IcuWordBoundaryTokenizerTest::stringProvider() - Code Metrics - Inspection of "Add SanitizerFactory" - onoi/tesa - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( 9401a4...2cc7a2 )

by mw

created 2016-08-02 14:22 UTC

IcuWordBoundaryTokenizerTest::stringProvider() B

↳ Parent: IcuWordBoundaryTokenizerTest

Complexity

Conditions	1
Paths	1

Size

Total Lines	88
Code Lines	53

Duplication

Lines	0
Ratio	0 %

Importance

Changes	1
Bugs	0	Features	1

Metric	Value
c	1
b	0
f	1
dl	0
loc	88
rs	8.6012
cc	1
eloc	53
nc	1
nop	0

How to fix Long Method

<?php

namespace Onoi\Tesa\Tests;

use Onoi\Tesa\Tokenizer\IcuWordBoundaryTokenizer;

/**
 * @covers \Onoi\Tesa\Tokenizer\IcuWordBoundaryTokenizer
 * @group onoi-tesa
 *
 * @license GNU GPL v2+
 * @since 0.1
 *
 * @author mwjames
 */
class IcuWordBoundaryTokenizerTest extends \PHPUnit_Framework_TestCase {

	protected function setUp() {
		$instance = new IcuWordBoundaryTokenizer();

		if ( !$instance->isAvailable() ) {
			$this->markTestSkipped( 'ICU extension is not available.' );
		}
	}

	public function testCanConstruct() {

		$this->assertInstanceOf(
			'\Onoi\Tesa\Tokenizer\IcuWordBoundaryTokenizer',
			new IcuWordBoundaryTokenizer()
		);
	}

	/**
	 * @dataProvider stringProvider
	 */
	public function testTokenize( $string, $expected ) {

		$instance = new IcuWordBoundaryTokenizer();

		$this->assertEquals(
			$expected,
			$instance->tokenize( $string )
		);
	}

	public function testSetOption() {

		$tokenizer = $this->getMockBuilder( '\Onoi\Tesa\Tokenizer\Tokenizer' )
			->disableOriginalConstructor()
			->getMockForAbstractClass();

		$tokenizer->expects( $this->once() )
			->method( 'setOption' );

		$instance = new IcuWordBoundaryTokenizer(
			$tokenizer
		);

		$instance->setOption(
			IcuWordBoundaryTokenizer::REGEX_EXEMPTION,
			array( 'Foo' )
		);
	}

	public function testGeneralSetters() {

		$tokenizer = $this->getMockBuilder( '\Onoi\Tesa\Tokenizer\Tokenizer' )
			->disableOriginalConstructor()
			->getMockForAbstractClass();

		$instance = new IcuWordBoundaryTokenizer(
			$tokenizer
		);

		$instance->setLocale( 'en' );
		$instance->setWordTokenizerAttribute( false );

		$this->assertFalse(
			$instance->isWordTokenizer()
		);
	}

	public function stringProvider() {

		$provider[] = array(
foreach ($collection as $item) {
    $myArray['foo'] = $item->getFoo();

    if ($item->hasBar()) {
        $myArray['bar'] = $item->getBar();
    }

    // do something with $myArray
}
			"安全テスト",
			array( '安全', 'テスト' )
		);

		// Would expect 'すもも', 'も', 'もも', 'も', 'もも', 'の', 'うち', '。'
		$provider[] = array(
			"すもももももももものうち。",
			array( 'すもも', 'も', 'も', 'も', 'も', 'も', 'もの', 'うち', '。' )
		);

		$provider[] = array(
			"李も桃も桃のうち。",
			array( '李', 'も', '桃', 'も', '桃', 'の', 'うち', '。' )
		);

		$provider[] = array(
			"إسرائيل",
			array( 'إسرائيل' )
		);

		$provider[] = array(
			"검색엔ㅇㅏ진",
			array( '검색엔', 'ㅇㅏ', '진' )
		);

		$provider[] = array(
			"검색엔ㅇㅏ진1234abcdfrA",
			array( '검색엔', 'ㅇㅏ', '진', '1234abcdfrA' )
		);

		$provider[] = array(
			"1234abcdfrA",
			array( '1234abcdfrA' )
		);

		$provider[] = array(
			"公明執ようなＳＮＳもストーカー行為の対象に",
			array(
				'公明', '執よう','な','ＳＮＳ', 'も',
				'ストーカー', '行為', 'の', '対象', 'に'
			)
		);

		$provider[] = array(
			"公明執",
			array( '公明', '執' )
		);

		$provider[] = array(
			"IQテスト",
			array( 'IQ', 'テスト' )
		);

		$provider[] = array(
			"foo テスト bar",
			array( 'foo', 'テスト', 'bar' )
		);

		$provider[] = array(
			"foo テスト bar 123abc ^&'",
			array( 'foo', 'テスト', 'bar', '123abc', '^', '&', "'" )
		);

		$provider[] = array(
			"was discovered in 1957 and first sold as a medication in 1971",
			array(
				'was', 'discovered', 'in', '1957', 'and',
				'first', 'sold', 'as', 'a', 'medication', 'in', '1971'
			)
		);

		// See JaTinySegmenterTokenizerTest for comparison
		$provider[] = array(
			'日本語の新聞記事であれば文字単位で95%程度の精度で分かち書きが行えます。 ',
			array(
				'日本語', 'の', '新聞', '記事', 'で',
				'あれ', 'ば',	'文字', '単位',
				'で', '95', '%', '程度',
				'の', '精度', 'で', '分かち書き',
				'が', '行', 'え', 'ます', '。'
			)
		);

