oracle展bom逻辑,oracle 求BOM树型展开的总用量
一个BOM展开求总使用量问题求助:
当然也请各路高手拔刀相助
1.建表:
表1 结构表
create table
MANUF_STRUCTURE_TAB
(PART_NO VARCHAR2(25) not null,
--父
COMPONENT_PART
VARCHAR2(25) not null, --子
QTY_PER_ASSEMBLY
NUMBER not null)
--子对父的构成数(即1个父由多少个子构成)
表2 分类表
create table
INVENTORY_PART_TAB
(PART_NO VARCHAR2(25) not null, --编码
SK_PART_CATEGORY_DB VARCHAR2(15) not null) --类别
2.说明:
表2中SK_PART_CATEGORY_DB分3类:
Product:成品
SemiProdudct:半成品
Component:原材料
3.想要结果:
1个成品所需要原材料的总使用量
PART_NO COMPONENT_PART T_QTY
......
测试数据及想要结果(想要结果未能列完)
想用sys_connect_by_path
把QTY_PER_ASSEMBLY用乘號串起來,
然後調用一個能執行動態SQL的function以得到結果.
简化一下你的数据：
create table
MANUF_STRUCTURE_TAB
(PART_NO VARCHAR2(25) not
null,
COMPONENT_PART
VARCHAR2(25) not
null,
QTY_PER_ASSEMBLY
NUMBER not
null
);
create table
INVENTORY_PART_TAB
(PART_NO VARCHAR2(25) not null,
SK_PART_CATEGORY_DB VARCHAR2(15) not null
);
INSERT INTO
MANUF_STRUCTURE_TAB VALUES ('A1','B',2);
INSERT INTO
MANUF_STRUCTURE_TAB VALUES ('A2','B',3);
INSERT INTO
MANUF_STRUCTURE_TAB VALUES ('B','C',4);
INSERT INTO
MANUF_STRUCTURE_TAB VALUES ('C','D',7);
INSERT INTO
MANUF_STRUCTURE_TAB VALUES ('A1','C',6);
INSERT INTO
INVENTORY_PART_TAB VALUES ('A1','Product');
INSERT INTO
INVENTORY_PART_TAB VALUES ('A2','Product');
INSERT INTO
INVENTORY_PART_TAB VALUES ('B','SemiProdudct');
INSERT INTO
INVENTORY_PART_TAB VALUES ('C','Component');
INSERT INTO
INVENTORY_PART_TAB VALUES ('D','Component');
三种方法(既然你不喜欢自定义函数就没写了)：
WITH t AS
(
SELECT
SYS_CONNECT_BY_PATH(a.part_no,'\')||'\'||a.component_part||'\' as
path
,qty_per_assembly
FROM
manuf_structure_tab a JOIN inventory_part_tab b
ON
a.part_no = b.part_no
START WITH
b.sk_part_category_db = 'Product'
CONNECT BY a.part_no = PRIOR
a.component_part
)
SELECT
root_part,part,SUM(qty) AS qty
FROM
(
SELECT
SUBSTR(t1.path,2,INSTR(t1.path,'\',1,2)-INSTR(t1.path,'\',1,1)-1)
as ROOT_PART
,SUBSTR(t1.path,INSTR(t1.path,'\',-1,2)+1,INSTR(t1.path,'\',-1,1)-INSTR(t1.path,'\',-1,2)-1)
as PART
,EXP(SUM(LN(t2.qty_per_assembly))) AS qty
FROM t
t1,t t2
WHERE t1.path LIKE
t2.path||'%'
GROUP BY
t1.path
)
GROUP BY
root_part,part;
WITH t AS
(
SELECT
SYS_CONNECT_BY_PATH(a.part_no,'\')||'\'||a.component_part||'\' as
path
,qty_per_assembly
FROM
manuf_structure_tab a JOIN inventory_part_tab b
ON
a.part_no = b.part_no
START WITH
b.sk_part_category_db = 'Product'
CONNECT BY a.part_no = PRIOR
