|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
/** |
|
4
|
|
|
* This software package is licensed under `AGPL-3.0-only, proprietary` license[s]. |
|
5
|
|
|
* |
|
6
|
|
|
* @package maslosoft/manganel |
|
7
|
|
|
* @license AGPL-3.0-only, proprietary |
|
8
|
|
|
* |
|
9
|
|
|
* @copyright Copyright (c) Peter Maselkowski <[email protected]> |
|
10
|
|
|
* @link https://maslosoft.com/manganel/ |
|
11
|
|
|
*/ |
|
12
|
|
|
|
|
13
|
|
|
namespace Maslosoft\Manganel\Options; |
|
14
|
|
|
|
|
15
|
|
|
use function is_array; |
|
16
|
|
|
use Maslosoft\Addendum\Interfaces\AnnotatedInterface; |
|
17
|
|
|
use Maslosoft\Mangan\Interfaces\DataProviderInterface; |
|
18
|
|
|
use Maslosoft\Manganel\Decorators\QueryBuilder\ConditionDecorator; |
|
19
|
|
|
use Maslosoft\Manganel\Helpers\Strings; |
|
20
|
|
|
use Maslosoft\Manganel\Helpers\TypeNamer; |
|
21
|
|
|
use Maslosoft\Manganel\Manganel; |
|
22
|
|
|
use redirect_interval; |
|
23
|
|
|
use ReflectionClass; |
|
24
|
|
|
|
|
25
|
|
|
/** |
|
26
|
|
|
* Holder for more like options |
|
27
|
|
|
* @package Maslosoft\Manganel\Options |
|
28
|
|
|
*/ |
|
29
|
|
|
class MoreLike |
|
30
|
|
|
{ |
|
31
|
|
|
/** |
|
32
|
|
|
* Models to search from similar |
|
33
|
|
|
* @var AnnotatedInterface[] |
|
34
|
|
|
*/ |
|
35
|
|
|
public $models = []; |
|
36
|
|
|
|
|
37
|
|
|
/** |
|
38
|
|
|
* A list of fields to fetch and analyze the text from. Defaults to the _all field for free text and to all |
|
39
|
|
|
* possible fields for document inputs. |
|
40
|
|
|
* @var array |
|
41
|
|
|
*/ |
|
42
|
|
|
public $fields = []; |
|
43
|
|
|
|
|
44
|
|
|
/** |
|
45
|
|
|
* |
|
46
|
|
|
* The unlike parameter is used in conjunction with like in order not to select terms found in a chosen set of |
|
47
|
|
|
* documents. In other words, we could ask for documents like: "Apple", but unlike: "cake crumble tree". The syntax |
|
48
|
|
|
* is the same as like. |
|
49
|
|
|
* @var array |
|
50
|
|
|
*/ |
|
51
|
|
|
public $unlike = []; |
|
52
|
|
|
|
|
53
|
|
|
/** |
|
54
|
|
|
* Extra text to search for |
|
55
|
|
|
* @var string[] |
|
56
|
|
|
*/ |
|
57
|
|
|
public $texts = []; |
|
58
|
|
|
|
|
59
|
|
|
/** |
|
60
|
|
|
* The maximum number of query terms that will be selected. Increasing this value gives greater accuracy at the |
|
61
|
|
|
* expense of query execution speed. |
|
62
|
|
|
* @var int |
|
63
|
|
|
*/ |
|
64
|
|
|
public $maxQueryTerms = 25; |
|
65
|
|
|
|
|
66
|
|
|
/** |
|
67
|
|
|
* The minimum term frequency below which the terms will be ignored from the input document. |
|
68
|
|
|
* @var int |
|
69
|
|
|
*/ |
|
70
|
|
|
public $minTermFreq = 2; |
|
71
|
|
|
|
|
72
|
|
|
/** |
|
73
|
|
|
* The minimum document frequency below which the terms will be ignored from the input document. |
|
74
|
|
|
* @var int |
|
75
|
|
|
*/ |
|
76
|
|
|
public $minDocFreq = 5; |
|
77
|
|
|
|
|
78
|
|
|
/** |
|
79
|
|
|
* The maximum document frequency above which the terms will be ignored from the input document. This could be |
|
80
|
|
|
* useful in order to ignore highly frequent words such as stop words |
|
81
|
|
|
* @var int |
|
82
|
|
|
*/ |
|
83
|
|
|
public $maxDocFreq = null; |
|
84
|
|
|
|
|
85
|
|
|
/** |
|
86
|
|
|
* The minimum word length below which the terms will be ignored. The old name min_word_len is deprecated. |
|
87
|
|
|
* @var int |
|
88
|
|
|
*/ |
|
89
|
|
|
public $minWordLength = 0; |
|
90
|
|
|
|
|
91
|
|
|
/** |
|
92
|
|
|
