Question

- 问题是：我有更多的公司，我想做id_shutter＆＃34; autoincrement＆＃34;每个公司从1开始（id_company和id_shutter是复合PK）

 CREATE FUNCTION insert_shutter() RETURNS TRIGGER AS $insert_shutter$
BEGIN
    IF exists(select 1 from shutter where id_company=new.id_company) then
    SELECT MAX(id_shutter) INTO new.id_shutter FROM shutter where id_company=new.id_company;
    new.id_shutter:=id_shutter+1;
    ELSE
    new.id_shutter=1;
    end if;
        RETURN NEW;

END;
$insert_shutter$ LANGUAGE plpgsql;

CREATE TRIGGER insert_shutter 
    BEFORE INSERT ON shutter
    FOR EACH ROW
    EXECUTE PROCEDURE insert_shutter();

Answer 1

偶尔这是必要的。但实际上，在表格中添加serial列（让我们称之为ShutterId）然后在检索数据时进行计算会更容易：

select s.*, row_number() over (partition by company order by id) as CompanyShutterSequence
from shutter s;

此外，调用带有重复值的id内容会产生误导。

Answer 2

抱歉我发现了，我无法使用：new.id_shutter：= id_shutter + 1; 我做了SELECT MAX（id_shutter）+1 INTO new.id_shutter FROM shutter其中id_company = new.id_company;

Postgresql在触发函数中使用max函数

2 个答案: