Oracle学习笔记
Oracle基本命令
OracleService<数据库名字> --数据库的服务
网络服务名 TNSNAMES 被客户端使用连库串
--------------------------------
连接命令
sqlplus shebao@dddddddd
sqlplus 用户@网络服务名
创建表空间
--数据文件的存放地-- 表空间
create tablespace [表空间]
datafile '[文件]'
size [大小]
autoextend on next [自增大小]
maxsize unlimited;
--用户
create user [用户]
identified by "[密码]"
default tablespace [表空间]
temporary tablespace temp
profile default;
--授权
grant connect,dba,resource to [用户]
连接字符串
jdbc.driverClassName=oracle.jdbc.driver.OracleDriver
jdbc.url=jdbc:oracle:thin:@localhost:1521:turing
权限
--直接权限和间接权限的区别 : 封装体内
revoke SELECT ANY TABLE from <user>;
grant SELECT ANY TABLE to <user>;
grant dba ,connect, resource to <user>;
select * from shebao.gsbx_dwxxb;
create view vvvv as
select * from shebao.gsbx_dwxxb;
declare
i number :=0;
begin
dbms_output.enable(999999);
for c in (select table_name c1,comments c2 from dictionary) loop
i := i + 1;
dbms_output.put_line(lpad(i,4,' ')||' '||c.c1);
end loop;
end;
create procedure print as
i number :=0;
begin
dbms_output.enable(999999);
for c in (select table_name c1,comments c2 from dictionary) loop
i := i + 1;
dbms_output.put_line(lpad(i,4,' ')||' '||c.c1);
end loop;
end;
begin
print;
end;
一般的数据库规范
1 单独指定4个超级用户的密码
2 创建好相应的数据库角色
A dba 备份 还原 导入 导出 建表 删表 改表 建包 修改profile 修改parameter
B manager 备份 还原 导入 导出 建表 删表 改表
c developer 增删改查 某些表 在某些特定包下 建立存储过程 在某些特定表下 建立触发器
3 创建数据库用户 授予角色
4 锁定4个超级用户 只允许本机登陆
数据的导入与导出
导出
exp 用户密码@网络服务名 file=文件地址 owner = 导谁的东西
exp shebaosa@dddddddd file=eshebao_bcode.dmp owner = shebao
exp 用户密码@网络服务名 file=文件地址 tables = 导哪些表
exp shebaosa@dddddddd file=eshebao_bcode.dmp tables=
导入
imp 用户密码@网络服务名 file=文件地址 fromuser = 从哪来 touser= 进哪去
imp shebaosa@dddddddd file=eshebao_bcode.dmp fromuser = shebao touser= liushengbo
类型：
number
默认是number(38) 总长不超过38就行
Number(3,4) 只能存储形如333.4444
Number(5,0)只能存储整数
varchar2
是一个最普通的变长的字符型，会将存储数据编码转换为本地数据库编码
char
Char是一定长的字符型，会用空格补齐长度
Db link
create public database link <名>
connect to <用户> identified by "<密码>" using '<网络服务名>'
select * from 远程表@<dblink的名字>
约束
--约束 保护数据库数据完整性 一致性的 数据库对象
1 primary key P
2 foreign key R
3 unique U
4 not null C
select * from user_constraints
函数
日期函数
日期函数
sysdate to_date months_between add_months next_day last_day
例句：(sysdate)
select sysdate ,'2011-12-15 9:04:40' b from dual
to_date(值,格式化字符)
例句：(to_date)
两个日期求差 单位是天 周时分秒 直接减
select (sysdate - to_date('1995-06-15','yyyy-mm-dd')) from dual
例句：(months_between)
两个日期求差 单位是月 年 months_between
select months_between(sysdate , to_date('1995-06-15','yyyy-mm-dd')) from dual
日期向前/后推算 单位是天 周时分秒 直接加减
select sysdate-26 from dual
日期向前/后推算 单位是月 年 add_months
例句：(add_months)
select add_months(sysdate,-26) from dual
例句：(next_day)
下一个星期几
select next_day(sysdate,'星期五') from dual
例句：(last_day)
某月最后一天
select last_day(to_date('2011-02-02','yyyy-mm-dd')) from dual
字符函数
例句：(to_char)
select to_char(sysdate,'Month ddth yyyy Day','nls_date_language=american') from dual
mi 分钟
hh 12小时制的小时
hh24 24小时制的小时
ss 本分第几秒
sssss 本天第几秒
w 本月第几周
ww 本年第几周
day 星期几
q 第几季
nls_date_language=形容词格的国家
2011年12月15号怎么显示?
