SQL - Where Select by Select Select语句？

Question

我有以下查询：

select r.people_code_id [People Code ID], r.resident_commuter [Campus6], 
c1.udormcom [AG], aR.RESIDENT_COMMUTER [AG Bridge], ar.ACADEMIC_SESSION, 
ar.ACADEMIC_TERM, ar.academic_year, ar.revision_date 
from RESIDENCY r 
left join AG_Common..CONTACT1 c1 on r.PEOPLE_CODE_ID=c1.key4 
left join AG_Common..CONTACT2 c2 on c1.ACCOUNTNO=c2.accountno 
left join AGPCBridge..ArchiveRESIDENCY aR on r.PEOPLE_CODE_ID=aR.PEOPLE_CODE_ID 
where r.ACADEMIC_YEAR='2010' 
and r.ACADEMIC_TERM='Fall' 
and SUBSTRING(c1.udormcom,1,1)<>r.resident_commuter 
and r.ACADEMIC_SESSION='Und 01' 
and aR.ACADEMIC_SESSION='Und 01' 
and aR.ACADEMIC_TERM='Fall' 
and aR.ACADEMIC_YEAR='2010' 
and SUBSTRING(c1.udormcom,1,1)=aR.RESIDENT_COMMUTER 

我需要一个条款添加到哪里段。我有这个查询：

select DISTINCT * from RESIDENCY where ACADEMIC_YEAR='2010' and 
ACADEMIC_TERM='Fall' and ACADEMIC_SESSION='Und 01' ORDER BY revision_date DESC 

这只获得每个人的最新行。我想做类似（伪代码）：

WHERE r.people_code_id and r.revision_date are in (select DISTINCT * from 
RESIDENCY where ACADEMIC_YEAR='2010' and ACADEMIC_TERM='Fall' and 
ACADEMIC_SESSION='Und 01' ORDER BY revision_date DESC) 

我在SQL 2000兼容模式下运行（尽管它实际上运行SQL 2008）。

Answer 1

我重新写你的查询的基础上，你想添加什么：

WITH residency_cte AS (
    SELECT TOP (1) 
      r.people_code_id, 
      r.resident_commuter, 
      r.academic_year, 
      r.academic_term, 
      r.academic_session 
     FROM RESIDENCY r 
     WHERE r.academic_year = '2010' 
     AND r.academic_term = 'Fall' 
     AND r.academic_session = 'Und 01' 
    ORDER BY revision_date DESC) 
    SELECT r.people_code_id, 
      r.resident_commuter [Campus6], 
      c1.udormcom [AG], 
      aR.RESIDENT_COMMUTER, 
      ar.ACADEMIC_SESSION, 
      ar.ACADEMIC_TERM, 
      ar.academic_year, 
      ar.revision_date 
    FROM residency_cte r 
LEFT JOIN AG_Common..CONTACT1 c1 ON c1.key4 = r.PEOPLE_CODE_ID 
           AND SUBSTRING(c1.udormcom, 1, 1) != r.resident_commuter 
LEFT JOIN AG_Common..CONTACT2 c2 ON c2.accountno = c1.ACCOUNTNO 
LEFT JOIN AGPCBridge..ArchiveRESIDENCY aR ON aR.PEOPLE_CODE_ID = r.PEOPLE_CODE_ID 
             AND aR.ACADEMIC_SESSION = r.academic_session 
             AND aR.ACADEMIC_TERM = r.academic_term 
             AND aR.ACADEMIC_YEAR = r.academic_year 
             AND SUBSTRING(c1.udormcom, 1, 1) = aR.RESIDENT_COMMUTER 

唯一的事情是udormcom列位置 - 一旦我知道它是什么表，我想移动条款成连接。我也更新了ArchiveRESIDENCY表的连接，所以你只需要在一个地方调整日期。

但请注意，使用子字符串在另一列上匹配将永远不会执行良好 - 直到数据模型更改为正确，永远不会真正优化。

Answer 2

你可以使用一个子查询的EXISTS

select 
    r.people_code_id [People Code ID], 
    r.resident_commuter [Campus6], 
    udormcom [AG], 
    aR.RESIDENT_COMMUTER [AG Bridge], 
    ar.ACADEMIC_SESSION, 
    ar.ACADEMIC_TERM, 
    ar.academic_year, 
    ar.revision_date 
from RESIDENCY r 
    left join AG_Common..CONTACT1 c1 
     on r.PEOPLE_CODE_ID=c1.key4 
    left join AG_Common..CONTACT2 c2 
     on c1.ACCOUNTNO=c2.accountno 
    left join AGPCBridge..ArchiveRESIDENCY aR 
     on r.PEOPLE_CODE_ID=aR.PEOPLE_CODE_ID 
where r.ACADEMIC_YEAR='2010' 
    and r.ACADEMIC_TERM='Fall' 
    and SUBSTRING(udormcom,1,1)<>r.resident_commuter 
    and r.ACADEMIC_SESSION='Und 01' 
    and aR.ACADEMIC_SESSION='Und 01' 
    and aR.ACADEMIC_TERM='Fall' 
    and aR.ACADEMIC_YEAR='2010' 
    and SUBSTRING(udormcom,1,1)=aR.RESIDENT_COMMUTER 
    and EXISTS 
    (
     select 1 
    FROM RESIDENCY r2 
    where 1=1 
     and r2.revision_date = ar.revision_date /* note the join here */ 
      and ACADEMIC_YEAR='2010' 
      and ACADEMIC_TERM='Fall' 
     and ACADEMIC_SESSION='Und 01' 

     /* the order by has been removed */ 
    ) 

Answer 3

WHERE r.people_code_id in (select DISTINCT people_code_id 
          from RESIDENCY 
          where ACADEMIC_YEAR='2010' 
          and ACADEMIC_TERM='Fall' 
          and ACADEMIC_SESSION='Und 01' 
          and revision_date = r.revision_date 
          ORDER BY revision_date DESC) 

我不认为你需要的“不同”或“按顺序”。去除这些应该会提高性能。