select [字段A],[字段B],MAX([字段C]) as [字段C],max([字段D]) as [字段D]
from test
where [字段A] not in(
select [字段A] from test where [字段B]='结束'
)
group by [字段A],[字段B]

------解决方案--------------------
SQL code
select
  a.字段A ,a.字段B,a.字段C,b.字段D
from tab a,tab b
where a.字段A = b.字段A
and not exists (
  select 1
  from tab
  where 字段A = a.字段A
  and 字段B = '结束'
  )
and not exists (
  select 1
  from tab
  where 字段A = a.字段A
  and 字段c > a.字段c
  )
and not exists (
  select 1
  from tab
  where 字段A = b.字段A
  and 字段c < b.字段c
  )

------解决方案--------------------

SQL code

IF EXISTS (SELECT 1 FROM SYSOBJECTS WHERE name = 'tba')
BEGIN
    DROP TABLE tba
END
GO
CREATE TABLE tba
(
    a VARCHAR(10),
    b VARCHAR(10),
    c INT,
    d VARCHAR(10)
)
GO
INSERT INTO tba
SELECT 'A','结束',3,'a' UNION
SELECT 'A','开始',2,'b' UNION
SELECT 'A','开始',1,'c' UNION
SELECT 'B','结束',7,'a' UNION
SELECT 'B','开始',6,'b' UNION
SELECT 'B','开始',5,'c' UNION
SELECT 'B','开始',4,'d' UNION
SELECT 'B','开始',3,'e' UNION
SELECT 'B','开始',2,'f' UNION
SELECT 'B','开始',1,'g' UNION
SELECT 'C','开始',2,'a' UNION
SELECT 'C','开始',1,'b' UNION
SELECT 'D','开始',3,'a' UNION
SELECT 'D','开始',2,'b' UNION
SELECT 'D','开始',1,'c'

SELECT A,b,MAX(c) AS C,MAX(d) AS D
FROM tba
WHERE a NOT IN (SELECT a from tba WHERE b = '结束')
GROUP BY a,b

A    b    C    D
C    开始    2    b
D    开始    3    c