Sanitizer::sanitizeWith() - Code Metrics - Inspection of "Test IcuWordBoundaryTokenizer only on ICU 54.1" - onoi/tesa - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( 2cc7a2...0dc8a0 )

by mw

created 2016-08-02 14:39 UTC

Sanitizer::sanitizeWith() D

↳ Parent: Sanitizer

Complexity

Conditions	10
Paths	10

Size

Total Lines	35
Code Lines	16

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	16
CRAP Score	10.0203

Importance

Changes	1
Bugs	0	Features	1

Metric	Value
c	1
b	0
f	1
dl	0
loc	35
ccs	16
cts	17
cp	0.9412
rs	4.8196
cc	10
eloc	16
nc	10
nop	3
crap	10.0203

How to fix Complexity

<?php

namespace Onoi\Tesa;

use Onoi\Tesa\Tokenizer\Tokenizer;
use Onoi\Tesa\Synonymizer\Synonymizer;
use Onoi\Tesa\StopwordAnalyzer\StopwordAnalyzer;
use RuntimeException;

/**
 * @license GNU GPL v2+
 * @since 0.1
 *
 * @author mwjames
 */
class Sanitizer {

	const WHITELIST = 'WHITELIST';
	const MIN_LENGTH = 'MIN_LENGTH';

	/**
	 * Any change to the content of its data files should be reflected in a
	 * version change (the version number does not necessarily correlate with
	 * the library version)
	 */
	const VERSION = '0.2';

	/**
	 * @var string
	 */
	private $string = '';

	/**
	 * @var array
	 */
	private $whiteList = array();

	/**
	 * @var array
	 */
	private $minLength = 3;

	/**
	 * @since 0.1
	 *
	 * @param string $string
	 */
	public function __construct( $string = '' ) {
		$this->setText( $string );
	}

	/**
	 * @since 1.0
	 *
	 * @param string $name
	 * @param mixed $value
	 */
	public function setOption( $name, $value ) {

		if ( $name === self::WHITELIST && is_array( $value ) && $value !== array() ) {
			$this->whiteList = array_fill_keys( $value, true );
		}

		if ( $name === self::MIN_LENGTH ) {
			$this->minLength = (int)$value;

		}
	}

	/**
	 * @since 0.1
	 *
	 * @param string $string
	 */
	public function setText( $string ) {
		$this->string = $string;
	}

	/**
	 * @since 0.1
	 *
	 * @param integer $flag
	 */
	public function applyTransliteration( $flag = Transliterator::DIACRITICS ) {
		$this->string = Normalizer::applyTransliteration( $this->string, $flag );
	}

	/**
	 * @see Localizer::convertDoubleWidth
	 *
	 * @since 0.1
	 *
	 * @param integer $flag
/**
 * @param array $germany
 * @param array $island
 * @param array $italy
 */
function finale($germany, $island) {
    return "2:1";
}
	 */
	public function convertDoubleWidth() {
		$this->string = Normalizer::convertDoubleWidth( $this->string );
	}

	/**
	 * @since 0.1
	 *
	 * @param Tokenizer $tokenizer
	 * @param StopwordAnalyzer $stopwordAnalyzer
	 *
	 * @return string
	 */
	public function sanitizeWith( Tokenizer $tokenizer, StopwordAnalyzer $stopwordAnalyzer, Synonymizer $synonymizer ) {

		// Treat non-words tokenizers (Ja,Zh*) differently
		$minLength = $tokenizer->isWordTokenizer() ? $this->minLength : 1;

		$words = $tokenizer->tokenize( $this->string );

		if ( !$words || !is_array( $words ) ) {
			return $this->string;
		}

		$index = array();
		$pos = 0;

		foreach ( $words as $key => $word ) {

			$word = $synonymizer->synonymize( $word );

			// If it is not an exemption and less than the required minimum length
			// or identified as stop word it is removed
			if ( !isset( $this->whiteList[$word] ) && ( mb_strlen( $word ) < $minLength || $stopwordAnalyzer->isStopWord( $word ) ) ) {
				continue;
			}

			// Simple proximity, check for same words appearing next to each other
			if ( isset( $index[$pos-1] ) && $index[$pos-1] === $word ) {
				continue;
			}

			$index[] = trim( $word );
			$pos++;
		}

		return implode( ' ' , $index );
	}

	/**
	 * @since 0.1
	 */
	public function toLowercase() {
		$this->string = Normalizer::toLowercase( $this->string );
	}

	/**
	 * @since 0.1
	 *
	 * @param integer $length
	 */
	public function reduceLengthTo( $length ) {
		$this->string = Normalizer::reduceLengthTo( $this->string, $length );
	}

	/**
	 * @see http://www.phpwact.org/php/i18n/utf-8#str_replace
	 * @since 0.1
	 *
	 * @param string $search
	 * @param string $replace
	 */
	public function replace( $search, $replace ) {
		$this->string = str_replace( $search, $replace, $this->string );
	}

	/**
	 * @since 0.1
	 *
	 * @return string
	 */
	public function __toString() {
		return $this->string;
	}