select to_char(sysdate,'yyyy"年"mm"月"dd"号"') from dual
数字函数
select mod(9,3) from dual
select power(5,3) from dual
select power(5,1/2) from dual
select power(5,-2) from dual
select ceil(3.14159) ,floor(3.14159) from dual
select round(3.14159,4),round(13.14159,-1),round(3.14159) from dual
select trunc(3.14159,4),trunc(19.14159,-1),trunc(3.14159),
trunc(sysdate),trunc(sysdate,'hh')
from dual
select abs(-333),log(2,32),ln(2.71828),sign(98874-555.32),
sin(3.1415926535/6)
from dual
sequence
sequence 序列 可以提供一个不断自增的整数(nextval)
例句：(创建序列)
create sequence seq_test
minvalue 1000000
maxvalue 9999999
start with 1000000
increment by 1
cache 20
例句：(使用序列)
select seq_test.nextval from dual
select seq_test.nextval,seq_test.nextval from scott.emp
在hibernate 中使用序列
<generator class="sequence">
<param name="sequence">seq_test</param>
</generator>
q 第几季
nls_date_language=形容词格的国家
2011年12月15号怎么显示?
select to_char(sysdate,'yyyy"年"mm"月"dd"号"') from dual
在hibernate 中使用序列
<generator class="sequence">
<param name="sequence">seq_test</param>
</generator>
日期类的所有函数
sysdate months_between add_months to_date() to_char()
select next_day(sysdate,'星期五') from dual
select last_day(to_date('2011-02-02','yyyy-mm-dd')) from dual
数字类
求余：mod(9,3)
求指数：power(5,3)
开方：power(5,0.5)
取整(顶)：ceil
取底floor
精确取位：round(33.11111,3)
数字类 加减乘除余平方开放
+ - * /
select mod(9,3) from dual
select power(5,3) from dual
select power(5,1/2) from dual
select power(5,-2) from dual
select ceil(3.14159) ,floor(3.14159) from dual
select round(3.14159,4),round(13.14159,-1),round(3.14159) from dual
select trunc(3.14159,4),trunc(19.14159,-1),trunc(3.14159),
trunc(sysdate),trunc(sysdate,'hh')
from dual
select abs(-333),log(2,32),ln(2.71828),sign(98874-555.32),
sin(3.1415926535/6)
from dual
字符函数
--upper lower initcap
select upper('Oracle_training') a,
lower('Oracle_training') b,
initcap('oraCle?traiNing') c from dual
--ascii chr ||
select ascii('A') , chr(38) from dual
select 'a'||chr(10)||'b' from dual
select '131'||'34563412' from dual
--lpad rpad
select lpad('3412',11,' ') from dual
--ltrim rtrim trim
select ltrim(' Oracle_training ') a1,rtrim(' Oracle_training ') a2,
' Oracle_training ' b from dual
select ltrim('Oracle_training','arO') a1,
rtrim('Oracle_tringingaining','ing') a1 from dual
select trim (' Oracle_training ') t1,
trim ('O' from 'OOOOracle_trainOngOOO') t2,
trim (leading 'O' from 'OOOOracle_trainOngOOO') t3 from dual
-- length instr substr replace
select length('汉字') from dual
select instr('abstractabdbaadbnaab3','a',-1,4) from dual
select substr('achdgjht',3,5),substr('achdgjht',-3,2) from dual
select substr('achdgjgt',4) from dual
select replace('ahdabaannarayarat','aa','Cow') from dual
select replace('ahdabaannarayarat','a') from dual
-通达街-西大桥-工大-铁路局-建设街-医大医院-烟厂-承德广场-
-承德广场-烟厂-医大医院-铁路局-工大-西大桥-通达街-
第一站
select substr(stops,1, 第一个"-" - 1) from busstop
select substr(stops,1, instr(stops,'-',1,1) - 1) from busstop
第三站
select substr(stops,第二个"-" 后面1个, 到第三个"-" - 第二个"-" -1) from busstop
select substr(stops,
instr(stops,'-',1,2) + 1,
instr(stops,'-',1,3) - instr(stops,'-',1,2) -1) a from busstop
最后一站
select substr(stops,最后一个"-" 后面一个) from busstop
select substr(stops,instr(stops,'-',-1,1)+1) from busstop
第七站
