831 lines
34 KiB
Plaintext
831 lines
34 KiB
Plaintext
<?php
|
|
|
|
/**
|
|
* @file
|
|
* Enables site-wide keyword searching.
|
|
*/
|
|
|
|
use Drupal\Component\Utility\Unicode;
|
|
use Symfony\Component\HttpFoundation\Request;
|
|
|
|
/**
|
|
* 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 CJK (Chinese, Japanese, Korean) letter-like characters.
|
|
*
|
|
* This list is derived from the "East Asian Scripts" section of
|
|
* http://www.unicode.org/charts/index.html, as well as a comment on
|
|
* http://unicode.org/reports/tr11/tr11-11.html listing some character
|
|
* ranges that are reserved for additional CJK ideographs.
|
|
*
|
|
* The character ranges do not include numbers, punctuation, or symbols, since
|
|
* these are handled separately in search. Note that radicals and strokes are
|
|
* considered symbols. (See
|
|
* http://www.unicode.org/Public/UNIDATA/extracted/DerivedGeneralCategory.txt)
|
|
*
|
|
* @see search_expand_cjk()
|
|
*/
|
|
define('PREG_CLASS_CJK', '\x{1100}-\x{11FF}\x{3040}-\x{309F}\x{30A1}-\x{318E}' .
|
|
'\x{31A0}-\x{31B7}\x{31F0}-\x{31FF}\x{3400}-\x{4DBF}\x{4E00}-\x{9FCF}' .
|
|
'\x{A000}-\x{A48F}\x{A4D0}-\x{A4FD}\x{A960}-\x{A97F}\x{AC00}-\x{D7FF}' .
|
|
'\x{F900}-\x{FAFF}\x{FF21}-\x{FF3A}\x{FF41}-\x{FF5A}\x{FF66}-\x{FFDC}' .
|
|
'\x{20000}-\x{2FFFD}\x{30000}-\x{3FFFD}');
|
|
|
|
/**
|
|
* Implements hook_help().
|
|
*/
|
|
function search_help($route_name, Request $request) {
|
|
switch ($route_name) {
|
|
case 'help.page.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/documentation/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 and its fields, along with text added to your content by other modules (such as comments from the core Comment module, and taxonomy terms from the core Taxonomy module). 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/pages'))) . '</dd>';
|
|
$output .= '<dt>' . t('Content reindexing') . '</dt>';
|
|
$output .= '<dd>' . t('Content-related actions on your site (creating, editing, or deleting content and comments) automatically cause affected content items to be marked for indexing or reindexing at the next cron run. When content is marked for reindexing, the previous content remains in the index until cron runs, at which time it is replaced by the new content. Unlike content-related actions, actions related to the structure of your site do not cause affected content to be marked for reindexing. Examples of structure-related actions that affect content include deleting or editing taxonomy terms, enabling or disabling modules that add text to content (such as Taxonomy, Comment, and field-providing modules), and modifying the fields or display parameters of your content types. If you take one of these actions and you want to ensure that the search index is updated to reflect your changed site structure, you can mark all content for reindexing by clicking the "Re-index site" button on the <a href="@searchsettings">Search settings page</a>. If you have a lot of content on your site, it may take several cron runs for the content to be reindexed.', array('@searchsettings' => url('admin/config/search/pages'))) . '</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 plugins</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/pages'))) . '</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="@porterstemmer_url">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="@solr_url">Apache Solr</a> or <a href="@sphinx_url">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', '@porterstemmer_url' => 'http://drupal.org/project/porterstemmer', '@solr_url' => 'http://drupal.org/project/apachesolr', '@sphinx_url' => 'http://drupal.org/project/sphinx')) . '</dd>';
|
|
$output .= '</dl>';
|
|
return $output;
|
|
|
|
case 'search.settings':
|
|
return '<p>' . t('The search engine maintains an index of words found in your site\'s content. To build and maintain this index, a correctly configured <a href="@cron">cron maintenance task</a> is required. Indexing behavior can be adjusted using the settings below.', array('@cron' => url('admin/reports/status'))) . '</p>';
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Implements hook_theme().
