|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
namespace Onoi\Tesa\Tests; |
|
4
|
|
|
|
|
5
|
|
|
use Onoi\Tesa\StopwordAnalyzer; |
|
6
|
|
|
|
|
7
|
|
|
/** |
|
8
|
|
|
* @covers \Onoi\Tesa\StopwordAnalyzer |
|
9
|
|
|
* @group onoi-tesa |
|
10
|
|
|
* |
|
11
|
|
|
* @license GNU GPL v2+ |
|
12
|
|
|
* @since 0.1 |
|
13
|
|
|
* |
|
14
|
|
|
* @author mwjames |
|
15
|
|
|
*/ |
|
16
|
|
|
class StopwordAnalyzerTest extends \PHPUnit_Framework_TestCase { |
|
17
|
|
|
|
|
18
|
|
|
public function testCanConstruct() { |
|
19
|
|
|
|
|
20
|
|
|
$this->assertInstanceOf( |
|
21
|
|
|
'\Onoi\Tesa\StopwordAnalyzer', |
|
22
|
|
|
new StopwordAnalyzer() |
|
23
|
|
|
); |
|
24
|
|
|
} |
|
25
|
|
|
|
|
26
|
|
|
public function testCustomStopwordList() { |
|
27
|
|
|
|
|
28
|
|
|
$instance = new StopwordAnalyzer(); |
|
29
|
|
|
|
|
30
|
|
|
$instance->setCustomStopwordList( |
|
31
|
|
|
array( |
|
32
|
|
|
'zoo' => array( 'Foo' ), |
|
33
|
|
|
'doo' => array( 'bar' ) |
|
34
|
|
|
) |
|
35
|
|
|
); |
|
36
|
|
|
|
|
37
|
|
|
$this->assertTrue( |
|
38
|
|
|
$instance->isStopWord( 'Foo' ) |
|
39
|
|
|
); |
|
40
|
|
|
|
|
41
|
|
|
$this->assertEquals( |
|
42
|
|
|
array( 'zoo', 'doo' ), |
|
43
|
|
|
$instance->getLanguageList() |
|
44
|
|
|
); |
|
45
|
|
|
} |
|
46
|
|
|
|
|
47
|
|
|
/** |
|
48
|
|
|
* @dataProvider defaultListStopWordsProvider |
|
49
|
|
|
*/ |
|
50
|
|
|
public function testUncachedIsWordCheckForDefaultList( $word, $expected ) { |
|
51
|
|
|
|
|
52
|
|
|
$instance = new StopwordAnalyzer(); |
|
53
|
|
|
$instance->loadListByDefaultLanguages(); |
|
54
|
|
|
|
|
55
|
|
|
$this->assertEquals( |
|
56
|
|
|
$expected, |
|
57
|
|
|
$instance->isStopWord( $word ) |
|
58
|
|
|
); |
|
59
|
|
|
} |
|
60
|
|
|
|
|
61
|
|
|
/** |
|
62
|
|
|
* @dataProvider listByLanguageStopWordsProvider |
|
63
|
|
|
*/ |
|
64
|
|
|
public function testListByLanguage( $languageCode, $word, $expected ) { |
|
65
|
|
|
|
|
66
|
|
|
$instance = new StopwordAnalyzer(); |
|
67
|
|
|
$instance->loadListByLanguage( $languageCode ); |
|
68
|
|
|
|
|
69
|
|
|
$this->assertEquals( |
|
70
|
|
|
$expected, |
|
71
|
|
|
$instance->isStopWord( $word ) |
|
72
|
|
|
); |
|
73
|
|
|
|
|
74
|
|
|
$this->assertEquals( |
|
75
|
|
|
(array)$languageCode, |
|
76
|
|
|
$instance->getLanguageList() |
|
77
|
|
|
); |
|
78
|
|
|
} |
|
79
|
|
|
|
|
80
|
|
|
/** |
|
81
|
|
|
* @dataProvider listByLanguageStopWordsProvider |
|
82
|
|
|
*/ |
|
83
|
|
|
public function testListFromCustomLocation( $languageCode, $word, $expected ) { |
|
84
|
|
|
|
|
85
|
|
|
$instance = new StopwordAnalyzer(); |
|
86
|
|
|
$instance->loadListFromCustomLocation( __DIR__ . '/../../../data/stopwords/', $languageCode ); |
|
87
|
|
|
|
|
88
|
|
|
$this->assertEquals( |
|
89
|
|
|
$expected, |
|
90
|
|
|
$instance->isStopWord( $word ) |
|
91
|
|
|
); |
|
92
|
|
|
} |
|
93
|
|
|
|
|
94
|
|
|
public function testTryToLoadUnlistedLanguage() { |