select substr(stops,
instr(stops,'-',1,6) + 1,
instr(stops,'-',1,7) - instr(stops,'-',1,6) -1) a from busstop
---查询时 数据库里的数据如果不符合规律 可以在SQL里进行补全 让其变得有规律
select substr('-'||stops||'-',
instr('-'||stops||'-','-',1,1) + 1,
instr('-'||stops||'-','-',1,2) - instr('-'||stops||'-','-',1,1) -1) a from busstop
public static void main(String[] args) {
// String a = "[email protected]";
// System.out.println(a.matches("[A-Za-z]+@[a-z0-9]{2,20}\\.com(\\.cn)?"));
String b = "承德广场-烟厂-医大医院-铁路局-工大-西大桥-通达街";
Pattern p = Pattern.compile("[^\\-]+");
Matcher m = p.matcher(b);
while (m.find()){
System.out.println(m.group());
}
}
//
正则表达式 抽象描述一组字符 的 表达式
[] 来表示 单个字符
a 穷举[wxyz] 四者之一
b 范围[a-z] 小写字母26之一
a+b [a-zORACLE]
c 取反[^A] 除了A什么都可以
c+a [^wxyz] 除了w,x,y,z什么都可以
c+b [^a-z] 除了小写字母什么都可以
{} 来表示 前一个表达式出现的次数
a 从到 {1,5} 1~5次都可以
b 从到无穷{1,}
转义
\[ \} \- \\ \^ \$ \? \+ \*
缩写
[A-Za-z0-9_] \w
[0-9] \d
{0,1} ?
{1,} +
{0,} *
---- 查询数学成绩比外语成绩高的同学的学号、姓名、数学成绩、外语成绩、差多少分。
decode 函数用法
decode 最丰
nvl(变量,替换值) = nvl2(变量,变量,替换值) = decode(变量,null,替换值,变量)
nvl2(变量,非空值1,空值2) = decode(变量,null,空值2,非空值1)
行转列：
decode case when then else end
分别用case和decode函数列出员工所在的部门，
deptno=10显示'部门10',
deptno=20显示'部门20'
deptno=30显示'部门30'
deptno=40显示'部门40'
否则为'其他部门'
select ename ,deptno,
decode(deptno,10,'部门10',20,'部门20',30,'部门30',40,'部门40','其它部门') a from scott.emp
select ename ,deptno, case deptno
when 10 then '部门10'
when 20 then '部门20'
when 30 then '部门30'
when 40 then '部门40'
else '其它部门' end a
from scott.emp
--2 套路 把一列数字拆分成几列的时候 decode(判断条件, ,1,0) 相当于对某个列使用where
select stu_id, stu_name, sum(decode(subject,'数学',score,0)) sc1 ,sum(decode(subject,'外语',score,0)) sc2 from score -- where subject = '数学'
group by stu_id, stu_name
递归题目
select * from (
select ename,(select count(*) from scott.emp
where ename <> e.ename
connect by mgr = prior empno
start with ename=e.ename ) c ,
(select sum(sal) from scott.emp
where ename <> e.ename
connect by mgr = prior empno
start with ename=e.ename )s,
(select max(sal) from scott.emp
where ename <> e.ename
connect by mgr = prior empno
start with ename=e.ename )ma,
(select min(sal) from scott.emp
where ename <> e.ename
connect by mgr = prior empno
start with ename=e.ename )mi from scott.emp e)
where c > 0
select sys_connect_by_path(ename,'-') path, --树上一支
level , --第几层
connect_by_isleaf, --是否还有子节点
connect_by_root(ename) --根节点的属性
empno, ename, job, mgr, hiredate, sal, comm, deptno
from scott.emp
connect by mgr = prior empno
start with mgr is null
rownum 分页
select * from
(
select rownum rn, empno, ename, job, mgr, hiredate, sal, comm, deptno from emp
where rownum < 7
)
where rn >3
--------------------------两层 / 三层分页---------------------------
select * from
(
select rownum rn,t.* from
(
select empno, ename, job, mgr, hiredate, sal, comm, deptno from emp
order by sal
)t where rownum < 7
)
where rn >3
-------------------------------------------------
select * from
(
select row_number() over(order by sal) rn,
empno, ename, job, mgr, hiredate, sal, comm, deptno from emp
)
where rn >3 and rn <7
---横向
select substr(mymonth,1,4) year,
sum(decode(substr(mymonth,5,2),'01',sell,0)) mon1,
sum(decode(substr(mymonth,5,2),'02',sell,0)) mon2,
sum(decode(substr(mymonth,5,2),'03',sell,0)) mon3,
sum(decode(substr(mymonth,5,2),'04',sell,0)) mon4,