|
|
*/
|
|
function search_theme() {
|
|
return array(
|
|
'search_result' => array(
|
|
'variables' => array('result' => NULL, 'plugin_id' => NULL),
|
|
'file' => 'search.pages.inc',
|
|
'template' => 'search-result',
|
|
),
|
|
);
|
|
}
|
|
|
|
/**
|
|
* Implements hook_permission().
|
|
*/
|
|
function search_permission() {
|
|
return array(
|
|
'administer search' => array(
|
|
'title' => t('Administer search'),
|
|
),
|
|
'search content' => array(
|
|
'title' => t('Use search'),
|
|
),
|
|
'use advanced search' => array(
|
|
'title' => t('Use advanced search'),
|
|
),
|
|
);
|
|
}
|
|
|
|
/**
|
|
* Implements hook_preprocess_HOOK() for block templates.
|
|
*/
|
|
function search_preprocess_block(&$variables) {
|
|
if ($variables['plugin_id'] == 'search_form_block') {
|
|
$variables['attributes']['role'] = 'search';
|
|
$variables['content_attributes']['class'][] = 'container-inline';
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Clears either a part of, or the entire search index.
|
|
*
|
|
* @param $sid
|
|
* (optional) The ID of the item to remove from the search index. If
|
|
* specified, $type must also be given. Omit both $sid and $type to clear
|
|
* the entire search index.
|
|
* @param $type
|
|
* (optional) The plugin ID or other machine-readable type for the item to
|
|
* remove from the search index.
|
|
* @param $langcode
|
|
* (optional) Language code for the operation. If not provided, all
|
|
* index records for the $sid and $type will be deleted.
|
|
*/
|
|
function search_reindex($sid = NULL, $type = NULL, $langcode = NULL) {
|
|
if ($type == NULL && $sid == NULL) {
|
|
/** @var $search_page_repository \Drupal\search\SearchPageRepositoryInterface */
|
|
$search_page_repository = \Drupal::service('search.search_page_repository');
|
|
foreach ($search_page_repository->getIndexableSearchPages() as $entity) {
|
|
$entity->getPlugin()->resetIndex();
|
|
}
|
|
}
|
|
else {
|
|
$query = db_delete('search_dataset')
|
|
->condition('sid', $sid)
|
|
->condition('type', $type);
|
|
if (!empty($langcode)) {
|
|
$query->condition('langcode', $langcode);
|
|
}
|
|
$query->execute();
|
|
|
|
$query = db_delete('search_index')
|
|
->condition('sid', $sid)
|
|
->condition('type', $type);
|
|
if (!empty($langcode)) {
|
|
$query->condition('langcode', $langcode);
|
|
}
|
|
$query->execute();
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Marks a word as "dirty" (changed), or retrieves the list of dirty words.
|
|
*
|
|
* This is used during indexing (cron). Words that are dirty have outdated
|
|
* total counts in the search_total table, and need to be recounted.
|
|
*/
|
|
function search_dirty($word = NULL) {
|
|
$dirty = &drupal_static(__FUNCTION__, array());
|
|
if ($word !== NULL) {
|
|
$dirty[$word] = TRUE;
|
|
}
|
|
else {
|
|
return $dirty;
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Implements hook_cron().
|
|
*
|
|
* Fires updateIndex() in the plugins for all indexable active search pages,
|
|
* 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.
|
|
drupal_register_shutdown_function('search_update_totals');
|
|
|
|
/** @var $search_page_repository \Drupal\search\SearchPageRepositoryInterface */
|
|
$search_page_repository = \Drupal::service('search.search_page_repository');
|
|
foreach ($search_page_repository->getIndexableSearchPages() as $entity) {
|
|
$entity->getPlugin()->updateIndex();
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Updates the {search_total} database table.
|
|
*
|
|
* 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('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.
|
|
*
|
|
* @param $text
|
|
* Text to simplify.
|
|
*
|
|
* @return
|
|
* Simplified text.
