DefaultFinder::getCurlCallback() - Code Metrics - Inspection of "Use cURL callback to only partially download remot..." - ArthurHoaro/web-thumbnailer - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Pull Request — master (#2)

by Arthur

created 2018-04-18 18:06 UTC

DefaultFinder::getCurlCallback() C

↳ Parent: DefaultFinder

Complexity

Conditions	13
Paths	1

Size

Total Lines	48
Code Lines	21

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	13
eloc	21
nc	1
nop	2
dl	0
loc	48
rs	5.0877
c	0
b	0
f	0

How to fix Complexity

<?php

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;

/**
 * Class DefaultFinder
 *
 * 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.
 *
 * @package WebThumbnailer\Finder
 */
class DefaultFinder extends FinderCommon
{
    /**
     * @var WebAccess instance.
     */
    protected $webAccess;

    /**
     * @inheritdoc
     */
    public function __construct($domain, $url, $rules, $options)
    {
        $this->webAccess = WebAccessFactory::getWebAccess($url);
        $this->url = $url;
        $this->domains = $domain;
    }

    /**
     * Generic finder.
     *
     * @inheritdoc
     */
    function find()

    {
        if (ImageUtils::isImageExtension(UrlUtils::getUrlFileExtension($this->url))) {
        if (ImageUtils::isImageExtension(/** @scrutinizer ignore-type */ UrlUtils::getUrlFileExtension($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,
            ConfigManager::get('settings.default.timeout', 30),
            /** @scrutinizer ignore-type */ ConfigManager::get('settings.default.timeout', 30),
            ConfigManager::get('settings.default.max_img_dl', 16777216),
            /** @scrutinizer ignore-type */ ConfigManager::get('settings.default.max_img_dl', 16777216),
            $callback,
            $content
        );

        if (empty($thumbnail) && ImageUtils::isImageString($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($content) ? self::extractMetaTag($content) : false;
    }

    /**
     * Get a callback for curl write function.
     *
     * @param string $content   A variable reference in which the downloaded content should be stored.
     * @param string $thumbnail A variable reference in which extracted thumb URL should be stored.
     *
     * @return \Closure CURLOPT_WRITEFUNCTION callback
     */
    protected function getCurlCallback(&$content, &$thumbnail)
    {
        $url = $this->url;
        /**
         * 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|bool length of $data or false if we need to stop the download
         */
        return function(&$ch, $data) use ($url, &$content, &$thumbnail) {
            $content .= $data;
            $responseCode = curl_getinfo($ch, CURLINFO_RESPONSE_CODE);
            if (!empty($responseCode) && $responseCode != 200) {
                return 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 = $url;
                return false;
            } else if (!empty($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 $content Downloaded HTML content
     *
     * @return string|bool Extracted thumb URL or false if not found.
     */
    public static function extractMetaTag($content)
    {
        $propertiesKey = ['property', 'name', 'itemprop'];
        // Try to retrieve OpenGraph image.
        $ogRegex = '#<meta[^>]+(?:'. implode('|', $propertiesKey) .')=["\']?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]+)[^>]+(?:'. implode('|', $propertiesKey) .')=["\']?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()
    {
        return true;
    }

    /**
     * @inheritdoc
     */
    function checkRules($rules)

    {
    }

    /**
     * @inheritdoc
     */
    function loadRules($rules)

    {
    }

    /**
     * @inheritdoc
     */
    function getName()

    {
        return 'default';
    }
}


ArthurHoaro / web-thumbnailer

Pull Request — master (#2)

DefaultFinder::getCurlCallback() C

Complexity

Size

Duplication

Importance

How to fix Complexity

Long Method

Duplication Side-by-Side

Filter issues like