yrewrite-domains

YRewrite Domains

YRewrite ties hostnames (like www.example.com, example.de, en.example.com) to a "mount category" in the article structure. Each domain can be limited to a subset of languages (clangs) and have its own start article, 404 article, and sitemap.

Configuring a domain

Domains are configured in the backend (YRewrite → Domains) and persisted to redaxo/data/addons/yrewrite/domains.php. Code can read them but typically doesn't write them.

Programmatic access:

$current = rex_yrewrite::getCurrentDomain();
echo $current->getName();              // 'www.example.com'
echo $current->getMountId();           // mount category ID
echo $current->getStartId();           // start article ID for this domain
echo $current->getNotfoundId();        // 404 article ID
echo $current->getDefaultClang();      // default language ID

$all = rex_yrewrite::getDomains();      // all domains by name
$de  = rex_yrewrite::getDomainByName('www.example.de');

Generating URLs

Always use YRewrite's helpers – never construct URLs by hand. The helpers know about domains, language slugs, and pretty paths.

// Full URL including scheme and domain
$url = rex_yrewrite::getFullUrlByArticleId($articleId, $clangId);

// Path-only URL (relative to current domain)
$path = rex_yrewrite::getPathByArticleId($articleId, $clangId);

// Convenience: same as rex_getUrl() but always uses YRewrite if installed
$url = rex_getUrl($articleId, $clangId);

Use getFullUrlByArticleId() for anything that leaves the page (canonical tags, og:url, hreflang alternates, sitemap entries, email confirmations). Use getPathByArticleId() or rex_getUrl() for in-page links.

Multi-language URLs

Each domain maps to one or more languages. The default language usually has no path prefix; alternate languages get a prefix matching their code (e.g. /en/about, /fr/about).

foreach (rex_clang::getAll(true) as $clang) {
    if (!rex_yrewrite::getCurrentDomain()->isClangSupported($clang->getId())) {
        continue;
    }
    $url = rex_yrewrite::getFullUrlByArticleId(
        rex_article::getCurrentId(),
        $clang->getId()
    );
    echo '<link rel="alternate" hreflang="' . $clang->getCode() . '" href="' . rex_escape($url, 'url') . '">';
}

Add a self-referencing canonical with the current language's code as well, plus an x-default:

$defaultClang = rex_yrewrite::getCurrentDomain()->getDefaultClang();
$defaultUrl = rex_yrewrite::getFullUrlByArticleId(rex_article::getCurrentId(), $defaultClang);
echo '<link rel="alternate" hreflang="x-default" href="' . rex_escape($defaultUrl, 'url') . '">';

Detecting the current request context

if (rex_yrewrite::getCurrentDomain()->getName() === 'shop.example.com') {
    // shop-specific frontend logic
}

For language-prefix detection, prefer rex_clang::getCurrentId() over parsing the URL yourself.

Switching domains in code (rare)

Generally you stay on whichever domain the user requested. If you need to redirect across domains:

$other = rex_yrewrite::getDomainByName('en.example.com');
if ($other) {
    $url = rex_yrewrite::getFullUrlByArticleId(
        $articleId,
        $other->getDefaultClang()
    );
    rex_response::sendRedirect($url, 302);
}

Common pitfalls

• Constructing URLs with rex::getServer() . '/' . $path – ignores domain mappings, hreflang prefixes, and per-domain start articles.
• Using rex_getUrl() and copy-pasting the result somewhere external – it's a relative path. Use rex_yrewrite::getFullUrlByArticleId() for anything that leaves the page.
• Hardcoding language codes in URL paths – fetch them via rex_clang::get($id)->getCode().
• Running multiple domains in dev without /etc/hosts entries – YRewrite picks the wrong domain and routing silently goes to the default.
• Forgetting to set up the 404 article per domain. Without one, REDAXO falls back to the global 404, which may be in the wrong language.