Oracle 递归

当对象存在父节点、子节点时，通过特定的方式获取父节点、子节点数据构建树状结构或其它形式结构时，通常都会使用递归，如:一个公司有多个部门、每个部门下可能有多个小部门，小部门下面又有组….为了数据容易管理和维护,通过构建合适的表结构存储这些数据,以下示例以省市县为例学习了解递归:
1.创建存储省市县数据表:

   1:  create table tb_distree

   2:  (

   3:  id number,

   4:  name varchar2(300),

   5:  pid number

   6:  )

   7:  /

   8:   remark 添加主外键

   9:  alter table tb_distree add (

  10:  constraints pk_id primary key(id),

  11:  constraints fk_pid foreign key(pid) references tb_distree(id)

  12:  )

  13:  /

.csharpcode, .csharpcode pre
{
font-size: small;
color: black;
font-family: consolas, "Courier New", courier, monospace;
background-color: #ffffff;
/*white-space: pre;*/
}
.csharpcode pre { margin: 0em; }
.csharpcode .rem { color: #008000; }
.csharpcode .kwrd { color: #0000ff; }
.csharpcode .str { color: #006080; }
.csharpcode .op { color: #0000c0; }
.csharpcode .preproc { color: #cc6633; }
.csharpcode .asp { background-color: #ffff00; }
.csharpcode .html { color: #800000; }
.csharpcode .attr { color: #ff0000; }
.csharpcode .alt
{
background-color: #f4f4f4;
width: 100%;
margin: 0em;
}
.csharpcode .lnum { color: #606060; }

2.初始化数据:

   1:  insert into tb_distree(id,name) values(1,'云南省');

   2:  insert into tb_distree(id,name,pid) values(2,'昆明市',1);

   3:  insert into tb_distree(id,name,pid) values(3,'临沧市',1);

   4:  insert into tb_distree(id,name,pid) values(4,'丽江市',1);

   5:  insert into tb_distree(id,name,pid) values(5,'云县',3);

   6:  insert into tb_distree(id,name,pid) values(6,'凤庆',3);

   7:  insert into tb_distree(id,name,pid) values(7,'幸福',3);

   8:  insert into tb_distree(id,name,pid) values(8,'盘龙区',2);

   9:  insert into tb_distree(id,name,pid) values(9,'五华区',2);

  10:  insert into tb_distree(id,name,pid) values(10,'西山区',2);

3.递归查询语法:

   1:  select column... from table_name

   2:  where ....  过滤条件

   3:  start with ...  递归开始点

   4:  connect by prior .... 优先级

4.递归查询数据:

4.1 从父节点开始查询出所有父节点和子节点:

SQL> select id,name,pid from tb_distree start with pid is null connect by prior id=pid;

        ID NAME                PID

---------- ------------ ----------

         1 云南省

         2 昆明市                1

         8 盘龙区                2

         9 五华区                2

        10 西山区                2

         3 临沧市                1

         5 云县                  3

         6 凤庆                  3

         7 幸福                  3

         4 丽江市                1

10 rows selected.

4.2 查询某个节点的父节点:

 SQL> select id,name,pid from tb_distree start with name='云县' connect by prior pid=id;

        ID NAME                PID

---------- ------------ ----------

         5 云县                  3

         3 临沧市                1

         1 云南省

在上例中"云县"属于"临沧市","临沧市"属于"云南省";对于从父节点递归到子节点,优先级条件为子节点id等于父节点id;对于从子节点到父节点递归,方向刚好相反;

4.3 通过层次查询出父节点和某个子节点:

SQL> select id,name,pid,level from tb_distree where level in(1,2) start with pid is null connect by prior id=pid;

        ID NAME                PID      LEVEL

---------- ------------ ---------- ----------

         1 云南省                           1

         2 昆明市                1          2

         3 临沧市                1          2

         4 丽江市                1          2

层次也很重要,某些时候要修改某个节点父节点或子节点时会很有用;

5. 递归查询效率:

SQL>select/*+ selectDG1 */ id,name,pid from tb_distree start with pid is null connect by prior id=pid;

SQL> select sql_id,sql_text from v$sql where sql_text like '%selectDG1%' ;

