Incorrect Use of Removing Query String From URLs

[amiga-500]

The remove_query() function is used to strip out WP's query string name-value pair from a URL. More specifically, it removes the ver=4.#.# held within a query string. However, some problems were identified:

• In DbCache_Environment.php and ObjectCache_Environment.php the line show below has an issue. This line generates a new URL with its own query string. However, it uses remove_query() which can still potentially return back a query string of its own. Recall that remove_query() only strips out ver=4.#.#. It seems the author's intention was to use remove_query_all(), which strips out a query string entirely, readying it for the new one: '?page=w3tc_dashboard&':

$url = basename( Util_Environment::remove_query( $_SERVER['REQUEST_URI'] ) ) . '?page=w3tc_dashboard&';

• When using the Prevent Caching of Objects After Settings Change feature for CSS & JS files, inside UriRewriter.php, which handles URIs contained in CSS files, there is a check to see if a query string exists. If so, then it follows by stripping ?ver=4.#.# and then attaching $_browserCacheId. The problem is that if the URI query string consisted only of ?ver=4.#.# the final URI will end up malformed (e.g., it would end up like this: test.css&x47202 instead of this: test.css?x47202).