我觉得楼上的办法是指整体，没有分段排序。
=======================================
说说咱的不成熟的看法(别笑咱啊，呵~~~)；看题目是：CW进行分组,按照Depth1进行排序，所以只用这两个字段
select cw, depth1,null ids into #b from table1
order by cw,depth1 asc
update #b set ids=1 from (select cw,min(depth1) depth1 from #b group by cw) m
where #b.cw=m.cw and #b.depth1=m.depth1
SELECT cw,count(*) cs into #cs from gd group by cw-- 循环次数
declare @a int ,@b int
set @a=2
select @b=max(cs) from #cs
while @a<=@b
begin
update #b set ids=@a from (select cw,min(depth1) depth1 from #b where ids is null group by cw) m
where #b.cw=m.cw and #b.depth1=m.depth1
set @a=@a+1
end
结果：select * from #b
呵呵，不好意思啊，下面的gd表就是table1表，忘记改了
SELECT cw,count(*) cs into #cs from gd group by cw-- 循环次数
declare @temptable table (
CW varchar(2) ,
Depth1 int,
Depth2 int
)
insert @temptable select 'A',2500,2512
union select 'B',2511,2513
union select 'A',2500,2512
union select 'A',2711,2080
union select 'A',2900,2901
union select 'B',1153,1787
select * from @temptable
select cw,Depth1,Depth2,ID = (select count(*) from @temptable b where a.depth1 >= b.depth1 and a.cw = b.cw group by cw) from @temptable a order by cw,Depth1