SQL_ID          SQL_TEXT

--------------- --------------------------------------------------

2wnu324ga4n0y   select sql_id,sql_text from v$sql where sql_text l

                ike '%selectDG1%'

d4g89bucsbvzd   select/*+ selectDG1 */ id,name,pid from tb_distree

                 start with pid is null connect by prior id=pid

SQL>select * from table(dbms_xplan.display_cursor('d4g89bucsbvzd',null,'advanced allstats last peeked_binds'));

PLAN_TABLE_OUTPUT

------------------------------------------------------------------------------------------------------------

SQL_ID  d4g89bucsbvzd, child number 0

-------------------------------------

select/*+ selectDG1 */ id,name,pid from tb_distree start with pid is

null connect by prior id=pid

Plan hash value: 1466399788

------------------------------------------------------------------------------------------------------------------------------------------------

| Id  | Operation                               | Name       | Starts | E-Rows |E-Bytes| Cost (%CPU)| E-Time   | A-Rows |   A-Time   | Buffers |

------------------------------------------------------------------------------------------------------------------------------------------------

|   0 | SELECT STATEMENT                        |            |      1 |        |       |    18 (100)|          |     10 |00:00:00.01 |      22 |

PLAN_TABLE_OUTPUT

----------------------------------------------------------------------------------------------------------------------------

|*  1 |  CONNECT BY NO FILTERING WITH START-WITH|            |      1 |        |       |            |          |     10 |00:00:00.01 |      22 |

|   2 |   TABLE ACCESS FULL                     | TB_DISTREE |      1 |     10 |  1780 |    17   (0)| 00:00:01 |     10 |00:00:00.01 |      22 |

------------------------------------------------------------------------------------------------------------------------------------------------

Query Block Name / Object Alias (identified by operation id):

-------------------------------------------------------------

   1 - SEL$1

   2 - SEL$2 / TB_DISTREE@SEL$2

Outline Data

PLAN_TABLE_OUTPUT

--------------------------------------------------------------------------------------------------------------------------------------------------

/*+

      BEGIN_OUTLINE_DATA

      IGNORE_OPTIM_EMBEDDED_HINTS

      OPTIMIZER_FEATURES_ENABLE('11.2.0.3')

      DB_VERSION('11.2.0.3')

      ALL_ROWS

      OUTLINE_LEAF(@"SEL$2")

      OUTLINE_LEAF(@"SEL$3")

      OUTLINE_LEAF(@"SEL$4")

PLAN_TABLE_OUTPUT

---------------------------------------------------------------------------------------------------------------------------------------------------

      OUTLINE_LEAF(@"SET$1")

      OUTLINE_LEAF(@"SEL$1")

      NO_ACCESS(@"SEL$1" "connect$_by$_work$_set$_006"@"SEL$1")

      NO_CONNECT_BY_FILTERING(@"SEL$1")

      CONNECT_BY_COMBINE_SW(@"SEL$1")

      FULL(@"SEL$4" "TB_DISTREE"@"SEL$4")

      FULL(@"SEL$3" "connect$_by$_pump$_002"@"SEL$3")

      FULL(@"SEL$3" "TB_DISTREE"@"SEL$3")

      LEADING(@"SEL$3" "connect$_by$_pump$_002"@"SEL$3" "TB_DISTREE"@"SEL$3")

      USE_HASH(@"SEL$3" "TB_DISTREE"@"SEL$3")

      FULL(@"SEL$2" "TB_DISTREE"@"SEL$2")

PLAN_TABLE_OUTPUT

----------------------------------------------------------------------------------------------------------------------------------------------

      END_OUTLINE_DATA

*/

Predicate Information (identified by operation id):

---------------------------------------------------

   1 - access("PID"=PRIOR NULL)

       filter("PID" IS NULL)

 10046 trace:

Rows (1st) Rows (avg) Rows (max)  Row Source Operation

 675 ---------- ---------- ----------  ---------------------------------------------------

 676         10         10         10  CONNECT BY NO FILTERING WITH START-WITH (cr=22 pr=0 pw=0 time=269 us)

 677         10         10         10   TABLE ACCESS FULL TB_DISTREE (cr=22 pr=0 pw=0 time=118 us cost=17 size=1780 card=10)

 680 Elapsed times include waiting on following events:

 681   Event waited on                             Times   Max. Wait  Total Waited

 682   ----------------------------------------   Waited  ----------  ------------

 683   row cache lock                                  3        0.00          0.00

 684   Disk file operations I/O                        1        0.00          0.00

 685   db file sequential read                         3        0.03          0.04

