Erlang--proplists结构解析
proplists 模块适用数据量较少的场景,处理配置文件和函数选项时常用.proplists对内部数据结构是Key-Value键值对形式,第一个元素做key用来查询和删除,如果一个key有多个值就会使用第一次出现的值,其它被忽略.proplists对于Key和Value的约束极其宽松,可以是任意term().甚至可以把{Atom,true}缩写成为Atom.也正是由于这样宽松的数据约束,proplists模块没有更新和追加数据项的方法,需要使用lists:replace/4.Key进行比较使用的是=:=精确等于,会判断类型和值.
5> proplists:get_value(1,[{1,a},{1.0,b},{1,c}]).
a
6> proplists:append_values(1,[{1,a},{1.0,b},{1,c}]).
[a,c]
8>
规范与压缩
上面提到Atom缩写的形式,atom的形式成为压缩格式,{Atom,true}的形式成为规范形式.这是Property的两种形式,其定义是 -type property() :: atom() | tuple().
模块里面有一个property方法专门进行数据规范化.
property({Key, true}) when is_atom(Key) ->
Key;
property(Property) ->
Property.
调用的形式:
16> proplists:property({a}).
{a}
17> proplists:property({a,true}).
a
压缩形式其实就是逐一对proplists内的元素执行property/1,即:[property(P) || P <- List].
10> proplists:compact( [{a, true}, {b, true}, {a, 3}, {c, true}, {a, [4]}]).
[a,b,{a,3},c,{a,[4]}]
计算有压缩就会有展开的函数
unfold([P | Ps]) ->
if is_atom(P) ->
[{P, true} | unfold(Ps)];
true ->
[P | unfold(Ps)]
end;
unfold([]) ->
[].
12> proplists:unfold([foo,bar,test,haha]).
[{foo,true},{bar,true},{test,true},{haha,true}]
13> proplists:unfold([foo,bar,{test,false},haha]).
[{foo,true},{bar,true},{test,false},{haha,true}]
14> proplists:unfold([foo,"zen",bar,{test,false},haha]).
[{foo,true},"zen",{bar,true},{test,false},{haha,true}]
15> proplists:unfold([foo,"zen",23,{test,false},haha]).
[{foo,true},"zen",23,{test,false},{haha,true}]
proplists 相关操作
下面看一下proplists的常规操作,有一些方法还是要注意一下细节的.
append_values (注意上图中少拼写了一个s )将所有key值相同的数据项,value整合在list中.
1> proplists:append_values(a, [{a, [1,2]}, {b, 0}, {a, 3}, {c, -1}, {a, [4]}]).
[1,2,3,4]
2> proplists:append_values(a, [{a, b}, {b, 0}, {a, 3}, {c, -1}, {a, [4]}]).
[b,3,4]
3> proplists:append_values(a, [{a, b}, {b, 0}, {a, 3}, {c, -1}, {a, [[4]]}]).
[b,3,[4]]
1> proplists:append_values(a, [{a, [1,2]}, {"zen", 0}, {a, 3}, {c, -1}, {a, [4]}
]).
[1,2,3,4]
2> proplists:append_values(a, [{a, [1,2]},b]).
[1,2]
3> proplists:append_values(b, [{a, [1,2]},b]).
[true]
4>
delete方法会删除所有等于Key值的数据项:
%% delete(Key, List) -> List
18> proplists:delete(a, [{a, true}, {b, true}, {a, 3}, {c, true}, {a, [4]}]).
[{b,true},{c,true}]
compact将数据进行压缩
10> proplists:compact( [{a, true}, {b, true}, {a, 3}, {c, true}, {a, [4]}]).
[a,b,{a,3},c,{a,[4]}]
get_all_values 获取所有等于Key值的数据项:
8> proplists:get_all_values(a, [{a, [1,2]}, {"zen", 0}, {a, 3}, {c, -1}, {a, [4]
}]).
