drupal/modules/search/search.module

<?php
// $Id$

/**
 * @file
 * Enables site-wide keyword searching.
 */

/**
 * Matches Unicode character classes to exclude from the search index.
 *
 * @see http://unicode.org/glossary
 *
 * Characters with the following General_category (gc) property values are
 * excluded from the search index. Also, they are used as word boundaries.
 * While this does not fully conform to the  Word Boundaries algorithm
 * described in http://unicode.org/reports/tr29, as PCRE does not contain the
 * Word_Break property table, this simpler algorithm has to do.
 * - Cc, Cf, Cn, Co, Cs: Other.
 * - Pc, Pd, Pe, Pf, Pi, Po, Ps: Punctuation.
 * - Sc, Sk, Sm, So: Symbols.
 * - Zl, Zp, Zs: Separators.
 *
 * Consequently, the index only contains characters with the following
 * General_category (gc) property values:
 * - Ll, Lm, Lo, Lt, Lu: Letters.
 * - Mc, Me, Mn: Combining Marks.
 * - Nd, Nl, No: Numbers.
 *
 * Note that the PCRE property matcher is not used because we wanted to be
 * compatible with Unicode 5.2.0 regardless of the PCRE version used (and any
 * bugs in PCRE property tables).
 */
define('PREG_CLASS_SEARCH_EXCLUDE',
  '\x{0}-\x{2F}\x{3A}-\x{40}\x{5B}-\x{60}\x{7B}-\x{A9}\x{AB}-\x{B1}\x{B4}' .
  '\x{B6}-\x{B8}\x{BB}\x{BF}\x{D7}\x{F7}\x{2C2}-\x{2C5}\x{2D2}-\x{2DF}' .   
  '\x{2E5}-\x{2EB}\x{2ED}\x{2EF}-\x{2FF}\x{375}\x{37E}-\x{385}\x{387}\x{3F6}' .
  '\x{482}\x{55A}-\x{55F}\x{589}-\x{58A}\x{5BE}\x{5C0}\x{5C3}\x{5C6}' .        
  '\x{5F3}-\x{60F}\x{61B}-\x{61F}\x{66A}-\x{66D}\x{6D4}\x{6DD}\x{6E9}' .       
  '\x{6FD}-\x{6FE}\x{700}-\x{70F}\x{7F6}-\x{7F9}\x{830}-\x{83E}' .             
  '\x{964}-\x{965}\x{970}\x{9F2}-\x{9F3}\x{9FA}-\x{9FB}\x{AF1}\x{B70}' .       
  '\x{BF3}-\x{BFA}\x{C7F}\x{CF1}-\x{CF2}\x{D79}\x{DF4}\x{E3F}\x{E4F}' .        
  '\x{E5A}-\x{E5B}\x{F01}-\x{F17}\x{F1A}-\x{F1F}\x{F34}\x{F36}\x{F38}' .
  '\x{F3A}-\x{F3D}\x{F85}\x{FBE}-\x{FC5}\x{FC7}-\x{FD8}\x{104A}-\x{104F}' .
  '\x{109E}-\x{109F}\x{10FB}\x{1360}-\x{1368}\x{1390}-\x{1399}\x{1400}' .
  '\x{166D}-\x{166E}\x{1680}\x{169B}-\x{169C}\x{16EB}-\x{16ED}' .
  '\x{1735}-\x{1736}\x{17B4}-\x{17B5}\x{17D4}-\x{17D6}\x{17D8}-\x{17DB}' .
  '\x{1800}-\x{180A}\x{180E}\x{1940}-\x{1945}\x{19DE}-\x{19FF}' .
  '\x{1A1E}-\x{1A1F}\x{1AA0}-\x{1AA6}\x{1AA8}-\x{1AAD}\x{1B5A}-\x{1B6A}' .
  '\x{1B74}-\x{1B7C}\x{1C3B}-\x{1C3F}\x{1C7E}-\x{1C7F}\x{1CD3}\x{1FBD}' .
  '\x{1FBF}-\x{1FC1}\x{1FCD}-\x{1FCF}\x{1FDD}-\x{1FDF}\x{1FED}-\x{1FEF}' .
  '\x{1FFD}-\x{206F}\x{207A}-\x{207E}\x{208A}-\x{208E}\x{20A0}-\x{20B8}' .
  '\x{2100}-\x{2101}\x{2103}-\x{2106}\x{2108}-\x{2109}\x{2114}' .
  '\x{2116}-\x{2118}\x{211E}-\x{2123}\x{2125}\x{2127}\x{2129}\x{212E}' .
  '\x{213A}-\x{213B}\x{2140}-\x{2144}\x{214A}-\x{214D}\x{214F}' .
  '\x{2190}-\x{244A}\x{249C}-\x{24E9}\x{2500}-\x{2775}\x{2794}-\x{2B59}' .
  '\x{2CE5}-\x{2CEA}\x{2CF9}-\x{2CFC}\x{2CFE}-\x{2CFF}\x{2E00}-\x{2E2E}' .
  '\x{2E30}-\x{3004}\x{3008}-\x{3020}\x{3030}\x{3036}-\x{3037}' .
  '\x{303D}-\x{303F}\x{309B}-\x{309C}\x{30A0}\x{30FB}\x{3190}-\x{3191}' .
  '\x{3196}-\x{319F}\x{31C0}-\x{31E3}\x{3200}-\x{321E}\x{322A}-\x{3250}' .
  '\x{3260}-\x{327F}\x{328A}-\x{32B0}\x{32C0}-\x{33FF}\x{4DC0}-\x{4DFF}' .
  '\x{A490}-\x{A4C6}\x{A4FE}-\x{A4FF}\x{A60D}-\x{A60F}\x{A673}\x{A67E}' .
  '\x{A6F2}-\x{A716}\x{A720}-\x{A721}\x{A789}-\x{A78A}\x{A828}-\x{A82B}' .
  '\x{A836}-\x{A839}\x{A874}-\x{A877}\x{A8CE}-\x{A8CF}\x{A8F8}-\x{A8FA}' .
  '\x{A92E}-\x{A92F}\x{A95F}\x{A9C1}-\x{A9CD}\x{A9DE}-\x{A9DF}' .
  '\x{AA5C}-\x{AA5F}\x{AA77}-\x{AA79}\x{AADE}-\x{AADF}\x{ABEB}' .
  '\x{D800}-\x{F8FF}\x{FB29}\x{FD3E}-\x{FD3F}\x{FDFC}-\x{FDFD}' .
  '\x{FE10}-\x{FE19}\x{FE30}-\x{FE6B}\x{FEFF}-\x{FF0F}\x{FF1A}-\x{FF20}' .
  '\x{FF3B}-\x{FF40}\x{FF5B}-\x{FF65}\x{FFE0}-\x{FFFD}');

