SearchResultAdCrawler::getAll() - Code Metrics - Inspection of "Add some filter" - rfussien/leboncoin-crawler - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — develop ( c88e50...0d5516 )

by Rémi

created 2017-04-07 13:35 UTC

SearchResultAdCrawler::getAll() A

↳ Parent: SearchResultAdCrawler

Complexity

Conditions	1
Paths	1

Size

Total Lines	14
Code Lines	11

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	11
CRAP Score	1

Importance

Changes	1
Bugs	0	Features	0

Metric	Value
c	1
b	0
f	0
dl	0
loc	14
ccs	11
cts	11
cp	1
rs	9.4285
cc	1
eloc	11
nc	1
nop	0
crap	1

<?php

namespace Lbc\Crawler;

use Lbc\Filter\DefaultSanitizer;
use Lbc\Filter\PrixSanitizer;
use Lbc\Parser\AdUrlParser;
use Lbc\Parser\SearchResultUrlParser;
use League\Uri\Schemes\Http;
use Symfony\Component\DomCrawler\Crawler;

/**
 * Class SearchResultAdCrawler
 * @package Lbc\Crawler
 */
class SearchResultAdCrawler extends CrawlerAbstract
{
    /**
     * @var AdUrlParser
     */
    protected $url;

    /**
     * @param $url
     * @return SearchResultUrlParser
     */
    protected function setUrlParser($url)
    {
        $this->url = new AdUrlParser($url);
    }

    /**
     * Return the Ad's ID
     *
     * @return string
     */
    public function getId()
    {
        return $this->url->getId();
    }


    /**
     * Return the title
     *
     * @return string
     */
    public function getTitle()
    {
        return DefaultSanitizer::clean($this->node->filter('h2')->text());
    }

    /**
     * Return the price
     *
     * @return int
     */
    public function getPrice()
    {
        if ($this->node->filter('*[itemprop=price]')->count()) {
            return PrixSanitizer::clean(
                $this->node->filter('*[itemprop=price]')->text()
            );
        }

        return 0;
    }

    /**
     * Return the Ad's URL
     *
     * @return string
     */
    public function getUrl()
    {
        return (string)Http::createFromString($this->url)->withScheme('https');
    }

    /**
     * Return the data and time the ad was created
     *
     * @return string
     */
    public function getCreatedAt()
    {
        return $this->node
            ->filter('*[itemprop=availabilityStarts]')
            ->first()
            ->attr('content');
    }

    /**
     * Return the thumb picture url
     *
     * @return null|string
     */
    public function getThumb()
    {
        $image = $this->node
            ->filter('.item_imagePic .lazyload[data-imgsrc]')
            ->first();

        if (0 === $image->count()) {
            return null;
        }

        $src = $image
            ->attr('data-imgsrc');

        return (string)Http::createFromString($src)->withScheme('https');
    }

    /**
     * Return the number of picture of the ad
     *
     * @return int
     */
    public function getNbImage()
    {
        $node = $this->node->filter('.item_imageNumber');

        return $this->getFieldValue($node, 0, function ($value) {
            return (int)trim($value);
        });
    }

    /**
     * @return mixed
     */
    public function getPlacement()

    {
        $node = $this->node->filter('*[itemprop=availableAtOrFrom]');

        return $this->getFieldValue($node, '', function ($value) {
            return preg_replace('/\s+/', ' ', trim($value));
        });
    }

    /**
     * @return mixed
     */
    public function getType()

    {
        $node = $this->node->filter('*[itemprop=category]');

        return $this->getFieldValue($node, false, function ($value) {
            if ('pro' === preg_replace('/[\s()]+/', '', $value)) {
                return 'pro';
            }

            return 'part';
        });
    }

    /**
     * @return object
     */
    public function getAll()
    {
        return (object)[
            'id'            => $this->getId(),
            'titre'         => $this->getTitle(),
            'prix'          => $this->getPrice(),
            'url'           => $this->getUrl(),
            'created_at'    => $this->getCreatedAt(),
            'images_thumbs' => $this->getThumb(),
            'nb_image'      => $this->getNbImage(),
            'placement'     => $this->getPlacement(),
            'type'          => $this->getType(),
        ];
    }

    /**
     * Return the field's value
     *
     * @param Crawler $node
     * @param mixed $defaultValue
     * @param \Closure $callback
     * @param string $funcName
     * @param string $funcParam
     *
     * @return mixed
     */
    private function getFieldValue(
        Crawler $node,
        $defaultValue,
        $callback,
        $funcName = 'text',
        $funcParam = ''
    ) {
        if ($node->count()) {
            return $callback($node->$funcName($funcParam));
        }