[[1,2],3,[4]]
9>
get_bool 这个方法还是有点陷阱的,其意图是看Key值第一次出现时的值是true|false.
get_bool(Key, [P | Ps]) ->
if is_atom(P), P =:= Key -> true;
tuple_size(P) >= 1, element(1, P) =:= Key ->
case P of
{_, true} ->
true;
_ ->
%% Don't continue the search!
false
end;
true ->
get_bool(Key, Ps)
end;
get_bool(_Key, []) ->
false.
9> proplists:get_bool(a, [{a, [1,2]}, {"zen", 0}, {a, 3}, {c, -1}, {a, [4]}]).
false
10> proplists:get_bool(a, [{a, [1,2]}, {"zen", 0}, {a, 3},a, {c, -1}, {a, [4]}]).
false
11> proplists:get_bool(a, [a,{a, [1,2]}, {"zen", 0}, {a, 3},a, {c, -1}, {a, [4]}]).
true
12> proplists:get_bool(a, [{a,true},{a, [1,2]}, {"zen", 0}, {a, 3},a, {c, -1}, {a, [4]}]).
true
13> proplists:get_bool(a, [{a,false},{a, [1,2]}, {"zen", 0}, {a, 3},a, {c, -1},{a, [4]}]).
false
14> proplists:get_bool(q, [{a,true},{a, [1,2]}, {"zen", 0}, {a, 3},a, {c, -1}, {a, [4]}]).
false
15> proplists:get_bool(q, ["abc",{a,true},{a, [1,2]}, {"zen", 0}, {a, 3},a, {c,-1}, {a, [4]}]).
false
16> proplists:get_bool("abc", ["abc",{a,true},{a, [1,2]}, {"zen", 0}, {a, 3},a,{c, -1}, {a, [4]}]).
false
17> proplists:get_bool("abc", [{"abc",true},{a,true},{a, [1,2]}, {"zen", 0}, {a,
3},a, {c, -1}, {a, [4]}]).
true
get_keys 获取所有不重复的keys
18> proplists:get_keys([{"abc",true},{a,true},{a, [1,2]}, {"zen", 0}, {a, 3},a,{c, -1}, {a, [4]}]).
["zen",a,c,"abc"]
19> proplists:get_keys([{a,true},{a,true},{a, [1,2]}, {"zen", 0}, {a, 3},a, {c,-1}, {a, [4]}]).
["zen",a,c]
get_value 按Key取值,取得是第一次出现的Value
get_value(Key, [P | Ps], Default) ->
if is_atom(P), P =:= Key ->
true;
tuple_size(P) >= 1, element(1, P) =:= Key ->
case P of
{_, Value} ->
Value;
_ ->
%% Don</code>t continue the search!
Default
end;
true ->
get_value(Key, Ps, Default)
end;
get_value(_Key, [], Default) ->
Default.
3> proplists:get_value([a,b], ["packet",[a,b],"login",22,2,s,f] , "none").
"none"
4> proplists:get_value("login", ["packet",[a,b],"login",22,2,s,f] , "none").
"none"
5> proplists:get_value(login, ["packet",[a,b],"login",22,2,s,f] , "none").
"none"
1> proplists:get_value([a,b], ["packet",{[a,b],bingo},"login",22,2,s,f] , "none").
bingo
2> proplists:get_value(s, ["packet",{[a,b],bingo},"login",22,2,s,f] , "none").
true
3>
look_up 与get_value不同的是这里返回的是{Key,Value}
6> proplists:lookup(a, [{a, b}, {b, 0}, {a, 3}, {c, -1}, {a, [[4]]}]).
{a,b}
7> proplists:lookup(a, [{a,1},{a, b}, {b, 0}, {a, 3}, {c, -1}, {a, [[4]]}]).
{a,1}
8>
lookup_all
8> proplists:lookup_all(a, [{a,1},{a, b}, {b, 0}, {a, 3}, {c, -1}, {a, [[4]]}]).
