Oracle中可以通过START WITH . . . CONNECT BY . . .子句来实现SQL的层次查询.
自从Oracle 9i开始,可以通过 SYS_CONNECT_BY_PATH 函数实现将父节点到当前行内容以“path”或者层次元素列表的形式显示出来。
自从Oracle 10g 中,还有其他更多关于层次查询的新特性 。例如,有的时候用户更关心的是每个层次分支中等级最低的内容。
那么你就可以利用伪列函数CONNECT_BY_ISLEAF来判断当前行是不是叶子。如果是叶子就会在伪列中显示“1”,
如果不是叶子而是一个分支(例如当前内容是其他行的父亲)就显示“0”。
在Oracle 10g 之前的版本中,如果在你的树中出现了环状循环(如一个孩子节点引用一个父亲节点),
Oracle 就会报出一个错误提示:“ ORA-01436: CONNECT BY loop in user data”。如果不删掉对父亲的引用就无法执行查询操作。
而在 Oracle 10g 中,只要指定“NOCYCLE”就可以进行任意的查询操作。与这个关键字相关的还有一个伪列——CONNECT_BY_ISCYCLE,
如果在当前行中引用了某个父亲节点的内容并在树中出现了循环,那么该行的伪列中就会显示“1”,否则就显示“0”。
The start with .. connect by clause can be used to select data that has a hierarchical relationship
(usually some sort of parent->child, boss->employee or thing->parts).
It is also being used when an sql execution plan is explained.
syntax:
select ... [start with initial-condition] connect by [nocycle] recurse-condition
level
With level it is possible to show the level in the hierarchical relation of all the data.
--oracle 9i
sys_connect_by_path
With sys_connect_by_path it is possible to show the entire path from the top level down to the 'actual' child.
--oracle 10g
connect_by_root
connect_by_root is a new operator that comes with Oracle 10g and enhances the ability to perform hierarchical queries.
connect_by_is_leaf
connect_by_isleaf is a new operator that comes with Oracle 10g and enhances the ability to perform hierarchical queries.
connect_by_iscycle
connect_by_is_cycle is a new operator that comes with Oracle 10g and enhances the ability to perform hierarchical queries.
下面举例说明:
[例1]创建一个部门表,这个表有4个字段,分别对应部门ID,部门名称,上级部门ID ,本部门直属员工人数。
create table DEP
(
DEPID number(10) not null,
DEPNAME varchar2(32),
UPPERDEPID number(10),
NUMOFEMP number(10)
) ;
下面插入一些数据,结果如下:
hr@MYTEST2> select * from dep;
DEPID DEPNAME UPPERDEPID NUMOFEMP
---------- --------------- ---------- ----------
0 Dev Center 2
1 DevA 0 3
2 DevA Team1 1 10
3 DevA Team2 1 8
4 DevB 0 2
5 DevB Team1 4 12
6 Test Center 1
7 Test Team1 6 5
8 Test Team2 6 5
9 rows selected.