/**
 * Matches all 'N' Unicode character classes (numbers)
 */
define('PREG_CLASS_NUMBERS',
  '\x{30}-\x{39}\x{b2}\x{b3}\x{b9}\x{bc}-\x{be}\x{660}-\x{669}\x{6f0}-\x{6f9}' .
  '\x{966}-\x{96f}\x{9e6}-\x{9ef}\x{9f4}-\x{9f9}\x{a66}-\x{a6f}\x{ae6}-\x{aef}' .
  '\x{b66}-\x{b6f}\x{be7}-\x{bf2}\x{c66}-\x{c6f}\x{ce6}-\x{cef}\x{d66}-\x{d6f}' .
  '\x{e50}-\x{e59}\x{ed0}-\x{ed9}\x{f20}-\x{f33}\x{1040}-\x{1049}\x{1369}-' .
  '\x{137c}\x{16ee}-\x{16f0}\x{17e0}-\x{17e9}\x{17f0}-\x{17f9}\x{1810}-\x{1819}' .
  '\x{1946}-\x{194f}\x{2070}\x{2074}-\x{2079}\x{2080}-\x{2089}\x{2153}-\x{2183}' .
  '\x{2460}-\x{249b}\x{24ea}-\x{24ff}\x{2776}-\x{2793}\x{3007}\x{3021}-\x{3029}' .
  '\x{3038}-\x{303a}\x{3192}-\x{3195}\x{3220}-\x{3229}\x{3251}-\x{325f}\x{3280}-' .
  '\x{3289}\x{32b1}-\x{32bf}\x{ff10}-\x{ff19}');

/**
 * Matches all 'P' Unicode character classes (punctuation)
 */
define('PREG_CLASS_PUNCTUATION',
  '\x{21}-\x{23}\x{25}-\x{2a}\x{2c}-\x{2f}\x{3a}\x{3b}\x{3f}\x{40}\x{5b}-\x{5d}' .
  '\x{5f}\x{7b}\x{7d}\x{a1}\x{ab}\x{b7}\x{bb}\x{bf}\x{37e}\x{387}\x{55a}-\x{55f}' .
  '\x{589}\x{58a}\x{5be}\x{5c0}\x{5c3}\x{5f3}\x{5f4}\x{60c}\x{60d}\x{61b}\x{61f}' .
  '\x{66a}-\x{66d}\x{6d4}\x{700}-\x{70d}\x{964}\x{965}\x{970}\x{df4}\x{e4f}' .
  '\x{e5a}\x{e5b}\x{f04}-\x{f12}\x{f3a}-\x{f3d}\x{f85}\x{104a}-\x{104f}\x{10fb}' .
  '\x{1361}-\x{1368}\x{166d}\x{166e}\x{169b}\x{169c}\x{16eb}-\x{16ed}\x{1735}' .
  '\x{1736}\x{17d4}-\x{17d6}\x{17d8}-\x{17da}\x{1800}-\x{180a}\x{1944}\x{1945}' .
  '\x{2010}-\x{2027}\x{2030}-\x{2043}\x{2045}-\x{2051}\x{2053}\x{2054}\x{2057}' .
  '\x{207d}\x{207e}\x{208d}\x{208e}\x{2329}\x{232a}\x{23b4}-\x{23b6}\x{2768}-' .
  '\x{2775}\x{27e6}-\x{27eb}\x{2983}-\x{2998}\x{29d8}-\x{29db}\x{29fc}\x{29fd}' .
  '\x{3001}-\x{3003}\x{3008}-\x{3011}\x{3014}-\x{301f}\x{3030}\x{303d}\x{30a0}' .
  '\x{30fb}\x{fd3e}\x{fd3f}\x{fe30}-\x{fe52}\x{fe54}-\x{fe61}\x{fe63}\x{fe68}' .
  '\x{fe6a}\x{fe6b}\x{ff01}-\x{ff03}\x{ff05}-\x{ff0a}\x{ff0c}-\x{ff0f}\x{ff1a}' .
  '\x{ff1b}\x{ff1f}\x{ff20}\x{ff3b}-\x{ff3d}\x{ff3f}\x{ff5b}\x{ff5d}\x{ff5f}-' .
  '\x{ff65}');

/**
 * Matches all CJK characters that are candidates for auto-splitting
 * (Chinese, Japanese, Korean).
 * Contains kana and BMP ideographs.
 */
define('PREG_CLASS_CJK', '\x{3041}-\x{30ff}\x{31f0}-\x{31ff}\x{3400}-\x{4db5}' .
'\x{4e00}-\x{9fbb}\x{f900}-\x{fad9}');

