在 Postgresql 中的列值上使用 where 子句更新表

Question

I have a table name table like this

| label_id| label_name | user_id|
----------------------------------
|    1    | insvt1     |   1    |
|    2    | invest2    |   1    |
|    3    | invest3    |   1    |
|    4    | ivsest3    |   2    |
|    5    | invest4    |   3    |

I want to update this user_id column for user_id 1,1,1 with user_id 2 if label_name not same in case label name same like row 3 and row 4 then no update will occur

The result should be like this after update

| label_id| label_name | user_id|
----------------------------------
|    1    | insvt1     |   2    |
|    2    | invest2    |   2    |
|    3    | invest3    |   1    |
|    4    | ivsest3    |   2    |
|    5    | invest4    |   3    |

I have tried this

UPDATE data_table t, data_table t1
   SET t.user_id = 2
 WHERE t.label_name <> t1.label_name (!= this also)
   AND (t.user_id = 1 or t1.user_id=2)

it's updating but also update where both lebel_name same

Any help will be appreciated

Answer 1

the code below should do it:

update data_table
SET [user_id] = 2
WHERE label_name IN ( SELECT label_name FROM data_table GROUP BY label_name HAVING COUNT(*) = 1) AND [user_id] = 1

Explanation: you update the data table, but ONLY those rows, for which the label description occurs only once. so that criteria i wrote in the label_name IN (....)

Answer 2

You code looks like MySQL. That suggests using JOIN for the logic:

UPDATE data_table t JOIN
       (SELECT label_name, COUNT(*) as cnt
        FROM data_table t1
        GROUP BY label_name
        HAVING COUNT(*) = 1
       ) tt
       ON t.label_name = tt.label_name
   SET t.user_id = 2
   WHERE t.user_id = 1;

In Postgres, this looks like:

UPDATE data_table t 
   SET t.user_id = 2
   FROM (SELECT label_name, COUNT(*) as cnt
         FROM data_table t1
         GROUP BY label_name
         HAVING COUNT(*) = 1
        ) tt
    WHERE t.label_name = tt.label_name AND t.user_id = 1;