【转】 oracle 层次查询判断叶子和根节点

Oracle 9i判断是叶子或根节点，是比较麻烦的一件事情，SQL演示脚本如下:

DROP TABLE idb_hierarchical;
create TABLE idb_hierarchical
(
id number,
parent_id number,
str varchar2(10)
);
insert into idb_hierarchical values(1,null,'A');
insert into idb_hierarchical values(2,1,'B');
insert into idb_hierarchical values(3,2,'C');
insert into idb_hierarchical values(4,3,'D');
insert into idb_hierarchical values(5,2,'E');
insert into idb_hierarchical values(6,2,'F');
insert into idb_hierarchical values(7,3,'G');
insert into idb_hierarchical values(8,4,'H');
insert into idb_hierarchical values(9,4,'I');
insert into idb_hierarchical values(10,null,'J');
insert into idb_hierarchical values(11,10,'K');
insert into idb_hierarchical values(12,11,'L');
insert into idb_hierarchical values(13,10,'M');

DROP TABLE idb_hierarchical;
create TABLE idb_hierarchical
(
id number,
parent_id number,
str varchar2(10)
);
insert into idb_hierarchical values(1,null,'A');
insert into idb_hierarchical values(2,1,'B');
insert into idb_hierarchical values(3,2,'C');
insert into idb_hierarchical values(4,3,'D');
insert into idb_hierarchical values(5,2,'E');
insert into idb_hierarchical values(6,2,'F');
insert into idb_hierarchical values(7,3,'G');
insert into idb_hierarchical values(8,4,'H');
insert into idb_hierarchical values(9,4,'I');
insert into idb_hierarchical values(10,null,'J');
insert into idb_hierarchical values(11,10,'K');
insert into idb_hierarchical values(12,11,'L');
insert into idb_hierarchical values(13,10,'M');

示例数据清单如下:

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL
FROM idb_hierarchical
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL
FROM idb_hierarchical
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

表1:数据清单
STR_LEVEL	ID	PARENT_ID	LVL
+..A	1		1
+….B	2	1	2
+……C	3	2	3
+……..D	4	3	4
+……….H	8	4	5
+……….I	9	4	5
+……..G	7	3	4
+……E	5	2	3
+……F	6	2	3
+..J	10		1
+….K	11	10	2
+……L	12	11	3
+….M	13	10	2

在表1中，ID为8、9、 7、5、6、12、13都没有子节点，因此称为叶节点。

1.oracle9i 查询叶节点

只显示叶子节点SQL

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL
FROM idb_hierarchical I
--在oracle 9i中显示叶节点，需要判断是否有子节点即可
WHERE NOT EXISTS(SELECT 1
FROM idb_hierarchical B
WHERE I.ID=B.PARENT_ID)
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL
FROM idb_hierarchical I
--在oracle 9i中显示叶节点，需要判断是否有子节点即可
WHERE NOT EXISTS(SELECT 1
FROM idb_hierarchical B
WHERE I.ID=B.PARENT_ID)
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

表2
STR_LEVEL	ID	PARENT_ID	LVL
+……….H	8	4	5
+……….I	9	4	5
+……..G	7	3	4
+……E	5	2	3
+……F	6	2	3
+……L	12	11	3
+….M	13	10	2

显示所有节点，标明该行是否为叶节点SQL

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL,
NVL((SELECT 'N'
FROM idb_hierarchical B
WHERE I.ID=B.PARENT_ID
AND ROWNUM < 2),'Y') IS_LEAF
FROM idb_hierarchical I
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL,
NVL((SELECT 'N'
FROM idb_hierarchical B
WHERE I.ID=B.PARENT_ID
AND ROWNUM < 2),'Y') IS_LEAF
FROM idb_hierarchical I
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

表3
STR_LEVEL	ID	PARENT_ID	LVL	IS_LEAF
+..A	1		1	N
+....B	2	1	2	N
+......C	3	2	3	N
+........D	4	3	4	N
+..........H	8	4	5	Y
+..........I	9	4	5	Y
+........G	7	3	4	Y
+......E	5	2	3	Y
+......F	6	2	3	Y
+..J	10		1	N
+....K	11	10	2	N
+......L	12	11	3	Y
+....M	13	10	2	Y

oracle 9i 查询根节点

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL
FROM idb_hierarchical I
START WITH id =2
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL
FROM idb_hierarchical I
START WITH id =2
CONNECT BY PARENT_ID = PRIOR ID;

表4
STR_LEVEL	ID	PARENT_ID	LVL
+..B	2	1	1
+....C	3	2	2
+......D	4	3	3
+........H	8	4	4
+........I	9	4	4
+......G	7	3	3
+....E	5	2	2
+....F	6	2	2

根节点ID应该为3、5、6，即lvl为1即可

查询根节点，只显示根节点SQL

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
LEVEL LVL,
(select b.str
from idb_hierarchical b
where level = 1
start with b.id = 2
connect by prior b.id = b.parent_id
) root_str
FROM idb_hierarchical I
where level = 1
START WITH id = 2
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
LEVEL LVL,
(select b.str
from idb_hierarchical b
where level = 1
start with b.id = 2
connect by prior b.id = b.parent_id
) root_str
FROM idb_hierarchical I
where level = 1
START WITH id = 2
CONNECT BY PARENT_ID = PRIOR ID;