		return $provider;
	}

}


1			<?php
2
3			namespace Onoi\Tesa\Tests;
4
5			use Onoi\Tesa\Tokenizer\IcuWordBoundaryTokenizer;
6
7			/**
8			* @covers \Onoi\Tesa\Tokenizer\IcuWordBoundaryTokenizer
9			* @group onoi-tesa
10			*
11			* @license GNU GPL v2+
12			* @since 0.1
13			*
14			* @author mwjames
15			*/
16			class IcuWordBoundaryTokenizerTest extends \PHPUnit_Framework_TestCase {
17
18			protected function setUp() {
19			$instance = new IcuWordBoundaryTokenizer();
20
21			if ( !$instance->isAvailable() ) {
22			$this->markTestSkipped( 'ICU extension is not available.' );
23			}
24			}
25
26			public function testCanConstruct() {
27
28			$this->assertInstanceOf(
29			'\Onoi\Tesa\Tokenizer\IcuWordBoundaryTokenizer',
30			new IcuWordBoundaryTokenizer()
31			);
32			}
33
34			/**
35			* @dataProvider stringProvider
36			*/
37			public function testTokenize( $string, $expected ) {
38
39			$instance = new IcuWordBoundaryTokenizer();
40
41			$this->assertEquals(
42			$expected,
43			$instance->tokenize( $string )
44			);
45			}
46
47			public function testSetOption() {
48
49			$tokenizer = $this->getMockBuilder( '\Onoi\Tesa\Tokenizer\Tokenizer' )
50			->disableOriginalConstructor()
51			->getMockForAbstractClass();
52
53			$tokenizer->expects( $this->once() )
54			->method( 'setOption' );
55
56			$instance = new IcuWordBoundaryTokenizer(
57			$tokenizer
58			);
59
60			$instance->setOption(
61			IcuWordBoundaryTokenizer::REGEX_EXEMPTION,
62			array( 'Foo' )
63			);
64			}
65
66			public function testGeneralSetters() {
67
68			$tokenizer = $this->getMockBuilder( '\Onoi\Tesa\Tokenizer\Tokenizer' )
69			->disableOriginalConstructor()
70			->getMockForAbstractClass();
71
72			$instance = new IcuWordBoundaryTokenizer(
73			$tokenizer
74			);
75
76			$instance->setLocale( 'en' );
77			$instance->setWordTokenizerAttribute( false );
78
79			$this->assertFalse(
80			$instance->isWordTokenizer()
81			);
82			}
83
84			public function stringProvider() {
85
86			$provider[] = array(
			0 ignored issues – show Coding Style Comprehensibility introduced 2016-08-02 14:35 UTC by Report Bug Copy Issue Report `$provider` was never initialized. Although not strictly required by PHP, it is generally a good practice to add `$provider = array();` before regardless. Adding an explicit array definition is generally preferable to implicit array definition as it guarantees a stable state of the code. Let’s take a look at an example: foreach ($collection as $item) { $myArray['foo'] = $item->getFoo(); if ($item->hasBar()) { $myArray['bar'] = $item->getBar(); } // do something with $myArray } As you can see in this example, the array `$myArray` is initialized the first time when the foreach loop is entered. You can also see that the value of the `bar` key 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. Loading history...
87			"安全テスト",
88			array( '安全', 'テスト' )
89			);
90
91			// Would expect 'すもも', 'も', 'もも', 'も', 'もも', 'の', 'うち', '。'
92			$provider[] = array(
93			"すもももももももものうち。",
94			array( 'すもも', 'も', 'も', 'も', 'も', 'も', 'もの', 'うち', '。' )
95			);
96
97			$provider[] = array(
98			"李も桃も桃のうち。",
99			array( '李', 'も', '桃', 'も', '桃', 'の', 'うち', '。' )
100			);
101
102			$provider[] = array(
103			"إسرائيل",
104			array( 'إسرائيل' )
105			);
106
107			$provider[] = array(
108			"검색엔ㅇㅏ진",
109			array( '검색엔', 'ㅇㅏ', '진' )
110			);
111
112			$provider[] = array(
113			"검색엔ㅇㅏ진1234abcdfrA",
114			array( '검색엔', 'ㅇㅏ', '진', '1234abcdfrA' )
115			);
116
117			$provider[] = array(
118			"1234abcdfrA",
119			array( '1234abcdfrA' )
120			);
121
122			$provider[] = array(
123			"公明執ようなＳＮＳもストーカー行為の対象に",
124			array(
125			'公明', '執よう','な','ＳＮＳ', 'も',
126			'ストーカー', '行為', 'の', '対象', 'に'
127			)
128			);
129
130			$provider[] = array(
131			"公明執",
132			array( '公明', '執' )
133			);
134
135			$provider[] = array(
136			"IQテスト",
137			array( 'IQ', 'テスト' )
138			);
139
140			$provider[] = array(
141			"foo テスト bar",
142			array( 'foo', 'テスト', 'bar' )
143			);
144
145			$provider[] = array(
146			"foo テスト bar 123abc ^&'",
147			array( 'foo', 'テスト', 'bar', '123abc', '^', '&', "'" )
148			);
149
150			$provider[] = array(
151			"was discovered in 1957 and first sold as a medication in 1971",
152			array(
153			'was', 'discovered', 'in', '1957', 'and',
154			'first', 'sold', 'as', 'a', 'medication', 'in', '1971'
155			)
156			);
157
158			// See JaTinySegmenterTokenizerTest for comparison
159			$provider[] = array(
160			'日本語の新聞記事であれば文字単位で95%程度の精度で分かち書きが行えます。 ',
161			array(
162			'日本語', 'の', '新聞', '記事', 'で',
163			'あれ', 'ば', '文字', '単位',
164			'で', '95', '%', '程度',
165			'の', '精度', 'で', '分かち書き',
166			'が', '行', 'え', 'ます', '。'
167			)
168			);
169
170			return $provider;
171			}
172
173			}
174

onoi / tesa

Push — master ( 9401a4...2cc7a2 )

IcuWordBoundaryTokenizerTest::stringProvider() B

Complexity

Size

Duplication

Importance

How to fix Long Method

Long Method

Duplication Side-by-Side

Filter issues like