现在根据“CONNECT BY”来实现树状查询:
column depname format a15;
column rootdep format a15;
column path format a30;
select rpad(' ', 2*(level-1), '-') || depname "DEPNAME",
connect_by_root depname "ROOTDEP",
connect_by_isleaf "ISLEAF",
level,
sys_connect_by_path(depname, '/') "PATH"
from dep
start with upperdepid is null
connect by prior depid = upperdepid
/
结果如下:
DEPNAME ROOTDEP ISLEAF LEVEL PATH
--------------- --------------- ---------- ---------- ---------------------------
Dev Center Dev Center 0 1 /Dev Center
-DevA Dev Center 0 2 /Dev Center/DevA
---DevA Team1 Dev Center 1 3 /Dev Center/DevA/DevA Team1
---DevA Team2 Dev Center 1 3 /Dev Center/DevA/DevA Team2
-DevB Dev Center 0 2 /Dev Center/DevB
---DevB Team1 Dev Center 1 3 /Dev Center/DevB/DevB Team1
Test Center Test Center 0 1 /Test Center
-Test Team1 Test Center 1 2 /Test Center/Test Team1
-Test Team2 Test Center 1 2 /Test Center/Test Team2
下面计算Dev Center和Test Center部门的总人数:
select ROOTDEPID, sum(numofemp) "TOTALEMP"
from (select connect_by_root depid "ROOTDEPID", numofemp from dep
start with upperdepid is null
connect by prior depid = upperdepid)
group by ROOTDEPID
/
结果如下:
ROOTDEPID TOTALEMP
---------- ----------
6 11
0 37
[例2]通过CONNECT BY用于十六进度转换为十进制
CREATE OR REPLACE FUNCTION f_hex_to_dec(p_str IN VARCHAR2) RETURN VARCHAR2 IS
----------------------------------------------------------------------------------------------------------------------
-- 对象名称: f_hex_to_dec
-- 对象描述: 十六进制转换十进制
-- 输入参数: p_str 十六进制字符串
-- 返回结果: 十进制字符串
-- 测试用例: SELECT f_hex_to_dec('78A') FROM dual;
----------------------------------------------------------------------------------------------------------------------
v_return VARCHAR2(4000);
BEGIN
SELECT SUM(DATA) INTO v_return
FROM (SELECT (CASE upper(substr(p_str, rownum, 1))
WHEN 'A' THEN '10'
WHEN 'B' THEN '11'
WHEN 'C' THEN '12'
WHEN 'D' THEN '13'
WHEN 'E' THEN '14'
WHEN 'F' THEN '15'
ELSE substr(p_str, rownum, 1)
END) * power(16, length(p_str) - rownum) DATA
FROM dual
CONNECT BY rownum <= length(p_str));
RETURN v_return;
EXCEPTION
WHEN OTHERS THEN
RETURN NULL;
END;
说明:
1. CONNECT BY rownum <= length(p_str))对输入的字符串进行逐个遍历
2. 通过CASE语句,来解析十六进制中的A-F对应的10进制值
测试结果如下:
hr@MYTEST2> variable dec varchar2(32);
hr@MYTEST2> exec :dec := f_hex_to_dec('1FF');
PL/SQL procedure successfully completed.
hr@MYTEST2> print dec
DEC
----------
511
[例3]通过CONNECT BY生成序列
对于connect by,现在大多数人已经很熟悉了,connect by中的条件就表示了父子之间的连接关系,比如 connect by id=prior pid。
但如果connect by中的条件没有表示记录之间的父子关系
那会出现什么情况?
常见的,connect by会在构造序列的时候使用
用select rownum from dual connect by rownum
我们注意到,dual是一个只有一条记录的表,如果表有多条记录,将会怎样?
下面开始实验
CREATE TABLE T(ID VARCHAR2(1 BYTE));
INSERT INTO T ( ID ) VALUES ( 'A');
INSERT INTO T ( ID ) VALUES ( 'B');
INSERT INTO T ( ID ) VALUES ( 'C');
COMMIT;
然后执行以下查询:
hr@MYTEST2> column id format a2;
hr@MYTEST2> select id,level from t connect by level<2;
ID LEVEL
-- ----------
A 1
B 1
C 1
hr@MYTEST2> select id,level from t connect by level<3;
ID LEVEL
-- ----------
A 1
A 2
B 2
C 2
B 1
A 2
B 2
C 2
C 1
A 2
B 2
C 2
12 rows selected.
无需多说,我们很快可以找到其中的规律,假设表中有N条记录
则记F(N,l)为 select id,level from t connect by level
那么,
F(N,1)=N
F(N,l) = F(N,l-1)*N+N
于是可以总结出
F(N,l)=∑power(N,p), p取值为[1,l)
要解释,也很容易。
当连接条件不能限制记录之间的关系时
每一条记录都可以作为自己或者其他记录的叶子
如下所示:
A 1A 2A 3
B 3
C 3
B 2A 3
B 3
C 3
C 2
A 3
B 3
C 3
在这里,我们看到的是
Oracle采用了深度优先的