高分求一个多表统计的SQL，来者有分，有答案马上结贴

[ABCatai]

现有３张表table1, table2，table3，每个表中分别有保存user_id的字段fld01, fld01, fld03, 并且

每张表中都有created字段，现在要统计在某一时间段内这三张表中保存的user的个数（其中同一个用

户只能统计一次，不能累加），如：
在2008-12-24到2008-12-25这两天：
table1保存的user id有：１，２，３
table2保存的user id有： ２，４，５
talbe3保存的user id有：７
则统计出的不重复user id有：１，２，３，４，５，７共６个，所以返回结果应该是6

请问各位老大这个SQL怎么写啊？

[ljking2006]

select count(user_id) from(
select fld01 user_id from table1 where created between '2008-12-24' and '2008-12-25'
union
select fld02 user_id from table2 where created between '2008-12-24' and '2008-12-25'
union
select fld03 user_id from table3 where created between '2008-12-24' and '2008-12-25'
)

[sll853]

学习ing

[hbwhwang]

[Quote=引用 7 楼 ljking2006 的回复:]
select count(user_id) from(
select fld01 user_id from table1 where created between '2008-12-24' and '2008-12-25'
union
select fld02 user_id from table2 where created between '2008-12-24' and '2008-12-25'
union
select fld03 user_id from table3 where created between '2008-12-24' and '2008-12-25'
)
[/Quote]
加个DISTINCT就对了
select count(distinct user_id) from
select fld01 user_id from table1 where created between '2008-12-24' and '2008-12-25'
union
select fld02 user_id from table2 where created between '2008-12-24' and '2008-12-25'
union
select fld03 user_id from table3 where created between '2008-12-24' and '2008-12-25'
)

[sunnylyy]

楼上，用了union就不用distinct了。union all则需要用distinct

[zh2208]

select count(distinct user_id) from
select fld01 user_id from table1 where created between '2008-12-24' and '2008-12-25'
union
select fld02 user_id from table2 where created between '2008-12-24' and '2008-12-25'
union
select fld03 user_id from table3 where created between '2008-12-24' and '2008-12-25'
)

正解

[spring_sun1]

来晚了。。。。

[guolimin1118]

恩
union就是可以过滤重复id
当然也可以用distinct
来去掉重复字段

[socool627]

来晚了 。。。

[possibleonline]

学习

[凌霄野鹤]

接分

[lnfszl]

一楼说的很好啊，怎么不结帖啊

[SylvanLiu]

用union就不用distinct；
union all需要用distinct

[yupengt]

向高手致敬

[weir75034]

呵呵，没分了吧

[我不是大明]

一楼是对的,CSDN高手多,

[joynet007]

select count(user_id) from(
select fld01 user_id from table1 where created between '2008-12-24' and '2008-12-25'
union
select fld02 user_id from table2 where created between '2008-12-24' and '2008-12-25'
union
select fld03 user_id from table3 where created between '2008-12-24' and '2008-12-25'
)
就可以了！

[kakanote]

Ssafsdfsfd

[jiang_jiajia10]

UP

[懒虎]

select count(distinct user_id) from
select fld01 user_id from table1 where created between '2008-12-24' and '2008-12-25'
union
select fld02 user_id from table2 where created between '2008-12-24' and '2008-12-25'
union
select fld03 user_id from table3 where created between '2008-12-24' and '2008-12-25'
)

这个应该是正确的

[Sky-Yang]

select count(distinct user_id)
from table1,table2,table3
where created between 'time1' and 'time2' 

加

union

的话效率高一点