|
|
*
|
|
* @see hook_search_preprocess()
|
|
*/
|
|
function search_simplify($text, $langcode = NULL) {
|
|
// 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, $langcode);
|
|
|
|
// Simple CJK handling
|
|
if (\Drupal::config('search.settings')->get('index.overlap_cjk')) {
|
|
$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);
|
|
|
|
// Multiple dot and dash groups are word boundaries and replaced with space.
|
|
// 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('/[.-]{2,}/', ' ', $text);
|
|
|
|
// The dot, underscore and dash are simply removed. This allows meaningful
|
|
// search behavior with acronyms and URLs. See unicode note directly above.
|
|
$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('/[' . Unicode::PREG_CLASS_WORD_BOUNDARY . ']+/u', ' ', $text);
|
|
|
|
// Truncate everything to 50 characters.
|
|
$words = explode(' ', $text);
|
|
array_walk($words, '_search_index_truncate');
|
|
$text = implode(' ', $words);
|
|
|
|
return $text;
|
|
}
|
|
|
|
/**
|
|
* Splits CJK (Chinese, Japanese, Korean) text into tokens.
|
|
*
|
|
* The Search module matches exact words, where a word is defined to be a
|
|
* sequence of characters delimited by spaces or punctuation. CJK languages are
|
|
* written in long strings of characters, though, not split up into words. So
|
|
* in order to allow search matching, we split up CJK text into tokens
|
|
* consisting of consecutive, overlapping sequences of characters whose length
|
|
* is equal to the 'minimum_word_size' variable. This tokenizing is only done
|
|
* if the 'overlap_cjk' variable is TRUE.
|
|
*
|
|
* @param $matches
|
|
* This function is a callback for preg_replace_callback(), which is called
|
|
* from search_simplify(). So, $matches is an array of regular expression
|
|
* matches, which means that $matches[0] contains the matched text -- a
|
|
* string of CJK characters to tokenize.
|
|
*
|
|
* @return
|
|
* Tokenized text, starting and ending with a space character.
|
|
*/
|
|
function search_expand_cjk($matches) {
|
|
$min = \Drupal::config('search.settings')->get('index.minimum_word_size');
|
|
$str = $matches[0];
|
|
$length = drupal_strlen($str);
|
|
// If the text is shorter than the minimum word size, don't tokenize it.
|
|
if ($length <= $min) {
|
|
return ' ' . $str . ' ';
|
|
}
|
|
$tokens = ' ';
|
|
// Build a FIFO queue of characters.
|
|
$chars = array();
|
|
for ($i = 0; $i < $length; $i++) {
|
|
// Add the next character off the beginning of the string to the queue.
|
|
$current = drupal_substr($str, 0, 1);
|
|
$str = substr($str, strlen($current));
|
|
$chars[] = $current;
|
|
if ($i >= $min - 1) {
|
|
// Make a token of $min characters, and add it to the token string.
|
|
$tokens .= implode('', $chars) . ' ';
|
|
// Shift out the first character in the queue.
|
|
array_shift($chars);
|
|
}
|
|
}
|
|
return $tokens;
|
|
}
|
|
|
|
/**
|
|
* Simplifies and splits a string into tokens for indexing.
|
|
*/
|
|
function search_index_split($text, $langcode = NULL) {
|
|
$last = &drupal_static(__FUNCTION__);
|
|
$lastsplit = &drupal_static(__FUNCTION__ . ':lastsplit');
|
|
|
|
if ($last == $text) {
|
|
return $lastsplit;
|
|
}
|
|
// Process words
|
|
$text = search_simplify($text, $langcode);
|
|
$words = explode(' ', $text);
|
|
|
|
// Save last keyword result
|
|
$last = $text;
|
|
$lastsplit = $words;
|
|
|
|
return $words;
|
|
}
|
|
|
|
/**
|
|
* Helper function for array_walk in search_index_split.
|
|
*/
|
|
function _search_index_truncate(&$text) {
|
|
if (is_numeric($text)) {
|
|
$text = ltrim($text, '0');
|
|
}
|
|
$text = truncate_utf8($text, 50);
|
|
}
|
|
|
|
/**
|
|
* Invokes hook_search_preprocess() in modules.
