主页 / 问题 / 601797
Accepted
Kamo Petrosyan
Asked:2020-12-10 13:31:44 +0000 UTC2020-12-10 13:31:44 +0000 UTC

摆脱重复的 INNER JOIN

  • 772

PostgreSQL 数据库

CREATE TABLE items (
    id SERIAL,
    user_id INTEGER,
    status SMALLINT
);

CREATE TABLE phones (
  phone VARCHAR(12),
  users INTEGER[]
);

其中 items.status 为 3、5 或 7

phone1, phone2... 列表需要选择下面的表格

phone1;COUNT(items.id) where status=3 ;COUNT(items.id) where status=7 phone2;COUNT(items.id) where status=3 ;COUNT(items.id) where status=7 ...

我的版本乘以重复然后总结

SELECT phones.phone as phone, COUNT(it1.id) as saled_count, COUNT(it2.id) as not_saled_count
FROM phones
INNER JOIN items it1 ON phones.users @> ARRAY[it1.user_id]::int[]
INNER JOIN items it2 ON phones.users @> ARRAY[it2.user_id]::int[]
WHERE phones.phone IN ('phone1', 'phone2')
    AND it1.status = 7 AND it2.status = 3
GROUP BY phones.phone;

结果分别

phone1    5    5
phone2    6    6
...

代替

phone1    3    2
phone2    5    1
...

那么，问题本身：如何提出请求？

sql

1 个回答

  1. Best Answer

    如果我正确理解问题

    select
    p.phone,
    sum(case i.status when 3 then 1 else 0 end) as status3,
    sum(case i.status when 5 then 1 else 0 end) as status5,
    sum(case i.status when 7 then 1 else 0 end) as status7
from phones as p
inner join items as i on i.user_id = any(p.users)
where p.phone in ('phone1', 'phone2')
group by p.phone;
    • 3

相关问题