如何獲取像這樣的重複數據？

Question

我有這樣的數據

和類別重複的是
如果存在重複1.Duplicate基礎代碼，結果是最後一步

2.Duplicate基地NUMBER1或者如果存在重複的基礎數字2，結果是最後一步

我需要像導致

個

感謝，

Answer 1

如果我明白你的要求，我相信這會達到你想要的。

它至少會產生指定的結果。

declare @Data table 
(
    ID int not null, 
    CODE varchar(3) not null, 
    NUMBER1 varchar(3) not null, 
    NUMBER2 varchar(3) not null, 
    STEP int not null 
) 

insert into @Data values 
    (1, 'XXA', '001', '009', 1), 
    (2, 'XXB', '001', '008', 2), 
    (3, 'XXC', '002', '009', 3), 
    (4, 'XXA', '002', '008', 4) 

-- this query returns the record with the highest STEP 
-- excluding any other records that have the same CODE, 
-- NUMBER1, or NUMBER2 and a lower STEP 
select 
    d.ID, 
    d.CODE, 
    d.NUMBER1, 
    d.NUMBER2, 
    d.STEP 
from @Data d 
where 
    -- there does not exist any other record that has 
    -- the same CODE, NUMBER1, or NUMBER2 and a higher STEP 
    not exists (
     select 1 
     from @Data duplicate 
     where 
      (
       duplicate.CODE = d.CODE 
       or duplicate.NUMBER1 = d.NUMBER1 
       or duplicate.NUMBER2 = d.NUMBER2 
      ) 
      and (
       duplicate.STEP > d.STEP 
       or (
        duplicate.STEP = d.STEP 
        and duplicate.ID > d.ID 
       ) 
      ) 
    ) 


-- this query returns the duplicate records with a lower STEP 
select 
    d.ID, 
    d.CODE, 
    d.NUMBER1, 
    d.NUMBER2, 
    d.STEP, 
    HigherStep.ID as HigherStepID 
from @Data d 
inner join (
    -- join to the duplicate record that has a higher STEP 
    select 
     ID, 
     CODE, 
     NUMBER1, 
     NUMBER2 
    from @Data HighestStep 
    where 
     -- there does not exist any other record that has 
     -- the same CODE, NUMBER1, or NUMBER2 and a higher STEP 
     not exists (
      select 1 
      from @Data duplicate 
      where 
       (
        duplicate.CODE = HighestStep.CODE 
        or duplicate.NUMBER1 = HighestStep.NUMBER1 
        or duplicate.NUMBER2 = HighestStep.NUMBER2 
       ) 
       and (
        duplicate.STEP > HighestStep.STEP 
        or (
         duplicate.STEP = HighestStep.STEP 
         and duplicate.ID > HighestStep.ID 
        ) 
       ) 
     ) 
) HigherStep on 
    HigherStep.ID <> d.ID -- don't match this record to itself 
    and (
     HigherStep.CODE = d.CODE 
     or HigherStep.NUMBER1 = d.NUMBER1 
     or HigherStep.NUMBER2 = d.NUMBER2 
    ) 

---

下面是一個下臺階的基礎上，玉山的評論檢索記錄修訂查詢。該查詢的一個問題是，它將無法捕獲那些記錄不是以最高的步驟立即重複記錄。

-- this query returns the duplicate records with a lower STEP 
select 
    d.ID, 
    d.CODE, 
    d.NUMBER1, 
    d.NUMBER2, 
    d.STEP, 
    HigherStep.ID as HigherStepID 
from @Data d 
inner join (
    -- join to the duplicate record that has a higher STEP 
    select 
     ID, 
     CODE, 
     NUMBER1, 
     NUMBER2, 
     STEP 
    from @Data 
) HigherStep on 
    (
     HigherStep.CODE = d.CODE 
     or HigherStep.NUMBER1 = d.NUMBER1 
     or HigherStep.NUMBER2 = d.NUMBER2 
    ) 
    and (
     HigherStep.STEP > d.STEP 
     or (
      HigherStep.STEP = d.STEP 
      and HigherStep.ID > d.ID 
     ) 
    ) 

Answer 2

我希望我的理解正確：

DECLARE @T1 TABLE 
(
    ID INT, 
    CODE VARCHAR(50), 
    NUMBER1 VARCHAR(50), 
    NUMBER2 VARCHAR(50), 
    STEP INT 
) 

INSERT INTO @T1(ID, CODE, NUMBER1, NUMBER2, STEP) 
VALUES(1, 'XXA', '001', '009', 1) 
INSERT INTO @T1(ID, CODE, NUMBER1, NUMBER2, STEP) 
VALUES(2, 'XXB', '001', '008', 2) 
INSERT INTO @T1(ID, CODE, NUMBER1, NUMBER2, STEP) 
VALUES(3, 'XXC', '002', '009', 3) 
INSERT INTO @T1(ID, CODE, NUMBER1, NUMBER2, STEP) 
VALUES(4, 'XXA', '002', '008', 4) 

SELECT  TAB.ID, 
      TAB.CODE, 
      TAB.NUMBER1, 
      TAB.NUMBER2, 
      TAB.STEP, 
      COALESCE(CAST(DUP_CODE_T.ID AS VARCHAR(50)) + ',', '') + 
      COALESCE(CAST(DUP_NUM1_T.ID AS VARCHAR(50)) + ',', '') + 
      COALESCE(CAST(DUP_NUM2_T.ID AS VARCHAR(50)), '') AS DUPLICATE_WITH 
FROM  @T1 AS TAB OUTER APPLY 
      (
       SELECT  TOP 1 DUP_CODE.ID 
       FROM  @T1 AS DUP_CODE 
       WHERE  DUP_CODE.CODE = TAB.CODE AND 
          DUP_CODE.ID <> TAB.ID 
      ) AS DUP_CODE_T OUTER APPLY 
      (
       SELECT  TOP 1 DUP_NUM1.ID 
       FROM  @T1 AS DUP_NUM1 
       WHERE  DUP_NUM1.NUMBER1 = TAB.NUMBER1 AND 
          DUP_NUM1.ID <> TAB.ID 
      ) AS DUP_NUM1_T OUTER APPLY 
      (
       SELECT  TOP 1 DUP_NUM2.ID 
       FROM  @T1 AS DUP_NUM2 
       WHERE  DUP_NUM2.NUMBER2 = TAB.NUMBER2 AND 
          DUP_NUM2.ID <> TAB.ID 
      ) AS DUP_NUM2_T 

您可以添加WHERE子句或TOP 1篩選結果。