Sitemap
=======

The ``SonataSeoBundle`` provides a support for static sitemap generation.
You can either create services or provide raw SQL queries.

Service Definition
------------------

The service must implement the ``SourceIteratorInterface`` from the
``sonata-project/exporter`` library. The iterator will be used to generate
the file. The iterator must not store all data as property to avoid memory issue.
Data must be fetch one by one using a buffered connection (from mysql and not
php which is not default behavior)::

    class MyCustomSitemapIterator implements SourceIteratorInterface
    {
        protected $key;

        protected $stop;

        protected $current;

        public function __construct($stop = 1000)
        {
            $this->stop = $stop;
        }

        public function current()
        {
            return $this->current;
        }

        public function next()
        {
            $this->key++;
            $this->current = [
                'url'  => '/the/path/to/target',
                'lastmod'    => '01.01.2020',
                'changefreq' => 'weekly',
                'priority'   => 0.5
            ];
        }

        public function key()
        {
            return $this->key;
        }

        public function valid()
        {
            return $this->key < $this->stop;
        }

        public function rewind()
        {
            $this->key = 0;
        }
    }

Raw Performance with Query
--------------------------

You will need to configure:
 - a Doctrine connection
 - a route
 - default parameters used by the route
 - the query which must contains all information required by a sitemap: lastmod, changefreq and priority

The url is generated by using ``route`` and ``parameters`` settings.

For performance reasons and memory usage there is no model hydrated while generating the sitemap. So if the sitemap requires
specific rules, they must be expressed in the ``WHERE`` condition. The query must select fields to be used in the route.

The following code is an extract of the query required to generate a valid sitemap for the ``SonataNewsBundle``

.. code-block:: sql

    SELECT
        CONCAT_WS('/', YEAR(created_at), MONTH(created_at), DAY(created_at), slug) as url ,
        DATETIME(updated_at) as lastmod,
        'weekly' as changefreq,
        '0.5' as priority
    FROM news__post
    WHERE
            enabled = 1
        AND (publication_date_start IS NULL OR publication_date_start <= NOW())


Please note: the ``changefreq`` and the ``priority`` fields are optional.

Configuration example
---------------------

Sitemap configuration obviously depends on the bundle, page types & custom routes you choose to expose.

.. code-block:: yaml

    service:
        app.my_custom_sitemap_service:
            class: MyCustomSitemapIterator

    sonata_seo:
        # ...
        sitemap:
            services:
                - app.my_custom_sitemap_service

            doctrine_orm:
                # media
                - types: [image]
                  connection: doctrine.dbal.default_connection
                  route: sonata_media_view
                  parameters: {id: null}
                  query: |
                      SELECT
                        id,
                        updated_at as lastmod,
                        'weekly' as changefreq,
                        '0.5' as priority
                      FROM media__media
                      WHERE enabled = true
                # blog post
                - group: "news"
                  connection: doctrine.dbal.default_connection
                  route: sonata_news_view
                  parameters: {url: null}
                  query: |
                      SELECT
                        CONCAT_WS('/', YEAR(created_at), MONTH(created_at), DAY(created_at), slug) as url ,
                        updated_at as lastmod,
                        'weekly' as changefreq,
                        '0.5' as priority
                      FROM news__post
                      WHERE enabled = true
                        AND (publication_date_start IS NULL OR publication_date_start <= NOW())
                # page - works only for one site, please adapt the code if required
                - connection: doctrine.dbal.default_connection
                  route: page_slug
                  parameters: {path: null}
                  query: |
                      SELECT
                        url as path,
                        updated_at as lastmod,
                        'weekly' as changefreq,
                        '0.5' as priority
                      FROM page__snapshot
                      WHERE route_name = 'page_slug'
                        AND enabled = true
                        AND (publication_date_start IS NULL OR publication_date_start <= NOW())
                        AND (publication_date_end IS NULL OR publication_date_end >= NOW())
                # product categories
                - connection: doctrine.dbal.default_connection
                  route: sonata_catalog_category
                  parameters: {category_id: null, category_slug: null}
                  query: |
                      SELECT
                        id as category_id,
                        slug as category_slug,
                        updated_at as lastmod,
                        'weekly' as changefreq,
                        '0.5' as priority
                      FROM classification__category
                      WHERE enabled = true
                # products
                - connection: doctrine.dbal.default_connection
                  route: sonata_product_view
                  parameters: {productId: null, slug: null}
                  query: |
                      SELECT
                        id as productId,
                        slug,
                        updated_at as lastmod,
                        'weekly' as changefreq,
                        '0.5' as priority
                      FROM product__product
                      WHERE enabled = true

Usage
-----

- Generate the sitemap::

    bin/console sonata:seo:sitemap web sonata-project.org

.. note::

    The command will generate all files in a temporary directory to
    avoid issue will files are indexed. Once the files are generated
    then the files will be copied to the ``web`` directory. The
    ``sonata-project.org`` argument will be used to prefix url with the
    provided domain.