search/documents/physical_doc.php

   1 <?php
   2 /**
   3 * Global Search Engine for Moodle
   4 *
   5 * @package search
   6 * @category core
   7 * @subpackage document_wrappers
   8 * @author Valery Fremaux [valery.fremaux@club-internet.fr] > 1.8
   9 * @date 2008/03/31
  10 * @license http://www.gnu.org/copyleft/gpl.html GNU Public License
  11 *
  12 * this is a format handler for getting text out of a proprietary binary format
  13 * so it can be indexed by Lucene search engine
  14 */
  15
  16 /**
  17 * MS Word extractor
  18 * @param object $resource
  19 * @uses CFG, USER
  20 */
  21 function get_text_for_indexing_doc(&$resource){
  22     global $CFG, $USER;
  23
  24     // SECURITY : do not allow non admin execute anything on system !!
  25     if (!isadmin($USER->id)) return;
  26
  27     $moodleroot = (@$CFG->block_search_usemoodleroot) ? "{$CFG->dirroot}/" : '' ;
  28
  29     // just call pdftotext over stdout and capture the output
  30     if (!empty($CFG->block_search_word_to_text_cmd)){
  31         if (!file_exists("{$moodleroot}{$CFG->block_search_word_to_text_cmd}")){
  32             mtrace('Error with MSWord to text converter command : exectuable not found.');
  33         }
  34         else{
  35             $file = escapeshellarg($CFG->dataroot.'/'.$resource->course.'/'.$resource->reference);
  36             $command = trim($CFG->block_search_word_to_text_cmd);
  37             $text_converter_cmd = "{$moodleroot}{$command} $file";
  38             if ($CFG->block_search_word_to_text_env){
  39                 putenv($CFG->block_search_word_to_text_env);
  40             }
  41             mtrace("Executing : $text_converter_cmd");
  42             $result = shell_exec($text_converter_cmd);
  43             if ($result){
  44                 return mb_convert_encoding($result, 'UTF8', 'auto');
  45             }
  46             else{
  47                 mtrace('Error with MSWord to text converter command : execution failed. ');
  48                 return '';
  49             }
  50         }
  51     }
  52     else {
  53         mtrace('Error with MSWord to text converter command : command not set up. Execute once search block configuration.');
  54         return '';
  55     }
  56 }
  57 ?>