oracle 批量insert into

看了ITPUB博主的文章也才知道oracle也这种操作，记录一下。。

1.for 循环批量insert

create table T_TABLE_SQL
(
  ids      NUMBER,
  col_name VARCHAR2(20)
);

insert into t_table_sql values(001,'trans');
insert into t_table_sql values(002,'json');

create table t_table_sql_2 as select * from t_table_sql where 1=2;

select * from t_table_sql;

begin
  for i in (select * from t_table_sql) loop
    --一直都以为oracle不支持把"行"变量直接和insert操作关联起来使用,这里"行"变量这个名字是我给起的,不知道是否准确,只要大家能理解就ok.
    insert into t_table_sql_2 values i;   --才晓得,oracle还有这种操作
    commit;
  end loop;
end;

select * from t_table_sql_2;

SQL> select * from t_table_sql_2;

       IDS COL_NAME
---------- --------------------
         1 trans
         2 json

2.游标批量insert

   2.1

truncate table t_table_sql_2;

select * from t_table_sql_2;

declare
cursor cur_data is
select * from t_table_sql;
v_tab t_table_sql%rowtype;
begin
 open cur_data;
  fetch cur_data into v_tab;
  while cur_data%found loop
    insert into t_table_sql_2 values v_tab;
    fetch cur_data into v_tab; 
    end loop;
 close cur_data;
 commit;
  end;

   2.2

truncate table t_table_sql_2;

select * from t_table_sql_2;

declare
cursor cur_data is select * from t_table_sql;
type t_tab_rec is record(
ids t_table_sql.ids%type,
col_name t_table_sql.col_name%type);
t_tab t_tab_rec;
begin
  open cur_data;
  fetch cur_data into t_tab;
  while cur_data%found loop
    insert into t_table_sql_2 values t_tab;
    fetch cur_data into t_tab;
    end loop;
  close cur_data;
  commit;
  end;

转:http://blog.itpub.net/19602/viewspace-1015318/