剑指offer 专项突破版 78、合并排序链表

题目链接

思路

• 最初的思路是上一题既然实现了归并两个链表，那么我们可以挨个归并~ 归并n-1次就可以，但是这样时间复杂度为O(k²n)

class Solution {
    public ListNode mergeKLists(ListNode[] lists) {
        if (0 == lists.length)
            return null;

        ListNode dummy = new ListNode(0, lists[0]);

        for (int i = 1; i < lists.length; i++) {
            ListNode head1 = dummy.next, head2 = lists[i], cur = dummy;

            while (null != head1 && null != head2) {
                if (head1.val < head2.val) {
                    cur.next = head1;
                    head1 = head1.next;
                } else {
                    cur.next = head2;
                    head2 = head2.next;
                }
                cur = cur.next;
            }
            cur.next = null == head1 ? head2 : head1;
        }

        return dummy.next;
    }
}

• 上面的思路之所以算的慢是因为，设每个长度为k，共有n个，第一次归并遍历k+k次；第二次归并遍历2k+k次；第三次归并遍历3k+k次，因此总共时间复杂度为O(k²n²)
• 但是我们可以第一次选择相邻两个归并，然后再相邻两个归并。也就是说，第一次是遍历n/2*(k+k) = nk次,第二次遍历n/4*(2k+2k)=nk次······因此时间复杂度为O(nklogn)

class Solution {

    private ListNode merge(ListNode head1, ListNode head2) {

        ListNode dummy = new ListNode(0), cur = dummy;

        while (null != head1 && null != head2) {
            if (head1.val < head2.val) {
                cur.next = head1;
                head1 = head1.next;
            } else {
                cur.next = head2;
                head2 = head2.next;
            }
            cur = cur.next;
        }
        cur.next = null == head1 ? head2 : head1;

        return dummy.next;
    }

    private ListNode mergeLists(ListNode[] lists, int start, int end) {
        if (start >= end)
            return lists[start];

        int mid = start + ((end - start) >> 1);

        ListNode head1 = mergeLists(lists, start, mid);
        ListNode head2 = mergeLists(lists, mid + 1, end);

        return merge(head1, head2);
    }

    public ListNode mergeKLists(ListNode[] lists) {
        if (0 == lists.length)
            return null;

        return mergeLists(lists,0,lists.length - 1);
    }
}