esun3d/app/code/Esun/Translation/Model/Html/HtmlTextExtractor.php

<?php

declare(strict_types=1);

namespace Esun\Translation\Model\Html;

use DOMDocument;
use DOMNode;
use DOMText;
use DOMXPath;

class HtmlTextExtractor
{
    private const ROOT_ID = 'esun-translation-root';
    private const DIRECTIVE_TOKEN_PREFIX = '__ESUN_HTML_DIRECTIVE_';

    public function extract(string $html): HtmlFragment
    {
        [$protectedHtml, $directivePlaceholders] = $this->protectMagentoDirectives($html);
        $document = $this->createDocument($protectedHtml);
        $xpath = new DOMXPath($document);
        $query = sprintf(
            '//*[@id="%s"]//text()[normalize-space(.) != "" and not(ancestor::script) and not(ancestor::style) and not(ancestor::code) and not(ancestor::pre)]',
            self::ROOT_ID
        );
        $nodes = $xpath->query($query);
        $textNodes = [];
        $texts = [];
        $index = 0;

        if ($nodes !== false) {
            foreach ($nodes as $node) {
                if (!$node instanceof DOMText) {
                    continue;
                }

                $text = $this->restoreMagentoDirectives($node->nodeValue ?? '', $directivePlaceholders);
                if (trim($text) === '') {
                    continue;
                }

                $id = 'html_' . $index++;
                $textNodes[$id] = $node;
                $texts[$id] = $text;
            }
        }

        $root = $document->getElementById(self::ROOT_ID);
        if (!$root) {
            throw new \RuntimeException('Unable to parse HTML fragment.');
        }

        return new HtmlFragment($document, $root, $textNodes, $texts, $directivePlaceholders);
    }

    private function createDocument(string $html): DOMDocument
    {
        $document = new DOMDocument('1.0', 'UTF-8');
        $wrappedHtml = sprintf(
            '<?xml encoding="UTF-8"><div id="%s">%s</div>',
            self::ROOT_ID,
            $html
        );

        $previous = libxml_use_internal_errors(true);
        $loaded = $document->loadHTML($wrappedHtml, LIBXML_HTML_NODEFDTD | LIBXML_HTML_NOIMPLIED);
        libxml_clear_errors();
        libxml_use_internal_errors($previous);

        if (!$loaded) {
            throw new \RuntimeException('Unable to parse HTML fragment.');
        }

        return $document;
    }

    /**
     * @return array{0: string, 1: array<string, string>}
     */
    private function protectMagentoDirectives(string $html): array
    {
        $placeholders = [];
        $index = 0;
        $protected = preg_replace_callback(
            '/\{\{.*?\}\}/s',
            function (array $matches) use (&$placeholders, &$index): string {
                $token = self::DIRECTIVE_TOKEN_PREFIX . $index++ . '__';
                $placeholders[$token] = $this->normalizeMagentoDirectiveQuotes($matches[0]);

                return $token;
            },
            $html
        );

        return [$protected ?? $html, $placeholders];
    }

    /**
     * @param array<string, string> $placeholders
     */
    private function restoreMagentoDirectives(string $html, array $placeholders): string
    {
        $restored = $placeholders === [] ? $html : strtr($html, $placeholders);

        return $this->normalizeMagentoDirectiveQuotes($restored);
    }

    private function normalizeMagentoDirectiveQuotes(string $html): string
    {
        return preg_replace_callback(
            '/\{\{.*?\}\}/s',
            static function (array $matches): string {
                return str_ireplace(
                    ['&quot;', '&#34;', '&#x22;', '&apos;', '&#39;', '&#x27;'],
                    ['"', '"', '"', "'", "'", "'"],
                    $matches[0]
                );
            },
            $html
        ) ?? $html;
    }
}