Файловый менеджер

Файловый менеджер - Редактировать - /home/wuectly/www/03cbe/indexer.tar

Назад
indexer.php��0000604��00000034654�15245703510�0006725 0��ustar�00��<?php /** * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; use Joomla\String\StringHelper; JLoader::register('FinderIndexerHelper', __DIR__ . '/helper.php'); JLoader::register('FinderIndexerParser', __DIR__ . '/parser.php'); JLoader::register('FinderIndexerStemmer', __DIR__ . '/stemmer.php'); JLoader::register('FinderIndexerTaxonomy', __DIR__ . '/taxonomy.php'); JLoader::register('FinderIndexerToken', __DIR__ . '/token.php'); jimport('joomla.filesystem.file'); /* * Main indexer class for the Finder indexer package. * * The indexer class provides the core functionality of the Finder * search engine. It is responsible for adding and updating the * content links table; extracting and scoring tokens; and maintaining * all referential information for the content. * * Note: All exceptions thrown from within this class should be caught * by the controller. * * @since 2.5 / abstract class FinderIndexer { /* * The title context identifier. * * @var integer * @since 2.5 / const TITLE_CONTEXT = 1; /* * The text context identifier. * * @var integer * @since 2.5 / const TEXT_CONTEXT = 2; /* * The meta context identifier. * * @var integer * @since 2.5 / const META_CONTEXT = 3; /* * The path context identifier. * * @var integer * @since 2.5 / const PATH_CONTEXT = 4; /* * The misc context identifier. * * @var integer * @since 2.5 / const MISC_CONTEXT = 5; /* * The indexer state object. * * @var JObject * @since 2.5 / public static $state; /* * The indexer profiler object. * * @var JProfiler * @since 2.5 / public static $profiler; /* * Database driver cache. * * @var JDatabaseDriver * @since 3.8.0 / protected $db; /* * Reusable Query Template. To be used with clone. * * @var JDatabaseQuery * @since 3.8.0 / protected $addTokensToDbQueryTemplate; /* * FinderIndexer constructor. * * @since 3.8.0 / public function __construct() { $this->db = JFactory::getDbo(); $db = $this->db; /* * Set up query template for addTokensToDb, we will be cloning this template when needed. * This is about twice as fast as calling the clear function or setting up a new object. / $this->addTokensToDbQueryTemplate = $db->getQuery(true)->insert($db->quoteName('#__finder_tokens')) ->columns( array( $db->quoteName('term'), $db->quoteName('stem'), $db->quoteName('common'), $db->quoteName('phrase'), $db->quoteName('weight'), $db->quoteName('context'), $db->quoteName('language') ) ); } /* * Returns a reference to the FinderIndexer object. * * @return FinderIndexer instance based on the database driver * * @since 3.0 * @throws RuntimeException if driver class for indexer not present. / public static function getInstance() { // Setup the adapter for the indexer. $serverType = JFactory::getDbo()->getServerType(); // For `mssql` server types, convert the type to `sqlsrv` if ($serverType === 'mssql') { $serverType = 'sqlsrv'; } $path = __DIR__ . '/driver/' . $serverType . '.php'; $class = 'FinderIndexerDriver' . ucfirst($serverType); // Check if a parser exists for the format. if (file_exists($path)) { // Instantiate the parser. JLoader::register($class, $path); return new $class; } // Throw invalid format exception. throw new RuntimeException(JText::sprintf('COM_FINDER_INDEXER_INVALID_DRIVER', $serverType)); } /* * Method to get the indexer state. * * @return object The indexer state object. * * @since 2.5 / public static function getState() { // First, try to load from the internal state. if ((bool) static::$state) { return static::$state; } // If we couldn't load from the internal state, try the session. $session = JFactory::getSession(); $data = $session->get('_finder.state', null); // If the state is empty, load the values for the first time. if (empty($data)) { $data = new JObject; // Load the default configuration options. $data->options = JComponentHelper::getParams('com_finder'); // Setup the weight lookup information. $data->weights = array( self::TITLE_CONTEXT => round($data->options->get('title_multiplier', 1.7), 2), self::TEXT_CONTEXT => round($data->options->get('text_multiplier', 0.7), 2), self::META_CONTEXT => round($data->options->get('meta_multiplier', 1.2), 2), self::PATH_CONTEXT => round($data->options->get('path_multiplier', 2.0), 2), self::MISC_CONTEXT => round($data->options->get('misc_multiplier', 0.3), 2) ); // Set the current time as the start time. $data->startTime = JFactory::getDate()->toSql(); // Set the remaining default values. $data->batchSize = (int) $data->options->get('batch_size', 50); $data->batchOffset = 0; $data->totalItems = 0; $data->pluginState = array(); } // Setup the profiler if debugging is enabled. if (JFactory::getApplication()->get('debug')) { static::$profiler = JProfiler::getInstance('FinderIndexer'); } // Setup the stemmer. if ($data->options->get('stem', 1) && $data->options->get('stemmer', 'porter_en')) { FinderIndexerHelper::$stemmer = FinderIndexerStemmer::getInstance($data->options->get('stemmer', 'porter_en')); } // Set the state. static::$state = $data; return static::$state; } /* * Method to set the indexer state. * * @param object $data A new indexer state object. * * @return boolean True on success, false on failure. * * @since 2.5 / public static function setState($data) { // Check the state object. if (empty($data) \|\| !$data instanceof JObject) { return false; } // Set the new internal state. static::$state = $data; // Set the new session state. JFactory::getSession()->set('_finder.state', $data); return true; } /* * Method to reset the indexer state. * * @return void * * @since 2.5 / public static function resetState() { // Reset the internal state to null. self::$state = null; // Reset the session state to null. JFactory::getSession()->set('_finder.state', null); } /* * Method to index a content item. * * @param FinderIndexerResult $item The content item to index. * @param string $format The format of the content. [optional] * * @return integer The ID of the record in the links table. * * @since 2.5 * @throws Exception on database error. / abstract public function index($item, $format = 'html'); /* * Method to remove a link from the index. * * @param integer $linkId The id of the link. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / public function remove($linkId) { $db = $this->db; $query = $db->getQuery(true); // Update the link counts and remove the mapping records. for ($i = 0; $i <= 15; $i++) { // Update the link counts for the terms. $query->clear() ->update($db->quoteName('#__finder_terms', 't')) ->join('INNER', $db->quoteName('#__finder_links_terms' . dechex($i), 'm') . ' ON ' . $db->quoteName('m.term_id') . ' = ' . $db->quoteName('t.term_id') ) ->set($db->quoteName('links') . ' = ' . $db->quoteName('links') . ' - 1') ->where($db->quoteName('m.link_id') . ' = ' . (int) $linkId); $db->setQuery($query)->execute(); // Remove all records from the mapping tables. $query->clear() ->delete($db->quoteName('#__finder_links_terms' . dechex($i))) ->where($db->quoteName('link_id') . ' = ' . (int) $linkId); $db->setQuery($query)->execute(); } // Delete all orphaned terms. $query->clear() ->delete($db->quoteName('#__finder_terms')) ->where($db->quoteName('links') . ' <= 0'); $db->setQuery($query)->execute(); // Delete the link from the index. $query->clear() ->delete($db->quoteName('#__finder_links')) ->where($db->quoteName('link_id') . ' = ' . (int) $linkId); $db->setQuery($query)->execute(); // Remove the taxonomy maps. FinderIndexerTaxonomy::removeMaps($linkId); // Remove the orphaned taxonomy nodes. FinderIndexerTaxonomy::removeOrphanNodes(); return true; } /* * Method to optimize the index. We use this method to remove unused terms * and any other optimizations that might be necessary. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / abstract public function optimize(); /* * Method to get a content item's signature. * * @param object $item The content item to index. * * @return string The content item's signature. * * @since 2.5 / protected static function getSignature($item) { // Get the indexer state. $state = static::getState(); // Get the relevant configuration variables. $config = array( $state->weights, $state->options->get('stem', 1), $state->options->get('stemmer', 'porter_en') ); return md5(serialize(array($item, $config))); } /* * Method to parse input, tokenize it, and then add it to the database. * * @param mixed $input String or resource to use as input. A resource input will automatically be chunked to conserve * memory. Strings will be chunked if longer than 2K in size. * @param integer $context The context of the input. See context constants. * @param string $lang The language of the input. * @param string $format The format of the input. * * @return integer The number of tokens extracted from the input. * * @since 2.5 / protected function tokenizeToDb($input, $context, $lang, $format) { $count = 0; $buffer = null; if (empty($input)) { return $count; } // If the input is a resource, batch the process out. if (is_resource($input)) { // Batch the process out to avoid memory limits. while (!feof($input)) { // Read into the buffer. $buffer .= fread($input, 2048); / * If we haven't reached the end of the file, seek to the last * space character and drop whatever is after that to make sure * we didn't truncate a term while reading the input. / if (!feof($input)) { // Find the last space character. $ls = strrpos($buffer, ' '); // Adjust string based on the last space character. if ($ls) { // Truncate the string to the last space character. $string = substr($buffer, 0, $ls); // Adjust the buffer based on the last space for the next iteration and trim. $buffer = StringHelper::trim(substr($buffer, $ls)); } // No space character was found. else { $string = $buffer; } } // We've reached the end of the file, so parse whatever remains. else { $string = $buffer; } // Parse, tokenise and add tokens to the database. $count = $this->tokenizeToDbShort($string, $context, $lang, $format, $count); unset($string, $tokens); } return $count; } // Parse, tokenise and add tokens to the database. $count = $this->tokenizeToDbShort($input, $context, $lang, $format, $count); return $count; } /* * Method to parse input, tokenise it, then add the tokens to the database. * * @param string $input String to parse, tokenise and add to database. * @param integer $context The context of the input. See context constants. * @param string $lang The language of the input. * @param string $format The format of the input. * @param integer $count The number of tokens processed so far. * * @return integer Cumulative number of tokens extracted from the input so far. * * @since 3.7.0 / private function tokenizeToDbShort($input, $context, $lang, $format, $count) { // Parse the input. $input = FinderIndexerHelper::parse($input, $format); // Check the input. if (empty($input)) { return $count; } // Tokenize the input. $tokens = FinderIndexerHelper::tokenize($input, $lang); // Add the tokens to the database. $count += $this->addTokensToDb($tokens, $context); // Check if we're approaching the memory limit of the token table. if ($count > static::$state->options->get('memory_table_limit', 30000)) { $this->toggleTables(false); } return $count; } /* * Method to add a set of tokens to the database. * * @param mixed $tokens An array or single FinderIndexerToken object. * @param mixed $context The context of the tokens. See context constants. [optional] * * @return integer The number of tokens inserted into the database. * * @since 2.5 * @throws Exception on database error. / protected function addTokensToDb($tokens, $context = '') { // Get the database object. $db = $this->db; $query = clone $this->addTokensToDbQueryTemplate; // Check if a single FinderIndexerToken object was given and make it to be an array of FinderIndexerToken objects $tokens = is_array($tokens) ? $tokens : array($tokens); // Count the number of token values. $values = 0; // Break into chunks of no more than 1000 items $chunks = array_chunk($tokens, 128); foreach ($chunks as $tokens) { $query->clear('values'); // Iterate through the tokens to create SQL value sets. foreach ($tokens as $token) { $query->values( $db->quote($token->term) . ', ' . $db->quote($token->stem) . ', ' . (int) $token->common . ', ' . (int) $token->phrase . ', ' . $db->escape((float) $token->weight) . ', ' . (int) $context . ', ' . $db->quote($token->language) ); ++$values; } $db->setQuery($query)->execute(); // Check if we're approaching the memory limit of the token table. if ($values > static::$state->options->get('memory_table_limit', 10000)) { $this->toggleTables(false); } } return $values; } /* * Method to switch the token tables from Memory tables to Disk tables * when they are close to running out of memory. * Since this is not supported/implemented in all DB-drivers, the default is a stub method, which simply returns true. * * @param boolean $memory Flag to control how they should be toggled. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / protected function toggleTables($memory) { return true; } } ��taxonomy.php��0000604��00000023457�15245703510�0007144 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; /* * Stemmer base class for the Finder indexer package. * * @since 2.5 / class FinderIndexerTaxonomy { /* * An internal cache of taxonomy branch data. * * @var array * @since 2.5 / public static $branches = array(); /* * An internal cache of taxonomy node data. * * @var array * @since 2.5 / public static $nodes = array(); /* * Method to add a branch to the taxonomy tree. * * @param string $title The title of the branch. * @param integer $state The published state of the branch. [optional] * @param integer $access The access state of the branch. [optional] * * @return integer The id of the branch. * * @since 2.5 * @throws Exception on database error. / public static function addBranch($title, $state = 1, $access = 1) { // Check to see if the branch is in the cache. if (isset(static::$branches[$title])) { return static::$branches[$title]->id; } // Check to see if the branch is in the table. $db = JFactory::getDbo(); $query = $db->getQuery(true) ->select('') ->from($db->quoteName('#__finder_taxonomy')) ->where($db->quoteName('parent_id') . ' = 1') ->where($db->quoteName('title') . ' = ' . $db->quote($title)); $db->setQuery($query); // Get the result. $result = $db->loadObject(); // Check if the database matches the input data. if ((bool) $result && $result->state == $state && $result->access == $access) { // The data matches, add the item to the cache. static::$branches[$title] = $result; return static::$branches[$title]->id; } /* * The database did not match the input. This could be because the * state has changed or because the branch does not exist. Let's figure * out which case is true and deal with it. / $branch = new JObject; if (empty($result)) { // Prepare the branch object. $branch->parent_id = 1; $branch->title = $title; $branch->state = (int) $state; $branch->access = (int) $access; } else { // Prepare the branch object. $branch->id = (int) $result->id; $branch->parent_id = (int) $result->parent_id; $branch->title = $result->title; $branch->state = (int) $result->title; $branch->access = (int) $result->access; $branch->ordering = (int) $result->ordering; } // Store the branch. static::storeNode($branch); // Add the branch to the cache. static::$branches[$title] = $branch; return static::$branches[$title]->id; } /* * Method to add a node to the taxonomy tree. * * @param string $branch The title of the branch to store the node in. * @param string $title The title of the node. * @param integer $state The published state of the node. [optional] * @param integer $access The access state of the node. [optional] * * @return integer The id of the node. * * @since 2.5 * @throws Exception on database error. / public static function addNode($branch, $title, $state = 1, $access = 1) { // Check to see if the node is in the cache. if (isset(static::$nodes[$branch][$title])) { return static::$nodes[$branch][$title]->id; } // Get the branch id, insert it if it does not exist. $branchId = static::addBranch($branch); // Check to see if the node is in the table. $db = JFactory::getDbo(); $query = $db->getQuery(true) ->select('') ->from($db->quoteName('#__finder_taxonomy')) ->where($db->quoteName('parent_id') . ' = ' . $db->quote($branchId)) ->where($db->quoteName('title') . ' = ' . $db->quote($title)); $db->setQuery($query); // Get the result. $result = $db->loadObject(); // Check if the database matches the input data. if ((bool) $result && $result->state == $state && $result->access == $access) { // The data matches, add the item to the cache. static::$nodes[$branch][$title] = $result; return static::$nodes[$branch][$title]->id; } /* * The database did not match the input. This could be because the * state has changed or because the node does not exist. Let's figure * out which case is true and deal with it. / $node = new JObject; if (empty($result)) { // Prepare the node object. $node->parent_id = (int) $branchId; $node->title = $title; $node->state = (int) $state; $node->access = (int) $access; } else { // Prepare the node object. $node->id = (int) $result->id; $node->parent_id = (int) $result->parent_id; $node->title = $result->title; $node->state = (int) $result->title; $node->access = (int) $result->access; $node->ordering = (int) $result->ordering; } // Store the node. static::storeNode($node); // Add the node to the cache. static::$nodes[$branch][$title] = $node; return static::$nodes[$branch][$title]->id; } /* * Method to add a map entry between a link and a taxonomy node. * * @param integer $linkId The link to map to. * @param integer $nodeId The node to map to. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / public static function addMap($linkId, $nodeId) { // Insert the map. $db = JFactory::getDbo(); $query = $db->getQuery(true) ->select($db->quoteName('link_id')) ->from($db->quoteName('#__finder_taxonomy_map')) ->where($db->quoteName('link_id') . ' = ' . (int) $linkId) ->where($db->quoteName('node_id') . ' = ' . (int) $nodeId); $db->setQuery($query); $db->execute(); $id = (int) $db->loadResult(); $map = new JObject; $map->link_id = (int) $linkId; $map->node_id = (int) $nodeId; if ($id) { $db->updateObject('#__finder_taxonomy_map', $map, array('link_id', 'node_id')); } else { $db->insertObject('#__finder_taxonomy_map', $map); } return true; } /* * Method to get the title of all taxonomy branches. * * @return array An array of branch titles. * * @since 2.5 * @throws Exception on database error. / public static function getBranchTitles() { $db = JFactory::getDbo(); // Set user variables $groups = implode(',', JFactory::getUser()->getAuthorisedViewLevels()); // Create a query to get the taxonomy branch titles. $query = $db->getQuery(true) ->select($db->quoteName('title')) ->from($db->quoteName('#__finder_taxonomy')) ->where($db->quoteName('parent_id') . ' = 1') ->where($db->quoteName('state') . ' = 1') ->where($db->quoteName('access') . ' IN (' . $groups . ')'); // Get the branch titles. $db->setQuery($query); return $db->loadColumn(); } /* * Method to find a taxonomy node in a branch. * * @param string $branch The branch to search. * @param string $title The title of the node. * * @return mixed Integer id on success, null on no match. * * @since 2.5 * @throws Exception on database error. / public static function getNodeByTitle($branch, $title) { $db = JFactory::getDbo(); // Set user variables $groups = implode(',', JFactory::getUser()->getAuthorisedViewLevels()); // Create a query to get the node. $query = $db->getQuery(true) ->select('t1.') ->from($db->quoteName('#__finder_taxonomy') . ' AS t1') ->join('INNER', $db->quoteName('#__finder_taxonomy') . ' AS t2 ON t2.id = t1.parent_id') ->where('t1.access IN (' . $groups . ')') ->where('t1.state = 1') ->where('t1.title LIKE ' . $db->quote($db->escape($title) . '%')) ->where('t2.access IN (' . $groups . ')') ->where('t2.state = 1') ->where('t2.title = ' . $db->quote($branch)); // Get the node. $db->setQuery($query, 0, 1); return $db->loadObject(); } /** * Method to remove map entries for a link. * * @param integer $linkId The link to remove. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / public static function removeMaps($linkId) { // Delete the maps. $db = JFactory::getDbo(); $query = $db->getQuery(true) ->delete($db->quoteName('#__finder_taxonomy_map')) ->where($db->quoteName('link_id') . ' = ' . (int) $linkId); $db->setQuery($query); $db->execute(); return true; } /* * Method to remove orphaned taxonomy nodes and branches. * * @return integer The number of deleted rows. * * @since 2.5 * @throws Exception on database error. / public static function removeOrphanNodes() { // Delete all orphaned nodes. $db = JFactory::getDbo(); $query = $db->getQuery(true) ->select($db->quoteName('t.id')) ->from($db->quoteName('#__finder_taxonomy', 't')) ->join('LEFT', $db->quoteName('#__finder_taxonomy_map', 'm') . ' ON ' . $db->quoteName('m.node_id') . '=' . $db->quoteName('t.id')) ->where($db->quoteName('t.parent_id') . ' > 1 ') ->where($db->quoteName('m.link_id') . ' IS NULL'); $db->setQuery($query); $ids = $db->loadColumn(); if (empty($ids)) { return 0; } $query->clear() ->delete($db->quoteName('#__finder_taxonomy')) ->where($db->quoteName('id') . ' IN (' . implode(',', $ids) . ')'); $db->setQuery($query); $db->execute(); return $db->getAffectedRows(); } /* * Method to store a node to the database. This method will accept either a branch or a node. * * @param object $item The item to store. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / protected static function storeNode($item) { $db = JFactory::getDbo(); // Check if we are updating or inserting the item. if (empty($item->id)) { // Insert the item. $db->insertObject('#__finder_taxonomy', $item, 'id'); } else { // Update the item. $db->updateObject('#__finder_taxonomy', $item, 'id'); } return true; } } ��query.php��0000604��00000105675�15245703510�0006436 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; use Joomla\Registry\Registry; use Joomla\String\StringHelper; use Joomla\Utilities\ArrayHelper; JLoader::register('FinderIndexerHelper', __DIR__ . '/helper.php'); JLoader::register('FinderIndexerTaxonomy', __DIR__ . '/taxonomy.php'); JLoader::register('FinderHelperRoute', JPATH_SITE . '/components/com_finder/helpers/route.php'); JLoader::register('FinderHelperLanguage', JPATH_ADMINISTRATOR . '/components/com_finder/helpers/language.php'); /* * Query class for the Finder indexer package. * * @since 2.5 / class FinderIndexerQuery { /* * Flag to show whether the query can return results. * * @var boolean * @since 2.5 / public $search; /* * The query input string. * * @var string * @since 2.5 / public $input; /* * The language of the query. * * @var string * @since 2.5 / public $language; /* * The query string matching mode. * * @var string * @since 2.5 / public $mode; /* * The included tokens. * * @var array * @since 2.5 / public $included = array(); /* * The excluded tokens. * * @var array * @since 2.5 / public $excluded = array(); /* * The tokens to ignore because no matches exist. * * @var array * @since 2.5 / public $ignored = array(); /* * The operators used in the query input string. * * @var array * @since 2.5 / public $operators = array(); /* * The terms to highlight as matches. * * @var array * @since 2.5 / public $highlight = array(); /* * The number of matching terms for the query input. * * @var integer * @since 2.5 / public $terms; /* * The static filter id. * * @var string * @since 2.5 / public $filter; /* * The taxonomy filters. This is a multi-dimensional array of taxonomy * branches as the first level and then the taxonomy nodes as the values. * * For example: * $filters = array( * 'Type' = array(10, 32, 29, 11, ...); * 'Label' = array(20, 314, 349, 91, 82, ...); * ... * ); * * @var array * @since 2.5 / public $filters = array(); /* * The start date filter. * * @var string * @since 2.5 / public $date1; /* * The end date filter. * * @var string * @since 2.5 / public $date2; /* * The start date filter modifier. * * @var string * @since 2.5 / public $when1; /* * The end date filter modifier. * * @var string * @since 2.5 / public $when2; /* * Method to instantiate the query object. * * @param array $options An array of query options. * * @since 2.5 * @throws Exception on database error. / public function __construct($options) { // Get the input string. $this->input = isset($options['input']) ? $options['input'] : ''; // Get the empty query setting. $this->empty = isset($options['empty']) ? (bool) $options['empty'] : false; // Get the input language. $this->language = !empty($options['language']) ? $options['language'] : FinderIndexerHelper::getDefaultLanguage(); $this->language = FinderIndexerHelper::getPrimaryLanguage($this->language); // Get the matching mode. $this->mode = 'AND'; // Initialize the temporary date storage. $this->dates = new Registry; // Populate the temporary date storage. if (!empty($options['date1'])) { $this->dates->set('date1', $options['date1']); } if (!empty($options['date2'])) { $this->dates->set('date2', $options['date2']); } if (!empty($options['when1'])) { $this->dates->set('when1', $options['when1']); } if (!empty($options['when2'])) { $this->dates->set('when2', $options['when2']); } // Process the static taxonomy filters. if (!empty($options['filter'])) { $this->processStaticTaxonomy($options['filter']); } // Process the dynamic taxonomy filters. if (!empty($options['filters'])) { $this->processDynamicTaxonomy($options['filters']); } // Get the date filters. $d1 = $this->dates->get('date1'); $d2 = $this->dates->get('date2'); $w1 = $this->dates->get('when1'); $w2 = $this->dates->get('when2'); // Process the date filters. if (!empty($d1) \|\| !empty($d2)) { $this->processDates($d1, $d2, $w1, $w2); } // Process the input string. $this->processString($this->input, $this->language, $this->mode); // Get the number of matching terms. foreach ($this->included as $token) { $this->terms += count($token->matches); } // Remove the temporary date storage. unset($this->dates); // Lastly, determine whether this query can return a result set. // Check if we have a query string. if (!empty($this->input)) { $this->search = true; } // Check if we can search without a query string. elseif ($this->empty && (!empty($this->filter) \|\| !empty($this->filters) \|\| !empty($this->date1) \|\| !empty($this->date2))) { $this->search = true; } // We do not have a valid search query. else { $this->search = false; } } /* * Method to convert the query object into a URI string. * * @param string $base The base URI. [optional] * * @return string The complete query URI. * * @since 2.5 / public function toUri($base = '') { // Set the base if not specified. if ($base === '') { $base = 'index.php?option=com_finder&view=search'; } // Get the base URI. $uri = JUri::getInstance($base); // Add the static taxonomy filter if present. if ((bool) $this->filter) { $uri->setVar('f', $this->filter); } // Get the filters in the request. $t = JFactory::getApplication()->input->request->get('t', array(), 'array'); // Add the dynamic taxonomy filters if present. if ((bool) $this->filters) { foreach ($this->filters as $nodes) { foreach ($nodes as $node) { if (!in_array($node, $t)) { continue; } $uri->setVar('t[]', $node); } } } // Add the input string if present. if (!empty($this->input)) { $uri->setVar('q', $this->input); } // Add the start date if present. if (!empty($this->date1)) { $uri->setVar('d1', $this->date1); } // Add the end date if present. if (!empty($this->date2)) { $uri->setVar('d2', $this->date2); } // Add the start date modifier if present. if (!empty($this->when1)) { $uri->setVar('w1', $this->when1); } // Add the end date modifier if present. if (!empty($this->when2)) { $uri->setVar('w2', $this->when2); } // Add a menu item id if one is not present. if (!$uri->getVar('Itemid')) { // Get the menu item id. $query = array( 'view' => $uri->getVar('view'), 'f' => $uri->getVar('f'), 'q' => $uri->getVar('q'), ); $item = FinderHelperRoute::getItemid($query); // Add the menu item id if present. if ($item !== null) { $uri->setVar('Itemid', $item); } } return $uri->toString(array('path', 'query')); } /* * Method to get a list of excluded search term ids. * * @return array An array of excluded term ids. * * @since 2.5 / public function getExcludedTermIds() { $results = array(); // Iterate through the excluded tokens and compile the matching terms. for ($i = 0, $c = count($this->excluded); $i < $c; $i++) { $results = array_merge($results, $this->excluded[$i]->matches); } // Sanitize the terms. $results = array_unique($results); return ArrayHelper::toInteger($results); } /* * Method to get a list of included search term ids. * * @return array An array of included term ids. * * @since 2.5 / public function getIncludedTermIds() { $results = array(); // Iterate through the included tokens and compile the matching terms. for ($i = 0, $c = count($this->included); $i < $c; $i++) { // Check if we have any terms. if (empty($this->included[$i]->matches)) { continue; } // Get the term. $term = $this->included[$i]->term; // Prepare the container for the term if necessary. if (!array_key_exists($term, $results)) { $results[$term] = array(); } // Add the matches to the stack. $results[$term] = array_merge($results[$term], $this->included[$i]->matches); } // Sanitize the terms. foreach ($results as $key => $value) { $results[$key] = array_unique($results[$key]); $results[$key] = ArrayHelper::toInteger($results[$key]); } return $results; } /* * Method to get a list of required search term ids. * * @return array An array of required term ids. * * @since 2.5 / public function getRequiredTermIds() { $results = array(); // Iterate through the included tokens and compile the matching terms. for ($i = 0, $c = count($this->included); $i < $c; $i++) { // Check if the token is required. if ($this->included[$i]->required) { // Get the term. $term = $this->included[$i]->term; // Prepare the container for the term if necessary. if (!array_key_exists($term, $results)) { $results[$term] = array(); } // Add the matches to the stack. $results[$term] = array_merge($results[$term], $this->included[$i]->matches); } } // Sanitize the terms. foreach ($results as $key => $value) { $results[$key] = array_unique($results[$key]); $results[$key] = ArrayHelper::toInteger($results[$key]); } return $results; } /* * Method to process the static taxonomy input. The static taxonomy input * comes in the form of a pre-defined search filter that is assigned to the * search form. * * @param integer $filterId The id of static filter. * * @return boolean True on success, false on failure. * * @since 2.5 * @throws Exception on database error. / protected function processStaticTaxonomy($filterId) { // Get the database object. $db = JFactory::getDbo(); // Initialize user variables $groups = implode(',', JFactory::getUser()->getAuthorisedViewLevels()); // Load the predefined filter. $query = $db->getQuery(true) ->select('f.data, f.params') ->from($db->quoteName('#__finder_filters') . ' AS f') ->where('f.filter_id = ' . (int) $filterId); $db->setQuery($query); $return = $db->loadObject(); // Check the returned filter. if (empty($return)) { return false; } // Set the filter. $this->filter = (int) $filterId; // Get a parameter object for the filter date options. $registry = new Registry($return->params); $params = $registry; // Set the dates if not already set. $this->dates->def('d1', $params->get('d1')); $this->dates->def('d2', $params->get('d2')); $this->dates->def('w1', $params->get('w1')); $this->dates->def('w2', $params->get('w2')); // Remove duplicates and sanitize. $filters = explode(',', $return->data); $filters = array_unique($filters); $filters = ArrayHelper::toInteger($filters); // Remove any values of zero. if (in_array(0, $filters, true) !== false) { unset($filters[array_search(0, $filters, true)]); } // Check if we have any real input. if (empty($filters)) { return true; } / * Create the query to get filters from the database. We do this for * two reasons: one, it allows us to ensure that the filters being used * are real; two, we need to sort the filters by taxonomy branch. / $query->clear() ->select('t1.id, t1.title, t2.title AS branch') ->from($db->quoteName('#__finder_taxonomy') . ' AS t1') ->join('INNER', $db->quoteName('#__finder_taxonomy') . ' AS t2 ON t2.id = t1.parent_id') ->where('t1.state = 1') ->where('t1.access IN (' . $groups . ')') ->where('t1.id IN (' . implode(',', $filters) . ')') ->where('t2.state = 1') ->where('t2.access IN (' . $groups . ')'); // Load the filters. $db->setQuery($query); $results = $db->loadObjectList(); // Sort the filter ids by branch. foreach ($results as $result) { $this->filters[$result->branch][$result->title] = (int) $result->id; } return true; } /* * Method to process the dynamic taxonomy input. The dynamic taxonomy input * comes in the form of select fields that the user chooses from. The * dynamic taxonomy input is processed AFTER the static taxonomy input * because the dynamic options can be used to further narrow a static * taxonomy filter. * * @param array $filters An array of taxonomy node ids. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / protected function processDynamicTaxonomy($filters) { // Initialize user variables $groups = implode(',', JFactory::getUser()->getAuthorisedViewLevels()); // Remove duplicates and sanitize. $filters = array_unique($filters); $filters = ArrayHelper::toInteger($filters); // Remove any values of zero. if (in_array(0, $filters, true) !== false) { unset($filters[array_search(0, $filters, true)]); } // Check if we have any real input. if (empty($filters)) { return true; } // Get the database object. $db = JFactory::getDbo(); $query = $db->getQuery(true); / * Create the query to get filters from the database. We do this for * two reasons: one, it allows us to ensure that the filters being used * are real; two, we need to sort the filters by taxonomy branch. / $query->select('t1.id, t1.title, t2.title AS branch') ->from($db->quoteName('#__finder_taxonomy') . ' AS t1') ->join('INNER', $db->quoteName('#__finder_taxonomy') . ' AS t2 ON t2.id = t1.parent_id') ->where('t1.state = 1') ->where('t1.access IN (' . $groups . ')') ->where('t1.id IN (' . implode(',', $filters) . ')') ->where('t2.state = 1') ->where('t2.access IN (' . $groups . ')'); // Load the filters. $db->setQuery($query); $results = $db->loadObjectList(); // Cleared filter branches. $cleared = array(); / * Sort the filter ids by branch. Because these filters are designed to * override and further narrow the items selected in the static filter, * we will clear the values from the static filter on a branch by * branch basis before adding the dynamic filters. So, if the static * filter defines a type filter of "articles" and three "category" * filters but the user only limits the category further, the category * filters will be flushed but the type filters will not. / foreach ($results as $result) { // Check if the branch has been cleared. if (!in_array($result->branch, $cleared, true)) { // Clear the branch. $this->filters[$result->branch] = array(); // Add the branch to the cleared list. $cleared[] = $result->branch; } // Add the filter to the list. $this->filters[$result->branch][$result->title] = (int) $result->id; } return true; } /* * Method to process the query date filters to determine start and end * date limitations. * * @param string $date1 The first date filter. * @param string $date2 The second date filter. * @param string $when1 The first date modifier. * @param string $when2 The second date modifier. * * @return boolean True on success. * * @since 2.5 / protected function processDates($date1, $date2, $when1, $when2) { // Clean up the inputs. $date1 = trim(StringHelper::strtolower($date1)); $date2 = trim(StringHelper::strtolower($date2)); $when1 = trim(StringHelper::strtolower($when1)); $when2 = trim(StringHelper::strtolower($when2)); // Get the time offset. $offset = JFactory::getApplication()->get('offset'); // Array of allowed when values. $whens = array('before', 'after', 'exact'); // The value of 'today' is a special case that we need to handle. if ($date1 === StringHelper::strtolower(JText::_('COM_FINDER_QUERY_FILTER_TODAY'))) { $date1 = JFactory::getDate('now', $offset)->format('%Y-%m-%d'); } // Try to parse the date string. $date = JFactory::getDate($date1, $offset); // Check if the date was parsed successfully. if ($date->toUnix() !== null) { // Set the date filter. $this->date1 = $date->toSql(); $this->when1 = in_array($when1, $whens, true) ? $when1 : 'before'; } // The value of 'today' is a special case that we need to handle. if ($date2 === StringHelper::strtolower(JText::_('COM_FINDER_QUERY_FILTER_TODAY'))) { $date2 = JFactory::getDate('now', $offset)->format('%Y-%m-%d'); } // Try to parse the date string. $date = JFactory::getDate($date2, $offset); // Check if the date was parsed successfully. if ($date->toUnix() !== null) { // Set the date filter. $this->date2 = $date->toSql(); $this->when2 = in_array($when2, $whens, true) ? $when2 : 'before'; } return true; } /* * Method to process the query input string and extract required, optional, * and excluded tokens; taxonomy filters; and date filters. * * @param string $input The query input string. * @param string $lang The query input language. * @param string $mode The query matching mode. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / protected function processString($input, $lang, $mode) { // Clean up the input string. $input = html_entity_decode($input, ENT_QUOTES, 'UTF-8'); $input = StringHelper::strtolower($input); $input = preg_replace('#\s+#mi', ' ', $input); $input = trim($input); $debug = JFactory::getConfig()->get('debug_lang'); / * First, we need to handle string based modifiers. String based * modifiers could potentially include things like "category:blah" or * "before:2009-10-21" or "type:article", etc. / $patterns = array( 'before' => JText::_('COM_FINDER_FILTER_WHEN_BEFORE'), 'after' => JText::_('COM_FINDER_FILTER_WHEN_AFTER'), ); // Add the taxonomy branch titles to the possible patterns. foreach (FinderIndexerTaxonomy::getBranchTitles() as $branch) { // Add the pattern. $patterns[$branch] = StringHelper::strtolower(JText::_(FinderHelperLanguage::branchSingular($branch))); } // Container for search terms and phrases. $terms = array(); $phrases = array(); // Cleared filter branches. $cleared = array(); / * Compile the suffix pattern. This is used to match the values of the * filter input string. Single words can be input directly, multi-word * values have to be wrapped in double quotes. / $quotes = html_entity_decode('‘’'', ENT_QUOTES, 'UTF-8'); $suffix = '(([\w\d' . $quotes . '-]+)\|\"([\w\d\s' . $quotes . '-]+)\")'; / * Iterate through the possible filter patterns and search for matches. * We need to match the key, colon, and a value pattern for the match * to be valid. / foreach ($patterns as $modifier => $pattern) { $matches = array(); if ($debug) { $pattern = substr($pattern, 2, -2); } // Check if the filter pattern is in the input string. if (preg_match('#' . $pattern . '\s:\s' . $suffix . '#mi', $input, $matches)) { // Get the value given to the modifier. $value = isset($matches[3]) ? $matches[3] : $matches[1]; // Now we have to handle the filter string. switch ($modifier) { // Handle a before and after date filters. case 'before': case 'after': { // Get the time offset. $offset = JFactory::getApplication()->get('offset'); // Array of allowed when values. $whens = array('before', 'after', 'exact'); // The value of 'today' is a special case that we need to handle. if ($value === StringHelper::strtolower(JText::_('COM_FINDER_QUERY_FILTER_TODAY'))) { $value = JFactory::getDate('now', $offset)->format('%Y-%m-%d'); } // Try to parse the date string. $date = JFactory::getDate($value, $offset); // Check if the date was parsed successfully. if ($date->toUnix() !== null) { // Set the date filter. $this->date1 = $date->toSql(); $this->when1 = in_array($modifier, $whens, true) ? $modifier : 'before'; } break; } // Handle a taxonomy branch filter. default: { // Try to find the node id. $return = FinderIndexerTaxonomy::getNodeByTitle($modifier, $value); // Check if the node id was found. if ($return) { // Check if the branch has been cleared. if (!in_array($modifier, $cleared, true)) { // Clear the branch. $this->filters[$modifier] = array(); // Add the branch to the cleared list. $cleared[] = $modifier; } // Add the filter to the list. $this->filters[$modifier][$return->title] = (int) $return->id; } break; } } // Clean up the input string again. $input = str_replace($matches[0], '', $input); $input = preg_replace('#\s+#mi', ' ', $input); $input = trim($input); } } / * Extract the tokens enclosed in double quotes so that we can handle * them as phrases. / if (StringHelper::strpos($input, '"') !== false) { $matches = array(); // Extract the tokens enclosed in double quotes. if (preg_match_all('#\"([^"]+)\"#m', $input, $matches)) { / * One or more phrases were found so we need to iterate through * them, tokenize them as phrases, and remove them from the raw * input string before we move on to the next processing step. / foreach ($matches[1] as $key => $match) { // Find the complete phrase in the input string. $pos = StringHelper::strpos($input, $matches[0][$key]); $len = StringHelper::strlen($matches[0][$key]); // Add any terms that are before this phrase to the stack. if (trim(StringHelper::substr($input, 0, $pos))) { $terms = array_merge($terms, explode(' ', trim(StringHelper::substr($input, 0, $pos)))); } // Strip out everything up to and including the phrase. $input = StringHelper::substr($input, $pos + $len); // Clean up the input string again. $input = preg_replace('#\s+#mi', ' ', $input); $input = trim($input); // Get the number of words in the phrase. $parts = explode(' ', $match); // Check if the phrase is longer than three words. if (count($parts) > 3) { / * If the phrase is longer than three words, we need to * break it down into smaller chunks of phrases that * are less than or equal to three words. We overlap * the chunks so that we can ensure that a match is * found for the complete phrase and not just portions * of it. / for ($i = 0, $c = count($parts); $i < $c; $i += 2) { // Set up the chunk. $chunk = array(); // The chunk has to be assembled based on how many // pieces are available to use. switch ($c - $i) { / * If only one word is left, we can break from * the switch and loop because the last word * was already used at the end of the last * chunk. / case 1: break 2; // If there words are left, we use them both as // the last chunk of the phrase and we're done. case 2: $chunk[] = $parts[$i]; $chunk[] = $parts[$i + 1]; break; // If there are three or more words left, we // build a three word chunk and continue on. default: $chunk[] = $parts[$i]; $chunk[] = $parts[$i + 1]; $chunk[] = $parts[$i + 2]; break; } // If the chunk is not empty, add it as a phrase. if (count($chunk)) { $phrases[] = implode(' ', $chunk); $terms[] = implode(' ', $chunk); } } } else { // The phrase is <= 3 words so we can use it as is. $phrases[] = $match; $terms[] = $match; } } } } // Add the remaining terms if present. if ((bool) $input) { $terms = array_merge($terms, explode(' ', $input)); } // An array of our boolean operators. $operator => $translation $operators = array( 'AND' => StringHelper::strtolower(JText::_('COM_FINDER_QUERY_OPERATOR_AND')), 'OR' => StringHelper::strtolower(JText::_('COM_FINDER_QUERY_OPERATOR_OR')), 'NOT' => StringHelper::strtolower(JText::_('COM_FINDER_QUERY_OPERATOR_NOT')), ); // If language debugging is enabled you need to ignore the debug strings in matching. if (JDEBUG) { $debugStrings = array('', '??'); $operators = str_replace($debugStrings, '', $operators); } / * Iterate through the terms and perform any sorting that needs to be * done based on boolean search operators. Terms that are before an * and/or/not modifier have to be handled in relation to their operator. / for ($i = 0, $c = count($terms); $i < $c; $i++) { // Check if the term is followed by an operator that we understand. if (isset($terms[$i + 1]) && in_array($terms[$i + 1], $operators, true)) { // Get the operator mode. $op = array_search($terms[$i + 1], $operators, true); // Handle the AND operator. if ($op === 'AND' && isset($terms[$i + 2])) { // Tokenize the current term. $token = FinderIndexerHelper::tokenize($terms[$i], $lang, true); // Todo: The previous function call may return an array, which seems not to be handled by the next one, which expects an object $token = $this->getTokenData($token); // Set the required flag. $token->required = true; // Add the current token to the stack. $this->included[] = $token; $this->highlight = array_merge($this->highlight, array_keys($token->matches)); // Skip the next token (the mode operator). $this->operators[] = $terms[$i + 1]; // Tokenize the term after the next term (current plus two). $other = FinderIndexerHelper::tokenize($terms[$i + 2], $lang, true); $other = $this->getTokenData($other); // Set the required flag. $other->required = true; // Add the token after the next token to the stack. $this->included[] = $other; $this->highlight = array_merge($this->highlight, array_keys($other->matches)); // Remove the processed phrases if possible. if (($pk = array_search($terms[$i], $phrases, true)) !