/**
 * Implements hook_help().
 */
function search_help($path, $arg) {
  switch ($path) {
    case 'admin/help#search':
      $output = '';
      $output .= '<h3>' . t('About') . '</h3>';
      $output .= '<p>' . t('The Search module provides the ability to index and search for content by exact keywords, and for users by username or e-mail. For more information, see the online handbook entry for <a href="@search-module">Search module</a>.', array('@search-module' => 'http://drupal.org/handbook/modules/search/', '@search' => url('search'))) . '</p>';
      $output .= '<h3>' . t('Uses') . '</h3>';
      $output .= '<dl>';
      $output .= '<dt>' . t('Searching content and users') . '</dt>';
      $output .= '<dd>' . t('Users with <em>Use search</em> permission can use the search block and <a href="@search">Search page</a>. Users with the <em>View published content</em> permission can search for content containing exact keywords. Users with the <em>View user profiles</em> permission can search for users containing the keyword anywhere in the user name, and users with the <em>Administer users</em> permission can search for users by email address. Additionally, users with <em>Use advanced search</em> permission can find content using more complex search methods and filtering by choosing the <em>Advanced search</em> option on the <a href="@search">Search page</a>.', array('@search' => url('search'))) . '</dd>';
      $output .= '<dt>' . t('Indexing content with cron') . '</dt>';
      $output .= '<dd>' . t('To provide keyword searching, the search engine maintains an index of words found in the content. To build and maintain this index, a correctly configured <a href="@cron">cron maintenance task</a> is required. Users with <em>Administer search</em> permission can further configure the cron settings on the <a href="@searchsettings">Search settings page</a>.', array('@cron' => 'http://drupal.org/cron', '@searchsettings' => url('admin/config/search/settings'))) . '</dd>';
      $output .= '<dt>' . t('Configuring search settings') . '</dt>';
      $output .= '<dd>' . t('Indexing behavior can be adjusted using the <a href="@searchsettings">Search settings page</a>. Users with <em>Administer search</em> permission can control settings such as the <em>Number of items to index per cron run</em>, <em>Indexing settings</em> (word length), <em>Active search modules</em>, and <em>Content ranking</em>, which lets you adjust the priority in which indexed content is returned in results.', array('@searchsettings' => url('admin/config/search/settings'))) . '</dd>';
      $output .= '<dt>' . t('Search block') . '</dt>';
      $output .= '<dd>' . t('The Search module includes a default <em>Search form</em> block, which can be enabled and configured on the <a href="@blocks">Blocks administration page</a>. The block is available to users with the <em>Search content</em> permission.', array('@blocks' => url('admin/structure/block'))) . '</dd>';
      $output .= '<dt>' . t('Extending Search module') . '</dt>';
      $output .= '<dd>' . t('By default, the Search module only supports exact keyword matching in content searches. You can modify this behavior by installing a language-specific stemming module for your language (such as <a href="http://drupal.org/project/porterstemmer">Porter Stemmer</a> for American English), which allows words such as walk, walking, and walked to be matched in the Search module. Another approach is to use a third-party search technology with stemming or partial word matching features built in, such as <a href="http://drupal.org/project/apachesolr">Apache Solr</a> or <a href="http://drupal.org/project/sphinx">Sphinx</a>. These and other <a href="@contrib-search">search-related contributed modules</a> can be downloaded by visiting Drupal.org.', array('@contrib-search' => 'http://drupal.org/project/modules?filters=tid%3A105')) . '</dd>';
      $output .= '</dl>';
      return $output;
    case 'admin/config/search/settings':
      return '<p>' . t('The search engine maintains an index of words found in your site\'s content. To build and maintain this index, a correctly configured <a href="@cron">cron maintenance task</a> is required. Indexing behavior can be adjusted using the settings below.', array('@cron' => url('admin/reports/status'))) . '</p>';
    case 'search#noresults':
      return t('<ul>
<li>Check if your spelling is correct.</li>
<li>Remove quotes around phrases to search for each word individually. <em>bike shed</em> will often show more results than <em>&quot;bike shed&quot;</em>.</li>
<li>Consider loosening your query with <em>OR</em>. <em>bike OR shed</em> will often show more results than <em>bike shed</em>.</li>
</ul>');
  }
}

/**
 * Implements hook_theme().
 */
function search_theme() {
  return array(
    'search_block_form' => array(
      'render element' => 'form',
      'template' => 'search-block-form',
    ),
    'search_result' => array(
      'variables' => array('result' => NULL, 'type' => NULL),
      'file' => 'search.pages.inc',
      'template' => 'search-result',
    ),
    'search_results' => array(
      'variables' => array('results' => NULL, 'type' => NULL),
      'file' => 'search.pages.inc',
      'template' => 'search-results',
    ),
    'search_results_listing' => array(
      'variables' => array('title' => NULL, 'content' => NULL),
    ),
  );
}

/**
 * Implements hook_permission().
 */
function search_permission() {
  return array(
    'administer search' => array(
      'title' => t('Administer search'),
    ),
    'search content' => array(
      'title' => t('Use search'),
    ),
    'use advanced search' => array(
      'title' => t('Use advanced search'),
    ),
  );
}

/**
 * Implements hook_block_info().
 */
function search_block_info() {
  $blocks['form']['info'] = t('Search form');
  // Not worth caching.
  $blocks['form']['cache'] = DRUPAL_NO_CACHE;
  return $blocks;
}

/**
 * Implements hook_block_view().
 */
function search_block_view($delta = '') {
  if (user_access('search content')) {
    $block['content'] = drupal_get_form('search_block_form');
    return $block;
  }
}

