第07章_單行函式

講師：尚矽谷-宋紅康（江湖人稱：康師傅）

官網：http://www.atguigu.com

1. 函式的理解

1.1 什麼是函式

函式在計算機語言的使用中貫穿始終，函式的作用是什麼呢？它可以把我們經常使用的程式碼封裝起來，需要的時候直接呼叫即可。這樣既提高了程式碼效率，又提高了可維護性。在 SQL 中我們也可以使用函式對檢索出來的資料進行函式操作。使用這些函式，可以極大地提高使用者對資料庫的管理效率。

從函式定義的角度出發，我們可以將函式分成內建函式和自定義函式。在 SQL 語言中，同樣也包括了內建函式和自定義函式。內建函式是系統內建的通用函式，而自定義函式是我們根據自己的需要編寫的，本章及下一章講解的是 SQL 的內建函式。

1.2 不同DBMS函式的差異

我們在使用 SQL 語言的時候，不是直接和這門語言打交道，而是通過它使用不同的資料庫軟體，即 DBMS。DBMS 之間的差異性很大，遠大於同一個語言不同版本之間的差異。實際上，只有很少的函式是被 DBMS 同時支援的。比如，大多數 DBMS 使用（||）或者（+）來做拼接符，而在 MySQL 中的字串拼接函式為concat()。大部分 DBMS 會有自己特定的函式，這就意味著採用 SQL 函式的程式碼可移植性是很差的，因此在使用函式的時候需要特別注意。

1.3 MySQL的內建函式及分類

MySQL提供了豐富的內建函式，這些函式使得資料的維護與管理更加方便，能夠更好地提供資料的分析與統計功能，在一定程度上提高了開發人員進行資料分析與統計的效率。

MySQL提供的內建函式從實現的功能角度可以分為數值函式、字串函式、日期和時間函式、流程控制函式、加密與解密函式、獲取MySQL資訊函式、聚合函式等。這裡，我將這些豐富的內建函式再分為兩類：單行函式、聚合函式（或分組函式）。

兩種SQL函式

單行函式

• 操作資料物件
• 接受引數返回一個結果
• 只對一行進行變換
• 每行返回一個結果
• 可以巢狀
• 引數可以是一列或一個值

2. 數值函式

2.1 基本函式

舉例：

SELECT ABS(-123),ABS(32),SIGN(-23),SIGN(43),PI(),CEIL(32.32),CEILING(-43.23),FLOOR(32.32),
FLOOR(-43.23),MOD(12,5)
FROM DUAL;

SELECT RAND(),RAND(),RAND(10),RAND(10),RAND(-1),RAND(-1)
FROM DUAL;

SELECT ROUND(12.33),ROUND(12.343,2),ROUND(12.324,-1),TRUNCATE(12.66,1),TRUNCATE(12.66,-1)
FROM DUAL;

2.2 角度與弧度互換函式

SELECT RADIANS(30),RADIANS(60),RADIANS(90),DEGREES(2*PI()),DEGREES(RADIANS(90))
FROM DUAL;

2.3 三角函式

舉例：

ATAN2(M,N)函式返回兩個引數的反正切值。
與ATAN(X)函式相比，ATAN2(M,N)需要兩個引數，例如有兩個點point(x1,y1)和point(x2,y2)，使用ATAN(X)函式計算反正切值為ATAN((y2-y1)/(x2-x1))，使用ATAN2(M,N)計算反正切值則為ATAN2(y2-y1,x2-x1)。由使用方式可以看出，當x2-x1等於0時，ATAN(X)函式會報錯，而ATAN2(M,N)函式則仍然可以計算。

ATAN2(M,N)函式的使用示例如下：

SELECT SIN(RADIANS(30)),DEGREES(ASIN(1)),TAN(RADIANS(45)),DEGREES(ATAN(1)),DEGREES(ATAN2(1,1))
FROM DUAL;

2.4 指數與對數

mysql> SELECT POW(2,5),POWER(2,4),EXP(2),LN(10),LOG10(10),LOG2(4)
    -> FROM DUAL;
+----------+------------+------------------+-------------------+-----------+---------+
| POW(2,5) | POWER(2,4) | EXP(2)           | LN(10)            | LOG10(10) | LOG2(4) |
+----------+------------+------------------+-------------------+-----------+---------+
|       32 |         16 | 7.38905609893065 | 2.302585092994046 |         1 |       2 |
+----------+------------+------------------+-------------------+-----------+---------+
1 row in set (0.00 sec)

2.5 進位制間的轉換

mysql> SELECT BIN(10),HEX(10),OCT(10),CONV(10,2,8)
    -> FROM DUAL;
+---------+---------+---------+--------------+
| BIN(10) | HEX(10) | OCT(10) | CONV(10,2,8) |
+---------+---------+---------+--------------+
| 1010    | A       | 12      | 2            |
+---------+---------+---------+--------------+
1 row in set (0.00 sec)

3. 字串函式

注意：MySQL中，字串的位置是從1開始的。

舉例：

mysql> SELECT FIELD('mm','hello','msm','amma'),FIND_IN_SET('mm','hello,mm,amma')
    -> FROM DUAL;
+----------------------------------+-----------------------------------+
| FIELD('mm','hello','msm','amma') | FIND_IN_SET('mm','hello,mm,amma') |
+----------------------------------+-----------------------------------+
|                                0 |                                 2 |
+----------------------------------+-----------------------------------+
1 row in set (0.00 sec)

mysql> SELECT NULLIF('mysql','mysql'),NULLIF('mysql', '');
+-------------------------+---------------------+
| NULLIF('mysql','mysql') | NULLIF('mysql', '') |
+-------------------------+---------------------+
| NULL                    | mysql               |
+-------------------------+---------------------+
1 row in set (0.00 sec)

4. 日期和時間函式(！！！1、2、5、6！！！)

4.1 獲取日期、時間

舉例：

SELECT CURDATE(),CURTIME(),NOW(),SYSDATE()+0,UTC_DATE(),UTC_DATE()+0,UTC_TIME(),UTC_TIME()+0
FROM DUAL;

4.2 日期與時間戳的轉換

舉例：

mysql> SELECT UNIX_TIMESTAMP(now());
+-----------------------+
| UNIX_TIMESTAMP(now()) |
+-----------------------+
|            1576380910 |
+-----------------------+
1 row in set (0.01 sec)

mysql> SELECT UNIX_TIMESTAMP(CURDATE());
+---------------------------+
| UNIX_TIMESTAMP(CURDATE()) |
+---------------------------+
|                1576339200 |
+---------------------------+
1 row in set (0.00 sec)

mysql> SELECT UNIX_TIMESTAMP(CURTIME());
+---------------------------+
| UNIX_TIMESTAMP(CURTIME()) |
+---------------------------+
|                1576380969 |
+---------------------------+
1 row in set (0.00 sec)

mysql> SELECT UNIX_TIMESTAMP('2011-11-11 11:11:11')
+---------------------------------------+
| UNIX_TIMESTAMP('2011-11-11 11:11:11') |
+---------------------------------------+
|                            1320981071 |
+---------------------------------------+
1 row in set (0.00 sec)

mysql> SELECT FROM_UNIXTIME(1576380910);
+---------------------------+
| FROM_UNIXTIME(1576380910) |
+---------------------------+
| 2019-12-15 11:35:10       |
+---------------------------+
1 row in set (0.00 sec)

4.3 獲取月份、星期、星期數、天數等函式

舉例：

SELECT YEAR(CURDATE()),MONTH(CURDATE()),DAY(CURDATE()),
HOUR(CURTIME()),MINUTE(NOW()),SECOND(SYSDATE())
FROM DUAL;