== false) { unset($phrases[$pk]); } if (($pk = array_search($terms[$i + 2], $phrases, true)) !== false) { unset($phrases[$pk]); } // Remove the processed terms. unset($terms[$i], $terms[$i + 1], $terms[$i + 2]); // Adjust the loop. $i += 2; continue; } // Handle the OR operator. elseif ($op === 'OR' && isset($terms[$i + 2])) { // Tokenize the current term. $token = FinderIndexerHelper::tokenize($terms[$i], $lang, true); $token = $this->getTokenData($token); // Set the required flag. $token->required = false; // Add the current token to the stack. if ((bool) $token->matches) { $this->included[] = $token; $this->highlight = array_merge($this->highlight, array_keys($token->matches)); } else { $this->ignored[] = $token; } // Skip the next token (the mode operator). $this->operators[] = $terms[$i + 1]; // Tokenize the term after the next term (current plus two). $other = FinderIndexerHelper::tokenize($terms[$i + 2], $lang, true); $other = $this->getTokenData($other); // Set the required flag. $other->required = false; // Add the token after the next token to the stack. if ((bool) $other->matches) { $this->included[] = $other; $this->highlight = array_merge($this->highlight, array_keys($other->matches)); } else { $this->ignored[] = $other; } // Remove the processed phrases if possible. if (($pk = array_search($terms[$i], $phrases, true)) !== false) { unset($phrases[$pk]); } if (($pk = array_search($terms[$i + 2], $phrases, true)) !== false) { unset($phrases[$pk]); } // Remove the processed terms. unset($terms[$i], $terms[$i + 1], $terms[$i + 2]); // Adjust the loop. $i += 2; continue; } } // Handle an orphaned OR operator. elseif (isset($terms[$i + 1]) && array_search($terms[$i], $operators, true) === 'OR') { // Skip the next token (the mode operator). $this->operators[] = $terms[$i]; // Tokenize the next term (current plus one). $other = FinderIndexerHelper::tokenize($terms[$i + 1], $lang, true); $other = $this->getTokenData($other); // Set the required flag. $other->required = false; // Add the token after the next token to the stack. if ((bool) $other->matches) { $this->included[] = $other; $this->highlight = array_merge($this->highlight, array_keys($other->matches)); } else { $this->ignored[] = $other; } // Remove the processed phrase if possible. if (($pk = array_search($terms[$i + 1], $phrases, true)) !== false) { unset($phrases[$pk]); } // Remove the processed terms. unset($terms[$i], $terms[$i + 1]); // Adjust the loop. $i++; continue; } // Handle the NOT operator. elseif (isset($terms[$i + 1]) && array_search($terms[$i], $operators, true) === 'NOT') { // Skip the next token (the mode operator). $this->operators[] = $terms[$i]; // Tokenize the next term (current plus one). $other = FinderIndexerHelper::tokenize($terms[$i + 1], $lang, true); $other = $this->getTokenData($other); // Set the required flag. $other->required = false; // Add the next token to the stack. if ((bool) $other->matches) { $this->excluded[] = $other; } else { $this->ignored[] = $other; } // Remove the processed phrase if possible. if (($pk = array_search($terms[$i + 1], $phrases, true)) !== false) { unset($phrases[$pk]); } // Remove the processed terms. unset($terms[$i], $terms[$i + 1]); // Adjust the loop. $i++; continue; } } / * Iterate through any search phrases and tokenize them. We handle * phrases as autonomous units and do not break them down into two and * three word combinations. / for ($i = 0, $c = count($phrases); $i < $c; $i++) { // Tokenize the phrase. $token = FinderIndexerHelper::tokenize($phrases[$i], $lang, true); $token = $this->getTokenData($token); // Set the required flag. $token->required = true; // Add the current token to the stack. $this->included[] = $token; $this->highlight = array_merge($this->highlight, array_keys($token->matches)); // Remove the processed term if possible. if (($pk = array_search($phrases[$i], $terms, true)) !== false) { unset($terms[$pk]); } // Remove the processed phrase. unset($phrases[$i]); } / * Handle any remaining tokens using the standard processing mechanism. / if ((bool) $terms) { // Tokenize the terms. $terms = implode(' ', $terms); $tokens = FinderIndexerHelper::tokenize($terms, $lang, false); // Make sure we are working with an array. $tokens = is_array($tokens) ? $tokens : array($tokens); // Get the token data and required state for all the tokens. foreach ($tokens as $token) { // Get the token data. $token = $this->getTokenData($token); // Set the required flag for the token. $token->required = $mode === 'AND' ? (!$token->phrase) : false; // Add the token to the appropriate stack. if ($token->required \|\| (bool) $token->matches) { $this->included[] = $token; $this->highlight = array_merge($this->highlight, array_keys($token->matches)); } else { $this->ignored[] = $token; } } } return true; } /* * Method to get the base and similar term ids and, if necessary, suggested * term data from the database. The terms ids are identified based on a * 'like' match in MySQL and/or a common stem. If no term ids could be * found, then we know that we will not be able to return any results for * that term and we should try to find a similar term to use that we can * match so that we can suggest the alternative search query to the user. * * @param FinderIndexerToken $token A FinderIndexerToken object. * * @return FinderIndexerToken A FinderIndexerToken object. * * @since 2.5 * @throws Exception on database error. / protected function getTokenData($token) { // Get the database object. $db = JFactory::getDbo(); // Create a database query to build match the token. $query = $db->getQuery(true) ->select('t.term, t.term_id') ->from('#__finder_terms AS t'); / * If the token is a phrase, the lookup process is fairly simple. If * the token is a word, it is a little more complicated. We have to * create two queries to lookup the term and the stem respectively, * then union the result sets together. This is MUCH faster than using * an or condition in the database query. / if ($token->phrase) { // Add the phrase to the query. $query->where('t.term = ' . $db->quote($token->term)) ->where('t.phrase = 1'); } else { // Add the term to the query. $query->where('t.term = ' . $db->quote($token->term)) ->where('t.phrase = 0'); // Clone the query, replace the WHERE clause. $sub = clone $query; $sub->clear('where'); $sub->where('t.stem = ' . $db->quote($token->stem)); $sub->where('t.phrase = 0'); // Union the two queries. $query->union($sub); } // Get the terms. $db->setQuery($query); $matches = $db->loadObjectList(); // Check the matching terms. if ((bool) $matches) { // Add the matches to the token. for ($i = 0, $c = count($matches); $i < $c; $i++) { $token->matches[$matches[$i]->term] = (int) $matches[$i]->term_id; } } // If no matches were found, try to find a similar but better token. if (empty($token->matches)) { // Create a database query to get the similar terms. // TODO: PostgreSQL doesn't support SOUNDEX out of the box $query->clear() ->select('DISTINCT t.term_id AS id, t.term AS term') ->from('#__finder_terms AS t') // ->where('t.soundex = ' . soundex($db->quote($token->term))) ->where('t.soundex = SOUNDEX(' . $db->quote($token->term) . ')') ->where('t.phrase = ' . (int) $token->phrase); // Get the terms. $db->setQuery($query); $results = $db->loadObjectList(); // Check if any similar terms were found. if (empty($results)) { return $token; } // Stack for sorting the similar terms. $suggestions = array(); // Get the levnshtein distance for all suggested terms. foreach ($results as $sk => $st) { // Get the levenshtein distance between terms. $distance = levenshtein($st->term, $token->term); // Make sure the levenshtein distance isn't over 50. if ($distance < 50) { $suggestions[$sk] = $distance; } } // Sort the suggestions. asort($suggestions, SORT_NUMERIC); // Get the closest match. $keys = array_keys($suggestions); $key = $keys[0]; // Add the suggested term. $token->suggestion = $results[$key]->term; } return $token; } } ��driver/mysql.php��0000604��00000046020�15245703510�0007715 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2012 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; jimport('joomla.filesystem.file'); /* * Indexer class supporting MySQL(i) for the Finder indexer package. * * The indexer class provides the core functionality of the Finder * search engine. It is responsible for adding and updating the * content links table; extracting and scoring tokens; and maintaining * all referential information for the content. * * Note: All exceptions thrown from within this class should be caught * by the controller. * * @since 3.0 / class FinderIndexerDriverMysql extends FinderIndexer { /* * Method to index a content item. * * @param FinderIndexerResult $item The content item to index. * @param string $format The format of the content. [optional] * * @return integer The ID of the record in the links table. * * @since 3.0 * @throws Exception on database error. / public function index($item, $format = 'html') { // Mark beforeIndexing in the profiler. static::$profiler ? static::$profiler->mark('beforeIndexing') : null; $db = $this->db; $nd = $db->getNullDate(); // Check if the item is in the database. $query = $db->getQuery(true) ->select($db->quoteName('link_id') . ', ' . $db->quoteName('md5sum')) ->from($db->quoteName('#__finder_links')) ->where($db->quoteName('url') . ' = ' . $db->quote($item->url)); // Load the item from the database. $db->setQuery($query); $link = $db->loadObject(); // Get the indexer state. $state = static::getState(); // Get the signatures of the item. $curSig = static::getSignature($item); $oldSig = isset($link->md5sum) ? $link->md5sum : null; // Get the other item information. $linkId = empty($link->link_id) ? null : $link->link_id; $isNew = empty($link->link_id) ? true : false; // Check the signatures. If they match, the item is up to date. if (!$isNew && $curSig == $oldSig) { return $linkId; } / * If the link already exists, flush all the term maps for the item. * Maps are stored in 16 tables so we need to iterate through and flush * each table one at a time. / if (!$isNew) { for ($i = 0; $i <= 15; $i++) { // Flush the maps for the link. $query->clear() ->delete($db->quoteName('#__finder_links_terms' . dechex($i))) ->where($db->quoteName('link_id') . ' = ' . (int) $linkId); $db->setQuery($query); $db->execute(); } // Remove the taxonomy maps. FinderIndexerTaxonomy::removeMaps($linkId); } // Mark afterUnmapping in the profiler. static::$profiler ? static::$profiler->mark('afterUnmapping') : null; // Perform cleanup on the item data. $item->publish_start_date = (int) $item->publish_start_date != 0 ? $item->publish_start_date : $nd; $item->publish_end_date = (int) $item->publish_end_date != 0 ? $item->publish_end_date : $nd; $item->start_date = (int) $item->start_date != 0 ? $item->start_date : $nd; $item->end_date = (int) $item->end_date != 0 ? $item->end_date : $nd; // Prepare the item description. $item->description = FinderIndexerHelper::parse($item->summary); / * Now, we need to enter the item into the links table. If the item * already exists in the database, we need to use an UPDATE query. * Otherwise, we need to use an INSERT to get the link id back. / if ($isNew) { $columnsArray = array( $db->quoteName('url'), $db->quoteName('route'), $db->quoteName('title'), $db->quoteName('description'), $db->quoteName('indexdate'), $db->quoteName('published'), $db->quoteName('state'), $db->quoteName('access'), $db->quoteName('language'), $db->quoteName('type_id'), $db->quoteName('object'), $db->quoteName('publish_start_date'), $db->quoteName('publish_end_date'), $db->quoteName('start_date'), $db->quoteName('end_date'), $db->quoteName('list_price'), $db->quoteName('sale_price') ); // Insert the link. $query->clear() ->insert($db->quoteName('#__finder_links')) ->columns($columnsArray) ->values( $db->quote($item->url) . ', ' . $db->quote($item->route) . ', ' . $db->quote($item->title) . ', ' . $db->quote($item->description) . ', ' . $query->currentTimestamp() . ', ' . '1, ' . (int) $item->state . ', ' . (int) $item->access . ', ' . $db->quote($item->language) . ', ' . (int) $item->type_id . ', ' . $db->quote(serialize($item)) . ', ' . $db->quote($item->publish_start_date) . ', ' . $db->quote($item->publish_end_date) . ', ' . $db->quote($item->start_date) . ', ' . $db->quote($item->end_date) . ', ' . (double) ($item->list_price ?: 0) . ', ' . (double) ($item->sale_price ?: 0) ); $db->setQuery($query); $db->execute(); // Get the link id. $linkId = (int) $db->insertid(); } else { // Update the link. $query->clear() ->update($db->quoteName('#__finder_links')) ->set($db->quoteName('route') . ' = ' . $db->quote($item->route)) ->set($db->quoteName('title') . ' = ' . $db->quote($item->title)) ->set($db->quoteName('description') . ' = ' . $db->quote($item->description)) ->set($db->quoteName('indexdate') . ' = ' . $query->currentTimestamp()) ->set($db->quoteName('state') . ' = ' . (int) $item->state) ->set($db->quoteName('access') . ' = ' . (int) $item->access) ->set($db->quoteName('language') . ' = ' . $db->quote($item->language)) ->set($db->quoteName('type_id') . ' = ' . (int) $item->type_id) ->set($db->quoteName('object') . ' = ' . $db->quote(serialize($item))) ->set($db->quoteName('publish_start_date') . ' = ' . $db->quote($item->publish_start_date)) ->set($db->quoteName('publish_end_date') . ' = ' . $db->quote($item->publish_end_date)) ->set($db->quoteName('start_date') . ' = ' . $db->quote($item->start_date)) ->set($db->quoteName('end_date') . ' = ' . $db->quote($item->end_date)) ->set($db->quoteName('list_price') . ' = ' . (double) ($item->list_price ?: 0)) ->set($db->quoteName('sale_price') . ' = ' . (double) ($item->sale_price ?: 0)) ->where('link_id = ' . (int) $linkId); $db->setQuery($query); $db->execute(); } // Set up the variables we will need during processing. $count = 0; // Mark afterLinking in the profiler. static::$profiler ? static::$profiler->mark('afterLinking') : null; // Truncate the tokens tables. $db->truncateTable('#__finder_tokens'); // Truncate the tokens aggregate table. $db->truncateTable('#__finder_tokens_aggregate'); / * Process the item's content. The items can customize their * processing instructions to define extra properties to process * or rearrange how properties are weighted. / foreach ($item->getInstructions() as $group => $properties) { // Iterate through the properties of the group. foreach ($properties as $property) { // Check if the property exists in the item. if (empty($item->$property)) { continue; } // Tokenize the property. if (is_array($item->$property)) { // Tokenize an array of content and add it to the database. foreach ($item->$property as $ip) { / * If the group is path, we need to a few extra processing * steps to strip the extension and convert slashes and dashes * to spaces. / if ($group === static::PATH_CONTEXT) { $ip = JFile::stripExt($ip); $ip = str_replace(array('/', '-'), ' ', $ip); } // Tokenize a string of content and add it to the database. $count += $this->tokenizeToDb($ip, $group, $item->language, $format); // Check if we're approaching the memory limit of the token table. if ($count > static::$state->options->get('memory_table_limit', 30000)) { $this->toggleTables(false); } } } else { / * If the group is path, we need to a few extra processing * steps to strip the extension and convert slashes and dashes * to spaces. / if ($group === static::PATH_CONTEXT) { $item->$property = JFile::stripExt($item->$property); $item->$property = str_replace('/', ' ', $item->$property); $item->$property = str_replace('-', ' ', $item->$property); } // Tokenize a string of content and add it to the database. $count += $this->tokenizeToDb($item->$property, $group, $item->language, $format); // Check if we're approaching the memory limit of the token table. if ($count > static::$state->options->get('memory_table_limit', 30000)) { $this->toggleTables(false); } } } } / * Process the item's taxonomy. The items can customize their * taxonomy mappings to define extra properties to map. / foreach ($item->getTaxonomy() as $branch => $nodes) { // Iterate through the nodes and map them to the branch. foreach ($nodes as $node) { // Add the node to the tree. $nodeId = FinderIndexerTaxonomy::addNode($branch, $node->title, $node->state, $node->access); // Add the link => node map. FinderIndexerTaxonomy::addMap($linkId, $nodeId); } } // Mark afterProcessing in the profiler. static::$profiler ? static::$profiler->mark('afterProcessing') : null; / * At this point, all of the item's content has been parsed, tokenized * and inserted into the #__finder_tokens table. Now, we need to * aggregate all the data into that table into a more usable form. The * aggregated data will be inserted into #__finder_tokens_aggregate * table. / $query = 'INSERT INTO ' . $db->quoteName('#__finder_tokens_aggregate') . ' (' . $db->quoteName('term_id') . ', ' . $db->quoteName('map_suffix') . ', ' . $db->quoteName('term') . ', ' . $db->quoteName('stem') . ', ' . $db->quoteName('common') . ', ' . $db->quoteName('phrase') . ', ' . $db->quoteName('term_weight') . ', ' . $db->quoteName('context') . ', ' . $db->quoteName('context_weight') . ', ' . $db->quoteName('total_weight') . ', ' . $db->quoteName('language') . ')' . ' SELECT' . ' COALESCE(t.term_id, 0), \'\', t1.term, t1.stem, t1.common, t1.phrase, t1.weight, t1.context,' . ' ROUND( t1.weight COUNT( t2.term ) * %F, 8 ) AS context_weight, 0, t1.language' . ' FROM (' . ' SELECT DISTINCT t1.term, t1.stem, t1.common, t1.phrase, t1.weight, t1.context, t1.language' . ' FROM ' . $db->quoteName('#__finder_tokens') . ' AS t1' . ' WHERE t1.context = %d' . ' ) AS t1' . ' JOIN ' . $db->quoteName('#__finder_tokens') . ' AS t2 ON t2.term = t1.term' . ' LEFT JOIN ' . $db->quoteName('#__finder_terms') . ' AS t ON t.term = t1.term' . ' WHERE t2.context = %d' . ' GROUP BY t1.term, t.term_id, t1.term, t1.stem, t1.common, t1.phrase, t1.weight, t1.context, t1.language' . ' ORDER BY t1.term DESC'; // Iterate through the contexts and aggregate the tokens per context. foreach ($state->weights as $context => $multiplier) { // Run the query to aggregate the tokens for this context.. $db->setQuery(sprintf($query, $multiplier, $context, $context)); $db->execute(); } // Mark afterAggregating in the profiler. static::$profiler ? static::$profiler->mark('afterAggregating') : null; /* * When we pulled down all of the aggregate data, we did a LEFT JOIN * over the terms table to try to find all the term ids that * already exist for our tokens. If any of the rows in the aggregate * table have a term of 0, then no term record exists for that * term so we need to add it to the terms table. / $db->setQuery( 'INSERT IGNORE INTO ' . $db->quoteName('#__finder_terms') . ' (' . $db->quoteName('term') . ', ' . $db->quoteName('stem') . ', ' . $db->quoteName('common') . ', ' . $db->quoteName('phrase') . ', ' . $db->quoteName('weight') . ', ' . $db->quoteName('soundex') . ', ' . $db->quoteName('language') . ')' . ' SELECT ta.term, ta.stem, ta.common, ta.phrase, ta.term_weight, SOUNDEX(ta.term), ta.language' . ' FROM ' . $db->quoteName('#__finder_tokens_aggregate') . ' AS ta' . ' WHERE ta.term_id = 0' . ' GROUP BY ta.term, ta.stem, ta.common, ta.phrase, ta.term_weight, SOUNDEX(ta.term), ta.language' ); $db->execute(); / * Now, we just inserted a bunch of new records into the terms table * so we need to go back and update the aggregate table with all the * new term ids. / $query = $db->getQuery(true) ->update($db->quoteName('#__finder_tokens_aggregate') . ' AS ta') ->join('INNER', $db->quoteName('#__finder_terms') . ' AS t ON t.term = ta.term') ->set('ta.term_id = t.term_id') ->where('ta.term_id = 0'); $db->setQuery($query); $db->execute(); // Mark afterTerms in the profiler. static::$profiler ? static::$profiler->mark('afterTerms') : null; / * After we've made sure that all of the terms are in the terms table * and the aggregate table has the correct term ids, we need to update * the links counter for each term by one. / $query->clear() ->update($db->quoteName('#__finder_terms') . ' AS t') ->join('INNER', $db->quoteName('#__finder_tokens_aggregate') . ' AS ta ON ta.term_id = t.term_id') ->set('t.' . $db->quoteName('links') . ' = t.links + 1'); $db->setQuery($query); $db->execute(); // Mark afterTerms in the profiler. static::$profiler ? static::$profiler->mark('afterTerms') : null; / * Before we can insert all of the mapping rows, we have to figure out * which mapping table the rows need to be inserted into. The mapping * table for each term is based on the first character of the md5 of * the first character of the term. In php, it would be expressed as * substr(md5(substr($token, 0, 1)), 0, 