        return $defaultValue;
    }
}


1			<?php
2
3			namespace Lbc\Crawler;
4
5			use Lbc\Filter\DefaultSanitizer;
6			use Lbc\Filter\PrixSanitizer;
7			use Lbc\Parser\AdUrlParser;
8			use Lbc\Parser\SearchResultUrlParser;
9			use League\Uri\Schemes\Http;
10			use Symfony\Component\DomCrawler\Crawler;
11
12			/**
13			* Class SearchResultAdCrawler
14			* @package Lbc\Crawler
15			*/
16			class SearchResultAdCrawler extends CrawlerAbstract
17			{
18			/**
19			* @var AdUrlParser
20			*/
21			protected $url;
22
23			/**
24			* @param $url
25			* @return SearchResultUrlParser
26			*/
27	10		protected function setUrlParser($url)
28			{
29	10		$this->url = new AdUrlParser($url);
30	10		}
31
32			/**
33			* Return the Ad's ID
34			*
35			* @return string
36			*/
37	10		public function getId()
38			{
39	10		return $this->url->getId();
40			}
41
42
43			/**
44			* Return the title
45			*
46			* @return string
47			*/
48	10		public function getTitle()
49			{
50	10		return DefaultSanitizer::clean($this->node->filter('h2')->text());
51			}
52
53			/**
54			* Return the price
55			*
56			* @return int
57			*/
58	10		public function getPrice()
59			{
60	10		if ($this->node->filter('*[itemprop=price]')->count()) {
61	10		return PrixSanitizer::clean(
62	10		$this->node->filter('*[itemprop=price]')->text()
63	10		);
64			}
65
66			return 0;
67			}
68
69			/**
70			* Return the Ad's URL
71			*
72			* @return string
73			*/
74	10		public function getUrl()
75			{
76	10		return (string)Http::createFromString($this->url)->withScheme('https');
77			}
78
79			/**
80			* Return the data and time the ad was created
81			*
82			* @return string
83			*/
84	10		public function getCreatedAt()
85			{
86	10		return $this->node
87	10		->filter('*[itemprop=availabilityStarts]')
88	10		->first()
89	10		->attr('content');
90			}
91
92			/**
93			* Return the thumb picture url
94			*
95			* @return null\|string
96			*/
97	10		public function getThumb()
98			{
99	10		$image = $this->node
100	10		->filter('.item_imagePic .lazyload[data-imgsrc]')
101	10		->first();
102
103	10		if (0 === $image->count()) {
104			return null;
105			}
106
107			$src = $image
108	10		->attr('data-imgsrc');
109
110	10		return (string)Http::createFromString($src)->withScheme('https');
111			}
112
113			/**
114			* Return the number of picture of the ad
115			*
116			* @return int
117			*/
118	10		public function getNbImage()
119			{
120	10		$node = $this->node->filter('.item_imageNumber');
121
122			return $this->getFieldValue($node, 0, function ($value) {
123	10		return (int)trim($value);
124	10		});
125			}
126
127			/**
128			* @return mixed
129			*/
130	10	View Code Duplication	public function getPlacement()
			0 ignored issues – show Duplication introduced 2015-12-21 14:14 UTC by Report Bug Copy Issue Report This method seems to be duplicated in your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
131			{
132	10		$node = $this->node->filter('*[itemprop=availableAtOrFrom]');
133
134			return $this->getFieldValue($node, '', function ($value) {
135	10		return preg_replace('/\s+/', ' ', trim($value));
136	10		});
137			}
138
139			/**
140			* @return mixed
141			*/
142	10	View Code Duplication	public function getType()
			0 ignored issues – show Duplication introduced 2015-12-21 14:14 UTC by Report Bug Copy Issue Report This method seems to be duplicated in your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
143			{
144	10		$node = $this->node->filter('*[itemprop=category]');
145
146	10		return $this->getFieldValue($node, false, function ($value) {
147	10		if ('pro' === preg_replace('/[\s()]+/', '', $value)) {
148	8		return 'pro';
149			}
150
151	10		return 'part';
152	10		});
153			}
154
155			/**
156			* @return object
157			*/
158	10		public function getAll()
159			{
160			return (object)[
161	10		'id' => $this->getId(),
162	10		'titre' => $this->getTitle(),
163	10		'prix' => $this->getPrice(),
164	10		'url' => $this->getUrl(),
165	10		'created_at' => $this->getCreatedAt(),
166	10		'images_thumbs' => $this->getThumb(),
167	10		'nb_image' => $this->getNbImage(),
168	10		'placement' => $this->getPlacement(),
169	10		'type' => $this->getType(),
170	10		];
171			}
172
173			/**
174			* Return the field's value
175			*
176			* @param Crawler $node
177			* @param mixed $defaultValue
178			* @param \Closure $callback
179			* @param string $funcName
180			* @param string $funcParam
181			*
182			* @return mixed
183			*/
184	10		private function getFieldValue(
185			Crawler $node,
186			$defaultValue,
187			$callback,
188			$funcName = 'text',
189			$funcParam = ''
190			) {
191	10		if ($node->count()) {
192	10		return $callback($node->$funcName($funcParam));
193			}
194
195			return $defaultValue;
196			}
197			}
198

rfussien / leboncoin-crawler

Push — develop ( c88e50...0d5516 )

SearchResultAdCrawler::getAll() A

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like