1

我将 sql server 脚本重写为 redshift 数据库查询

我有外部应用结构

OUTER APPLY
   (
   SELECT  q.*
   FROM    (
           SELECT  ROW_NUMBER() OVER(ORDER BY ca.Id DESC) AS rn,
                   ca.StateProvince,
                   ca.ZipPostalCode,
                   ca.ContactId
           FROM    public.contact_addresses ca 
           WHERE   ca.OrganizationId = <Parameters.DemographicsOrgId>
                   AND ca.DeletedDate IS NULL
                   AND ca.TypeId = 7
                   AND ca.ContactId = cc.Id
           ) q
   WHERE   q.rn = 1
   ) ca

但是 Redshift 没有外部应用。我如何正确地用 LEFT JOIN 重写它？

更新

我想像这样重写它

LEFT JOIN
   (
   SELECT  q.*,
           q.rn = 1
   FROM    (
           SELECT  ROW_NUMBER() OVER(ORDER BY ca.Id DESC) AS rn,
                   ca.StateProvince,
                   ca.ZipPostalCode,
                   ca.ContactId
           FROM    public.contact_addresses ca
           WHERE   ca.OrganizationId = <Parameters.DemographicsOrgId>
                   AND ca.DeletedDate IS NULL
                   AND ca.TypeId = 7
                   AND ca.ContactId = cc.Id
           ) q
   GROUP BY q.rn
   ) ca
ON ca.rn = 1

但这是正确的吗？

4

2 回答 2

3

不，它看起来不对。我猜：

LEFT JOIN
(SELECT ca.OrganizationId,
        ROW_NUMBER() OVER (ORDER BY ca.Id DESC) AS rn,
        ca.StateProvince,
        ca.ZipPostalCode,
        ca.ContactId
 FROM  public.contact_addresses ca 
 WHERE ca.DeletedDate IS NULL AND
       ca.TypeId = 7
 GROUP BY ca.OrganizationId, ca.ContactId
) ca
ON ca.ContactId = cc.ID AND
   ca.OrganizationId = <Parameters.DemographicsOrgId> AND
   ca.rn = 1

基本上，您需要按相关条件（如果它们相等）进行聚合，然后将它们用于外部ON条件。

于 2018-10-29T10:58:20.710 回答
2

OUTER APPLY 运算符返回左表表达式中的所有行，而不管它是否与右表表达式匹配。对于那些在右表表达式中没有对应匹配的行，它在右表表达式的列中包含 NULL 值

所以你的方法是正确的

于 2018-10-29T07:59:37.737 回答