[MySQL]多表關聯查詢技巧

示例表A:

author_id	author_name
1	Kimmy
2	Abel
3	Bill
4	Berton

示例表B:

book_id	author_id	start_date	end_date
9	1	2017-09-25 21:16:04	2017-09-25 21:16:06
10	3
11	2	2017-09-25 21:21:46	2017-09-25 21:21:47
12	1
13	8

示例表C:

order_id	book_id	price	order_date
1	9	0.2	2017-09-24 21:21:46
2	9	0.6	2017-09-25 21:16:04
3	11	0.1	2017-09-25 21:21:46

在以上表中執行AB表關聯

SELECT `authors`.*, `books`.book_id FROM `authors` 
LEFT JOIN `books` ON `authors`.author_id = `books`.author_id 

結果

author_id	author_name	book_id
1	Kimmy	9
3	Bill	10
2	Abel	11
1	Kimmy	12
4	Berton

結果出現了2條author_id爲1的記錄，由於右表中存在了兩條關聯author_id=1的行

右邊出現N條關聯左邊的記錄，結果就會相應出現N條關聯了右表出現的記錄

在以上表中執行ABC表關聯

SELECT `authors`.*, `books`.book_id, `orders`.order_id, `orders`.price FROM `authors` 
LEFT JOIN `books` ON `authors`.author_id = `books`.author_id 
LEFT JOIN `orders` ON `books`.book_id = `orders`.book_id

結果

author_id	author_name	book_id	order_id	order_price
1	Kimmy	9	1	0.2
1	Kimmy	9	2	0.6
2	Abel	11	3	0.1
3	Bill	10
1	Kimmy	12
4	Berton

結果出現了3條author_id=1的記錄，由於authors第一次關聯了books表book_id爲9和12的book關聯了author_id爲1的做者，而book_id爲9的書本則關聯了兩個orders記錄，因此結果集包含3條author_id爲1的記錄

能夠運用

count(),sum()

等函數經過

group by

來統計結果

SELECT `authors`.*, sum(`orders`.price) FROM `authors` 
LEFT JOIN `books` ON `authors`.author_id = `books`.author_id 
LEFT JOIN `orders` ON `books`.book_id = `orders`.book_id
GROUP BY `books`.book_id

結果集會基於book_id來統計每一本書的訂單總額

author_id	author_name	book_id	sum(order_price)
4	Berton
1	Kimmy	9	0.80
3	Bill	10
2	Abel	11	0.10
1	Kimmy	12

 

book_id爲9的訂單總額爲0.80，而且9的記錄從多條合併爲1條。

 

多條件join

SELECT `authors`.*, `books`.book_id, `orders`.order_id, sum(`orders`.price) FROM `authors` 
LEFT JOIN `books` ON `authors`.author_id = `books`.author_id 
LEFT JOIN `orders` ON `books`.book_id = `orders`.book_id AND `orders`.order_date >= `books`.start_date AND `orders`.order_date <= `books`.end_date
GROUP BY `books`.book_id

選取在必定時間區間範圍內的order訂單，能夠看到訂單order_id爲1的訂單再也不歸入book_id爲9的統計當中，由於它的時間區間不符合join條件

author_id	author_name	book_id	order_id	sum(`order`.price)
4	Berton
1	Kimmy	9	2	0.60
3	Bill	10
2	Abel	11	3	0.10
1	Kimmy	12

關於where的使用，看下面示範

SELECT `authors`.*, `books`.book_id, `orders`.order_id, sum(`orders`.price) AS prices FROM `authors`
LEFT JOIN `books` ON `authors`.author_id = `books`.author_id
LEFT JOIN `orders` ON `books`.book_id = `orders`.book_id AND `orders`.order_date >= `books`.start_date AND `orders`.order_date <= `books`.end_date 
WHERE prices is not NULL
GROUP BY `books`.book_id

以上語句假設選取price不爲空的記錄，致使了一個錯誤的出現

[Err] 1054 - Unknown column 'prices' in 'where clause'

由於where不能用於選取列的AS別名判斷，MYSQL的處理機制是先進行選取，再進行篩選，在選取階段就啓用了where條件，由於這時並不存在prices的篩選結果後才產生的字段，因此這裏會拋出錯誤

咱們能夠這樣作

SELECT `authors`.*, `books`.book_id, `orders`.order_id, sum(`orders`.price) AS prices FROM `authors`
LEFT JOIN `books` ON `authors`.author_id = `books`.author_id
LEFT JOIN `orders` ON `books`.book_id = `orders`.book_id AND `orders`.order_date >= `books`.start_date AND `orders`.order_date <= `books`.end_date 
WHERE `orders`.price is not NULL
GROUP BY `books`.book_id

選取階段order表是存在price字段的，因此只有price不爲空的記錄纔會被選取

author_id	author_name	book_id	order_id	prices
2	Abel	11	3	0.10
1	Kimmy	9	2	0.60

運用

having

對那些沒法進行 WHERE 的AS別名的字段進行一些篩選查詢

SELECT `authors`.*, `books`.book_id, sum(`orders`.price)AS prices FROM `authors` 
LEFT JOIN `books` ON `authors`.author_id = `books`.author_id 
LEFT JOIN `orders` ON `books`.book_id = `orders`.book_id
GROUP BY `books`.book_id 
HAVING prices > 0.1

這時只有sum爲0.8的結果被選中 

author_id	author_name	book_id	sum(order_price)
1	Kimmy	9	0.80

對於組合其餘語法查詢，也是沒問題的

SELECT `authors`.*, `books`.book_id, sum(`orders`.price)AS prices FROM `authors` 
LEFT JOIN `books` ON `authors`.author_id = `books`.author_id 
LEFT JOIN `orders` ON `books`.book_id = `orders`.book_id
GROUP BY `books`.book_id 
HAVING prices >= 0.1 
ORDER BY prices asc
LIMIT 1,1