117. Populating Next Right Pointers in Each Node II

解题思路：
和Populating Next Right Pointers in Each Node（https://leetcode.com/problems/populating-next-right-pointers-in-each-node/）一样采用dfs，但是注意在dfs的时候需要多做一些next节点是否存在的判断。并且需要注意的是，DFS时要先遍历right son， 再遍历left son。
如下例：
2
1 3
0 7 9 1
1 1 0 8 8
若先搜索left son，再搜索right son，在遍历到值为7的节点，9->1的next连接还没有建立。

代码如下：

public class Solution {
    //most right son in the next generation
    public TreeLinkNode mostRightSon(TreeLinkNode root){
        if (root == null)
            return null;
        else
            return (root.right == null) ? root.left : root.right;
    }

    //most left son of root.next(or root.next.next...) in the next generation
    public TreeLinkNode mostLeftSon(TreeLinkNode root){
        TreeLinkNode node = root;
        while(node != null){
            if (node.left != null)
                return node.left;
            else if (node.right != null)
                return node.right;
            else
                node = node.next;
        }
        return null;
    }

    public void connect(TreeLinkNode root) {
        if (root == null)
            return;
        if (root.left != null && root.right != null){
            root.left.next = root.right;
        }

        TreeLinkNode mostRightSon = mostRightSon(root);
        if (mostRightSon != null){
            mostRightSon.next = mostLeftSon(root.next);
        }

        //root.right must be processed before left, because this is a depth first search so that the next of your right cousin may not have been processed.
        connect(root.right);
        connect(root.left);
    }
}