GoPress Docs
Docs / URL and SEO

URL and SEO

GoPress treats URLs and SEO metadata as framework-level concerns. Themes provide templates and content presentation; core owns rewrite resolution, canonical URLs, sitemap output, redirects, and SEO metadata assembly.

Rewrite Engine

The rewrite engine resolves incoming URLs against registered content types, taxonomy archives, static pages, and custom theme routes. Theme-declared content types define archive behavior, rewrite slugs, and optional template mapping in theme.toml.

Database-driven 301 and 302 rules are resolved before normal rewrite handling, cached in memory, and track hit counts for administration.

Example:

[[content_types]]
name = "product"
has_archive = true
rewrite_slug = "products"

This produces archive and single-content URLs such as /products and /products/example-product.

product is not special to core. A theme can declare any content type name and any public URL base:

[[content_types]]
name = "module"
has_archive = true
rewrite_slug = "modules"
templates = { archive = "products", single = "product-detail" }

In this example the data model is module, the public URLs are /modules and /modules/{slug}, and the archive/detail pages reuse templates/pages/products.tmpl and templates/pages/product-detail.tmpl. This keeps routing, admin CRUD, REST API exposure, sitemap entries, and frontend rendering driven by the same registry entry.

Theme templates should use registry-aware helpers for internal content links:

{{archiveURL "module"}}
{{contentURL . "module"}}
{{taxonomyURL "category" .Slug}}

archiveURL returns the current archive URL for a content type. contentURL uses an item's existing URL field when present, otherwise combines the item's Type/Slug with the rewrite registry and falls back to the supplied type. taxonomyURL builds the canonical term archive path used by internal links and sitemap entries.

SEO Builder

The SEO builder creates page-level metadata for:

  • Home pages.
  • Archive pages.
  • Single content pages.
  • Taxonomy archives.

The resulting metadata includes the document title, meta description, self-referencing canonical URL, Open Graph title/description/image/type, robots directive, and page-appropriate JSON-LD such as Article or WebSite.

Themes render the common output through {{seoHeadFor .}} and select the document title through {{pageTitleFor . $fallbackTitle}}; see Theme SEO Integration.

Taxonomy archives receive their own self-referencing canonical URL. Request-aware SEO rendering also keeps canonical URLs in the current non-default language, matching the language-prefixed internal URL.

Site-wide values such as site_name, site_description, and site_icon are applied as final fallbacks from system settings. site_icon renders /favicon.ico first, then a typed image icon and Apple touch icon when present.

For normal internal paths, templates should use {{langPrefixURL .Ctx "/path"}}. Together with the rewrite-aware helpers, this keeps links in the current non-default language instead of dropping visitors back to the default language.

Plugin Overrides

The seo.content.meta filter allows plugins to modify SEO output after core builds the default metadata. The built-in seo-extras plugin uses this to provide per-content SEO title, description, Open Graph image, and robots overrides.

Sitemap and Redirects

Sitemap generation reads registered content types and their rewrite configuration, taxonomy URLs, and route transformers. Redirect rules are stored separately and are resolved before normal rewrite handling.

Core's sitemap transformer contract can add xhtml:link rel="alternate" entries and derived localized URLs. A multilingual extension uses that generic contract to publish hreflang translation groups without requiring changes in core or themes.

/sitemap.xml is served dynamically by the active site process and supports both GET and HEAD. The admin "Generate Sitemap" action writes a static copy to the active site's public/ directory, for example sites/example.com/public/sitemap.xml, so multiple sites can share one application root without overwriting each other's generated files. Future site-scoped public artifacts such as robots.txt or llms.txt should use the same directory.