<?php
defined( 'ABSPATH' ) || exit;

/**
 * XML Sitemap generator with image support, news sitemap, and video sitemap.
 */
class AWBSEO_Sitemap {

    public static function init() {
        // Disable WordPress Core's built-in sitemap - AWBSEO provides its own.
        add_filter( 'wp_sitemaps_enabled', '__return_false' );

        add_action( 'init',                  [ __CLASS__, 'register_rewrites' ] );
        add_filter( 'query_vars',            [ __CLASS__, 'add_query_vars' ] );
        add_action( 'template_redirect',     [ __CLASS__, 'serve_sitemap' ] );
        add_action( 'save_post',             [ __CLASS__, 'flush_cache' ] );
        add_action( 'deleted_post',          [ __CLASS__, 'flush_cache' ] );
    }

    public static function register_rewrites() {
        add_rewrite_rule( '^sitemap\.xml$',              'index.php?awb_sitemap=index',    'top' );
        add_rewrite_rule( '^sitemap-posts\.xml$',        'index.php?awb_sitemap=posts',    'top' );
        add_rewrite_rule( '^sitemap-pages\.xml$',        'index.php?awb_sitemap=pages',    'top' );
        add_rewrite_rule( '^sitemap-products\.xml$',     'index.php?awb_sitemap=products', 'top' );
        add_rewrite_rule( '^sitemap-taxonomies\.xml$',   'index.php?awb_sitemap=tax',      'top' );
        add_rewrite_rule( '^sitemap-images\.xml$',       'index.php?awb_sitemap=images',   'top' );
        add_rewrite_rule( '^sitemap-news\.xml$',         'index.php?awb_sitemap=news',     'top' );

        // Register a sub-sitemap URL for every public custom post type, whether
        // or not it is currently checked "include in sitemap" in Settings, so
        // toggling that checkbox never requires a permalinks re-save. Post,
        // Page, and Product already have their own dedicated rules above
        // (with legacy plural URLs), so they're skipped here.
        foreach ( self::get_extra_post_types() as $pt ) {
            add_rewrite_rule( '^sitemap-' . $pt . '\.xml$', 'index.php?awb_sitemap=cpt_' . $pt, 'top' );
        }
    }

    /**
     * Public post types other than post/page/product/attachment - candidates
     * for their own "sitemap-{post_type}.xml" sub-sitemap.
     *
     * @return string[]
     */
    private static function get_extra_post_types() {
        $skip = [ 'post', 'page', 'product', 'attachment' ];
        return array_values( array_diff( get_post_types( [ 'public' => true ], 'names' ), $skip ) );
    }

    /**
     * Post types actually checked "include in sitemap" in Settings → Sitemap,
     * limited to ones that still exist.
     *
     * @return string[]
     */
    private static function get_configured_post_types() {
        $settings = get_option( 'awbseo_settings', AWBSEO_Core::default_settings() );
        $types    = (array) ( $settings['sitemap_post_types'] ?? [ 'post', 'page' ] );
        return array_values( array_filter( $types, 'post_type_exists' ) );
    }

    public static function add_query_vars( $vars ) {
        $vars[] = 'awb_sitemap';
        return $vars;
    }

    public static function serve_sitemap() {
        $type = get_query_var( 'awb_sitemap' );
        if ( ! $type ) return;

        $cached = get_transient( 'awbseo_sitemap_' . $type );
        if ( $cached ) {
            self::output( $cached );
            exit;
        }

        switch ( $type ) {
            case 'index':    $xml = self::build_index();      break;
            case 'posts':    $xml = self::build_posts();      break;
            case 'pages':    $xml = self::build_pages();      break;
            case 'products': $xml = self::build_products();   break;
            case 'tax':      $xml = self::build_taxonomies(); break;
            case 'images':   $xml = self::build_images();     break;
            case 'news':     $xml = self::build_news();       break;
            default:
                // Custom post type sub-sitemaps use a shared "cpt_{post_type}"
                // query var value (see register_rewrites()).
                $xml = ( 0 === strpos( $type, 'cpt_' ) )
                    ? self::build_cpt( substr( $type, 4 ) )
                    : null;
                break;
        }

