SQL面试题练习 —— 查询最近一笔有效订单

目录

  • [1 题目](#1 题目)
  • [2 建表语句](#2 建表语句)
  • [3 题解](#3 题解)

题目来源:字节跳动。

1 题目

现有订单表t_order,包含订单ID,订单时间,下单用户,当前订单是否有效,请查询出每个用户每笔订单的上一笔有效订单

+---------+----------------------+----------+-----------+
| ord_id  |       ord_time       | user_id  | is_valid  |
+---------+----------------------+----------+-----------+
| 1       | 2023-12-11 12:01:03  | a        | 1         |
| 2       | 2023-12-11 12:02:06  | a        | 0         |
| 3       | 2023-12-11 12:03:15  | a        | 0         |
| 4       | 2023-12-11 12:04:20  | a        | 1         |
| 5       | 2023-12-11 12:05:03  | a        | 1         |
| 6       | 2023-12-11 12:01:02  | b        | 1         |
| 7       | 2023-12-11 12:03:03  | b        | 0         |
| 8       | 2023-12-11 12:04:01  | b        | 1         |
| 9       | 2023-12-11 12:07:03  | b        | 1         |
+---------+----------------------+----------+-----------+

期望查询结果如下:

+---------+----------------------+----------+-----------+--------------------+
| ord_id  |       ord_time       | user_id  | is_valid  | last_valid_ord_id  |
+---------+----------------------+----------+-----------+--------------------+
| 1       | 2023-12-11 12:01:03  | a        | 1         | NULL               |
| 2       | 2023-12-11 12:02:06  | a        | 0         | 1                  |
| 3       | 2023-12-11 12:03:15  | a        | 0         | 1                  |
| 4       | 2023-12-11 12:04:20  | a        | 1         | 1                  |
| 5       | 2023-12-11 12:05:03  | a        | 1         | 4                  |
| 6       | 2023-12-11 12:01:02  | b        | 1         | NULL               |
| 7       | 2023-12-11 12:03:03  | b        | 0         | 6                  |
| 8       | 2023-12-11 12:04:01  | b        | 1         | 6                  |
| 9       | 2023-12-11 12:07:03  | b        | 1         | 8                  |
+---------+----------------------+----------+-----------+--------------------+

2 建表语句

sql 复制代码
--建表语句
create table t_order
(
ord_id bigint COMMENT '订单ID',
ord_time string COMMENT '订单时间',
user_id string COMMENT '用户',
is_valid bigint COMMENT '订单是否有效'
) COMMENT '订单记录表'
stored as orc
;
-- 数据插入
insert into t_order(ord_id,ord_time,user_id,is_valid)
values
(1,'2023-12-11 12:01:03','a',1),
(2,'2023-12-11 12:02:06','a',0),
(3,'2023-12-11 12:03:15','a',0),
(4,'2023-12-11 12:04:20','a',1),
(5,'2023-12-11 12:05:03','a',1),
(6,'2023-12-11 12:01:02','b',1),
(7,'2023-12-11 12:03:03','b',0),
(8,'2023-12-11 12:04:01','b',1),
(9,'2023-12-11 12:07:03','b',1);

3 题解

(1)先查询出有效订单,然后计算出每笔有效订单的上一单有效订单;

sql 复制代码
select ord_id,
       ord_time,
       user_id,
       is_valid,
       lag(ord_id) over (partition by user_id order by ord_time asc) as last_valid_ord_id
from (select ord_id,
             ord_time,
             user_id,
             is_valid
      from t_order
      where is_valid = 1) t

执行结果

+---------+----------------------+----------+-----------+--------------------+
| ord_id  |       ord_time       | user_id  | is_valid  | last_valid_ord_id  |
+---------+----------------------+----------+-----------+--------------------+
| 1       | 2023-12-11 12:01:03  | a        | 1         | NULL               |
| 4       | 2023-12-11 12:04:20  | a        | 1         | 1                  |
| 5       | 2023-12-11 12:05:03  | a        | 1         | 4                  |
| 6       | 2023-12-11 12:01:02  | b        | 1         | NULL               |
| 8       | 2023-12-11 12:04:01  | b        | 1         | 6                  |
| 9       | 2023-12-11 12:07:03  | b        | 1         | 8                  |
+---------+----------------------+----------+-----------+--------------------+

(2)原始的明细数据与新的有效订单表按照用户进行关联,有效订单表的订单时间大于等于原始订单表;

sql 复制代码
with tmp as (
    -- 有效订单及其上一单有效记录
    select ord_id,
           ord_time,
           user_id,
           is_valid,
           lag(ord_id) over (partition by user_id order by ord_time asc) as last_valid_ord_id
    from (select ord_id,
                 ord_time,
                 user_id,
                 is_valid
          from t_order
          where is_valid = 1) t)
select t1.*,
       t2.*
from t_order t1
left join tmp t2
on t1.user_id = t2.user_id
where t1.ord_time <= t2.ord_time

