高分求教一sql语的优化

[silentcross]

共3张表

表A｛a1,a2,a3,a4,a5｝
表B {b1,b2,b3,b4,b5}
表C {c1,c2,c3,c4 ,c5}

连接条件
1： A.a1 = B.b1(+) and A.a2 = B.b2(+) and A.a3 = B.b3(+)
2: A.a4 = C.c1(+) and A.a5 = C.c2(+)
3: 麻烦的是这个条件，C中的数据是按c1,c2分组的,条件要求c中的数据所在组中的c3字段值最大的那条数据里对应的c4大于指定的值(假设100)

效果如下：
SELECT
B.b4, C.c5
FROM
A,
B,
C,
(SELECT C.c1,C.c2,C.c4
FROM C,
(SELECT c1, c2, max(c3) c3
FROM C
GROUP BY c1, c2) T2
WHERE
C.c1 = T2.c1 AND C.c2 = T2.c2 AND C.c3 = T2.c3) T1
WHERE
A.a1 = B.b1(+) AND
A.a2 = B.b2(+) AND
A.a3 = B.b3(+) AND
A.a4 = C.c1(+) AND
A.a5 = C.c2(+) AND
C.c1 = T1.c1 AND
C.c2 = T1.c2 AND
T1.c4 > 100

问题是这么写效率太低了，苦思一下午没想出什么办法，求高手指点

[silentcross]

补充说明一个：需要使用标准sql

[fangcheng007]

SELECT
B.b4, C.c5
FROM
A,
B,
C,
(SELECT C.c1,C.c2,C.c4
FROM C,
(SELECT c1, c2, max(c3) c3
FROM C

GROUP BY c1, c2) T2
WHERE
C.c1 = T2.c1 AND C.c2 = T2.c2 AND C.c3 = T2.c3) T1
WHERE
A.a1 = B.b1(+) AND
A.a2 = B.b2(+) AND
A.a3 = B.b3(+) AND
A.a4 = C.c1(+) AND
A.a5 = C.c2(+) AND
C.c1 = T1.c1 AND
C.c2 = T1.c2 AND
T1.c4 > 100

[fangcheng007]

SELECT
B.b4, C.c5
FROM
A,
B,
C,
(SELECT c1, c2, max(c3) c3
FROM C
where c4>100
GROUP BY c1, c2) T1
WHERE
A.a1 = B.b1(+) AND
A.a2 = B.b2(+) AND
A.a3 = B.b3(+) AND
A.a4 = C.c1(+) AND
A.a5 = C.c2(+) AND
C.c1 = T1.c1 AND
C.c2 = T1.c2 AND

[fangcheng007]

SELECT
B.b4, C.c5
FROM
A,
B,
C,
(SELECT c1, c2, max(c3) c3
FROM C
where c4>100
GROUP BY c1, c2) T1
WHERE
A.a1 = B.b1(+) AND
A.a2 = B.b2(+) AND
A.a3 = B.b3(+) AND
A.a4 = C.c1(+) AND
A.a5 = C.c2(+) AND
C.c1 = T1.c1 AND
C.c2 = T1.c2

注意建索引

[silentcross]

感谢城城相助，将c4条件提前的确极大缩短了执行时间，从80秒降到3秒了
你给我的代码结果上有点不一致，3层select还是不能避免
虽然实际上c4的条件没有缩小数据规模，但是效果却如此惊人，真实很奇怪
不管怎么说现在已经完全可以接受了，非常感谢