        if ( $xml ) {
            set_transient( 'awbseo_sitemap_' . $type, $xml, HOUR_IN_SECONDS );
            self::output( $xml );
        }
        exit;
    }

    public static function flush_cache() {
        $types = [ 'index', 'posts', 'pages', 'products', 'tax', 'images', 'news' ];
        foreach ( self::get_extra_post_types() as $pt ) {
            $types[] = 'cpt_' . $pt;
        }
        foreach ( $types as $t ) {
            delete_transient( 'awbseo_sitemap_' . $t );
        }
    }

    /**
     * Return URL counts for each sub-sitemap for the admin preview page.
     * Uses fast COUNT queries - does not generate XML.
     *
     * @return array {
     *     @type int posts      Published posts (non-noindex)
     *     @type int pages      Published pages
     *     @type int taxonomies Term URLs across configured taxonomies
     *     @type int images     Posts/pages that have at least one image in content
     *     @type int news       Posts published in the last 2 days
     *     @type array cpt_counts Published count per included custom post type, keyed by slug
     *     @type int total      Grand total
     * }
     */
    public static function get_url_counts() {
        global $wpdb;

        $settings = get_option( 'awbseo_settings', [] );

        // Posts - exclude only those explicitly set to noindex.
        // Use IS NULL OR != noindex so posts with no meta row are included.
        $posts_count = (int) $wpdb->get_var(
            "SELECT COUNT(*) FROM {$wpdb->posts} p
             LEFT JOIN {$wpdb->postmeta} m
                 ON m.post_id = p.ID AND m.meta_key = '_awb_seo_robots_index'
             WHERE p.post_type = 'post'
               AND p.post_status = 'publish'
               AND ( m.meta_value IS NULL OR m.meta_value != 'noindex' )"
        );

        // Pages.
        // Use WP function for accuracy (handles page exclusions etc.).
        $pages_count = (int) wp_count_posts( 'page' )->publish;

        // Taxonomy terms.
        $tax_count  = 0;
        $configured_taxes = $settings['sitemap_taxonomies'] ?? [ 'category' ];
        foreach ( $configured_taxes as $tax ) {
            $terms = get_terms( [ 'taxonomy' => $tax, 'hide_empty' => true, 'fields' => 'count' ] );
            if ( ! is_wp_error( $terms ) ) {
                $tax_count += (int) $terms;
            }
        }

        // Images - posts/pages that contain at least one <img> tag.
        $images_count = (int) $wpdb->get_var(
            "SELECT COUNT(*) FROM {$wpdb->posts}
             WHERE post_type IN ('post','page')
               AND post_status = 'publish'
               AND post_content LIKE '%<img%'"
        );

        // News - posts in last 2 days.
        $two_days_ago = gmdate( 'Y-m-d H:i:s', strtotime( '-2 days' ) );
        $news_count   = (int) $wpdb->get_var(
            $wpdb->prepare(
                "SELECT COUNT(*) FROM {$wpdb->posts}
                 WHERE post_type = 'post'
                   AND post_status = 'publish'
                   AND post_date_gmt >= %s",
                $two_days_ago
            )
        );

        // Custom post types checked "include in sitemap" in Settings → Sitemap
        // (post/page already counted above; product is WooCommerce's own).
        $cpt_counts = [];
        $configured = self::get_configured_post_types();
        foreach ( self::get_extra_post_types() as $pt ) {
            if ( in_array( $pt, $configured, true ) ) {
                $cpt_counts[ $pt ] = (int) wp_count_posts( $pt )->publish;
            }
        }
        $cpt_total = array_sum( $cpt_counts );

        $total = $posts_count + $pages_count + $tax_count + $images_count + $news_count + $cpt_total;

        return compact( 'posts_count', 'pages_count', 'tax_count', 'images_count', 'news_count', 'cpt_counts', 'total' );
    }