1) / $query->clear() ->update($db->quoteName('#__finder_tokens_aggregate')) ->set($db->quoteName('map_suffix') . ' = SUBSTR(MD5(SUBSTR(' . $db->quoteName('term') . ', 1, 1)), 1, 1)'); $db->setQuery($query); $db->execute(); / * At this point, the aggregate table contains a record for each * term in each context. So, we're going to pull down all of that * data while grouping the records by term and add all of the * sub-totals together to arrive at the final total for each token for * this link. Then, we insert all of that data into the appropriate * mapping table. / for ($i = 0; $i <= 15; $i++) { // Get the mapping table suffix. $suffix = dechex($i); / * We have to run this query 16 times, one for each link => term * mapping table. / $db->setQuery( 'INSERT INTO ' . $db->quoteName('#__finder_links_terms' . $suffix) . ' (' . $db->quoteName('link_id') . ', ' . $db->quoteName('term_id') . ', ' . $db->quoteName('weight') . ')' . ' SELECT ' . (int) $linkId . ', ' . $db->quoteName('term_id') . ',' . ' ROUND(SUM(' . $db->quoteName('context_weight') . '), 8)' . ' FROM ' . $db->quoteName('#__finder_tokens_aggregate') . ' WHERE ' . $db->quoteName('map_suffix') . ' = ' . $db->quote($suffix) . ' GROUP BY ' . $db->quoteName('term') . ', ' . $db->quoteName('term_id') . ' ORDER BY ' . $db->quoteName('term') . ' DESC' ); $db->execute(); } // Mark afterMapping in the profiler. static::$profiler ? static::$profiler->mark('afterMapping') : null; // Update the signature. $object = serialize($item); $query->clear() ->update($db->quoteName('#__finder_links')) ->set($db->quoteName('md5sum') . ' = ' . $db->quote($curSig)) ->set($db->quoteName('object') . ' = ' . $db->quote($object)) ->where($db->quoteName('link_id') . ' = ' . $db->quote($linkId)); $db->setQuery($query); $db->execute(); // Mark afterSigning in the profiler. static::$profiler ? static::$profiler->mark('afterSigning') : null; // Truncate the tokens tables. $db->truncateTable('#__finder_tokens'); // Truncate the tokens aggregate table. $db->truncateTable('#__finder_tokens_aggregate'); // Toggle the token tables back to memory tables. $this->toggleTables(true); // Mark afterTruncating in the profiler. static::$profiler ? static::$profiler->mark('afterTruncating') : null; return $linkId; } /* * Method to optimize the index. We use this method to remove unused terms * and any other optimizations that might be necessary. * * @return boolean True on success. * * @since 3.0 * @throws Exception on database error. / public function optimize() { // Get the database object. $db = $this->db; $query = $db->getQuery(true); // Delete all orphaned terms. $query->delete($db->quoteName('#__finder_terms')) ->where($db->quoteName('links') . ' <= 0'); $db->setQuery($query); $db->execute(); // Optimize the links table. $db->setQuery('OPTIMIZE TABLE ' . $db->quoteName('#__finder_links')); $db->execute(); for ($i = 0; $i <= 15; $i++) { // Optimize the terms mapping table. $db->setQuery('OPTIMIZE TABLE ' . $db->quoteName('#__finder_links_terms' . dechex($i))); $db->execute(); } // Optimize the filters table. $db->setQuery('OPTIMIZE TABLE ' . $db->quoteName('#__finder_filters')); $db->execute(); // Optimize the terms common table. $db->setQuery('OPTIMIZE TABLE ' . $db->quoteName('#__finder_terms_common')); $db->execute(); // Optimize the types table. $db->setQuery('OPTIMIZE TABLE ' . $db->quoteName('#__finder_types')); $db->execute(); // Remove the orphaned taxonomy nodes. FinderIndexerTaxonomy::removeOrphanNodes(); // Optimize the taxonomy mapping table. $db->setQuery('OPTIMIZE TABLE ' . $db->quoteName('#__finder_taxonomy_map')); $db->execute(); // Optimize the taxonomy table. $db->setQuery('OPTIMIZE TABLE ' . $db->quoteName('#__finder_taxonomy')); $db->execute(); return true; } /* * Method to switch the token tables from Memory tables to MyISAM tables * when they are close to running out of memory. * * @param boolean $memory Flag to control how they should be toggled. * * @return boolean True on success. * * @since 3.0 * @throws Exception on database error. / protected function toggleTables($memory) { static $state; // Get the database adapter. $db = $this->db; // Check if we are setting the tables to the Memory engine. if ($memory === true && $state !== true) { // Set the tokens table to Memory. $db->setQuery('ALTER TABLE ' . $db->quoteName('#__finder_tokens') . ' ENGINE = MEMORY'); $db->execute(); // Set the tokens aggregate table to Memory. $db->setQuery('ALTER TABLE ' . $db->quoteName('#__finder_tokens_aggregate') . ' ENGINE = MEMORY'); $db->execute(); // Set the internal state. $state = $memory; } // We must be setting the tables to the MyISAM engine. elseif ($memory === false && $state !== false) { // Set the tokens table to MyISAM. $db->setQuery('ALTER TABLE ' . $db->quoteName('#__finder_tokens') . ' ENGINE = MYISAM'); $db->execute(); // Set the tokens aggregate table to MyISAM. $db->setQuery('ALTER TABLE ' . $db->quoteName('#__finder_tokens_aggregate') . ' ENGINE = MYISAM'); $db->execute(); // Set the internal state. $state = $memory; } return true; } } ��driver/postgresql.php��0000604��00000037550�15245703510�0010763 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2012 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; use Joomla\CMS\Factory; use Joomla\String\StringHelper; jimport('joomla.filesystem.file'); /* * Indexer class supporting PostgreSQL for the Finder indexer package. * * @since 3.0 / class FinderIndexerDriverPostgresql extends FinderIndexer { /* * Method to index a content item. * * @param FinderIndexerResult $item The content item to index. * @param string $format The format of the content. [optional] * * @return integer The ID of the record in the links table. * * @since 3.0 * @throws Exception on database error. / public function index($item, $format = 'html') { // Mark beforeIndexing in the profiler. static::$profiler ? static::$profiler->mark('beforeIndexing') : null; $db = $this->db; $nd = $db->getNullDate(); // Check if the item is in the database. $query = $db->getQuery(true) ->select($db->quoteName('link_id') . ', ' . $db->quoteName('md5sum')) ->from($db->quoteName('#__finder_links')) ->where($db->quoteName('url') . ' = ' . $db->quote($item->url)); // Load the item from the database. $db->setQuery($query); $link = $db->loadObject(); // Get the indexer state. $state = static::getState(); // Get the signatures of the item. $curSig = static::getSignature($item); $oldSig = isset($link->md5sum) ? $link->md5sum : null; // Get the other item information. $linkId = empty($link->link_id) ? null : $link->link_id; $isNew = empty($link->link_id) ? true : false; // Check the signatures. If they match, the item is up to date. if (!$isNew && $curSig === $oldSig) { return $linkId; } / * If the link already exists, flush all the term maps for the item. * Maps are stored in 16 tables so we need to iterate through and flush * each table one at a time. / if (!$isNew) { for ($i = 0; $i <= 15; $i++) { // Flush the maps for the link. $query->clear() ->delete($db->quoteName('#__finder_links_terms' . dechex($i))) ->where($db->quoteName('link_id') . ' = ' . (int) $linkId); $db->setQuery($query); $db->execute(); } // Remove the taxonomy maps. FinderIndexerTaxonomy::removeMaps($linkId); } // Mark afterUnmapping in the profiler. static::$profiler ? static::$profiler->mark('afterUnmapping') : null; // Perform cleanup on the item data. $item->publish_start_date = (int) $item->publish_start_date != 0 ? $item->publish_start_date : $nd; $item->publish_end_date = (int) $item->publish_end_date != 0 ? $item->publish_end_date : $nd; $item->start_date = (int) $item->start_date != 0 ? $item->start_date : $nd; $item->end_date = (int) $item->end_date != 0 ? $item->end_date : $nd; // Prepare the item description. $item->description = FinderIndexerHelper::parse($item->summary); / * Now, we need to enter the item into the links table. If the item * already exists in the database, we need to use an UPDATE query. * Otherwise, we need to use an INSERT to get the link id back. / $entry = new stdClass; $entry->url = $item->url; $entry->route = $item->route; $entry->title = $item->title; // We are shortening the description in order to not run into length issues with this field $entry->description = StringHelper::substr($item->description, 0, 32000); $entry->indexdate = Factory::getDate()->toSql(); $entry->state = (int) $item->state; $entry->access = (int) $item->access; $entry->language = $item->language; $entry->type_id = (int) $item->type_id; $entry->object = ''; $entry->publish_start_date = $item->publish_start_date; $entry->publish_end_date = $item->publish_end_date; $entry->start_date = $item->start_date; $entry->end_date = $item->end_date; $entry->list_price = (double) ($item->list_price ?: 0); $entry->sale_price = (double) ($item->sale_price ?: 0); if ($isNew) { // Insert the link and get its id. $db->insertObject('#__finder_links', $entry); $linkId = (int) $db->insertid(); } else { // Update the link. $entry->link_id = $linkId; $db->updateObject('#__finder_links', $entry, 'link_id'); } // Set up the variables we will need during processing. $count = 0; // Mark afterLinking in the profiler. static::$profiler ? static::$profiler->mark('afterLinking') : null; // Truncate the tokens tables. $db->truncateTable('#__finder_tokens'); // Truncate the tokens aggregate table. $db->truncateTable('#__finder_tokens_aggregate'); / * Process the item's content. The items can customize their * processing instructions to define extra properties to process * or rearrange how properties are weighted. / foreach ($item->getInstructions() as $group => $properties) { // Iterate through the properties of the group. foreach ($properties as $property) { // Check if the property exists in the item. if (empty($item->$property)) { continue; } // Tokenize the property. if (is_array($item->$property)) { // Tokenize an array of content and add it to the database. foreach ($item->$property as $ip) { / * If the group is path, we need to a few extra processing * steps to strip the extension and convert slashes and dashes * to spaces. / if ($group === static::PATH_CONTEXT) { $ip = JFile::stripExt($ip); $ip = str_replace(array('/', '-'), ' ', $ip); } // Tokenize a string of content and add it to the database. $count += $this->tokenizeToDb($ip, $group, $item->language, $format); // Check if we're approaching the memory limit of the token table. if ($count > static::$state->options->get('memory_table_limit', 30000)) { $this->toggleTables(false); } } } else { / * If the group is path, we need to a few extra processing * steps to strip the extension and convert slashes and dashes * to spaces. / if ($group === static::PATH_CONTEXT) { $item->$property = JFile::stripExt($item->$property); $item->$property = str_replace('/', ' ', $item->$property); $item->$property = str_replace('-', ' ', $item->$property); } // Tokenize a string of content and add it to the database. $count += $this->tokenizeToDb($item->$property, $group, $item->language, $format); // Check if we're approaching the memory limit of the token table. if ($count > static::$state->options->get('memory_table_limit', 30000)) { $this->toggleTables(false); } } } } / * Process the item's taxonomy. The items can customize their * taxonomy mappings to define extra properties to map. / foreach ($item->getTaxonomy() as $branch => $nodes) { // Iterate through the nodes and map them to the branch. foreach ($nodes as $node) { // Add the node to the tree. $nodeId = FinderIndexerTaxonomy::addNode($branch, $node->title, $node->state, $node->access); // Add the link => node map. FinderIndexerTaxonomy::addMap($linkId, $nodeId); } } // Mark afterProcessing in the profiler. static::$profiler ? static::$profiler->mark('afterProcessing') : null; / * At this point, all of the item's content has been parsed, tokenized * and inserted into the #__finder_tokens table. Now, we need to * aggregate all the data into that table into a more usable form. The * aggregated data will be inserted into #__finder_tokens_aggregate * table. / $query = 'INSERT INTO ' . $db->quoteName('#__finder_tokens_aggregate') . ' (' . $db->quoteName('term_id') . ', ' . $db->quoteName('map_suffix') . ', ' . $db->quoteName('term') . ', ' . $db->quoteName('stem') . ', ' . $db->quoteName('common') . ', ' . $db->quoteName('phrase') . ', ' . $db->quoteName('term_weight') . ', ' . $db->quoteName('context') . ', ' . $db->quoteName('context_weight') . ', ' . $db->quoteName('total_weight') . ', ' . $db->quoteName('language') . ')' . ' SELECT' . ' COALESCE(t.term_id, 0), \'\', t1.term, t1.stem, t1.common, t1.phrase, t1.weight, t1.context,' . ' ROUND( t1.weight COUNT( t2.term ) * %F, 8 ) AS context_weight, 0, t1.language' . ' FROM (' . ' SELECT DISTINCT t1.term, t1.stem, t1.common, t1.phrase, t1.weight, t1.context, t1.language' . ' FROM ' . $db->quoteName('#__finder_tokens') . ' AS t1' . ' WHERE t1.context = %d' . ' ) AS t1' . ' JOIN ' . $db->quoteName('#__finder_tokens') . ' AS t2 ON t2.term = t1.term AND t2.language = t1.language' . ' LEFT JOIN ' . $db->quoteName('#__finder_terms') . ' AS t ON t.term = t1.term ' . ' WHERE t2.context = %d' . ' GROUP BY t1.term, t.term_id, t1.stem, t1.common, t1.phrase, t1.weight, t1.context, t1.language' . ' ORDER BY t1.term DESC'; // Iterate through the contexts and aggregate the tokens per context. foreach ($state->weights as $context => $multiplier) { // Run the query to aggregate the tokens for this context.. $db->setQuery(sprintf($query, $multiplier, $context, $context)); $db->execute(); } // Mark afterAggregating in the profiler. static::$profiler ? static::$profiler->mark('afterAggregating') : null; /* * When we pulled down all of the aggregate data, we did a LEFT JOIN * over the terms table to try to find all the term ids that * already exist for our tokens. If any of the rows in the aggregate * table have a term of 0, then no term record exists for that * term so we need to add it to the terms table. / $db->setQuery( 'INSERT INTO ' . $db->quoteName('#__finder_terms') . ' (' . $db->quoteName('term') . ', ' . $db->quoteName('stem') . ', ' . $db->quoteName('common') . ', ' . $db->quoteName('phrase') . ', ' . $db->quoteName('weight') . ', ' . $db->quoteName('soundex') . ', ' . $db->quoteName('language') . ')' . ' SELECT ta.term, ta.stem, ta.common, ta.phrase, ta.term_weight, SOUNDEX(ta.term), ta.language' . ' FROM ' . $db->quoteName('#__finder_tokens_aggregate') . ' AS ta' . ' WHERE ta.term_id = 0' . ' GROUP BY ta.term, ta.stem, ta.common, ta.phrase, ta.term_weight, SOUNDEX(ta.term), ta.language' ); $db->execute(); / * Now, we just inserted a bunch of new records into the terms table * so we need to go back and update the aggregate table with all the * new term ids. / $query = $db->getQuery(true) ->update($db->quoteName('#__finder_tokens_aggregate') . ' AS ta') ->join('INNER', $db->quoteName('#__finder_terms') . ' AS t ON t.term = ta.term') ->set('term_id = t.term_id') ->where('ta.term_id = 0'); $db->setQuery($query); $db->execute(); // Mark afterTerms in the profiler. static::$profiler ? static::$profiler->mark('afterTerms') : null; / * After we've made sure that all of the terms are in the terms table * and the aggregate table has the correct term ids, we need to update * the links counter for each term by one. / $query->clear() ->update($db->quoteName('#__finder_terms') . ' AS t') ->join('INNER', $db->quoteName('#__finder_tokens_aggregate') . ' AS ta ON ta.term_id = t.term_id') ->set($db->quoteName('links') . ' = t.links + 1'); $db->setQuery($query); $db->execute(); // Mark afterTerms in the profiler. static::$profiler ? static::$profiler->mark('afterTerms') : null; / * Before we can insert all of the mapping rows, we have to figure out * which mapping table the rows need to be inserted into. The mapping * table for each term is based on the first character of the md5 of * the first character of the term. In php, it would be expressed as * substr(md5(substr($token, 0, 1)), 0, 1) / $query->clear() ->update($db->quoteName('#__finder_tokens_aggregate')) ->set($db->quoteName('map_suffix') . ' = SUBSTR(MD5(SUBSTR(' . $db->quoteName('term') . ', 1, 1)), 1, 1)'); $db->setQuery($query); $db->execute(); / * At this point, the aggregate table contains a record for each * term in each context. So, we're going to pull down all of that * data while grouping the records by term and add all of the * sub-totals together to arrive at the final total for each token for * this link. Then, we insert all of that data into the appropriate * mapping table. / for ($i = 0; $i <= 15; $i++) { // Get the mapping table suffix. $suffix = dechex($i); / * We have to run this query 16 times, one for each link => term * mapping table. / $db->setQuery( 'INSERT INTO ' . $db->quoteName('#__finder_links_terms' . $suffix) . ' (' . $db->quoteName('link_id') . ', ' . $db->quoteName('term_id') . ', ' . $db->quoteName('weight') . ')' . ' SELECT ' . (int) $linkId . ', ' . $db->quoteName('term_id') . ',' . ' ROUND(SUM(' . $db->quoteName('context_weight') . '), 8)' . ' FROM ' . $db->quoteName('#__finder_tokens_aggregate') . ' WHERE ' . $db->quoteName('map_suffix') . ' = ' . $db->quote($suffix) . ' GROUP BY ' . $db->quoteName('term') . ', ' . $db->quoteName('term_id') . ' ORDER BY ' . $db->quoteName('term') . ' DESC' ); $db->execute(); } // Mark afterMapping in the profiler. static::$profiler ? static::$profiler->mark('afterMapping') : null; // Update the signature. $object = serialize($item); $query->clear() ->update($db->quoteName('#__finder_links')) ->set($db->quoteName('md5sum') . ' = ' . $db->quote($curSig)) ->set($db->quoteName('object') . ' = ' . $db->quote(pg_escape_bytea($object))) ->where($db->quoteName('link_id') . ' = ' . $db->quote($linkId)); $db->setQuery($query); $db->execute(); // Mark afterSigning in the profiler. static::$profiler ? static::$profiler->mark('afterSigning') : null; // Truncate the tokens tables. $db->truncateTable('#__finder_tokens'); // Truncate the tokens aggregate table. $db->truncateTable('#__finder_tokens_aggregate'); // Toggle the token tables back to memory tables. $this->toggleTables(true); // Mark afterTruncating in the profiler. static::$profiler ? static::$profiler->mark('afterTruncating') : null; return $linkId; } /* * Method to optimize the index. We use this method to remove unused terms * and any other optimizations that might be necessary. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / public function optimize() { // Get the database object. $db = $this->db; $query = $db->getQuery(true); // Delete all orphaned terms. $query->delete($db->quoteName('#__finder_terms')) ->where($db->quoteName('links') . ' <= 0'); $db->setQuery($query); $db->execute(); // Optimize the links table. $db->setQuery('VACUUM ' . $db->quoteName('#__finder_links')); $db->execute(); $db->setQuery('REINDEX TABLE ' . $db->quoteName('#__finder_links')); $db->execute(); for ($i = 0; $i <= 15; $i++) { // Optimize the terms mapping table. $db->setQuery('VACUUM ' . $db->quoteName('#__finder_links_terms' . dechex($i))); $db->execute(); $db->setQuery('REINDEX TABLE ' . $db->quoteName('#__finder_links_terms' . dechex($i))); $db->execute(); } // Optimize the filters table. $db->setQuery('REINDEX TABLE ' . $db->quoteName('#__finder_filters')); $db->execute(); // Optimize the terms common table. $db->setQuery('REINDEX TABLE ' . $db->quoteName('#__finder_terms_common')); $db->execute(); // Optimize the types table. $db->setQuery('REINDEX TABLE ' . $db->quoteName('#__finder_types')); $db->execute(); // Remove the orphaned taxonomy nodes. FinderIndexerTaxonomy::removeOrphanNodes(); // Optimize the taxonomy mapping table. $db->setQuery('REINDEX TABLE ' . $db->quoteName('#__finder_taxonomy_map')); $db->execute(); // Optimize the taxonomy table. $db->setQuery('REINDEX TABLE ' . $db->quoteName('#__finder_taxonomy')); $db->execute(); return true; } } ��driver/sqlsrv.php��0000604��00000043472�15245703510�0010112 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2013 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; jimport('joomla.filesystem.file'); /* * Indexer class supporting SQL Server for the Finder indexer package. * * The indexer class provides the core functionality of the Finder * search engine. It is responsible for adding and updating the * content links table; extracting and scoring tokens; and maintaining * all referential information for the content. * * Note: All exceptions thrown from within this class should be caught * by the controller. * * @since 3.1 / class FinderIndexerDriverSqlsrv extends FinderIndexer { /* * Method to index a content item. * * @param FinderIndexerResult $item The content item to index. * @param string $format The format of the content. [optional] * * @return integer The ID of the record in the links table. * * @since 3.1 * @throws Exception on database error. / public function index($item, $format = 'html') { // Mark beforeIndexing in the profiler. static::$profiler ? static::$profiler->mark('beforeIndexing') : null; $db = $this->db; $nd = $db->getNullDate(); // Check if the item is in the database. $query = $db->getQuery(true) ->select($db->quoteName('link_id') . ', ' . $db->quoteName('md5sum')) ->from($db->quoteName('#__finder_links')) ->where($db->quoteName('url') . ' = ' . $db->quote($item->url)); // Load the item from the database. $db->setQuery($query); $link = $db->loadObject(); // Get the indexer state. $state = static::getState(); // Get the signatures of the item. $curSig = static::getSignature($item); $oldSig = isset($link->md5sum) ? $link->md5sum : null; // Get the other item information. $linkId = empty($link->link_id) ? null : $link->link_id; $isNew = empty($link->link_id) ? true : false; // Check the signatures. If they match, the item is up to date. if (!$isNew && $curSig === $oldSig) { return $linkId; } / * If the link already exists, flush all the term maps for the item. * Maps are stored in 16 tables so we need to iterate through and flush * each table one at a time. / if (!