double-term.php

<?php
/*
Plugin Name: double-term
Description: Custom code for oieniuz's auto sales installation. Sets up sitemap links using two taxonomy terms. Requires Yoast SEO plugin to be active.
Author: bcworkz
Version: 0.10 beta
License: GPL 2.0
*/

/**
 *  Requires Yoast WP SEO plugin to be installed and activated. This is not verified.
 *  Adds custom sitemaps to Yoast sitemaps that include links with double terms. This sitemap type
 *  in Yoast is 'dblterm', the sitemap URLS are something like dblterm-sitemap3.xml . The '3' could
 *  be any integer. They are referenced in the root sitemap-index.xml generated by Yoast
 *  Sitemap URLs are shown to be published 1/2 week ago, regardless of actual publish dates.
 */

// Constants for use in custom Yoast sitemaps - adjust values as required
// This plugin ignores the max. items set in Yoast. The value here is used.
// To flush Yoast sitemap cache, change the Max items on Yoast's sitemap screen, then save
// You must also reload the retrieved sitemap to flush browser cached versions

/**
 * Max items. Per sitemap specifications, max is 50,000 minus $location_count
 *  - maybe assume 48k max ?? (there may also be memory constraints)
 *  - script will output more than specified to complete a full set of locations per model
 */
define( 'MAX_ITEMS', 1000 );

// Taxonomies for term sources
define('MODELS', 'vehicle_model');
define('LOCATIONS', 'vehicle_location');

// Register "dblterm" sitemap type with Yoast sitemap object
//  this allows Yoast to handle redirects and sitemap caching :)
add_action('init', 'dt_register_map');
function dt_register_map() {
    if ( array_key_exists('wpseo_sitemaps', $GLOBALS )) {
        $sitemaps = $GLOBALS['wpseo_sitemaps'];
        $sitemaps->register_sitemap( 'dblterm', 'dt_gen_sitemap_dbl_terms');
    }
}

/**
 * generates XML sitemap content from 2 taxonomies/terms
 * used as callback to action 'wpseo_do_sitemap_dblterm'
 * ignores Yoast max. entries - The max. is hardcoded as constant MAX_ITEMS above.
 * filename to request: dblterm-sitemap#.xml where # is page number (any places: #, ##, ###, ...)
 */
function dt_gen_sitemap_dbl_terms() {
    // assign constants to local vars
    $max_items = MAX_ITEMS;
    $vehicle_model = MODELS;
    $vehicle_location = LOCATIONS;

    $page = get_query_var( 'sitemap_n' );   //Yoast sitemap page # taken from filename requested
    if ( '' == $page ) $page = 1;           // Assume page 1 if no query var
    // Yoast will send headers, set <?xml version, encoding, stylesheets, templates ? >
    $pages = '<urlset xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:image="http://www.google.com/schemas/sitemap-image/1.1" xsi:schemaLocation="http://www.sitemaps.org/schemas/sitemap/0.9 http://www.sitemaps.org/schemas/sitemap/0.9/sitemap.xsd" xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">' . "\n";

    // get terms through SQL queries
    $models = dt_alt_get_terms( $vehicle_model ); // each hides empty terms by default
    $locations = dt_alt_get_terms( $vehicle_location );
    $model_ct = count( $models );
    $location_ct = count( $locations );

    // determine parameters so page breaks occur after full set of locations are output
    $models_per_page = floor(($max_items/$location_ct) + 1 );
    $offset = ($page-1) * $models_per_page;
    if ( 0 > $offset ) $offset = 0;         // safety function just in case something odd happens
    $limit = $offset + ($models_per_page * $location_ct);
    // do not surpass number of models available
    $limit = ( $model_ct < $limit ) ? $model_ct: $limit;

    $count = 0;
    // Loop through all term permutaions
    for ( $i=$offset; $i<$limit; $i++ ) {
        $model = $models[$i];
        foreach ( $locations as $location ) {
            $location2 = home_url( '/used-cars/location/' . $location . '/model/' . $model .'/' );
            // Assume all published 1/2 week ago
            $date = date_i18n('o-m-d\TH:i:sP', current_time('timestamp')-( 3.5 * DAY_IN_SECONDS ));
            $pages .= "\t<url>\n\t\t<loc>$location2</loc>\n\t\t<lastmod>$date</lastmod>\n\t</url>\n";
            $count++;
        }
        if ( $count >= $max_items) break;
    }

    // Get Yoast sitemap object and give it our sitemap
    $sitemaps = $GLOBALS['wpseo_sitemaps'];
    $sitemaps->set_sitemap( $pages . '</urlset>');
    return;
}

