afterattack now returns a flag if it's reasonable to suspect the user intends to act on an item (#72320)

Necessary for #72292 to work effectively, and probably not very useful
out of that context. Split out of its own PR because this is long and
boring.

I want to make sure that we're catching actual mistakes there, and not
just experiencing side effects of how shitty the attack chain is.

code/datums/elements/openspace_item_click_handler.dm

@@ -22,3 +22,4 @@
 	var/turf/turf_above = get_step_multiz(target, UP)
 	if(turf_above?.z == user.z)
 		INVOKE_ASYNC(source, TYPE_PROC_REF(/obj/item, handle_openspace_click), turf_above, user, user.CanReach(turf_above, source), click_parameters)
+	return COMPONENT_AFTERATTACK_PROCESSED_ITEM