有关于MYSQL回滚的存储过程问题

Question

BEGIN
	DECLARE t_error INTEGER DEFAULT 0;
	DECLARE CONTINUE HANDLER FOR 1062 SET t_error=1;
	set autocommit=0;
	START TRANSACTION;
	call put_playerbehavior(i_playerid,i_gameid,i_channelid,i_acttime,@a);
	IF t_error=1 then
		rollback;
		call  put_playerbehavior(i_playerid,i_gameid,i_channelid,i_acttime,@a);
	END IF;
	COMMIT;
END

begin
 DECLARE v_playerid int;
 DECLARE v_gameid int;
 DECLARE v_channelid int;
 DECLARE done int;
 DECLARE v_acttime datetime;
 DECLARE v_endday datetime;
 DECLARE cur_behavior CURSOR FOR select playerid,gameid,channelid,acttime from playerbehavior where acttime >= i_startday and acttime <  date_add(i_endday,INTERVAL 1 day);
 declare continue handler FOR SQLSTATE '02000' SET done = 1;


 set o_out=0;
 set v_endday=date_add(i_endday,INTERVAL 1 day);


 open cur_behavior;
 repeat
 FETCH  cur_behavior INTO v_playerid,v_gameid,v_channelid,v_acttime;
 CALL test_roll(v_playerid,v_gameid,v_channelid,v_acttime,@a);
 until done end repeat;
 close cur_behavior;
 set o_out=1;
end

第二个存储过程调用第一个存储过程，功能是把playerbehavior表中符合条件的数据用test_roll存储过程进行分析处理，讲结果放到一张表中，但是实际运行过程中，符合条件的最后一条数据会重复运行一次，不知道是什么情况~求大神解惑~分析的存储过程是没错的，就是以上2个存储过程~

Answer 1

谢谢~~那有什么办法能解决么~

Answer 2

FETCH  之后会触发handler进行SET done = 1;

	但FETCH 与 until中间的代码会继续执行，从而引起最后一条数据被重复运行

Answer 3

大神，膜拜一下

Answer 4

rp_1:repeat

    fetch cur** into **;

    if  done=1 then

        leave rp_1;

    end if;

    call procedure(********);

until  done end repeat rp_1: