如何根据某些列值从 SQL 服务器 table 中查找重复条目？

Question

这是我的 SQL 服务器 table - 我需要根据客户姓名、出生日期、父亲姓名列查找重复的客户代码：

Customer code   Customer Name   Date of birth   Father Name
-------------------------------------------------------------
0001            Md. Alam        1991-10-20      Sr. Alam 
0002            Alam            1991-10-20      Sr. alam
0004            Hasan           1990-01-01      Sr. Hasan
0005            Karim           1988-01-01      Sr. Karim
0006            Karim           1988-01-01      S Karim
0007            Kalam           1985-01-01      Sr. Kalam

输出看起来像：

0001,0002,0005,0006 客户重复，因为客户姓名、出生日期和父亲姓名列相似。

Customer code   Customer Name   Date of birth   Father Name
------------------------------------------------------------
0001            Md. Alam        1991-10-20      Sr. Alam 
0002            Alam            1991-10-20      Sr. alam
0005            Karim           1988-01-01      Sr. Karim
0006            Karim           1988-01-01      S Karim

请帮我找出一个有效的方法

Answer 1

select * from customer where code in(
select c1.code from customer c1 inner join customer c2 on c1.dob = c2.dob and 
(c1.father = c2.father or c1.name = c2.name)
group by c1.code having count(c1.code) > 1

Answer 2

SELECT t.Customer_Code
    ,t.Customer_Name
FROM tb2 AS t
WHERE t.Customer_Name IN (
        SELECT t.Customer_Name
        FROM tb2 AS t
        GROUP BY t.Customer_Name
        HAVING COUNT(t.Customer_Name) > 1
        )
ORDER BY t.Customer_Code ASC

Answer 3

请尝试以下解决方案。

1st CTE 正在清理两列 CustomerName 和 FatherName.
2nd CTE 正在根据 3 列的组合创建存储桶： CustomerName、DOB 和 FatherName.
3rd CTE 正在创建存储桶中行的计数器，即查找重复项。
最终 SELECT 正在加入原始 table 并过滤掉不重复行。

SQL

-- DDL and sample data population, start
DECLARE @tbl TABLE (CustomerCode CHAR(4), CustomerName VARCHAR(20), DOB DATE, FatherName VARCHAR(20));
INSERT INTO @tbl (CustomerCode, CustomerName, DOB, FatherName) VALUES
('0001', 'Md. Alam', '1991-10-20', 'Sr. Alam'),
('0002', 'Alam', '1991-10-20', 'Sr. alam'),
('0004', 'Hasan', '1990-01-01', 'Sr. Hasan'),
('0005', 'Karim', '1988-01-01', 'Sr. Karim'),
('0006', 'Karim', '1988-01-01', 'S Karim'),
('0007', 'Kalam', '1985-01-01', 'Sr. Kalam');
-- DDL and sample data population, end

WITH rs AS
(
    SELECT CustomerCode, DOB
        , RIGHT(CustomerName, LEN(CustomerName) - c.pos) AS CustomerName
        , RIGHT(FatherName, LEN(FatherName) - f.pos) AS FatherName
    FROM @tbl
        CROSS APPLY (SELECT CHARINDEX(SPACE(1), CustomerName)) AS c(pos)
        CROSS APPLY (SELECT CHARINDEX(SPACE(1), FatherName)) AS f(pos)
), cte AS
(
    SELECT * 
    , ROW_NUMBER() OVER (ORDER BY CustomerCode) -
     ROW_NUMBER() OVER (PARTITION BY rs.CustomerName, rs.DOB, rs.FatherName ORDER BY rs.CustomerCode) AS bucket
    FROM rs
), cte2 AS
(
    SELECT CustomerCode, bucket, COUNT(bucket) OVER (PARTITION BY cte.bucket) AS [counter]
    FROM cte
    GROUP BY  CustomerCode, bucket
)
SELECT t.* FROM @tbl AS t
INNER JOIN cte2 ON cte2.CustomerCode = t.CustomerCode
WHERE cte2.counter > 1;

输出

+--------------+--------------+------------+------------+
| CustomerCode | CustomerName |    DOB     | FatherName |
+--------------+--------------+------------+------------+
|         0001 | Md. Alam     | 1991-10-20 | Sr. Alam   |
|         0002 | Alam         | 1991-10-20 | Sr. alam   |
|         0005 | Karim        | 1988-01-01 | Sr. Karim  |
|         0006 | Karim        | 1988-01-01 | S Karim    |
+--------------+--------------+------------+------------+

Answer 4

如果我没理解错的话，当客户名称在整体上明显不匹配时，您认为客户 0001 和 0002 是重复的。

我相信您在这里需要某种模糊搜索，它可以匹配 2 个字符串并根据您可以确定的 2 个条目是否重复给出匹配百分比。

此处回答了类似的问题。 SQL Server Fuzzy Search with Percentage of match

如何根据某些列值从 SQL 服务器 table 中查找重复条目？

How to find duplicate entries from a SQL Server table based on some column value?

sql-server

sql-server-2014