使用MERGE语句并在ROWID伪列上关联：

MERGE INTO test_csv dst
USING (
  SELECT row_number() over (partition by user_id order by user_id) as seq 
  FROM   test_csv
) src
ON (src.ROWID = dst.ROWID)
WHEN MATCHED THEN
  UPDATE
  SET id = seq;

或相关的UPDATE：

UPDATE test_csv dst
SET id = (
           SELECT seq
           FROM   (
             SELECT row_number() over (partition by user_id order by user_id) as seq 
             FROM   test_csv
           ) src
           WHERE src.ROWID = dst.ROWID  
         );

其中，对于样本数据：

CREATE TABLE test_csv (User_id, User_name, ID) AS
SELECT 123, 'ABC', CAST(NULL AS NUMBER) FROM DUAL UNION ALL
SELECT 456, 'BCD', NULL FROM DUAL UNION ALL
SELECT 123, 'ABC', NULL FROM DUAL UNION ALL
SELECT 123, 'ABC', NULL FROM DUAL UNION ALL
SELECT 234, 'XYZ', NULL FROM DUAL;

然后，在MERGE或UPDATE之后，该表包含：
| 用户名|ID| ID |
| - -----|- -----| ------------ |
| ABC|一个| 1 |
| BCD|一个| 1 |
| ABC| 2| 2 |
| ABC|三个| 3 |
| XYZ|一个| 1 |
fiddle