esun3d/app/code/Esun/BlogImport/Model/HtmlProcessor.php

<?php
declare(strict_types=1);

namespace Esun\BlogImport\Model;

use Magento\Framework\Exception\LocalizedException;

class HtmlProcessor
{
    private const ESUN_MEDIA_HOSTS = ['esun3d.com', 'www.esun3d.com'];

    private const MEDIA_EXTENSIONS = [
        'avif', 'bmp', 'gif', 'jpeg', 'jpg', 'm4v', 'mov', 'mp4', 'ogg', 'png', 'svg', 'tif', 'tiff', 'webm', 'webp',
    ];

    /**
     * @return array{html: string, media: string[]}
     * @throws LocalizedException
     */
    public function process(string $htmlPath, string $mediaSubdirectory): array
    {
        $html = file_get_contents($htmlPath);
        if ($html === false || trim($html) === '') {
            throw new LocalizedException(__('HTML file is empty or unreadable: %1', $htmlPath));
        }

        $previous = libxml_use_internal_errors(true);
        $document = new \DOMDocument('1.0', 'UTF-8');
        $loaded = $document->loadHTML(
            '<?xml encoding="UTF-8"><div id="esun-import-root">' . $html . '</div>',
            LIBXML_HTML_NOIMPLIED | LIBXML_HTML_NODEFDTD
        );
        libxml_clear_errors();
        libxml_use_internal_errors($previous);
        if (!$loaded) {
            throw new LocalizedException(__('Unable to parse HTML file: %1', $htmlPath));
        }

        $xpath = new \DOMXPath($document);
        $article = $xpath->query(
            '//*[contains(concat(" ", normalize-space(@class), " "), " blog-article ")]'
        )->item(0);
        $contentRoot = $article instanceof \DOMElement
            ? $article
            : $xpath->query('//*[@id="esun-import-root"]')->item(0);
        if (!$contentRoot instanceof \DOMElement) {
            throw new LocalizedException(__('No importable HTML content found in %1.', $htmlPath));
        }

        $removeQueries = [
            './/h1[contains(concat(" ", normalize-space(@class), " "), " news-title ")]',
            './/*[contains(concat(" ", normalize-space(@class), " "), " news-time ")]',
            './/script',
            './/*[contains(concat(" ", normalize-space(@class), " "), " picker ")]',
            './*[contains(concat(" ", normalize-space(@class), " "), " clear ")]',
        ];
        foreach ($removeQueries as $query) {
            $nodes = iterator_to_array($xpath->query($query, $contentRoot));
            foreach ($nodes as $node) {
                $node->parentNode?->removeChild($node);
            }
        }

        $media = [];
        $mediaReplacements = [];
        foreach (iterator_to_array($xpath->query('.//*', $contentRoot)) as $element) {
            if (!$element instanceof \DOMElement) {
                continue;
            }
            foreach (['src', 'href', 'poster'] as $attribute) {
                if (!$element->hasAttribute($attribute)) {
                    continue;
                }
                $relativeMediaPath = $this->getLegacyMediaPath($element->getAttribute($attribute));
                if ($relativeMediaPath === null) {
                    continue;
                }
                $media[$relativeMediaPath] = $relativeMediaPath;
                $placeholder = sprintf('__ESUN_IMPORTED_MEDIA_%d__', count($mediaReplacements));
                $mediaReplacements[$placeholder] = sprintf(
                    "{{media url='%s/%s'}}",
                    $mediaSubdirectory,
                    str_replace("'", '&#039;', $relativeMediaPath)
                );
                $element->setAttribute($attribute, $placeholder);
            }
        }

        $output = '';
        foreach ($contentRoot->childNodes as $child) {
            $output .= $document->saveHTML($child);
        }
        $output = trim(strtr($output, $mediaReplacements));
        if ($output === '') {
            throw new LocalizedException(__('HTML content became empty after processing: %1', $htmlPath));
        }

        return ['html' => $output, 'media' => array_values($media)];
    }

    private function getLegacyMediaPath(string $url): ?string
    {
        $url = html_entity_decode(trim($url), ENT_QUOTES | ENT_HTML5);
        $path = '';
        if (str_starts_with($url, '/uploads/')) {
            $embeddedUrl = rawurldecode(substr($url, strlen('/uploads/')));
            if (preg_match('#^(?:https?:)?//#iu', $embeddedUrl)) {
                return null;
            }
            $path = $url;
        } elseif (preg_match(
            '#^(?:https?:)?//([^/]+)(/.*)$#iu',
            $url,
            $matches
        )) {
            if (!in_array(strtolower($matches[1]), self::ESUN_MEDIA_HOSTS, true)) {
                return null;
            }
            $path = $matches[2];
        }
        if ($path === '') {
            return null;
        }
        $path = (string)preg_split('/[?#]/', $path, 2)[0];
        $filename = rawurldecode(basename($path));
        if ($filename === '' || $filename === '.' || $filename === '..') {
            return null;
        }
        $extension = strtolower(pathinfo($filename, PATHINFO_EXTENSION));
        if (!in_array($extension, self::MEDIA_EXTENSIONS, true)) {
            return null;
        }

        return $filename;
    }
}