|
95
|
|
|
|
|
96
|
|
|
$instance = new StopwordAnalyzer(); |
|
97
|
|
|
$instance->loadListByLanguage( 'foo' ); |
|
98
|
|
|
|
|
99
|
|
|
$this->assertFalse( |
|
100
|
|
|
$instance->isStopWord( 'Bar' ) |
|
101
|
|
|
); |
|
102
|
|
|
} |
|
103
|
|
|
|
|
104
|
|
|
public function testToLoadFromCache() { |
|
105
|
|
|
|
|
106
|
|
|
$cache = $this->getMockBuilder( '\Onoi\Cache\Cache' ) |
|
107
|
|
|
->disableOriginalConstructor() |
|
108
|
|
|
->setMethods( array( 'contains', 'fetch' ) ) |
|
109
|
|
|
->getMockForAbstractClass(); |
|
110
|
|
|
|
|
111
|
|
|
$cache->expects( $this->once() ) |
|
112
|
|
|
->method( 'contains' ) |
|
113
|
|
|
->with( $this->stringContains( 'onoi:tesa:stopword' ) ) |
|
114
|
|
|
->will( $this->returnValue( true ) ); |
|
115
|
|
|
|
|
116
|
|
|
$cache->expects( $this->once() ) |
|
117
|
|
|
->method( 'fetch' ); |
|
118
|
|
|
|
|
119
|
|
|
$instance = new StopwordAnalyzer( $cache ); |
|
120
|
|
|
$instance->loadListByLanguage( 'foo' ); |
|
121
|
|
|
} |
|
122
|
|
|
|
|
123
|
|
|
public function defaultListStopWordsProvider() { |
|
124
|
|
|
|
|
125
|
|
|
$provider[] = array( |
|
|
|
|
|
|
126
|
|
|
'Foo', |
|
127
|
|
|
false |
|
128
|
|
|
); |
|
129
|
|
|
|
|
130
|
|
|
// en-list |
|
131
|
|
|
$provider[] = array( |
|
132
|
|
|
'about', |
|
133
|
|
|
true |
|
134
|
|
|
); |
|
135
|
|
|
|
|
136
|
|
|
// de-list |
|
137
|
|
|
$provider[] = array( |
|
138
|
|
|
'über', |
|
139
|
|
|
true |
|
140
|
|
|
); |
|
141
|
|
|
|
|
142
|
|
|
// fr-list |
|
143
|
|
|
$provider[] = array( |
|
144
|
|
|
'avoir', |
|
145
|
|
|
true |
|
146
|
|
|
); |
|
147
|
|
|
|
|
148
|
|
|
// es-list |
|
149
|
|
|
$provider[] = array( |
|
150
|
|
|
'aquellos', |
|
151
|
|
|
true |
|
152
|
|
|
); |
|
153
|
|
|
|
|
154
|
|
|
// ja-list (not default) |
|
155
|
|
|
$provider[] = array( |
|
156
|
|
|
'かつて', |
|
157
|
|
|
false |
|
158
|
|
|
); |
|
159
|
|
|
|
|
160
|
|
|
return $provider; |
|
161
|
|
|
} |
|
162
|
|
|
|
|
163
|
|
|
public function listByLanguageStopWordsProvider() { |
|
164
|
|
|
|
|
165
|
|
|
// ja-list |
|
166
|
|
|
$provider[] = array( |
|
|
|
|
|
|
167
|
|
|
'ja', |
|
168
|
|
|
'かつて', |
|
169
|
|
|
true |
|
170
|
|
|
); |
|
171
|
|
|
|
|
172
|
|
|
return $provider; |
|
173
|
|
|
} |
|
174
|
|
|
|
|
175
|
|
|
} |
|
176
|
|
|
|
Adding an explicit array definition is generally preferable to implicit array definition as it guarantees a stable state of the code.
Let’s take a look at an example:
As you can see in this example, the array
$myArrayis initialized the first time when the foreach loop is entered. You can also see that the value of thebarkey is only written conditionally; thus, its value might result from a previous iteration.This might or might not be intended. To make your intention clear, your code more readible and to avoid accidental bugs, we recommend to add an explicit initialization $myArray = array() either outside or inside the foreach loop.