|
|
*/
|
|
function search_invoke_preprocess(&$text, $langcode = NULL) {
|
|
foreach (\Drupal::moduleHandler()->getImplementations('search_preprocess') as $module) {
|
|
$text = \Drupal::moduleHandler()->invoke($module, 'search_preprocess', array($text, $langcode));
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Updates the full-text search index for a particular item.
|
|
*
|
|
* @param $sid
|
|
* An ID number identifying this particular item (e.g., node ID).
|
|
* @param $type
|
|
* The plugin ID or other machine-readable type of this item,
|
|
* which should be less than 64 bytes.
|
|
* @param $text
|
|
* The content of this item. Must be a piece of HTML or plain text.
|
|
* @param $langcode
|
|
* Language code for text being indexed.
|
|
*
|
|
* @ingroup search
|
|
*/
|
|
function search_index($sid, $type, $text, $langcode) {
|
|
$minimum_word_size = \Drupal::config('search.settings')->get('index.minimum_word_size');
|
|
|
|
// Multipliers for scores of words inside certain HTML tags. The weights are
|
|
// stored in config so that modules can overwrite the default weights.
|
|
// Note: 'a' must be included for link ranking to work.
|
|
$tags = \Drupal::config('search.settings')->get('index.tag_weights');
|
|
|
|
// Strip off all ignored tags to speed up processing, but insert space before
|
|
// and 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.
|
|
$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
|
|
|
|
$scored_words = array(); // Accumulator for words for index
|
|
|
|
foreach ($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)]);
|
|
}
|
|
}
|
|
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];
|
|
}
|
|
}
|
|
// A tag change occurred, reset counter.
|
|
$tagwords = 0;
|
|
}
|
|
else {
|
|
// Note: use of PREG_SPLIT_DELIM_CAPTURE above will introduce empty values
|
|
if ($value != '') {
|
|
$words = search_index_split($value, $langcode);
|
|
foreach ($words as $word) {
|
|
// Add word to accumulator
|
|
$accum .= $word . ' ';
|
|
// Check wordlength
|
|
if (is_numeric($word) || drupal_strlen($word) >= $minimum_word_size) {
|
|
if (!isset($scored_words[$word])) {
|
|
$scored_words[$word] = 0;
|
|
}
|
|
$scored_words[$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.
|
|
$focus = min(1, .01 + 3.5 / (2 + count($scored_words) * .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, $langcode);
|
|
|
|
// Insert cleaned up data into dataset
|
|
db_insert('search_dataset')
|
|
->fields(array(
|
|
'sid' => $sid,
|
|
'langcode' => $langcode,
|
|
'type' => $type,
|
|
'data' => $accum,
|
|
'reindex' => 0,
|
|
))
|
|
->execute();
|
|
|
|
// Insert results into search index
|
|
foreach ($scored_words 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')
|
|
->keys(array(
|
|
'word' => $word,
|
|
'sid' => $sid,
|
|
'langcode' => $langcode,
|
|
'type' => $type,
|
|
))
|
|
->fields(array('score' => $score))
|
|
->expression('score', 'score + :score', array(':score' => $score))
|
|
->execute();
|
|
search_dirty($word);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Changes the timestamp on an indexed item to 'now' to force reindexing.
|
|
*
|
|
* @param $type
|
|
* The plugin ID or other machine-readable type of this item.
|
|
* @param $sid
|
|
* An ID number identifying this particular item (e.g., node ID).
|
|
*/
|
|
function search_mark_for_reindex($type, $sid) {
|
|
db_update('search_dataset')
|
|
->fields(array('reindex' => REQUEST_TIME))
|
|
->condition('type', $type)
|
|
->condition('sid', $sid)
|
|
->execute();
|
|
}
|
|
|
|
/**
|
|
* @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 the Search module, so this must be
|
|
* enabled for all of the search features to work.
|
|
*
|
|
* To be discovered, the plugins must implement
|
|
* \Drupal\search\Plugin\SearchInterface and be annotated as
|
|
* \Drupal\search\Annotation\SearchPlugin plugins. Defining a plugin will allow
|
|
* administrators to set up one or more search pages using this plugin.
