批处理 动态sql

１.
DECLARE
TYPE ref_cursor_type IS ref CURSOR;
v_mycursor ref_cursor_type;
TYPE id_list IS TABLE OF integer;
TYPE name_list IS TABLE OF varchar2(30);
v_tabid id_list:=id_list();
v_tabname name_list:=name_list();
sql_str varchar2(200);
BEGIN
--查询所以行,放在集合里
sql_str:='select empno,ename from emp';
sql_str:=sql_str||' order by empno desc';
execute immediate sql_str BULK COLLECT INTO v_tabid,v_tabname;
FOR c IN v_tabid.first..v_tabid.last LOOP
dbms_output.put_line('empno为'||v_tabid(c)||' 记录的NAME为'||v_tabname(c));
END LOOP;
dbms_output.put_line('---------------------------------');
--更新(返回更新后的值)
sql_str:='update emp set empno=1+empno,ename=''a'' where rownum=1 RETURNING empno,ename into :1,:2 ';
execute immediate sql_str RETURNING BULK COLLECT INTO v_tabid, v_tabname;
FOR c IN v_tabid.first..v_tabid.last LOOP
dbms_output.put_line('empno为'||v_tabid(c)||' 记录的NAME为'||v_tabname(c));
END LOOP;
dbms_output.put_line('---------------------------------');
--删除(返回被删除的行)
sql_str:='delete from emp where rownum<=2 RETURNING empno,ename into :1,:2 ';
execute immediate sql_str RETURNING BULK COLLECT INTO v_tabid, v_tabname;
FOR c IN v_tabid.first..v_tabid.last LOOP
dbms_output.put_line('empno为'||v_tabid(c)||' 记录的NAME为'||v_tabname(c));
END LOOP;
dbms_output.put_line('---------------------------------');
--插入(返回插入的行)
sql_str:='insert into emp(empno,ename) values(1,''abc'') RETURNING empno,ename into :1,:2 ';
execute immediate sql_str RETURNING BULK COLLECT INTO v_tabid, v_tabname;
FOR c IN v_tabid.first..v_tabid.last LOOP
dbms_output.put_line('empno为'||v_tabid(c)||' 记录的NAME为'||v_tabname(c));
END LOOP;
dbms_output.put_line('---------------------------------');
/ 批fetch
语法:
fetch dynamic_cursor
bulk collect into define_variable[,define_variable...]
/
sql_str:='select empno,ename from emp';
sql_str:=sql_str||' order by empno desc';
OPEN v_mycursor FOR sql_str;
--取
FETCH v_mycursor BULK COLLECT INTO v_tabid,v_tabname;
--关
CLOSE v_mycursor;
--输
FOR c IN v_tabid.first..v_tabid.last LOOP
dbms_output.put_line('empno为'||v_tabid(c)||' 记录的NAME为'||v_tabname(c));
END LOOP;
dbms_output.put_line('---------------------------------');
END;
２.－－－－－－－
forall
DECLARE
/批forall
语法动态字符串必须为insert/update/delete,不能为select
forall index in lower..upper
execute immediate dynamic_string
using bind |bind(index)[,bind |bind(index)...]
[{returning|return} bulk collect into bind_argument[,bind_argument...]];
/
TYPE sal_list IS TABLE OF number(8,2);
TYPE name_list IS TABLE OF varchar2(30);
TYPE dept_list IS VARRAY(15) OF integer;
v_depts dept_list:=dept_list(10,20,30,40,50,60,70,80);
v_tabsal sal_list:=sal_list();
v_tabname name_list:=name_list();
sql_str varchar2(200);
BEGIN
sql_str:='update emp set sal=sal:arg1 where DEPTNO=:arg2';
sql_str:=sql_str||' returning ename,sal into :arg3,:arg4';
--给前面4个部门加薪1０％，并返回结果到集合.
FORALL j IN 1..4
execute immediate sql_str
using 1.10,v_depts(j)
RETURNING BULK COLLECT INTO v_tabname,v_tabsal;
--显示结果
FOR j IN v_tabname.first..v_tabname.last LOOP
dbms_output.put_line('雇员'||v_tabname(j)
||'　的薪水被提到'||v_tabsal(j));
END LOOP;
dbms_output.put_line('---------------------------------');
--给后面4个部门加薪２０％，并返回结果到集合.
FORALL j IN 5..8
execute immediate sql_str
using 1.20,v_depts(j)
RETURNING BULK COLLECT INTO v_tabname,v_tabsal;
--显示结果(用notfound判断是否有结果集）
IF SQL%NOTFOUND THEN
dbms_output.put_line('无数据更新');
ELSE
FOR j IN v_tabname.first..v_tabname.last LOOP
dbms_output.put_line('雇员'||v_tabname(j)
||'　的薪水被提到'||v_tabsal(j));
END LOOP;
END IF;
END;
３.用一个值绑定绑定名称相同的值.
把sql语句用begin end括起来就能实现
如:
execute immediate 'begin calc_stats(:x,:x,:y,:x,:y); end;' using a,b;
将Ａ与Ｘ绑定，当第二次出来不同名称时，与Ｂ绑定，以此类推