1、Mnesia提供了如下几个函数遍历所有记录

     mnesia:foldl(Fun, Acc0, Tab) -> NewAcc | transaction abort
     mnesia:foldr(Fun, Acc0, Tab) -> NewAcc | transaction abort
     mnesia:foldl(Fun, Acc0, Tab, LockType) -> NewAcc | transaction abort
     mnesia:foldr(Fun, Acc0, Tab, LockType) -> NewAcc | transaction abort
 这些函数会将Fun遍历应用到表Tab表上，并把结构放入累计集Acc0中，可以按需要指定锁类型。Fun有两个参数，第一个是从表中取出的记录，第二个是累计集。

 例如要查找薪资级别在10以下的员工：

   find_low_salaries() ->
     Constraint = 
          fun(Emp, Acc) when Emp#employee.salary < 10 ->
                 [Emp | Acc];

               (_, Acc) ->
                    Acc
             end,
        Find = fun() -> mnesia:foldl(Constraint, [], employee) end,
        mnesia:transaction(Find)

将薪资上调到10级，返回所有涨薪和：

      increase_low_salaries() ->
         Increase = 
             fun(Emp, Acc) when Emp#employee.salary < 10 ->
                    OldS = Emp#employee.salary,
                    ok = mnesia:write(Emp#employee{salary = 10}),
                    Acc + 10 - OldS;
                (_, Acc) ->
                    Acc
             end,
        IncLow = fun() -> mnesia:foldl(Increase, 0, employee, write) end,
        mnesia:transaction(IncLow).
在遍历中可以做很多事情，但是要特别留意性能和内存消耗。

2、firecat的实践一，查询数据库表的某一列，并求和

get_onlinenum(<<"ALL_DEVICES">>) -> Mnesia迭代求和
    Sum =
	fun(Emp, Acc) when Emp#?ONLINE_TAB.onlinenum /= 0 ->
	V = Emp#?ONLINE_TAB.onlinenum,
        Acc + V;
	(_, Acc) ->
	Acc
	end,
    Find = fun() -> mnesia:foldl(Sum, 0, ?ONLINE_TAB) end,
    mnesia:transaction(Find);

([email protected])1> test:get_onlinenum(<<"ALL_DEVICES">>).
{atomic,2}

实践二，修改数据

clear_onlinenum() ->
    Clear =
	fun(Emp, Acc) when Emp#?ONLINE_TAB.onlinenum /= 0 ->
	   ok = mnesia:write(Emp#?ONLINE_TAB{onlinenum = 0});
	   (_, Acc) ->
	   Acc
	end,
    Find = fun() -> mnesia:foldl(Clear, [], ?ONLINE_TAB) end,
    ret(mnesia:transaction(Find)).

扫描二维码关注公众号，回复： 3441912 查看本文章