|
|
*
|
|
* 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 during node viewing. This is
|
|
* usually sufficient. You should only use this mechanism if you want
|
|
* additional, non-visible data to be indexed.
|
|
* - Define a plugin implementing \Drupal\search\Plugin\SearchInterface and
|
|
* annotated as \Drupal\search\Annotation\SearchPlugin. This will create a
|
|
* search page type that users can use to set up one or more search pages.
|
|
* Each of these corresponds to a tab on the /search page, which can be
|
|
* used to perform searches. You will also need to implement the execute()
|
|
* method from the interface to perform the search. A base class is provided
|
|
* in \Drupal\search\Plugin\SearchPluginBase.
|
|
*
|
|
* If your module needs to provide a more complicated search form, then you
|
|
* need to implement it yourself. In that case, you may wish to define it as a
|
|
* local task (tab) under the /search page (e.g. /search/mymodule) so that users
|
|
* can easily find it.
|
|
*/
|
|
|
|
/**
|
|
* Returns snippets from a piece of text, with search keywords highlighted.
|
|
*
|
|
* Used for formatting search results.
|
|
*
|
|
* @param string $keys
|
|
* A string containing a search query.
|
|
* @param string $text
|
|
* The text to extract fragments from.
|
|
*
|
|
* @return string
|
|
* A string containing HTML for the excerpt.
|
|
*/
|
|
function search_excerpt($keys, $text, $langcode = NULL) {
|
|
// We highlight around non-indexable or CJK characters.
|
|
$boundary = '(?:(?<=[' . Unicode::PREG_CLASS_WORD_BOUNDARY . PREG_CLASS_CJK . '])|(?=[' . Unicode::PREG_CLASS_WORD_BOUNDARY . PREG_CLASS_CJK . ']))';
|
|
|
|
// Extract positive keywords and phrases.
|
|
preg_match_all('/ ("([^"]+)"|(?!OR)([^" ]+))/', ' ' . $keys, $matches);
|
|
$keys = array_merge($matches[2], $matches[3]);
|
|
|
|
// Prepare text by stripping HTML tags and decoding HTML entities.
|
|
$text = strip_tags(str_replace(array('<', '>'), array(' <', '> '), $text));
|
|
$text = decode_entities($text);
|
|
|
|
// Make a list of unique keywords that are actually found in the text,
|
|
// which could be items in $keys or replacements that are equivalent through
|
|
// search_simplify().
|
|
$temp_keys = array();
|
|
foreach ($keys as $key) {
|
|
$key = _search_find_match_with_simplify($key, $text, $boundary, $langcode);
|
|
if (isset($key)) {
|
|
// Quote slashes so they can be used in regular expressions.
|
|
$temp_keys[] = preg_quote($key, '/');
|
|
}
|
|
}
|
|
// Several keywords could have simplified down to the same thing, so pick
|
|
// out the unique ones.
|
|
$keys = array_unique($temp_keys);
|
|
|
|
// Extract fragments of about 60 characters around keywords, bounded by word
|
|
// boundary characters. Try to reach 256 characters, using second occurrences
|
|
// if necessary.
|
|
$ranges = array();
|
|
$length = 0;
|
|
$look_start = array();
|
|
$remaining_keys = $keys;
|
|
|
|
while ($length < 256 && !empty($remaining_keys)) {
|
|
$found_keys = array();
|
|
foreach ($remaining_keys as $key) {
|
|
if ($length >= 256) {
|
|
break;
|
|
}
|
|
|
|
// Remember where we last found $key, in case we are coming through a
|
|
// second time.
|
|
if (!isset($look_start[$key])) {
|
|
$look_start[$key] = 0;
|
|
}
|
|
|
|
// See if we can find $key after where we found it the last time. Since
|
|
// we are requiring a match on a word boundary, make sure $text starts
|
|
// and ends with a space.
|
|
$matches = array();
|
|
if (preg_match('/' . $boundary . $key . $boundary . '/iu', ' ' . $text . ' ', $matches, PREG_OFFSET_CAPTURE, $look_start[$key])) {
|
|
$found_position = $matches[0][1];
|
|
$look_start[$key] = $found_position + 1;
|
|
// Keep track of which keys we found this time, in case we need to
|
|
// pass through again to find more text.