    /**
     * Check whether the sitemap index is reachable (basic HTTP check).
     *
     * @return array { reachable: bool, code: int|null, latency_ms: int }
     */
    public static function check_reachability() {
        $url   = home_url( '/sitemap.xml' );
        $start = microtime( true );
        $resp  = wp_remote_head( $url, [ 'timeout' => 5, 'sslverify' => false ] );
        $ms    = (int) round( ( microtime( true ) - $start ) * 1000 );
        if ( is_wp_error( $resp ) ) {
            return [ 'reachable' => false, 'code' => null, 'latency_ms' => $ms ];
        }
        $code = (int) wp_remote_retrieve_response_code( $resp );
        return [ 'reachable' => $code >= 200 && $code < 400, 'code' => $code, 'latency_ms' => $ms ];
    }

    private static function output( $xml ) {
        header( 'Content-Type: application/xml; charset=UTF-8' );
        header( 'X-Robots-Tag: noindex, follow' );
        echo $xml;
    }

    private static function header( $extra_ns = '' ) {
        return '<?xml version="1.0" encoding="UTF-8"?>' . "\n"
            . '<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"'
            . ' xmlns:image="http://www.google.com/schemas/sitemap-image/1.1"'
            . ' xmlns:news="http://www.google.com/schemas/sitemap-news/0.9"'
            . $extra_ns . ">\n";
    }

    private static function footer() {
        return "</urlset>\n";
    }

    private static function build_index() {
        $sitemaps = [
            home_url( '/sitemap-posts.xml' ),
            home_url( '/sitemap-pages.xml' ),
            home_url( '/sitemap-taxonomies.xml' ),
            home_url( '/sitemap-images.xml' ),
            home_url( '/sitemap-news.xml' ),
        ];

        // Include products sub-sitemap only when WooCommerce is active.
        if ( post_type_exists( 'product' ) ) {
            $sitemaps[] = home_url( '/sitemap-products.xml' );
        }

        // Include one sub-sitemap per additional custom post type that is
        // checked "include in sitemap" in Settings → Sitemap.
        $configured = self::get_configured_post_types();
        foreach ( self::get_extra_post_types() as $pt ) {
            if ( in_array( $pt, $configured, true ) ) {
                $sitemaps[] = home_url( '/sitemap-' . $pt . '.xml' );
            }
        }

        $xml  = '<?xml version="1.0" encoding="UTF-8"?>' . "\n";
        $xml .= '<sitemapindex xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">' . "\n";
        foreach ( $sitemaps as $url ) {
            $xml .= "  <sitemap>\n    <loc>" . esc_url( $url ) . "</loc>\n  </sitemap>\n";
        }
        $xml .= '</sitemapindex>';
        return $xml;
    }

    private static function build_posts() {
        // No meta_query filter here - posts often have no _awb_seo_robots_index
        // meta row at all, which would cause a != compare to exclude them.
        // build_url_set() already skips noindex posts in PHP.
        $posts = get_posts( [
            'post_type'      => 'post',
            'post_status'    => 'publish',
            'posts_per_page' => -1,
        ] );
        return self::build_url_set( $posts );
    }

    private static function build_pages() {
        $pages = get_posts( [
            'post_type'      => 'page',
            'post_status'    => 'publish',
            'posts_per_page' => -1,
        ] );
        return self::build_url_set( $pages );
    }

    private static function build_url_set( $posts, $with_images = false ) {
        $xml = self::header();
        foreach ( $posts as $post ) {
            $robots = AWBSEO_Meta_Manager::get( $post->ID, 'robots_index', 'index' );
            if ( $robots === 'noindex' ) continue;
            $xml .= "  <url>\n";
            $xml .= '    <loc>' . esc_url( get_permalink( $post->ID ) ) . "</loc>\n";
            $xml .= '    <lastmod>' . get_the_modified_date( 'c', $post->ID ) . "</lastmod>\n";
            $xml .= "    <changefreq>weekly</changefreq>\n";
            $xml .= "    <priority>0.8</priority>\n";

