http://www.oschina.net/question/89964_65912
即使你认为自己已对 MySQL 的 LEFT JOIN 理解深刻,但我敢打赌,这篇文章肯定能让你学会点东西!
- ON 子句与 WHERE 子句的不同
- 一种更好地理解带有 WHERE ... IS NULL 子句的复杂匹配条件的简单方法
- Matching-Conditions 与 Where-conditions 的不同
关于 “A LEFT JOIN B ON 条件表达式” 的一点提醒
ON 条件(“A LEFT JOIN B ON 条件表达式”中的ON)用来决定如何从 B 表中检索数据行。
如果 B 表中没有任何一行数据匹配 ON 的条件,将会额外生成一行所有列为 NULL 的数据
在匹配阶段 WHERE 子句的条件都不会被使用。仅在匹配阶段完成以后,WHERE 子句条件才会被使用。它将从匹配阶段产生的数据中检索过滤。因此where子句中使用的字段前面select必须有。
让我们看一个 LFET JOIN 示例:
01 | mysql> CREATE TABLE `product` ( |
02 | `id` int (10) unsigned NOT NULL auto_increment, |
03 | `amount` int (10) unsigned default NULL , |
04 | PRIMARY KEY (`id`) |
05 | ) ENGINE=MyISAM AUTO_INCREMENT=5 DEFAULT CHARSET=latin1 |
06 | |
07 | mysql> CREATE TABLE `product_details` ( |
08 | `id` int (10) unsigned NOT NULL , |
09 | `weight` int (10) unsigned default NULL , |
10 | `exist` int (10) unsigned default NULL , |
11 | PRIMARY KEY (`id`) |
12 | ) ENGINE=MyISAM DEFAULT CHARSET=latin1 |
13 | |
14 | mysql> INSERT INTO product (id,amount) |
15 | VALUES (1,100),(2,200),(3,300),(4,400); |
16 | Query OK, 4 rows affected (0.00 sec) |
17 | Records: 4 Duplicates: 0 Warnings: 0 |
18 | |
19 | mysql> INSERT INTO product_details (id,weight,exist) |
20 | VALUES (2,22,0),(4,44,1),(5,55,0),(6,66,1); |
21 | Query OK, 4 rows affected (0.00 sec) |
22 | Records: 4 Duplicates: 0 Warnings: 0 |
23 | |
24 | mysql> SELECT * FROM product; |
25 | + ----+--------+ |
26 | | id | amount | |
27 | + ----+--------+ |
28 | | 1 | 100 | |
29 | | 2 | 200 | |
30 | | 3 | 300 | |
31 | | 4 | 400 | |
32 | + ----+--------+ |
33 | 4 rows in set (0.00 sec) |
34 | |
35 | mysql> SELECT * FROM product_details; |
36 | + ----+--------+-------+ |
37 | | id | weight | exist | |
38 | + ----+--------+-------+ |
39 | | 2 | 22 | 0 | |
40 | | 4 | 44 | 1 | |
41 | | 5 | 55 | 0 | |
42 | | 6 | 66 | 1 | |
43 | + ----+--------+-------+ |
44 | 4 rows in set (0.00 sec) |
45 | |
46 | mysql> SELECT * FROM product LEFT JOIN product_details |
47 | ON (product.id = product_details.id); |
48 | + ----+--------+------+--------+-------+ |
49 | | id | amount | id | weight | exist | |
50 | + ----+--------+------+--------+-------+ |
51 | | 1 | 100 | NULL | NULL | NULL | |
52 | | 2 | 200 | 2 | 22 | 0 | |
53 | | 3 | 300 | NULL | NULL | NULL | |
54 | | 4 | 400 | 4 | 44 | 1 | |
55 | + ----+--------+------+--------+-------+ |
56 | 4 rows in set (0.00 sec) |
ON 子句和 WHERE 子句有什么不同?
一个问题:下面两个查询的结果集有什么不同么?
1 | 1. SELECT * FROM product LEFT JOIN product_details |
2 | ON (product.id = product_details.id) |
3 | AND product_details.id=2; |
4 | 2. SELECT * FROM product LEFT JOIN product_details |
5 | ON (product.id = product_details.id) |
6 | WHERE product_details.id=2; |
用例子来理解最好不过了:
01 | mysql> SELECT * FROM product LEFT JOIN product_details |
02 | ON (product.id = product_details.id) |
03 | AND product_details.id=2; |
04 | + ----+--------+------+--------+-------+ |
05 | | id | amount | id | weight | exist | |
06 | + ----+--------+------+--------+-------+ |
07 | | 1 | 100 | NULL | NULL | NULL | |
08 | | 2 | 200 | 2 | 22 | 0 | |
09 | | 3 | 300 | NULL | NULL | NULL | |
10 | | 4 | 400 | NULL | NULL | NULL | |
11 | + ----+--------+------+--------+-------+ |
12 | 4 rows in set (0.00 sec) |
13 | |
14 | mysql> SELECT * FROM product LEFT JOIN product_details |
15 | ON (product.id = product_details.id) |
16 | WHERE product_details.id=2; |
17 | + ----+--------+----+--------+-------+ |
18 | | id | amount | id | weight | exist | |
19 | + ----+--------+----+--------+-------+ |
20 | | 2 | 200 | 2 | 22 | 0 | |
21 | + ----+--------+----+--------+-------+ |
22 | 1 row in set (0.01 sec) |
第一条查询使用 ON 条件决定了从 LEFT JOIN的 product_details表中检索符合的所有数据行。
第二条查询做了简单的LEFT JOIN,然后使用 WHERE 子句从 LEFT JOIN的数据中过滤掉不符合条件的数据行。
再来看一些示例:
01 | mysql> |
02 | mysql> SELECT * FROM product LEFT JOIN product_details |
03 | ON product.id = product_details.id |
04 | AND product.amount=100; |
05 | + ----+--------+------+--------+-------+ |
06 | | id | amount | id | weight | exist | |
07 | + ----+--------+------+--------+-------+ |
08 | | 1 | 100 | NULL | NULL | NULL | |
09 | | 2 | 200 | NULL | NULL | NULL | |
10 | | 3 | 300 | NULL | NULL | NULL | |
11 | | 4 | 400 | NULL | NULL | NULL | |
12 | + ----+--------+------+--------+-------+ |
13 | 4 rows in set (0.00 sec) |
所有来自product表的数据行都被检索到了,但没有在product_details表中匹配到记录(product.id = product_details.id AND product.amount=100 条件并没有匹配到任何数据)
01 | mysql> SELECT * FROM product LEFT JOIN product_details |
02 | ON (product.id = product_details.id) |
03 | AND product.amount=200; |
04 | + ----+--------+------+--------+-------+ |
05 | | id | amount | id | weight | exist | |
06 | + ----+--------+------+--------+-------+ |
07 | | 1 | 100 | NULL | NULL | NULL | |
08 | | 2 | 200 | 2 | 22 | 0 | |
09 | | 3 | 300 | NULL | NULL | NULL | |
10 | | 4 | 400 | NULL | NULL | NULL | |
11 | + ----+--------+------+--------+-------+ |
12 | 4 rows in set (0.01 sec) |
同样,所有来自product表的数据行都被检索到了,有一条数据匹配到了。
使用 WHERE ... IS NULL 子句的 LEFT JOIN
当你使用 WHERE ... IS NULL 子句时会发生什么呢?
如前所述,WHERE 条件查询发生在 匹配阶段之后,这意味着 WHERE ... IS NULL 子句将从匹配阶段后的数据中过滤掉不满足匹配条件的数据行。
纸面上看起来很清楚,但是当你在 ON 子句中使用多个条件时就会感到困惑了。
我总结了一种简单的方式来理解上述情况:
- 将 IS NULL 作为否定匹配条件
- 使用 !(A and B) == !A OR !B 逻辑判断
看看下面的示例:
01 | mysql> SELECT a.* FROM product a LEFT JOIN product_details b |
02 | ON a.id=b.id AND b.weight!=44 AND b.exist=0 |
03 | WHERE b.id IS NULL ; |
04 | + ----+--------+ |
05 | | id | amount | |
06 | + ----+--------+ |
07 | | 1 | 100 | |
08 | | 3 | 300 | |
09 | | 4 | 400 | |
10 | + ----+--------+ |
11 | 3 rows in set (0.00 sec) |
让我们检查一下 ON 匹配子句:
1 | (a.id=b.id) AND (b.weight!=44) AND (b.exist=0) |
我们可以把 IS NULL 子句 看作是否定匹配条件。
这意味着我们将检索到以下行:
1 | !( exist(b.id that equals to a.id) AND b.weight !=44 AND b.exist=0 ) |
2 | !exist(b.id that equals to a.id) || !(b.weight !=44) || !(b.exist=0) |
3 | !exist(b.id that equals to a.id) || b.weight =44 || b.exist=1 |
就像在C语言中的逻辑 AND 和 逻辑 OR表达式一样,其操作数是从左到右求值的。如果第一个参数做够判断操作结果,那么第二个参数便不会被计算求值(短路效果)
看看别的示例:
01 | mysql> SELECT a.* FROM product a LEFT JOIN product_details b |
02 | ON a.id=b.id AND b.weight!=44 AND b.exist=1 |
03 | WHERE b.id IS NULL ; |
04 | + ----+--------+ |
05 | | id | amount | |
06 | + ----+--------+ |
07 | | 1 | 100 | |
08 | | 2 | 200 | |
09 | | 3 | 300 | |
10 | | 4 | 400 | |
11 | + ----+--------+ |
12 | 4 rows in set (0.00 sec) |
Matching-Conditions 与 Where-conditions 之战
如果你吧基本的查询条件放在 ON 子句中,把剩下的否定条件放在 WHERE 子句中,那么你会获得相同的结果。
例如,你可以不这样写:
1 | SELECT a.* FROM product a LEFT JOIN product_details b |
2 | ON a.id=b.id AND b.weight!=44 AND b.exist=0 |
3 | WHERE b.id IS NULL ; |
你可以这样写:
1 | SELECT a.* FROM product a LEFT JOIN product_details b |
2 | ON a.id=b.id |
3 | WHERE b.id is null OR b.weight=44 OR b.exist=1; |
01 | mysql> SELECT a.* FROM product a LEFT JOIN product_details b |
02 | ON a.id=b.id |
03 | WHERE b.id is null OR b.weight=44 OR b.exist=1; |
04 | + ----+--------+ |
05 | | id | amount | |
06 | + ----+--------+ |
07 | | 1 | 100 | |
08 | | 3 | 300 | |
09 | | 4 | 400 | |
10 | + ----+--------+ |
11 | 3 rows in set (0.00 sec) |
你可以不这样写:
1 | SELECT a.* FROM product a LEFT JOIN product_details b |
2 | ON a.id=b.id AND b.weight!=44 AND b.exist!=0 |
3 | WHERE b.id IS NULL ; |
可以这样写:
1 | SELECT a.* FROM product a LEFT JOIN product_details b |
2 | ON a.id=b.id |
3 | WHERE b.id is null OR b.weight=44 OR b.exist=0; |
01 | mysql> SELECT a.* FROM product a LEFT JOIN product_details b |
02 | ON a.id=b.id |
03 | WHERE b.id is null OR b.weight=44 OR b.exist=0; |
04 | + ----+--------+ |
05 | | id | amount | |
06 | + ----+--------+ |
07 | | 1 | 100 | |
08 | | 2 | 200 | |
09 | | 3 | 300 | |
10 | | 4 | 400 | |
11 | + ----+--------+ |
12 | 4 rows in set (0.00 sec) |
这些查询真的效果一样?
如果你只需要第一个表中的数据的话,这些查询会返回相同的结果集。有一种情况就是,如果你从 LEFT JOIN的表中检索数据时,查询的结果就不同了。
如前所属,WHERE 子句是在匹配阶段之后用来过滤的。
例如:
01 | mysql> SELECT * FROM product a LEFT JOIN product_details b |
02 | ON a.id=b.id AND b.weight!=44 AND b.exist=1 |
03 | WHERE b.id is null ; |
04 | + ----+--------+------+--------+-------+ |
05 | | id | amount | id | weight | exist | |
06 | + ----+--------+------+--------+-------+ |
07 | | 1 | 100 | NULL | NULL | NULL | |
08 | | 2 | 200 | NULL | NULL | NULL | |
09 | | 3 | 300 | NULL | NULL | NULL | |
10 | | 4 | 400 | NULL | NULL | NULL | |
11 | + ----+--------+------+--------+-------+ |
12 | 4 rows in set (0.00 sec) |
13 | |
14 | mysql> SELECT * FROM product a LEFT JOIN product_details b |
15 | ON a.id=b.id |
16 | WHERE b.id IS NULL OR b.weight=44 OR b.exist=0; |
17 | + ----+--------+------+--------+-------+ |
18 | | id | amount | id | weight | exist | |
19 | + ----+--------+------+--------+-------+ |
20 | | 1 | 100 | NULL | NULL | NULL | |
21 | | 2 | 200 | 2 | 22 | 0 | |
22 | | 3 | 300 | NULL | NULL | NULL | |
23 | | 4 | 400 | 4 | 44 | 1 | |
24 | + ----+--------+------+--------+-------+ |
25 | 4 rows in set (0.00 sec) |
总附注:
如果你使用 LEFT JOIN 来寻找在一些表中不存在的记录,你需要做下面的测试:WHERE 部分的col_name IS NULL(其中 col_name 列被定义为 NOT NULL),MYSQL 在查询到一条匹配 LEFT JOIN 条件后将停止搜索更多行(在一个特定的组合键下)。
/