* The maximum word length above which the terms will be ignored. The old name max_word_len is deprecated. |
|
93
|
|
|
* @var int |
|
94
|
|
|
*/ |
|
95
|
|
|
public $maxWordLength = null; |
|
96
|
|
|
|
|
97
|
|
|
/** |
|
98
|
|
|
* An array of stop words. Any word in this set is considered "uninteresting" and ignored. If the analyzer allows |
|
99
|
|
|
* for stop words, you might want to tell MLT to explicitly ignore them, as for the purposes of document similarity |
|
100
|
|
|
* it seems reasonable to assume that "a stop word is never interesting". |
|
101
|
|
|
* @var array |
|
102
|
|
|
*/ |
|
103
|
|
|
public $stopWords = []; |
|
104
|
|
|
|
|
105
|
|
|
/** |
|
106
|
|
|
* MoreLikeOptions constructor. |
|
107
|
|
|
* @param AnnotatedInterface|AnnotatedInterface[]|DataProviderInterface|null $models |
|
108
|
|
|
*/ |
|
109
|
|
|
public function __construct($models = null) |
|
110
|
|
|
{ |
|
111
|
|
|
if ($models instanceof DataProviderInterface) |
|
112
|
|
|
{ |
|
113
|
|
|
$this->models = $models->getData(); |
|
114
|
|
|
} |
|
115
|
|
|
elseif ($models instanceof AnnotatedInterface) |
|
116
|
|
|
{ |
|
117
|
|
|
$this->models = [$models]; |
|
118
|
|
|
} |
|
119
|
|
|
elseif (is_array($models)) |
|
120
|
|
|
{ |
|
121
|
|
|
$this->models = $models; |
|
122
|
|
|
} |
|
123
|
|
|
} |
|
124
|
|
|
|
|
125
|
|
|
public function toArray() |
|
126
|
|
|
{ |
|
127
|
|
|
$options = []; |
|
128
|
|
|
$like = []; |
|
129
|
|
|
if (!empty($this->texts)) |
|
130
|
|
|
{ |
|
131
|
|
|
foreach ((array)$this->texts as $text) |
|
132
|
|
|
{ |
|
133
|
|
|
$like[] = $text; |
|
134
|
|
|
} |
|
135
|
|
|
} |
|
136
|
|
|
if (!empty($this->models)) |
|
137
|
|
|
{ |
|
138
|
|
|
foreach ($this->models as $model) |
|
139
|
|
|
{ |
|
140
|
|
|
$like[] = [ |
|
141
|
|
|
'_index' => Manganel::create($model)->index, |
|
142
|
|
|
'_type' => TypeNamer::nameType($model), |
|
143
|
|
|
'_id' => (string)$model->_id |
|
144
|
|
|
]; |
|
145
|
|
|
} |
|
146
|
|
|
} |
|
147
|
|
|
|
|
148
|
|
|
if (!empty($like)) |
|
149
|
|
|
{ |
|
150
|
|
|
$options['like'] = $like; |
|
151
|
|
|
} |
|
152
|
|
|
|
|
153
|
|
|
if (!empty($this->fields)) |
|
154
|
|
|
{ |
|
155
|
|
|
// TODO Must decorate, ie `title` to be `title.en` etc. |
|
156
|
|
|
$options['fields'] = $this->fields; |
|
157
|
|
|
} |
|
158
|
|
|
|
|
159
|
|
|
if(!empty($this->unlike)) |
|
160
|
|
|
{ |
|
161
|
|
|
// TODO Must consider also models array etc. |
|
162
|
|
|
$options['unlike'] = $this->unlike; |
|
163
|
|
|
} |
|
164
|
|
|
|
|
165
|
|
|
$keys = [ |
|
166
|
|
|
'maxQueryTerms', |
|
167
|
|
|
'minTermFreq', |
|
168
|
|
|
'minDocFreq', |
|
169
|
|
|
'maxDocFreq', |
|
170
|
|
|
'minWordLength', |
|
171
|
|
|
'maxWordLength', |
|
172
|
|
|
'stopWords', |
|
173
|
|
|
]; |
|
174
|
|
|
$info = new ReflectionClass($this); |
|
175
|
|
|
$defaults = $info->getDefaultProperties(); |
|
176
|
|
|
|
|
177
|
|
|
foreach ($keys as $name) |
|
178
|
|
|
{ |
|
179
|
|
|
$value = $this->$name; |
|
180
|
|
|
if (null === $value) |
|
181
|
|
|
{ |
|
182
|
|
|
continue; |
|
183
|
|
|
} |
|
184
|
|
|
if (is_array($value) && empty($value)) |
|
185
|
|
|
{ |
|
186
|
|
|
continue; |
|
187
|
|
|
} |
|
188
|
|
|
if ($defaults[$name] === $value) |
|
189
|
|
|
{ |
|
190
|
|
|
continue; |
|
191
|
|
|
} |
|
192
|
|
|
$key = Strings::decamelize($name, '_'); |
|
193
|
|
|
$options[$key] = $value; |
|
194
|
|
|
} |
|
195
|
|
|
return $options; |
|
196
|
|
|
} |
|
197
|
|
|
} |
Maslosoft /
Manganel