// Add dblterm-sitemap links to sitemap-index
add_filter( 'wpseo_sitemap_index', 'dt_add_to_index' );
function dt_add_to_index() {
    // assign constants to local vars
    $max_items = MAX_ITEMS;
    $vehicle_model = MODELS;
    $vehicle_location = LOCATIONS;

    // get term counts through SQL queries
    $models = dt_alt_get_terms( $vehicle_model ); // each hides empty terms by default
    $locations = dt_alt_get_terms( $vehicle_location );
    $model_ct = count( $models );
    $location_ct = count( $locations );

    // determine parameters so page breaks occur after full set of locations are output
    $models_per_page = floor(( $max_items / $location_ct ) + 1 );
    $max_page_idx = floor(( $model_ct / $models_per_page ) + 1 );

    $links = '';
    // Assume all published 1/2 week ago
    $date = date_i18n('o-m-d\TH:i:sP', current_time('timestamp')-( 3.5 * DAY_IN_SECONDS ));
    for ( $i=1; $i<=$max_page_idx; $i++ ) {
        $url = get_site_url( null, '/dblterm-sitemap' . (string)$i . '.xml');
        $links .= "\t<sitemap>\n\t\t<loc>$url</loc>\n\t\t<lastmod>$date</lastmod>\n\t</sitemap>\n";
    }
    return $links;
}

/**
 * Gets all terms for taxonomy $tax
 * Hides empty terms unless hiding algorithm is commented out (@ line 144)
 * Returns an array of term slugs
 */
function dt_alt_get_terms( $tax ) {
/*  // The original alt_get_terms(). Direct query ought to be a bit faster??
    $hide_empty = true;
    $term_query = new WP_Term_Query(['taxonomy'=> $tax, 'fields'=>'id=>slug', 'hide_empty'=> $hide_empty,]);
    $terms = $term_query->get_terms();
    $no_empties = array();
    foreach ( $terms as $term ) { $no_empties[] = $term->slug; }
*/
    global $wpdb;
    $terms = $wpdb->get_results( $wpdb->prepare("SELECT t.term_id, t.slug, tt.count, tt.taxonomy FROM $wpdb->terms AS t
        INNER JOIN $wpdb->term_taxonomy AS tt ON t.term_id = tt.term_id
        WHERE tt.taxonomy IN ('%s') ORDER BY t.slug ASC;",
        $tax
    ));
    // Remove empty terms unless children are not empty (hide_empty = true)
    // Algorithm copied from WordPress core code (it's OK, WP is GPL licensed)
    // To show empties add astrisk between two slashes at beginning of this line
    if ( is_array( $terms ) ) {
        foreach ( $terms as $k => $term ) {
            if ( ! $term->count ) {
                $children = get_term_children( $term->term_id, $term->taxonomy );
                if ( is_array( $children ) ) {
                    foreach ( $children as $child_id ) {
                        $child = get_term( $child_id, $term->taxonomy );
                        if ( $child->count ) {
                            continue 2;
                        }
                    }
                }
                // It really is empty.
                unset( $terms[ $k ] );
            }
        }
    }
    // end remove empty terms------------------------ */

    //make array indices sequential, save value of slug only
    $no_empties = array();
    foreach ( $terms as $term ) { $no_empties[] = $term->slug; }
    return $no_empties;
}

/**
 * Transient Garbage Collection - WP only does this on DB upgrades
 *
 * Deletes all expired transients.
 * The multi-table delete syntax is used to delete the transient record from table a,
 * and the corresponding transient_timeout record from table b, though in fact both
 * are in the same table.
 */

// only setup event on activation
register_activation_hook(__FILE__, 'dt_activation');

function dt_activation() {
    if (! wp_next_scheduled ( 'dt_event' )) {
        wp_schedule_event(time(), 'daily', 'dt_event');
    }
}

add_action('dt_event', 'dt_do_this_daily');
function dt_do_this_daily() {
    global $wpdb;
    $time = time();
    //from upgrade_network()
    $wpdb->query("DELETE a, b FROM $wpdb->options a, $wpdb->options b WHERE
        a.option_name LIKE '\_transient\_%' AND
        a.option_name NOT LIKE '\_transient\_timeout\_%' AND
        b.option_name = CONCAT( '_transient_timeout_', SUBSTRING( a.option_name, 17 ) )
        AND b.option_value < $time");
    return;
}

// remove event hook on deactivation
register_deactivation_hook(__FILE__, 'dt_deactivation');
function dt_deactivation() {
    wp_clear_scheduled_hook('dt_event');
}