执行结果

+------------+----------------------+-------------+--------------+------------+----------------------+-------------+--------------+-----------------------+
| t1.ord_id  |     t1.ord_time      | t1.user_id  | t1.is_valid  | t2.ord_id  |     t2.ord_time      | t2.user_id  | t2.is_valid  | t2.last_valid_ord_id  |
+------------+----------------------+-------------+--------------+------------+----------------------+-------------+--------------+-----------------------+
| 1          | 2023-12-11 12:01:03  | a           | 1            | 1          | 2023-12-11 12:01:03  | a           | 1            | NULL                  |
| 1          | 2023-12-11 12:01:03  | a           | 1            | 4          | 2023-12-11 12:04:20  | a           | 1            | 1                     |
| 2          | 2023-12-11 12:02:06  | a           | 0            | 4          | 2023-12-11 12:04:20  | a           | 1            | 1                     |
| 3          | 2023-12-11 12:03:15  | a           | 0            | 4          | 2023-12-11 12:04:20  | a           | 1            | 1                     |
| 4          | 2023-12-11 12:04:20  | a           | 1            | 4          | 2023-12-11 12:04:20  | a           | 1            | 1                     |
| 1          | 2023-12-11 12:01:03  | a           | 1            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     |
| 2          | 2023-12-11 12:02:06  | a           | 0            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     |
| 3          | 2023-12-11 12:03:15  | a           | 0            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     |
| 4          | 2023-12-11 12:04:20  | a           | 1            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     |
| 5          | 2023-12-11 12:05:03  | a           | 1            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     |
| 6          | 2023-12-11 12:01:02  | b           | 1            | 6          | 2023-12-11 12:01:02  | b           | 1            | NULL                  |
| 6          | 2023-12-11 12:01:02  | b           | 1            | 8          | 2023-12-11 12:04:01  | b           | 1            | 6                     |
| 7          | 2023-12-11 12:03:03  | b           | 0            | 8          | 2023-12-11 12:04:01  | b           | 1            | 6                     |
| 8          | 2023-12-11 12:04:01  | b           | 1            | 8          | 2023-12-11 12:04:01  | b           | 1            | 6                     |
| 6          | 2023-12-11 12:01:02  | b           | 1            | 9          | 2023-12-11 12:07:03  | b           | 1            | 8                     |
| 7          | 2023-12-11 12:03:03  | b           | 0            | 9          | 2023-12-11 12:07:03  | b           | 1            | 8                     |
| 8          | 2023-12-11 12:04:01  | b           | 1            | 9          | 2023-12-11 12:07:03  | b           | 1            | 8                     |
| 9          | 2023-12-11 12:07:03  | b           | 1            | 9          | 2023-12-11 12:07:03  | b           | 1            | 8                     |
+------------+----------------------+-------------+--------------+------------+----------------------+-------------+--------------+-----------------------+

(3)使用row_number,原始订单记录表中的user_id、ord_id进行分组,按照有效订单表的时间排序,增加分组排序

sql 复制代码
with tmp as (
    -- 有效订单及其上一单有效记录
    select ord_id,
           ord_time,
           user_id,
           is_valid,
           lag(ord_id) over (partition by user_id order by ord_time asc) as last_valid_ord_id
    from (select ord_id,
                 ord_time,
                 user_id,
                 is_valid
          from t_order
          where is_valid = 1) t)
select t1.*,
       t2.*,
       row_number() over (partition by t1.ord_id,t1.user_id order by t2.ord_time asc) as rn
from t_order t1
left join tmp t2
on t1.user_id = t2.user_id
where t1.ord_time <= t2.ord_time

