引用相同的 SQL 列以在 CASE 子句中使用

Question

我需要一些关于使用 QMF for Qrkstation 在 DB2 中进行 SQL 查询的帮助 z/OS V11.1 修复包 3。

我正在查询来自不同 table 的几个字段。我也显示了一些时间戳。我需要一个使用这些时间戳之一的 CASE 子句。

想法是其中一个时间戳可以有数据或没有数据（空），因为我正在使用 LEFT JOIN 和那个 table（如果值则带上它，如果没有则为空）所以 CASE 子句会需要检查 column/field 并且如果它有数据（IS NOT NULL）然后输入值 'Delivered'。如下图所示，此逻辑对我不起作用。

select distinct
    z.po_id, 
    max(CASE 
           WHEN dcus.ship_evnt_tms IS NOT NULL THEN 'Delivered'
           WHEN days(CURRENT_timestamp) - days(Z.CAD_TMS) <= 0 THEN 'On time CAD'
           WHEN days(CURRENT_timestamp) - days(Z.CAD_TMS) between 1 and 2 THEN 'Less than 3 days after CAD'
           WHEN days(CURRENT_timestamp) - days(Z.CAD_TMS) between 3 and 5 THEN 'Between 3 & 5 days after CAD'
           WHEN days(CURRENT_timestamp) - days(Z.CAD_TMS) between 6 and 10 THEN 'Between 6 & 10 days after CAD'
           WHEN days(CURRENT_timestamp) - days(Z.CAD_TMS) >= 10 THEN 'More than 10 days after CAD'
           ELSE 'Non physical' 
       END) AS Aging,
    max(z.cad_tms) as cad_tms, 
    max(dcus.ship_evnt_tms) as DELV_DATE_CUS
from
    SCHEMA.TABLE z
left join 
    SCHEMA.TABLE1 A ON z.po_id = a.po_id
left join
    SCHEMA.TABLE2 scus ON (A.SHIP_ID = scus.SHIP_ID AND A.SHIP_TO_LOC_CODE = scus.SHIP_TO_LOC_CODE and scus.loc_type = 'CUS')
left join 
    SCHEMA.TABLE3 dcus ON (scus.SHIP_ID = dcus.SHIP_ID_856 AND scus.SHIP_TO_LOC_CODE = dcus.SHIP_TO_LOC_CD856 
AND dcus.SHIP_EVNT_CD = 'D')
left join
    SCHEMA.TABLE2 scdc ON (A.SHIP_ID = scdc.SHIP_ID AND A.SHIP_TO_LOC_CODE = scdc.SHIP_TO_LOC_CODE and scdc.loc_type = 'CDC')
left join 
    SCHEMA.TABLE3 dcdc ON (scdc.SHIP_ID = dcdc.SHIP_ID_856 AND scdc.SHIP_TO_LOC_CODE = dcdc.SHIP_TO_LOC_CD856 
AND dcdc.SHIP_EVNT_CD = 'D')

如图所示，前几行在最后一列有数据 tms，根据 CASE 子句逻辑应该显示为已交付（至少据我所知）而且，根据其他 WHEN 语句，最后的记录显示正确的计算和结果。

我尝试删除 table 别名 (dcus) 以便 "reference" 到列值本身，但它给了我一个 SQL 错误（模糊引用）

如果有人知道不同的方法或发现条款中的某种错误，请告诉我。

谢谢！！

Answer 1

我认为主要问题是您选择的是 case 语句为每个 id 返回的最大值，而不是 CAS 语句的最大日期值。所以这些值的顺序是（准时加元、非实物、超过 10 天...、少于 3 天...、已交付、6 日之间...、3 日之间）。因此，如果所有这些 id 都有一条记录，其中 ship_evnt_tms 为空，则该情况下的最大 vbalue 将为 More than 10 days...。您使用 distinct with aggregation 也很奇怪。相反，您可以在 po_id 上分组。也许试试：

select
z.po_id, 
CASE 
  WHEN max(dcus.ship_evnt_tms) is not null then 'Delivered'
  WHEN days(CURRENT_timestamp) - days(max(Z.CAD_TMS)) <= 0 then 'On time CAD'
  WHEN days(CURRENT_timestamp) - days(max(Z.CAD_TMS)) between 1 and 2 then 'Less than 3 days after CAD'
  WHEN days(CURRENT_timestamp) - days(max(Z.CAD_TMS)) between 3 and 5 then 'Between 3 & 5 days after CAD'
  WHEN days(CURRENT_timestamp) - days(max(Z.CAD_TMS)) between 6 and 10 then 'Between 6 & 10 days after CAD'
  WHEN days(CURRENT_timestamp) - days(max(Z.CAD_TMS)) >= 10 then 'More than 10 days after CAD'
  ELSE 'Non physical' 
END AS Aging,

max(z.cad_tms) as cad_tms, 
max(dcus.ship_evnt_tms) as DELV_DATE_CUS

from SCHEMA.TABLE z

left join SCHEMA.TABLE1 A ON z.po_id = a.po_id

left JOIN SCHEMA.TABLE2 scus ON (A.SHIP_ID = scus.SHIP_ID AND A.SHIP_TO_LOC_CODE = scus.SHIP_TO_LOC_CODE and scus.loc_type = 'CUS')

left join SCHEMA.TABLE3 dcus ON (scus.SHIP_ID = dcus.SHIP_ID_856 AND scus.SHIP_TO_LOC_CODE = dcus.SHIP_TO_LOC_CD856 
AND dcus.SHIP_EVNT_CD = 'D')

left JOIN SCHEMA.TABLE2 scdc ON (A.SHIP_ID = scdc.SHIP_ID AND A.SHIP_TO_LOC_CODE = scdc.SHIP_TO_LOC_CODE and scdc.loc_type = 'CDC')

left join SCHEMA.TABLE3 dcdc ON (scdc.SHIP_ID = dcdc.SHIP_ID_856 AND scdc.SHIP_TO_LOC_CODE = dcdc.SHIP_TO_LOC_CD856 
AND dcdc.SHIP_EVNT_CD = 'D')

group by z.po_id

引用相同的 SQL 列以在 CASE 子句中使用

Reference same SQL column to use in CASE clause

sql

db2

switch-statement

case-when