表5
STR_LEVEL	ID	PARENT_ID	LVL	ROOT_STR
+..B	2	1	1	B

标明根节点SQL

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
DECODE(LEVEL, 1, 'Y', 'N') is_root,
LEVEL LVL,
(select b.str
from idb_hierarchical b
where level = 1
start with b.id = 2
connect by prior b.id = b.parent_id) root_str
FROM idb_hierarchical I
START WITH id = 2
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
DECODE(LEVEL, 1, 'Y', 'N') is_root,
LEVEL LVL,
(select b.str
from idb_hierarchical b
where level = 1
start with b.id = 2
connect by prior b.id = b.parent_id) root_str
FROM idb_hierarchical I
START WITH id = 2
CONNECT BY PARENT_ID = PRIOR ID;

表6
STR_LEVEL	ID	PARENT_ID	IS_ROOT	LVL	ROOT_STR
+..B	2	1	Y	1	B
+....C	3	2	N	2	B
+......D	4	3	N	3	B
+........H	8	4	N	4	B
+........I	9	4	N	4	B
+......G	7	3	N	3	B
+....E	5	2	N	2	B
+....F	6	2	N	2	B

在oracle 10g提供了connect_by_isleaf和connect_by_root

oracle 10g用connect_by_isleaf判断叶节点

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL
FROM idb_hierarchical I
where connect_by_isleaf=1
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL
FROM idb_hierarchical I
where connect_by_isleaf=1
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

表7
STR_LEVEL	ID	PARENT_ID	LVL
+..........H	8	4	5
+..........I	9	4	5
+........G	7	3	4
+......E	5	2	3
+......F	6	2	3
+......L	12	11	3
+....M	13	10	2

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL,
decode(connect_by_isleaf,1,'Y','N') IS_LEAF
FROM idb_hierarchical I
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+',LEVEL*2+1,'.')||STR STR_LEVEL,ID,PARENT_ID,LEVEL LVL,
decode(connect_by_isleaf,1,'Y','N') IS_LEAF
FROM idb_hierarchical I
START WITH PARENT_ID IS NULL
CONNECT BY PARENT_ID = PRIOR ID;

表8
STR_LEVEL	ID	PARENT_ID	LVL	IS_LEAF
+..A	1		1	N
+....B	2	1	2	N
+......C	3	2	3	N
+........D	4	3	4	N
+..........H	8	4	5	Y
+..........I	9	4	5	Y
+........G	7	3	4	Y
+......E	5	2	3	Y
+......F	6	2	3	Y
+..J	10		1	N
+....K	11	10	2	N
+......L	12	11	3	Y
+....M	13	10	2	Y

oracle 10g用connect_by_root判断根节点

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
LEVEL LVL,
connect_by_root STR ROOT_STR
FROM idb_hierarchical I
START WITH id = 2
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
LEVEL LVL,
connect_by_root STR ROOT_STR
FROM idb_hierarchical I
START WITH id = 2
CONNECT BY PARENT_ID = PRIOR ID;

表9
STR_LEVEL	ID	PARENT_ID	LVL	ROOT_STR
+..B	2	1	1	B
+....C	3	2	2	B
+......D	4	3	3	B
+........H	8	4	4	B
+........I	9	4	4	B
+......G	7	3	3	B
+....E	5	2	2	B
+....F	6	2	2	B

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
DECODE(LEVEL, 1, 'Y', 'N') is_root,
LEVEL LVL,
connect_by_root STR ROOT_STR
FROM idb_hierarchical I
START WITH id = 3
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
DECODE(LEVEL, 1, 'Y', 'N') is_root,
LEVEL LVL,
connect_by_root STR ROOT_STR
FROM idb_hierarchical I
START WITH id = 3
CONNECT BY PARENT_ID = PRIOR ID;

表10
STR_LEVEL	ID	PARENT_ID	IS_ROOT	LVL	ROOT_STR
+..C	3	2	Y	1	C
+....D	4	3	N	2	C
+......H	8	4	N	3	C
+......I	9	4	N	3	C
+....G	7	3	N	2	C

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
DECODE(LEVEL, 1, 'Y', 'N') is_root,
LEVEL LVL,
connect_by_root STR ROOT_STR
FROM idb_hierarchical I
START WITH PARENT_ID = 2
CONNECT BY PARENT_ID = PRIOR ID;

SELECT RPAD('+', LEVEL * 2 + 1, '.') || STR STR_LEVEL,
ID,
PARENT_ID,
DECODE(LEVEL, 1, 'Y', 'N') is_root,
LEVEL LVL,
connect_by_root STR ROOT_STR
FROM idb_hierarchical I
START WITH PARENT_ID = 2
CONNECT BY PARENT_ID = PRIOR ID;