a.component_part
)
SELECT
root_part,part,SUM(qty) AS qty
FROM
(
SELECT
SUBSTR(t1.path,2,INSTR(t1.path,'\',1,2)-INSTR(t1.path,'\',1,1)-1)
as ROOT_PART
,SUBSTR(t1.path,INSTR(t1.path,'\',-1,2)+1,INSTR(t1.path,'\',-1,1)-INSTR(t1.path,'\',-1,2)-1)
as PART
,(SELECT
EXP(SUM(LN(t2.qty_per_assembly))) FROM t t2 WHERE t1.path LIKE
t2.path||'%') AS qty
FROM t
t1
GROUP BY
t1.path
)
GROUP BY
root_part,part;
WITH t AS
(
SELECT
SYS_CONNECT_BY_PATH(a.part_no,'\')||'\'||a.component_part||'\' as
path
,SYS_CONNECT_BY_PATH(a.qty_per_assembly,'\')||'\' as
qty_path
,qty_per_assembly
FROM
manuf_structure_tab a JOIN inventory_part_tab b
ON
a.part_no = b.part_no
START WITH
b.sk_part_category_db = 'Product'
CONNECT BY a.part_no = PRIOR
a.component_part
)
SELECT
root_part,part,SUM(qty) AS qty
FROM
(
SELECT
SUBSTR(path,2,INSTR(path,'\',1,2)-INSTR(path,'\',1,1)-1) as
ROOT_PART
,SUBSTR(path,INSTR(path,'\',-1,2)+1,INSTR(path,'\',-1,1)-INSTR(path,'\',-1,2)-1)
as PART
,(SELECT
EXP(SUM(LN(SUBSTR(qty_path,INSTR(qty_path,'\',1,rn)+1,INSTR(qty_path,'\',1,rn+1)-INSTR(qty_path,'\',1,rn)-1))))
FROM (SELECT ROWNUM RN FROM DUAL CONNECT BY
ROWNUM<=100)
WHERE
SUBSTR(qty_path,INSTR(qty_path,'\',1,rn)+1,INSTR(qty_path,'\',1,rn+1)-INSTR(qty_path,'\',1,rn)-1)
IS NOT NULL
) AS
qty
FROM
t
)
GROUP BY
root_part,part;
如果最后只需要原材料，你自己过滤一下。
等会有空我用10G的MODEL来写一个。
用11GR2的递归WITH子查询应该是易如反掌，但我现在没环境测试。
---------------
发现有大量重复的，这是怎么回事？父、子应该构成唯一键。如果不唯一要先DISTINCT.
经测试第三种效果最好，和10G的MODEL差不多。
上面的方法都不新鲜了，这个10G的MODEL方法好像没见过：
WITH t AS
(
SELECT
SYS_CONNECT_BY_PATH(a.part_no,'\')||'\'||a.component_part as
path
,qty_per_assembly
,LEVEL
lvl
FROM
manuf_structure_tab a JOIN inventory_part_tab b
ON
a.part_no = b.part_no
START WITH
b.sk_part_category_db = 'Product'
CONNECT BY a.part_no = PRIOR
a.component_part
)
SELECT
root_part,part,SUM(qty) AS qty
FROM
(
SELECT
SUBSTR(path,2,INSTR(path,'\',1,2)-INSTR(path,'\',1,1)-1) as
ROOT_PART
,SUBSTR(path,INSTR(path,'\',-1,1)+1) as PART
,qty
FROM
t
MODEL IGNORE NAV RETURN
UPDATED ROWS
DIMENSION BY (path)
MEASURES (qty_per_assembly, 1
qty,lvl)
RULES AUTOMATIC ORDER
(
qty[any] order by path=(CASE
WHEN lvl[cv()]=1 THEN
qty_per_assembly[cv()]
ELSE qty_per_assembly[cv()]
*
qty[SUBSTR(cv(),1,INSTR(cv(),'\',-1,1)-1)]
END)
)
)
GROUP BY
root_part,part;
我的manuf_structure_tab有15万记录,INVENORY_PART_TAB有8W.
目前执行了5MIN还没反应.