[{a,1},{a,b},{a,3},{a,[[4]]}]
is_defined 是否存在特定Key值
6> proplists:is_defined(s, ["packet",{s,kill},{[a,b],bingo},"login",22,2,s,f]).
true
7> proplists:is_defined(p, ["packet",{s,kill},{[a,b],bingo},"login",22,2,s,f]).
false
8>
split 按照Key值进行数据分组
9> proplists:split([{c, 2}, {e, 1}, a, {c, 3, 4}, d, {b, 5}, b], [a, b, c]).
{[[a],[{b,5},b],[{c,2},{c,3,4}]],[{e,1},d]}
10> proplists:split([{c, 2}, {c,23},{a,false},{e, 1}, a, {c, 3, 4}, d, {b, 5}, b
], [a, b, c]).
{[[{a,false},a],[{b,5},b],[{c,2},{c,23},{c,3,4}]],[{e,1},d]}
11>
单独一组
下面这几个方法我们放在一起看
expand 做的是把list中的Key替换成对应Value ,注意, 这个方法展开的对象是Property
8> proplists: expand([{foo, [bar, baz]}],[fie, foo, fum]).
[fie,bar,baz,fum]
9> proplists: expand([{foo, [bar, baz]},{fie,ok},{fum,100}],[fie, foo, fum]).
[ok,bar,baz,100]
10> proplists: expand([{foo, [bar, baz]},{fie,[[ok]]},{fum,"100"}],[fie, foo, fum]).
[[ok],bar,baz,49,48,48]
12> proplists: expand([{"fie",23},{1,{1}},{1.0,{29}},{foo, [bar, baz]},{fie,[[ok]]},{fum,"100"}],["fie",1, foo, fum]).
[102,105,101,1,bar,baz,49,48,48]
13> ${.
123
14> [102,105,101].
"fie"
15>
substitute_aliases 将对应的key值替换为别名
1> proplists:substitute_aliases([{zen,"ligaoren"},{0,zero}],[zen,{zen,zen},{abc,zen},{zen,tick},0,{0,1},{23,0}]).
[{"ligaoren",true},{"ligaoren",zen},{abc,zen},{"ligaoren",tick},0,{zero,1},{23,0}]
2>
substitute_negations key值替换,value取反
2> proplists:substitute_negations([{zen,"ligaoren"},{0,zero}],[zen,{zen,zen},{abc,zen},{zen,tick},0,{0,1},{23,0}]).
[{"ligaoren",false},{"ligaoren",true},{abc,zen},{"ligaoren",true},0,zero,{23,0}]
3> proplists:substitute_negations([{zen,"ligaoren"},{0,zero}],[zen,{zen,zen},{abc,zen},{zen,tick},0,{0,true},{23,0}]).
[{"ligaoren",false},{"ligaoren",true},{abc,zen},{"ligaoren",true},0,{zero,false},{23,0}]
4> proplists:substitute_negations([{zen,"ligaoren"},{0,zero}],[zen,{zen,zen},{abc,zen},{zen,tick},0,{0,false},{23,0}]).
[{"ligaoren",false},{"ligaoren",true},{abc,zen},{"ligaoren",true},0,zero,{23,0}]
normalize 整合了substitute_aliases substitute_negations expand
2> proplists:normalize( [a,b,c,d,e,f,g],[{aliases, [{b,b2},{e,email}]} ]).
[a,b2,c,d,email,f,g]
3> proplists:normalize([a,b,c,d,e,f,g],[{aliases, [{b,b2},{e,email}]} ]).
[a,b2,c,d,email,f,g]
4> proplists:normalize([a,b,c,d,e,f,g],[{aliases, [ {negations, [a,f]}]} ]).
[a,b,c,d,e,f,g]
5> proplists:normalize([a,b,c,d,e,f,g],[ {expand, [{d,do_it_by_yourself},{g,1000}]}]).
[a,b,c,do_it_by_yourself,e,f,1000]
应用举例
6> test:module_info(compile).
[{options,[{outdir,"/zen/temp"}]},
{version,"4.8"},
{time,{2012,6,15,2,3,23}},
{source,"/zen/temp/test.erl"}]
7> proplists:get_value(time,test:module_info(compile)).