$isNew) { for ($i = 0; $i <= 15; $i++) { // Flush the maps for the link. $query->clear() ->delete($db->quoteName('#__finder_links_terms' . dechex($i))) ->where($db->quoteName('link_id') . ' = ' . (int) $linkId); $db->setQuery($query); $db->execute(); } // Remove the taxonomy maps. FinderIndexerTaxonomy::removeMaps($linkId); } // Mark afterUnmapping in the profiler. static::$profiler ? static::$profiler->mark('afterUnmapping') : null; // Perform cleanup on the item data. $item->publish_start_date = (int) $item->publish_start_date != 0 ? $item->publish_start_date : $nd; $item->publish_end_date = (int) $item->publish_end_date != 0 ? $item->publish_end_date : $nd; $item->start_date = (int) $item->start_date != 0 ? $item->start_date : $nd; $item->end_date = (int) $item->end_date != 0 ? $item->end_date : $nd; // Prepare the item description. $item->description = FinderIndexerHelper::parse($item->summary); / * Now, we need to enter the item into the links table. If the item * already exists in the database, we need to use an UPDATE query. * Otherwise, we need to use an INSERT to get the link id back. / if ($isNew) { $columnsArray = array( $db->quoteName('url'), $db->quoteName('route'), $db->quoteName('title'), $db->quoteName('description'), $db->quoteName('indexdate'), $db->quoteName('published'), $db->quoteName('state'), $db->quoteName('access'), $db->quoteName('language'), $db->quoteName('type_id'), $db->quoteName('object'), $db->quoteName('publish_start_date'), $db->quoteName('publish_end_date'), $db->quoteName('start_date'), $db->quoteName('end_date'), $db->quoteName('list_price'), $db->quoteName('sale_price') ); // Insert the link. $query->clear() ->insert($db->quoteName('#__finder_links')) ->columns($columnsArray) ->values( $db->quote($item->url) . ', ' . $db->quote($item->route) . ', ' . $db->quote($item->title) . ', ' . $db->quote($item->description) . ', ' . $query->currentTimestamp() . ', ' . '1, ' . (int) $item->state . ', ' . (int) $item->access . ', ' . $db->quote($item->language) . ', ' . (int) $item->type_id . ', ' . $db->quote(serialize($item)) . ', ' . $db->quote($item->publish_start_date) . ', ' . $db->quote($item->publish_end_date) . ', ' . $db->quote($item->start_date) . ', ' . $db->quote($item->end_date) . ', ' . (double) ($item->list_price ?: 0) . ', ' . (double) ($item->sale_price ?: 0) ); $db->setQuery($query); $db->execute(); // Get the link id. $linkId = (int) $db->insertid(); } else { // Update the link. $query->clear() ->update($db->quoteName('#__finder_links')) ->set($db->quoteName('route') . ' = ' . $db->quote($item->route)) ->set($db->quoteName('title') . ' = ' . $db->quote($item->title)) ->set($db->quoteName('description') . ' = ' . $db->quote($item->description)) ->set($db->quoteName('indexdate') . ' = ' . $query->currentTimestamp()) ->set($db->quoteName('state') . ' = ' . (int) $item->state) ->set($db->quoteName('access') . ' = ' . (int) $item->access) ->set($db->quoteName('language') . ' = ' . $db->quote($item->language)) ->set($db->quoteName('type_id') . ' = ' . (int) $item->type_id) ->set($db->quoteName('object') . ' = ' . $db->quote(serialize($item))) ->set($db->quoteName('publish_start_date') . ' = ' . $db->quote($item->publish_start_date)) ->set($db->quoteName('publish_end_date') . ' = ' . $db->quote($item->publish_end_date)) ->set($db->quoteName('start_date') . ' = ' . $db->quote($item->start_date)) ->set($db->quoteName('end_date') . ' = ' . $db->quote($item->end_date)) ->set($db->quoteName('list_price') . ' = ' . (double) ($item->list_price ?: 0)) ->set($db->quoteName('sale_price') . ' = ' . (double) ($item->sale_price ?: 0)) ->where('link_id = ' . (int) $linkId); $db->setQuery($query); $db->execute(); } // Set up the variables we will need during processing. $count = 0; // Mark afterLinking in the profiler. static::$profiler ? static::$profiler->mark('afterLinking') : null; // Truncate the tokens tables. $db->truncateTable('#__finder_tokens'); // Truncate the tokens aggregate table. $db->truncateTable('#__finder_tokens_aggregate'); / * Process the item's content. The items can customize their * processing instructions to define extra properties to process * or rearrange how properties are weighted. / foreach ($item->getInstructions() as $group => $properties) { // Iterate through the properties of the group. foreach ($properties as $property) { // Check if the property exists in the item. if (empty($item->$property)) { continue; } // Tokenize the property. if (is_array($item->$property)) { // Tokenize an array of content and add it to the database. foreach ($item->$property as $ip) { / * If the group is path, we need to a few extra processing * steps to strip the extension and convert slashes and dashes * to spaces. / if ($group === static::PATH_CONTEXT) { $ip = JFile::stripExt($ip); $ip = str_replace(array('/', '-'), ' ', $ip); } // Tokenize a string of content and add it to the database. $count += $this->tokenizeToDb($ip, $group, $item->language, $format); // Check if we're approaching the memory limit of the token table. if ($count > static::$state->options->get('memory_table_limit', 30000)) { $this->toggleTables(false); } } } else { / * If the group is path, we need to a few extra processing * steps to strip the extension and convert slashes and dashes * to spaces. / if ($group === static::PATH_CONTEXT) { $item->$property = JFile::stripExt($item->$property); $item->$property = str_replace('/', ' ', $item->$property); $item->$property = str_replace('-', ' ', $item->$property); } // Tokenize a string of content and add it to the database. $count += $this->tokenizeToDb($item->$property, $group, $item->language, $format); // Check if we're approaching the memory limit of the token table. if ($count > static::$state->options->get('memory_table_limit', 30000)) { $this->toggleTables(false); } } } } / * Process the item's taxonomy. The items can customize their * taxonomy mappings to define extra properties to map. / foreach ($item->getTaxonomy() as $branch => $nodes) { // Iterate through the nodes and map them to the branch. foreach ($nodes as $node) { // Add the node to the tree. $nodeId = FinderIndexerTaxonomy::addNode($branch, $node->title, $node->state, $node->access); // Add the link => node map. FinderIndexerTaxonomy::addMap($linkId, $nodeId); } } // Mark afterProcessing in the profiler. static::$profiler ? static::$profiler->mark('afterProcessing') : null; / * At this point, all of the item's content has been parsed, tokenized * and inserted into the #__finder_tokens table. Now, we need to * aggregate all the data into that table into a more usable form. The * aggregated data will be inserted into #__finder_tokens_aggregate * table. / $query = 'INSERT INTO ' . $db->quoteName('#__finder_tokens_aggregate') . ' (' . $db->quoteName('term_id') . ', ' . $db->quoteName('term') . ', ' . $db->quoteName('stem') . ', ' . $db->quoteName('common') . ', ' . $db->quoteName('phrase') . ', ' . $db->quoteName('term_weight') . ', ' . $db->quoteName('context') . ', ' . $db->quoteName('context_weight') . ', ' . $db->quoteName('language') . ')' . ' SELECT' . ' t.term_id, t1.term, t1.stem, t1.common, t1.phrase, t1.weight, t1.context,' . ' ROUND( t1.weight COUNT( t2.term ) * %F, 8 ) AS context_weight, t1.language' . ' FROM (' . ' SELECT DISTINCT t1.term, t1.stem, t1.common, t1.phrase, t1.weight, t1.context, t1.language' . ' FROM ' . $db->quoteName('#__finder_tokens') . ' AS t1' . ' WHERE t1.context = %d' . ' ) AS t1' . ' JOIN ' . $db->quoteName('#__finder_tokens') . ' AS t2 ON t2.term = t1.term' . ' LEFT JOIN ' . $db->quoteName('#__finder_terms') . ' AS t ON t.term = t1.term' . ' WHERE t2.context = %d' . ' GROUP BY t1.term, t.term_id, t1.term, t1.stem, t1.common, t1.phrase, t1.weight, t1.context, t1.language' . ' ORDER BY t1.term DESC'; // Iterate through the contexts and aggregate the tokens per context. foreach ($state->weights as $context => $multiplier) { // Run the query to aggregate the tokens for this context.. $db->setQuery(sprintf($query, $multiplier, $context, $context)); $db->execute(); } // Mark afterAggregating in the profiler. static::$profiler ? static::$profiler->mark('afterAggregating') : null; /* * When we pulled down all of the aggregate data, we did a LEFT JOIN * over the terms table to try to find all the term ids that * already exist for our tokens. If any of the rows in the aggregate * table have a term of 0, then no term record exists for that * term so we need to add it to the terms table. / $db->setQuery( 'INSERT INTO ' . $db->quoteName('#__finder_terms') . ' (' . $db->quoteName('term') . ', ' . $db->quoteName('stem') . ', ' . $db->quoteName('common') . ', ' . $db->quoteName('phrase') . ', ' . $db->quoteName('weight') . ', ' . $db->quoteName('soundex') . ')' . ' SELECT ta.term, ta.stem, ta.common, ta.phrase, ta.term_weight, SOUNDEX(ta.term)' . ' FROM ' . $db->quoteName('#__finder_tokens_aggregate') . ' AS ta' . ' WHERE ta.term_id IS NULL' . ' GROUP BY ta.term, ta.stem, ta.common, ta.phrase, ta.term_weight' ); $db->execute(); / * Now, we just inserted a bunch of new records into the terms table * so we need to go back and update the aggregate table with all the * new term ids. / $query = $db->getQuery(true) ->update('ta') ->set('ta.term_id = t.term_id from #__finder_tokens_aggregate AS ta INNER JOIN #__finder_terms AS t ON t.term = ta.term') ->where('ta.term_id IS NULL'); $db->setQuery($query); $db->execute(); // Mark afterTerms in the profiler. static::$profiler ? static::$profiler->mark('afterTerms') : null; / * After we've made sure that all of the terms are in the terms table * and the aggregate table has the correct term ids, we need to update * the links counter for each term by one. / $query->clear() ->update('t') ->set('t.links = t.links + 1 FROM #__finder_terms AS t INNER JOIN #__finder_tokens_aggregate AS ta ON ta.term_id = t.term_id'); $db->setQuery($query); $db->execute(); // Mark afterTerms in the profiler. static::$profiler ? static::$profiler->mark('afterTerms') : null; / * Before we can insert all of the mapping rows, we have to figure out * which mapping table the rows need to be inserted into. The mapping * table for each term is based on the first character of the md5 of * the first character of the term. In php, it would be expressed as * substr(md5(substr($token, 0, 1)), 0, 1) / $query->clear() ->update($db->quoteName('#__finder_tokens_aggregate')) ->set($db->quoteName('map_suffix') . " = SUBSTRING(HASHBYTES('MD5', SUBSTRING(" . $db->quoteName('term') . ', 1, 1)), 1, 1)'); $db->setQuery($query); $db->execute(); / * At this point, the aggregate table contains a record for each * term in each context. So, we're going to pull down all of that * data while grouping the records by term and add all of the * sub-totals together to arrive at the final total for each token for * this link. Then, we insert all of that data into the appropriate * mapping table. / for ($i = 0; $i <= 15; $i++) { // Get the mapping table suffix. $suffix = dechex($i); / * We have to run this query 16 times, one for each link => term * mapping table. / $db->setQuery( 'INSERT INTO ' . $db->quoteName('#__finder_links_terms' . $suffix) . ' (' . $db->quoteName('link_id') . ', ' . $db->quoteName('term_id') . ', ' . $db->quoteName('weight') . ')' . ' SELECT ' . (int) $linkId . ', ' . $db->quoteName('term_id') . ',' . ' ROUND(SUM(' . $db->quoteName('context_weight') . '), 8)' . ' FROM ' . $db->quoteName('#__finder_tokens_aggregate') . ' WHERE ' . $db->quoteName('map_suffix') . ' = ' . $db->quote($suffix) . ' GROUP BY term, term_id' . ' ORDER BY ' . $db->quoteName('term') . ' DESC' ); $db->execute(); } // Mark afterMapping in the profiler. static::$profiler ? static::$profiler->mark('afterMapping') : null; // Update the signature. $query->clear() ->update($db->quoteName('#__finder_links')) ->set($db->quoteName('md5sum') . ' = ' . $db->quote($curSig)) ->where($db->quoteName('link_id') . ' = ' . $db->quote($linkId)); $db->setQuery($query); $db->execute(); // Mark afterSigning in the profiler. static::$profiler ? static::$profiler->mark('afterSigning') : null; // Truncate the tokens tables. $db->truncateTable('#__finder_tokens'); // Truncate the tokens aggregate table. $db->truncateTable('#__finder_tokens_aggregate'); // Toggle the token tables back to memory tables. $this->toggleTables(true); // Mark afterTruncating in the profiler. static::$profiler ? static::$profiler->mark('afterTruncating') : null; return $linkId; } /* * Method to remove a link from the index. * * @param integer $linkId The id of the link. * * @return boolean True on success. * * @since 3.1 * @throws Exception on database error. / public function remove($linkId) { $db = $this->db; $query = $db->getQuery(true); // Update the link counts and remove the mapping records. for ($i = 0; $i <= 15; $i++) { // Update the link counts for the terms. $query->update('t') ->set('t.links = t.links - 1 from #__finder_terms AS t INNER JOIN #__finder_links_terms' . dechex($i) . ' AS m ON m.term_id = t.term_id') ->where('m.link_id = ' . $db->quote((int) $linkId)); $db->setQuery($query); $db->execute(); // Remove all records from the mapping tables. $query->clear() ->delete($db->quoteName('#__finder_links_terms' . dechex($i))) ->where($db->quoteName('link_id') . ' = ' . (int) $linkId); $db->setQuery($query); $db->execute(); } // Delete all orphaned terms. $query->clear() ->delete($db->quoteName('#__finder_terms')) ->where($db->quoteName('links') . ' <= 0'); $db->setQuery($query); $db->execute(); // Delete the link from the index. $query->clear() ->delete($db->quoteName('#__finder_links')) ->where($db->quoteName('link_id') . ' = ' . $db->quote((int) $linkId)); $db->setQuery($query); $db->execute(); // Remove the taxonomy maps. FinderIndexerTaxonomy::removeMaps($linkId); // Remove the orphaned taxonomy nodes. FinderIndexerTaxonomy::removeOrphanNodes(); return true; } /* * Method to optimize the index. We use this method to remove unused terms * and any other optimizations that might be necessary. * * @return boolean True on success. * * @since 3.1 * @throws Exception on database error. / public function optimize() { // Get the database object. $db = $this->db; $query = $db->getQuery(true); // Delete all orphaned terms. $query->delete($db->quoteName('#__finder_terms')) ->where($db->quoteName('links') . ' <= 0'); $db->setQuery($query); $db->execute(); // Remove the orphaned taxonomy nodes. FinderIndexerTaxonomy::removeOrphanNodes(); return true; } } ��stemmer/porter_en.php��0000604��00000023413�15245703510�0010727 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; JLoader::register('FinderIndexerStemmer', dirname(__DIR__) . '/stemmer.php'); /* * Porter English stemmer class for the Finder indexer package. * * This class was adapted from one written by Richard Heyes. * See copyright and link information above. * * @since 2.5 / class FinderIndexerStemmerPorter_En extends FinderIndexerStemmer { /* * Regex for matching a consonant. * * @var string * @since 2.5 / private static $regex_consonant = '(?:[bcdfghjklmnpqrstvwxz]\|(?<=[aeiou])y\|^y)'; /* * Regex for matching a vowel * * @var string * @since 2.5 / private static $regex_vowel = '(?:[aeiou]\|(?<![aeiou])y)'; /* * Method to stem a token and return the root. * * @param string $token The token to stem. * @param string $lang The language of the token. * * @return string The root token. * * @since 2.5 / public function stem($token, $lang) { // Check if the token is long enough to merit stemming. if (strlen($token) <= 2) { return $token; } // Check if the language is English or All. if ($lang !== 'en' && $lang !== '') { return $token; } // Stem the token if it is not in the cache. if (!isset($this->cache[$lang][$token])) { // Stem the token. $result = $token; $result = self::step1ab($result); $result = self::step1c($result); $result = self::step2($result); $result = self::step3($result); $result = self::step4($result); $result = self::step5($result); // Add the token to the cache. $this->cache[$lang][$token] = $result; } return $this->cache[$lang][$token]; } /** * Step 1 * * @param string $word The token to stem. * * @return string * * @since 2.5 / private static function step1ab($word) { // Part a if (substr($word, -1) === 's') { self::replace($word, 'sses', 'ss') \|\| self::replace($word, 'ies', 'i') \|\| self::replace($word, 'ss', 'ss') \|\| self::replace($word, 's', ''); } // Part b if (substr($word, -2, 1) !== 'e' \|\| !self::replace($word, 'eed', 'ee', 0)) { // First rule $v = self::$regex_vowel; // Words ending with ing and ed // Note use of && and OR, for precedence reasons if (preg_match("#$v+#", substr($word, 0, -3)) && self::replace($word, 'ing', '') \|\| preg_match("#$v+#", substr($word, 0, -2)) && self::replace($word, 'ed', '')) { // If one of above two test successful if (!self::replace($word, 'at', 'ate') && !self::replace($word, 'bl', 'ble') && !self::replace($word, 'iz', 'ize')) { // Double consonant ending $wordSubStr = substr($word, -2); if ($wordSubStr !== 'll' && $wordSubStr !== 'ss' && $wordSubStr !== 'zz' && self::doubleConsonant($word)) { $word = substr($word, 0, -1); } elseif (self::m($word) === 1 && self::cvc($word)) { $word .= 'e'; } } } } return $word; } /* * Step 1c * * @param string $word The token to stem. * * @return string * * @since 2.5 / private static function step1c($word) { $v = self::$regex_vowel; if (substr($word, -1) === 'y' && preg_match("#$v+#", substr($word, 0, -1))) { self::replace($word, 'y', 'i'); } return $word; } /* * Step 2 * * @param string $word The token to stem. * * @return string * * @since 2.5 / private static function step2($word) { switch (substr($word, -2, 1)) { case 'a': self::replace($word, 'ational', 'ate', 0) \|\| self::replace($word, 'tional', 'tion', 0); break; case 'c': self::replace($word, 'enci', 'ence', 0) \|\| self::replace($word, 'anci', 'ance', 0); break; case 'e': self::replace($word, 'izer', 'ize', 0); break; case 'g': self::replace($word, 'logi', 'log', 0); break; case 'l': self::replace($word, 'entli', 'ent', 0) \|\| self::replace($word, 'ousli', 'ous', 0) \|\| self::replace($word, 'alli', 'al', 0) \|\| self::replace($word, 'bli', 'ble', 0) \|\| self::replace($word, 'eli', 'e', 0); break; case 'o': self::replace($word, 'ization', 'ize', 0) \|\| self::replace($word, 'ation', 'ate', 0) \|\| self::replace($word, 'ator', 'ate', 0); break; case 's': self::replace($word, 'iveness', 'ive', 0) \|\| self::replace($word, 'fulness', 'ful', 0) \|\| self::replace($word, 'ousness', 'ous', 0) \|\| self::replace($word, 'alism', 'al', 0); break; case 't': self::replace($word, 'biliti', 'ble', 0) \|\| self::replace($word, 'aliti', 'al', 0) \|\| self::replace($word, 'iviti', 'ive', 0); break; } return $word; } /* * Step 3 * * @param string $word The token to stem. * * @return string * * @since 2.5 / private static function step3($word) { switch (substr($word, -2, 1)) { case 'a': self::replace($word, 'ical', 'ic', 0); break; case 's': self::replace($word, 'ness', '', 0); break; case 't': self::replace($word, 'icate', 'ic', 0) \|\| self::replace($word, 'iciti', 'ic', 0); break; case 'u': self::replace($word, 'ful', '', 0); break; case 'v': self::replace($word, 'ative', '', 0); break; case 'z': self::replace($word, 'alize', 'al', 0); break; } return $word; } /* * Step 4 * * @param string $word The token to stem. * * @return string * * @since 2.5 / private static function step4($word) { switch (substr($word, -2, 1)) { case 'a': self::replace($word, 'al', '', 1); break; case 'c': self::replace($word, 'ance', '', 1) \|\| self::replace($word, 'ence', '', 1); break; case 'e': self::replace($word, 'er', '', 1); break; case 'i': self::replace($word, 'ic', '', 1); break; case 'l': self::replace($word, 'able', '', 1) \|\| self::replace($word, 'ible', '', 1); break; case 'n': self::replace($word, 'ant', '', 1) \|\| self::replace($word, 'ement', '', 1) \|\| self::replace($word, 'ment', '', 1) \|\| self::replace($word, 'ent', '', 1); break; case 'o': $wordSubStr = substr($word, -4); if ($wordSubStr === 'tion' \|\| $wordSubStr === 'sion') { self::replace($word, 'ion', '', 1); } else { self::replace($word, 'ou', '', 1); } break; case 's': self::replace($word, 'ism', '', 1); break; case 't': self::replace($word, 'ate', '', 1) \|\| self::replace($word, 'iti', '', 1); break; case 'u': self::replace($word, 'ous', '', 1); break; case 'v': self::replace($word, 'ive', '', 1); break; case 'z': self::replace($word, 'ize', '', 1); break; } return $word; } /* * Step 5 * * @param string $word The token to stem. * * @return string * * @since 2.5 / private static function step5($word) { // Part a if (substr($word, -1) === 'e') { if (self::m(substr($word, 0, -1)) > 1) { self::replace($word, 'e', ''); } elseif (self::m(substr($word, 0, -1)) === 1) { if (!self::cvc(substr($word, 0, -1))) { self::replace($word, 'e', ''); } } } // Part b if (self::m($word) > 1 && self::doubleConsonant($word) && substr($word, -1) === 'l') { $word = substr($word, 0, -1); } return $word; } /* * Replaces the first string with the second, at the end of the string. If third * arg is given, then the preceding string must match that m count at least. * * @param string $str String to check * @param string $check Ending to check for * @param string $repl Replacement string * @param integer $m Optional minimum number of m() to meet * * @return boolean Whether the $check string was at the end * of the $str string. True does not necessarily mean * that it was replaced. * * @since 2.5 / private static function replace(&$str, $check, $repl, $m = null) { $len = 0 - strlen($check); if (substr($str, $len) === $check) { $substr = substr($str, 0, $len); if ($m === null \|\| self::m($substr) > $m) { $str = $substr . $repl; } return true; } return false; } /* * m() measures the number of consonant sequences in $str. if c is * a consonant sequence and v a vowel sequence, and <..