Source of file Wildcard.php
Size: 12,075 Bytes - Last Modified: 2021-12-23T10:29:30+00:00
/var/www/docs.ssmods.com/process/src/thirdparty/Zend/Search/Lucene/Search/Query/Wildcard.php
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367 | <?php /** * Zend Framework * * LICENSE * * This source file is subject to the new BSD license that is bundled * with this package in the file LICENSE.txt. * It is also available through the world-wide-web at this URL: * http://framework.zend.com/license/new-bsd * If you did not receive a copy of the license and are unable to * obtain it through the world-wide-web, please send an email * to license@zend.com so we can send you a copy immediately. * * @category Zend * @package Zend_Search_Lucene * @subpackage Search * @copyright Copyright (c) 2005-2010 Zend Technologies USA Inc. (http://www.zend.com) * @license http://framework.zend.com/license/new-bsd New BSD License * @version $Id: Wildcard.php 20096 2010-01-06 02:05:09Z bkarwin $ */ /** * Zend_Search_Lucene_Search_Query */ require_once 'Zend/Search/Lucene/Search/Query.php'; /** * @category Zend * @package Zend_Search_Lucene * @subpackage Search * @copyright Copyright (c) 2005-2010 Zend Technologies USA Inc. (http://www.zend.com) * @license http://framework.zend.com/license/new-bsd New BSD License */ class Zend_Search_Lucene_Search_Query_Wildcard extends Zend_Search_Lucene_Search_Query { /** * Search pattern. * * Field has to be fully specified or has to be null * Text may contain '*' or '?' symbols * * @var Zend_Search_Lucene_Index_Term */ private $_pattern; /** * Matched terms. * * Matched terms list. * It's filled during the search (rewrite operation) and may be used for search result * post-processing * * Array of Zend_Search_Lucene_Index_Term objects * * @var array */ private $_matches = null; /** * Minimum term prefix length (number of minimum non-wildcard characters) * * @var integer */ private static $_minPrefixLength = 3; /** * Zend_Search_Lucene_Search_Query_Wildcard constructor. * * @param Zend_Search_Lucene_Index_Term $pattern */ public function __construct(Zend_Search_Lucene_Index_Term $pattern) { $this->_pattern = $pattern; } /** * Get minimum prefix length * * @return integer */ public static function getMinPrefixLength() { return self::$_minPrefixLength; } /** * Set minimum prefix length * * @param integer $minPrefixLength */ public static function setMinPrefixLength($minPrefixLength) { self::$_minPrefixLength = $minPrefixLength; } /** * Get terms prefix * * @param string $word * @return string */ private static function _getPrefix($word) { $questionMarkPosition = strpos($word, '?'); $astrericPosition = strpos($word, '*'); if ($questionMarkPosition !== false) { if ($astrericPosition !== false) { return substr($word, 0, min($questionMarkPosition, $astrericPosition)); } return substr($word, 0, $questionMarkPosition); } else if ($astrericPosition !== false) { return substr($word, 0, $astrericPosition); } return $word; } /** * Re-write query into primitive queries in the context of specified index * * @param Zend_Search_Lucene_Interface $index * @return Zend_Search_Lucene_Search_Query * @throws Zend_Search_Lucene_Exception */ public function rewrite(Zend_Search_Lucene_Interface $index) { $this->_matches = array(); if ($this->_pattern->field === null) { // Search through all fields $fields = $index->getFieldNames(true /* indexed fields list */); } else { $fields = array($this->_pattern->field); } $prefix = self::_getPrefix($this->_pattern->text); $prefixLength = strlen($prefix); $matchExpression = '/^' . str_replace(array('\\?', '\\*'), array('.', '.*'), preg_quote($this->_pattern->text, '/')) . '$/'; if ($prefixLength < self::$_minPrefixLength) { include_once 'Zend/Search/Lucene/Exception.php'; throw new Zend_Search_Lucene_Exception('At least ' . self::$_minPrefixLength . ' non-wildcard characters are required at the beginning of pattern.'); } /** * @todo check for PCRE unicode support may be performed through Zend_Environment in some future */ if (@preg_match('/\pL/u', 'a') == 1) { // PCRE unicode support is turned on // add Unicode modifier to the match expression $matchExpression .= 'u'; } $maxTerms = Zend_Search_Lucene::getTermsPerQueryLimit(); foreach ($fields as $field) { $index->resetTermsStream(); include_once 'Zend/Search/Lucene/Index/Term.php'; if ($prefix != '') { $index->skipTo(new Zend_Search_Lucene_Index_Term($prefix, $field)); while ($index->currentTerm() !== null && $index->currentTerm()->field == $field && substr($index->currentTerm()->text, 0, $prefixLength) == $prefix) { if (preg_match($matchExpression, $index->currentTerm()->text) === 1) { $this->_matches[] = $index->currentTerm(); if ($maxTerms != 0 && count($this->_matches) > $maxTerms) { include_once 'Zend/Search/Lucene/Exception.php'; throw new Zend_Search_Lucene_Exception('Terms per query limit is reached.'); } } $index->nextTerm(); } } else { $index->skipTo(new Zend_Search_Lucene_Index_Term('', $field)); while ($index->currentTerm() !