HtmlSplitter::split() - Code Metrics - algolia/scout-extended - Measure and Improve Code Quality continuously with Scrutinizer

HtmlSplitter::split() A
last analyzed 2020-03-03 14:38 UTC

↳ Parent: HtmlSplitter

Complexity

Conditions	5
Paths	4

Size

Total Lines	17
Code Lines	9

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	9
CRAP Score	5.025

Importance

Changes	1
Bugs	0	Features	1

Metric	Value
cc	5
eloc	9
c	1
b	0
f	1
nc	4
nop	2
dl	0
loc	17
ccs	9
cts	10
cp	0.9
crap	5.025
rs	9.6111

<?php

declare(strict_types=1);

/**
 * This file is part of Scout Extended.
 *
 * (c) Algolia Team <[email protected]>
 *
 *  For the full copyright and license information, please view the LICENSE
 *  file that was distributed with this source code.
 */

namespace Algolia\ScoutExtended\Splitters;

use Algolia\ScoutExtended\Contracts\SplitterContract;
use DOMDocument;

class HtmlSplitter implements SplitterContract
{
    /**
     * The list of html tags.
     *
     * @var string[]
     */
    protected $tags = [
        'h1',
        'h2',
        'h3',
        'h4',
        'h5',
        'p',
    ];

    /**
     * Creates a new instance of the class.
     *
     * @param array $tags
     *
     * @return void
     */
    public function __construct(array $tags = null)
    {
        if ($tags !== null) {
            $this->tags = $tags;
        }
    }

    /**
     * Acts a static factory.
     *
     * @param  string|array $tags
     *
     * @return static
     */
    public static function by($tags)
    {
        return new static((array) $tags);
    }

    /**
     * Splits the given value.
     *
     * @param  object $searchable
     * @param  string $value
     *
     * @return array
     */
    public function split($searchable, $value): array
    {
        $dom = new DOMDocument();
        $dom->loadHTML($value);
        $values = [];

        foreach ($this->tags as $tag) {
            foreach ($dom->getElementsByTagName($tag) as $node) {
                $values[] = $node->textContent;

                while (($node = $node->nextSibling) && $node->nodeName !== $tag) {
                    $values[] = $node->textContent;
                }
            }
        }

        return $values;
    }
}


1		<?php
2
3		declare(strict_types=1);
4
5		/**
6		* This file is part of Scout Extended.
7		*
8		* (c) Algolia Team <[email protected]>
9		*
10		* For the full copyright and license information, please view the LICENSE
11		* file that was distributed with this source code.
12		*/
13
14		namespace Algolia\ScoutExtended\Splitters;
15
16		use Algolia\ScoutExtended\Contracts\SplitterContract;
17		use DOMDocument;
18
19		class HtmlSplitter implements SplitterContract
20		{
21		/**
22		* The list of html tags.
23		*
24		* @var string[]
25		*/
26		protected $tags = [
27		'h1',
28		'h2',
29		'h3',
30		'h4',
31		'h5',
32		'p',
33		];
34
35		/**
36		* Creates a new instance of the class.
37		*
38		* @param array $tags
39		*
40		* @return void
41		*/
42	3	public function __construct(array $tags = null)
43		{
44	3	if ($tags !== null) {
45	2	$this->tags = $tags;
46		}
47	3	}
48
49		/**
50		* Acts a static factory.
51		*
52		* @param string\|array $tags
53		*
54		* @return static
55		*/
56	2	public static function by($tags)
57		{
58	2	return new static((array) $tags);
59		}
60
61		/**
62		* Splits the given value.
63		*
64		* @param object $searchable
65		* @param string $value
66		*
67		* @return array
68		*/
69	3	public function split($searchable, $value): array
70		{
71	3	$dom = new DOMDocument();
72	3	$dom->loadHTML($value);
73	3	$values = [];
74
75	3	foreach ($this->tags as $tag) {
76	3	foreach ($dom->getElementsByTagName($tag) as $node) {
77	3	$values[] = $node->textContent;
78
79	3	while (($node = $node->nextSibling) && $node->nodeName !== $tag) {
80		$values[] = $node->textContent;
81		}
82		}
83		}
84
85	3	return $values;
86		}
87		}
88

algolia / scout-extended

HtmlSplitter::split() A last analyzed 2020-03-03 14:38 UTC

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like

HtmlSplitter::split() A
last analyzed 2020-03-03 14:38 UTC