
  18 /**
  19  * Filter converting URLs in the text to HTML links
  20  *
  21  * @package    filter
  22  * @subpackage urltolink
  23  * @copyright  2010 David Mudrak <david@moodle.com>
  24  * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
  25  */
  26
  27 defined('MOODLE_INTERNAL') || die();
  28
  29 class filter_urltolink extends moodle_text_filter {
  30
  31     /**
  32      * @var array global configuration for this filter
  33      *
  34      * This might be eventually moved into parent class if we found it
  35      * useful for other filters, too.
  36      */
  37     protected static $globalconfig;
  38
  39     /**
  40      * Apply the filter to the text
  41      *
  42      * @see filter_manager::apply_filter_chain()
  43      * @param string $text to be processed by the text
  44      * @param array $options filter options
  45      * @return string text after processing
  46      */
  47     public function filter($text, array $options = array()) {
  48         if (!isset($options['originalformat'])) {
  49             // if the format is not specified, we are probably called by {@see format_string()}
  50             // in that case, it would be dangerous to replace URL with the link because it could
  51             // be stripped. therefore, we do nothing
  52             return $text;
  53         }
  54         if (in_array($options['originalformat'], explode(',', get_config('filter_urltolink', 'formats')))) {
  55             $this->convert_urls_into_links($text);
  56         }
  57         return $text;
  58     }
  59
  60     ////////////////////////////////////////////////////////////////////////////
  61     // internal implementation starts here
  62     ////////////////////////////////////////////////////////////////////////////
  63
  64     /**
  65      * Given some text this function converts any URLs it finds into HTML links
  66      *
  67      * @param string $text Passed in by reference. The string to be searched for urls.
  68      */
  69     protected function convert_urls_into_links(&$text) {
  70         //I've added img tags to this list of tags to ignore.
  71         //See MDL-21168 for more info. A better way to ignore tags whether or not
  72         //they are escaped partially or completely would be desirable. For example:
  73         //<a href="blah">
  74         //&lt;a href="blah"&gt;
  75         //&lt;a href="blah">
  76         $filterignoretagsopen  = array('<a\s[^>]+?>');
  77         $filterignoretagsclose = array('</a>');
  78         filter_save_ignore_tags($text,$filterignoretagsopen,$filterignoretagsclose,$ignoretags);
  79
  80         // Check if we support unicode modifiers in regular expressions. Cache it.
  81         // TODO: this check should be a environment requirement in Moodle 2.0, as far as unicode
  82         // chars are going to arrive to URLs officially really soon (2010?)
  83         // Original RFC regex from: http://www.bytemycode.com/snippets/snippet/796/
  84         // Various ideas from: http://alanstorm.com/url_regex_explained
  85         // Unicode check, negative assertion and other bits from Moodle.
  86         static $unicoderegexp;
  87         if (!isset($unicoderegexp)) {
  88             $unicoderegexp = @preg_match('/\pL/u', 'a'); // This will fail silently, returning false,
  89         }
  90
  91         // TODO MDL-21296 - use of unicode modifiers may cause a timeout
  92         $urlstart = '(?:http(s)?://|(?<!://)(www\.))';
  93         $domainsegment = '(?:[\pLl0-9][\pLl0-9-]*[\pLl0-9]|[\pLl0-9])';
  94         $numericip = '(?:(?:[0-9]{1,3}\.){3}[0-9]{1,3})';
  95         $port = '(?::\d*)';
  96         $pathchar = '(?:[\pL0-9\.!$&\'\(\)*+,;=_~:@-]|%[a-f0-9]{2})';
  97         $path = "(?:/$pathchar*)*";
  98         $querystring = '(?:\?(?:[\pL0-9\.!$&\'\(\)*+,;=_~:@/?-]|%[a-fA-F0-9]{2})*)';
  99         $fragment = '(?:\#(?:[\pL0-9\.!$&\'\(\)*+,;=_~:@/?-]|%[a-fA-F0-9]{2})*)';
 100
 101         // Lookbehind assertions.
 102         // Is not HTML attribute or CSS URL property. Unfortunately legit text like "url(http://...)" will not be a link.
 103         $lookbehindend = "(?<![]),.;])";
 104
 105         $regex = "$urlstart((?:$domainsegment\.)+$domainsegment|$numericip)" .
 106                 "($port?$path$querystring?$fragment?)$lookbehindend";
 107         if ($unicoderegexp) {
 108             $regex = '#' . $regex . '#ui';
 109         } else {
 110             $regex = '#' . preg_replace(array('\pLl', '\PL'), 'a-z', $regex) . '#i';
 111         }
 112
 113         // Locate any HTML tags.
 114         $matches = preg_split('/(<[^<|>]*>)/i', $text, -1, PREG_SPLIT_NO_EMPTY | PREG_SPLIT_DELIM_CAPTURE);
 115
 116         // Iterate through the tokenized text to handle chunks (html and content).
 117         foreach ($matches as $idx => $chunk) {
 118             // Nothing to do. We skip completely any html chunk.
 119             if (strpos(trim($chunk), '<') === 0) {
 120                 continue;
 121             }
 122
 123             // Nothing to do. We skip any content chunk having any of these attributes.
 124             if (preg_match('#(background=")|(action=")|(style="background)|(href=")|(src=")|(url [(])#', $chunk)) {
 125                 continue;
 126             }
 127
 128             // Arrived here, we want to process every word in this chunk.
 129             $text = $chunk;
 130             $words = explode(' ', $text);
 131
 132             foreach ($words as $idx2 => $word) {
 133                 // ReDoS protection. Stop processing if a word is too large.
 134                 if (strlen($word) < 4096) {
 135                     $words[$idx2] = preg_replace($regex, '<a href="http$1://$2$3$4" class="_blanktarget">$0</a>', $word);
 136                 }
 137             }
 138             $text = implode(' ', $words);
 139
 140             // Copy the result back to the array.
 141             $matches[$idx] = $text;
 142         }
 143
 144         $text = implode('', $matches);
 145
 146         if (!empty($ignoretags)) {
 147             $ignoretags = array_reverse($ignoretags); /// Reversed so "progressive" str_replace() will solve some nesting problems.
 148             $text = str_replace(array_keys($ignoretags),$ignoretags,$text);
 149         }
 150
 151         if (get_config('filter_urltolink', 'embedimages')) {
 152             // now try to inject the images, this code was originally in the mediapluing filter
 153             // this may be useful only if somebody relies on the fact the links in FORMAT_MOODLE get converted
 154             // to URLs which in turn change to real images
 155             $search = '/<a href="([^"]+\.(jpg|png|gif))" class="_blanktarget">([^>]*)<\/a>/is';
 156             $text = preg_replace_callback($search, 'filter_urltolink_img_callback', $text);
 157         }
 158     }
 159 }
 160
 161
 162 /**
 163  * Change links to images into embedded images.
 164  *
 165  * This plugin is intended for automatic conversion of image URLs when FORMAT_MOODLE used.
 166  *
 167  * @param  $link
 168  * @return string
 169  */
 170 function filter_urltolink_img_callback($link) {
 171     if ($link[1] !== $link[3]) {
 172         // this is not a link created by this filter, because the url does not match the text
 173         return $link[0];
 174     }
 175     return '<img class="filter_urltolink_image" alt="" src="'.$link[1].'" />';
 176 }
 177