/**
 * Implements hook_menu().
 */
function search_menu() {
  $items['search'] = array(
    'title' => 'Search',
    'page callback' => 'search_view',
    'access arguments' => array('search content'),
    'type' => MENU_SUGGESTED_ITEM,
    'file' => 'search.pages.inc',
  );
  $items['admin/config/search/settings'] = array(
    'title' => 'Search settings',
    'description' => 'Configure relevance settings for search and other indexing options.',
    'page callback' => 'drupal_get_form',
    'page arguments' => array('search_admin_settings'),
    'access arguments' => array('administer search'),
    'type' => MENU_NORMAL_ITEM,
    'file' => 'search.admin.inc',
  );
  $items['admin/config/search/settings/reindex'] = array(
    'title' => 'Clear index',
    'page callback' => 'drupal_get_form',
    'page arguments' => array('search_reindex_confirm'),
    'access arguments' => array('administer search'),
    'type' => MENU_CALLBACK,
    'file' => 'search.admin.inc',
  );
  $items['admin/reports/search'] = array(
    'title' => 'Top search phrases',
    'description' => 'View most popular search phrases.',
    'page callback' => 'dblog_top',
    'page arguments' => array('search'),
    'access arguments' => array('access site reports'),
    'file path' => drupal_get_path('module', 'dblog'),
    'file' => 'dblog.admin.inc',
  );
  drupal_static_reset('search_get_info');
  $search_hooks = search_get_info();
  foreach(variable_get('search_active_modules', array('node', 'user')) as $module) {
    if (isset($search_hooks[$module])) {
      $items['search/' . $search_hooks[$module]['path'] . '/%menu_tail'] = array(
        'title' => $search_hooks[$module]['title'],
        'page callback' => 'search_view',
        'page arguments' => array($module),
        'access callback' => '_search_menu_access',
        'access arguments' => array($module),
        'type' => MENU_LOCAL_TASK,
        'file' => 'search.pages.inc',
      );
    }
  }
  return $items;
}

/**
 * Get information about all available search hooks.
 */
function search_get_info() {
  $search_hooks = &drupal_static(__FUNCTION__);

  if (!isset($search_hooks)) {
    foreach (module_implements('search_info') as $module) {
      $search_hooks[$module] = call_user_func($module . '_search_info');
      // Use module name as the default.
      $search_hooks[$module] += array('title' => $module, 'path' => $module);
    }
  }
  return $search_hooks;
}

/**
 * Access callback for search tabs.
 */
function _search_menu_access($name) {
  return user_access('search content') && (!function_exists($name . '_search_access') || module_invoke($name, 'search_access'));
}

/**
 * Wipes a part of or the entire search index.
 *
 * @param $sid
 *  (optional) The SID of the item to wipe. If specified, $type must be passed
 *  too.
 * @param $type
 *  (optional) The type of item to wipe.
 */
function search_reindex($sid = NULL, $type = NULL, $reindex = FALSE) {
  if ($type == NULL && $sid == NULL) {
    module_invoke_all('search_reset');
  }
  else {
    db_delete('search_dataset')
      ->condition('sid', $sid)
      ->condition('type', $type)
      ->execute();
    db_delete('search_index')
      ->condition('sid', $sid)
      ->condition('type', $type)
      ->execute();
    // Don't remove links if re-indexing.
    if (!$reindex) {
      db_delete('search_node_links')
        ->condition('sid', $sid)
        ->condition('type', $type)
        ->execute();
    }
  }
}

/**
 * Marks a word as dirty (or retrieves the list of dirty words). This is used
 * during indexing (cron). Words which are dirty have outdated total counts in
 * the search_total table, and need to be recounted.
 */
function search_dirty($word = NULL) {
  $dirty = &drupal_static(__FUNCTION__, array());
  if ($word !== NULL) {
    $dirty[$word] = TRUE;
  }
  else {
    return $dirty;
  }
}

/**
 * Implements hook_cron().
 *
 * Fires hook_update_index() in all modules and cleans up dirty words (see
 * search_dirty).
 */
function search_cron() {
  // We register a shutdown function to ensure that search_total is always up
  // to date.
  register_shutdown_function('search_update_totals');

  foreach(variable_get('search_active_modules', array('node', 'user')) as $module) {
    // Update word index
    module_invoke($module, 'update_index');
  }
}

/**
 * This function is called on shutdown to ensure that search_total is always
 * up to date (even if cron times out or otherwise fails).
 */
function search_update_totals() {
  // Update word IDF (Inverse Document Frequency) counts for new/changed words.
  foreach (search_dirty() as $word => $dummy) {
    // Get total count
    $total = db_query("SELECT SUM(score) FROM {search_index} WHERE word = :word", array(':word' => $word), array('target' => 'slave'))->fetchField();
    // Apply Zipf's law to equalize the probability distribution.
    $total = log10(1 + 1/(max(1, $total)));
    db_merge('search_total')
      ->key(array('word' => $word))
      ->fields(array('count' => $total))
      ->execute();
  }
  // Find words that were deleted from search_index, but are still in
  // search_total. We use a LEFT JOIN between the two tables and keep only the
  // rows which fail to join.
  $result = db_query("SELECT t.word AS realword, i.word FROM {search_total} t LEFT JOIN {search_index} i ON t.word = i.word WHERE i.word IS NULL", array(), array('target' => 'slave'));
  $or = db_or();
  foreach ($result as $word) {
    $or->condition('word', $word->realword);
  }
  if (count($or) > 0) {
    db_delete('search_total')
      ->condition($or)
      ->execute();
  }
}

/**
 * Simplifies a string according to indexing rules.
 */
function search_simplify($text) {
  // Decode entities to UTF-8
  $text = decode_entities($text);

  // Lowercase
  $text = drupal_strtolower($text);

  // Call an external processor for word handling.
  search_invoke_preprocess($text);

  // Simple CJK handling
  if (variable_get('overlap_cjk', TRUE)) {
    $text = preg_replace_callback('/[' . PREG_CLASS_CJK . ']+/u', 'search_expand_cjk', $text);
  }

