如何在分层查询中删除不以叶节点结尾的列。可能有人共享示例
SELECT cust_id AS child_id
FROM customer
WHERE LEVEL = v_cnt
START WITH cust_id = 12660
CONNECT BY PRIOR cust_id = cust_cust_id
我为此编写了一个存储过程。
CREATE OR REPLACE PROCEDURE FRONTROOM.SP_CUST_HIER_DENORM_2
IS
L_ERR_MSG VARCHAR2(1000);
l_cust_id number(10);
v_cnt number(1);
BEGIN
SP_LOG_ENTRY(CURRENT_TIMESTAMP,'SP_CUST_HIER_DENORM_2','START',NULL,NULL,NULL);
-- select cust_id into l_cust_id from customer where true_gcdb_source_key ='9900000001';
delete CUST_HIER_DNORM_2;
for v_cnt in 2..6
loop
-- v_cnt := 6;
For j in ( SELECT cust_id AS child_id
FROM customer
WHERE LEVEL = v_cnt
START WITH cust_id = 12660
CONNECT BY PRIOR cust_id = cust_cust_id)
LOOP
begin
INSERT INTO CUST_HIER_DNORM_2
(LVL_1,LVL_2,LVL_3,LVL_4,LVL_5,LVL_6,LVL_7,LVL_8 )
SELECT SUM(DECODE(LEVEL,v_cnt,cust_id)) LVL_1 ,
SUM(DECODE(LEVEL,v_cnt-1,cust_id)) LVL_2 ,
SUM(DECODE(LEVEL,v_cnt-2,cust_id)) LVL_3 ,
SUM(DECODE(LEVEL,v_cnt-3,cust_id)) LVL_4 ,
SUM(DECODE(LEVEL,v_cnt-4,cust_id)) LVL_5 ,
SUM(DECODE(LEVEL,v_cnt-5,cust_id)) LVL_6 ,
SUM(DECODE(LEVEL,v_cnt-6,cust_id)) LVL_7 ,
SUM(DECODE(LEVEL,v_cnt-7,cust_id)) LVL_8
FROM CUSTOMER
WHERE LEVEL <=v_cnt
START WITH cust_id = j.child_id
CONNECT BY cust_id = PRIOR cust_cust_id ;
EXCEPTION
WHEN OTHERS
THEN
SP_LOG_ENTRY (CURRENT_TIMESTAMP,
'SP_CUST_HIER_DENORM_2',
'ERROR WHILE INSERTING ',
SQLERRM,
'* ' || j.CHILD_ID || ' *',NULL
);
END;
END LOOP ;
end loop;
COMMIT;
UPDATE cust_hier_dnorm_2 cd
SET LVL_1_DESC =
(SELECT c.true_gcdb_source_key
FROM customer c
WHERE cd.lvl_1 = C.CUST_ID ),
LVL_2_DESC =
(SELECT c.true_gcdb_source_key
FROM customer c
WHERE cd.lvl_2 = C.CUST_ID ),
LVL_3_DESC =
(SELECT c.true_gcdb_source_key
FROM customer c
WHERE cd.lvl_3 = C.CUST_ID ),
LVL_4_DESC =
(SELECT c.true_gcdb_source_key
FROM customer c
WHERE cd.lvl_4 = C.CUST_ID ),
LVL_5_DESC =
(SELECT c.true_gcdb_source_key
FROM customer c
WHERE cd.lvl_5 = C.CUST_ID ),
LVL_6_DESC =
(SELECT c.true_gcdb_source_key
FROM customer c
WHERE cd.lvl_6 = C.CUST_ID ),
LVL_7_DESC =
(SELECT c.true_gcdb_source_key
FROM customer c
WHERE cd.lvl_7 = C.CUST_ID ),
LVL_8_DESC =
(SELECT c.true_gcdb_source_key
FROM customer c
WHERE cd.lvl_8 = C.CUST_ID );
commit;
SP_LOG_ENTRY(CURRENT_TIMESTAMP,'SP_CUST_HIER_DENORM_2','START',NULL,NULL,NULL);
EXCEPTION
WHEN OTHERS THEN
L_ERR_MSG :=SQLERRM;
SP_LOG_ENTRY(CURRENT_TIMESTAMP,'SP_LOAD_CUST_HIER_DENORM_2',NULL,L_ERR_MSG,NULL,NULL);
END;
/
目前我的输出采用这种格式
lvl 1 lvl2 lvl3 lvl4 lvl5 lvl6
12660 12389 12517 12476 12557 3953(leaf node)
12660 12389 12517 12476 12557 3978(leaf node)
12660 12389 12517 12476 12557 3992(leaf node)
12660 12389 12517 12476 12557
12660 12389 12517 12476
12660 12389 12517
desired output
lvl 1 lvl2 lvl3 lvl4 lvl5 lvl6
12660 12389 12517 12476 12557 3953(leaf node)
12660 12389 12517 12476 12557 3978(leaf node)
12660 12389 12517 12476 12557 3992(leaf node)
表示删除不以叶节点结尾的列。
它是一个denorm table.all将所有记录插入到非规范化表中
答案 0 :(得分：0)
目前还不太清楚你想要得到什么。如果要仅提取叶节点，可以使用connect_by_isleaf：
SQL> select mgr, ename, empno, sal, connect_by_isleaf from emp
2 start with mgr is null
3 connect by prior empno = mgr
4 /
MGR ENAME EMPNO SAL CONNECT_BY_ISLEAF
---------- ---------- ---------- ---------- -----------------
KING 7839 5000 0
7839 JONES 7566 2975 0
7566 FORD 7902 3000 0
7902 SMITH 7369 800 1
7839 BLAKE 7698 2850 0
7698 ALLEN 7499 1600 1
7698 WARD 7521 1250 1
7698 MARTIN 7654 1250 1
7698 TURNER 7844 1500 1
7698 JAMES 7900 950 1
7839 CLARK 7782 2450 0
7782 MILLER 7934 1300 1
SQL> select mgr, ename, sal from emp
2 where connect_by_isleaf = 1
3 start with mgr is null
4 connect by prior empno = mgr
5 /
MGR ENAME SAL
---------- ---------- ----------
7902 SMITH 800
7698 ALLEN 1600
7698 WARD 1250
7698 MARTIN 1250
7698 TURNER 1500
7698 JAMES 950
7782 MILLER 1300
如果你想只留下有叶子的分支，你可以使用类似的东西：
SQL> select mgr,ename, sal from (
2 -- Just branches
3 select mgr, ename, empno, sal from emp
4 where connect_by_isleaf = 0
5 start with mgr is null
6 connect by prior empno = mgr
7 )
8 -- low-level branches
9 where connect_by_isleaf = 1
10 start with mgr is null
11 connect by prior empno = mgr
12 /
MGR ENAME SAL
---------- ---------- ----------
7566 FORD 3000
7839 BLAKE 2850
7839 CLARK 2450