sum(decode(substr(mymonth,5,2),'05',sell,0)) mon5,
sum(decode(substr(mymonth,5,2),'06',sell,0)) mon6,
sum(decode(substr(mymonth,5,2),'07',sell,0)) mon7,
sum(decode(substr(mymonth,5,2),'08',sell,0)) mon8,
sum(decode(substr(mymonth,5,2),'12',sell,0)) mon12 from sale
group by substr(mymonth,1,4) order by substr(mymonth,1,4)
select mymonth, sell ,
decode(substr(mymonth,5,2),'01',sell,0) mon1,
decode(substr(mymonth,5,2),'02',sell,0) mon2,
decode(substr(mymonth,5,2),'03',sell,0) mon3,
decode(substr(mymonth,5,2),'04',sell,0) mon4,
decode(substr(mymonth,5,2),'12',sell,0) mon12 from sale
--纵向
select substr(mymonth,1,4) year, sum(sell) from (
select to_char(n,'yyyymm') mymonth, nvl(sell,0) sell from sale
right join (select add_months(to_date('199912','yyyymm'),level) n from dual connect by level <= 48)t
on to_char(n,'yyyymm')= mymonth )
group by substr(mymonth,1,4)
union all
select to_char(n,'yyyymm') mymonth, nvl(sell,0) sell from sale
right join (select add_months(to_date('199912','yyyymm'),level) n from dual connect by level <= 48)t
on to_char(n,'yyyymm')= mymonth
order by year
横向拆分数据 sum(decode(**,,1,0))
纵向拆分数据 group by
rowid 行标记 在表空间内不重复 可以做最后的定位
delete from emp where rowid not in
(
select max(rowid)
from emp group by empno, ename, job, mgr, hiredate, sal, comm, deptno
)
PL-SQL
基本语法
declare
begin
dbms_output.put_line('Helo,World!'); --System.out.println();
end;
例句：
抛异常
declare
v number(10);
begin --try{
--select 1 into v from dual where 1=2;
dbms_output.put_line(1/0);
exception --}catch (Exception e){
when others then
dbms_output.put_line(sqlerrm); -- e.printStackTrace();
end;
例句1：(循环_for_打印金字塔)
：
declare --pl-sql Procedural Language Structed Query Language
begin
for h in 1..5 loop
for k in 0..5-h-1 loop
dbms_output.put(' ');
end loop;
for o in 0..2*h-1-1 loop
dbms_output.put('o');
end loop;
dbms_output.put_line('');
end loop;
end;
例句2：(循环_while_蛤蟆爬井)
declare
gao number(38,0) := 0; --声明区 赋值用 :=
bw boolean := true;
begin
while gao < 10 loop
if bw = true then --判断用 =
gao := gao + 5;
bw := false;
else
gao := gao - 4;
bw := true;
end if;
dbms_output.put_line(gao);
end loop;
end;
例句3：(循环_do..while_)
declare
i number(10,0) :=0 ;
begin
loop
exit when i>6 ;
dbms_output.put_line(i);
i := i+1;
end loop;
end;
例句3：(选择_if else)
选择
declare
s number(10,0) := 4 ;
begin
if s = 1 then
dbms_output.put_line('31');
elsif s=2 then
dbms_output.put_line('28');
elsif s in( 3,5,7,8,10,12) then
dbms_output.put_line('31');
else
dbms_output.put_line('30');
end if;
end;
例句4：(选择_case)
declare
s number(10,0) := 11 ;
begin
case s
when 1 then
dbms_output.put_line('31');
when 2 then
dbms_output.put_line('28');
when 3 then
dbms_output.put_line('31');
when 4 then
dbms_output.put_line('30');
when 5 then
dbms_output.put_line('31');
when 6 then
dbms_output.put_line('30');
when 7 then
dbms_output.put_line('31');
when 8 then
dbms_output.put_line('31');
when 9 then
dbms_output.put_line('30');
when 10 then
dbms_output.put_line('31');
when 11 then
dbms_output.put_line('30');
when 12 then
dbms_output.put_line('31');
end case;
end;
游标
游标 - 数据库内部对象 -- java : ResultSet
1 某个结果集的路标(内存地址) cursor c is select type, subtype from tongji;
2 操作结果集 读取值的方法 fetch <游标> into <变量>,....
