pl/sql batch update

组里有同事说有一个需要大批量跟新数据库中的功能，批量修改50万数据很慢，于是我改造了一下她的代码，从打开游标的单条修改，改成了批量修改。性能得到了满足。

function F_updatebankdata
(
    i_Bankbatchs          IN          T_bankdatacollect.Bankbatchs%type, --批次号
    o_returnstr           out         varchar2
)
return number--0成功1失败
is
       type detailid_tab is table of varchar2(20);
       type bankcode_tab is table of varchar2(3);
       type bankaccount_tab is table of varchar2(50);
       type bankno_tab is table of varchar2(2);

       detailids                           detailid_tab;
       bankcodes                           bankcode_tab;
       bankaccounts                        bankaccount_tab;
       banknos                             bankno_tab;

       cursor c_T_bankdatadetail is
         select b.Bankdatadetailid,b.bankcode,b.bankaccount,b.bankno
           from T_bankdatadetail b
          where b.bankbatchs = i_Bankbatchs;
begin
    l_effrow := 0;
    open c_T_bankdatadetail;
        loop
          fetch c_T_bankdatadetail 
          bulk collect into detailids,bankcodes,bankaccounts,banknos  limit 10000;
          exit when detailids.count() = 0;
                 l_effrow := l_effrow + detailids.count();
                 forall i in detailids.first .. detailids.last
                      update T_bankcyclebills a
                         set  a.bankdatadetailid = detailids(i)
                       where  a.bankbatchs = i_Bankbatchs
                         and  a.bankcode = bankcodes(i)
                         and  a.bankaccount = bankaccounts(i)
                         and  a.bankno = banknos(i);
                  commit;     
        end loop;
     close c_T_bankdatadetail;
    return 0;
exception
     when others then
       return 1;
end;