htmldocument.php

<?php
 
namespace Bitrix\Main\UrlPreview;
 
use Bitrix\Main\Context;
use Bitrix\Main\Text\Encoding;
use Bitrix\Main\Type\DateTime;
use Bitrix\Main\Web\HttpClient;
use Bitrix\Main\Web\Uri;
use Bitrix\Main\Web\MimeType;
 
class HtmlDocument
{
    const MAX_IMAGES = 4;
    const MAX_IMAGE_URL_LENGTH = 2000;
    const MAX_HTML_LENGTH = 1048576; // 1 MB
 
    protected $uri;
 
    protected $html;
 
    protected $htmlEncoding;
 
    protected $metadata = array(
        "TITLE" => null,
        "DESCRIPTION" => null,
        "IMAGE" => null,
        "EMBED" => null,
        "DATE_EXPIRE" => null,
    );
 
    protected $metaElements = array();
 
    protected $linkElements = array();
 
    public function __construct($html, Uri $uri)
    {
        $this->html = substr($html, 0, self::MAX_HTML_LENGTH);
        $this->uri = $uri;
    }
 
    public function getUri()
    {
        return $this->uri;
    }
 
    public function getHtml()
    {
        return $this->html;
    }
 
    public function checkMetadata()
    {
        $result = (    $this->metadata['TITLE'] != ''
                    && $this->metadata['DESCRIPTION'] != ''
                    && $this->metadata['IMAGE'] != '');
 
        if ($this->isEmbeddingAllowed())
        {
            $result = $result && $this->metadata['EMBED'] != '';
        }
 
        return $result;
    }
 
    public function getMetadata()
    {
        return $this->metadata;
    }
 
    public function getTitle()
    {
        return $this->metadata['TITLE'];
    }
 
    public function setTitle($title)
    {
        if ($title <> '')
        {
            $this->metadata['TITLE'] = $this->filterString($title);
        }
    }
 
    public function getDescription()
    {
        return $this->metadata['DESCRIPTION'];
    }
 
    public function setDescription($description)
    {
        if ($description <> '')
        {
            $this->metadata['DESCRIPTION'] = $this->filterString($description);
        }
    }
 
    public function getImage()
    {
        return $this->metadata['IMAGE'];
    }
 
    public function setImage($image)
    {
        if ($image <> '')
        {
            $imageUrl = $this->normalizeImageUrl($image);
            if (!is_null($imageUrl) && $this->validateImage($imageUrl, true))
            {
                $this->metadata['IMAGE'] = $imageUrl;
            }
        }
    }
 
    public function getEmdbed()
    {
        return $this->metadata['EMBED'];
    }
 
    public function setEmbed($embed)
    {
        if ($this->isEmbeddingAllowed())
        {
            $this->metadata['EMBED'] = $embed;
        }
    }
 
    public function setExtraField($fieldName, $fieldValue)
    {
        if ($fieldName == 'FAVICON')
        {
            $this->metadata['EXTRA'][$fieldName] = $this->convertRelativeUriToAbsolute($fieldValue);
        }
        elseif ($fieldName == 'IMAGES')
        {
            if (is_array($fieldValue))
            {
                $this->metadata['EXTRA']['IMAGES'] = array();
                foreach($fieldValue as $image)
                {
                    $image = $this->normalizeImageUrl($image);
                    if ($image)
                    {
                        $this->metadata['EXTRA']['IMAGES'][] = $image;
                    }
 
                    if (count($this->metadata['EXTRA']['IMAGES']) >= self::MAX_IMAGES)
                    {
                        break;
                    }
                }
            }
        }
        else
        {
            $this->metadata['EXTRA'][$fieldName] = $this->filterString($fieldValue);
        }
    }
 
    public function getExtraField($fieldName)
    {
        return $this->metadata['EXTRA'][$fieldName] ?? null;
    }
 
    public function setDateExpire(DateTime $dateExpire)
    {
        if (!isset($this->metadata['DATE_EXPIRE']) || $this->metadata['DATE_EXPIRE']->getTimestamp() > $dateExpire->getTimestamp())
        {
            $this->metadata['DATE_EXPIRE'] = $dateExpire;
        }
    }
 
    public function getDateExpire(): ?DateTime
    {
        return $this->metadata['DATE_EXPIRE'];
    }
 
    public function setEncoding($encoding)
    {
        $encoding = trim($encoding, " \t\n\r\0\x0B'\"");
        $this->htmlEncoding = $encoding;
    }
 
    public function getEncoding()
    {
        if ($this->htmlEncoding <> '')
        {
            return $this->htmlEncoding;
        }
 
