简单的 KeyValue 数据库
此源代码基于 map
Erlang 数据结构创建简单的键/值存储服务。首先,我们需要定义有关 gen_server
的所有信息:
-module(cache).
-behaviour(gen_server).
% our API
-export([start_link/0]).
-export([get/1, put/2, state/0, delete/1, stop/0]).
% our handlers
-export([init/1, handle_call/3, handle_cast/2, handle_info/2,
terminate/2, code_change/3]).
% Defining our function to start `cache` process:
start_link() ->
gen_server:start_link({local, ?MODULE}, ?MODULE, [], []).
%%%%%%%%%%%%%%
% API
% Key/Value database is a simple store, value indexed by an unique key.
% This implementation is based on map, this datastructure is like hash
# in Perl or dictionaries in Python.
% put/2
% put a value indexed by a key. We assume the link is stable
% and the data will be written, so, we use an asynchronous call with
% gen_server:cast/2.
put(Key, Value) ->
gen_server:cast(?MODULE, {put, {Key, Value}}).
% get/1
% take one argument, a key and will a return the value indexed
% by this same key. We use a synchronous call with gen_server:call/2.
get(Key) ->
gen_server:call(?MODULE, {get, Key}).
% delete/1
% like `put/1`, we assume the data will be removed. So, we use an
% asynchronous call with gen_server:cast/2.
delete(Key) ->
gen_server:cast(?MODULE, {delete, Key}).
% state/0
% This function will return the current state (here the map who contain all
% indexed values), we need a synchronous call.
state() ->
gen_server:call(?MODULE, {get_state}).
% stop/0
% This function stop cache server process.
stop() ->
gen_server:stop(?MODULE).
%%%%%%%%%%%%%%%
% Handlers
% init/1
% Here init/1 will initialize state with simple empty map datastructure.
init([]) ->
{ok, #{}}.
% handle_call/3
% Now, we need to define our handle. In a cache server we need to get our
% value from a key, this feature need to be synchronous, so, using
% handle_call seems a good choice:
handle_call({get, Key}, From, State) ->
Response = maps:get(Key, State, undefined),
{reply, Response, State};
% We need to check our current state, like get_fea
handle_call({get_state}, From, State) ->
Response = {current_state, State},
{reply, Response, State};
% All other messages will be dropped here.
handle_call(_Request, _From, State) ->
Reply = ok,
{reply, Reply, State}.
% handle_cast/2
% put/2 will execute this function.
handle_cast({put, {Key, Value}}, State) ->
NewState = maps:put(Key, Value, State),
{noreply, NewState};
% delete/1 will execute this function.
handle_cast({delete, Key}, State) ->
NewState = maps:remove(Key, State),
{noreply, NewState};
% All other messages are dropped here.
handle_cast(_Msg, State) ->
{noreply, State}.
%%%%%%%%%%%%%%%%
% other handlers
% We don't need other features, other handlers do nothing.
handle_info(_Info, State) ->
{noreply, State}.
terminate(_Reason, _State) ->
ok.
code_change(_OldVsn, State, _Extra) ->
{ok, State}.
使用我们的缓存服务器
我们现在可以编译我们的代码并开始使用 erl
。
% compile cache
c(cache).
% starting cache server
cache:start_link().
% get current store
% will return:
% #{}
cache:state().
% put some data
cache:put(1, one).
cache:put(hello, bonjour).
cache:put(list, []).
% get current store
% will return:
% #{1 => one,hello => bonjour,list => []}
cache:state().
% delete a value
cache:delete(1).
cache:state().
% #{1 => one,hello => bonjour,list => []}
% stopping cache server
cache:stop().