> indicates arbitrary * presence, * * <c><v> gives 0 * <c>vc<v> gives 1 * <c>vcvc<v> gives 2 * <c>vcvcvc<v> gives 3 * * @param string $str The string to return the m count for * * @return integer The m count * * @since 2.5 / private static function m($str) { $c = self::$regex_consonant; $v = self::$regex_vowel; $str = preg_replace("#^$c+#", '', $str); $str = preg_replace("#$v+$#", '', $str); preg_match_all("#($v+$c+)#", $str, $matches); return count($matches[1]); } /* * Returns true/false as to whether the given string contains two * of the same consonant next to each other at the end of the string. * * @param string $str String to check * * @return boolean Result * * @since 2.5 / private static function doubleConsonant($str) { $c = self::$regex_consonant; return preg_match("#$c{2}$#", $str, $matches) && $matches[0][0] === $matches[0][1]; } /* * Checks for ending CVC sequence where second C is not W, X or Y * * @param string $str String to check * * @return boolean Result * * @since 2.5 / private static function cvc($str) { $c = self::$regex_consonant; $v = self::$regex_vowel; return preg_match("#($c$v$c)$#", $str, $matches) && strlen($matches[1]) === 3 && $matches[1][2] !== 'w' && $matches[1][2] !== 'x' && $matches[1][2] !== 'y'; } } ��stemmer/fr.php��0000604��00000024133�15245703510�0007341 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2012 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; JLoader::register('FinderIndexerStemmer', dirname(__DIR__) . '/stemmer.php'); /* * French stemmer class for Smart Search indexer. * * First contributed by Eric Sanou (bobotche@hotmail.fr) * This class is inspired in Alexis Ulrich's French stemmer code (http://alx2002.free.fr) * * @since 3.0 / class FinderIndexerStemmerFr extends FinderIndexerStemmer { /* * Stemming rules. * * @var array * @since 3.0 / private static $stemRules; /* * Method to stem a token and return the root. * * @param string $token The token to stem. * @param string $lang The language of the token. * * @return string The root token. * * @since 3.0 / public function stem($token, $lang) { // Check if the token is long enough to merit stemming. if (strlen($token) <= 2) { return $token; } // Check if the language is French or All. if ($lang !== 'fr' && $lang !== '') { return $token; } // Stem the token if it is not in the cache. if (!isset($this->cache[$lang][$token])) { // Stem the token. $result = self::getStem($token); // Add the token to the cache. $this->cache[$lang][$token] = $result; } return $this->cache[$lang][$token]; } /** * French stemmer rules variables. * * @return array The rules * * @since 3.0 / protected static function getStemRules() { if (self::$stemRules) { return self::$stemRules; } $vars = array(); // French accented letters in ISO-8859-1 encoding $vars['accents'] = chr(224) . chr(226) . chr(232) . chr(233) . chr(234) . chr(235) . chr(238) . chr(239) . chr(244) . chr(251) . chr(249) . chr(231); // The rule patterns include all accented words for french language $vars['rule_pattern'] = '/^([a-z' . $vars['accents'] . '])(\){0,1}(\d)([a-z' . $vars['accents'] . '])([.\|>])/'; // French vowels (including y) in ISO-8859-1 encoding $vars['vowels'] = chr(97) . chr(224) . chr(226) . chr(101) . chr(232) . chr(233) . chr(234) . chr(235) . chr(105) . chr(238) . chr(239) . chr(111) . chr(244) . chr(117) . chr(251) . chr(249) . chr(121); // The French rules in ISO-8859-1 encoding $vars['rules'] = array( 'esre1>', 'esio1>', 'siol1.', 'siof0.', 'sioe0.', 'sio3>', 'st1>', 'sf1>', 'sle1>', 'slo1>', 's' . chr(233) . '1>', chr(233) . 'tuae5.', chr(233) . 'tuae2.', 'tnia0.', 'tniv1.', 'tni3>', 'suor1.', 'suo0.', 'sdrail5.', 'sdrai4.', 'er' . chr(232) . 'i1>', 'sesue3x>', 'esuey5i.', 'esue2x>', 'se1>', 'er' . chr(232) . 'g3.', 'eca1>', 'esiah0.', 'esi1>', 'siss2.', 'sir2>', 'sit2>', 'egan' . chr(233) . '1.', 'egalli6>', 'egass1.', 'egas0.', 'egat3.', 'ega3>', 'ette4>', 'ett2>', 'etio1.', 'tio' . chr(231) . '4c.', 'tio0.', 'et1>', 'eb1>', 'snia1>', 'eniatnau8>', 'eniatn4.', 'enia1>', 'niatnio3.', 'niatg3.', 'e' . chr(233) . '1>', chr(233) . 'hcat1.', chr(233) . 'hca4.', chr(233) . 'tila5>', chr(233) . 'tici5.', chr(233) . 'tir1.', chr(233) . 'ti3>', chr(233) . 'gan1.', chr(233) . 'ga3>', chr(233) . 'tehc1.', chr(233) . 'te3>', chr(233) . 'it0.', chr(233) . '1>', 'eire4.', 'eirue5.', 'eio1.', 'eia1.', 'ei1>', 'eng1.', 'xuaessi7.', 'xuae1>', 'uaes0.', 'uae3.', 'xuave2l.', 'xuav2li>', 'xua3la>', 'ela1>', 'lart2.', 'lani2>', 'la' . chr(233) . '2>', 'siay4i.', 'siassia7.', 'siarv1.', 'sia1>', 'tneiayo6i.', 'tneiay6i.', 'tneiassia9.', 'tneiareio7.', 'tneia5>', 'tneia4>', 'tiario4.', 'tiarim3.', 'tiaria3.', 'tiaris3.', 'tiari5.', 'tiarve6>', 'tiare5>', 'iare4>', 'are3>', 'tiay4i.', 'tia3>', 'tnay4i.', 'em' . chr(232) . 'iu5>', 'em' . chr(232) . 'i4>', 'tnaun3.', 'tnauqo3.', 'tnau4>', 'tnaf0.', 'tnat' . chr(233) . '2>', 'tna3>', 'tno3>', 'zeiy4i.', 'zey3i.', 'zeire5>', 'zeird4.', 'zeirio4.', 'ze2>', 'ssiab0.', 'ssia4.', 'ssi3.', 'tnemma6>', 'tnemesuey9i.', 'tnemesue8>', 'tnemevi7.', 'tnemessia5.', 'tnemessi8.', 'tneme5>', 'tnemia4.', 'tnem' . chr(233) . '5>', 'el2l>', 'lle3le>', 'let' . chr(244) . '0.', 'lepp0.', 'le2>', 'srei1>', 'reit3.', 'reila2.', 'rei3>', 'ert' . chr(226) . 'e5.', 'ert' . chr(226) . chr(233) . '1.', 'ert' . chr(226) . '4.', 'drai4.', 'erdro0.', 'erute5.', 'ruta0.', 'eruta1.', 'erutiov1.', 'erub3.', 'eruh3.', 'erul3.', 'er2r>', 'nn1>', 'r' . chr(232) . 'i3.', 'srev0.', 'sr1>', 'rid2>', 're2>', 'xuei4.', 'esuei5.', 'lbati3.', 'lba3>', 'rueis0.', 'ruehcn4.', 'ecirta6.', 'ruetai6.', 'rueta5.', 'rueir0.', 'rue3>', 'esseti6.', 'essere6>', 'esserd1.', 'esse4>', 'essiab1.', 'essia5.', 'essio1.', 'essi4.', 'essal4.', 'essa1>', 'ssab1.', 'essurp1.', 'essu4.', 'essi1.', 'ssor1.', 'essor2.', 'esso1>', 'ess2>', 'tio3.', 'r' . chr(232) . 's2re.', 'r' . chr(232) . '0e.', 'esn1.', 'eu1>', 'sua0.', 'su1>', 'utt1>', 'tu' . chr(231) . '3c.', 'u' . chr(231) . '2c.', 'ur1.', 'ehcn2>', 'ehcu1>', 'snorr3.', 'snoru3.', 'snorua3.', 'snorv3.', 'snorio4.', 'snori5.', 'snore5>', 'snortt4>', 'snort' . chr(238) . 'a7.', 'snort3.', 'snor4.', 'snossi6.', 'snoire6.', 'snoird5.', 'snoitai7.', 'snoita6.', 'snoits1>', 'noits0.', 'snoi4>', 'noitaci7>', 'noitai6.', 'noita5.', 'noitu4.', 'noi3>', 'snoya0.', 'snoy4i.', 'sno' . chr(231) . 'a1.', 'sno' . chr(231) . 'r1.', 'snoe4.', 'snosiar1>', 'snola1.', 'sno3>', 'sno1>', 'noll2.', 'tnennei4.', 'ennei2>', 'snei1>', 'sne' . chr(233) . '1>', 'enne' . chr(233) . '5e.', 'ne' . chr(233) . '3e.', 'neic0.', 'neiv0.', 'nei3.', 'sc1.', 'sd1.', 'sg1.', 'sni1.', 'tiu0.', 'ti2.', 'sp1>', 'sna1>', 'sue1.', 'enn2>', 'nong2.', 'noss2.', 'rioe4.', 'riot0.', 'riorc1.', 'riovec5.', 'rio3.', 'ric2.', 'ril2.', 'tnerim3.', 'tneris3>', 'tneri5.', 't' . chr(238) . 'a3.', 'riss2.', 't' . chr(238) . '2.', 't' . chr(226) . '2>', 'ario2.', 'arim1.', 'ara1.', 'aris1.', 'ari3.', 'art1>', 'ardn2.', 'arr1.', 'arua1.', 'aro1.', 'arv1.', 'aru1.', 'ar2.', 'rd1.', 'ud1.', 'ul1.', 'ini1.', 'rin2.', 'tnessiab3.', 'tnessia7.', 'tnessi6.', 'tnessni4.', 'sini2.', 'sl1.', 'iard3.', 'iario3.', 'ia2>', 'io0.', 'iule2.', 'i1>', 'sid2.', 'sic2.', 'esoi4.', 'ed1.', 'ai2>', 'a1>', 'adr1.', 'tner' . chr(232) . '5>', 'evir1.', 'evio4>', 'evi3.', 'fita4.', 'fi2>', 'enie1.', 'sare4>', 'sari4>', 'sard3.', 'sart2>', 'sa2.', 'tnessa6>', 'tnessu6>', 'tnegna3.', 'tnegi3.', 'tneg0.', 'tneru5>', 'tnemg0.', 'tnerni4.', 'tneiv1.', 'tne3>', 'une1.', 'en1>', 'nitn2.', 'ecnay5i.', 'ecnal1.', 'ecna4.', 'ec1>', 'nn1.', 'rit2>', 'rut2>', 'rud2.', 'ugn1>', 'eg1>', 'tuo0.', 'tul2>', 't' . chr(251) . '2>', 'ev1>', 'v' . chr(232) . '2ve>', 'rtt1>', 'emissi6.', 'em1.', 'ehc1.', 'c' . chr(233) . 'i2c' . chr(232) . '.', 'libi2l.', 'llie1.', 'liei4i.', 'xuev1.', 'xuey4i.', 'xueni5>', 'xuell4.', 'xuere5.', 'xue3>', 'rb' . chr(233) . '3rb' . chr(232) . '.', 'tur2.', 'rir' . chr(233) . '4re.', 'rir2.', 'c' . chr(226) . '2ca.', 'snu1.', 'rt' . chr(238) . 'a4.', 'long2.', 'vec2.', chr(231) . '1c>', 'ssilp3.', 'silp2.', 't' . chr(232) . 'hc2te.', 'n' . chr(232) . 'm2ne.', 'llepp1.', 'tan2.', 'rv' . chr(232) . '3rve.', 'rv' . chr(233) . '3rve.', 'r' . chr(232) . '2re.', 'r' . chr(233) . '2re.', 't' . chr(232) . '2te.', 't' . chr(233) . '2te.', 'epp1.', 'eya2i.', 'ya1i.', 'yo1i.', 'esu1.', 'ugi1.', 'tt1.', 'end0.' ); self::$stemRules = $vars; return self::$stemRules; } /* * Returns the number of the first rule from the rule number * that can be applied to the given reversed input. * returns -1 if no rule can be applied, ie the stem has been found * * @param string $reversedInput The input to check in reversed order * @param integer $ruleNumber The rule number to check * * @return integer Number of the first rule * * @since 3.0 / private static function getFirstRule($reversedInput, $ruleNumber) { $vars = static::getStemRules(); $nb_rules = count($vars['rules']); for ($i = $ruleNumber; $i < $nb_rules; $i++) { // Gets the letters from the current rule $rule = $vars['rules'][$i]; $rule = preg_replace($vars['rule_pattern'], "\\1", $rule); if (strncasecmp(utf8_decode($rule), $reversedInput, strlen(utf8_decode($rule))) == 0) { return $i; } } return -1; } /* * Check the acceptability of a stem for French language * * @param string $reversedStem The stem to check in reverse form * * @return boolean True if stem is acceptable * * @since 3.0 / private static function check($reversedStem) { $vars = static::getStemRules(); if (preg_match('/[' . $vars['vowels'] . ']$/', utf8_encode($reversedStem))) { // If the form starts with a vowel then at least two letters must remain after stemming (e.g.: "etaient" --> "et") return (strlen($reversedStem) > 2); } else { // If the reversed stem starts with a consonant then at least two letters must remain after stemming if (strlen($reversedStem) <= 2) { return false; } // And at least one of these must be a vowel or "y" return preg_match('/[' . $vars['vowels'] . ']/', utf8_encode($reversedStem)); } } /* * Paice/Husk stemmer which returns a stem for the given $input * * @param string $input The word for which we want the stem in UTF-8 * * @return string The stem * * @since 3.0 / private static function getStem($input) { $vars = static::getStemRules(); $reversed_input = strrev(utf8_decode($input)); $rule_number = 0; // This loop goes through the rules' array until it finds an ending one (ending by '.') or the last one ('end0.') while (true) { $rule_number = self::getFirstRule($reversed_input, $rule_number); if ($rule_number === -1) { // No other rule can be applied => the stem has been found break; } $rule = $vars['rules'][$rule_number]; preg_match($vars['rule_pattern'], $rule, $matches); $reversed_stem = utf8_decode($matches[4]) . substr($reversed_input, $matches[3]); if (self::check($reversed_stem)) { $reversed_input = $reversed_stem; if ($matches[5] === '.') { break; } } else { // Go to another rule $rule_number++; } } return utf8_encode(strrev($reversed_input)); } } ��stemmer/snowball.php��0000604��00000005320�15245703510�0010550 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; JLoader::register('FinderIndexerStemmer', dirname(__DIR__) . '/stemmer.php'); /* * Snowball stemmer class for the Finder indexer package. * * @since 2.5 / class FinderIndexerStemmerSnowball extends FinderIndexerStemmer { /* * Method to stem a token and return the root. * * @param string $token The token to stem. * @param string $lang The language of the token. * * @return string The root token. * * @since 2.5 / public function stem($token, $lang) { // Language to use if All is specified. static $defaultLang = ''; // If language is All then try to get site default language. if ($lang === '' && $defaultLang === '') { $languages = JLanguageHelper::getLanguages(); $defaultLang = isset($languages[0]->sef) ? $languages[0]->sef : ''; $lang = $defaultLang; } // Stem the token if it is not in the cache. if (!isset($this->cache[$lang][$token])) { // Get the stem function from the language string. switch ($lang) { // Danish stemmer. case 'da': $function = 'stem_danish'; break; // German stemmer. case 'de': $function = 'stem_german'; break; // English stemmer. default: case 'en': $function = 'stem_english'; break; // Spanish stemmer. case 'es': $function = 'stem_spanish'; break; // Finnish stemmer. case 'fi': $function = 'stem_finnish'; break; // French stemmer. case 'fr': $function = 'stem_french'; break; // Hungarian stemmer. case 'hu': $function = 'stem_hungarian'; break; // Italian stemmer. case 'it': $function = 'stem_italian'; break; // Norwegian stemmer. case 'nb': $function = 'stem_norwegian'; break; // Dutch stemmer. case 'nl': $function = 'stem_dutch'; break; // Portuguese stemmer. case 'pt': $function = 'stem_portuguese'; break; // Romanian stemmer. case 'ro': $function = 'stem_romanian'; break; // Russian stemmer. case 'ru': $function = 'stem_russian_unicode'; break; // Swedish stemmer. case 'sv': $function = 'stem_swedish'; break; // Turkish stemmer. case 'tr': $function = 'stem_turkish_unicode'; break; } // Stem the word if the stemmer method exists. $this->cache[$lang][$token] = function_exists($function) ? $function($token) : $token; } return $this->cache[$lang][$token]; } } ��result.php��0000604��00000021021�15245703510�0006565 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; JLoader::register('FinderIndexer', __DIR__ . '/indexer.php'); /* * Result class for the Finder indexer package. * * This class uses magic __get() and __set() methods to prevent properties * being added that might confuse the system. All properties not explicitly * declared will be pushed into the elements array and can be accessed * explicitly using the getElement() method. * * @since 2.5 / class FinderIndexerResult { /* * An array of extra result properties. * * @var array * @since 2.5 / protected $elements = array(); /* * This array tells the indexer which properties should be indexed and what * weights to use for those properties. * * @var array * @since 2.5 / protected $instructions = array( FinderIndexer::TITLE_CONTEXT => array('title', 'subtitle', 'id'), FinderIndexer::TEXT_CONTEXT => array('summary', 'body'), FinderIndexer::META_CONTEXT => array('meta', 'list_price', 'sale_price'), FinderIndexer::PATH_CONTEXT => array('path', 'alias'), FinderIndexer::MISC_CONTEXT => array('comments'), ); /* * The indexer will use this data to create taxonomy mapping entries for * the item so that it can be filtered by type, label, category, * or whatever. * * @var array * @since 2.5 / protected $taxonomy = array(); /* * The content URL. * * @var string * @since 2.5 / public $url; /* * The content route. * * @var string * @since 2.5 / public $route; /* * The content title. * * @var string * @since 2.5 / public $title; /* * The content description. * * @var string * @since 2.5 / public $description; /* * The published state of the result. * * @var integer * @since 2.5 / public $published; /* * The content published state. * * @var integer * @since 2.5 / public $state; /* * The content access level. * * @var integer * @since 2.5 / public $access; /* * The content language. * * @var string * @since 2.5 / public $language = ''; /** * The publishing start date. * * @var string * @since 2.5 / public $publish_start_date; /* * The publishing end date. * * @var string * @since 2.5 / public $publish_end_date; /* * The generic start date. * * @var string * @since 2.5 / public $start_date; /* * The generic end date. * * @var string * @since 2.5 / public $end_date; /* * The item list price. * * @var mixed * @since 2.5 / public $list_price; /* * The item sale price. * * @var mixed * @since 2.5 / public $sale_price; /* * The content type id. This is set by the adapter. * * @var integer * @since 2.5 / public $type_id; /* * The default language for content. * * @var string * @since 3.0.2 / public $defaultLanguage; /* * Constructor * * @since 3.0.3 / public function __construct() { $this->defaultLanguage = JComponentHelper::getParams('com_languages')->get('site', 'en-GB'); } /* * The magic set method is used to push additional values into the elements * array in order to preserve the cleanliness of the object. * * @param string $name The name of the element. * @param mixed $value The value of the element. * * @return void * * @since 2.5 / public function __set($name, $value) { $this->setElement($name, $value); } /* * The magic get method is used to retrieve additional element values from the elements array. * * @param string $name The name of the element. * * @return mixed The value of the element if set, null otherwise. * * @since 2.5 / public function __get($name) { return $this->getElement($name); } /* * The magic isset method is used to check the state of additional element values in the elements array. * * @param string $name The name of the element. * * @return boolean True if set, false otherwise. * * @since 2.5 / public function __isset($name) { return isset($this->elements[$name]); } /* * The magic unset method is used to unset additional element values in the elements array. * * @param string $name The name of the element. * * @return void * * @since 2.5 / public function __unset($name) { unset($this->elements[$name]); } /* * Method to retrieve additional element values from the elements array. * * @param string $name The name of the element. * * @return mixed The value of the element if set, null otherwise. * * @since 2.5 / public function getElement($name) { // Get the element value if set. if (array_key_exists($name, $this->elements)) { return $this->elements[$name]; } return null; } /* * Method to retrieve all elements. * * @return array The elements * * @since 3.8.3 / public function getElements() { return $this->elements; } /* * Method to set additional element values in the elements array. * * @param string $name The name of the element. * @param mixed $value The value of the element. * * @return void * * @since 2.5 / public function setElement($name, $value) { $this->elements[$name] = $value; } /* * Method to get all processing instructions. * * @return array An array of processing instructions. * * @since 2.5 / public function getInstructions() { return $this->instructions; } /* * Method to add a processing instruction for an item property. * * @param string $group The group to associate the property with. * @param string $property The property to process. * * @return void * * @since 2.5 / public function addInstruction($group, $property) { // Check if the group exists. We can't add instructions for unknown groups. // Check if the property exists in the group. if (array_key_exists($group, $this->instructions) && !in_array($property, $this->instructions[$group], true)) { // Add the property to the group. $this->instructions[$group][] = $property; } } /* * Method to remove a processing instruction for an item property. * * @param string $group The group to associate the property with. * @param string $property The property to process. * * @return void * * @since 2.5 / public function removeInstruction($group, $property) { // Check if the group exists. We can't remove instructions for unknown groups. if (array_key_exists($group, $this->instructions)) { // Search for the property in the group. $key = array_search($property, $this->instructions[$group]); // If the property was found, remove it. if ($key !== false) { unset($this->instructions[$group][$key]); } } } /* * Method to get the taxonomy maps for an item. * * @param string $branch The taxonomy branch to get. [optional] * * @return array An array of taxonomy maps. * * @since 2.5 / public function getTaxonomy($branch = null) { // Get the taxonomy branch if available. if ($branch !== null && isset($this->taxonomy[$branch])) { // Filter the input. $branch = preg_replace('#[^\pL\pM\pN\p{Pi}\p{Pf}\'+-.,_]+#mui', ' ', $branch); return $this->taxonomy[$branch]; } return $this->taxonomy; } /* * Method to add a taxonomy map for an item. * * @param string $branch The title of the taxonomy branch to add the node to. * @param string $title The title of the taxonomy node. * @param integer $state The published state of the taxonomy node. [optional] * @param integer $access The access level of the taxonomy node. [optional] * * @return void * * @since 2.5 / public function addTaxonomy($branch, $title, $state = 1, $access = 1) { // Filter the input. $branch = preg_replace('#[^\pL\pM\pN\p{Pi}\p{Pf}\'+-.,_]+#mui', ' ', $branch); // Create the taxonomy node. $node = new JObject; $node->title = $title; $node->state = (int) $state; $node->access = (int) $access; // Add the node to the taxonomy branch. $this->taxonomy[$branch][$node->title] = $node; } /* * Method to set the item language * * @return void * * @since 3.0 / public function setLanguage() { if ($this->language == '') { $this->language = $this->defaultLanguage; } } } ��token.php��0000604��00000007657�15245703510�0006412 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; use Joomla\String\StringHelper; /* * Token class for the Finder indexer package. * * @since 2.5 / class FinderIndexerToken { /* * This is the term that will be referenced in the terms table and the * mapping tables. * * @var string * @since 2.5 / public $term; /* * The stem is used to match the root term and produce more potential * matches when searching the index. * * @var string * @since 2.5 / public $stem; /* * If the token is numeric, it is likely to be short and uncommon so the * weight is adjusted to compensate for that situation. * * @var boolean * @since 2.5 / public $numeric; /* * If the token is a common term, the weight is adjusted to compensate for * the higher frequency of the term in relation to other terms. * * @var boolean * @since 2.5 / public $common; /* * Flag for phrase tokens. * * @var boolean * @since 2.5 / public $phrase; /* * The length is used to calculate the weight of the token. * * @var integer * @since 2.5 / public $length; /* * The weight is calculated based on token size and whether the token is * considered a common term. * * @var integer * @since 2.5 / public $weight; /* * The simple language identifier for the token. * * @var string * @since 2.5 / public $language; /* * The container for matches. * * @var array * @since 3.8.12 / public $matches = array(); /* * Is derived token (from individual words) * * @var boolean * @since 3.8.12 / public $derived; /* * The suggested term * * @var string * @since 3.8.12 / public $suggestion; /* * Method to construct the token object. * * @param mixed $term The term as a string for words or an array for phrases. * @param string $lang The simple language identifier. * @param string $spacer The space separator for phrases. [optional] * * @since 2.5 / public function __construct($term, $lang, $spacer = ' ') { $this->language = $lang; // Tokens can be a single word or an array of words representing a phrase. if (is_array($term)) { // Populate the token instance. $this->term = implode($spacer, $term); $this->stem = implode($spacer, array_map(array('FinderIndexerHelper', 'stem'), $term, array($lang))); $this->numeric = false; $this->common = false; $this->phrase = true; $this->length = StringHelper::strlen($this->term); / * Calculate the weight of the token. * * 1. Length of the token up to 30 and divide by 30, add 1. * 2. Round weight to 4 decimal points. / $this->weight = (($this->length >= 30 ? 