}


1		<?php
2
3		namespace Onoi\Tesa;
4
5		use Onoi\Tesa\Tokenizer\Tokenizer;
6		use Onoi\Tesa\Synonymizer\Synonymizer;
7		use Onoi\Tesa\StopwordAnalyzer\StopwordAnalyzer;
8		use RuntimeException;
9
10		/**
11		* @license GNU GPL v2+
12		* @since 0.1
13		*
14		* @author mwjames
15		*/
16		class Sanitizer {
17
18		const WHITELIST = 'WHITELIST';
19		const MIN_LENGTH = 'MIN_LENGTH';
20
21		/**
22		* Any change to the content of its data files should be reflected in a
23		* version change (the version number does not necessarily correlate with
24		* the library version)
25		*/
26		const VERSION = '0.2';
27
28		/**
29		* @var string
30		*/
31		private $string = '';
32
33		/**
34		* @var array
35		*/
36		private $whiteList = array();
37
38		/**
39		* @var array
40		*/
41		private $minLength = 3;
42
43		/**
44		* @since 0.1
45		*
46		* @param string $string
47		*/
48	13	public function __construct( $string = '' ) {
49	13	$this->setText( $string );
50	13	}
51
52		/**
53		* @since 1.0
54		*
55		* @param string $name
56		* @param mixed $value
57		*/
58	1	public function setOption( $name, $value ) {
59
60	1	if ( $name === self::WHITELIST && is_array( $value ) && $value !== array() ) {
61	1	$this->whiteList = array_fill_keys( $value, true );
62	1	}
63
64	1	if ( $name === self::MIN_LENGTH ) {
65	1	$this->minLength = (int)$value;
		0 ignored issues – show Documentation Bug introduced 2015-12-08 22:51 UTC by Report Bug Copy Issue Report It seems like `(int) $value` of type `integer` is incompatible with the declared type `array` of property `$minLength`. Our type inference engine has found an assignment to a property that is incompatible with the declared type of that property. Either this assignment is in error or the assigned type should be added to the documentation/type hint for that property.. Loading history...
66	1	}
67	1	}
68
69		/**
70		* @since 0.1
71		*
72		* @param string $string
73		*/
74	13	public function setText( $string ) {
75	13	$this->string = $string;
76	13	}
77
78		/**
79		* @since 0.1
80		*
81		* @param integer $flag
82		*/
83	1	public function applyTransliteration( $flag = Transliterator::DIACRITICS ) {
84	1	$this->string = Normalizer::applyTransliteration( $this->string, $flag );
85	1	}
86
87		/**
88		* @see Localizer::convertDoubleWidth
89		*
90		* @since 0.1
91		*
92		* @param integer $flag
		0 ignored issues – show Bug introduced 2016-08-02 14:35 UTC by Report Bug Copy Issue Report There is no parameter named `$flag`. Was it maybe removed? This check looks for PHPDoc comments describing methods or function parameters that do not exist on the corresponding method or function. Consider the following example. The parameter `$italy` is not defined by the method `finale(...)`. /** * @param array $germany * @param array $island * @param array $italy */ function finale($germany, $island) { return "2:1"; } The most likely cause is that the parameter was removed, but the annotation was not. Loading history...
93		*/
94		public function convertDoubleWidth() {
95		$this->string = Normalizer::convertDoubleWidth( $this->string );
96		}
97
98		/**
99		* @since 0.1
100		*
101		* @param Tokenizer $tokenizer
102		* @param StopwordAnalyzer $stopwordAnalyzer
103		*
104		* @return string
105		*/
106	7	public function sanitizeWith( Tokenizer $tokenizer, StopwordAnalyzer $stopwordAnalyzer, Synonymizer $synonymizer ) {
107
108		// Treat non-words tokenizers (Ja,Zh*) differently
109	7	$minLength = $tokenizer->isWordTokenizer() ? $this->minLength : 1;
110
111	7	$words = $tokenizer->tokenize( $this->string );
112
113	7	if ( !$words \|\| !is_array( $words ) ) {
114		return $this->string;
115		}
116
117	7	$index = array();
118	7	$pos = 0;
119
120	7	foreach ( $words as $key => $word ) {
121
122	7	$word = $synonymizer->synonymize( $word );
123
124		// If it is not an exemption and less than the required minimum length
125		// or identified as stop word it is removed
126	7	if ( !isset( $this->whiteList[$word] ) && ( mb_strlen( $word ) < $minLength \|\| $stopwordAnalyzer->isStopWord( $word ) ) ) {
127	6	continue;
128		}
129
130		// Simple proximity, check for same words appearing next to each other
131	7	if ( isset( $index[$pos-1] ) && $index[$pos-1] === $word ) {
132	2	continue;
133		}
134
135	7	$index[] = trim( $word );
136	7	$pos++;
137	7	}
138
139	7	return implode( ' ' , $index );
140		}
141
142		/**
143		* @since 0.1
144		*/
145	5	public function toLowercase() {
146	5	$this->string = Normalizer::toLowercase( $this->string );
147	5	}
148
149		/**
150		* @since 0.1
151		*
152		* @param integer $length
153		*/
154	3	public function reduceLengthTo( $length ) {
155	3	$this->string = Normalizer::reduceLengthTo( $this->string, $length );
156	3	}
157
158		/**
159		* @see http://www.phpwact.org/php/i18n/utf-8#str_replace
160		* @since 0.1
161		*
162		* @param string $search
163		* @param string $replace
164		*/
165	1	public function replace( $search, $replace ) {
166	1	$this->string = str_replace( $search, $replace, $this->string );
167	1	}
168
169		/**
170		* @since 0.1
171		*
172		* @return string
173		*/
174	6	public function __toString() {
175	6	return $this->string;
176		}
177
178		}
179

onoi / tesa

Push — master ( 2cc7a2...0dc8a0 )

Sanitizer::sanitizeWith() D

Complexity

Size

Duplication

Code Coverage

Importance

How to fix Complexity

Long Method

Duplication Side-by-Side

Filter issues like