3 此结果集的状态参数 %notfound %found %rowcount
select =============> 替代游标循环
where =============> 替代选择结构
函数 =============> 替代拼接,计算
sql =====思路简化===== pl-sql
动态游标
静态游标不好使，有了关键位置变量的时候使用动态游标
作用：1,、动态游标可以封装
2、可以做返回（可以用存储过程实现）—程序复杂到无法用HQL语句来完成
3、
使用方式
动态游标 Ref Cursor ==> sys_refcursor
declare
c sys_refcursor; --声明动态游标类型
n number(10,0):=18;
tbname varchar2(40) := 'month';
--cursor c is select code,codename from shebao.tbname where rownum <= n;
v_code varchar2(40);
v_codename varchar2(40);
begin
open c for 'select code,codename from shebao.'||tbname||' where rownum <= '||n||'';
loop
fetch c into v_code,v_codename;
exit when c%notfound;
dbms_output.put_line(v_code||' '||v_codename);
end loop;
close c;
end;
例句：(创建自己的代码表)__绑定变量
declare
c sys_refcursor;
n number(10,0):=11;
tbname varchar2(40) := 'month';
--cursor c is select code,codename from shebao.tbname where rownum <= n;
v_code varchar2(40);
v_codename varchar2(40);
begin
execute immediate
'create table '||tbname||' (code varchar2(40),codename varchar2(40))';
open c for 'select code,codename from shebao.'||tbname||' where rownum <= :nn' using n;
loop
fetch c into v_code,v_codename;
exit when c%notfound;
execute immediate
'insert into '||tbname||' values(:p1,:p2)' using v_code,v_codename;
end loop;
commit;
close c;
end;
例句：(用于返回)__分页存储过程
create or replace procedure pro_page_cur(v_sql in varchar,cong in number,dao in number, rs out sys_refcursor) as
str_pre varchar2(99) := 'select * from (select rownum rn,t.* from (';
str_nxt varchar2(99) := ' )t where rownum <= :dao) where rn >:cong ';
begin
open rs for str_pre||v_sql||str_nxt using dao,cong;
end;
———java中调用代码：
import java.sql.*;
public class RunPro{
public static void main(String[] args) throws SQLException,ClassNotFoundException{
Class.forName("oracle.jdbc.driver.OracleDriver");
Connection con = DriverManager.getConnection("jdbc:oracle:thin:@localhost:1521:turing","liushengbo000","sa");
//pro_page_cur(v_sql in varchar,cong in number,dao in number, rs out sys_refcursor)
CallableStatement cst = con.prepareCall("{call pro_page_cur(?,?,?,?)}");
cst.setString(1,"select * from nation");
cst.setInt(2,3);
cst.setInt(3,10);
cst.registerOutParameter(4,-10);//oracle.sql.OracleType.CURSOR
cst.execute();
ResultSet rs = (ResultSet)cst.getObject(4);
while (rs.next()){
System.out.println(rs.getInt(1)+" "+rs.getString(2)+" "+rs.getString(3));
}
rs.close();
cst.close();
con.close();
}
}
隐式调用
begin
for hang in (select type, subtype from tongji where rownum <=20) loop
dbms_output.put_line(hang.type||' '||hang.subtype);
end loop;
end;
------------------- 以上 称为 隐式调用 ----------
显示调用
declare
cursor c is select type, subtype from tongji;
col1 varchar2(40);
col2 varchar2(400);
begin
open c;
loop
fetch c into col1,col2;
exit when c%notfound;
dbms_output.put_line(col1||' '||col2);
end loop;
close c;
end;
例题
例句1：(其它数据库转移过来的大小写混合的表)
应用1 其它数据库转移过来的大小写混合的表
select ct_id, code, codename, fathercode from "AdjustCause"
alter table "AdjustCause" rename to AdjustCause
declare
sqlstr varchar2(400);
begin
for hang in (select table_name from user_tables where table_name <> upper(table_name) ) loop
sqlstr := 'alter table "'||hang.table_name||'" rename to '||hang.table_name||'';