30 : $this->length) / 30) + 1; $this->weight = round($this->weight, 4); } else { // Populate the token instance. $this->term = $term; $this->stem = FinderIndexerHelper::stem($this->term, $lang); $this->numeric = (is_numeric($this->term) \|\| (bool) preg_match('#^[0-9,.\-\+]+$#', $this->term)); $this->common = $this->numeric ? false : FinderIndexerHelper::isCommon($this->term, $lang); $this->phrase = false; $this->length = StringHelper::strlen($this->term); / * Calculate the weight of the token. * * 1. Length of the token up to 15 and divide by 15. * 2. If common term, divide weight by 8. * 3. If numeric, multiply weight by 1.5. * 4. Round weight to 4 decimal points. / $this->weight = ($this->length >= 15 ? 15 : $this->length) / 15; $this->weight = $this->common === true ? $this->weight / 8 : $this->weight; $this->weight = $this->numeric === true ? $this->weight 1.5 : $this->weight; $this->weight = round($this->weight, 4); } } } ��parser/txt.php��0000604��00000001336�15245703510�0007371 0��ustar�00��<?php /** * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; JLoader::register('FinderIndexerParser', dirname(__DIR__) . '/parser.php'); /* * Text Parser class for the Finder indexer package. * * @since 2.5 / class FinderIndexerParserTxt extends FinderIndexerParser { /* * Method to process Text input and extract the plain text. * * @param string $input The input to process. * * @return string The plain text input. * * @since 2.5 / protected function process($input) { return $input; } } ��parser/html.php��0000604��00000010556�15245703510�0007522 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; JLoader::register('FinderIndexerParser', dirname(__DIR__) . '/parser.php'); /* * HTML Parser class for the Finder indexer package. * * @since 2.5 / class FinderIndexerParserHtml extends FinderIndexerParser { /* * Method to parse input and extract the plain text. Because this method is * called from both inside and outside the indexer, it needs to be able to * batch out its parsing functionality to deal with the inefficiencies of * regular expressions. We will parse recursively in 2KB chunks. * * @param string $input The input to parse. * * @return string The plain text input. * * @since 2.5 / public function parse($input) { // Strip invalid UTF-8 characters. $input = iconv('utf-8', 'utf-8//IGNORE', $input); // Remove anything between <head> and </head> tags. Do this first // because there might be <script> or <style> tags nested inside. $input = $this->removeBlocks($input, '<head>', '</head>'); // Convert <style> and <noscript> tags to <script> tags // so we can remove them efficiently. $search = array( '<style', '</style', '<noscript', '</noscript', ); $replace = array( '<script', '</script', '<script', '</script', ); $input = str_replace($search, $replace, $input); // Strip all script blocks. $input = $this->removeBlocks($input, '<script', '</script>'); // Decode HTML entities. $input = html_entity_decode($input, ENT_QUOTES, 'UTF-8'); // Convert entities equivalent to spaces to actual spaces. $input = str_replace(array(' ', ' '), ' ', $input); // Add a space before both the OPEN and CLOSE tags of BLOCK and LINE BREAKING elements, // e.g. 'all<h1><em>m</em>obile List</h1>' will become 'all mobile List' $input = preg_replace('/(<\|<\/)(' . 'address\|article\|aside\|blockquote\|br\|canvas\|dd\|div\|dl\|dt\|' . 'fieldset\|figcaption\|figure\|footer\|form\|h1\|h2\|h3\|h4\|h5\|h6\|header\|hgroup\|hr\|li\|' . 'main\|nav\|noscript\|ol\|output\|p\|pre\|section\|table\|tfoot\|ul\|video' . ')\b/i', ' $1$2', $input ); // Strip HTML tags. $input = strip_tags($input); return parent::parse($input); } /* * Method to process HTML input and extract the plain text. * * @param string $input The input to process. * * @return string The plain text input. * * @since 2.5 / protected function process($input) { // Replace any amount of white space with a single space. return preg_replace('#\s+#u', ' ', $input); } /* * Method to remove blocks of text between a start and an end tag. * Each block removed is effectively replaced by a single space. * * Note: The start tag and the end tag must be different. * Note: Blocks must not be nested. * Note: This method will function correctly with multi-byte strings. * * @param string $input String to be processed. * @param string $startTag String representing the start tag. * @param string $endTag String representing the end tag. * * @return string with blocks removed. * * @since 3.4 / private function removeBlocks($input, $startTag, $endTag) { $return = ''; $offset = 0; $startTagLength = strlen($startTag); $endTagLength = strlen($endTag); // Find the first start tag. $start = stripos($input, $startTag); // If no start tags were found, return the string unchanged. if ($start === false) { return $input; } // Look for all blocks defined by the start and end tags. while ($start !== false) { // Accumulate the substring up to the start tag. $return .= substr($input, $offset, $start - $offset) . ' '; // Look for an end tag corresponding to the start tag. $end = stripos($input, $endTag, $start + $startTagLength); // If no corresponding end tag, leave the string alone. if ($end === false) { // Fix the offset so part of the string is not duplicated. $offset = $start; break; } // Advance the start position. $offset = $end + $endTagLength; // Look for the next start tag and loop. $start = stripos($input, $startTag, $offset); } // Add in the final substring after the last end tag. $return .= substr($input, $offset); return $return; } } ��parser/rtf.php��0000604��00000002040�15245703510�0007336 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; JLoader::register('FinderIndexerParser', dirname(__DIR__) . '/parser.php'); /* * RTF Parser class for the Finder indexer package. * * @since 2.5 / class FinderIndexerParserRtf extends FinderIndexerParser { /* * Method to process RTF input and extract the plain text. * * @param string $input The input to process. * * @return string The plain text input. * * @since 2.5 / protected function process($input) { // Remove embedded pictures. $input = preg_replace('#{\\\pict[^}]}#mi', '', $input); // Remove control characters. $input = str_replace(array('{', '}', "\\\n"), array(' ', ' ', "\n"), $input); $input = preg_replace('#\\\([^;]+?);#m', ' ', $input); $input = preg_replace('#\\\[\'a-zA-Z0-9]+#mi', ' ', $input); return $input; } } ��helper.php��0000604��00000034247�15245703510�0006544 0��ustar�00��<?php /** * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; use Joomla\Registry\Registry; use Joomla\String\StringHelper; JLoader::register('FinderIndexerParser', __DIR__ . '/parser.php'); JLoader::register('FinderIndexerStemmer', __DIR__ . '/stemmer.php'); JLoader::register('FinderIndexerToken', __DIR__ . '/token.php'); /* * Helper class for the Finder indexer package. * * @since 2.5 / class FinderIndexerHelper { /* * The token stemmer object. The stemmer is set by whatever class * wishes to use it but it must be an instance of FinderIndexerStemmer. * * @var FinderIndexerStemmer * @since 2.5 / public static $stemmer; /* * A state flag, in order to not constantly check if the stemmer is an instance of FinderIndexerStemmer * * @var boolean * @since 3.7.0 / protected static $stemmerOK; /* * Method to parse input into plain text. * * @param string $input The raw input. * @param string $format The format of the input. [optional] * * @return string The parsed input. * * @since 2.5 * @throws Exception on invalid parser. / public static function parse($input, $format = 'html') { // Get a parser for the specified format and parse the input. return FinderIndexerParser::getInstance($format)->parse($input); } /* * Method to tokenize a text string. * * @param string $input The input to tokenize. * @param string $lang The language of the input. * @param boolean $phrase Flag to indicate whether input could be a phrase. [optional] * * @return array\|FinderIndexerToken An array of FinderIndexerToken objects or a single FinderIndexerToken object. * * @since 2.5 / public static function tokenize($input, $lang, $phrase = false) { static $cache; $store = StringHelper::strlen($input) < 128 ? md5($input . '::' . $lang . '::' . $phrase) : null; // Check if the string has been tokenized already. if ($store && isset($cache[$store])) { return $cache[$store]; } $tokens = array(); $quotes = html_entity_decode('‘’'', ENT_QUOTES, 'UTF-8'); // Get the simple language key. $lang = static::getPrimaryLanguage($lang); / * Parsing the string input into terms is a multi-step process. * * Regexes: * 1. Remove everything except letters, numbers, quotes, apostrophe, plus, dash, period, and comma. * 2. Remove plus, dash, period, and comma characters located before letter characters. * 3. Remove plus, dash, period, and comma characters located after other characters. * 4. Remove plus, period, and comma characters enclosed in alphabetical characters. Ungreedy. * 5. Remove orphaned apostrophe, plus, dash, period, and comma characters. * 6. Remove orphaned quote characters. * 7. Replace the assorted single quotation marks with the ASCII standard single quotation. * 8. Remove multiple space characters and replaces with a single space. / $input = StringHelper::strtolower($input); $input = preg_replace('#[^\pL\pM\pN\p{Pi}\p{Pf}\'+-.,]+#mui', ' ', $input); $input = preg_replace('#(^\|\s)[+-.,]+([\pL\pM]+)#mui', ' $1', $input); $input = preg_replace('#([\pL\pM\pN]+)[+-.,]+(\s\|$)#mui', '$1 ', $input); $input = preg_replace('#([\pL\pM]+)[+.,]+([\pL\pM]+)#muiU', '$1 $2', $input); $input = preg_replace('#(^\|\s)[\'+-.,]+(\s\|$)#mui', ' ', $input); $input = preg_replace('#(^\|\s)[\p{Pi}\p{Pf}]+(\s\|$)#mui', ' ', $input); $input = preg_replace('#[' . $quotes . ']+#mui', '\'', $input); $input = preg_replace('#\s+#mui', ' ', $input); $input = trim($input); // Explode the normalized string to get the terms. $terms = explode(' ', $input); / * If we have Unicode support and are dealing with Chinese text, Chinese * has to be handled specially because there are not necessarily any spaces * between the "words". So, we have to test if the words belong to the Chinese * character set and if so, explode them into single glyphs or "words". / if ($lang === 'zh') { // Iterate through the terms and test if they contain Chinese. for ($i = 0, $n = count($terms); $i < $n; $i++) { $charMatches = array(); $charCount = preg_match_all('#[\p{Han}]#mui', $terms[$i], $charMatches); // Split apart any groups of Chinese characters. for ($j = 0; $j < $charCount; $j++) { $tSplit = StringHelper::str_ireplace($charMatches[0][$j], '', $terms[$i], false); if ((bool) $tSplit) { $terms[$i] = $tSplit; } else { unset($terms[$i]); } $terms[] = $charMatches[0][$j]; } } // Reset array keys. $terms = array_values($terms); } / * If we have to handle the input as a phrase, that means we don't * tokenize the individual terms and we do not create the two and three * term combinations. The phrase must contain more than one word! / if ($phrase === true && count($terms) > 1) { // Create tokens from the phrase. $tokens[] = new FinderIndexerToken($terms, $lang); } else { // Create tokens from the terms. for ($i = 0, $n = count($terms); $i < $n; $i++) { $tokens[] = new FinderIndexerToken($terms[$i], $lang); } // Create two and three word phrase tokens from the individual words. for ($i = 0, $n = count($tokens); $i < $n; $i++) { // Setup the phrase positions. $i2 = $i + 1; $i3 = $i + 2; // Create the two word phrase. if ($i2 < $n && isset($tokens[$i2])) { // Tokenize the two word phrase. $token = new FinderIndexerToken( array( $tokens[$i]->term, $tokens[$i2]->term ), $lang, $lang === 'zh' ? '' : ' ' ); $token->derived = true; // Add the token to the stack. $tokens[] = $token; } // Create the three word phrase. if ($i3 < $n && isset($tokens[$i3])) { // Tokenize the three word phrase. $token = new FinderIndexerToken( array( $tokens[$i]->term, $tokens[$i2]->term, $tokens[$i3]->term ), $lang, $lang === 'zh' ? '' : ' ' ); $token->derived = true; // Add the token to the stack. $tokens[] = $token; } } } if ($store) { $cache[$store] = count($tokens) > 1 ? $tokens : array_shift($tokens); return $cache[$store]; } else { return count($tokens) > 1 ? $tokens : array_shift($tokens); } } /* * Method to get the base word of a token. This method uses the public * {@link FinderIndexerHelper::$stemmer} object if it is set. If no stemmer is set, * the original token is returned. * * @param string $token The token to stem. * @param string $lang The language of the token. * * @return string The root token. * * @since 2.5 / public static function stem($token, $lang) { // Trim apostrophes at either end of the token. $token = trim($token, '\''); // Trim everything after any apostrophe in the token. if ($res = explode('\'', $token)) { $token = $res[0]; } if (static::$stemmerOK === true) { return static::$stemmer->stem($token, $lang); } else { // Stem the token if we have a valid stemmer to use. if (static::$stemmer instanceof FinderIndexerStemmer) { static::$stemmerOK = true; return static::$stemmer->stem($token, $lang); } } return $token; } /* * Method to add a content type to the database. * * @param string $title The type of content. For example: PDF * @param string $mime The mime type of the content. For example: PDF [optional] * * @return integer The id of the content type. * * @since 2.5 * @throws Exception on database error. / public static function addContentType($title, $mime = null) { static $types; $db = JFactory::getDbo(); $query = $db->getQuery(true); // Check if the types are loaded. if (empty($types)) { // Build the query to get the types. $query->select('') ->from($db->quoteName('#__finder_types')); // Get the types. $db->setQuery($query); $types = $db->loadObjectList('title'); } // Check if the type already exists. if (isset($types[$title])) { return (int) $types[$title]->id; } // Add the type. $query->clear() ->insert($db->quoteName('#__finder_types')) ->columns(array($db->quoteName('title'), $db->quoteName('mime'))) ->values($db->quote($title) . ', ' . $db->quote($mime)); $db->setQuery($query); $db->execute(); // Return the new id. return (int) $db->insertid(); } /** * Method to check if a token is common in a language. * * @param string $token The token to test. * @param string $lang The language to reference. * * @return boolean True if common, false otherwise. * * @since 2.5 / public static function isCommon($token, $lang) { static $data; static $default; $langCode = $lang; // If language requested is wildcard, use the default language. if ($default === null && $lang === '') { $default = strstr(self::getDefaultLanguage(), '-', true); $langCode = $default; } // Load the common tokens for the language if necessary. if (!isset($data[$langCode])) { $data[$langCode] = self::getCommonWords($langCode); } // Check if the token is in the common array. return in_array($token, $data[$langCode], true); } /** * Method to get an array of common terms for a language. * * @param string $lang The language to use. * * @return array Array of common terms. * * @since 2.5 * @throws Exception on database error. / public static function getCommonWords($lang) { $db = JFactory::getDbo(); // Create the query to load all the common terms for the language. $query = $db->getQuery(true) ->select($db->quoteName('term')) ->from($db->quoteName('#__finder_terms_common')) ->where($db->quoteName('language') . ' = ' . $db->quote($lang)); // Load all of the common terms for the language. $db->setQuery($query); return $db->loadColumn(); } /* * Method to get the default language for the site. * * @return string The default language string. * * @since 2.5 / public static function getDefaultLanguage() { static $lang; // We need to go to com_languages to get the site default language, it's the best we can guess. if (empty($lang)) { $lang = JComponentHelper::getParams('com_languages')->get('site', 'en-GB'); } return $lang; } /* * Method to parse a language/locale key and return a simple language string. * * @param string $lang The language/locale key. For example: en-GB * * @return string The simple language string. For example: en * * @since 2.5 / public static function getPrimaryLanguage($lang) { static $data; // Only parse the identifier if necessary. if (!isset($data[$lang])) { if (is_callable(array('Locale', 'getPrimaryLanguage'))) { // Get the language key using the Locale package. $data[$lang] = Locale::getPrimaryLanguage($lang); } else { // Get the language key using string position. $data[$lang] = StringHelper::substr($lang, 0, StringHelper::strpos($lang, '-')); } } return $data[$lang]; } /* * Method to get the path (SEF route) for a content item. * * @param string $url The non-SEF route to the content item. * * @return string The path for the content item. * * @since 2.5 * @deprecated 4.0 / public static function getContentPath($url) { static $router; // Only get the router once. if (!($router instanceof JRouter)) { // Get and configure the site router. $config = JFactory::getConfig(); $router = JRouter::getInstance('site'); $router->setMode($config->get('sef', 1)); } // Build the relative route. $uri = $router->build($url); $route = $uri->toString(array('path', 'query', 'fragment')); $route = str_replace(JUri::base(true) . '/', '', $route); return $route; } /* * Method to get extra data for a content before being indexed. This is how * we add Comments, Tags, Labels, etc. that should be available to Finder. * * @param FinderIndexerResult $item The item to index as a FinderIndexerResult object. * * @return boolean True on success, false on failure. * * @since 2.5 * @throws Exception on database error. / public static function getContentExtras(FinderIndexerResult $item) { // Get the event dispatcher. $dispatcher = JEventDispatcher::getInstance(); // Load the finder plugin group. JPluginHelper::importPlugin('finder'); // Trigger the event. $results = $dispatcher->trigger('onPrepareFinderContent', array(&$item)); // Check the returned results. This is for plugins that don't throw // exceptions when they encounter serious errors. if (in_array(false, $results)) { throw new Exception($dispatcher->getError(), 500); } return true; } /* * Method to process content text using the onContentPrepare event trigger. * * @param string $text The content to process. * @param Registry $params The parameters object. [optional] * @param FinderIndexerResult $item The item which get prepared. [optional] * * @return string The processed content. * * @since 2.5 / public static function prepareContent($text, $params = null, FinderIndexerResult $item = null) { static $loaded; // Get the dispatcher. $dispatcher = JEventDispatcher::getInstance(); // Load the content plugins if necessary. if (empty($loaded)) { JPluginHelper::importPlugin('content'); $loaded = true; } // Instantiate the parameter object if necessary. if (!