WITH t AS
(
SELECT
SYS_CONNECT_BY_PATH(a.part_no,'\')||'\'||a.component_part||'\' as
path
,qty_per_assembly
FROM
manuf_structure_tab a,inventory_part_tab
b
where
a.part_no = b.part_no
and
a.contract=b.contract
and
b.contract='SKY'
START WITH
b.sk_part_category_db = 'Product'
CONNECT BY a.part_no = PRIOR
a.component_part
)
SELECT
root_part,part,SUM(qty) AS qty
FROM
(
SELECT
SUBSTR(t1.path,2,INSTR(t1.path,'\',1,2)-INSTR(t1.path,'\',1,1)-1)
as ROOT_PART
,SUBSTR(t1.path,INSTR(t1.path,'\',-1,2)+1,INSTR(t1.path,'\',-1,1)-INSTR(t1.path,'\',-1,2)-1)
as PART
,(SELECT
EXP(SUM(LN(t2.qty_per_assembly))) FROM t t2 WHERE t1.path LIKE
t2.path||'%') AS qty
FROM t
t1
GROUP BY
t1.path
)
GROUP BY
root_part,part
这样呢？
WITH t AS
(
SELECT
SYS_CONNECT_BY_PATH(a.part_no,'\')||'\' as path
,component_part as
part
,LEVEL
lvl
,SYS_CONNECT_BY_PATH(a.qty_per_assembly,'\')||'\' as
qty_path
FROM
manuf_structure_tab a JOIN inventory_part_tab b
ON
a.part_no = b.part_no
START WITH
b.sk_part_category_db = 'Product'
CONNECT BY a.part_no = PRIOR
a.component_part
)
SELECT
root_part,part,SUM(qty) AS qty
FROM
(
SELECT
SUBSTR(path,2,INSTR(path,'\',1,2)-2) as
ROOT_PART
,PART
,(SELECT
EXP(SUM(LN(SUBSTR(qty_path,INSTR(qty_path,'\',1,rn)+1,INSTR(qty_path,'\',1,rn+1)-INSTR(qty_path,'\',1,rn)-1))))
FROM (SELECT ROWNUM RN FROM DUAL CONNECT BY
ROWNUM<=100)
WHERE rn<=lvl
) AS
qty
FROM
t
)
GROUP BY
root_part,part;
---------------------------------------------------------------------------------------------------------------
| Id |
Operation | Name | Rows | Bytes |TempSpc| Cost
(%CPU)| Time |
---------------------------------------------------------------------------------------------------------------
| 0 |
SELECT STATEMENT |
| 804 |
3163K| | 1216 (1)|
00:00:15 |
| 1 |
SORT AGGREGATE | | 1 | 13 |
|
|
|
|* 2 |
VIEW | |
1 | 13 | | 2
(0)| 00:00:01 |
| 3 |
COUNT | |
|
|
|
|
|
| 4 |
CONNECT BY WITHOUT FILTERING|
| | | | | |
| 5 |
FAST DUAL
|
| 1 |
|
|
2 (0)|
00:00:01 |
| 6 |
HASH GROUP BY | |
804 | 3163K|
13M| 1216
(1)| 00:00:15 |
| 7 |
VIEW | |
2587 | 9M|
|
9 (12)|
00:00:01 |
|* 8 |
CONNECT BY WITH FILTERING
| |
|
|
|
|
|
|* 9 |
FILTER | | | | | | |
| 10 |
COUNT
| |
|
|
|
|
|
|* 11 | HASH JOIN | | 2587 |
126K| | 9
(12)| 00:00:01 |
| 12 |
TABLE ACCESS FULL |
INVENTORY_PART_TAB | 887 |
20401 | |
3 (0)|
00:00:01 |
| 13 |
TABLE ACCESS FULL |
MANUF_STRUCTURE_TAB | 2587 | 69849 |
|
5 (0)|
00:00:01 |
|* 14 | HASH JOIN | | | | | | |
| 15 |
CONNECT BY
PUMP |
| | | | | |
| 16 |
COUNT
| |
|
|
|
|
|
|* 17 | HASH JOIN | | 2587 |
126K| | 9
(12)| 00:00:01 |
| 18 |
TABLE ACCESS FULL |
INVENTORY_PART_TAB | 887 |
20401 | |
3 (0)|
00:00:01 |
| 19 |
TABLE ACCESS FULL |
MANUF_STRUCTURE_TAB | 2587 | 69849 |
|
5 (0)|
00:00:01 |
| 20 |
COUNT | | | | | | |
|* 21 | HASH JOIN | |
2587 | 126K|
|
9 (12)|
00:00:01 |
| 22 |
TABLE
ACCESS FULL | INVENTORY_PART_TAB
| 887 | 20401 |
|
3 (0)|
00:00:01 |
| 23 |
TABLE
ACCESS FULL | MANUF_STRUCTURE_TAB |
2587 | 69849 | | 5 (0)| 00:00:01
|
---------------------------------------------------------------------------------------------------------------
从你的数据量来看，遍历深度应该很浅啊？如果单独挑几种产品(在START
WITH中指定)，速度怎么样？结果对吗？
单独把WITH里面的查询CREATE TABLE
AS...要花多少时间？生成的结果多大？这一步完成后尽量过滤(比如只保留你要的原料)，让后面的乘法计算更少。
SELECT xmlquery('1*2*3*4'
returning content).getnumberval() FROM DUAL;
但是10G不支持变量，只能常量：
SELECT xmlquery(str returning
content).getnumberval()
FROM (SELECT '1*2*3*4' as str
FROM DUAL);
ERROR at line
1:
ORA-19102: XQuery string
literal expected
----
11GR2:
WITH t(root_part,part,qty) AS
(
SELECT a.part_no as
root_part, component_part as part,qty_per_assembly as
qty
FROM
manuf_structure_tab a JOIN inventory_part_tab b
ON
a.part_no = b.part_no
WHERE b.sk_part_category_db =
'Product'
UNION ALL
SELECT
a.root_part,b.component_part,a.qty*b.qty_per_assembly
FROM t a,
manuf_structure_tab b
WHERE a.part =
b.part_no
)
SELECT
root_part,part,SUM(qty) AS qty
FROM
t
GROUP BY
root_part,part;
ROOT_PART PART QTY
-------------------------
------------------------- ----------
A2 C 12
A1 D 98
A1 B 2
A2 B 3
A1 C 14
A2 D 84
oracle展bom逻辑,oracle 求BOM树型展开的总用量相关推荐
- 计算机的拓扑 树状结构图,树型网络拓扑结构
树型网络拓扑结构[编辑] 树型网络拓扑结构简介:树型网络拓扑结构是网络节点呈树状排列,整体看来就象一棵朝上的树.树型网络拓扑结构是总线型结构的扩展,它是在总线网上加上分支形成的,其传输介质可有多条分支 ...