{2012,6,15,2,3,23}
8>
mochiweb项目解析Header Cookie多处使用了proplist:
parse_form_outer(eof, _, Acc) ->
lists:reverse(Acc);
parse_form_outer({headers, H}, FileHandler, State) ->
{"form-data", H1} = proplists:get_value("content-disposition", H),
Name = proplists:get_value("name", H1),
Filename = proplists:get_value("filename", H1),
case Filename of
undefined ->
fun (Next) ->
parse_form_value(Next, {Name, []}, FileHandler, State)
end;
_ ->
ContentType = proplists:get_value("content-type", H),
Handler = FileHandler(Filename, ContentType),
fun (Next) ->
parse_form_file(Next, {Name, Handler}, FileHandler, State)
end
end.
解析选项一例:
%%% 18> proplists:lookup(loop,[{ip, "127.0.0.1"},{loop, {mochiweb_http, default_body}}]).
%%% {loop,{mochiweb_http,default_body}}
parse_options(Options) ->
{loop, HttpLoop} = proplists:lookup(loop, Options),
Loop = fun (S) ->
?MODULE:loop(S, HttpLoop)
end,
Options1 = [{loop, Loop} | proplists:delete(loop, Options)],
mochilists:set_defaults(?DEFAULTS, Options1).
最后:估计90%的情况下,我们只使用proplists:get_value : )
2012-8-22更新
proplists:get_value的性能要比lists:keyfind差很多,
lists的下面几个方法都是BIF实现:%% Bifs: keymember/3, keysearch/3, keyfind/3
而proplists:get_value是Erlang实现,我觉得这是产生性能差异的根本原因;
下面有一个相关讨论基本上是同样的判断: http://www.ostinelli.net/erlang-listskeyfind-or-proplistsget_value/
-module(pvsl).
-define(LIST_SIZES, [10000, 100000, 1000000]).
-define(RETRIES, 1000).
-compile(export_all).
start() ->
% test for different list sizes
lists:foreach(fun(N) -> test_list(N) end, ?LIST_SIZES).
test_list(ListSize) ->
% generate a list of size ListSize of {Key, Val} entries
KeyList = [{K, K} || K <- lists:seq(1, ListSize)],
% test this list against both functions
lists:foreach(fun(Type) -> get_val(Type, now(), KeyList, ListSize, ?RETRIES) end,
[proplists, lists]).
% test getting values, compute necessary time and output print results
get_val(Type, Start, _KeyList, ListSize, 0) ->
T = timer:now_diff(now(), Start),
io:format("computed ~p random key searches on a ~p-sized list in ~p ms using ~p~n",
[?RETRIES, ListSize, T/1000, Type]);
get_val(proplists, Start, KeyList, ListSize, Tries) ->
proplists:get_value(random:uniform(ListSize), KeyList),
get_val(proplists, Start, KeyList, ListSize, Tries - 1);
get_val(lists, Start, KeyList, ListSize, Tries) ->
lists:keyfind(random:uniform(ListSize), 1, KeyList),
get_val(lists, Start, KeyList, ListSize, Tries - 1).
I ran this test on my MacBook Pro, Intel Core i5 2.4GHz with 4GB Memory, and Erlang R13B04, with Kernel Polling enabled. These are the results.
roberto$ erl +K true +P 1000000
Erlang R13B04 (erts-5.7.5) [source] [smp:4:4] [rq:4] [async-threads:0] [hipe] [kernel-poll:true]
Eshell V5.7.5 (abort with ^G)
1> c(pvsl).
{ok,pvsl}
2> pvsl:start().
computed 1000 random key searches on a 10000-sized list in 323.373 ms using proplists
computed 1000 random key searches on a 10000-sized list in 12.897 ms using lists
computed 1000 random key searches on a 100000-sized list in 3273.973 ms using proplists
computed 1000 random key searches on a 100000-sized list in 130.592 ms using lists
computed 1000 random key searches on a 1000000-sized list in 34131.905 ms using proplists
computed 1000 random key searches on a 1000000-sized list in 2050.627 ms using lists
ok
3>
