DefaultFinder::getCurlCallback() - Code Metrics - ArthurHoaro/web-thumbnailer - Measure and Improve Code Quality continuously with Scrutinizer

DefaultFinder::getCurlCallback() C
last analyzed 2022-11-24 19:40 UTC

↳ Parent: DefaultFinder

Complexity

Conditions	17
Paths	1

Size

Total Lines	60
Code Lines	28

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	17
eloc	28
c	0
b	0
f	0
nc	1
nop	2
dl	0
loc	60
rs	5.2166

How to fix Long Method Complexity

<?php

declare(strict_types=1);

namespace WebThumbnailer\Finder;

use WebThumbnailer\Application\ConfigManager;
use WebThumbnailer\Application\WebAccess\WebAccess;
use WebThumbnailer\Application\WebAccess\WebAccessCUrl;
use WebThumbnailer\Application\WebAccess\WebAccessFactory;
use WebThumbnailer\Utils\ImageUtils;
use WebThumbnailer\Utils\UrlUtils;

/**
 * This finder isn't linked to any domain.
 * It will return the resource if it is an image (by extension, or by content).
 * Otherwise, it'll try to retrieve an OpenGraph resource.
 */
class DefaultFinder extends FinderCommon
{
    /** @var WebAccess instance. */
    protected $webAccess;

    /**
     * @inheritdoc
     * @param mixed[]|null $rules   All existing rules loaded from JSON files.
     * @param mixed[]|null $options Options provided by the user to retrieve a thumbnail.
     */
    public function __construct(string $domain, string $url, ?array $rules, ?array $options)
    {
        $this->webAccess = WebAccessFactory::getWebAccess($url);
        $this->url = $url;
        $this->domain = $domain;
    }

    /**
     * Generic finder.
     *
     * @inheritdoc
     */
    public function find()
    {
        if (ImageUtils::isImageExtension(UrlUtils::getUrlFileExtension($this->url))) {
            return $this->url;
        }

        $content = $thumbnail = null;
        $callback = $this->webAccess instanceof WebAccessCUrl
            ? $this->getCurlCallback($content, $thumbnail)
            : null;
        list($headers, $content) = $this->webAccess->getContent(
            $this->url,
            (int) ConfigManager::get('settings.default.timeout', 30),
            (int) ConfigManager::get('settings.default.max_img_dl', 16777216),
            $callback,
            $content
        );

        if (empty($thumbnail) && !empty($content) && ImageUtils::isImageString($content)) {
            return $this->url;
        }

        if (empty($thumbnail) && ! empty($headers) && strpos($headers[0], '200') === false) {
            return false;
        }

        // With curl, the thumb is extracted during the download
        if ($this->webAccess instanceof WebAccessCUrl && ! empty($thumbnail)) {
            return $thumbnail;
        }

        return ! empty($content) ? static::extractMetaTag($content) : false;
    }

    /**
     * Get a callback for curl write function.
     *
     * @param string|null $content   A variable reference in which the downloaded content should be stored.
     * @param string|null $thumbnail A variable reference in which extracted thumb URL should be stored.
     *
     * @return callable CURLOPT_WRITEFUNCTION callback
     */
    protected function getCurlCallback(?string &$content, ?string &$thumbnail): callable
    {
        $url = $this->url;
        $isRedirected = false;

        /**
         * cURL callback function for CURLOPT_WRITEFUNCTION (called during the download).
         *
         * While downloading the remote page, we check that the HTTP code is 200 and content type is 'html/text'
         * Then we extract the title and the charset and stop the download when it's done.
         *
         * Note that when using CURLOPT_WRITEFUNCTION, we have to manually handle the content retrieved,
         * hence the $content reference variable.
         *
         * @param resource $ch   cURL resource
         * @param string   $data chunk of data being downloaded
         *
         * @return int|false length of $data or false if we need to stop the download
         */
        return function ($ch, $data) use ($url, &$content, &$thumbnail, &$isRedirected) {
            $content .= $data;
            $responseCode = curl_getinfo($ch, CURLINFO_RESPONSE_CODE);
            if (!empty($responseCode) && in_array($responseCode, [301, 302])) {
                $isRedirected = true;
                return strlen($data);
            }
            if (!empty($responseCode) && $responseCode !== 200) {
                return false;
            }
            // After a redirection, the content type will keep the previous request value
            // until it finds the next content-type header.
            if (! $isRedirected || strpos(strtolower($data), 'content-type') !== false) {
                $contentType = curl_getinfo($ch, CURLINFO_CONTENT_TYPE);
            }
            // we look for image, and ignore application/octet-stream,
            // which is a the default content type for any binary
            // @see https://developer.mozilla.org/fr/docs/Web/HTTP/Basics_of_HTTP/MIME_types
            if (
                !empty($contentType)
                && strpos($contentType, 'image/') !== false
                && strpos($contentType, 'application/octet-stream') === false
            ) {
                $thumbnail = $url;
                return false;
            } elseif (
                !empty($contentType)
                && strpos($contentType, 'text/html') === false
                && strpos($contentType, 'application/octet-stream') === false
            ) {
                return false;
            }
            if (empty($thumbnail)) {
                $thumbnail = DefaultFinder::extractMetaTag($data);
            }
            // We got everything we want, stop the download.
            if (!empty($responseCode) && !empty($contentType) && !empty($thumbnail)) {
                return false;
            }