== null && $index->currentTerm()->field == $field) { if (preg_match($matchExpression, $index->currentTerm()->text) === 1) { $this->_matches[] = $index->currentTerm(); if ($maxTerms != 0 && count($this->_matches) > $maxTerms) { include_once 'Zend/Search/Lucene/Exception.php'; throw new Zend_Search_Lucene_Exception('Terms per query limit is reached.'); } } $index->nextTerm(); } } $index->closeTermsStream(); } if (count($this->_matches) == 0) { include_once 'Zend/Search/Lucene/Search/Query/Empty.php'; return new Zend_Search_Lucene_Search_Query_Empty(); } else if (count($this->_matches) == 1) { include_once 'Zend/Search/Lucene/Search/Query/Term.php'; return new Zend_Search_Lucene_Search_Query_Term(reset($this->_matches)); } else { include_once 'Zend/Search/Lucene/Search/Query/MultiTerm.php'; $rewrittenQuery = new Zend_Search_Lucene_Search_Query_MultiTerm(); foreach ($this->_matches as $matchedTerm) { $rewrittenQuery->addTerm($matchedTerm); } return $rewrittenQuery; } } /** * Optimize query in the context of specified index * * @param Zend_Search_Lucene_Interface $index * @return Zend_Search_Lucene_Search_Query */ public function optimize(Zend_Search_Lucene_Interface $index) { include_once 'Zend/Search/Lucene/Exception.php'; throw new Zend_Search_Lucene_Exception('Wildcard query should not be directly used for search. Use $query->rewrite($index)'); } /** * Returns query pattern * * @return Zend_Search_Lucene_Index_Term */ public function getPattern() { return $this->_pattern; } /** * Return query terms * * @return array * @throws Zend_Search_Lucene_Exception */ public function getQueryTerms() { if ($this->_matches === null) { include_once 'Zend/Search/Lucene/Exception.php'; throw new Zend_Search_Lucene_Exception('Search has to be performed first to get matched terms'); } return $this->_matches; } /** * Constructs an appropriate Weight implementation for this query. * * @param Zend_Search_Lucene_Interface $reader * @return Zend_Search_Lucene_Search_Weight * @throws Zend_Search_Lucene_Exception */ public function createWeight(Zend_Search_Lucene_Interface $reader) { include_once 'Zend/Search/Lucene/Exception.php'; throw new Zend_Search_Lucene_Exception('Wildcard query should not be directly used for search. Use $query->rewrite($index)'); } /** * Execute query in context of index reader * It also initializes necessary internal structures * * @param Zend_Search_Lucene_Interface $reader * @param Zend_Search_Lucene_Index_DocsFilter|null $docsFilter * @throws Zend_Search_Lucene_Exception */ public function execute(Zend_Search_Lucene_Interface $reader, $docsFilter = null) { include_once 'Zend/Search/Lucene/Exception.php'; throw new Zend_Search_Lucene_Exception('Wildcard query should not be directly used for search. Use $query->rewrite($index)'); } /** * Get document ids likely matching the query * * It's an array with document ids as keys (performance considerations) * * @return array * @throws Zend_Search_Lucene_Exception */ public function matchedDocs() { include_once 'Zend/Search/Lucene/Exception.php'; throw new Zend_Search_Lucene_Exception('Wildcard query should not be directly used for search. Use $query->rewrite($index)'); } /** * Score specified document * * @param integer $docId * @param Zend_Search_Lucene_Interface $reader * @return float * @throws Zend_Search_Lucene_Exception */ public function score($docId, Zend_Search_Lucene_Interface $reader) { include_once 'Zend/Search/Lucene/Exception.php'; throw new Zend_Search_Lucene_Exception('Wildcard query should not be directly used for search. Use $query->rewrite($index)'); } /** * Query specific matches highlighting * * @param Zend_Search_Lucene_Search_Highlighter_Interface $highlighter Highlighter object (also contains doc for highlighting) */ protected function _highlightMatches(Zend_Search_Lucene_Search_Highlighter_Interface $highlighter) { $words = array(); $matchExpression = '/^' . str_replace(array('\\?', '\\*'), array('.', '.*'), preg_quote($this->_pattern->text, '/')) . '$/'; if (@preg_match('/\pL/u', 'a') == 1) { // PCRE unicode support is turned on // add Unicode modifier to the match expression $matchExpression .= 'u'; } $docBody = $highlighter->getDocument()->getFieldUtf8Value('body'); include_once 'Zend/Search/Lucene/Analysis/Analyzer.php'; $tokens = Zend_Search_Lucene_Analysis_Analyzer::getDefault()->tokenize($docBody, 'UTF-8'); foreach ($tokens as $token) { if (preg_match($matchExpression, $token->getTermText()) === 1) { $words[] = $token->getTermText(); } } $highlighter->highlight($words); } /** * Print a query * * @return string */ public function __toString() { // It's used only for query visualisation, so we don't care about characters escaping if ($this->_pattern->field !== null) { $query = $this->_pattern->field . ':'; } else { $query = ''; } $query .= $this->_pattern->text; if ($this->getBoost() != 1) { $query = $query . '^' . round($this->getBoost(), 4); } return $query; } } |