SQL search and destroy duplicates

I have a table with fields (simplified):

id, fld1, fld2, fld3.

id is a numeric primary key field.

There are duplicates: id differs but fld1, fld2 and fld3 are identical over 2 or more rows. There are also entries where the values occur only once, i.e. non-duplicates, of course.

Of each set of duplicate entries, I want to retain only the entry with the highest ID. I was planning to first list the doomed rows and then to delete them.

My first stab at it was this:

SELECT * FROM tab1 t1 WHERE EXISTS (
  SELECT COUNT(*) FROM tab1 t2
  WHERE t1.fld1 = t2.fld1 AND t1.fld2 = t2.fld2 AND t1.fld3 = t2.fld3
    AND t1.id < MAX(t2.id)
  HAVING COUNT(*) > 1
  GROUP BY t2.fld1, t2.fld2, t2.fld3)

But (in Oracle) I'm getting a Missi开发者_Python百科ng right parenthesis error message. I think this needs a new approach altogether, but my SQL-fu is not up to the task. Help appreciated!

---

Edit:

With 'real' data fields:

select x.leg_id, x.airline_des, x.flight_nr, x.suffix, x.flight_id_date, x.lt_flight_id_date
from fdb_leg x
join ( select max(t.leg_id) 'max_id', 
t.airline_des, t.flight_nr, t.suffix, t.flight_id_date, t.lt_flight_id_date
from fdb_leg t
group by t.airline_des, t.flight_nr, t.suffix, t.flight_id_date, t.lt_flight_id_date
having count(*) > 1) y on y.max_id > x.leg_id
and y.airline_des = x.airline_des and y.flight_nr = x.flight_nr and y.suffix = x.suffix
and y.flight_id_date = x.flight_id_date and x.lt_flight_id_date = y.lt_flight_id_date

Response is:

ORA-00923: FROM keyword not found where expected

---

Oracle 9i+, Using WITH:

---

To get the list of doomed entries, use:

WITH keepers AS (
   SELECT MAX(t.id) 'max_id',
          t.fld1, t.fld2, t.fld3
     FROM TABLE_1 t
 GROUP BY t.fld1, t.fld2, t.fld3
   HAVING COUNT(*) > 1)
SELECT x.id,
       x.fld1, x.fld2, x.fld3
  FROM TABLE_1 x
  JOIN keepers y ON y.max_id > x.id
                AND y.fld1 = x.fld1
                AND y.fld2 = x.fld2
                AND y.fld3 = x.fld3

Non-WITH Equivalent:

---

To get the list of doomed entries, use:

SELECT x.id,
       x.fld1, x.fld2, x.fld3
  FROM TABLE_1 x
  JOIN (SELECT MAX(t.id) 'max_id',
               t.fld1, t.fld2, t.fld3
          FROM TABLE_1 t
      GROUP BY t.fld1, t.fld2, t.fld3
        HAVING COUNT(*) > 1) y ON y.max_id > x.id
                              AND y.fld1 = x.fld1
                              AND y.fld2 = x.fld2
                              AND y.fld3 = x.fld3

---

You can delete them in one shot, like this:

SQL> create table mytable (id, fld1, fld2, fld3)
  2  as
  3  select 1, 1, 1, 1 from dual union all
  4  select 2, 1, 1, 1 from dual union all
  5  select 3, 2, 2, 2 from dual union all
  6  select 4, 2, 3, 2 from dual union all
  7  select 5, 2, 3, 2 from dual union all
  8  select 6, 2, 3, 2 from dual
  9  /

Table created.

SQL> delete mytable
  2   where id not in
  3         ( select max(id)
  4             from mytable
  5            group by fld1
  6                , fld2
  7                , fld3
  8         )
  9  /

3 rows deleted.

SQL> select * from mytable
  2  /

        ID       FLD1       FLD2       FLD3
---------- ---------- ---------- ----------
         2          1          1          1
         3          2          2          2
         6          2          3          2

3 rows selected.

Regards, Rob.

---

Ugh, I get it. Scratch that.

This will identify the ID's needed to delete.

Select 
       fld1
     , fld2
     , fld3
     , Max(ID)
From table_name
Group By
       fld1
     , fld2
     , fld3