Updating multiple rows with different primary key in one query in PostgreSQL?

Question

I have to update many columns in many rows in PostgreSQL 9.1. I'm currently doing it with many different UPDATE queries, each one that works on a different row (based on the primary key):

UPDATE mytable SET column_a = 12, column_b = 6 WHERE id = 1;
UPDATE mytable SET column_a = 1, column_b = 45 WHERE id = 2;
UPDATE mytable SET column_a = 56, column_b = 3 WHERE id = 3;

我必须做几千个这样的查询.

无论如何,我可以在PostgreSQL的一个查询中"批量更新"大量行吗？如果你正在使用INSERT,你可以一次插入多行:( INSERT INTO mytable (column_a, column_b) VALUES ( (12, 6), (1, 45) );),有类似的东西UPDATE吗？

就像是:

UPDATE mytable SET (id, column_a, column_b) FROM VALUES ( (1, 12, 6), (2, 1, 45), (3, 56, 3), … )

??

重要的一点是每个'VALUE'只会更新一行(基于WHERE id =).每行都有相同的,固定数量的列需要更新,但每行对每列都有不同的值,因此UPDATE mytable SET column_a = 12, column_b = 6 WHERE id IN (1, 2, 3);不起作用.

Answer 1

是的,您可以(通常在SQL中首选)一次更新多个行.有几种方法可以做到这一点,但我认为最可读和优雅的是使用带有id和值的派生表:

update mytable as m set
    column_a = c.column_a,
    column_b = c.column_b
from (values
    (1, 12, 6),
    (2, 1, 45),
    (3, 56, 3)
) as c(id, column_a, column_b)
where c.id = m.id

不太可读,但更明显的解决方案是使用case:

update mytable set
    column_a = case id when 1 then 12 when 2 then 1 when 3 then 56 end,
    column_b = case id when 1 then 6 when 2 then 45 when 3 then 3 end
where id in (1, 2, 3)

Answer 2

如果这适用于您的情况,您可以使用它.

create table test(id int, a int, b int);

insert into test(id, a, b)
values
(1, 1, 1),
(2, 1, 1),
(3, 1, 1),
(4, 1, 1),
(5, 1, 1),
(6, 1, 1),
(7, 1, 1);


update test as d
set a = s.a, b = s.b
from 
(
  values
  (1, 2, 2),
  (2, 2, 2)
) as s(id, a, b)
where d.id = s.id

SQL FIDDLE DEMO