  // To improve searching for numerical data such as dates, IP addresses
  // or version numbers, we consider a group of numerical characters
  // separated only by punctuation characters to be one piece.
  // This also means that searching for e.g. '20/03/1984' also returns
  // results with '20-03-1984' in them.
  // Readable regexp: ([number]+)[punctuation]+(?=[number])
  $text = preg_replace('/([' . PREG_CLASS_NUMBERS . ']+)[' . PREG_CLASS_PUNCTUATION . ']+(?=[' . PREG_CLASS_NUMBERS . '])/u', '\1', $text);

  // The dot, underscore and dash are simply removed. This allows meaningful
  // search behavior with acronyms and URLs. No need to use the unicode modifer
  // here because 0-127 ASCII characters can't match higher UTF-8 characters as
  // the leftmost bit of those are 1.
  $text = preg_replace('/[._-]+/', '', $text);

  // With the exception of the rules above, we consider all punctuation,
  // marks, spacers, etc, to be a word boundary.
  $text = preg_replace('/[' . PREG_CLASS_SEARCH_EXCLUDE . ']+/u', ' ', $text);

  return $text;
}

/**
 * Basic CJK tokenizer. Simply splits a string into consecutive, overlapping
 * sequences of characters ('minimum_word_size' long).
 */
function search_expand_cjk($matches) {
  $min = variable_get('minimum_word_size', 3);
  $str = $matches[0];
  $l = drupal_strlen($str);
  // Passthrough short words
  if ($l <= $min) {
    return ' ' . $str . ' ';
  }
  $tokens = ' ';
  // FIFO queue of characters
  $chars = array();
  // Begin loop
  for ($i = 0; $i < $l; ++$i) {
    // Grab next character
    $current = drupal_substr($str, 0, 1);
    $str = substr($str, strlen($current));
    $chars[] = $current;
    if ($i >= $min - 1) {
      $tokens .= implode('', $chars) . ' ';
      array_shift($chars);
    }
  }
  return $tokens;
}

/**
 * Splits a string into tokens for indexing.
 */
function search_index_split($text) {
  $last = &drupal_static(__FUNCTION__);
  $lastsplit = &drupal_static(__FUNCTION__ . ':lastsplit');

  if ($last == $text) {
    return $lastsplit;
  }
  // Process words
  $text = search_simplify($text);
  $words = explode(' ', $text);
  array_walk($words, '_search_index_truncate');

  // Save last keyword result
  $last = $text;
  $lastsplit = $words;

  return $words;
}

/**
 * Helper function for array_walk in search_index_split.
 */
function _search_index_truncate(&$text) {
  $text = truncate_utf8($text, 50);
}

/**
 * Invokes hook_search_preprocess() in modules.
 */
function search_invoke_preprocess(&$text) {
  foreach (module_implements('search_preprocess') as $module) {
    $text = module_invoke($module, 'search_preprocess', $text);
  }
}

/**
 * Update the full-text search index for a particular item.
 *
 * @param $sid
 *   A number identifying this particular item (e.g. node id).
 *
 * @param $type
 *   A string defining this type of item (e.g. 'node')
 *
 * @param $text
 *   The content of this item. Must be a piece of HTML text.
 *
 * @ingroup search
 */
function search_index($sid, $type, $text) {
  $minimum_word_size = variable_get('minimum_word_size', 3);

  // Link matching
  global $base_url;
  $node_regexp = '@href=[\'"]?(?:' . preg_quote($base_url, '@') . '/|' . preg_quote(base_path(), '@') . ')(?:\?q=)?/?((?![a-z]+:)[^\'">]+)[\'">]@i';

  // Multipliers for scores of words inside certain HTML tags. The weights are stored
  // in a variable so that modules can overwrite the default weights.
  // Note: 'a' must be included for link ranking to work.
  $tags = variable_get('search_tag_weights', array(
    'h1' => 25,
    'h2' => 18,
    'h3' => 15,
    'h4' => 12,
    'h5' => 9,
    'h6' => 6,
    'u' => 3,
    'b' => 3,
    'i' => 3,
    'strong' => 3,
    'em' => 3,
    'a' => 10));

  // Strip off all ignored tags to speed up processing, but insert space before/after
  // them to keep word boundaries.
  $text = str_replace(array('<', '>'), array(' <', '> '), $text);
  $text = strip_tags($text, '<' . implode('><', array_keys($tags)) . '>');

  // Split HTML tags from plain text.
  $split = preg_split('/\s*<([^>]+?)>\s*/', $text, -1, PREG_SPLIT_DELIM_CAPTURE);
  // Note: PHP ensures the array consists of alternating delimiters and literals
  // and begins and ends with a literal (inserting $null as required).

  $tag = FALSE; // Odd/even counter. Tag or no tag.
  $link = FALSE; // State variable for link analyzer
  $score = 1; // Starting score per word
  $accum = ' '; // Accumulator for cleaned up data
  $tagstack = array(); // Stack with open tags
  $tagwords = 0; // Counter for consecutive words
  $focus = 1; // Focus state

  $results = array(0 => array()); // Accumulator for words for index

  foreach ($split as $value) {
    if ($tag) {
      // Increase or decrease score per word based on tag
      list($tagname) = explode(' ', $value, 2);
      $tagname = drupal_strtolower($tagname);
      // Closing or opening tag?
      if ($tagname[0] == '/') {
        $tagname = substr($tagname, 1);
        // If we encounter unexpected tags, reset score to avoid incorrect boosting.
        if (!count($tagstack) || $tagstack[0] != $tagname) {
          $tagstack = array();
          $score = 1;
        }
        else {
          // Remove from tag stack and decrement score
          $score = max(1, $score - $tags[array_shift($tagstack)]);
        }
        if ($tagname == 'a') {
          $link = FALSE;
        }
      }
      else {
        if (isset($tagstack[0]) && $tagstack[0] == $tagname) {
          // None of the tags we look for make sense when nested identically.
          // If they are, it's probably broken HTML.
          $tagstack = array();
          $score = 1;
        }
        else {
          // Add to open tag stack and increment score
          array_unshift($tagstack, $tagname);
          $score += $tags[$tagname];
        }
        if ($tagname == 'a') {
          // Check if link points to a node on this site
          if (preg_match($node_regexp, $value, $match)) {
            $path = drupal_get_normal_path($match[1]);
            if (preg_match('!(?:node|book)/(?:view/)?([0-9]+)!i', $path, $match)) {
              $linknid = $match[1];
              if ($linknid > 0) {
                $node = db_query('SELECT title, nid, vid FROM {node} WHERE nid = :nid', array(':nid' => $linknid), array('target' => 'slave'))->fetchObject();
                $link = TRUE;
                $linktitle = $node->title;
              }
            }
          }
        }
      }
      // A tag change occurred, reset counter.
      $tagwords = 0;
    }
    else {
      // Note: use of PREG_SPLIT_DELIM_CAPTURE above will introduce empty values
      if ($value != '') {
        if ($link) {
          // Check to see if the node link text is its URL. If so, we use the target node title instead.
          if (preg_match('!^https?://!i', $value)) {
            $value = $linktitle;
          }
        }
        $words = search_index_split($value);
        foreach ($words as $word) {
          // Add word to accumulator
          $accum .= $word . ' ';
          $num = is_numeric($word);
          // Check wordlength
          if ($num || drupal_strlen($word) >= $minimum_word_size) {
            // Normalize numbers
            if ($num) {
              $word = (int)ltrim($word, '-0');
            }

            // Links score mainly for the target.
            if ($link) {
              if (!isset($results[$linknid])) {
                $results[$linknid] = array();
              }
              $results[$linknid][] = $word;
              // Reduce score of the link caption in the source.
              $focus *= 0.2;
            }
            // Fall-through
            if (!isset($results[0][$word])) {
              $results[0][$word] = 0;
            }
            $results[0][$word] += $score * $focus;

            // Focus is a decaying value in terms of the amount of unique words up to this point.
            // From 100 words and more, it decays, to e.g. 0.5 at 500 words and 0.3 at 1000 words.
            $focus = min(1, .01 + 3.5 / (2 + count($results[0]) * .015));
          }
          $tagwords++;
          // Too many words inside a single tag probably mean a tag was accidentally left open.
          if (count($tagstack) && $tagwords >= 15) {
            $tagstack = array();
            $score = 1;
          }
        }
      }
    }
    $tag = !$tag;
  }

  search_reindex($sid, $type, TRUE);

  // Insert cleaned up data into dataset
  db_insert('search_dataset')
    ->fields(array(
      'sid' => $sid,
      'type' => $type,
      'data' => $accum,
      'reindex' => 0,
    ))
    ->execute();

  // Insert results into search index
  foreach ($results[0] as $word => $score) {
    // If a word already exists in the database, its score gets increased
    // appropriately. If not, we create a new record with the appropriate
    // starting score.
    db_merge('search_index')
      ->key(array(
        'word' => $word,
        'sid' => $sid,
        'type' => $type,
      ))
      ->fields(array('score' => $score))
      ->expression('score', 'score + :score', array(':score' => $score))
      ->execute();
    search_dirty($word);
  }
  unset($results[0]);

  // Get all previous links from this item.
  $result = db_query("SELECT nid, caption FROM {search_node_links} WHERE sid = :sid AND type = :type", array(
    ':sid' => $sid,
    ':type' => $type
  ), array('target' => 'slave'));
  $links = array();
  foreach ($result as $link) {
    $links[$link->nid] = $link->caption;
  }

  // Now store links to nodes.
  foreach ($results as $nid => $words) {
    $caption = implode(' ', $words);
    if (isset($links[$nid])) {
      if ($links[$nid] != $caption) {
        // Update the existing link and mark the node for reindexing.
        db_update('search_node_links')
          ->fields(array('caption' => $caption))
          ->condition('sid', $sid)
          ->condition('type', $type)
          ->condition('nid', $nid)
          ->execute();
        search_touch_node($nid);
      }
      // Unset the link to mark it as processed.
      unset($links[$nid]);
    }
    else {
      // Insert the existing link and mark the node for reindexing.
      db_insert('search_node_links')
        ->fields(array(
          'caption' => $caption,
          'sid' => $sid,
          'type' => $type,
          'nid' => $nid,
        ))
        ->execute();
      search_touch_node($nid);
    }
  }
  // Any left-over links in $links no longer exist. Delete them and mark the nodes for reindexing.
  foreach ($links as $nid => $caption) {
    db_delete('search_node_links')
      ->condition('sid', $sid)
      ->condition('type', $type)
      ->condition('nid', $nid)
      ->execute();
    search_touch_node($nid);
  }
}

/**
 * Change a node's changed timestamp to 'now' to force reindexing.
 *
 * @param $nid
 *   The nid of the node that needs reindexing.
 */
function search_touch_node($nid) {
  db_update('search_dataset')
    ->fields(array('reindex' => REQUEST_TIME))
    ->condition('type', 'node')
    ->condition('sid', $nid)
    ->execute();
}

/**
 * Implements hook_node_update_index().
 */
function search_node_update_index($node) {
  // Transplant links to a node into the target node.
  $result = db_query("SELECT caption FROM {search_node_links} WHERE nid = :nid", array(':nid' => $node->nid), array('target' => 'slave'));
  $output = array();
  foreach ($result as $link) {
    $output[] = $link->caption;
  }
  if (count($output)) {
    return '<a>(' . implode(', ', $output) . ')</a>';
  }
}

/**
 * Implements hook_node_update().
 */
function search_node_update($node) {
  // Reindex the node when it is updated. The node is automatically indexed
  // when it is added, simply by being added to the node table.
  search_touch_node($node->nid);
}