        $this->htmlEncoding = $this->detectEncoding();
        return $this->htmlEncoding;
    }
 
    public function detectEncoding()
    {
        $result = '';
        if (empty($this->metaElements))
        {
            $this->metaElements = $this->extractElementAttributes('meta');
        }
 
        foreach($this->metaElements as $metaElement)
        {
            if (isset($metaElement['http-equiv']) && mb_strtolower($metaElement['http-equiv']) == 'content-type')
            {
                if (preg_match('/charset=([\w\-]+)/', $metaElement['content'], $matches))
                {
                    $result = $matches[1];
                    break;
                }
            }
            elseif (isset($metaElement['charset']))
            {
                $result = $metaElement['charset'];
                break;
            }
        }
 
        return $result;
    }
 
    public function extractElementAttributes($tagName)
    {
        $results = array();
        preg_match_all("/<$tagName.+?>/mis", $this->html, $elements);
 
        foreach($elements[0] as $element)
        {
            preg_match_all('/(?:([\w\-_]+)=([\'"])(.*?)\g{-2}\s*)/mis', $element, $matches);
 
            $elementAttributes = array();
            foreach($matches[1] as $k => $attributeName)
            {
                $attributeName = mb_strtolower($attributeName);
                $attributeValue = $matches[3][$k];
                $elementAttributes[$attributeName] = $attributeValue;
            }
 
            $results[] = $elementAttributes;
        }
 
        return $results;
    }
 
    public function getMetaContent($name)
    {
        if (empty($this->metaElements))
        {
            $this->metaElements = $this->extractElementAttributes('meta');
        }
        $name = mb_strtolower($name);
 
        foreach ($this->metaElements as $metaElement)
        {
            if ((isset($metaElement['name']) && mb_strtolower($metaElement['name']) === $name
                || isset($metaElement['property']) && mb_strtolower($metaElement['property']) === $name)
                && $metaElement['content'] <> '')
            {
                return $metaElement['content'];
            }
        }
 
        return null;
    }
 
    public function getLinkHref($rel)
    {
        if (empty($this->linkElements))
        {
            $this->linkElements = $this->extractElementAttributes('link');
        }
        $rel = mb_strtolower($rel);
 
        foreach ($this->linkElements as $linkElement)
        {
            if (isset($linkElement['rel'])
                && mb_strtolower($linkElement['rel']) == $rel
                && $linkElement['href'] <> '')
            {
                return $linkElement['href'];
            }
        }
 
        return null;
    }
 
    protected function filterString($str)
    {
        $str = html_entity_decode($str, ENT_QUOTES, $this->getEncoding());
        $str = Encoding::convertEncoding($str, $this->getEncoding(), Context::getCurrent()->getCulture()->getCharset());
        $str = trim($str);
        $str = strip_tags($str);
 
        return $str;
    }
 
    protected function convertRelativeUriToAbsolute($uri)
    {
        if (strpos($uri, '//') === 0)
        {
            $uri = $this->uri->getScheme().":".$uri;
        }
 
        if (preg_match('#^https?://#', $uri))
        {
            return $uri;
        }
 
        $pars = parse_url($uri);
        if ($pars === false)
        {
            return null;
        }
 
        if (isset($pars['host']))
        {
            $result = $uri;
        }
        elseif (isset($pars['path']))
        {
            if (mb_substr($pars['path'], 0, 1) !== '/')
            {
                $pathPrefix = preg_replace('/^(.+?)([^\/]*)$/', '$1', $this->uri->getPath());
                $pars['path'] = $pathPrefix.$pars['path'];
            }
 
            $uriPort = '';
            if ($this->uri->getScheme() === 'http' && $this->uri->getPort() != '80'
                || $this->uri->getScheme() === 'https' && $this->uri->getPort() != '443')
            {
                $uriPort = ':'.$this->uri->getPort();
            }
 
            $result = $this->uri->getScheme().'://'
                .$this->uri->getHost()
                .$uriPort
                .$pars['path']
                .(isset($pars['query']) ? '?'.$pars['query'] : '')
                .(isset($pars['fragment']) ? '#'.$pars['fragment'] : '');
        }
        else
        {
            $result = null;
        }
 
        return $result;
    }
 
    protected function normalizeImageUrl($url): ?string
    {
        $url = $this->convertRelativeUriToAbsolute($url);
        if (mb_strlen($url) > self::MAX_IMAGE_URL_LENGTH)
        {
            $url = null;
        }
        return $url;
    }
 
    protected function validateImage($url, $skipForPrivateIp = false)
    {
        $httpClient = new HttpClient();
        $httpClient->setTimeout(5);
        $httpClient->setStreamTimeout(5);
        $httpClient->setPrivateIp(false);
        $httpClient->setHeader('User-Agent', UrlPreview::USER_AGENT);
 
        if (!$httpClient->query('HEAD', $url))
        {
            $errorCode = array_key_first($httpClient->getError());
            return ($skipForPrivateIp && $errorCode === 'PRIVATE_IP');
        }
 
        if ($httpClient->getStatus() !== 200)
        {
            return false;
        }
 
        $contentType = $httpClient->getHeaders()->getContentType();
 
        return MimeType::isImage($contentType);
    }
 
    protected function isEmbeddingAllowed()
    {
        return UrlPreview::isHostTrusted($this->uri);
    }
}