1、测试表及数据准备
create table T_update01(ID int ,infoname varchar2(32),sys_guid varchar2(36));
create table T_update02(ID int ,infoname varchar2(32),sys_guid varchar2(36));
insert into T_update01
select 1,N'1_updateName',sys_guid() from dual
union
select 2,N'2_updateName',sys_guid() from dual;
commit;
insert into T_update02
select 1,N'update_set_exists',sys_guid() from dual;
insert into T_update02
select 2,N'update_set_cursor',sys_guid() from dual;
insert into T_update02
select 3,N'3_Name',sys_guid() from dual;
commit;
-- 查询表T_update01、T_update02
select * from T_update01;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 1_updateName 189F5A1099BF6606E0639C0AA8C0F15E
2 2_updateName 189F5A1099C06606E0639C0AA8C0F15E
select * from T_update02;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 update_set_exists 189F5A1099C46606E0639C0AA8C0F15E
2 update_set_cursor 189F5A1099C56606E0639C0AA8C0F15E
3 3_Name 189F5A1099C66606E0639C0AA8C0F15E
2、update set column … where exists
2.1、update set 单列字段
-- update set 单列字段，更新满足关联条件的所有数据
update T_update01 T1
set infoname=(select T2.infoname from T_update02 T2 where T2.ID=T1.ID)
where exists (select 1 from T_update02 T2 where T2.ID=T1.ID );
-- update set 单列字段 ，更新满足特定条件ID=1的数据
update T_update01 T1
set infoname=(select T2.infoname from T_update02 T2 where T2.ID=T1.ID)
where T1.ID=1;
-- 本次执行更新满足特定条件T_update01表的ID=1
SCOTT@prod02> select * from T_update01;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 update_set_exists 189F5A1099BF6606E0639C0AA8C0F15E
2 2_updateName 189F5A1099C06606E0639C0AA8C0F15E
2.2、update set 多列字段
-- T_update01表多插入一行数据
insert into T_update01
select 3,N'insert03',sys_guid() from dual;
commit;
select * from T_update01;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 update_set_exists 189F5A1099BF6606E0639C0AA8C0F15E
2 2_updateName 189F5A1099C06606E0639C0AA8C0F15E
3 insert03 189F5A1099C76606E0639C0AA8C0F15E
update T_update01 T1
set (sys_guid,infoname) = (select T2.sys_guid,T2.infoname from T_update02 T2 where T2.ID=T1.ID)
where exists (select 1 from T_update02 T2 where T2.ID=T1.ID );
commit;
-- 更新后检查，sys_guid,infoname两列的值和T_update02一样了
select * from T_update01;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 update_set_exists 189F5A1099C46606E0639C0AA8C0F15E
2 update_set_cursor 189F5A1099C56606E0639C0AA8C0F15E
3 3_Name 189F5A1099C66606E0639C0AA8C0F15E
select * from T_update02;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 update_set_exists 189F5A1099C46606E0639C0AA8C0F15E
2 update_set_cursor 189F5A1099C56606E0639C0AA8C0F15E
3 3_Name 189F5A1099C66606E0639C0AA8C0F15E
3、使用游标
-- T_update02数据更新一下，方便使用游标更新的结果显示
update T_update02 set INFONAME='cursor is select' where id>=2;
commit;
select * from T_update02;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 update_set_exists 189F5A1099C46606E0639C0AA8C0F15E
2 cursor is select 189F5A1099C56606E0639C0AA8C0F15E
3 cursor is select 189F5A1099C66606E0639C0AA8C0F15E
-- 使用用游标更新T_update01的INFONAME字段，使其和T_update02 where id>=2
declare
cursor cur_my_source is select infoname,id from T_update02;
begin
for cur_my_target in cur_my_source loop
update T_update01 set infoname=cur_my_target.infoname where id=cur_my_target.id;
end loop;
commit;
end;
/
-- 检查查询结果
