Oracle的层次查询（oracle树查询）

备注:测试数据库版本为Oracle 11g R2

这个blog我们来聊聊Oracle的层次查询

级联查询的一些语法:

select … from tablename
start with 条件1
connect by 条件2
where 条件3;

start with: 指定起始节点的条件
connect by: 指定父子行的条件关系
prior: 查询父行的限定符，格式: prior column1 = column2 or column1 = prior column2 and ... ，
nocycle: 若数据表中存在循环行，那么不添加此关键字会报错，添加关键字后，便不会报错，但循环的两行只会显示其中的第一条
connect_by_iscycle: 前置条件:在使用了nocycle之后才能使用此关键字，用于表示是否是循环行，0表示否，1 表示是
connect_by_isleaf: 是否是叶子节点，0表示否，1 表示是
level: level伪列,表示层级，值越小层级越高，level=1为层级最高节点

如需要scott用户下建表及录入数据语句，可参考:
scott建表及录入数据sql脚本

这里，我们熟悉下emp的表结构
emp表中每个员工都有一个mgr(上级领导,最顶层的没有)
每个mgr其实也是emp表中的一员，通过父子ID关联，可以构造员工表的上下级关系

一.start with ...... connect by语句

显示员工表各个职员的级别
因为职位为PRESIDENT的mgr是空的，可以认为是最高的一级

有两种显示 上下级关系的办法，一种是通过 prior的顺序，一个是通过 parent key和child key 调换顺序。
--找到起始职位的下级以及下级的下级
connect by prior empno= mgr
connect by mgr = prior empno
--找到起始职位的上级及上级的上级
connect by prior mgr= empno
connect by empno = prior mgr

select empno, ename, job, mgr, deptno, level from emp start with job='PRESIDENT' connect by prior empno= mgr;
select empno, ename, job, mgr, deptno, level from emp start with job='PRESIDENT' connect by  mgr = prior empno;

SQL> select empno, ename, job, mgr, deptno, level from emp start with job='PRESIDENT' connect by prior empno= mgr;
EMPNO ENAME      JOB         MGR DEPTNO      LEVEL
----- ---------- --------- ----- ------ ----------
 7839 KING       PRESIDENT           10          1
 7566 JONES      MANAGER    7839     20          2
 7788 SCOTT      ANALYST    7566     20          3
 7876 ADAMS      CLERK      7788     20          4
 7902 FORD       ANALYST    7566     20          3
 7369 SMITH      CLERK      7902     20          4
 7698 BLAKE      MANAGER    7839     30          2
 7499 ALLEN      SALESMAN   7698     30          3
 7521 WARD       SALESMAN   7698     30          3
 7654 MARTIN     SALESMAN   7698     30          3
 7844 TURNER     SALESMAN   7698     30          3
 7900 JAMES      CLERK      7698     30          3
 7782 CLARK      MANAGER    7839     10          2
 7934 MILLER     CLERK      7782     10          3
14 rows selected

SQL> select empno, ename, job, mgr, deptno, level from emp start with job='PRESIDENT' connect by  mgr = prior empno;
EMPNO ENAME      JOB         MGR DEPTNO      LEVEL
----- ---------- --------- ----- ------ ----------
 7839 KING       PRESIDENT           10          1
 7566 JONES      MANAGER    7839     20          2
 7788 SCOTT      ANALYST    7566     20          3
 7876 ADAMS      CLERK      7788     20          4
 7902 FORD       ANALYST    7566     20          3
 7369 SMITH      CLERK      7902     20          4
 7698 BLAKE      MANAGER    7839     30          2
 7499 ALLEN      SALESMAN   7698     30          3
 7521 WARD       SALESMAN   7698     30          3
 7654 MARTIN     SALESMAN   7698     30          3
 7844 TURNER     SALESMAN   7698     30          3
 7900 JAMES      CLERK      7698     30          3
 7782 CLARK      MANAGER    7839     10          2
 7934 MILLER     CLERK      7782     10          3
14 rows selected

用lpad进行填充，看起来更有层次感

select empno, cast(lpad(' ', level*2-1,' ')||ename as varchar2(20)) ename, job, mgr, deptno, level 
from emp 
start with job='PRESIDENT' connect by prior empno= mgr;