/**
 * Implements hook_comment_insert().
 */
function search_comment_insert($comment) {
  // Reindex the node when comments are added.
  search_touch_node($comment->nid);
}

/**
 * Implements hook_comment_update().
 */
function search_comment_update($comment) {
  // Reindex the node when comments are changed.
  search_touch_node($comment->nid);
}

/**
 * Implements hook_comment_delete().
 */
function search_comment_delete($comment) {
  // Reindex the node when comments are deleted.
  search_touch_node($comment->nid);
}

/**
 * Implements hook_comment_publish().
 */
function search_comment_publish($comment) {
  // Reindex the node when comments are published.
  search_touch_node($comment->nid);
}

/**
 * Implements hook_comment_unpublish().
 */
function search_comment_unpublish($comment) {
  // Reindex the node when comments are unpublished.
  search_touch_node($comment->nid);
}

/**
 * Extract a module-specific search option from a search query. e.g. 'type:book'
 */
function search_expression_extract($keys, $option) {
  if (preg_match('/(^| )' . $option . ':([^ ]*)( |$)/i', $keys, $matches)) {
    return $matches[2];
  }
}

/**
 * Return a query with the given module-specific search option inserted in.
 * e.g. 'type:book'.
 */
function search_expression_insert($keys, $option, $value = '') {
  if (search_expression_extract($keys, $option)) {
    $keys = trim(preg_replace('/(^| )' . $option . ':[^ ]*/i', '', $keys));
  }
  if ($value != '') {
    $keys .= ' ' . $option . ':' . $value;
  }
  return $keys;
}

/**
 * Helper function for grabbing search keys.
 */
function search_get_keys() {
  $return = &drupal_static(__FUNCTION__);
  if (!isset($return)) {
    // Extract keys as remainder of path
    // Note: support old GET format of searches for existing links.
    $path = explode('/', $_GET['q'], 3);
    $keys = empty($_REQUEST['keys']) ? '' : $_REQUEST['keys'];
    $return = count($path) == 3 ? $path[2] : $keys;
  }
  return $return;
}

/**
 * @defgroup search Search interface
 * @{
 * The Drupal search interface manages a global search mechanism.
 *
 * Modules may plug into this system to provide searches of different types of
 * data. Most of the system is handled by search.module, so this must be enabled
 * for all of the search features to work.
 *
 * There are three ways to interact with the search system:
 * - Specifically for searching nodes, you can implement hook_node_update_index()
 *   and hook_node_search_result(). However, note that the search system already
 *   indexes all visible output of a node, i.e. everything displayed normally
 *   by hook_view() and hook_node_view(). This is usually sufficient. You should
 *   only use this mechanism if you want additional, non-visible data to be
 *   indexed.
 * - Implement hook_search_info(). This will create a search tab for your module on
 *   the /search page with a simple keyword search form. You will also need to
 *   implement hook_search_execute() to perform the search.
 * - Implement hook_update_index(). This allows your module to use Drupal's
 *   HTML indexing mechanism for searching full text efficiently.
 *
 * If your module needs to provide a more complicated search form, then you need
 * to implement it yourself without hook_search(). In that case, you should
 * define it as a local task (tab) under the /search page (e.g. /search/mymodule)
 * so that users can easily find it.
 */

/**
 * Render a search form.
 *
 * @param $action
 *   Form action. Defaults to "search".
 * @param $keys
 *   The search string entered by the user, containing keywords for the search.
 * @param $type
 *   The type of search to render the node for. Must be the name of module
 *   which implements hook_search(). Defaults to 'node'.
 * @param $prompt
 *   A piece of text to put before the form (e.g. "Enter your keywords")
 * @return
 *   An HTML string containing the search form.
 */
function search_form($form, &$form_state, $action = '', $keys = '', $type = NULL, $prompt = NULL) {
  // Add CSS
  drupal_add_css(drupal_get_path('module', 'search') . '/search.css', array('preprocess' => FALSE));

  if (!$action) {
    $action = url('search/' . $type);
  }
  if (is_null($prompt)) {
    $prompt = t('Enter your keywords');
  }

  $form['#action'] = $action;
  $form['#attributes']['class'][] = 'search-form';
  $form['module'] = array('#type' => 'value', '#value' => $type);
  $form['basic'] = array('#type' => 'item', '#title' => $prompt, '#id' => 'edit-keys');
  $form['basic']['inline'] = array('#prefix' => '<div class="container-inline">', '#suffix' => '</div>');
  $form['basic']['inline']['keys'] = array(
    '#type' => 'textfield',
    '#title' => '',
    '#default_value' => $keys,
    '#size' => $prompt ? 40 : 20,
    '#maxlength' => 255,
  );
  // processed_keys is used to coordinate keyword passing between other forms
  // that hook into the basic search form.
  $form['basic']['inline']['processed_keys'] = array('#type' => 'value', '#value' => array());
  $form['basic']['inline']['submit'] = array('#type' => 'submit', '#value' => t('Search'));

  return $form;
}

/**
 * Form builder; Output a search form for the search block's search box.
 *
 * @ingroup forms
 * @see search_box_form_submit()
 * @see search-theme-form.tpl.php
 * @see search-block-form.tpl.php
 */
function search_box($form, &$form_state, $form_id) {
  $form[$form_id] = array(
    '#type' => 'textfield',
    '#size' => 15,
    '#default_value' => '',
    '#attributes' => array('title' => t('Enter the terms you wish to search for.')),
  );
  $form['submit'] = array('#type' => 'submit', '#value' => t('Search'));
  $form['#submit'][] = 'search_box_form_submit';

  return $form;
}

/**
 * Process a block search form submission.
 */
function search_box_form_submit($form, &$form_state) {
  // The search form relies on control of the redirect destination for its
  // functionality, so we override any static destination set in the request,
  // for example by drupal_access_denied() or drupal_not_found()
  // (see http://drupal.org/node/292565).
  if (isset($_GET['destination'])) {
    unset($_GET['destination']);
  }

