我的要求似乎并不那么困难,但我不确定最好的方法。
我有下表:
userID file
1 1
1 2
1 3
2 1
2 3
3 2
4 1
4 2
我想选择仅具有文件编号2的UserID。在我的示例中,结果仅为3.
答案 0 :(得分:5)
SELECT userID
FROM tableName a
WHERE file = 2
GROUP BY userID
HAVING COUNT(*) =
(
SELECT COUNT(*)
FROM tableName b
WHERE a.userID = b.userID
)
答案 1 :(得分:1)
您可以使用WHERE NOT EXISTS
(SQL Fiddle):
SELECT UserId
FROM theTable t
WHERE file = 2
AND NOT EXISTS(SELECT 1 FROM theTable tt WHERE t.UserId = tt.UserId AND file <> 2)
GROUP BY UserId
或自我反连接(SQL Fiddle):
SELECT t.UserId
FROM theTable t
LEFT OUTER JOIN theTable tt ON t.UserId = tt.UserId AND tt.file <> 2
WHERE t.file = 2
AND tt.UserId IS NULL
GROUP BY t.UserId
答案 2 :(得分:0)
尝试此查询 -
SELECT userID FROM table
GROUP BY userID
HAVING COUNT(IF(file = 2, 1, NULL)) = COUNT(*)
+--------+
| userID |
+--------+
| 3 |
+--------+
答案 3 :(得分:0)
一种方法是通过一个标志IF的标志,除了用户的总记录外,他们还有一个#2文件......如果两者都是1,那么你就是好的。但是,此版本会查询所有用户。
select
userID
from
YourTable
group by
UserID
having
sum( if( file = 2, 1, 0 )) = 1
AND count(*) = 1
此版本将预先查询应用于仅列出那些最低限度拥有#2文件的人,然后获取总计数
select
PreQuery.UserID
from
( select distinct YT1.userID
from YourTable YT1
where YT1.File = 2 ) as PreQuery
JOIN YourTable YT2
on PreQuery.UserID = YT2.UserID
group by
PreQuery.UserID
having
count(*) = 1
我显然不知道您的数据大小(记录计数),但如果您有10k用户和100k文件,但只有250个可以访问文件#2,那么第二个查询将只关注250个用户,然后只开出那些只有HAVING count = 1的单一显式文件库的那些。