search/Zend/Search/Lucene/Field.php

   1 <?php
   2 /**
   3  * Zend Framework
   4  *
   5  * LICENSE
   6  *
   7  * This source file is subject to the new BSD license that is bundled
   8  * with this package in the file LICENSE.txt.
   9  * It is also available through the world-wide-web at this URL:
  10  * http://framework.zend.com/license/new-bsd
  11  * If you did not receive a copy of the license and are unable to
  12  * obtain it through the world-wide-web, please send an email
  13  * to license@zend.com so we can send you a copy immediately.
  14  *
  15  * @category   Zend
  16  * @package    Zend_Search_Lucene
  17  * @subpackage Document
  18  * @copyright  Copyright (c) 2005-2007 Zend Technologies USA Inc. (http://www.zend.com)
  19  * @license    http://framework.zend.com/license/new-bsd     New BSD License
  20  */
  21
  22
  23 /**
  24  * A field is a section of a Document.  Each field has two parts,
  25  * a name and a value. Values may be free text or they may be atomic
  26  * keywords, which are not further processed. Such keywords may
  27  * be used to represent dates, urls, etc.  Fields are optionally
  28  * stored in the index, so that they may be returned with hits
  29  * on the document.
  30  *
  31  * @category   Zend
  32  * @package    Zend_Search_Lucene
  33  * @subpackage Document
  34  * @copyright  Copyright (c) 2005-2007 Zend Technologies USA Inc. (http://www.zend.com)
  35  * @license    http://framework.zend.com/license/new-bsd     New BSD License
  36  */
  37 class Zend_Search_Lucene_Field
  38 {
  39     /**
  40      * Field name
  41      *
  42      * @var string
  43      */
  44     public $name;
  45
  46
  47     public $value;
  48     public $isStored    = false;
  49     public $isIndexed   = true;
  50     public $isTokenized = true;
  51     public $isBinary    = false;
  52
  53     public $storeTermVector = false;
  54
  55     /**
  56      * Field boos factor
  57      * It's not stored directly in the index, but affects on normalizetion factor
  58      *
  59      * @var float
  60      */
  61     public $boost = 1.0;
  62
  63     /**
  64      * Field value encoding.
  65      *
  66      * @var string
  67      */
  68     public $encoding;
  69
  70     /**
  71      * Object constructor
  72      *
  73      * @param string $name
  74      * @param string $value
  75      * @param string $encoding
  76      * @param boolean $isStored
  77      * @param boolean $isIndexed
  78      * @param boolean $isTokenized
  79      * @param boolean $isBinary
  80      */
  81     public function __construct($name, $value, $encoding, $isStored, $isIndexed, $isTokenized, $isBinary = false)
  82     {
  83         $this->name  = $name;
  84         $this->value = $value;
  85
  86         if (!$isBinary) {
  87             $this->encoding    = $encoding;
  88             $this->isTokenized = $isTokenized;
  89         } else {
  90             $this->encoding    = '';
  91             $this->isTokenized = false;
  92         }
  93
  94         $this->isStored  = $isStored;
  95         $this->isIndexed = $isIndexed;
  96         $this->isBinary  = $isBinary;
  97
  98         $this->storeTermVector = false;
  99         $this->boost           = 1.0;
 100     }
 101
 102
 103     /**
 104      * Constructs a String-valued Field that is not tokenized, but is indexed
 105      * and stored.  Useful for non-text fields, e.g. date or url.
 106      *
 107      * @param string $name
 108      * @param string $value
 109      * @param string $encoding
 110      * @return Zend_Search_Lucene_Field
 111      */
 112     public static function Keyword($name, $value, $encoding = '')
 113     {
 114         return new self($name, $value, $encoding, true, true, false);
 115     }
 116
 117
 118     /**
 119      * Constructs a String-valued Field that is not tokenized nor indexed,
 120      * but is stored in the index, for return with hits.
 121      *
 122      * @param string $name
 123      * @param string $value
 124      * @param string $encoding
 125      * @return Zend_Search_Lucene_Field
 126      */
 127     public static function UnIndexed($name, $value, $encoding = '')
 128     {
 129         return new self($name, $value, $encoding, true, false, false);
 130     }
 131
 132
 133     /**
 134      * Constructs a Binary String valued Field that is not tokenized nor indexed,
 135      * but is stored in the index, for return with hits.
 136      *
 137      * @param string $name
 138      * @param string $value
 139      * @param string $encoding
 140      * @return Zend_Search_Lucene_Field
 141      */
 142     public static function Binary($name, $value)
 143     {
 144         return new self($name, $value, '', true, false, false, true);
 145     }
 146
 147     /**
 148      * Constructs a String-valued Field that is tokenized and indexed,
 149      * and is stored in the index, for return with hits.  Useful for short text
 150      * fields, like "title" or "subject". Term vector will not be stored for this field.
 151      *
 152      * @param string $name
 153      * @param string $value
 154      * @param string $encoding
 155      * @return Zend_Search_Lucene_Field
 156      */
 157     public static function Text($name, $value, $encoding = '')
 158     {
 159         return new self($name, $value, $encoding, true, true, true);
 160     }
 161
 162
 163     /**
 164      * Constructs a String-valued Field that is tokenized and indexed,
 165      * but that is not stored in the index.
 166      *
 167      * @param string $name
 168      * @param string $value
 169      * @param string $encoding
 170      * @return Zend_Search_Lucene_Field
 171      */
 172     public static function UnStored($name, $value, $encoding = '')
 173     {
 174         return new self($name, $value, $encoding, false, true, true);
 175     }
 176
 177     /**
 178      * Get field value in UTF-8 encoding
 179      *
 180      * @return string
 181      */
 182     public function getUtf8Value()
 183     {
 184         if (strcasecmp($this->encoding, 'utf8' ) == 0  ||
 185             strcasecmp($this->encoding, 'utf-8') == 0 ) {
 186                 return $this->value;
 187         } else {
 188             return iconv($this->encoding, 'UTF-8', $this->value);
 189         }
 190     }
 191 }
 192