WITH TEST AS
(
SELECT '1' ORD, 'CHAR' GRP, 'A' code FROM dual
UNION ALL SELECT '1' ORD,'CHAR' GRP, 'B' code FROM dual
UNION ALL SELECT '1' ORD,'CHAR' GRP, 'C' code FROM dual
UNION ALL SELECT '2' ORD,'NUM' GRP, '1' code FROM dual
UNION ALL SELECT '2' ORD,'NUM' GRP, '2' code FROM dual
UNION ALL SELECT '3' ORD,'KOR' GRP, '일' code FROM dual
UNION ALL SELECT '3' ORD,'KOR' GRP, '이' code FROM dual
UNION ALL SELECT '3' ORD,'KOR' GRP, '삼' code FROM dual
UNION ALL SELECT '3' ORD,'KOR' GRP, '사' code FROM dual
)
SELECT a.code || '-' || b.code || '-' || c.code code
  FROM (SELECT *
          FROM TEST
         WHERE ord = 1) a
     , (SELECT *
          FROM TEST
         WHERE ord = 2) b
     , (SELECT *
          FROM TEST
         WHERE ord = 3) c

일, 이, 삼, 사 같은 경우에는 그냥 소트하면 한글 순이기 때문에 이것에 대한 처리를 좀 해주셔야할 것 같네요.

SELECT a.code || '-' || a.ord || '-' || b.code
  FROM test a
     , test b

SELECT SUBSTR(SYS_CONNECT_BY_PATH(code, '-'), 2) x
  FROM test
 WHERE CONNECT_BY_ISLEAF = 1
 START WITH ord = 1
 CONNECT BY PRIOR ord + 1 = ord
;

 SELECT CONNECT_BY_ROOT ( code ) || '-' || PRIOR code || '-' || code
   FROM test
  WHERE CONNECT_BY_ISLEAF = 1
  START WITH ord = 1
CONNECT BY PRIOR ord = ord - 1

WITH TEST AS
 (
 SELECT '1' ORD, 'CHAR' GRP, 'A' code FROM dual
 UNION ALL SELECT '1' ORD,'CHAR' GRP, 'B' code FROM dual
 UNION ALL SELECT '1' ORD,'CHAR' GRP, 'C' code FROM dual
 UNION ALL SELECT '2' ORD,'NUM' GRP, '1' code FROM dual
 UNION ALL SELECT '2' ORD,'NUM' GRP, '2' code FROM dual
 UNION ALL SELECT '3' ORD,'KOR' GRP, '일' code FROM dual
 UNION ALL SELECT '3' ORD,'KOR' GRP, '이' code FROM dual
 UNION ALL SELECT '3' ORD,'KOR' GRP, '삼' code FROM dual
 UNION ALL SELECT '3' ORD,'KOR' GRP, '사' code FROM dual
 )
 select c.code||'-'||n.code||'-'||k.code
 from
 (select
 grp,code
 from test
 where grp='CHAR') c,
 (select
 grp,code
 from test
 where grp='NUM') n,
 (select
 grp,code
 from test
 where grp='KOR') k
 order by c.code,n.code,k.code;