 686   SQL*Net message to client                       2        0.00          0.00

 687   SQL*Net message from client                     2        0.00          0.00

 688 ********************************************************************************

在递归表上创建索引是没有用的,都必须进行全表扫描,当然通常这样的表也不会很大,如果数据量的确很大,建议将表中的节点拆分成多个表提高效率;

Oracle 递归的更多相关文章

Oracle 递归的写法（start with) 以及where条件作用域
先转一个讲Oracle递归讲得非常透彻的文章: http://blog.csdn.net/weiwenhp/article/details/8218091 前言:嗯,这也是一个前人挖坑,后人来填的故事 ...
oracle 递归应用（挺复杂的）
最近做数据过滤觉得很有必要记录下整个过程,说不定下次就不知道了. 废话不多说开始: 表结构: 企业表(自关联,采用树的形式记录分子公司) 区域表(自关联,采用树的形式记录省/市/县/乡,数据量大) 公 ...
oracle 递归和connect by【转】
oracle递归查询(单表包含多级上下级关系) http://www.cnblogs.com/walk-the-Line/p/4882866.html -- 查找所有第一层(包括)下的所有子节点 -- ...
oracle递归层级查询 start with connect by prior
递归层级查询:start with connect by prior 以部门表作为解析表结构:dept{id:'主键',name:'部门名称',parent_id:'父亲id'} select * ...
Oracle递归sql笔记
查询一个机构下所辖机构: select * from t00_organ t start with t.organkey=#uporgankey# connect by prior t.organke ...
Oracle递归 start with...connect by...prior
prior一侧是父节点另一侧是子节点 --查询region_id等于4519的节点下面的所有子节点查找出给定节点的所有子节点 SELECT sr.* FROM spc_region sr wher ...
Oracle递归操作
需求:找出代理商中没有挂商家的代理商简单SQL如下: select * from t_proxy tp where tp.id not in (SELECT tp.id as p_id FROM t ...
[oracle] 递归追溯完整部门名称函数
create or replace function fn_DeptWholeName2(objectid in number) return nvarchar2 is wholename nvarc ...
oracle 分页(rownum的理解) 以及树节点的查询
1:什么是rownum, rownum的生成, rownum相关的符号操作 Rownum是oracle生成结果集时得到的一个伪列, 按照读出行的顺序, 第一条rownum=1, 第二条=2. 对于 O ...

随机推荐

codechef January Lunchtime 2017简要题解
题目地址https://www.codechef.com/LTIME44 Nothing in Common 签到题,随便写个求暴力交集就行了 Sealing up 完全背包算出得到长度≥x的最小花费 ...
(转) docker跨主机 macvlan 网络配置
原文链接 https://github.com/alfredhuang211/study-docker-doc/blob/master/docker%E8%B7%A8%E4%B8%BB%E6%9C%B ...
解决： Project facet Java version 1.8 is not supported
背景从别处Import一个Java project之后,Eclipse提示“Project facet Java version 1.8 is not supported”. 分析从错误的描述来看 ...
python学习 (三十二) 异常处理
1 异常: def exceptionHandling(): try: a = b = d = a / b print(d) except ZeroDivisionError as ex: print ...
Mysql--产品支持的平台
【CVE】CVE-2018-4304 Apple多个操作系统函数拒绝服务漏洞
TextImpact: Processing a maliciously crafted text file may lead to adenial of serviceDescription: A ...
bat执行sqlplus语句，省去@xx.sql过程
bat文件中执行写sqlplus连接,再@调用自己,sql登录成功后,会忽略掉第一行sqlplus xxx,转而执行下方的sql语句 --------------------------------- ...
linux网络监视器
vnstat iftop nethogs dig ipcalc
【airflow实战系列】基于 python 的调度和监控工作流的平台
简介 airflow 是一个使用python语言编写的data pipeline调度和监控工作流的平台.Airflow被Airbnb内部用来创建.监控和调整数据管道.任何工作流都可以在这个使用Pyth ...
Linux监控和安全运维 1.9 zabbix增加客户端监控
1.在客户端安装 zabbix20与服务器相对应. yum install zabbix20-agent 2.修改配置文件 vim /etc/zabbix_agentd.conf Server=127 ...

Oracle 递归

Oracle 递归的更多相关文章

随机推荐

热门专题