SELECT MONTHNAME('2021-10-26'),DAYNAME('2021-10-26'),WEEKDAY('2021-10-26'),
QUARTER(CURDATE()),WEEK(CURDATE()),DAYOFYEAR(NOW()),
DAYOFMONTH(NOW()),DAYOFWEEK(NOW())
FROM DUAL;

4.4 日期的操作函式

EXTRACT(type FROM date)函式中type的取值與含義：

SELECT EXTRACT(MINUTE FROM NOW()),EXTRACT( WEEK FROM NOW()),
EXTRACT( QUARTER FROM NOW()),EXTRACT( MINUTE_SECOND FROM NOW())
FROM DUAL;

4.5 時間和秒鐘轉換的函式

舉例：

mysql> SELECT TIME_TO_SEC(NOW());
+--------------------+
| TIME_TO_SEC(NOW()) |
+--------------------+
|               78774 |
+--------------------+
1 row in set (0.00 sec)

mysql> SELECT SEC_TO_TIME(78774);
+--------------------+
| SEC_TO_TIME(78774) |
+--------------------+
| 21:52:54            |
+--------------------+
1 row in set (0.12 sec)

4.6 計算日期和時間的函式

第1組：

上述函式中type的取值：

舉例：

SELECT DATE_ADD(NOW(), INTERVAL 1 DAY) AS col1,DATE_ADD('2021-10-21 23:32:12',INTERVAL 1 SECOND) AS col2,
ADDDATE('2021-10-21 23:32:12',INTERVAL 1 SECOND) AS col3,
DATE_ADD('2021-10-21 23:32:12',INTERVAL '1_1' MINUTE_SECOND) AS col4,
DATE_ADD(NOW(), INTERVAL -1 YEAR) AS col5, #可以是負數
DATE_ADD(NOW(), INTERVAL '1_1' YEAR_MONTH) AS col6 #需要單引號
FROM DUAL;

SELECT DATE_SUB('2021-01-21',INTERVAL 31 DAY) AS col1,
SUBDATE('2021-01-21',INTERVAL 31 DAY) AS col2,
DATE_SUB('2021-01-21 02:01:01',INTERVAL '1 1' DAY_HOUR) AS col3
FROM DUAL;

第2組：

舉例：

SELECT ADDTIME(NOW(),20),SUBTIME(NOW(),30),SUBTIME(NOW(),'1:1:3'),DATEDIFF(NOW(),'2021-10-01'),
TIMEDIFF(NOW(),'2021-10-25 22:10:10'),FROM_DAYS(366),TO_DAYS('0000-12-25'),
LAST_DAY(NOW()),MAKEDATE(YEAR(NOW()),12),MAKETIME(10,21,23),PERIOD_ADD(20200101010101,10)
FROM DUAL;

mysql> SELECT ADDTIME(NOW(), 50);
+---------------------+
| ADDTIME(NOW(), 50)  |
+---------------------+
| 2019-12-15 22:17:47 |
+---------------------+
1 row in set (0.00 sec)

mysql> SELECT ADDTIME(NOW(), '1:1:1');
+-------------------------+
| ADDTIME(NOW(), '1:1:1') |
+-------------------------+
| 2019-12-15 23:18:46     |
+-------------------------+
1 row in set (0.00 sec)

mysql> SELECT SUBTIME(NOW(), '1:1:1');
+-------------------------+
| SUBTIME(NOW(), '1:1:1') |
+-------------------------+
| 2019-12-15 21:23:50     |
+-------------------------+
1 row in set (0.00 sec)

mysql> SELECT SUBTIME(NOW(), '-1:-1:-1'); 
+----------------------------+
| SUBTIME(NOW(), '-1:-1:-1') |
+----------------------------+
| 2019-12-15 22:25:11        |
+----------------------------+
1 row in set, 1 warning (0.00 sec)

mysql> SELECT FROM_DAYS(366);
+----------------+
| FROM_DAYS(366) |
+----------------+
| 0001-01-01     |
+----------------+
1 row in set (0.00 sec)