select * from T_update01;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 update_set_exists 189F5A1099C46606E0639C0AA8C0F15E
2 cursor is select 189F5A1099C56606E0639C0AA8C0F15E
3 cursor is select 189F5A1099C66606E0639C0AA8C0F15E
4、merge into子句
create table T_merg01(ID int ,infoname varchar2(32),sys_guid varchar2(36));
create table T_merg02(ID int ,infoname varchar2(32),sys_guid varchar2(36));
insert into T_merg01
select 1,N'1_Name',sys_guid() from dual
union
select 2,N'2_Name',sys_guid() from dual;
commit;
select * from T_merg01;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 1_Name 189F5A1099BB6606E0639C0AA8C0F15E
2 2_Name 189F5A1099BC6606E0639C0AA8C0F15E
insert into T_merg02
select 1,N'merge_into_Name1',sys_guid() from dual;
insert into T_merg02
select 3,N'3_Name',sys_guid() from dual;
select * from T_merg02;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 merge_into_Name1 189F5A1099BD6606E0639C0AA8C0F15E
3 3_Name 189F5A1099BE6606E0639C0AA8C0F15E
merge into T_merg01 T1
using T_merg02 T2 on (T1.id=T2.id)
when matched then update set infoname=T2.infoname
when not matched then insert (ID,infoname,sys_guid) values(T2.ID ,T2.infoname,T2.sys_guid);
commit;
select * from T_merg01;
ID INFONAME SYS_GUID
---------- ------------------------------ ------------------------------------
1 merge_into_Name1 189F5A1099BB6606E0639C0AA8C0F15E
2 2_Name 189F5A1099BC6606E0639C0AA8C0F15E
3 3_Name 189F5A1099BE6606E0639C0AA8C0F15E
-- 可以发现T_merg01表的ID=1的INFONAME=merge_into_Name1和T_merg02表ID=1的值一样了
-- 可以发现T_merg01表多了一行数据是T_merg02表ID=3的这一行数据
5、Oracle 23c/AI 新特性
不论是已发版本Oracle23c free还是最终发布的长期支持的Oracle23Ai，表关联更新update和删除delete语句易用且更加优雅，类似SQLServer的关联更新
以下操作基于的环境
SQL*Plus: Release 23.0.0.0.0 – Developer-Release on Fri May 17 11:17:54 2024
Version 23.2.0.0.0
5.1、关联更新update
TESTUSER@FREEPDB1> create table t_emp as select EMPLOYEE_ID,DEPARTMENT_ID,SALARY from employees;
Table created.
TESTUSER@FREEPDB1> desc t_emp;
Name Null? Type
----------------------------------------- -------- ----------------------------
EMPLOYEE_ID NUMBER(6)
DEPARTMENT_ID NUMBER(4)
SALARY NUMBER(8,2)
TESTUSER@FREEPDB1> select * from t_emp where DEPARTMENT_ID=110;
EMPLOYEE_ID DEPARTMENT_ID SALARY
----------- ------------- ----------
205 110 12008
206 110 8300
TESTUSER@FREEPDB1> update t_emp set DEPARTMENT_ID=null,SALARY=null where DEPARTMENT_ID=110;
2 rows updated.
TESTUSER@FREEPDB1> commit;
Commit complete.
TESTUSER@FREEPDB1> select * from t_emp where DEPARTMENT_ID is null;
EMPLOYEE_ID DEPARTMENT_ID SALARY
----------- ------------- ----------
178 7000
205
206
-- oracle 23c SQL增强 表关联更新
TESTUSER@FREEPDB1> update t_emp t1 set t1.DEPARTMENT_ID=t2.DEPARTMENT_ID,t1.SALARY=t2.SALARY from employees t2 where t2.EMPLOYEE_ID=t1.EMPLOYEE_ID and t1.DEPARTMENT_ID is null;
3 row updated.
TESTUSER@FREEPDB1> commit;
Commit complete.
TESTUSER@FREEPDB1> select t1.* from t_emp t1 where t1.DEPARTMENT_ID=110;
EMPLOYEE_ID DEPARTMENT_ID SALARY
----------- ------------- ----------
205 110 12008
206 110 8300
5.2、关联删除delete
TESTUSER@FREEPDB1> delete t_emp t1 from employees t2 where t2.EMPLOYEE_ID=t1.EMPLOYEE_ID and t2.DEPARTMENT_ID=110; 45 rows deleted. TESTUSER@FREEPDB1> commit; Commit complete. TESTUSER@FREEPDB1> select t1.* from t_emp t1 where t1.DEPARTMENT_ID=110; no rows selected
以上就是Oracle表关联更新几种方法小结的详细内容，更多关于Oracle表关联更新的资料请关注IT俱乐部其它相关文章！
