SQL刷题总结 SQL Leetcode Review

---

Design a Relational Database

Designing a Relational Database and Creating an Entity Relationship Diagram

https://towardsdatascience.com/designing-a-relational-database-and-creating-an-entity-relationship-diagram-89c1c19320b2

---

join

inner join

SELECT column_name(s)
FROM table1
INNER JOIN table2
ON table1.column_name = table2.column_name;

或者

SELECT column_name(s)
FROM table1, table2
where table1.column_name = table2.column_name;

left join

SELECT column_name(s)
FROM table1
LEFT JOIN table2
ON table1.column_name = table2.column_name;

eg. 181

---

cross join

笛卡尔积

select *
from A cross join B

eg. 1280

---

group by having

SELECT column_name(s)
FROM table_name
WHERE condition
GROUP BY column_name(s)
HAVING condition
ORDER BY column_name(s);

eg. 182

delete

DELETE xxx FROM table_name WHERE condition;

eg. 196

distinct

从“列”中取出不重复的元素

SELECT DISTINCT column1, column2, ...
FROM table_name;

eg. 176

Case

CASE
    WHEN condition1 THEN result1
    WHEN condition2 THEN result2
    WHEN conditionN THEN resultN
    ELSE result
END;

Union

Union all 保留重复行
Union 不保留重复行

各种Rank

https://blog.csdn.net/shaiguchun9503/article/details/82349050

1. rank_number()
   将select查询到的数据进行排序，每一条数据加一个序号
   他不能用做于学生成绩的排名，一般多用于分页查询，比如查询前10个 查询10-100个学生。

2. rank()
   简单来说rank函数就是对查询出来的记录进行排名。
   与row_number函数不同的是，rank函数考虑到了over子句中排序字段值相同的情况，
   如果使用rank函数来生成序号，over子句中排序字段值相同的序号是一样的，后面字段值不相同的序号将跳过相同的排名号排下一个。

3. dense_rank()
   DENSE_RANK()密集的排名他和RANK()区别在于，排名的连续性，DENSE_RANK()排名是连续的，RANK()是跳跃的排名，所以一般情况下用的排名函数就是RANK()。

DENSE_RANK ( ) OVER ( [ <partition_by_clause> ] < order_by_clause > )  

eg. 178, 1321

4. ntile
   NTILE()函数是将有序分区中的行分发到指定数目的组中，各个组有编号，编号从1开始，就像我们说的’分区’一样 ，分为几个区，一个区会有多少个。

分区 - PARTITION BY

分区内排名
https://www.cnblogs.com/hxfcodelife/p/10226934.html

rank() over(partition by xxx order by aaa desc)

ex: 1082

Limit, offset

https://www.sqltutorial.org/sql-limit/

Limit: 限制输出行数
offset: 跳过多少行

注意:

• MySQL里offset必须和limit连在一起用(limit可以单独使用),如果要达到同样的效果,MySQL最好用row_number()
• SQL server 可以单独使用offset

SELECT 
    employee_id, first_name, last_name
FROM
    employees
ORDER BY first_name
LIMIT 5 OFFSET 3;

或者

SELECT 
    employee_id, 
    first_name, 
    last_name
FROM
    employees
ORDER BY 
	first_name
LIMIT 3 , 5;

Update

eg. 627

UPDATE, 
UPDATE table_name
SET column1 = value1, column2 = value2, ...
WHERE condition;

累计求和

两种方式

1、sum(weight) over(order by id asc)
也可以加上 between unbounded preceding and current row
2、利用不等式join，group by

eg. 534, 1204

---

Find the Start / End Number of Continuous Ranges

ROW_NUMBER()

eg. 1285

---

First_value()

SELECT DISTINCT
    player_id,
    FIRST_VALUE(device_id) OVER (
        PARTITION BY player_id
        ORDER BY event_date ASC
    ) AS device_id
FROM Activity
ORDER BY player_id;

eg. 512

null, ‘’, 0

这三个值是不同的。

对于0，直接用=检查即可。
对于“”，也可以直接用=检查。但是需要注意的是，0 = “” 是成立的。所以需要单独删除0的情况。
对于Null，直接使用 is null 或者 is not null

eg: 176

with temp  as (
    select distinct salary, dense_rank() over(order by salary desc) as r
    from Employee
)

select IF(
    (select temp.salary from temp where temp.r = 2) = '' and (select temp.salary from temp where temp.r = 2) != 0, 
    null, 
    (select temp.salary from temp where temp.r = 2)) as SecondHighestSalary

---

null, ifnull

isnull(value1,value2) (SQL server)
ifnull(value1,value2) (MySQL)