mysql> SELECT MAKEDATE(2020,1);
+------------------+
| MAKEDATE(2020,1) |
+------------------+
| 2020-01-01       |
+------------------+
1 row in set (0.00 sec)

mysql> SELECT MAKEDATE(2020,32);
+-------------------+
| MAKEDATE(2020,32) |
+-------------------+
| 2020-02-01        |
+-------------------+
1 row in set (0.00 sec)

mysql> SELECT MAKETIME(1,1,1);
+-----------------+
| MAKETIME(1,1,1) |
+-----------------+
| 01:01:01        |
+-----------------+
1 row in set (0.00 sec)

mysql> SELECT PERIOD_ADD(20200101010101,1);
+------------------------------+
| PERIOD_ADD(20200101010101,1) |
+------------------------------+
|               20200101010102 |
+------------------------------+
1 row in set (0.00 sec)

mysql> SELECT TO_DAYS(NOW());
+----------------+
| TO_DAYS(NOW()) |
+----------------+
|          737773 |
+----------------+
1 row in set (0.00 sec)

舉例：查詢 7 天內的新增使用者數有多少？

SELECT COUNT(*) as num FROM new_user WHERE TO_DAYS(NOW())-TO_DAYS(regist_time)<=7

!!!4.7 日期的格式化與解析!!!

SELECT DATE_FORMAT(NOW(),GET_FORMAT(DATETIME,'ISO'))

上述非GET_FORMAT函式中fmt引數常用的格式符：

GET_FORMAT函式中date_type和format_type引數取值如下：

舉例：

mysql> SELECT DATE_FORMAT(NOW(), '%H:%i:%s');
+--------------------------------+
| DATE_FORMAT(NOW(), '%H:%i:%s') |
+--------------------------------+
| 22:57:34                        |
+--------------------------------+
1 row in set (0.00 sec)

SELECT STR_TO_DATE('09/01/2009','%m/%d/%Y')
FROM DUAL;

SELECT STR_TO_DATE('20140422154706','%Y%m%d%H%i%s')
FROM DUAL;

SELECT STR_TO_DATE('2014-04-22 15:47:06','%Y-%m-%d %H:%i:%s')
FROM DUAL;

mysql> SELECT GET_FORMAT(DATE, 'USA');
+-------------------------+
| GET_FORMAT(DATE, 'USA') |
+-------------------------+
| %m.%d.%Y                |
+-------------------------+
1 row in set (0.00 sec)

SELECT DATE_FORMAT(NOW(),GET_FORMAT(DATE,'USA')),
FROM DUAL;

mysql> SELECT STR_TO_DATE('2020-01-01 00:00:00','%Y-%m-%d'); 
+-----------------------------------------------+
| STR_TO_DATE('2020-01-01 00:00:00','%Y-%m-%d') |
+-----------------------------------------------+
| 2020-01-01                                    |
+-----------------------------------------------+
1 row in set, 1 warning (0.00 sec)

5. 流程控制函式

流程處理函式可以根據不同的條件，執行不同的處理流程，可以在SQL語句中實現不同的條件選擇。MySQL中的流程處理函式主要包括IF()、IFNULL()和CASE()函式。

SELECT IF(1 > 0,'正確','錯誤')    
->正確

SELECT IFNULL(null,'Hello Word')
->Hello Word

SELECT CASE 
　　WHEN 1 > 0
　　THEN '1 > 0'
　　WHEN 2 > 0
　　THEN '2 > 0'
　　ELSE '3 > 0'
　　END
->1 > 0

SELECT CASE 1 
　　WHEN 1 THEN '我是1'
　　WHEN 2 THEN '我是2'
ELSE '你是誰'

SELECT employee_id,salary, CASE WHEN salary>=15000 THEN '高薪' 
				  WHEN salary>=10000 THEN '潛力股'  
				  WHEN salary>=8000 THEN '屌絲' 
				  ELSE '草根' END  "描述"
FROM employees; 