SQL> select empno, cast(lpad(' ', level*2-1,' ')||ename as varchar2(20)) ename, job, mgr, deptno, level from emp start with job='PRESIDENT' connect by prior empno= mgr;
EMPNO ENAME                JOB         MGR DEPTNO      LEVEL
----- -------------------- --------- ----- ------ ----------
 7839  KING                PRESIDENT           10          1
 7566    JONES             MANAGER    7839     20          2
 7788      SCOTT           ANALYST    7566     20          3
 7876        ADAMS         CLERK      7788     20          4
 7902      FORD            ANALYST    7566     20          3
 7369        SMITH         CLERK      7902     20          4
 7698    BLAKE             MANAGER    7839     30          2
 7499      ALLEN           SALESMAN   7698     30          3
 7521      WARD            SALESMAN   7698     30          3
 7654      MARTIN          SALESMAN   7698     30          3
 7844      TURNER          SALESMAN   7698     30          3
 7900      JAMES           CLERK      7698     30          3
 7782    CLARK             MANAGER    7839     10          2
 7934      MILLER          CLERK      7782     10          3
14 rows selected

也可以从下往上查找
查看雇员scott及其上级的相关信息

select empno, cast(lpad(' ', level*2-1,' ')||ename as varchar2(20)) ename, job, mgr, deptno, level 
from emp 
start with empno=7788 connect by empno= prior mgr;

SQL> select empno, cast(lpad(' ', level*2-1,' ')||ename as varchar2(20)) ename, job, mgr, deptno, level from emp start with empno=7788 connect by empno= prior mgr;
EMPNO ENAME                JOB         MGR DEPTNO      LEVEL
----- -------------------- --------- ----- ------ ----------
 7788  SCOTT               ANALYST    7566     20          1
 7566    JONES             MANAGER    7839     20          2
 7839      KING            PRESIDENT           10          3

二.sys_connect_by_path函数

--语法：SYS_CONNECT_BY_PATH ( column , char )
--column和char可以是VARCHAR2、CHAR、NCHAR,、 NVARCHAR2中的一种
--返回值是VARCHAR2
--返回值的字符集是与column的字符集一样的
--只能用于层次查询中

从名为KING的雇员开始，查出所有下级雇员相关信息以及雇员名路径（用'/'连接）

select empno,
       cast(lpad(' ', level * 2 - 1, ' ') || ename as varchar2(20)) ename,
       job,
       mgr,
       deptno,
       level,
       cast(sys_connect_by_path(ename, '/') as varchar2(40)) namepath
  from emp
 start with ename = 'KING'
connect by prior empno = mgr;

SQL> select empno,
  2         cast(lpad(' ', level * 2 - 1, ' ') || ename as varchar2(20)) ename,
  3         job,
  4         mgr,
  5         deptno,
  6         level,
  7         cast(sys_connect_by_path(ename, '/') as varchar2(40)) namepath
  8    from emp
  9   start with ename = 'KING'
 10  connect by prior empno = mgr;
EMPNO ENAME                JOB         MGR DEPTNO      LEVEL NAMEPATH
----- -------------------- --------- ----- ------ ---------- ----------------------------------------
 7839  KING                PRESIDENT           10          1 /KING
 7566    JONES             MANAGER    7839     20          2 /KING/JONES
 7788      SCOTT           ANALYST    7566     20          3 /KING/JONES/SCOTT
 7876        ADAMS         CLERK      7788     20          4 /KING/JONES/SCOTT/ADAMS
 7902      FORD            ANALYST    7566     20          3 /KING/JONES/FORD
 7369        SMITH         CLERK      7902     20          4 /KING/JONES/FORD/SMITH
 7698    BLAKE             MANAGER    7839     30          2 /KING/BLAKE
 7499      ALLEN           SALESMAN   7698     30          3 /KING/BLAKE/ALLEN
 7521      WARD            SALESMAN   7698     30          3 /KING/BLAKE/WARD
 7654      MARTIN          SALESMAN   7698     30          3 /KING/BLAKE/MARTIN
 7844      TURNER          SALESMAN   7698     30          3 /KING/BLAKE/TURNER
 7900      JAMES           CLERK      7698     30          3 /KING/BLAKE/JAMES
 7782    CLARK             MANAGER    7839     10          2 /KING/CLARK
 7934      MILLER          CLERK      7782     10          3 /KING/CLARK/MILLER
14 rows selected

三.connect_by_root 语句

connect_by_root 查找根节点

select empno,
       cast(lpad(' ', level * 2 - 1, ' ') || ename as varchar2(20)) ename,
       job,
       mgr,
       deptno,
       level,
       cast(sys_connect_by_path(ename, '/') as varchar2(40)) namepath,
       connect_by_root(empno) rootno
  from emp
 start with ename in ('JONES','BLAKE','CLARK')
connect by prior empno = mgr;

SQL> select empno,
  2         cast(lpad(' ', level * 2 - 1, ' ') || ename as varchar2(20)) ename,
  3         job,
  4         mgr,
  5         deptno,
  6         level,
  7         cast(sys_connect_by_path(ename, '/') as varchar2(40)) namepath,
  8         connect_by_root(empno) rootno
  9    from emp
 10   start with ename in ('JONES','BLAKE','CLARK')
 11  connect by prior empno = mgr;
EMPNO ENAME                JOB         MGR DEPTNO      LEVEL NAMEPATH                                     ROOTNO
----- -------------------- --------- ----- ------ ---------- ---------------------------------------- ----------
 7566  JONES               MANAGER    7839     20          1 /JONES                                         7566
 7788    SCOTT             ANALYST    7566     20          2 /JONES/SCOTT                                   7566
 7876      ADAMS           CLERK      7788     20          3 /JONES/SCOTT/ADAMS                             7566
 7902    FORD              ANALYST    7566     20          2 /JONES/FORD                                    7566
 7369      SMITH           CLERK      7902     20          3 /JONES/FORD/SMITH                              7566
 7698  BLAKE               MANAGER    7839     30          1 /BLAKE                                         7698
 7499    ALLEN             SALESMAN   7698     30          2 /BLAKE/ALLEN                                   7698
 7521    WARD              SALESMAN   7698     30          2 /BLAKE/WARD                                    7698
 7654    MARTIN            SALESMAN   7698     30          2 /BLAKE/MARTIN                                  7698
 7844    TURNER            SALESMAN   7698     30          2 /BLAKE/TURNER                                  7698
 7900    JAMES             CLERK      7698     30          2 /BLAKE/JAMES                                   7698
 7782  CLARK               MANAGER    7839     10          1 /CLARK                                         7782
 7934    MILLER            CLERK      7782     10          2 /CLARK/MILLER                                  7782
13 rows selected

四.NOCYCLE、CONNECT_BY_ISCYCLE语句

测试数据:
备注:改动点，原表emp表职位为PRESIDENT的mgr是空，是最高级的领导
这个地方，我把他的mgr设置为SMITH的empno 7369

drop table emp2;

create table EMP2
(
  empno    NUMBER(4) not null,
  ename    VARCHAR2(10),
  job      VARCHAR2(9),
  mgr      NUMBER(4),
  hiredate DATE,
  sal      NUMBER(7,2),
  comm     NUMBER(7,2),
  deptno   NUMBER(2)
);

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7369', 'SMITH', 'CLERK', '7902', to_date('17-12-1980', 'dd-mm-yyyy'), '800', null, '20');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7499', 'ALLEN', 'SALESMAN', '7698', to_date('20-02-1981', 'dd-mm-yyyy'), '1600', '300', '30');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7521', 'WARD', 'SALESMAN', '7698', to_date('22-02-1981', 'dd-mm-yyyy'), '1250', '500', '30');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7566', 'JONES', 'MANAGER', '7839', to_date('02-04-1981', 'dd-mm-yyyy'), '2975', null, '20');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7654', 'MARTIN', 'SALESMAN', '7698', to_date('28-09-1981', 'dd-mm-yyyy'), '1250', '1400', '30');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7698', 'BLAKE', 'MANAGER', '7839', to_date('01-05-1981', 'dd-mm-yyyy'), '2850', null, '30');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7782', 'CLARK', 'MANAGER', '7839', to_date('09-06-1981', 'dd-mm-yyyy'), '2450', null, '10');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7788', 'SCOTT', 'ANALYST', '7566', to_date('13-06-0187', 'dd-mm-yyyy'), '3000', null, '20');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7839', 'KING', 'PRESIDENT', 7369, to_date('17-11-1981', 'dd-mm-yyyy'), '5000', null, '10');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7844', 'TURNER', 'SALESMAN', '7698', to_date('08-09-1981', 'dd-mm-yyyy'), '1500', '0', '30');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7876', 'ADAMS', 'CLERK', '7788', to_date('13-06-0187', 'dd-mm-yyyy'), '1100', null, '20');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7900', 'JAMES', 'CLERK', '7698', to_date('03-12-1981', 'dd-mm-yyyy'), '950', null, '30');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7902', 'FORD', 'ANALYST', '7566', to_date('03-12-1981', 'dd-mm-yyyy'), '3000', null, '20');

insert into emp2 (EMPNO, ENAME, JOB, MGR, HIREDATE, SAL, COMM, DEPTNO)
values ('7934', 'MILLER', 'CLERK', '7782', to_date('23-01-1982', 'dd-mm-yyyy'), '1300', null, '10');
commit;