  $form_id = $form['form_id']['#value'];
  $form_state['redirect'] = 'search/node/' . trim($form_state['values'][$form_id]);
}

/**
 * Process variables for search-block-form.tpl.php.
 *
 * The $variables array contains the following arguments:
 * - $form
 *
 * @see search-block-form.tpl.php
 */
function template_preprocess_search_block_form(&$variables) {
  $variables['search'] = array();
  $hidden = array();
  // Provide variables named after form keys so themers can print each element independently.
  foreach (element_children($variables['form']) as $key) {
    $type = $variables['form'][$key]['#type'];
    if ($type == 'hidden' || $type == 'token') {
      $hidden[] = drupal_render($variables['form'][$key]);
    }
    else {
      $variables['search'][$key] = drupal_render($variables['form'][$key]);
    }
  }
  // Hidden form elements have no value to themers. No need for separation.
  $variables['search']['hidden'] = implode($hidden);
  // Collect all form elements to make it easier to print the whole form.
  $variables['search_form'] = implode($variables['search']);
}

/**
 * Perform a standard search on the given keys, and return the formatted results.
 */
function search_data($keys = NULL, $type = 'node') {

  if (isset($keys)) {
    if (module_hook($type, 'search_execute')) {
      $results = module_invoke($type, 'search_execute', $keys);
      if (module_hook($type, 'search_page')) {
        return module_invoke($type, 'search_page', $results);
      }
      else {
        return theme('search_results', array('results' => $results, 'type' => $type));
      }
    }
  }
}

/**
 * Returns snippets from a piece of text, with certain keywords highlighted.
 * Used for formatting search results.
 *
 * @param $keys
 *   A string containing a search query.
 *
 * @param $text
 *   The text to extract fragments from.
 *
 * @return
 *   A string containing HTML for the excerpt.
 */
function search_excerpt($keys, $text) {
  // We highlight around non-indexable or CJK characters.
  $boundary = '(?:(?<=[' . PREG_CLASS_SEARCH_EXCLUDE . PREG_CLASS_CJK . '])|(?=[' . PREG_CLASS_SEARCH_EXCLUDE . PREG_CLASS_CJK . ']))';

  // Extract positive keywords and phrases
  preg_match_all('/ ("([^"]+)"|(?!OR)([^" ]+))/', ' ' . $keys, $matches);
  $keys = array_merge($matches[2], $matches[3]);

  // Prepare text
  $text = ' ' . strip_tags(str_replace(array('<', '>'), array(' <', '> '), $text)) . ' ';
  array_walk($keys, '_search_excerpt_replace');
  $workkeys = $keys;

  // Extract a fragment per keyword for at most 4 keywords.
  // First we collect ranges of text around each keyword, starting/ending
  // at spaces.
  // If the sum of all fragments is too short, we look for second occurrences.
  $ranges = array();
  $included = array();
  $length = 0;
  while ($length < 256 && count($workkeys)) {
    foreach ($workkeys as $k => $key) {
      if (strlen($key) == 0) {
        unset($workkeys[$k]);
        unset($keys[$k]);
        continue;
      }
      if ($length >= 256) {
        break;
      }
      // Remember occurrence of key so we can skip over it if more occurrences
      // are desired.
      if (!isset($included[$key])) {
        $included[$key] = 0;
      }
      // Locate a keyword (position $p), then locate a space in front (position
      // $q) and behind it (position $s)
      if (preg_match('/' . $boundary . $key . $boundary . '/iu', $text, $match, PREG_OFFSET_CAPTURE, $included[$key])) {
        $p = $match[0][1];
        if (($q = strpos($text, ' ', max(0, $p - 60))) !== FALSE) {
          $end = substr($text, $p, 80);
          if (($s = strrpos($end, ' ')) !== FALSE) {
            $ranges[$q] = $p + $s;
            $length += $p + $s - $q;
            $included[$key] = $p + 1;
          }
          else {
            unset($workkeys[$k]);
          }
        }
        else {
          unset($workkeys[$k]);
        }
      }
      else {
        unset($workkeys[$k]);
      }
    }
  }

  // If we didn't find anything, return the beginning.
  if (count($ranges) == 0) {
    return truncate_utf8($text, 256) . ' ...';
  }

  // Sort the text ranges by starting position.
  ksort($ranges);

  // Now we collapse overlapping text ranges into one. The sorting makes it O(n).
  $newranges = array();
  foreach ($ranges as $from2 => $to2) {
    if (!isset($from1)) {
      $from1 = $from2;
      $to1 = $to2;
      continue;
    }
    if ($from2 <= $to1) {
      $to1 = max($to1, $to2);
    }
    else {
      $newranges[$from1] = $to1;
      $from1 = $from2;
      $to1 = $to2;
    }
  }
  $newranges[$from1] = $to1;

  // Fetch text
  $out = array();
  foreach ($newranges as $from => $to) {
    $out[] = substr($text, $from, $to - $from);
  }
  $text = (isset($newranges[0]) ? '' : '... ') . implode(' ... ', $out) . ' ...';

  // Highlight keywords. Must be done at once to prevent conflicts ('strong' and '<strong>').
  $text = preg_replace('/' . $boundary . '(' . implode('|', $keys) . ')' . $boundary . '/iu', '<strong>\0</strong>', $text);
  return $text;
}

/**
 * @} End of "defgroup search".
 */

/**
 * Helper function for array_walk in search_except.
 */
function _search_excerpt_replace(&$text) {
  $text = preg_quote($text, '/');
}

function search_forms() {
  $forms['search_block_form']= array(
    'callback' => 'search_box',
    'callback arguments' => array('search_block_form'),
  );
  return $forms;
}