- oracle planning bom,OracleERP表结构--BOM模块
BOM模块常用表结构 表名: bom.bom_bill_of_materials 说明: BOM清单父项目 BILL_SEQUENCE_ID NUMBER 清单序号(关键字) ASSEMBLY_ITE ...
- oracle的逻辑结构包括,oracle逻辑结构分析
oracle数据库在逻辑上是由多个表间组成的,表空间中存储的对象叫段,比如数据段,索引段,和回退段.段由区组成,区是磁盘分配的最小单位.段的增大是通过增加区的个数来实现的.每个区的大小是数据块大小的整 ...
- oracle中更新一列分组的均值,oracle 分组平均后又求平均值的方法
oracle 分组平均后再求平均值的方法 请教大家一个问题,是分组平均后再平均的方法 有一张表,记录物品的价格 名称 销售日期 价格 A 1日 2 A 2日 ...
- ORACLE表有逻辑坏块时EXPDP导出报错排查
某用户反馈EXPDP导出脚本异常中断,接入排查后发现导出某表时报错,后台ALERT日志中也有相应报错:最终确认排查为逻辑坏块问题,经沟通得知为某个时间段的备份表,后续排除此表后重新设置EXPDP脚本, ...
- Oracle 物理读 逻辑读 一致性读 当前模式读总结浅析
Oracle 物理读 逻辑读 一致性读 当前模式读总结浅析 在ORACLE数据库中有物理读(Physical Reads).逻辑读(Logical Reads).一致性读(Consistant Get ...
- mysql start with connect by_MYSQL实现Oracle的Start with…Connect By递归树查询
最近接触一个老项目,最初适配的数据库是oracle 后来迁移到mysql .这个迁移是由一个已经离职的新手做的.这个新手对于很多oracle特有的函数及存储过程都没有进行迁移.导致系统很多功能都BUG ...
- ORACLE 树型层次结构查询
为什么80%的码农都做不了架构师?>>> ORACLE中的SELECT语句可以用 START WITH...CONNECT BY PRIOR 子句实现递归查询,CONNECT ...
- 数据导不进oracle数据库,学习笔记:Oracle逻辑导出/导入 数据逻辑导出时跳过指定表不进行导出...
天萃荷净 Oracle数据库逻辑exp导出时,跳过指定某些表,对其表中数据不进行导出 有一个需求,某个用户有很多张表,但是只能使用exp导出,而且想跳过其中某几张表,其他对象包括依赖关系都需要.针对这 ...
最新文章
- 17 个品牌，113 款 5G 手机，5G 离我们越来越近
- c语言构造满二叉树,递归创建二叉树c语言实现+详细解释
- Linux下 运行Jmeter （含一个jmeter简单示例）
- 2020蓝桥杯省赛---java---A---2(既分数组)
- ViceDinoSpecCtrlDlg.cpp
- [转]notepad++各种插件
- eplan步进电机_干货丨三张接线图搞定步进电机接线问题！
- 【预训练模型】预训练语言模型的前世今生之风起云涌
- 数据结构常见算法机试题
- pytdx 获取板块指数_是否可以获得板块指数8806XX成分股？
- 开放微博社区使用OAUTH协议简介
- 计算机属性资源管理器已停止工作,windows资源管理器已停止工作,教您windows资源管理器已停止工作如何解决...
- UE4 UMG多行文本
- 智慧树期末考试可以切换页面吗_智慧树考试可以切换界面吗？中途可以退出吗...
- 第二章 马尔可夫决策过程及表格型方法
- WinCC 7.3 + SQL server（杂）
- 三国杀 中的概率计算 几何分布的期望和方差分析
- 微信公众号开发----生成带参数的临时二维码
- xp系统创建wifi连接服务器,xp笔记本连接无线wifi简便教程
- 作为学生借助 CODECHINA 这样学习 Git 知识