执行结果

+------------+----------------------+-------------+--------------+------------+----------------------+-------------+--------------+-----------------------+-----+
| t1.ord_id  |     t1.ord_time      | t1.user_id  | t1.is_valid  | t2.ord_id  |     t2.ord_time      | t2.user_id  | t2.is_valid  | t2.last_valid_ord_id  | rn  |
+------------+----------------------+-------------+--------------+------------+----------------------+-------------+--------------+-----------------------+-----+
| 1          | 2023-12-11 12:01:03  | a           | 1            | 1          | 2023-12-11 12:01:03  | a           | 1            | NULL                  | 1   |
| 1          | 2023-12-11 12:01:03  | a           | 1            | 4          | 2023-12-11 12:04:20  | a           | 1            | 1                     | 2   |
| 1          | 2023-12-11 12:01:03  | a           | 1            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     | 3   |
| 2          | 2023-12-11 12:02:06  | a           | 0            | 4          | 2023-12-11 12:04:20  | a           | 1            | 1                     | 1   |
| 2          | 2023-12-11 12:02:06  | a           | 0            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     | 2   |
| 3          | 2023-12-11 12:03:15  | a           | 0            | 4          | 2023-12-11 12:04:20  | a           | 1            | 1                     | 1   |
| 3          | 2023-12-11 12:03:15  | a           | 0            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     | 2   |
| 4          | 2023-12-11 12:04:20  | a           | 1            | 4          | 2023-12-11 12:04:20  | a           | 1            | 1                     | 1   |
| 4          | 2023-12-11 12:04:20  | a           | 1            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     | 2   |
| 5          | 2023-12-11 12:05:03  | a           | 1            | 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     | 1   |
| 6          | 2023-12-11 12:01:02  | b           | 1            | 6          | 2023-12-11 12:01:02  | b           | 1            | NULL                  | 1   |
| 6          | 2023-12-11 12:01:02  | b           | 1            | 8          | 2023-12-11 12:04:01  | b           | 1            | 6                     | 2   |
| 6          | 2023-12-11 12:01:02  | b           | 1            | 9          | 2023-12-11 12:07:03  | b           | 1            | 8                     | 3   |
| 7          | 2023-12-11 12:03:03  | b           | 0            | 8          | 2023-12-11 12:04:01  | b           | 1            | 6                     | 1   |
| 7          | 2023-12-11 12:03:03  | b           | 0            | 9          | 2023-12-11 12:07:03  | b           | 1            | 8                     | 2   |
| 8          | 2023-12-11 12:04:01  | b           | 1            | 8          | 2023-12-11 12:04:01  | b           | 1            | 6                     | 1   |
| 8          | 2023-12-11 12:04:01  | b           | 1            | 9          | 2023-12-11 12:07:03  | b           | 1            | 8                     | 2   |
| 9          | 2023-12-11 12:07:03  | b           | 1            | 9          | 2023-12-11 12:07:03  | b           | 1            | 8                     | 1   |
+------------+----------------------+-------------+--------------+------------+----------------------+-------------+--------------+-----------------------+-----+

(4)去除冗余字段,筛选rn=1 的记录

sql 复制代码
with tmp as (
    -- 有效订单及其上一单有效记录
    select ord_id,
           ord_time,
           user_id,
           is_valid,
           lag(ord_id) over (partition by user_id order by ord_time asc) as last_valid_ord_id
    from (select ord_id,
                 ord_time,
                 user_id,
                 is_valid
          from t_order
          where is_valid = 1) t)
select *
from (select t1.*,
             t2.last_valid_ord_id,
             row_number() over (partition by t1.ord_id,t1.user_id order by t2.ord_time asc) as rn
      from t_order t1
      left join tmp t2
      on t1.user_id = t2.user_id
      where t1.ord_time <= t2.ord_time) tt
where rn = 1

执行结果

+------------+----------------------+-------------+--------------+-----------------------+--------+
| tt.ord_id  |     tt.ord_time      | tt.user_id  | tt.is_valid  | tt.last_valid_ord_id  | tt.rn  |
+------------+----------------------+-------------+--------------+-----------------------+--------+
| 1          | 2023-12-11 12:01:03  | a           | 1            | NULL                  | 1      |
| 2          | 2023-12-11 12:02:06  | a           | 0            | 1                     | 1      |
| 3          | 2023-12-11 12:03:15  | a           | 0            | 1                     | 1      |
| 4          | 2023-12-11 12:04:20  | a           | 1            | 1                     | 1      |
| 5          | 2023-12-11 12:05:03  | a           | 1            | 4                     | 1      |
| 6          | 2023-12-11 12:01:02  | b           | 1            | NULL                  | 1      |
| 7          | 2023-12-11 12:03:03  | b           | 0            | 6                     | 1      |
| 8          | 2023-12-11 12:04:01  | b           | 1            | 6                     | 1      |
| 9          | 2023-12-11 12:07:03  | b           | 1            | 8                     | 1      |
+------------+----------------------+-------------+--------------+-----------------------+--------+
相关推荐
Viktor_Ye12 分钟前
高效集成易快报与金蝶应付单的方案
java·前端·数据库
努力算法的小明1 小时前
SQL 复杂查询
数据库·sql
斗-匕1 小时前
MySQL 三大日志详解
数据库·mysql·oracle
代码中の快捷键1 小时前
MySQL数据库存储引擎
数据库·mysql
只因在人海中多看了你一眼1 小时前
数据库体系
数据库
尘浮生1 小时前
Java项目实战II基于微信小程序的电影院买票选座系统(开发文档+数据库+源码)
java·开发语言·数据库·微信小程序·小程序·maven·intellij-idea
六月闻君2 小时前
MySQL 报错:1137 - Can‘t reopen table
数据库·mysql
SelectDB技术团队2 小时前
兼顾高性能与低成本,浅析 Apache Doris 异步物化视图原理及典型场景
大数据·数据库·数据仓库·数据分析·doris
inventecsh2 小时前
mongodb基础操作
数据库·mongodb
白云如幻2 小时前
SQL99版链接查询语法
数据库·sql·mysql