--- /dev/null
+<?php\r
+/**\r
+ * Zend Framework\r
+ *\r
+ * LICENSE\r
+ *\r
+ * This source file is subject to the new BSD license that is bundled\r
+ * with this package in the file LICENSE.txt.\r
+ * It is also available through the world-wide-web at this URL:\r
+ * http://framework.zend.com/license/new-bsd\r
+ * If you did not receive a copy of the license and are unable to\r
+ * obtain it through the world-wide-web, please send an email\r
+ * to license@zend.com so we can send you a copy immediately.\r
+ *\r
+ * @category Zend\r
+ * @package Zend_Search_Lucene\r
+ * @subpackage Search\r
+ * @copyright Copyright (c) 2005-2009 Zend Technologies USA Inc. (http://www.zend.com)\r
+ * @license http://framework.zend.com/license/new-bsd New BSD License\r
+ * @version $Id: Phrase.php 16971 2009-07-22 18:05:45Z mikaelkael $\r
+ */\r
+\r
+\r
+/** Zend_Search_Lucene_Search_Query_Processing */\r
+require_once 'Zend/Search/Lucene/Search/Query/Preprocessing.php';\r
+\r
+/** Zend_Search_Lucene_Search_Query_Phrase */\r
+require_once 'Zend/Search/Lucene/Search/Query/Phrase.php';\r
+\r
+/** Zend_Search_Lucene_Search_Query_Insignificant */\r
+require_once 'Zend/Search/Lucene/Search/Query/Insignificant.php';\r
+\r
+/** Zend_Search_Lucene_Search_Query_Empty */\r
+require_once 'Zend/Search/Lucene/Search/Query/Empty.php';\r
+\r
+/** Zend_Search_Lucene_Search_Query_Term */\r
+require_once 'Zend/Search/Lucene/Search/Query/Term.php';\r
+\r
+/** Zend_Search_Lucene_Index_Term */\r
+require_once 'Zend/Search/Lucene/Index/Term.php';\r
+\r
+\r
+/**\r
+ * It's an internal abstract class intended to finalize ase a query processing after query parsing.\r
+ * This type of query is not actually involved into query execution.\r
+ *\r
+ * @category Zend\r
+ * @package Zend_Search_Lucene\r
+ * @subpackage Search\r
+ * @internal\r
+ * @copyright Copyright (c) 2005-2009 Zend Technologies USA Inc. (http://www.zend.com)\r
+ * @license http://framework.zend.com/license/new-bsd New BSD License\r
+ */\r
+class Zend_Search_Lucene_Search_Query_Preprocessing_Phrase extends Zend_Search_Lucene_Search_Query_Preprocessing\r
+{\r
+ /**\r
+ * Phrase to find.\r
+ *\r
+ * @var string\r
+ */\r
+ private $_phrase;\r
+\r
+ /**\r
+ * Phrase encoding (field name is always provided using UTF-8 encoding since it may be retrieved from index).\r
+ *\r
+ * @var string\r
+ */\r
+ private $_phraseEncoding;\r
+\r
+\r
+ /**\r
+ * Field name.\r
+ *\r
+ * @var string\r
+ */\r
+ private $_field;\r
+\r
+ /**\r
+ * Sets the number of other words permitted between words in query phrase.\r
+ * If zero, then this is an exact phrase search. For larger values this works\r
+ * like a WITHIN or NEAR operator.\r
+ *\r
+ * The slop is in fact an edit-distance, where the units correspond to\r
+ * moves of terms in the query phrase out of position. For example, to switch\r
+ * the order of two words requires two moves (the first move places the words\r
+ * atop one another), so to permit re-orderings of phrases, the slop must be\r
+ * at least two.\r
+ * More exact matches are scored higher than sloppier matches, thus search\r
+ * results are sorted by exactness.\r
+ *\r
+ * The slop is zero by default, requiring exact matches.\r
+ *\r
+ * @var integer\r
+ */\r
+ private $_slop;\r
+\r
+ /**\r
+ * Class constructor. Create a new preprocessing object for prase query.\r
+ *\r
+ * @param string $phrase Phrase to search.\r
+ * @param string $phraseEncoding Phrase encoding.\r
+ * @param string $fieldName Field name.\r
+ */\r
+ public function __construct($phrase, $phraseEncoding, $fieldName)\r
+ {\r
+ $this->_phrase = $phrase;\r
+ $this->_phraseEncoding = $phraseEncoding;\r
+ $this->_field = $fieldName;\r
+ }\r
+\r
+ /**\r
+ * Set slop\r
+ *\r
+ * @param integer $slop\r
+ */\r
+ public function setSlop($slop)\r
+ {\r
+ $this->_slop = $slop;\r
+ }\r
+\r
+\r
+ /**\r
+ * Get slop\r
+ *\r
+ * @return integer\r
+ */\r
+ public function getSlop()\r
+ {\r
+ return $this->_slop;\r
+ }\r
+\r
+ /**\r
+ * Re-write query into primitive queries in the context of specified index\r
+ *\r
+ * @param Zend_Search_Lucene_Interface $index\r
+ * @return Zend_Search_Lucene_Search_Query\r
+ */\r
+ public function rewrite(Zend_Search_Lucene_Interface $index)\r
+ {\r
+// Allow to use wildcards within phrases\r
+// They are either removed by text analyzer or used as a part of keyword for keyword fields\r
+//\r
+// if (strpos($this->_phrase, '?') !== false || strpos($this->_phrase, '*') !== false) {\r
+// require_once 'Zend/Search/Lucene/Search/QueryParserException.php';\r
+// throw new Zend_Search_Lucene_Search_QueryParserException('Wildcards are only allowed in a single terms.');\r
+// }\r
+\r
+ // Split query into subqueries if field name is not specified\r
+ if ($this->_field === null) {\r
+ $query = new Zend_Search_Lucene_Search_Query_Boolean();\r
+ $query->setBoost($this->getBoost());\r
+\r
+ if (Zend_Search_Lucene::getDefaultSearchField() === null) {\r
+ $searchFields = $index->getFieldNames(true);\r
+ } else {\r
+ $searchFields = array(Zend_Search_Lucene::getDefaultSearchField());\r
+ }\r
+\r
+ foreach ($searchFields as $fieldName) {\r
+ $subquery = new Zend_Search_Lucene_Search_Query_Preprocessing_Phrase($this->_phrase,\r
+ $this->_phraseEncoding,\r
+ $fieldName);\r
+ $subquery->setSlop($this->getSlop());\r
+\r
+ $query->addSubquery($subquery->rewrite($index));\r
+ }\r
+\r
+ $this->_matches = $query->getQueryTerms();\r
+ return $query;\r
+ }\r
+\r
+ // Recognize exact term matching (it corresponds to Keyword fields stored in the index)\r
+ // encoding is not used since we expect binary matching\r
+ $term = new Zend_Search_Lucene_Index_Term($this->_phrase, $this->_field);\r
+ if ($index->hasTerm($term)) {\r
+ $query = new Zend_Search_Lucene_Search_Query_Term($term);\r
+ $query->setBoost($this->getBoost());\r
+\r
+ $this->_matches = $query->getQueryTerms();\r
+ return $query;\r
+ }\r
+\r
+\r
+ // tokenize phrase using current analyzer and process it as a phrase query\r
+ $tokens = Zend_Search_Lucene_Analysis_Analyzer::getDefault()->tokenize($this->_phrase, $this->_phraseEncoding);\r
+\r
+ if (count($tokens) == 0) {\r
+ $this->_matches = array();\r
+ return new Zend_Search_Lucene_Search_Query_Insignificant();\r
+ }\r
+\r
+ if (count($tokens) == 1) {\r
+ $term = new Zend_Search_Lucene_Index_Term($tokens[0]->getTermText(), $this->_field);\r
+ $query = new Zend_Search_Lucene_Search_Query_Term($term);\r
+ $query->setBoost($this->getBoost());\r
+\r
+ $this->_matches = $query->getQueryTerms();\r
+ return $query;\r
+ }\r
+\r
+ //It's non-trivial phrase query\r
+ $position = -1;\r
+ $query = new Zend_Search_Lucene_Search_Query_Phrase();\r
+ foreach ($tokens as $token) {\r
+ $position += $token->getPositionIncrement();\r
+ $term = new Zend_Search_Lucene_Index_Term($token->getTermText(), $this->_field);\r
+ $query->addTerm($term, $position);\r
+ $query->setSlop($this->getSlop());\r
+ }\r
+ $this->_matches = $query->getQueryTerms();\r
+ return $query;\r
+ }\r
+\r
+ /**\r
+ * Query specific matches highlighting\r
+ *\r
+ * @param Zend_Search_Lucene_Search_Highlighter_Interface $highlighter Highlighter object (also contains doc for highlighting)\r
+ */\r
+ protected function _highlightMatches(Zend_Search_Lucene_Search_Highlighter_Interface $highlighter)\r
+ {\r
+ /** Skip fields detection. We don't need it, since we expect all fields presented in the HTML body and don't differentiate them */\r
+\r
+ /** Skip exact term matching recognition, keyword fields highlighting is not supported */\r
+\r
+ /** Skip wildcard queries recognition. Supported wildcards are removed by text analyzer */\r
+\r
+ // tokenize phrase using current analyzer and process it as a phrase query\r
+ $tokens = Zend_Search_Lucene_Analysis_Analyzer::getDefault()->tokenize($this->_phrase, $this->_phraseEncoding);\r
+\r
+ if (count($tokens) == 0) {\r
+ // Do nothing\r
+ return;\r
+ }\r
+\r
+ if (count($tokens) == 1) {\r
+ $highlighter->highlight($tokens[0]->getTermText());\r
+ return;\r
+ }\r
+\r
+ //It's non-trivial phrase query\r
+ $words = array();\r
+ foreach ($tokens as $token) {\r
+ $words[] = $token->getTermText();\r
+ }\r
+ $highlighter->highlight($words);\r
+ }\r
+\r
+ /**\r
+ * Print a query\r
+ *\r
+ * @return string\r
+ */\r
+ public function __toString()\r
+ {\r
+ // It's used only for query visualisation, so we don't care about characters escaping\r
+ if ($this->_field !== null) {\r
+ $query = $this->_field . ':';\r
+ } else {\r
+ $query = '';\r
+ }\r
+\r
+ $query .= '"' . $this->_phrase . '"';\r
+\r
+ if ($this->_slop != 0) {\r
+ $query .= '~' . $this->_slop;\r
+ }\r
+\r
+ if ($this->getBoost() != 1) {\r
+ $query .= '^' . round($this->getBoost(), 4);\r
+ }\r
+\r
+ return $query;\r
+ }\r
+}\r