玩轉Mysql系列 – 第13篇:細說NULL導致的神坑,讓人防不勝防

  • 2019 年 10 月 7 日
  • 筆記

這是Mysql系列第11篇。

環境:mysql5.7.25,cmd命令中進行演示。

當數據的值為NULL的時候,可能出現各種意想不到的效果,讓人防不勝防,我們來看看NULL導致的各種神坑,如何避免?比較運算符中使用NULL

認真看下面的效果

mysql> select 1>NULL;  +--------+  | 1>NULL |  +--------+  |   NULL |  +--------+  1 row in set (0.00 sec)    mysql> select 1<NULL;  +--------+  | 1<NULL |  +--------+  |   NULL |  +--------+  1 row in set (0.00 sec)    mysql> select 1<>NULL;  +---------+  | 1<>NULL |  +---------+  |    NULL |  +---------+  1 row in set (0.00 sec)    mysql> select 1>NULL;  +--------+  | 1>NULL |  +--------+  |   NULL |  +--------+  1 row in set (0.00 sec)    mysql> select 1<NULL;  +--------+  | 1<NULL |  +--------+  |   NULL |  +--------+  1 row in set (0.00 sec)    mysql> select 1>=NULL;  +---------+  | 1>=NULL |  +---------+  |    NULL |  +---------+  1 row in set (0.00 sec)    mysql> select 1<=NULL;  +---------+  | 1<=NULL |  +---------+  |    NULL |  +---------+  1 row in set (0.00 sec)    mysql> select 1!=NULL;  +---------+  | 1!=NULL |  +---------+  |    NULL |  +---------+  1 row in set (0.00 sec)    mysql> select 1<>NULL;  +---------+  | 1<>NULL |  +---------+  |    NULL |  +---------+  1 row in set (0.00 sec)    mysql> select NULL=NULL,NULL!=NULL;  +-----------+------------+  | NULL=NULL | NULL!=NULL |  +-----------+------------+  |      NULL |       NULL |  +-----------+------------+  1 row in set (0.00 sec)    mysql> select 1 in (null),1 not in (null),null in (null),null not in (null);  +-------------+-----------------+----------------+--------------------+  | 1 in (null) | 1 not in (null) | null in (null) | null not in (null) |  +-------------+-----------------+----------------+--------------------+  |        NULL |            NULL |           NULL |               NULL |  +-------------+-----------------+----------------+--------------------+  1 row in set (0.00 sec)    mysql> select 1=any(select null),null=any(select null);  +--------------------+-----------------------+  | 1=any(select null) | null=any(select null) |  +--------------------+-----------------------+  |               NULL |                  NULL |  +--------------------+-----------------------+  1 row in set (0.00 sec)    mysql> select 1=all(select null),null=all(select null);  +--------------------+-----------------------+  | 1=all(select null) | null=all(select null) |  +--------------------+-----------------------+  |               NULL |                  NULL |  +--------------------+-----------------------+  1 row in set (0.00 sec)  

結論:任何值和NULL使用運算符(>、<、>=、<=、!=、<>)或者(in、not in、any/some、all)比較時,返回值都為NULL,NULL作為布爾值的時候,不為1也不為0。

準備數據

mysql> create table test1(a int,b int);  Query OK, 0 rows affected (0.01 sec)    mysql> insert into test1 values (1,1),(1,null),(null,null);  Query OK, 3 rows affected (0.00 sec)  Records: 3  Duplicates: 0  Warnings: 0    mysql> select * from test1;  +------+------+  | a    | b    |  +------+------+  |    1 |    1 |  |    1 | NULL |  | NULL | NULL |  +------+------+  3 rows in set (0.00 sec)  

上面3條數據,認真看一下,特別是注意上面NULL的記錄。

IN、NOT IN和NULL比較

IN和NULL比較

mysql> select * from test1;  +------+------+  | a    | b    |  +------+------+  |    1 |    1 |  |    1 | NULL |  | NULL | NULL |  +------+------+  3 rows in set (0.00 sec)    mysql> select * from test1 where a in (null);  Empty set (0.00 sec)    mysql> select * from test1 where a in (null,1);  +------+------+  | a    | b    |  +------+------+  |    1 |    1 |  |    1 | NULL |  +------+------+  2 rows in set (0.00 sec)  

結論:當IN和NULL比較時,無法查詢出為NULL的記錄。

NOT IN 和NULL比較

mysql> select * from test1 where a not in (1);  Empty set (0.00 sec)    mysql> select * from test1 where a not in (null);  Empty set (0.00 sec)    mysql> select * from test1 where a not in (null,2);  Empty set (0.00 sec)    mysql> select * from test1 where a not in (2);  +------+------+  | a    | b    |  +------+------+  |    1 |    1 |  |    1 | NULL |  +------+------+  2 rows in set (0.00 sec)  