|
|
$found_keys[] = $key;
|
|
|
|
// Locate a space before and after this match, leaving about 60
|
|
// characters of context on each end.
|
|
$before = strpos(' ' . $text, ' ', max(0, $found_position - 61));
|
|
if ($before !== FALSE && $before <= $found_position) {
|
|
$after = strrpos(' ' . $text . ' ', ' ', min($found_position + 61, strlen($text) + 1));
|
|
if ($after !== FALSE && $after > $found_position) {
|
|
// Account for the spaces we added.
|
|
$before = max($before - 1, 0);
|
|
$after = min($after - 1, strlen($text));
|
|
if ($before < $after) {
|
|
// Save this range.
|
|
$ranges[$before] = $after;
|
|
$length += $after - $before;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
// Next time through this loop, only look for keys we found this time,
|
|
// if any.
|
|
$remaining_keys = $found_keys;
|
|
}
|
|
|
|
if (empty($ranges)) {
|
|
// We didn't find any keyword matches, so just return the first part of the
|
|
// text. We also need to re-encode any HTML special characters that we
|
|
// entity-decoded above.
|
|
return check_plain(truncate_utf8($text, 256, TRUE, TRUE));
|
|
}
|
|
|
|
// Sort the text ranges by starting position.
|
|
ksort($ranges);
|
|
|
|
// Collapse overlapping text ranges into one. The sorting makes it O(n).
|
|
$new_ranges = array();
|
|
$max_end = 0;
|
|
foreach ($ranges as $this_from => $this_to) {
|
|
$max_end = max($max_end, $this_to);
|
|
if (!isset($working_from)) {
|
|
// This is the first time through this loop: initialize.
|
|
$working_from = $this_from;
|
|
$working_to = $this_to;
|
|
continue;
|
|
}
|
|
if ($this_from <= $working_to) {
|
|
// The ranges overlap: combine them.
|
|
$working_to = max($working_to, $this_to);
|
|
}
|
|
else {
|
|
// The ranges do not overlap: save the working range and start a new one.
|
|
$new_ranges[$working_from] = $working_to;
|
|
$working_from = $this_from;
|
|
$working_to = $this_to;
|
|
}
|
|
}
|
|
// Save the remaining working range.
|
|
$new_ranges[$working_from] = $working_to;
|
|
|
|
// Fetch text within the combined ranges we found.
|
|
$out = array();
|
|
foreach ($new_ranges as $from => $to) {
|
|
$out[] = substr($text, $from, $to - $from);
|
|
}
|
|
|
|
// Combine the text chunks with "..." separators. The "..." needs to be
|
|
// translated. Let translators have the ... separator text as one chunk.
|
|
$dots = explode('!excerpt', t('... !excerpt ... !excerpt ...'));
|
|
$text = (isset($new_ranges[0]) ? '' : $dots[0]) . implode($dots[1], $out) . (($max_end < strlen($text) - 1) ? $dots[2] : '');
|
|
$text = check_plain($text);
|
|
|
|
// Highlight keywords. Must be done at once to prevent conflicts ('strong'
|
|
// and '<strong>').
|
|
$text = trim(preg_replace('/' . $boundary . '(?:' . implode('|', $keys) . ')' . $boundary . '/iu', '<strong>\0</strong>', ' ' . $text . ' '));
|
|
return $text;
|
|
}
|
|
|
|
/**
|
|
* @} End of "defgroup search".
|
|
*/
|
|
|
|
/**
|
|
* Finds an appropriate keyword in text.
|
|
*
|
|
* @param $key
|
|
* The keyword to find.
|
|
* @param $text
|
|
* The text to search for the keyword.
|
|
* @param $boundary
|
|
* Regular expression for boundary characters between words.
|
|
* @param $langcode
|
|
* Language code.
|
|
*
|
|
* @return
|
|
* A segment of $text that is between word boundary characters that either
|
|
* matches $key directly, or matches $key when both this text segment and
|
|
* $key are processed by search_simplify(). If a matching text segment is
|
|
* not located, NULL is returned.