1、假如value1为null的话，返回value2，否则返回value1
2、value1与value2的数据类型必须一致。

如果希望有 “” 的时候，用null代替，则可以加一些表达式。比如max

ifnull 也会删除“”（empty）
https://stackoverflow.com/questions/17832906/how-to-check-if-field-is-null-or-empty-in-mysql

eg: 176

• NULL 不参与 > < = 的运算, 使用这些符号会输出null，需要注意使用前进行null的检查
• 用 >, <, =, != 进行筛选的时候，原数据中的 null 不会被保留 eg 577，584
• count：使用count的时候，如果用count(*)，那么就是数行数，这时不管有没有null都不影响输出。如果是count(var)，那么var中的null是不参与计数的。

blank & Null

有时候会select不会东西来,结果就是一个空.
但是我们不希望有空, 而希望是Null, 这个时候可以使用一些函数在结果上,比如max(), 然后结果就会变成Null

eg. 176

with temp as (
    select distinct salary, dense_rank() over(order by salary desc) as r
    from Employee
)

select max(salary) as SecondHighestSalary
from temp
where r = 2

---

Date

between

between表示的是闭区间，但是需要注意如果包涵具体时间的话，按照以下方式。

a between '2020-01-01' and '2020-01-10'
a >= 2020-01-01 00:00:00
a <= 2020-01-10 00:00:00

如果需要具体选择，有以下几种方式。
https://stackoverflow.com/questions/16347649/sql-between-not-inclusive/16347680

比如需要选取 created_at looks like 2013-05-01 22:25:19

• 使用cast，将原本的 datetime 全部改成 date

SELECT *
FROM Cases
WHERE cast(created_at as date) BETWEEN '2013-05-01' AND '2013-05-01'

---

• 直接用新一天的开区间选择

SELECT *
FROM Cases
WHERE created_at >= '2013-05-01' AND created_at < '2013-05-02'

• 注意：如果只有年月日，那么就是默认时间是“00:00:00”。比如“2019-12-01”，也就是“2019-12-01 00:00:00”

eg. 1294

---

选择year/month/date

• postgresql

https://www.postgresqltutorial.com/postgresql-extract/

EXTRACT(field FROM source)

source: DAY, QUARTER, MONTH …

• MySQL

SELECT YEAR("2017-06-15");
SELECT QUARTER("2017-06-15");
SELECT MONTH("2017-06-15");
SELECT DAY("2017-06-15");

选择“year-month”

DATE_FORMAT(trans_date, '%Y-%m')

LEFT(trans_date, 7)

eg. 1193

日期相减

• interval 1 day (不能直接减数字，或者两个日期直接相减和数字比较, 因为相差一年也会出现相减等于 1)
• TO_DAYS: TO_DAYS(wt1.DATE) - TO_DAYS(wt2.DATE)=1
• DATEDIFF: MySQL: DATEDIFF(a.Date, b.Date)=1，a.Date - b.Date = 1，
• DATEDIFF: SQL Server：DATEDIFF(day, a.Date, b.Date)=1, b.Date - a.Date = 1

DATE_DIFF

DATE_DIFF(time1, time2, day)

注意：将时间全部转成date，然后计算日期的间隔；和直接用timestamp算会有差异

SELECT
  DATE_DIFF(
      DATE "2021-12-30", 
      EXTRACT(date
    FROM
      CAST("2021-12-29 23:59:59" AS TIMESTAMP)), day) AS day1,

  DATE_DIFF(
      CAST("2021-12-30" AS TIMESTAMP), 
      CAST("2021-12-29 00:00:01" AS TIMESTAMP), day) AS day2

Split, UNNEST

  SELECT
    split_tags,
    tags
  FROM
    `bigquery-public-data.stackoverflow.posts_questions`,
    UNNEST(SPLIT(tags, '|')) AS split_tags
  WHERE
    EXTRACT(year
    FROM
      creation_date) = 2020 )

Pivot

• MySQL没有pivot函数，需要用group by + case来构建每一列
• SQL Server可以用pivot函数

https://www.techonthenet.com/sql_server/pivot.php

eg. 1179

function 函数

eg. 2205

CREATE FUNCTION getUserIDs(startDate DATE, endDate DATE, minAmount INT) RETURNS INT
BEGIN
  RETURN (
      # Write your MySQL query statement below.
      select count(distinct user_id) as user_cnt
          from Purchases
          where time_stamp >= startDate and time_stamp <= endDate and amount >= minAmount
  );
END

eg 2230

CREATE PROCEDURE getUserIDs(startDate DATE, endDate DATE, minAmount INT)
BEGIN
	# Write your MySQL query statement below.
	select distinct user_id as user_id
    from Purchases
    where time_stamp between startDate and endDate and amount >= minAmount
    order by user_id asc;
END