結論:當NOT IN 後面有NULL值時,不論什麼情況下,整個sql的查詢結果都為空。

EXISTS、NOT EXISTS和NULL比較

mysql> select * from test2;  +------+------+  | a    | b    |  +------+------+  |    1 |    1 |  |    1 | NULL |  | NULL | NULL |  +------+------+  3 rows in set (0.00 sec)    mysql> select * from test1 t1 where exists (select * from test2 t2 where t1.a = t2.a);  +------+------+  | a    | b    |  +------+------+  |    1 |    1 |  |    1 | NULL |  +------+------+  2 rows in set (0.00 sec)    mysql> select * from test1 t1 where not exists (select * from test2 t2 where t1.a = t2.a);  +------+------+  | a    | b    |  +------+------+  | NULL | NULL |  +------+------+  1 row in set (0.00 sec)  

上面我們複製了表test1創建了表test2。

查詢語句中使用exists、not exists對比test1.a=test2.a,因為=不能比較NULL,結果和預期一致。

判斷NULL只能用IS NULL、IS NOT NULL

mysql> select 1 is not null;  +---------------+  | 1 is not null |  +---------------+  |             1 |  +---------------+  1 row in set (0.00 sec)    mysql> select 1 is null;  +-----------+  | 1 is null |  +-----------+  |         0 |  +-----------+  1 row in set (0.00 sec)    mysql> select null is null;  +--------------+  | null is null |  +--------------+  |            1 |  +--------------+  1 row in set (0.00 sec)    mysql> select null is not null;  +------------------+  | null is not null |  +------------------+  |                0 |  +------------------+  1 row in set (0.00 sec)  

看上面的效果,返回的結果為1或者0。

結論:判斷是否為空只能用IS NULL、IS NOT NULL。

聚合函數中NULL的坑

示例

mysql> select count(a),count(b),count(*) from test1;  +----------+----------+----------+  | count(a) | count(b) | count(*) |  +----------+----------+----------+  |        2 |        1 |        3 |  +----------+----------+----------+  1 row in set (0.00 sec)  

count(a)返回了2行記錄,a欄位為NULL的沒有統計出來。

count(b)返回了1行記錄,為NULL的2行記錄沒有統計出來。

count(*)可以統計所有數據,不論欄位的數據是否為NULL。

再繼續看

mysql> select * from test1 where a is null;  +------+------+  | a    | b    |  +------+------+  | NULL | NULL |  +------+------+  1 row in set (0.00 sec)    mysql> select count(a) from test1 where a is null;  +----------+  | count(a) |  +----------+  |        0 |  +----------+  1 row in set (0.00 sec)  

上面第1個sql使用is null查詢出了結果,第2個sql中count(a)返回的是0行。

結論:count(欄位)無法統計欄位為NULL的值,count(*)可以統計值為null的行。

NULL不能作為主鍵的值

mysql> create table test3(a int primary key,b int);  Query OK, 0 rows affected (0.01 sec)    mysql> insert into test3 values (null,1);  ERROR 1048 (23000): Column 'a' cannot be null  

上面我們創建了一個表test3,欄位a未指定不能為空,插入了一條NULL的數據,報錯原因:a 欄位的值不能為NULL,我們看一下表的創建語句:

mysql> show create table test3;  +-------+------------+  | Table | Create Table      |  +-------+------------+  | test3 | CREATE TABLE `test3` (    `a` int(11) NOT NULL,    `b` int(11) DEFAULT NULL,    PRIMARY KEY (`a`)  ) ENGINE=InnoDB DEFAULT CHARSET=utf8  +-------+------------+  1 row in set (0.00 sec)  

從上面的腳本可以看出,當欄位為主鍵的時候,欄位會自動設置為not null

結論:當欄位為主鍵的時候,欄位會自動設置為not null。

看了上面這些還是比較暈,NULL的情況確實比較難以處理,容易出錯,最有效的方法就是避免使用NULL。所以,強烈建議創建欄位的時候欄位不允許為NULL,設置一個默認值。

總結

  • NULL作為布爾值的時候,不為1也不為0
  • 任何值和NULL使用運算符(>、<、>=、<=、!=、<>)或者(in、not in、any/some、all),返回值都為NULL
  • 當IN和NULL比較時,無法查詢出為NULL的記錄
  • 當NOT IN 後面有NULL值時,不論什麼情況下,整個sql的查詢結果都為空
  • 判斷是否為空只能用IS NULL、IS NOT NULL
  • count(欄位)無法統計欄位為NULL的值,count(*)可以統計值為null的行
  • 當欄位為主鍵的時候,欄位會自動設置為not null
  • NULL導致的坑讓人防不勝防,強烈建議創建欄位的時候欄位不允許為NULL,給個默認值