Question

我正在努力压扁我的部分SQL查询。我不确定是否需要PIVOT，或者应使用OVER对子查询进行此操作。

现在，我无法使用我的真实表名，因此我创建了一个代表我们真实系统的非常懊悔的例子。它看起来像是家庭作业，但事实并非如此。

这是复制品（*注意：我试图让它成为一个SqlFiddle，但它还没有工作）：

-- Some fake courses/subjects.
DECLARE @Courses TABLE (
    CourseId INTEGER IDENTITY NOT NULL,
    Name VARCHAR(100) NOT NULL)

INSERT INTO @Courses VALUES ('CS100')
INSERT INTO @Courses VALUES ('PY300')
INSERT INTO @Courses VALUES ('AG201')

-- Some teachers that teach the subject/course.
-- Notice there is no ORDER column, so the first teacher is considered
-- The main teacher. the 2nd result is the backup teacher.
-- First teacher = ORDER BY TeacherId ASC (per course)
DECLARE @Teachers TABLE (
    TeacherId INTEGER IDENTITY NOT NULL,
    CourseId INTEGER NOT NULL,
    Name VARCHAR(100)
)

INSERT INTO @Teachers VALUES (1, 'Jane')
INSERT INTO @Teachers VALUES (1, 'John')
INSERT INTO @Teachers VALUES (2, 'Bill')
INSERT INTO @Teachers VALUES (2, 'Anne')
INSERT INTO @Teachers VALUES (3, 'R2D2')

/*
-- Expected output:

CourseId | Cource Name | Primary Teacher | Backup Teacher
----------------------------------------------------------
1        | CS100       | Jane            | John
2        | PY300       | Bill            | Anne
3        | AG201       | R2D2            | <NULL>
*/

有人可以提供一些建议吗？

最后，请不要提出架构建议，比如老师可以在自己的桌子上，然后有Courses和Teachers的加入/链接表（因为同一位老师可以教一个以上的课，对吗？）。

Answer 1

您可以尝试此查询：

;with cte as(
select *, row_number() over(partition by CourseId order by TeacherId ASC) as rn
  from @Teachers)

select c.CourseId
     , c.Name as [Cource Name]
     , t1.Name as [Primary Teacher]
     , t2.Name as [Backup Teacher]
  from @Courses c 
      left outer join cte t1 on c.CourseId = t1.CourseId and t1.rn = 1
      left outer join cte t2 on c.CourseId = t2.CourseId and t2.rn = 2

如何压缩此Sql查询的一部分？这需要PIVOT吗？

1 个答案: