Question

我的桌子看起来像这样：

+------+-------+-------------+
ID_Loc  Type    Data
+------+-------+-------------+
ABC     RMKS    Hello
ABC     NAM     Joe Smith
ABD     NAM     Peter Hill
ABD     RMKS    Bye Bye
ABD     NAM     Freddy Tall
ABE     NAM     Loran Bennett
ABE     RMKS    Bye Bye
ABF     NAM     Liv Claris
ABF     RMKS    Bye Bye
+------+-------+-------------+

我需要选择所有ID_Loc WHERE DATA NOT LIKE'Hello'。我试过的时候：

SELECT distinct ID_loc FROM data_full WHERE DATA NOT LIKE '% Hello'

这也选择ID_Loc：'ABC'，其中包含Data中的'Hello'。另外，由于这会影响相当多的行或行，如果我可以将查询指向仅查看使用了类型RMKS的行，那将会很好。

我正在使用MS SQL Server 2008

SQL小提琴地址是： http://sqlfiddle.com/#!6/38130/6

任何帮助都会非常感激。

Answer 1

如果您需要选择没有与ID_Loc模式匹配的记录的'%Hello'值，请执行以下查询：

SELECT ID_loc
FROM data_full 
group by ID_Loc
having max(case 
             when DATA LIKE '%Hello' then 1
             else 0
           end) = 0;

结果如下：http://sqlfiddle.com/#!6/38130/33

如果您还需要应用Type = 'RMKS'过滤器，则可以在WHERE子句（sqlfiddle）中执行此操作：

SELECT ID_loc
FROM data_full 
where type = 'RMKS'
group by ID_Loc
having max(case 
             when DATA LIKE '%Hello' then 1
             else 0
           end) = 0;

Answer 2

您需要删除空格，同时添加了RMKS过滤器。

SELECT distinct ID_loc 
FROM data_full 
WHERE Type= 'RMKS'
and DATA NOT LIKE '%Hello'  --'%Hello%' if it can be between other text

编辑：如果ID_Loc,Type不是唯一的：

SELECT distinct ID_loc 
FROM data_full 
WHERE ID_loc NOT IN (SELECT ID_loc FROM  data_full WHERE Type= 'RMKS'
and DATA LIKE '%Hello')

http://sqlfiddle.com/#!6/59e9a/8

Answer 3

您需要删除空格并按类型

过滤

SELECT distinct ID_loc 
FROM data_full 
WHERE DATA NOT LIKE '%Hello%' And [TYPE] = 'RMKS'

Answer 4

首先从LIKE子句中删除空格，然后，您可以使用EXISTS子句过滤符合条件的结果：

SQL Fiddle Demo

<强>查询：

SELECT DISTINCT t.ID_loc 
FROM data_full t
WHERE NOT EXISTS (SELECT ID_loc 
              FROM data_full
              WHERE Data LIKE '%Hello' AND ID_loc = t.ID_loc )
      AND 
      EXISTS (SELECT ID_loc 
              FROM data_full
              WHERE Type = 'RMKS' AND ID_loc = t.ID_loc )

<强> Results ：

| ID_loc |
|--------|
|    ABD |
|    ABE |
|    ABF |

Answer 5

使用反向逻辑的简单解决方案，您需要where not exists

SELECT distinct t.ID_loc FROM data_full t 
where not exists (
   select 1 from data_full t2 
   where t2.ID_loc=t.ID_loc and t2.type='RMKS' and t2.data like '%Hello')
ORDER BY 1;

SQLFIDDLE DEMO

SQL选择WHERE列NOT LIKE跨多行

5 个答案:

SQL Fiddle Demo