Improve generate_cache_key to reduce cache miss while avoiding collision

[kt-12]

Trac ticket: https://core.trac.wordpress.org/ticket/59516

---

This Pull Request is for code review only. Please keep all other discussion in the Trac ticket. Do not merge this Pull Request. See GitHub Pull Requests for Code Review in the Core Handbook for more details.

[peterwilsoncc]

When generating the cache key, the SQL statement is normalized to use the same fields regardless of the processed SQL statement:

wordpress-develop/src/wp-includes/class-wp-query.php

Lines 3174 to 3175 in e486ac0

	$new_request = str_replace( $fields, "{$wpdb->posts}.*", $this->request );
	$cache_key = $this->generate_cache_key( $q, $new_request );

However, in the following lines the actual cached data an be altered depending on the arguments used in the query in the code immediately following the lines linked above.

You're right, it might be possible to optimize the cache key generation to increase the number of cache hits but the cache key will need to consider some of the options used. It may be worth considering:

• normalizing arguments a little more, for example the post type argument 'post' and [ 'post' ] are equivalent; as are [ 'page', 'post' ] and [ 'post', 'page' ]
• whether additional arguments can be removed from the cache key

I think we'd need a bunch of tests to make sure there aren't cache collisions too.