Select 具有一个值但没有来自同一个表的另一个值的行

Question

I have a table like below in MySQL. I want all the rows that have name A but not name B.

I want a result like this. Thank you for helping.

I have tried the following query. It is not working.

Select fk_id from TableA
where  name  not like 'B'
group by fk_id
having  name like 'A'

Answer 1

Your question isn't particularly clear but it sounds like you want fk_id having an associated name value of A and where any record with the same fk_id and name of B does not exist.

Try this

SELECT fk_id FROM TableA a
WHERE a.name = 'A'
AND NOT EXISTS (
  SELECT 1 FROM TableA b
  WHERE b.name = 'B'
  AND a.fk_id = b.fk_id
)

Demo ~ http://sqlfiddle.com/#!9/ba0df2/2

Answer 2

WITH b_rows AS(
    SELECT fk_id  AS fk_id_b,
           '1'    AS flg_b
    FROM TableA
    WHERE name  = 'B'
)
SELECT fk_id
FROM TableA
    LEFT JOIN b_rows ON TableA.fk_id = b_rows.fk_id_b
WHERE name  = 'A'
    AND flg_b <> '1' 

Answer 3

How about using GROUP_CONCAT() in HAVING like:

SELECT fk_id  
FROM TableA
GROUP BY fk_id
HAVING GROUP_CONCAT(name) LIKE '%A%' 
   AND GROUP_CONCAT(name) NOT LIKE '%B%';

Basically, the result of GROUP_CONCAT(name) with GROUP BY fk_id returns the following:

+--------------------+-------+
| GROUP_CONCAT(name) | fk_id |
+--------------------+-------+
| C                  | l     |
| A,B                | m     |
| A,C                | n     |
+--------------------+-------+

So we can make use of that result to do the filtering in HAVING . This query is similar to:

SELECT fk_id
  FROM
   (SELECT GROUP_CONCAT(name) AS names, fk_id
    FROM TableA
     GROUP BY fk_id) A
  WHERE names LIKE '%A%'
    AND names NOT LIKE '%B%';

This instead is using a sub-query then doing the filtering afterward using WHERE .

Demo Fiddle