| 1 |
lars |
1 |
<?php
|
|
|
2 |
/**
|
|
|
3 |
* Zend Framework
|
|
|
4 |
*
|
|
|
5 |
* LICENSE
|
|
|
6 |
*
|
|
|
7 |
* This source file is subject to version 1.0 of the Zend Framework
|
|
|
8 |
* license, that is bundled with this package in the file LICENSE, and
|
|
|
9 |
* is available through the world-wide-web at the following URL:
|
|
|
10 |
* http://www.zend.com/license/framework/1_0.txt. If you did not receive
|
|
|
11 |
* a copy of the Zend Framework license and are unable to obtain it
|
|
|
12 |
* through the world-wide-web, please send a note to license@zend.com
|
|
|
13 |
* so we can mail you a copy immediately.
|
|
|
14 |
*
|
|
|
15 |
* @package Zend_Search_Lucene
|
|
|
16 |
* @subpackage Analysis
|
|
|
17 |
* @copyright Copyright (c) 2005-2006 Zend Technologies USA Inc. (http://www.zend.com)
|
|
|
18 |
* @license http://www.zend.com/license/framework/1_0.txt Zend Framework License version 1.0
|
|
|
19 |
*/
|
|
|
20 |
|
|
|
21 |
|
|
|
22 |
/** Zend_Search_Lucene_Analysis_Token */
|
|
|
23 |
require_once 'Zend/Search/Lucene/Analysis/Token.php';
|
|
|
24 |
|
|
|
25 |
/** Zend_Search_Lucene_Analysis_Analyzer_Common_Text */
|
|
|
26 |
require_once 'Zend/Search/Lucene/Analysis/Analyzer/Common/Text.php';
|
|
|
27 |
|
|
|
28 |
/** Zend_Search_Lucene_Analysis_Analyzer_Common_Text_CaseInsensitive */
|
|
|
29 |
require_once 'Zend/Search/Lucene/Analysis/Analyzer/Common/Text/CaseInsensitive.php';
|
|
|
30 |
|
|
|
31 |
|
|
|
32 |
|
|
|
33 |
/**
|
|
|
34 |
* An Analyzer is used to analyze text.
|
|
|
35 |
* It thus represents a policy for extracting index terms from text.
|
|
|
36 |
*
|
|
|
37 |
* Note:
|
|
|
38 |
* Lucene Java implementation is oriented to streams. It provides effective work
|
|
|
39 |
* with a huge documents (more then 20Mb).
|
|
|
40 |
* But engine itself is not oriented such documents.
|
|
|
41 |
* Thus Zend_Search_Lucene analysis API works with data strings and sets (arrays).
|
|
|
42 |
*
|
|
|
43 |
* @package Zend_Search_Lucene
|
|
|
44 |
* @subpackage Analysis
|
|
|
45 |
* @copyright Copyright (c) 2005-2006 Zend Technologies USA Inc. (http://www.zend.com)
|
|
|
46 |
* @license http://www.zend.com/license/framework/1_0.txt Zend Framework License version 1.0
|
|
|
47 |
*/
|
|
|
48 |
|
|
|
49 |
abstract class Zend_Search_Lucene_Analysis_Analyzer
|
|
|
50 |
{
|
|
|
51 |
/**
|
|
|
52 |
* The Analyzer implementation used by default.
|
|
|
53 |
*
|
|
|
54 |
* @var Zend_Search_Lucene_Analysis_Analyzer
|
|
|
55 |
*/
|
|
|
56 |
static private $_defaultImpl;
|
|
|
57 |
|
|
|
58 |
/**
|
|
|
59 |
* Tokenize text to a terms
|
|
|
60 |
* Returns array of Zend_Search_Lucene_Analysis_Token objects
|
|
|
61 |
*
|
|
|
62 |
* @param string $data
|
|
|
63 |
* @return array
|
|
|
64 |
*/
|
|
|
65 |
abstract public function tokenize($data);
|
|
|
66 |
|
|
|
67 |
|
|
|
68 |
/**
|
|
|
69 |
* Set the default Analyzer implementation used by indexing code.
|
|
|
70 |
*
|
|
|
71 |
* @param Zend_Search_Lucene_Analysis_Analyzer $similarity
|
|
|
72 |
*/
|
|
|
73 |
static public function setDefault(Zend_Search_Lucene_Analysis_Analyzer $analyzer)
|
|
|
74 |
{
|
|
|
75 |
self::$_defaultImpl = $analyzer;
|
|
|
76 |
}
|
|
|
77 |
|
|
|
78 |
|
|
|
79 |
/**
|
|
|
80 |
* Return the default Analyzer implementation used by indexing code.
|
|
|
81 |
*
|
|
|
82 |
* @return Zend_Search_Lucene_Analysis_Analyzer
|
|
|
83 |
*/
|
|
|
84 |
static public function getDefault()
|
|
|
85 |
{
|
|
|
86 |
if (!self::$_defaultImpl instanceof Zend_Search_Lucene_Analysis_Analyzer) {
|
|
|
87 |
self::$_defaultImpl = new Zend_Search_Lucene_Analysis_Analyzer_Common_Text_CaseInsensitive();
|
|
|
88 |
}
|
|
|
89 |
|
|
|
90 |
return self::$_defaultImpl;
|
|
|
91 |
}
|
|
|
92 |
|
|
|
93 |
}
|
|
|
94 |
|