            return strlen($data);
        };
    }

    /**
     * Applies the regexp on the HTML $content to extract the thumb URL.
     *
     * @param string $content Downloaded HTML content
     *
     * @return string|false Extracted thumb URL or false if not found.
     */
    public static function extractMetaTag(string $content)
    {
        $propertiesKey = ['property', 'name', 'itemprop'];
        $properties = implode('|', $propertiesKey);
        // Try to retrieve OpenGraph image.
        $ogRegex = '#<meta[^>]+(?:' . $properties . ')=["\']?og:image["\'\s][^>]*content=["\']?(.*?)["\'\s>]#';
        // If the attributes are not in the order property => content (e.g. Github)
        // New regex to keep this readable... more or less.
        $ogRegexReverse = '#<meta[^>]+content=["\']?([^"\'\s]+)[^>]+(?:' . $properties . ')=["\']?og:image["\'\s/>]#';

        if (
            preg_match($ogRegex, $content, $matches) > 0
            || preg_match($ogRegexReverse, $content, $matches) > 0
        ) {
            return $matches[1];
        }

        return false;
    }

    /** @inheritdoc */
    public function isHotlinkAllowed(): bool
    {
        return true;
    }

    /** @inheritdoc */
    public function checkRules(?array $rules): bool
    {
        return true;
    }

    /** @inheritdoc */
    public function loadRules(?array $rules): void
    {
    }

    /** @inheritdoc */
    public function getName(): string
    {
        return 'default';
    }
}


