First, note that all graphs formed by permutations are composed of a bunch of different cycles. Also, we can only create new components by disconnecting existing cycles, because nodes in the query set ~S~ cannot attach to themselves or each other to create new ones. For all ~1 \le i \le N~, let node ~i~ be pointing towards node ~P_i~. Consider a node ~x~ in ~S~. We can delete the edge from ~x~ to ~P_x~ if we reroute it backwards to the node which is pointing to ~x~, essentially making the edge meaningless. Deleting ~D~ edges in a cycle will create ~D-1~ new components. However, we can only delete an edge from ~x~ to ~P_x~ if the node pointing to ~x~ is not in the query set, because ~2~ nodes in ~S~ cannot be connected with an edge. So if the node pointing to ~x~ is also in ~S~, then ~x~ cannot disconnect the component further, and we just attach ~x~ to another existing component. This does not affect the answer. Thus, we can find how many new components we can create from a cycle by counting the number of edges we can delete.

Subtask 1

The initial graph is ~1~ big cycle, so our initial answer is ~1~ component. Let ~D~ be the number of edges we can delete. We know that deleting ~D~ edges in a cycle will create ~D-1~ new components. Adding ~D-1~ to the initial answer ~1~, we get ~D~ as the final answer.

Subtask 2

Let the initial number of cycles be ~C~. For each component that contains a node in the query set ~S~, we can find ~D~, and add ~D-1~ to the answer. However, we do not actually need to consider each component independently. Notice that for each component, we always just subtract ~1~. So to find the number of new components, we can just find the total ~D~ value then subtract the number of components affected. Add this to the initial answer ~C~ to get the final answer.