运行最上面的sql，只是将包名从emp调整为emp2
此时报错，提示connect by循环了

SELECT empno, ename, job, mgr, deptno, LEVEL
  FROM emp2
 START WITH job = 'PRESIDENT'
CONNECT BY PRIOR empno = mgr;

SQL> SELECT empno, ename, job, mgr, deptno, LEVEL
  2    FROM emp2
  3   START WITH job = 'PRESIDENT'
  4  CONNECT BY PRIOR empno = mgr;
SELECT empno, ename, job, mgr, deptno, LEVEL
  FROM emp2
 START WITH job = 'PRESIDENT'
CONNECT BY PRIOR empno = mgr
ORA-01436: 用户数据中的 CONNECT BY 循环

加上 nocycle，让语句不循环，就可以正常运行了
CONNECT_BY_ISCYCLE 为1表示是循环的点

SELECT empno, ename, job, mgr, deptno, LEVEL,CONNECT_BY_ISCYCLE iscycle
  FROM emp2
 START WITH job = 'PRESIDENT'
CONNECT BY nocycle PRIOR empno = mgr;

SQL> SELECT empno, ename, job, mgr, deptno, LEVEL,CONNECT_BY_ISCYCLE iscycle
  2    FROM emp2
  3   START WITH job = 'PRESIDENT'
  4  CONNECT BY nocycle PRIOR empno = mgr;
EMPNO ENAME      JOB         MGR DEPTNO      LEVEL    ISCYCLE
----- ---------- --------- ----- ------ ---------- ----------
 7839 KING       PRESIDENT  7369     10          1          0
 7566 JONES      MANAGER    7839     20          2          0
 7788 SCOTT      ANALYST    7566     20          3          0
 7876 ADAMS      CLERK      7788     20          4          0
 7902 FORD       ANALYST    7566     20          3          0
 7369 SMITH      CLERK      7902     20          4          1
 7698 BLAKE      MANAGER    7839     30          2          0
 7499 ALLEN      SALESMAN   7698     30          3          0
 7521 WARD       SALESMAN   7698     30          3          0
 7654 MARTIN     SALESMAN   7698     30          3          0
 7844 TURNER     SALESMAN   7698     30          3          0
 7900 JAMES      CLERK      7698     30          3          0
 7782 CLARK      MANAGER    7839     10          2          0
 7934 MILLER     CLERK      7782     10          3          0
14 rows selected

五.connect_by_isleaf语句

connect_by_isleaf 是否叶子节点,1为是，0为否
叶子节点可以理解为最后一层

select empno,
       cast(lpad(' ', level * 2 - 1, ' ') || ename as varchar2(20)) ename,
       job,
       mgr,
       deptno,
       level,
       cast(sys_connect_by_path(ename, '/') as varchar2(40)) namepath,
       CONNECT_BY_ISLEAF ISLEAF
  from emp
 start with ename = 'KING'
connect by prior empno = mgr;

SQL> select empno,
  2         cast(lpad(' ', level * 2 - 1, ' ') || ename as varchar2(20)) ename,
  3         job,
  4         mgr,
  5         deptno,
  6         level,
  7         cast(sys_connect_by_path(ename, '/') as varchar2(40)) namepath,
  8         CONNECT_BY_ISLEAF ISLEAF
  9    from emp
 10   start with ename = 'KING'
 11  connect by prior empno = mgr;
EMPNO ENAME                JOB         MGR DEPTNO      LEVEL NAMEPATH                                     ISLEAF
----- -------------------- --------- ----- ------ ---------- ---------------------------------------- ----------
 7839  KING                PRESIDENT           10          1 /KING                                             0
 7566    JONES             MANAGER    7839     20          2 /KING/JONES                                       0
 7788      SCOTT           ANALYST    7566     20          3 /KING/JONES/SCOTT                                 0
 7876        ADAMS         CLERK      7788     20          4 /KING/JONES/SCOTT/ADAMS                           1
 7902      FORD            ANALYST    7566     20          3 /KING/JONES/FORD                                  0
 7369        SMITH         CLERK      7902     20          4 /KING/JONES/FORD/SMITH                            1
 7698    BLAKE             MANAGER    7839     30          2 /KING/BLAKE                                       0
 7499      ALLEN           SALESMAN   7698     30          3 /KING/BLAKE/ALLEN                                 1
 7521      WARD            SALESMAN   7698     30          3 /KING/BLAKE/WARD                                  1
 7654      MARTIN          SALESMAN   7698     30          3 /KING/BLAKE/MARTIN                                1
 7844      TURNER          SALESMAN   7698     30          3 /KING/BLAKE/TURNER                                1
 7900      JAMES           CLERK      7698     30          3 /KING/BLAKE/JAMES                                 1
 7782    CLARK             MANAGER    7839     10          2 /KING/CLARK                                       0
 7934      MILLER          CLERK      7782     10          3 /KING/CLARK/MILLER                                1
14 rows selected

