Question

I am inserting data into a table by selecting from another table which may have duplicates. I thought my query was handling this by checking if the row already exists but I am getting a unique constraint violation.

Here is the query:

INSERT INTO FOLDER_USER (FOLDER_ID, USER_ID) 
    SELECT DECODE(FOLDERID,'F10', '1','F565','2','F11', '3','F81', '4','0'), USERID
    FROM DATA1.FOLDERS F1
    WHERE UPPER(OWNER) = 'ADMIN'
    AND NOT EXISTS 
        (SELECT 1 FROM FOLDER_USER F2  
         WHERE DECODE(FOLDERID,'F10', '1','F565','2','F11', '3','F81', '4','0')= F2.FOLDER_ID 
         AND F1.USERID = F2.USER_ID);

• Table FOLDER_USER contains 2 columns FOLDER_ID (number), USER_ID
  (varchar) and combined they make up the primary key
• Table FOLDERS contains 2 columns FOLDERID and USERID (both varchars). The value
  in FOLDERID needs to be decoded into a number depending on its value
  before being inserted into the new table

Answer 1

Perhaps the rows you are selecting are not unique within themselves.

Try:

INSERT INTO folder_user( folder_id,
                         user_id)
   SELECT UNIQUE 
          DECODE( folderid,  'F10', '1',  'F565', '2',  'F11', '3',  'F81', '4',  '0' ),
          userid
     FROM data1.folders f1
    WHERE UPPER( owner ) = 'ADMIN'
          AND NOT EXISTS
                 (SELECT 1
                    FROM folder_user f2
                   WHERE DECODE( folderid,  'F10', '1',  'F565', '2',  'F11', '3',  'F81', '4',  '0' ) = f2.folder_id
                         AND f1.userid = f2.user_id);