--dbms_output.put_line(sqlstr) ;
execute immediate sqlstr;
end loop;
end ;
例句2：(写一段 PL-SQL 把当前用户名下大小写混合的字段改为大写)
declare
sqlstr varchar2(400);
begin
for c in (select table_name ,column_name ,data_type from user_tab_cols
where column_name <> upper(column_name)) loop
sqlstr := 'alter table '||c.table_name||' rename column "'||c.column_name||'" to '||c.column_name||'';
--dbms_output.put_line(sqlstr);
execute immediate sqlstr;
end loop;
end;
例句3：(写一段 PL-SQL 把当前用户名下表中的空格去掉)
方法1
declare
sqlstr varchar2(4000);
begin
dbms_output.enable(1000000);
for c in (select 'update '||table_name||' set '||wm_concat(column_name||'=trim('||column_name||')') str
from user_tab_cols where data_type='VARCHAR2'
group by table_name) loop
sqlstr := c.str;
dbms_output.put_line(sqlstr);
--execute immediate sqlstr;
end loop;
end;
方法2
declare
sqlstr varchar2(4000);
shu number;
begin
dbms_output.enable(1000000);
for c in (select table_name from user_tab_cols group by table_name) loop
sqlstr := 'update '||c.table_name||' set ';
for d in (select column_name from user_tab_cols where table_name = c.table_name and data_type='VARCHAR2' ) loop
sqlstr := sqlstr||d.column_name||'=trim('||d.column_name||'),';
end loop;
shu := instr(sqlstr,',',1,1);
if shu <> 0 then
sqlstr := rtrim(sqlstr,',');
--dbms_output.put_line(sqlstr);
execute immediate sqlstr;
end if;
end loop;
commit;
end;
动态sql (execute immediate)
如果其它位置使用变量 需要使用动态SQL execute immediate
--动态语句的使用方式
1如果其它位置使用变量 需要使用动态SQL execute immediate
a insert/delete/update/create/drop/alter 直接 execute immediate 'i/d/u'
b select into 直接 execute immediate 'select ....' into
2如果使用DDL语句(create drop alter purge) 无论有没有变量都必须动态SQL
3如果表(其它对象)是在当前语句块中创建出来的,
对于语句块里的其它语句,它们在编译时是不存在的
动态创建对象,后面的操作语句必须动态sql
例句1：
declare
tabname varchar2(40) := 'aa_test';
begin
execute immediate
'create table '||tabname||'(id number,name varchar2(40))';
execute immediate
'insert into aa_test(id, name)values(1, ''张三'')';
--dbms_output.put_line('insert into aa_test(id, name)values(1, ''张三'')');
end;
存储过程
创建
create [or replace] procedure <名>(<参数名> <进出> <参数类型>[,.....]) as
--declare 去掉
例句：
创建存储过程
create or replace procedure pro_yue1(mon in number) as
--declare
s number(10,0) := mon ;
begin
case
when s in (1,3,5,7,8,10,12) then
dbms_output.put_line('31');
when s in (4,6,9,11) then
dbms_output.put_line('30');
when s = 2 then
dbms_output.put_line('28');
end case;
end;
带输出存储过程
create or replace procedure pro_yue2(mon in number,dds out number) as
--declare
s number := mon;
begin
case
when s in (1,3,5,7,8,10,12) then
dds := 31;
when s in (4,6,9,11) then
dds := 30;
when s = 2 then
dds := 28;
end case;
end;
执行过程
数据库语句：
declare
begin
pro_yue1(2);
end;
java环境中调用（不带参数）
package a.b.c;
import java.sql.*;
public class 存储过程{
public static void main(String[] args) throws SQLException,ClassNotFoundException{
Class.forName("oracle.jdbc.driver.OracleDriver");
Connection con = DriverManager.getConnection("jdbc:oracle:thin:@localhost:1521:turing","dba_01","sa");
CallableStatement cst = con.prepareCall("{call pro_upper_tableName}");
cst.execute();
cst.close();
con.close();
}
}
Java环境中调用（带参数）
package a.b.c;
import java.sql.*;