SELECT oid,`status`, CASE `status` WHEN 1 THEN '未付款' 
								   WHEN 2 THEN '已付款' 
								   WHEN 3 THEN '已發貨'  
								   WHEN 4 THEN '確認收貨'  
								   ELSE '無效訂單' END 
FROM t_order;

mysql> SELECT CASE WHEN 1 > 0 THEN 'yes' WHEN 1 <= 0 THEN 'no' ELSE 'unknown' END;
+---------------------------------------------------------------------+
| CASE WHEN 1 > 0 THEN 'yes' WHEN 1 <= 0 THEN 'no' ELSE 'unknown' END |
+---------------------------------------------------------------------+
| yes                                                                  |
+---------------------------------------------------------------------+
1 row in set (0.00 sec)

mysql> SELECT CASE WHEN 1 < 0 THEN 'yes' WHEN 1 = 0 THEN 'no' ELSE 'unknown' END;  
+--------------------------------------------------------------------+
| CASE WHEN 1 < 0 THEN 'yes' WHEN 1 = 0 THEN 'no' ELSE 'unknown' END |
+--------------------------------------------------------------------+
| unknown                                                             |
+--------------------------------------------------------------------+
1 row in set (0.00 sec)

mysql> SELECT CASE 1 WHEN 0 THEN 0 WHEN 1 THEN 1 ELSE -1 END;
+------------------------------------------------+
| CASE 1 WHEN 0 THEN 0 WHEN 1 THEN 1 ELSE -1 END |
+------------------------------------------------+
|                                               1 |
+------------------------------------------------+
1 row in set (0.00 sec)

mysql> SELECT CASE -1 WHEN 0 THEN 0 WHEN 1 THEN 1 ELSE -1 END;
+-------------------------------------------------+
| CASE -1 WHEN 0 THEN 0 WHEN 1 THEN 1 ELSE -1 END |
+-------------------------------------------------+
|                                               -1 |
+-------------------------------------------------+
1 row in set (0.00 sec)

SELECT employee_id,12 * salary * (1 + IFNULL(commission_pct,0))
FROM employees;

SELECT last_name, job_id, salary,
       CASE job_id WHEN 'IT_PROG'  THEN  1.10*salary
                   WHEN 'ST_CLERK' THEN  1.15*salary
                   WHEN 'SA_REP'   THEN  1.20*salary
       			   ELSE      salary END     "REVISED_SALARY"
FROM   employees;

練習：查詢部門號為 10,20, 30 的員工資訊, 若部門號為 10, 則列印其工資的 1.1 倍, 20 號部門, 則列印其工資的 1.2 倍, 30 號部門列印其工資的 1.3 倍數。

6. 加密與解密函式

加密與解密函式主要用於對資料庫中的資料進行加密和解密處理，以防止資料被他人竊取。這些函式在保證資料庫安全時非常有用。

可以看到，ENCODE(value,password_seed)函式與DECODE(value,password_seed)函式互為反函式。

舉例：

mysql> SELECT PASSWORD('mysql'), PASSWORD(NULL);
+-------------------------------------------+----------------+
| PASSWORD('mysql')                         | PASSWORD(NULL) |
+-------------------------------------------+----------------+
| *E74858DB86EBA20BC33D0AECAE8A8108C56B17FA |                |
+-------------------------------------------+----------------+
1 row in set, 1 warning (0.00 sec)

SELECT md5('123')
->202cb962ac59075b964b07152d234b70

SELECT SHA('Tom123')
->c7c506980abc31cc390a2438c90861d0f1216d50

mysql> SELECT ENCODE('mysql', 'mysql');
+--------------------------+
| ENCODE('mysql', 'mysql') |
+--------------------------+
| íg　¼　ìÉ                  |
+--------------------------+
1 row in set, 1 warning (0.01 sec)

mysql> SELECT DECODE(ENCODE('mysql','mysql'),'mysql');
+-----------------------------------------+
| DECODE(ENCODE('mysql','mysql'),'mysql') |
+-----------------------------------------+
| mysql                                   |
+-----------------------------------------+
1 row in set, 2 warnings (0.00 sec)