|
|
*/
|
|
function _search_find_match_with_simplify($key, $text, $boundary, $langcode = NULL) {
|
|
// See if $key appears as-is. When testing, make sure $text starts/ends with
|
|
// a space, because we require $key to be surrounded by word boundary
|
|
// characters.
|
|
$temp = trim($key);
|
|
if ($temp == '') {
|
|
return NULL;
|
|
}
|
|
if (preg_match('/' . $boundary . preg_quote($temp, '/') . $boundary . '/iu', ' ' . $text . ' ')) {
|
|
return $key;
|
|
}
|
|
|
|
// Run both text and key through search_simplify.
|
|
$simplified_key = trim(search_simplify($key, $langcode));
|
|
$simplified_text = trim(search_simplify($text, $langcode));
|
|
if ($simplified_key == '' || $simplified_text == '' || strpos($simplified_text, $simplified_key) === FALSE) {
|
|
// The simplfied keyword and text do not match at all, or are empty.
|
|
return NULL;
|
|
}
|
|
|
|
// Split $text into words, keeping track of where the word boundaries are.
|
|
$words = preg_split('/' . $boundary . '/iu', $text, NULL, PREG_SPLIT_OFFSET_CAPTURE);
|
|
// Add an entry pointing to the end of the string, for the loop below.
|
|
$words[] = array('', strlen($text));
|
|
$num_words = count($words);
|
|
|
|
// Find the smallest segment of complete words in $text that we can simplify
|
|
// to match $simplified_key.
|
|
$start_position = 0;
|
|
$word_end = 0;
|
|
for ($word_index = 0; $word_index < $num_words; $word_index++) {
|
|
// See if we can move the starting position out from our previously-saved
|
|
// best position to here and still have a match.
|
|
$trial_position = $words[$word_index][1];
|
|
if ($trial_position < strlen($text)) {
|
|
$candidate = substr($text, $trial_position);
|
|
$test_text = trim(search_simplify($candidate, $langcode));
|
|
if (strpos($test_text, $simplified_key) !== FALSE) {
|
|
$start_position = $trial_position;
|
|
$word_end = $trial_position + strlen($words[$word_index][0]);
|
|
continue;
|
|
}
|
|
}
|
|
|
|
// See if we can end at our currently-saved word-ending position and still
|
|
// match, in which case this is the minimal matching string.
|
|
if ($word_end > $start_position) {
|
|
$candidate = substr($text, $start_position, $word_end - $start_position);
|
|
$test_text = trim(search_simplify($candidate, $langcode));
|
|
if (strpos($test_text, $simplified_key) !== FALSE) {
|
|
return $candidate;
|
|
}
|
|
}
|
|
|
|
// Save the end position of this word for the next time through this loop.
|
|
$word_end = $trial_position + strlen($words[$word_index][0]);
|
|
}
|
|
|
|
// If we get here, we couldn't find a match.
|
|
return NULL;
|
|
}
|
|
|
|
/**
|
|
* Implements hook_module_preinstall().
|
|
*/
|
|
function search_module_preinstall() {
|
|
// @todo Remove in https://drupal.org/node/2155635.
|
|
\Drupal::service('plugin.manager.search')->clearCachedDefinitions();
|
|
}
|
|
|
|
/**
|
|
* Implements hook_form_FORM_ID_alter() for the search_block_form form.
|
|
*
|
|
* Since the exposed form is a GET form, we don't want it to send the form
|
|
* tokens. However, you cannot make this happen in the form builder function
|
|
* itself, because the tokens are added to the form after the builder function
|
|
* is called. So, we have to do it in a form_alter.
|
|
*
|
|
* @see \Drupal\search\Form\SearchBlockForm
|
|
*/
|
|
function search_form_search_block_form_alter(&$form, &$form_state) {
|
|
$form['form_build_id']['#access'] = FALSE;
|
|
$form['form_token']['#access'] = FALSE;
|
|
$form['form_id']['#access'] = FALSE;
|
|
}
|