            if ( $with_images ) {
                preg_match_all( '/<img[^>]+src=["\']([^"\']+)["\'][^>]*>/i', $post->post_content, $imgs );
                foreach ( $imgs[1] as $img_url ) {
                    $xml .= "    <image:image>\n";
                    $xml .= '      <image:loc>' . esc_url( $img_url ) . "</image:loc>\n";
                    $xml .= "    </image:image>\n";
                }
            }
            $xml .= "  </url>\n";
        }
        return $xml . self::footer();
    }

    private static function build_taxonomies() {
        $settings = get_option( 'awbseo_settings', AWBSEO_Core::default_settings() );
        $xml      = self::header();
        foreach ( ( $settings['sitemap_taxonomies'] ?? [ 'category' ] ) as $tax ) {
            $terms = get_terms( [ 'taxonomy' => $tax, 'hide_empty' => true ] );
            if ( is_wp_error( $terms ) ) continue;
            foreach ( $terms as $term ) {
                $xml .= "  <url>\n";
                $xml .= '    <loc>' . esc_url( get_term_link( $term ) ) . "</loc>\n";
                $xml .= "    <changefreq>weekly</changefreq>\n";
                $xml .= "    <priority>0.6</priority>\n";
                $xml .= "  </url>\n";
            }
        }
        return $xml . self::footer();
    }

    private static function build_products() {
        if ( ! post_type_exists( 'product' ) ) {
            return self::build_url_set( [] );
        }
        // No meta_query filter here - products often have no _awb_seo_robots_index
        // meta row at all, which would cause a != compare to exclude them.
        // build_url_set() already skips noindex posts in PHP.
        $products = get_posts( [
            'post_type'      => 'product',
            'post_status'    => 'publish',
            'posts_per_page' => -1,
        ] );
        return self::build_url_set( $products );
    }

    /**
     * Sub-sitemap for one custom post type. Returns an empty <urlset> (not
     * null) when the type is disabled/missing so a direct hit on the URL
     * behaves like any other empty sitemap rather than a hard failure.
     */
    private static function build_cpt( $post_type ) {
        if ( ! post_type_exists( $post_type ) || ! in_array( $post_type, self::get_configured_post_types(), true ) ) {
            return self::build_url_set( [] );
        }
        $posts = get_posts( [
            'post_type'      => $post_type,
            'post_status'    => 'publish',
            'posts_per_page' => -1,
        ] );
        return self::build_url_set( $posts );
    }

    private static function build_images() {
        $post_types = [ 'post', 'page' ];
        if ( post_type_exists( 'product' ) ) {
            $post_types[] = 'product';
        }
        $posts = get_posts( [ 'post_type' => $post_types, 'post_status' => 'publish', 'posts_per_page' => -1 ] );
        return self::build_url_set( $posts, true );
    }

    private static function build_news() {
        // Last 2 days for Google News.
        $posts = get_posts( [
            'post_type'      => 'post',
            'post_status'    => 'publish',
            'posts_per_page' => 1000,
            'date_query'     => [ [ 'after' => '2 days ago' ] ],
        ] );
        $xml = '<?xml version="1.0" encoding="UTF-8"?>' . "\n"
            . '<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"'
            . ' xmlns:news="http://www.google.com/schemas/sitemap-news/0.9">' . "\n";
        foreach ( $posts as $post ) {
            $xml .= "  <url>\n";
            $xml .= '    <loc>' . esc_url( get_permalink( $post->ID ) ) . "</loc>\n";
            $xml .= "    <news:news>\n";
            $xml .= '      <news:publication><news:name>' . esc_html( get_bloginfo( 'name' ) ) . "</news:name><news:language>en</news:language></news:publication>\n";
            $xml .= '      <news:publication_date>' . get_the_date( 'c', $post->ID ) . "</news:publication_date>\n";
            $xml .= '      <news:title>' . esc_html( get_the_title( $post->ID ) ) . "</news:title>\n";
            $xml .= "    </news:news>\n";
            $xml .= "  </url>\n";
        }
        $xml .= "</urlset>\n";
        return $xml;
    }
}