我有可以有多个地址与它们相关联的记录,我将所有这些归一化到一个表中,以便我可以看到属于记录的每个地址并识别重复项(基于地址栏1)。在SQL Server 2008表上识别重复项的行数
我想要识别的是具有相同地址两次的记录。我使用CTE完成了这项工作,并为每次发生的地址添加了一个行号,但我忘记了一个重要部分。对于每个记录,都有一个主要地址,并且该地址(如果它是重复的)应始终具有最低的行号(RN)。
请告诉我偶尔会发生的事情是我的主地址相匹配的非主地址和非主地址获取RN = 1,我的主要获取RN> 1
因为我要停用这些我期待对于所有不是主要地址的RN>'1'的地址,我最终会丢失一堆结果。我添加了一个名为'PRIMARY_ADDR'的列,但我不确定如何将此字段并入CTE中。
表
CREATE TABLE [dbo].[ALL_Address_Table]
(
[RECORD_ID] [varchar](50) NULL,
[EXT_ID] [varchar](50) NULL,
[ADDR_ID] [varchar](50) NULL,
[ADDR_LINE_1] [varchar](80) NULL,
[ADDR_LINE_2] [varchar](80) NULL,
[ADDR_ZIP] [varchar](50) NULL,
[ADDR_STATUS] [varchar](50) NULL,
[RN] [varchar](50) NULL,
[PRIMARY_ADDR] [varchar](50) NULL
) ON [PRIMARY]
INSERT INTO ALL_Address_Table
([RECORD_ID], [EXT_ID], [ADDR_ID], [ADDR_LINE_1], [ADDR_LINE_2], [ADDR_ZIP], [ADDR_STATUS], [RN], [PRIMARY_ADDR])
VALUES
(555678, '900120', '555678-100001', '123 MAIN ST','','58865','ACTIVE','','No'),
(555678, '900125', '555678-100002', '123 MAIN ST','','58865','ACTIVE','','Yes'),
(555678, '900145', '555678-100003', '123 MAIN ST','','58865','ACTIVE','','No'),
(555678, '900115', '555678-100004', '500 W TRAIN ST','','58873','ACTIVE','','No')
;
CTE
WITH cte AS
(
SELECT
[RECORD_ID], [EXT_ID], [ADDR_ID], [ADDR_LINE_1],
row_number() OVER(PARTITION BY RECORD_ID, ADDR_Line_1 ORDER BY RECORD_ID) AS [rn]
FROM ALL_Address_Table
)
update ALL_Address_Table
set RN = CTE.rn
from CTE
join ALL_Address_Table on CTE.RECORD_ID = ALL_Address_Table.RECORD_ID
where CTE.ADDR_ID = All_Address_Table.ADDR_ID
--This should not return a row with Primary_ADDR = 'Yes' but must be included in the count.
--I cannot just remove all the primary addresses or else I will miss the ones that have
--duplicates
select * from ALL_Address_Table where RN > '1'
SQL FIDDLE http://sqlfiddle.com/#!3/52a0f/3
感谢您的任何建议。
谢谢,我不知道我为什么试图像改变顺序那样简单。 – Jeremy 2015-02-09 18:47:15