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) 2005-2010 Zend Technologies USA Inc. (http://www.zend.com)

    19  * @license    http://framework.zend.com/license/new-bsd     New BSD License

    20  * @version    $Id: Field.php 20096 2010-01-06 02:05:09Z bkarwin $

    21  */

    22 

    23 

    24 /**

    25  * A field is a section of a Document.  Each field has two parts,

    26  * a name and a value. Values may be free text or they may be atomic

    27  * keywords, which are not further processed. Such keywords may

    28  * be used to represent dates, urls, etc.  Fields are optionally

    29  * stored in the index, so that they may be returned with hits

    30  * on the document.

    31  *

    32  * @category   Zend

    33  * @package    Zend_Search_Lucene

    34  * @subpackage Document

    35  * @copyright  Copyright (c) 2005-2010 Zend Technologies USA Inc. (http://www.zend.com)

    36  * @license    http://framework.zend.com/license/new-bsd     New BSD License

    37  */

    38 class Zend_Search_Lucene_Field

    39 {

    40     /**

    41      * Field name

    42      *

    43      * @var string

    44      */

    45     public $name;

    46 

    47     /**

    48      * Field value

    49      *

    50      * @var boolean

    51      */

    52     public $value;

    53 

    54     /**

    55      * Field is to be stored in the index for return with search hits.

    56      *

    57      * @var boolean

    58      */

    59     public $isStored    = false;

    60 

    61     /**

    62      * Field is to be indexed, so that it may be searched on.

    63      *

    64      * @var boolean

    65      */

    66     public $isIndexed   = true;

    67 

    68     /**

    69      * Field should be tokenized as text prior to indexing.

    70      *

    71      * @var boolean

    72      */

    73     public $isTokenized = true;

    74     /**

    75      * Field is stored as binary.

    76      *

    77      * @var boolean

    78      */

    79     public $isBinary    = false;

    80 

    81     /**

    82      * Field are stored as a term vector

    83      *

    84      * @var boolean

    85      */

    86     public $storeTermVector = false;

    87 

    88     /**

    89      * Field boost factor

    90      * It's not stored directly in the index, but affects on normalization factor

    91      *

    92      * @var float

    93      */

    94     public $boost = 1.0;

    95 

    96     /**

    97      * Field value encoding.

    98      *

    99      * @var string

   100      */

   101     public $encoding;

   102 

   103     /**

   104      * Object constructor

   105      *

   106      * @param string $name

   107      * @param string $value

   108      * @param string $encoding

   109      * @param boolean $isStored

   110      * @param boolean $isIndexed

   111      * @param boolean $isTokenized

   112      * @param boolean $isBinary

   113      */

   114     public function __construct($name, $value, $encoding, $isStored, $isIndexed, $isTokenized, $isBinary = false)

   115     {

   116         $this->name  = $name;

   117         $this->value = $value;

   118 

   119         if (!$isBinary) {

   120             $this->encoding    = $encoding;

   121             $this->isTokenized = $isTokenized;

   122         } else {

   123             $this->encoding    = '';

   124             $this->isTokenized = false;

   125         }

   126 

   127         $this->isStored  = $isStored;

   128         $this->isIndexed = $isIndexed;

   129         $this->isBinary  = $isBinary;

   130 

   131         $this->storeTermVector = false;

   132         $this->boost           = 1.0;

   133     }

   134 

   135 

   136     /**

   137      * Constructs a String-valued Field that is not tokenized, but is indexed

   138      * and stored.  Useful for non-text fields, e.g. date or url.

   139      *

   140      * @param string $name

   141      * @param string $value

   142      * @param string $encoding

   143      * @return Zend_Search_Lucene_Field

   144      */

   145     public static function keyword($name, $value, $encoding = '')

   146     {

   147         return new self($name, $value, $encoding, true, true, false);

   148     }

   149 

   150 

   151     /**

   152      * Constructs a String-valued Field that is not tokenized nor indexed,

   153      * but is stored in the index, for return with hits.

   154      *

   155      * @param string $name

   156      * @param string $value

   157      * @param string $encoding

   158      * @return Zend_Search_Lucene_Field

   159      */

   160     public static function unIndexed($name, $value, $encoding = '')

   161     {

   162         return new self($name, $value, $encoding, true, false, false);

   163     }

   164 

   165 

   166     /**

   167      * Constructs a Binary String valued Field that is not tokenized nor indexed,

   168      * but is stored in the index, for return with hits.

   169      *

   170      * @param string $name

   171      * @param string $value

   172      * @param string $encoding

   173      * @return Zend_Search_Lucene_Field

   174      */

   175     public static function binary($name, $value)

   176     {

   177         return new self($name, $value, '', true, false, false, true);

   178     }

   179 

   180     /**

   181      * Constructs a String-valued Field that is tokenized and indexed,

   182      * and is stored in the index, for return with hits.  Useful for short text

   183      * fields, like "title" or "subject". Term vector will not be stored for this field.

   184      *

   185      * @param string $name

   186      * @param string $value

   187      * @param string $encoding

   188      * @return Zend_Search_Lucene_Field

   189      */

   190     public static function text($name, $value, $encoding = '')

   191     {

   192         return new self($name, $value, $encoding, true, true, true);

   193     }

   194 

   195 

   196     /**

   197      * Constructs a String-valued Field that is tokenized and indexed,

   198      * but that is not stored in the index.

   199      *

   200      * @param string $name

   201      * @param string $value

   202      * @param string $encoding

   203      * @return Zend_Search_Lucene_Field

   204      */

   205     public static function unStored($name, $value, $encoding = '')

   206     {

   207         return new self($name, $value, $encoding, false, true, true);

   208     }

   209 

   210     /**

   211      * Get field value in UTF-8 encoding

   212      *

   213      * @return string

   214      */

   215     public function getUtf8Value()

   216     {

   217         if (strcasecmp($this->encoding, 'utf8' ) == 0  ||

   218             strcasecmp($this->encoding, 'utf-8') == 0 ) {

   219                 return $this->value;

   220         } else {

   221 

   222             return (PHP_OS != 'AIX') ? iconv($this->encoding, 'UTF-8', $this->value) : iconv('ISO8859-1', 'UTF-8', $this->value);

   223         }

   224     }

   225 }

   226