ABAP CDS 中的多个 SUM 聚合

Question

我有一张具有不同状态和多个工厂的工单表，如下所示：

植物	订单ID	状态
植物1	订单1	打开
植物1	订单2	已发布
植物1	订单3	已完成
植物1	订单4	打开
植物1	订单5	已发布
植物1	订单6	已完成
植物2	订单7	已发布
植物2	订单8	已完成

我想创建一个 CDS 来获取每个工厂的总计：“总订单”加上每个状态聚合的列。预期结果如下：

植物	总订单	打开	已发布	已完成
植物1	6	2	2	2
植物2	2	0	1	1

由于子查询是不可能的，我能想到的唯一解决方案是每个聚合的 CDS。 1 表示工厂总数，1 表示每个工厂状态。最终获得具有预期结果的顶级 CDS。

但这对于应该相对简单的事情来说似乎有很多 CDS，因为实际上我有 7 种不同的状态，而不仅仅是 3 种。我的解决方案有更好的替代方案吗？

Answer 1

在字段列表中使用 CASE...WHEN...THEN...ELSE 表达式 将

Status

转换为多个列，其中包含整数 0 或整数 1，具体取决于

Status

的值:

define view VIEWNAME as select from DATABASE_TABLE {
    key OrderId,
    Plant,
    CASE Status WHEN 'OPEN' THEN 1 ELSE 0 END AS Open,
    CASE Status WHEN 'RELEASED' THEN 1 ELSE 0 END AS Released,
    CASE Status WHEN 'COMPLETED' THEN 1 ELSE 0 END AS Completed
}

现在您可以创建一个 second 视图来查询此 first 视图。使视图

GROUP BY Plant

。然后使用

COUNT( * )

获取每个工厂的订单总数，并使用 Open、Released 和 Completed 字段上的 SUM 获取各自的计数：

define view VIEWNAME2 as select from VIEWNAME {
    key Plant,
    COUNT( * ) as TotalOrders,
    SUM( Open ) AS Open,
    SUM( Released ) AS Released,
    SUM( Completed ) AS Completed
}
GROUP BY Plant

Answer 2

实现旋转的另一种方法是使用 UNION，遵循我此处针对 HANA CDS 描述的相同方法。 这是菲利普使用方式的另一种风格。

联合状态总和的第一个视图：

define view ZORDERS (plant, TOTAL, OPENED, RELEASED, COMPLETED) as 
select from zplants {
    plant, count(distinct orderid ) as TOTAL, 0 as OPENED, 0 as RELEASED, 0 as COMPLETED
}  group by plant
union
select from zplants {
    plant, 0 as TOTAL, count(distinct orderid ) as OPENED, 0 as RELEASED, 0 as COMPLETED
} where status = 'OPENED' group by plant
union
select from zplants {
    plant, 0 as TOTAL, 0 as OPENED, count(distinct orderid ) as RELEASED, 0 as COMPLETED
} where status = 'RELEASED' group by plant
union
select from zplants {
    plant, 0 as TOTAL, 0 as OPENED, 0 as RELEASED, count(distinct orderid ) as COMPLETED
} where status = 'COMPLETED' group by plant

聚合它们的第二个视图：

define view ZPIVOT as select from ZORDERS {
    plant,
    sum ( TOTAL ) as TOTAL, 
    sum ( OPENED ) as OPENED, 
    sum ( RELEASED ) as RELEASED,
    sum ( COMPLETED ) as COMPLETED
} group by plant 

注意：这里我使用了

OPENED

，因为

OPEN

是 ABAP CDS 中的保留字。

Answer 3

您可以通过使用自连接来使用一个 CDS 来完成此操作，如示例所示：

define view VIEWNAME as select from DB_TAB
left outer join DB_TAB as DB_TAB1 on DB_TAB.Plant = DB_TAB1.Plant
                                 and DB_TAB.OrderID = DB_TAB1.OrderID
                                 and DB_TAB1.Status = 'OPEN'
left outer join DB_TAB as DB_TAB2 on DB_TAB.Plant = DB_TAB2.Plant
                                 and DB_TAB.OrderID = DB_TAB2.OrderID
                                 and DB_TAB2.Status = 'RELEASED'
left outer join DB_TAB as DB_TAB3 on DB_TAB.Plant = DB_TAB3.Plant
                                 and DB_TAB.OrderID = DB_TAB3.OrderID
                                 and DB_TAB3.Status = 'COMPLETED'
{
   key Plant,
   count( * ) as Total_Orders,
   count( distinct DB_TAB1.OrderID ) as Open,
   count( distinct DB_TAB2.OrderID ) as Released,
   count( distinct DB_TAB3.OrderID ) as Completed,
}
group by Plant