Oracle 删除重复记录

DELETE FROM TABLE WHERE ROWID NOT IN
(SELECT MAX(rowID) FROM TABLE GROUP BY NEWCODE)
解释：
      1：SELECT MAX(rowID) FROM TABLE GROUP BY NEWCODE，按照新的字段（newcode）分组；

      2：然后查找分组后的rowId，其中rowId可以为MIN(rowId)或MAX（rowId）；
     
      3：DELETE FROM TABLE WHERE ROWID NOT IN (SELECT MAX(rowID) FROM TABLE GROUP BY NEWCODE) 进行删除重复记录
      4：该语句执行效果：当newCode中有一列属性有重复值时，就会删除；
     5：游标是对多行数据结果的一个操作入口，通过遍历游标，我们可以对访问到每行记录的数据，并通过游标取出值，从而进行操作。当select出多行结果集时就会用到cousor.不过游标的缺点就是消耗资源。

项目中遇到的实际问题

declare
  v_sql   varchar2(4000);
  v_table varchar2(4000);
  cursor CUR_DETAILID is
    select tname from tab where tname like 'X\_%' escape '\';
begin
  delete from C_DETAIL_FEATURE_RE;
  open CUR_DETAILID;
  loop
    fetch CUR_DETAILID
      into v_table;
    exit when CUR_DETAILID%notfound;
 
    v_sql := 'insert into C_DETAIL_FEATURE_RE select  S_C_DETAIL_FEATURE_RE.NEXTVAL,D_DETAILID,D_FNO,D_FID FROM ' ||
             v_table;
 
    dbms_output.put_line(v_sql);
    execute immediate v_sql;
    commit;
  end loop;
  close CUR_DETAILID;
  ----去除重复
  delete from C_DETAIL_FEATURE_RE
   where rowid not in (select max(rowid)
                         from C_DETAIL_FEATURE_RE
                        group by (d_fid, d_detailid, d_fno));
commit;
end;