You are given a directed graph with n nodes/cities, labeled from 0 to n - 1. The graph is represented by a 0-indexed 2D integer array graph, where graph[i] is an integer array of cities directly reachable from city i, meaning there is a direct road (edge) from city i to each of the cities in graph[i].

Destination city: A city (node) that has no outgoing roads (edges).

Safe city: A city where every possible route starting from that city eventually leads to a destination city or another safe city, and no route can get stuck in a cycle.

Return an array of all the safe cities in the graph (sorted in ascending order).

Example 1:

Input: graph = [[1,2],[2,3],[5],[0],[5],[],[]]
Output: [2,4,5,6]
Explanation: The given graph is shown above.
Nodes 5 and 6 are destination cities because they have no outgoing roads (edges).
All paths starting from cities 2, 4, 5, and 6 eventually lead to either node 5 or node 6.

Example 2:

Input: graph = [[1], [2], [3], [4], [5], []]
Output: [0, 1, 2, 3, 4, 5]
Explanation:
Node 5 is a destination city, so it's safe. 
Node 4 leads to node 5 (a destination city), so it's safe.
Node 3 leads to node 4 (a safe city), so it's safe.
Node 2 leads to node 3 (a safe city), so it's safe.
Node 1 leads to node 2 (a safe city), so it's safe.
Node 0 leads to node 1 (a safe city), which leads to node 2, which leads to node 3, which leads to node 4, which leads to node 5, so it's safe.

Example 3:

Input: graph = [[1,2,3,4],[1,2],[3,4],[0,4],[]]
Output: [4]
Explanation: Node 4 is the only destination city.

Constraints

n == graph.length

1 <= n <= 10⁴

0 <= graph[0].length <= n

0 <= graph[0][0] <= n - 1

graph[0] is sorted in a strictly increasing order.

The graph may contain self-loops.

The number of edges in the graph will be in the range [0, 4 * $10^4$ ].

91. Safe Cities