7. MySQL資訊函式

MySQL中內建了一些可以查詢MySQL資訊的函式，這些函式主要用於幫助資料庫開發或運維人員更好地對資料庫進行維護工作。

舉例：

mysql> SELECT DATABASE();
+------------+
| DATABASE() |
+------------+
| test       |
+------------+
1 row in set (0.00 sec)

mysql> SELECT DATABASE();
+------------+
| DATABASE() |
+------------+
| test       |
+------------+
1 row in set (0.00 sec)

mysql> SELECT USER(), CURRENT_USER(), SYSTEM_USER(),SESSION_USER();
+----------------+----------------+----------------+----------------+
| USER()         | CURRENT_USER() | SYSTEM_USER()  | SESSION_USER() |
+----------------+----------------+----------------+----------------+
| root@localhost | root@localhost | root@localhost | root@localhost |
+----------------+----------------+----------------+----------------+

mysql> SELECT CHARSET('ABC');
+----------------+
| CHARSET('ABC') |
+----------------+
| utf8mb4        |
+----------------+
1 row in set (0.00 sec)

mysql> SELECT COLLATION('ABC');
+--------------------+
| COLLATION('ABC')   |
+--------------------+
| utf8mb4_general_ci |
+--------------------+
1 row in set (0.00 sec)

8. 其他函式

MySQL中有些函式無法對其進行具體的分類，但是這些函式在MySQL的開發和運維過程中也是不容忽視的。

舉例：

# 如果n的值小於或者等於0，則只保留整數部分
mysql> SELECT FORMAT(123.123, 2), FORMAT(123.523, 0), FORMAT(123.123, -2); 
+--------------------+--------------------+---------------------+
| FORMAT(123.123, 2) | FORMAT(123.523, 0) | FORMAT(123.123, -2) |
+--------------------+--------------------+---------------------+
| 123.12             | 124                | 123                 |
+--------------------+--------------------+---------------------+
1 row in set (0.00 sec)

mysql> SELECT CONV(16, 10, 2), CONV(8888,10,16), CONV(NULL, 10, 2);
+-----------------+------------------+-------------------+
| CONV(16, 10, 2) | CONV(8888,10,16) | CONV(NULL, 10, 2) |
+-----------------+------------------+-------------------+
| 10000           | 22B8             | NULL              |
+-----------------+------------------+-------------------+
1 row in set (0.00 sec)

mysql> SELECT INET_ATON('192.168.1.100');
+----------------------------+
| INET_ATON('192.168.1.100') |
+----------------------------+
|                 3232235876 |
+----------------------------+
1 row in set (0.00 sec)

# 以“192.168.1.100”為例，計算方式為192乘以256的3次方，加上168乘以256的2次方，加上1乘以256，再加上100。

mysql> SELECT INET_NTOA(3232235876);
+-----------------------+
| INET_NTOA(3232235876) |
+-----------------------+
| 192.168.1.100         |
+-----------------------+
1 row in set (0.00 sec)

mysql> SELECT BENCHMARK(1, MD5('mysql'));
+----------------------------+
| BENCHMARK(1, MD5('mysql')) |
+----------------------------+
|                          0 |
+----------------------------+
1 row in set (0.00 sec)

mysql> SELECT BENCHMARK(1000000, MD5('mysql')); 
+----------------------------------+
| BENCHMARK(1000000, MD5('mysql')) |
+----------------------------------+
|                                0 |
+----------------------------------+
1 row in set (0.20 sec)

mysql> SELECT CHARSET('mysql'), CHARSET(CONVERT('mysql' USING 'utf8'));
+------------------+----------------------------------------+
| CHARSET('mysql') | CHARSET(CONVERT('mysql' USING 'utf8')) |
+------------------+----------------------------------------+
| utf8mb4          | utf8                                   |
+------------------+----------------------------------------+
1 row in set, 1 warning (0.00 sec)