public class 存储过程带参数{
public static void main(String[] args) throws SQLException,ClassNotFoundException{
Class.forName("oracle.jdbc.driver.OracleDriver");
Connection con = DriverManager.getConnection("jdbc:oracle:thin:@localhost:1521:turing","liushengbo000","sa");
//pro_yue2(mon in number,dds out number) as
CallableStatement cst = con.prepareCall("{call pro_yue2(?,?)}");
cst.setInt(1,2);
cst.registerOutParameter(2,java.sql.Types.INTEGER);//java.sql.Types.INTEGER=4
cst.execute();
System.out.println(cst.getInt(2));
cst.close();
con.close();
}
}
自定义函数（function）
必须有返回值
创建
create or replace function fun_yue(mon in number)return number as
--declare
s number(10,0) := mon ;
begin
case
when s in (1,3,5,7,8,10,12) then
return 31; --必须有返回值
when s in (4,6,9,11) then
return 30;
when s = 2 then
return 28;
else
return null;
end case;
end;
使用
select fun_yue(100) from dual
包
包内函数 包内过程
将功能相近/结构相似/相互之间有调用依赖关系的
type constant exception function procedure
封装到一起 就是包 package
目录(directory)
目录 directory 对硬盘上文件夹的一个映射 绝大部分时候用来配合Oracle文件读写包操作
创建：
create directory dir_me as 'E:\ora';
使用：
写文件
declare
f utl_file.file_type ; --文件读写包
begin
f := utl_file.fopen('DIR_ME','test.txt','a'); --w - writable r - readonly a – all
--第一个参数必须完全大写，2个参数文件名，3读写方式
utl_file.put_line(f,'内容内容内容内容内容内容内容');
utl_file.fclose(f);
end;
读文件
declare
f utl_file.file_type ;
neirong varchar2(999);
begin
f := utl_file.fopen('DIR_ME','Tqws.java','r'); --w - writable r - readonly a - all
loop
utl_file.get_line(f,neirong);
dbms_output.put_line(neirong);
end loop;
exception
when no_data_found then
dbms_output.put_line('--正常结束--');
utl_file.fclose(f);
when others then
dbms_output.put_line('--非正常结束--');
utl_file.fclose(f);
end;
触发器（trigger）
作用：
监听 捕获到指定的事件时 自动执行 无法调用 没有参数
触发器可以针对用户的数据库操作 补全业务逻辑 简化用户的数据库操作逻辑
创建：
create or replace trigger tri_casade
--
例句
例句1：(级联删除)
--级联删除
create trigger tri_casade
before delete on dept for each row
declare
begin
delete from emp where deptno = :old.deptno;
end;
例句2：(生成主键)
--生成CT_ID
create or replace trigger tri_enterprisetype
before insert on shebao.enterprisetype for each row
declare
entityName varchar2(40) := 'enterprisetype';
maxid varchar2(40);
seq number;
clen number;
ctype varchar2(40);
finalid varchar2(40);
begin
execute immediate 'select max(ct_id) from shebao.'||entityName||'' into maxid;
select codelength,codetype into clen,ctype from shebao.codenavigation where upper(codetype) = upper(entityName);
if maxid is null then
seq := 1;
else
seq := substr(maxid,-clen);
seq := seq + 1;
end if;
finalid := ctype||lpad(seq,clen,'0');
:new.ct_id := finalid;
end;
教师机 10:02:44 (多人发送)
--级联删除
create trigger tri_casade
before delete on dept for each row
declare
begin
delete from emp where deptno = :old.deptno;
end;
例句3：()
create or replace trigger tri_casade
before delete or insert or update of loc on dept for each row
declare
f utl_file.file_type ;
begin
if inserting then
null;
elsif updating then
f := utl_file.fopen('DIR_ME','test.log','a');
utl_file.put_line(f,'老部门是:'||:old.dname);
utl_file.put_line(f,'新部门是:'||:new.dname);
:new.dname := :old.dname;
utl_file.fclose(f);
elsif deleting then
delete from emp where deptno = :old.deptno;
end if;
end;