Oracle SQL - 自定义排序

Oracle SQL - Custom Sort

我有一个场景,我有以下数据:

Table:地点

ID      TYPE
------------------
1000    STORE
11001   STORE
20000   STORE
1181    WAREHOUSE
12002   STORE

我想以一种方式排序,所有以“0000”结尾的 ID 应该首先排序,然后是类型 'Warehouse',然后是商店的其余部分。

期望的输出应该像

ID      TYPE
------------------
10000   STORE
20000   STORE
1181    WAREHOUSE
11001   STORE
12002   STORE

如何进行自定义排序?

您可以在排序中使用 case 表达式:

order by (case when id like '%0000' then 1
               when type = 'WAREHOUSE' then 2
               else 3
          end), id

这也使用id在三组内排序。

注意:如果id是数字而不是字符串,我建议:

order by (case when mod(id, 10000) = 0 then 1
               when type = 'WAREHOUSE' then 2
               else 3
          end), id

[LF 编辑]

这是您的 ORDER BY returns,这不是 OP 想要的:

SQL> with locations (id, type) as
  2    (select 1000 , 'STORE'     from dual union all
  3     select 11001, 'STORE'     from dual union all
  4     select 20000, 'STORE'     from dual union all
  5     select 1181 , 'WAREHOUSE' from dual union all
  6     select 12002, 'STORE'     from dual
  7    )
  8  select id, type
  9  from locations
 10  order by (case when id like '%0000' then 1
 11                 when type = 'WAREHOUSE' then 2
 12                 else 3
 13            end), id;

        ID TYPE
---------- ---------
     20000 STORE
      1181 WAREHOUSE
      1000 STORE
     11001 STORE
     12002 STORE

SQL>

Gordon 的评论:如果第 1000 行是 10000,以上应该有效。

我是这样理解问题的;样本数据直到第 7 行;查询从第 8 行开始。

SQL> with locations (id, type) as
  2    (select 1000 , 'STORE'     from dual union all
  3     select 11001, 'STORE'     from dual union all
  4     select 20000, 'STORE'     from dual union all
  5     select 1181 , 'WAREHOUSE' from dual union all
  6     select 12002, 'STORE'     from dual
  7    )
  8  select id, type
  9  from locations
 10  order by case when substr(to_char(id), -3) = '000' then 1 end,
 11           case when type = 'WAREHOUSE' then 2 end,
 12           type;

        ID TYPE
---------- ---------
      1000 STORE
     20000 STORE
      1181 WAREHOUSE
     12002 STORE
     11001 STORE

SQL>