Question

有5个表

用户个人数据（用户_记录= 120000）。
工作的用户（users_positions _ record = 150000）。
组织（组织_记录= 2500）。
分部（分部_记录= 3500）。
职位（职位_记录= 4500）。

我想加入这些表，但最好的查询是什么？

我写了这个查询，但速度很慢：

SELECT 
  * 
FROM
  (SELECT 
    p.* 
  FROM
    `user_positions` p 
    LEFT JOIN `organizations` u_o 
      ON u_o.`id` = p.`org_title` 
    LEFT JOIN `divisions` u_d 
      ON u_d.`id` = p.`division` 
    LEFT JOIN `positions` u_p 
      ON u_p.`id` = p.`position` 
      WHERE u_o.`status` = 1 
  GROUP BY p.`PRN`) u_a 
  LEFT JOIN `users` u 
    ON u.`PRN` = u_a.`PRN`

示例 - 简单结构

table users
-- PRN --- FirstName --- Age....
-- 001 --- david --- 24 ....
-- 052 --- george --- 27 ....


table users_positions
-- PRN --- Organization_ID --- Division_ID --- Positionn_ID....
-- 001 --- 1 --- 5 --- 1....
-- 001 --- 2 --- 10 --- 5....
-- 052 --- 1 --- 1 --- 1....

table organizations
-- ID --- Name ....
-- 1 --- test 1 ....
-- 2 --- test 2 ....

table divisions
-- ID --- Name --- Age....
-- 1 --- IT Department ....
-- 5 --- Sale Department ....
-- 10 --- Administration Department ....

table positions
-- ID --- Name --- Age....
-- 1 --- Programmer ....
-- 5 --- Manager ....

Answer 1

我不确定您在查询中使用GROUP BY的原因。当您在查询中使用聚合函数时，通常会使用GROUP BY。但我会说你可以使用以下内容：

select *
from users u
left join users_positions up
  on u.PRN = up.PRN
left join positions p
  on up.id = p.position
left join organizations o
  on p.org_title = o.id
left join divisions d
  on p.division = d.id

使用您提供的示例数据进行编辑，以下是您可以使用的查询示例：

select u.prn,
  u.firstname,
  u.age,
  o.name orgName,
  p.name positionName,
  d.name divisionName
from users u
left join users_positions up
  on u.PRN = up.PRN
left join organizations o
  on up.Organization_ID = o.id
left join positions p
  on up.Positionn_ID = p.id
left join divisions d
  on up.Division_ID = d.id

请参阅SQL Fiddle with Demo

此查询的结果是：

| PRN | FIRSTNAME | AGE | ORGNAME | POSITIONNAME |              DIVISIONNAME |
------------------------------------------------------------------------------
| 001 |     david |  24 |  test 1 |   Programmer |           Sale Department |
| 001 |     david |  24 |  test 2 |      Manager | Administration Department |
| 052 |    george |  27 |  test 1 |   Programmer |             IT Department |

如果您只想为每个用户返回一个组织等，那么您需要决定如何确定要返回的正确结果。例如，如果您只想返回每个用户max(Organization_ID)，那么您可以使用与此类似的内容：

select u.prn,
  u.firstname,
  u.age,
  o.name orgName,
  p.name positionName,
  d.name divisionName
from users u
left join
(
  select max(Organization_ID) Organization_ID, PRN
  from users_positions
  group by PRN
) up1
  on u.PRN = up1.PRN
left join users_positions up2
  on up1.PRN = up2.prn
  and up1.Organization_ID = up2.Organization_ID
left join organizations o
  on up2.Organization_ID = o.id
left join positions p
  on up2.Positionn_ID = p.id
left join divisions d
  on up2.Division_ID = d.id

请参阅SQL Fiddle with Demo

结果：

| PRN | FIRSTNAME | AGE | ORGNAME | POSITIONNAME |              DIVISIONNAME |
------------------------------------------------------------------------------
| 001 |     david |  24 |  test 2 |      Manager | Administration Department |
| 052 |    george |  27 |  test 1 |   Programmer |             IT Department |

Mysql加入表 - 最佳实践？

1 个答案: