KantSites/src/Content/SiteSettings.php

<?php

namespace App\Content;

use App\Domain\InvalidPage;
use Doctrine\DBAL\Connection;

/**
 * Company data and site wide switches. The structured data the pages emit is built from exactly
 * these values, so what a search engine reads is what an editor maintains here – never a separate,
 * invisible copy.
 */
final class SiteSettings
{
    public const DEFAULTS = [
        'legalName' => '',
        'description' => '',
        'phone' => '',
        'email' => '',
        'street' => '',
        'postalCode' => '',
        'city' => '',
        'country' => 'DE',
        'sameAs' => [],
        'localBusiness' => false,
        'llmsTxt' => false,
        'robotsExtra' => '',
        // Which stored page answers at the site root. Everything else derives from this.
        'homePath' => 'home',
    ];

    public function __construct(private Connection $db)
    {
    }

    public function load(): array
    {
        $raw = $this->db->fetchOne('SELECT settings FROM website WHERE id = ?', ['default']);
        $stored = is_string($raw) ? json_decode($raw, true, 512, JSON_THROW_ON_ERROR) : [];

        return [...self::DEFAULTS, ...(is_array($stored) ? $stored : [])];
    }

    public function save(array $input): void
    {
        $values = self::DEFAULTS;
        foreach (['legalName' => 200, 'description' => 300, 'phone' => 40, 'email' => 120, 'street' => 120, 'postalCode' => 16, 'city' => 120] as $key => $max) {
            $values[$key] = mb_substr(trim((string) ($input[$key] ?? '')), 0, $max);
        }
        if ('' !== $values['email'] && false === filter_var($values['email'], FILTER_VALIDATE_EMAIL)) {
            throw new InvalidPage('Bitte gib eine gültige E-Mail-Adresse an.');
        }
        if ('' !== $values['phone'] && 1 !== preg_match('/^\+?[0-9 ()\/-]{3,32}$/D', $values['phone'])) {
            throw new InvalidPage('Bitte gib eine gültige Telefonnummer an.');
        }
        $country = strtoupper(trim((string) ($input['country'] ?? 'DE')));
        if (1 !== preg_match('/^[A-Z]{2}$/D', $country)) {
            throw new InvalidPage('Das Land wird als zweibuchstabiger Code angegeben, zum Beispiel DE.');
        }
        $values['country'] = $country;
        foreach (preg_split('/[\r\n,]+/', (string) ($input['sameAs'] ?? '')) ?: [] as $profile) {
            $profile = trim($profile);
            if ('' === $profile) {
                continue;
            }
            if (!str_starts_with($profile, 'https://') || false === filter_var($profile, FILTER_VALIDATE_URL)) {
                throw new InvalidPage(sprintf('„%s“ ist keine HTTPS-Adresse.', $profile));
            }
            $values['sameAs'][] = $profile;
        }
        $values['sameAs'] = array_slice(array_unique($values['sameAs']), 0, 20);
        $values['localBusiness'] = (bool) ($input['localBusiness'] ?? false);
        if ($values['localBusiness'] && ('' === $values['street'] || '' === $values['city'])) {
            throw new InvalidPage('Ein lokales Unternehmen braucht mindestens Straße und Ort.');
        }
        $values['llmsTxt'] = (bool) ($input['llmsTxt'] ?? false);
        // Only comment and simple directive lines; nothing that could smuggle in a second policy.
        $extra = [];
        foreach (preg_split('/\R/', (string) ($input['robotsExtra'] ?? '')) ?: [] as $line) {
            $line = trim($line);
            if ('' !== $line && 1 === preg_match('/^(#.{0,180}|(?:Disallow|Allow|Crawl-delay|User-agent):[ \t][^\s][^\r\n]{0,180})$/D', $line)) {
                $extra[] = $line;
            }
        }
        $values['robotsExtra'] = implode("\n", array_slice($extra, 0, 30));

        $home = trim(trim((string) ($input['homePath'] ?? self::DEFAULTS['homePath'])), '/');
        if (1 !== preg_match('#^[a-z0-9]+(?:[-/][a-z0-9]+)*$#D', $home)) {
            throw new InvalidPage('Die Startseite braucht eine gültige Adresse.');
        }
        // A root that points nowhere would answer every visit with 404, so it must exist.
        if (false === $this->db->fetchOne('SELECT id FROM page WHERE website_id = ? AND path = ? AND deleted_at IS NULL', ['default', $home])) {
            throw new InvalidPage(sprintf('Es gibt keine Seite mit der Adresse /%s.', $home));
        }
        $values['homePath'] = $home;

        $this->db->executeStatement('UPDATE website SET settings = ? WHERE id = ?', [
            json_encode($values, JSON_THROW_ON_ERROR | JSON_UNESCAPED_UNICODE),
            'default',
        ]);
    }

    /** The stored path of the page that answers at the site root. */
    public function homePath(): string
    {
        return $this->load()['homePath'];
    }

    /**
     * Turns a stored page path into its public address. The root page answers at "/" and must not
     * also be reachable under its own path, or the same content would live at two addresses.
     */
    public function publicPath(string $storedPath, ?string $homePath = null): string
    {
        return $storedPath === ($homePath ?? $this->homePath()) ? '/' : '/'.$storedPath;
    }

    /**
     * Structured data for the organisation, built from the maintained values only. Returns null
     * while nothing is maintained: an empty Organization block would be noise, not data.
     */
    public function organisation(array $settings, string $origin, string $siteName): ?array
    {
        $hasAddress = '' !== $settings['street'] && '' !== $settings['city'];
        if ('' === $settings['legalName'] && !$hasAddress && '' === $settings['phone']) {
            return null;
        }
        $data = [
            '@context' => 'https://schema.org',
            '@type' => $settings['localBusiness'] && $hasAddress ? 'LocalBusiness' : 'Organization',
            'name' => '' !== $settings['legalName'] ? $settings['legalName'] : $siteName,
            'url' => $origin.'/',
        ];
        foreach (['description' => 'description', 'phone' => 'telephone', 'email' => 'email'] as $key => $property) {
            if ('' !== $settings[$key]) {
                $data[$property] = $settings[$key];
            }
        }
        if ($hasAddress) {
            $data['address'] = array_filter([
                '@type' => 'PostalAddress',
                'streetAddress' => $settings['street'],
                'postalCode' => $settings['postalCode'],
                'addressLocality' => $settings['city'],
                'addressCountry' => $settings['country'],
            ], static fn (string $value): bool => '' !== $value);
        }
        if ([] !== $settings['sameAs']) {
            $data['sameAs'] = $settings['sameAs'];
        }

        return $data;
    }
}