1			<?php
2
3			declare(strict_types=1);
4
5			namespace WebThumbnailer\Finder;
6
7			use WebThumbnailer\Application\ConfigManager;
8			use WebThumbnailer\Application\WebAccess\WebAccess;
9			use WebThumbnailer\Application\WebAccess\WebAccessCUrl;
10			use WebThumbnailer\Application\WebAccess\WebAccessFactory;
11			use WebThumbnailer\Utils\ImageUtils;
12			use WebThumbnailer\Utils\UrlUtils;
13
14			/**
15			* This finder isn't linked to any domain.
16			* It will return the resource if it is an image (by extension, or by content).
17			* Otherwise, it'll try to retrieve an OpenGraph resource.
18			*/
19			class DefaultFinder extends FinderCommon
20			{
21			/** @var WebAccess instance. */
22			protected $webAccess;
23
24			/**
25			* @inheritdoc
26			* @param mixed[]\|null $rules All existing rules loaded from JSON files.
27			* @param mixed[]\|null $options Options provided by the user to retrieve a thumbnail.
28			*/
29			public function __construct(string $domain, string $url, ?array $rules, ?array $options)
30			{
31			$this->webAccess = WebAccessFactory::getWebAccess($url);
32			$this->url = $url;
33			$this->domain = $domain;
34			}
35
36			/**
37			* Generic finder.
38			*
39			* @inheritdoc
40			*/
41			public function find()
42			{
43			if (ImageUtils::isImageExtension(UrlUtils::getUrlFileExtension($this->url))) {
44			return $this->url;
45			}
46
47			$content = $thumbnail = null;
48			$callback = $this->webAccess instanceof WebAccessCUrl
49			? $this->getCurlCallback($content, $thumbnail)
50			: null;
51			list($headers, $content) = $this->webAccess->getContent(
52			$this->url,
53			(int) ConfigManager::get('settings.default.timeout', 30),
54			(int) ConfigManager::get('settings.default.max_img_dl', 16777216),
55			$callback,
56			$content
57			);
58
59			if (empty($thumbnail) && !empty($content) && ImageUtils::isImageString($content)) {
60			return $this->url;
61			}
62
63			if (empty($thumbnail) && ! empty($headers) && strpos($headers[0], '200') === false) {
64			return false;
65			}
66
67			// With curl, the thumb is extracted during the download
68			if ($this->webAccess instanceof WebAccessCUrl && ! empty($thumbnail)) {
69			return $thumbnail;
70			}
71
72			return ! empty($content) ? static::extractMetaTag($content) : false;
73			}
74
75			/**
76			* Get a callback for curl write function.
77			*
78			* @param string\|null $content A variable reference in which the downloaded content should be stored.
79			* @param string\|null $thumbnail A variable reference in which extracted thumb URL should be stored.
80			*
81			* @return callable CURLOPT_WRITEFUNCTION callback
82			*/
83			protected function getCurlCallback(?string &$content, ?string &$thumbnail): callable
84			{
85			$url = $this->url;
86			$isRedirected = false;
87
88			/**
89			* cURL callback function for CURLOPT_WRITEFUNCTION (called during the download).
90			*
91			* While downloading the remote page, we check that the HTTP code is 200 and content type is 'html/text'
92			* Then we extract the title and the charset and stop the download when it's done.
93			*
94			* Note that when using CURLOPT_WRITEFUNCTION, we have to manually handle the content retrieved,
95			* hence the $content reference variable.
96			*
97			* @param resource $ch cURL resource
98			* @param string $data chunk of data being downloaded
99			*
100			* @return int\|false length of $data or false if we need to stop the download
101			*/
102			return function ($ch, $data) use ($url, &$content, &$thumbnail, &$isRedirected) {
103			$content .= $data;
104			$responseCode = curl_getinfo($ch, CURLINFO_RESPONSE_CODE);
105			if (!empty($responseCode) && in_array($responseCode, [301, 302])) {
106			$isRedirected = true;
107			return strlen($data);
108			}
109			if (!empty($responseCode) && $responseCode !== 200) {
110			return false;
111			}
112			// After a redirection, the content type will keep the previous request value
113			// until it finds the next content-type header.
114			if (! $isRedirected \|\| strpos(strtolower($data), 'content-type') !== false) {
115			$contentType = curl_getinfo($ch, CURLINFO_CONTENT_TYPE);
116			}
117			// we look for image, and ignore application/octet-stream,
118			// which is a the default content type for any binary
119			// @see https://developer.mozilla.org/fr/docs/Web/HTTP/Basics_of_HTTP/MIME_types
120			if (
121			!empty($contentType)
122			&& strpos($contentType, 'image/') !== false
123			&& strpos($contentType, 'application/octet-stream') === false
124			) {
125			$thumbnail = $url;
126			return false;
127			} elseif (
128			!empty($contentType)
129			&& strpos($contentType, 'text/html') === false
130			&& strpos($contentType, 'application/octet-stream') === false
131			) {
132			return false;
133			}
134			if (empty($thumbnail)) {
135			$thumbnail = DefaultFinder::extractMetaTag($data);
136			}
137			// We got everything we want, stop the download.
138			if (!empty($responseCode) && !empty($contentType) && !empty($thumbnail)) {
139			return false;
140			}
141
142			return strlen($data);
143			};
144			}
145
146			/**
147			* Applies the regexp on the HTML $content to extract the thumb URL.
148			*
149			* @param string $content Downloaded HTML content
150			*
151			* @return string\|false Extracted thumb URL or false if not found.
152			*/
153			public static function extractMetaTag(string $content)
154			{
155			$propertiesKey = ['property', 'name', 'itemprop'];
156			$properties = implode('\|', $propertiesKey);
157			// Try to retrieve OpenGraph image.
158			$ogRegex = '#<meta[^>]+(?:' . $properties . ')=["\']?og:image["\'\s][^>]content=["\']?(.?)["\'\s>]#';
159			// If the attributes are not in the order property => content (e.g. Github)
160			// New regex to keep this readable... more or less.
161			$ogRegexReverse = '#<meta[^>]+content=["\']?([^"\'\s]+)[^>]+(?:' . $properties . ')=["\']?og:image["\'\s/>]#';
162
163			if (
164			preg_match($ogRegex, $content, $matches) > 0
165			\|\| preg_match($ogRegexReverse, $content, $matches) > 0
166			) {
167			return $matches[1];
168			}
169
170			return false;
171			}
172
173			/** @inheritdoc */
174			public function isHotlinkAllowed(): bool
175			{
176			return true;
177			}
178
179			/** @inheritdoc */
180			public function checkRules(?array $rules): bool
181			{
182			return true;
183			}
184
185			/** @inheritdoc */
186			public function loadRules(?array $rules): void
187			{
188			}
189
190			/** @inheritdoc */
191			public function getName(): string
192			{
193			return 'default';
194			}
195			}
196

ArthurHoaro / web-thumbnailer

DefaultFinder::getCurlCallback() C last analyzed 2022-11-24 19:40 UTC

Complexity

Size

Duplication

Importance

How to fix Long Method Complexity

Long Method

Duplication Side-by-Side

Filter issues like

DefaultFinder::getCurlCallback() C
last analyzed 2022-11-24 19:40 UTC