表11
STR_LEVEL	ID	PARENT_ID	IS_ROOT	LVL	ROOT_STR
+..C	3	2	Y	1	C
+....D	4	3	N	2	C
+......H	8	4	N	3	C
+......I	9	4	N	3	C
+....G	7	3	N	2	C
+..E	5	2	Y	1	E
+..F	6	2	Y	1	F

【转】 oracle 层次查询判断叶子和根节点的更多相关文章

Oracle 层次查询 connect by
oracle 层次查询语法: SELECT ... FROM [WHERE condition] --过 ...
带您了解Oracle层次查询
http://database.51cto.com/art/201010/231539.htm Oracle层次查询(connect by )是结构化查询中用到的,下面就为您介绍Oracle层次查询的 ...
Oracle层次查询
Oracle层次查询的语法如下: 下面根据两道“烧脑”的题具体来体现: 1. 根据时间先后顺序,十二星座的英文名称用逗号串起来为'Aries,Taurus,Gemini,Cancer,Leo,Virg ...
Oracle - 层次查询
如果表中含有层次数据,可以通过使用层次查询有序地查看层次数据. 语法: condition:指一个或多个表达式和逻辑(布尔)运算符的组合,并返回TRUE.FALSE或UNKNOWNstart with ...
Oracle层次查询start with connect by
博客参考:https://www.cnblogs.com/jerryxing/articles/2339352.html start with connect by 层次查询(Hierarchical ...
【转载】Oracle层次查询和分析函数
摘要一组连续的数,去掉中间一些数,如何求出剩下的数的区间(即号段)?知道号段的起止,如何求出该号段内所有的数?知道一个大的号段范围和已经取过的号段,如何求出可用的号段?利用Oracle提供的强大的查 ...
Oracle层次查询和分析函数在号段选取中的应用
转自:http://www.itpub.net/thread-719692-1-1.html 摘要一组连续的数,去掉中间一些数,如何求出剩下的数的区间(即号段)?知道号段的起止,如何求出该号段内所有的 ...
oracle层次查询的陷阱
今天开发组同事找到我,说一个简单的层次查询非常慢,业务就是有一个存设备表连接关系的表,从node1连入,从node2连出,现在要找出node2的连出顺序,sql类似于: SELECT LEVEL ID ...
Oracle层次查询和with函数的使用
开发中大家应该都做过什么类似部门管理这样的功能,一般情况下一个部门下面还有下一级部门(子部门),这个层级就类似一棵树.这种情况下一般会把父级部门和子级部门分成2个或者多个表,这种算是比较常规的做法:有 ...

随机推荐

SQL2000的系统表sysproperties在SQL2005中无效的问题
有两种解决办法方法一.是我在网上找的:将原来的sysproperties改成sys.extended_properties并且对应关系如下 sys.extended_properties left ...
golang学习笔记 ---dup
对文件做拷贝.打印.搜索.排序.统计或类似事情的程序都有一个差不多的程序结构:一个处理输入的循环,在每个元素上执行计算处理,在处理的同时或最后产生输出.我们会展示一个名为 dup 的程序的三个版本 p ...
Innodb的内存结构
1.缓冲池从1.0.x版本开始,允许有多个缓冲池实例. mysql> show variables like 'innodb_buffer_pool_size'\G ************** ...
ArcGIS 10.2数字化线状要素时自己主动拼接成一条线
一開始一直和捕捉和经典捕捉过不去.最后才发现找错了地方,原来要Customize.自己定义加入"Continue Feature"工具,将其拖动到界面上的随意工具条内就可以.如图1 ...
数据库事务隔离级别+Spring 声明性事务隔离级别
数据库事务隔离级别数据库提供了四种事务隔离级别, 不同的隔离级别采用不同的锁类开来实现. 在四种隔离级别中, Serializable的级别最高, Read Uncommited级别最低. 大多数数 ...
How to get current timestamps in Java
How to get current timestamps in Java Timestamp timestamp = new Timestamp(System.currentTimeMillis() ...
用MATLAB做T检验（ttest）
t-检验: t-检验,又称student‘s t-test,可以用于比较两组数据是否来自同一分布(可以用于比较两组数据的区分度),假设了数据的正态性,并反应两组数据的方差在统计上是否有显著差异. ma ...
WPF中动态改变控件显示位置
转自 http://blog.csdn.net/lassewang/article/details/6928897 测试环境: Windows XP/Windows 7 开发环境: Microsoft ...
MySql（十）：MySQL性能调优——MySQL Server性能优化
本章主要通过针对MySQL Server( mysqld)相关实现机制的分析,得到一些相应的优化建议.主要涉及MySQL的安装以及相关参数设置的优化,但不包括mysqld之外的比如存储引擎相关的参数优 ...
PCIE xilinx v5 IP核使用前的研究
外带一个月前啃的一个星期,加本星期心无旁骛,啃出些心得,广惠后人.但愿有用. trn信号是数据链路层的信号 TLP包是数据链路层传给transaction层的包解包需要一个transaction的协 ...