SimpleParseOpenGraph.php

<?php
/**
 * Project codeigniter-basic-helper
 * Created by PhpStorm
 * User: 713uk13m <dev@nguyenanhung.com>
 * Copyright: 713uk13m <dev@nguyenanhung.com>
 * Date: 20/07/2023
 * Time: 22:56
 */

namespace nguyenanhung\CodeIgniter\BasicHelper;

class SimpleParseOpenGraph extends BaseHelper implements \Iterator
{
    /**
     * There are base schema's based on type, this is just
     * a map so that the schema can be obtained
     *
     */
    public static $TYPES = array(
        'activity' => array('activity', 'sport'),
        'business' => array('bar', 'company', 'cafe', 'hotel', 'restaurant'),
        'group' => array('cause', 'sports_league', 'sports_team'),
        'organization' => array('band', 'government', 'non_profit', 'school', 'university'),
        'person' => array('actor', 'athlete', 'author', 'director', 'musician', 'politician', 'public_figure'),
        'place' => array('city', 'country', 'landmark', 'state_province'),
        'product' => array('album', 'book', 'drink', 'food', 'game', 'movie', 'product', 'song', 'tv_show'),
        'website' => array('blog', 'website'),
    );

    /**
     * Holds all the Open Graph values we've parsed from a page
     *
     */
    private $_values = array();

    /**
     * Fetches a URI and parses it for Open Graph data, returns
     * false on error.
     *
     * @param mixed $URI URI to page to parse for Open Graph data
     * @return SimpleParseOpenGraph|bool
     */
    static public function fetch($URI)
    {
        $curl = curl_init($URI);

        curl_setopt($curl, CURLOPT_FAILONERROR, true);
        curl_setopt($curl, CURLOPT_FOLLOWLOCATION, true);
        curl_setopt($curl, CURLOPT_RETURNTRANSFER, true);
        curl_setopt($curl, CURLOPT_TIMEOUT, 15);
        curl_setopt($curl, CURLOPT_SSL_VERIFYHOST, false);
        curl_setopt($curl, CURLOPT_SSL_VERIFYPEER, false);
        if (isset($_SERVER['HTTP_USER_AGENT'])) {
            curl_setopt($curl, CURLOPT_USERAGENT, $_SERVER['HTTP_USER_AGENT']);
        } else {
            curl_setopt(
                $curl,
                CURLOPT_USERAGENT,
                "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/115.0.0.0 Safari/537.36"
            );
        }


        $response = curl_exec($curl);

        curl_close($curl);

        if (!empty($response)) {
            return self::_parse($response);
        }

        return false;
    }

    /**
     * Parses HTML and extracts Open Graph data, this assumes
     * the document is at least well formed.
     *
     * @param mixed $HTML HTML to parse
     * @return SimpleParseOpenGraph
     */
    static private function _parse($HTML)
    {
        $old_libxml_error = libxml_use_internal_errors(true);

        $doc = new \DOMDocument();
        $doc->loadHTML($HTML);

        libxml_use_internal_errors($old_libxml_error);

        $tags = $doc->getElementsByTagName('meta');
        if (!$tags || $tags->length === 0) {
            return false;
        }

        $page = new self();

        $nonOgDescription = null;

        foreach ($tags as $tag) {
            if ($tag->hasAttribute('property') &&
                mb_strpos($tag->getAttribute('property'), 'og:') === 0) {
                $key = strtr(mb_substr($tag->getAttribute('property'), 3), '-', '_');
                $page->_values[$key] = $tag->getAttribute('content');
            }

            //Added this if loop to retrieve description values from sites like the New York Times who have malformed it.
            if ($tag->hasAttribute('value') && $tag->hasAttribute('property') &&
                mb_strpos($tag->getAttribute('property'), 'og:') === 0) {
                $key = strtr(mb_substr($tag->getAttribute('property'), 3), '-', '_');
                $page->_values[$key] = $tag->getAttribute('value');
            }
            //Based on modifications at https://github.com/bashofmann/opengraph/blob/master/src/OpenGraph/OpenGraph.php
            if ($tag->hasAttribute('name') && $tag->getAttribute('name') === 'description') {
                $nonOgDescription = $tag->getAttribute('content');
            }
        }
        //Based on modifications at https://github.com/bashofmann/opengraph/blob/master/src/OpenGraph/OpenGraph.php
        if (!isset($page->_values['title'])) {
            $titles = $doc->getElementsByTagName('title');
            if ($titles->length > 0) {
                $page->_values['title'] = $titles->item(0)->textContent;
            }
        }
        if (!isset($page->_values['description']) && $nonOgDescription) {
            $page->_values['description'] = $nonOgDescription;
        }

        //Fallback to use image_src if ogp::image isn't set.
        if (!isset($page->values['image'])) {
            $domxpath = new \DOMXPath($doc);
            $elements = $domxpath->query("//link[@rel='image_src']");

            if ($elements->length > 0) {
                $domattr = $elements->item(0)->attributes->getNamedItem('href');
                if ($domattr) {
                    $page->_values['image'] = $domattr->value;
                    $page->_values['image_src'] = $domattr->value;
                }
            }
        }

        if (empty($page->_values)) {
            return false;
        }

        return $page;
    }

    /**
     * Helper method to access attributes directly
     * Example:
     * $graph->title
     *
     * @param mixed $key Key to fetch from the lookup
     */
    public function __get($key)
    {
        if (array_key_exists($key, $this->_values)) {
            return $this->_values[$key];
        }

        if ($key === 'schema') {
            foreach (self::$TYPES as $schema => $types) {
                if (array_search($this->_values['type'], $types)) {
                    return $schema;
                }
            }
        }
    }

    /**
     * Return all the keys found on the page
     *
     * @return array
     */
    public function keys()
    {
        return array_keys($this->_values);
    }

    /**
     * Helper method to check an attribute exists
     *
     * @param mixed $key
     */
    public function __isset($key)
    {
        return array_key_exists($key, $this->_values);
    }

    /**
     * Will return true if the page has location data embedded
     *
     * @return boolean Check if the page has location data
     */
    public function hasLocation()
    {
        if (array_key_exists('latitude', $this->_values) && array_key_exists('longitude', $this->_values)) {
            return true;
        }

        $address_keys = array('street_address', 'locality', 'region', 'postal_code', 'country_name');
        $valid_address = true;
        foreach ($address_keys as $key) {
            $valid_address = ($valid_address && array_key_exists($key, $this->_values));
        }
        return $valid_address;
    }

    /**
     * Iterator code
     */
    private $_position = 0;

    public function rewind()
    {
        reset($this->_values);
        $this->_position = 0;
    }

    public function current()
    {
        return current($this->_values);
    }

    public function key()
    {
        return key($this->_values);
    }

    public function next()
    {
        next($this->_values);
        ++$this->_position;
    }

    public function valid()
    {
        return $this->_position < sizeof($this->_values);
    }
}