六.connect by构造数据

--connect by构造等差数列
select rownum rn from dual connect by rownum<5;
--调整起始值
select rownum+15 rn from dual connect by rownum<5;
--调整间隔
select 3*rownum-9 rn from dual connect by rownum<5;
--等比数列
select power(2,rownum) rn from dual connect by rownum<5;

SQL> --connect by构造等差数列
SQL> select rownum rn from dual connect by rownum<5;
        RN
----------
         1
         2
         3
         4
SQL> --调整起始值
SQL> select rownum+15 rn from dual connect by rownum<5;
        RN
----------
        16
        17
        18
        19
SQL> --调整间隔
SQL> select 3*rownum-9 rn from dual connect by rownum<5;
        RN
----------
        -6
        -3
         0
         3
SQL> --等比数列
SQL> select power(2,rownum) rn from dual connect by rownum<5;
        RN
----------
         2
         4
         8
        16

--用connect by构造 26个英文字母
with t as (select ascii('A')+rownum-1 aa from dual connect by rownum<=26),
 u as (select chr(aa) from t)
select * from u;

SQL> --用connect by构造 26个英文字母
SQL> with t as (select ascii('A')+rownum-1 aa from dual connect by rownum<=26),
  2   u as (select chr(aa) from t)
  3  select * from u;
CHR(AA)
-------
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
W
X
Y
Z
26 rows selected

百钱买鸡兔
小母鸡4块钱5只
老母鸡3块钱1只
大白兔2块钱1只
小白兔3块钱4只

花费的钱总数是100
小母鸡、老母鸡、大白兔、小白兔的总数也刚好是100

with tmp1 as
 (select rownum n from dual connect by rownum <= 100-3)
select t1.n xmj, t2.n lmj, t3.n dbt, t4.n xbt
    from tmp1 t1, tmp1 t2, tmp1 t3, tmp1 t4
    where 1 = 1
    and   t1.n*4/5 + t2.n*3/1 + t3.n*2/1 + t4.n*3/4 = 100
    and   t1.n + t2.n + t3.n + t4.n = 100;

SQL> with tmp1 as
  2   (select rownum n from dual connect by rownum <= 100-3)
  3  select t1.n xmj, t2.n lmj, t3.n dbt, t4.n xbt
  4      from tmp1 t1, tmp1 t2, tmp1 t3, tmp1 t4
  5      where 1 = 1
  6      and   t1.n*4/5 + t2.n*3/1 + t3.n*2/1 + t4.n*3/4 = 100
  7      and   t1.n + t2.n + t3.n + t4.n = 100;
  
       XMJ        LMJ        DBT        XBT
---------- ---------- ---------- ----------
         5          1         18         76
         5          6          9         80
        10          2         16         72
        10          7          7         76
        15          3         14         68
        15          8          5         72
        20          4         12         64
        20          9          3         68
        25          5         10         60
        25         10          1         64
        30          1         17         52
        30          6          8         56
        35          2         15         48
        35          7          6         52
        40          3         13         44
        40          8          4         48
        45          4         11         40
        45          9          2         44
        50          5          9         36
        55          1         16         28
        55          6          7         32
        60          2         14         24
        60          7          5         28
        65          3         12         20
        65          8          3         24
        70          4         10         16
        70          9          1         20
        75          5          8         12
        80          1         15          4
        80          6          6          8
        85          7          4          4
31 rows selected

作者：只是甲
链接：https://www.jianshu.com/p/5225a48b4754
来源：简书
著作权归作者所有。商业转载请联系作者获得授权，非商业转载请注明出处。