($params instanceof Registry)) { $registry = new Registry($params); $params = $registry; } // Create a mock content object. $content = JTable::getInstance('Content'); $content->text = $text; if ($item) { $content->bind((array) $item); $content->bind($item->getElements()); } if ($item && !empty($item->context)) { $content->context = $item->context; } // Fire the onContentPrepare event. $dispatcher->trigger('onContentPrepare', array('com_finder.indexer', &$content, &$params, 0)); return $content->text; } } ��adapter.php��0000604��00000052672�15245703510�0006707 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; use Joomla\Utilities\ArrayHelper; JLoader::register('FinderIndexer', __DIR__ . '/indexer.php'); JLoader::register('FinderIndexerHelper', __DIR__ . '/helper.php'); JLoader::register('FinderIndexerResult', __DIR__ . '/result.php'); JLoader::register('FinderIndexerTaxonomy', __DIR__ . '/taxonomy.php'); /* * Prototype adapter class for the Finder indexer package. * * @since 2.5 / abstract class FinderIndexerAdapter extends JPlugin { /* * The context is somewhat arbitrary but it must be unique or there will be * conflicts when managing plugin/indexer state. A good best practice is to * use the plugin name suffix as the context. For example, if the plugin is * named 'plgFinderContent', the context could be 'Content'. * * @var string * @since 2.5 / protected $context; /* * The extension name. * * @var string * @since 2.5 / protected $extension; /* * The sublayout to use when rendering the results. * * @var string * @since 2.5 / protected $layout; /* * The mime type of the content the adapter indexes. * * @var string * @since 2.5 / protected $mime; /* * The access level of an item before save. * * @var integer * @since 2.5 / protected $old_access; /* * The access level of a category before save. * * @var integer * @since 2.5 / protected $old_cataccess; /* * The type of content the adapter indexes. * * @var string * @since 2.5 / protected $type_title; /* * The type id of the content. * * @var integer * @since 2.5 / protected $type_id; /* * The database object. * * @var object * @since 2.5 / protected $db; /* * The table name. * * @var string * @since 2.5 / protected $table; /* * The indexer object. * * @var FinderIndexer * @since 3.0 / protected $indexer; /* * The field the published state is stored in. * * @var string * @since 2.5 / protected $state_field = 'state'; /* * Method to instantiate the indexer adapter. * * @param object $subject The object to observe. * @param array $config An array that holds the plugin configuration. * * @since 2.5 / public function __construct(&$subject, $config) { // Get the database object. $this->db = JFactory::getDbo(); // Call the parent constructor. parent::__construct($subject, $config); // Get the type id. $this->type_id = $this->getTypeId(); // Add the content type if it doesn't exist and is set. if (empty($this->type_id) && !empty($this->type_title)) { $this->type_id = FinderIndexerHelper::addContentType($this->type_title, $this->mime); } // Check for a layout override. if ($this->params->get('layout')) { $this->layout = $this->params->get('layout'); } // Get the indexer object $this->indexer = FinderIndexer::getInstance(); } /* * Method to get the adapter state and push it into the indexer. * * @return void * * @since 2.5 * @throws Exception on error. / public function onStartIndex() { // Get the indexer state. $iState = FinderIndexer::getState(); // Get the number of content items. $total = (int) $this->getContentCount(); // Add the content count to the total number of items. $iState->totalItems += $total; // Populate the indexer state information for the adapter. $iState->pluginState[$this->context]['total'] = $total; $iState->pluginState[$this->context]['offset'] = 0; // Set the indexer state. FinderIndexer::setState($iState); } /* * Method to prepare for the indexer to be run. This method will often * be used to include dependencies and things of that nature. * * @return boolean True on success. * * @since 2.5 * @throws Exception on error. / public function onBeforeIndex() { // Get the indexer and adapter state. $iState = FinderIndexer::getState(); $aState = $iState->pluginState[$this->context]; // Check the progress of the indexer and the adapter. if ($iState->batchOffset == $iState->batchSize \|\| $aState['offset'] == $aState['total']) { return true; } // Run the setup method. return $this->setup(); } /* * Method to index a batch of content items. This method can be called by * the indexer many times throughout the indexing process depending on how * much content is available for indexing. It is important to track the * progress correctly so we can display it to the user. * * @return boolean True on success. * * @since 2.5 * @throws Exception on error. / public function onBuildIndex() { // Get the indexer and adapter state. $iState = FinderIndexer::getState(); $aState = $iState->pluginState[$this->context]; // Check the progress of the indexer and the adapter. if ($iState->batchOffset == $iState->batchSize \|\| $aState['offset'] == $aState['total']) { return true; } // Get the batch offset and size. $offset = (int) $aState['offset']; $limit = (int) ($iState->batchSize - $iState->batchOffset); // Get the content items to index. $items = $this->getItems($offset, $limit); // Iterate through the items and index them. for ($i = 0, $n = count($items); $i < $n; $i++) { // Index the item. $this->index($items[$i]); // Adjust the offsets. $offset++; $iState->batchOffset++; $iState->totalItems--; } // Update the indexer state. $aState['offset'] = $offset; $iState->pluginState[$this->context] = $aState; FinderIndexer::setState($iState); return true; } /* * Method to change the value of a content item's property in the links * table. This is used to synchronize published and access states that * are changed when not editing an item directly. * * @param string $id The ID of the item to change. * @param string $property The property that is being changed. * @param integer $value The new value of that property. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / protected function change($id, $property, $value) { // Check for a property we know how to handle. if ($property !== 'state' && $property !== 'access') { return true; } // Get the URL for the content id. $item = $this->db->quote($this->getUrl($id, $this->extension, $this->layout)); // Update the content items. $query = $this->db->getQuery(true) ->update($this->db->quoteName('#__finder_links')) ->set($this->db->quoteName($property) . ' = ' . (int) $value) ->where($this->db->quoteName('url') . ' = ' . $item); $this->db->setQuery($query); $this->db->execute(); return true; } /* * Method to index an item. * * @param FinderIndexerResult $item The item to index as a FinderIndexerResult object. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / abstract protected function index(FinderIndexerResult $item); /* * Method to reindex an item. * * @param integer $id The ID of the item to reindex. * * @return void * * @since 2.5 * @throws Exception on database error. / protected function reindex($id) { // Run the setup method. $this->setup(); // Remove the old item. $this->remove($id); // Get the item. $item = $this->getItem($id); // Index the item. $this->index($item); } /* * Method to remove an item from the index. * * @param string $id The ID of the item to remove. * * @return boolean True on success. * * @since 2.5 * @throws Exception on database error. / protected function remove($id) { // Get the item's URL $url = $this->db->quote($this->getUrl($id, $this->extension, $this->layout)); // Get the link ids for the content items. $query = $this->db->getQuery(true) ->select($this->db->quoteName('link_id')) ->from($this->db->quoteName('#__finder_links')) ->where($this->db->quoteName('url') . ' = ' . $url); $this->db->setQuery($query); $items = $this->db->loadColumn(); // Check the items. if (empty($items)) { return true; } // Remove the items. foreach ($items as $item) { $this->indexer->remove($item); } return true; } /* * Method to setup the adapter before indexing. * * @return boolean True on success, false on failure. * * @since 2.5 * @throws Exception on database error. / abstract protected function setup(); /* * Method to update index data on category access level changes * * @param JTable $row A JTable object * * @return void * * @since 2.5 / protected function categoryAccessChange($row) { $query = clone $this->getStateQuery(); $query->where('c.id = ' . (int) $row->id); // Get the access level. $this->db->setQuery($query); $items = $this->db->loadObjectList(); // Adjust the access level for each item within the category. foreach ($items as $item) { // Set the access level. $temp = max($item->access, $row->access); // Update the item. $this->change((int) $item->id, 'access', $temp); // Reindex the item $this->reindex($row->id); } } /* * Method to update index data on category access level changes * * @param array $pks A list of primary key ids of the content that has changed state. * @param integer $value The value of the state that the content has been changed to. * * @return void * * @since 2.5 / protected function categoryStateChange($pks, $value) { / * The item's published state is tied to the category * published state so we need to look up all published states * before we change anything. / foreach ($pks as $pk) { $query = clone $this->getStateQuery(); $query->where('c.id = ' . (int) $pk); // Get the published states. $this->db->setQuery($query); $items = $this->db->loadObjectList(); // Adjust the state for each item within the category. foreach ($items as $item) { // Translate the state. $temp = $this->translateState($item->state, $value); // Update the item. $this->change($item->id, 'state', $temp); // Reindex the item $this->reindex($item->id); } } } /* * Method to check the existing access level for categories * * @param JTable $row A JTable object * * @return void * * @since 2.5 / protected function checkCategoryAccess($row) { $query = $this->db->getQuery(true) ->select($this->db->quoteName('access')) ->from($this->db->quoteName('#__categories')) ->where($this->db->quoteName('id') . ' = ' . (int) $row->id); $this->db->setQuery($query); // Store the access level to determine if it changes $this->old_cataccess = $this->db->loadResult(); } /* * Method to check the existing access level for items * * @param JTable $row A JTable object * * @return void * * @since 2.5 / protected function checkItemAccess($row) { $query = $this->db->getQuery(true) ->select($this->db->quoteName('access')) ->from($this->db->quoteName($this->table)) ->where($this->db->quoteName('id') . ' = ' . (int) $row->id); $this->db->setQuery($query); // Store the access level to determine if it changes $this->old_access = $this->db->loadResult(); } /* * Method to get the number of content items available to index. * * @return integer The number of content items available to index. * * @since 2.5 * @throws Exception on database error. / protected function getContentCount() { $return = 0; // Get the list query. $query = $this->getListQuery(); // Check if the query is valid. if (empty($query)) { return $return; } // Tweak the SQL query to make the total lookup faster. if ($query instanceof JDatabaseQuery) { $query = clone $query; $query->clear('select') ->select('COUNT()') ->clear('order'); } // Get the total number of content items to index. $this->db->setQuery($query); return (int) $this->db->loadResult(); } /** * Method to get a content item to index. * * @param integer $id The id of the content item. * * @return FinderIndexerResult A FinderIndexerResult object. * * @since 2.5 * @throws Exception on database error. / protected function getItem($id) { // Get the list query and add the extra WHERE clause. $query = $this->getListQuery(); $query->where('a.id = ' . (int) $id); // Get the item to index. $this->db->setQuery($query); $row = $this->db->loadAssoc(); // Convert the item to a result object. $item = ArrayHelper::toObject((array) $row, 'FinderIndexerResult'); // Set the item type. $item->type_id = $this->type_id; // Set the item layout. $item->layout = $this->layout; return $item; } /* * Method to get a list of content items to index. * * @param integer $offset The list offset. * @param integer $limit The list limit. * @param JDatabaseQuery $query A JDatabaseQuery object. [optional] * * @return array An array of FinderIndexerResult objects. * * @since 2.5 * @throws Exception on database error. / protected function getItems($offset, $limit, $query = null) { $items = array(); // Get the content items to index. $this->db->setQuery($this->getListQuery($query), $offset, $limit); $rows = $this->db->loadAssocList(); // Convert the items to result objects. foreach ($rows as $row) { // Convert the item to a result object. $item = ArrayHelper::toObject((array) $row, 'FinderIndexerResult'); // Set the item type. $item->type_id = $this->type_id; // Set the mime type. $item->mime = $this->mime; // Set the item layout. $item->layout = $this->layout; // Set the extension if present if (isset($row->extension)) { $item->extension = $row->extension; } // Add the item to the stack. $items[] = $item; } return $items; } /* * Method to get the SQL query used to retrieve the list of content items. * * @param mixed $query A JDatabaseQuery object. [optional] * * @return JDatabaseQuery A database object. * * @since 2.5 / protected function getListQuery($query = null) { // Check if we can use the supplied SQL query. return $query instanceof JDatabaseQuery ? $query : $this->db->getQuery(true); } /* * Method to get the plugin type * * @param integer $id The plugin ID * * @return string The plugin type * * @since 2.5 / protected function getPluginType($id) { // Prepare the query $query = $this->db->getQuery(true) ->select($this->db->quoteName('element')) ->from($this->db->quoteName('#__extensions')) ->where($this->db->quoteName('extension_id') . ' = ' . (int) $id); $this->db->setQuery($query); return $this->db->loadResult(); } /* * Method to get a SQL query to load the published and access states for * an article and category. * * @return JDatabaseQuery A database object. * * @since 2.5 / protected function getStateQuery() { $query = $this->db->getQuery(true); // Item ID $query->select('a.id'); // Item and category published state $query->select('a.' . $this->state_field . ' AS state, c.published AS cat_state'); // Item and category access levels $query->select('a.access, c.access AS cat_access') ->from($this->table . ' AS a') ->join('LEFT', '#__categories AS c ON c.id = a.catid'); return $query; } /* * Method to get the query clause for getting items to update by time. * * @param string $time The modified timestamp. * * @return JDatabaseQuery A database object. * * @since 2.5 / protected function getUpdateQueryByTime($time) { // Build an SQL query based on the modified time. $query = $this->db->getQuery(true) ->where('a.modified >= ' . $this->db->quote($time)); return $query; } /* * Method to get the query clause for getting items to update by id. * * @param array $ids The ids to load. * * @return JDatabaseQuery A database object. * * @since 2.5 / protected function getUpdateQueryByIds($ids) { // Build an SQL query based on the item ids. $query = $this->db->getQuery(true) ->where('a.id IN(' . implode(',', $ids) . ')'); return $query; } /* * Method to get the type id for the adapter content. * * @return integer The numeric type id for the content. * * @since 2.5 * @throws Exception on database error. / protected function getTypeId() { // Get the type id from the database. $query = $this->db->getQuery(true) ->select($this->db->quoteName('id')) ->from($this->db->quoteName('#__finder_types')) ->where($this->db->quoteName('title') . ' = ' . $this->db->quote($this->type_title)); $this->db->setQuery($query); return (int) $this->db->loadResult(); } /* * Method to get the URL for the item. The URL is how we look up the link * in the Finder index. * * @param integer $id The id of the item. * @param string $extension The extension the category is in. * @param string $view The view for the URL. * * @return string The URL of the item. * * @since 2.5 / protected function getUrl($id, $extension, $view) { return 'index.php?option=' . $extension . '&view=' . $view . '&id=' . $id; } /* * Method to get the page title of any menu item that is linked to the * content item, if it exists and is set. * * @param string $url The URL of the item. * * @return mixed The title on success, null if not found. * * @since 2.5 * @throws Exception on database error. / protected function getItemMenuTitle($url) { $return = null; // Set variables $user = JFactory::getUser(); $groups = implode(',', $user->getAuthorisedViewLevels()); // Build a query to get the menu params. $query = $this->db->getQuery(true) ->select($this->db->quoteName('params')) ->from($this->db->quoteName('#__menu')) ->where($this->db->quoteName('link') . ' = ' . $this->db->quote($url)) ->where($this->db->quoteName('published') . ' = 1') ->where($this->db->quoteName('access') . ' IN (' . $groups . ')'); // Get the menu params from the database. $this->db->setQuery($query); $params = $this->db->loadResult(); // Check the results. if (empty($params)) { return $return; } // Instantiate the params. $params = json_decode($params); // Get the page title if it is set. if (isset($params->page_title) && $params->page_title) { $return = $params->page_title; } return $return; } /* * Method to update index data on access level changes * * @param JTable $row A JTable object * * @return void * * @since 2.5 / protected function itemAccessChange($row) { $query = clone $this->getStateQuery(); $query->where('a.id = ' . (int) $row->id); // Get the access level. $this->db->setQuery($query); $item = $this->db->loadObject(); // Set the access level. $temp = max($row->access, $item->cat_access); // Update the item. $this->change((int) $row->id, 'access', $temp); } /* * Method to update index data on published state changes * * @param array $pks A list of primary key ids of the content that has changed state. * @param integer $value The value of the state that the content has been changed to. * * @return void * * @since 2.5 / protected function itemStateChange($pks, $value) { / * The item's published state is tied to the category * published state so we need to look up all published states * before we change anything. / foreach ($pks as $pk) { $query = clone $this->getStateQuery(); $query->where('a.id = ' . (int) $pk); // Get the published states. $this->db->setQuery($query); $item = $this->db->loadObject(); // Translate the state. $temp = $this->translateState($value, $item->cat_state); // Update the item. $this->change($pk, 'state', $temp); // Reindex the item $this->reindex($pk); } } /* * Method to update index data when a plugin is disabled * * @param array $pks A list of primary key ids of the content that has changed state. * * @return void * * @since 2.5 / protected function pluginDisable($pks) { // Since multiple plugins may be disabled at a time, we need to check first // that we're handling the appropriate one for the context foreach ($pks as $pk) { if ($this->getPluginType($pk) == strtolower($this->context)) { // Get all of the items to unindex them $query = clone $this->getStateQuery(); $this->db->setQuery($query); $items = $this->db->loadColumn(); // Remove each item foreach ($items as $item) { $this->remove($item); } } } } /* * Method to translate the native content states into states that the * indexer can use. * * @param integer $item The item state. * @param integer $category The category state. [optional] * * @return integer The translated indexer state. * * @since 2.5 / protected function translateState($item, $category = null) { // If category is present, factor in its states as well if ($category !== null && $category == 0) { $item = 0; } // Translate the state switch ($item) { // Published and archived items only should return a published state case 1; case 2: return 1; // All other states should return an unpublished state default: case 0: return 0; } } } ��parser.php��0000604��00000005717�15245703510�0006561 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; /* * Parser base class for the Finder indexer package. * * @since 2.5 / abstract class FinderIndexerParser { /* * Method to get a parser, creating it if necessary. * * @param string $format The type of parser to load. * * @return FinderIndexerParser A FinderIndexerParser instance. * * @since 2.5 * @throws Exception on invalid parser. / public static function getInstance($format) { static $instances; // Only create one parser for each format. if (isset($instances[$format])) { return $instances[$format]; } // Create an array of instances if necessary. if (!is_array($instances)) { $instances = array(); } // Setup the adapter for the parser. $format = JFilterInput::getInstance()->clean($format, 'cmd'); $path = __DIR__ . '/parser/' . $format . '.php'; $class = 'FinderIndexerParser' . ucfirst($format); // Check if a parser exists for the format. if (!file_exists($path)) { // Throw invalid format exception. throw new Exception(JText::sprintf('COM_FINDER_INDEXER_INVALID_PARSER', $format)); } // Instantiate the parser. JLoader::register($class, $path); $instances[$format] = new $class; return $instances[$format]; } /* * Method to parse input and extract the plain text. Because this method is * called from both inside and outside the indexer, it needs to be able to * batch out its parsing functionality to deal with the inefficiencies of * regular expressions. We will parse recursively in 2KB chunks. * * @param string $input The input to parse. * * @return string The plain text input. * * @since 2.5 / public function parse($input) { // If the input is less than 2KB we can parse it in one go. if (strlen($input) <= 2048) { return $this->process($input); } // Input is longer than 2Kb so parse it in chunks of 2Kb or less. $start = 0; $end = strlen($input); $chunk = 2048; $return = null; while ($start < $end) { // Setup the string. $string = substr($input, $start, $chunk); // Find the last space character if we aren't at the end. $ls = (($start + $chunk) < $end ? strrpos($string, ' ') : false); // Truncate to the last space character. if ($ls !== false) { $string = substr($string, 0, $ls); } // Adjust the start position for the next iteration. $start += ($ls !== false ? ($ls + 1 - $chunk) + $chunk : $chunk); // Parse the chunk. $return .= $this->process($string); } return $return; } /* * Method to process input and extract the plain text. * * @param string $input The input to process. * * @return string The plain text input. * * @since 2.5 / abstract protected function process($input); } ��stemmer.php��0000604��00000003566�15245703510�0006741 0��ustar�00��<?php /* * @package Joomla.Administrator * @subpackage com_finder * * @copyright (C) 2011 Open Source Matters, Inc. <https://www.joomla.org> * @license GNU General Public License version 2 or later; see LICENSE.txt / defined('_JEXEC') or die; /* * Stemmer base class for the Finder indexer package. * * @since 2.5 / abstract class FinderIndexerStemmer { /* * An internal cache of stemmed tokens. * * @var array * @since 2.5 / public $cache = array(); /* * Method to get a stemmer, creating it if necessary. * * @param string $adapter The type of stemmer to load. * * @return FinderIndexerStemmer A FinderIndexerStemmer instance. * * @since 2.5 * @throws Exception on invalid stemmer. / public static function getInstance($adapter) { static $instances; // Only create one stemmer for each adapter. if (isset($instances[$adapter])) { return $instances[$adapter]; } // Create an array of instances if necessary. if (!is_array($instances)) { $instances = array(); } // Setup the adapter for the stemmer. $adapter = JFilterInput::getInstance()->clean($adapter, 'cmd'); $path = __DIR__ . '/stemmer/' . $adapter . '.php'; $class = 'FinderIndexerStemmer' . ucfirst($adapter); // Check if a stemmer exists for the adapter. if (!file_exists($path)) { // Throw invalid adapter exception. throw new Exception(JText::sprintf('COM_FINDER_INDEXER_INVALID_STEMMER', $adapter)); } // Instantiate the stemmer. JLoader::register($class, $path); $instances[$adapter] = new $class; return $instances[$adapter]; } /* * Method to stem a token and return the root. * * @param string $token The token to stem. * @param string $lang The language of the token. * * @return string The root token. * * @since 2.5 */ abstract public function stem($token, $lang); } ��

| ver. 1.4 | Github | . | PHP 8.1.34 | Генерация страницы: 0 | proxy | phpinfo | Настройка