初步实现了书上案例第二，三问的要求，对输出结果有部分偏差，没有实现对已完成任务状态的记录，因此已完成任务输出无论如何都是0，明天会在record中加一个字段进行已完成任务状态的记录
(2) 添加一个名为job_centre:statistics()的统计函数，让它报告队列内、进行中和已完
成任务的状态。
(3) 添加监视工人进程的代码。如果某个工人进程挂了，请确保它所执行的任务被返回到等待完成的任务池里。

-module(gen_server_test).

-export([start_link/0, add_job/2, work_wanted/0, job_done/1,statistics/0]).
-export([init/1, handle_call/3, handle_cast/2, handle_info/2,
  terminate/2, code_change/3]).
-export([test_job_centre/0]).
-define(SERVER, ?MODULE).

%%创建一个记录，记录中包含了一个工作队列，与下一个工作任务编号
-record(state, 
{
  jobs = queue:new(), 
  next_job_number = 1,
  workers = gb_sets:new()
}).

start_link() ->
  gen_server:start_link({local, ?SERVER}, ?MODULE, [], []).
add_job(Fun,WorkerPid) ->
  gen_server:call(?SERVER, {add_job, Fun,WorkerPid}).
work_wanted() ->
  gen_server:call(?SERVER, work_wanted).
job_done(JobNumber) ->
  gen_server:call(?SERVER, {job_done, JobNumber}).
  %%调用该函数统计让它报告队列内、进行中和已完成任务的状态。
statistics() ->
  gen_server:call(?SERVER, statistics).
%%---------------------------------------------------------------------------
init([]) ->
  {ok, #state{}}.

handle_call({add_job, Fun,WorkerPid}, _From, State) ->
  %%向队列中添加元素,队列的任务为执行Fun中的操作
  NewJobs = queue:in({State#state.next_job_number, Fun}, State#state.jobs),
  NewWorkers = gb_sets:add(WorkerPid, State#state.workers),
  % 建立与工人的连接
  link(WorkerPid),
  {reply, State#state.next_job_number,
    State#state{jobs = NewJobs, next_job_number = State#state.next_job_number + 1, workers = NewWorkers}};
handle_call(work_wanted, _From, State) ->
  IsEmpty = queue:is_empty(State#state.jobs),
  case IsEmpty of
    true ->
      {reply, no, State};
    false ->
      {{value, {JobNumber, Fun}}, NewJobs} = queue:out(State#state.jobs),
      {reply, {JobNumber, Fun}, State#state{jobs = NewJobs}}
  end;
handle_call({job_done, _}, _From, State) ->
  {reply, ok, State};
handle_call(statistics, _From, State) ->
  {reply, #{waiting => queue:len(State#state.jobs), in_progress => gb_sets:size(State#state.workers), completed => 0}, State};
handle_call(Request, _From, State) ->
  {reply, {error, {unknown_request, Request}}, State}.

handle_cast(_Msg, State) ->
  {noreply, State}.
handle_info({exit, WorkerPid, _Reason}, State) ->
  %% 如果工人进程崩溃，则从workers集合中移除，并将任务重新添加到队列
  NewWorkers = gb_sets:delete(WorkerPid, State#state.workers),
  %% 查找并重新添加任务
  {ok, Fun} = find_task_for_worker(WorkerPid, State#state.jobs),
  NewJobs = queue:in({State#state.next_job_number - 1, Fun}, State#state.jobs),
  {noreply, State#state{workers = NewWorkers, jobs = NewJobs}};
handle_info(_Info, State) ->
  {noreply, State}.

terminate(_Reason, _State) ->
  ok.

code_change(_OldVsn, State, _Extra) ->
  {ok, State}.


find_task_for_worker(WorkerPid, Jobs) ->
  case gb_sets:is_member(WorkerPid, gb_sets:from_list(Jobs)) of
    true ->
      %% 找到对应的任务
      {value, {_, Fun}} = queue:out(Jobs),
      {ok, Fun};
    false ->
      {error, not_found}
  end.
test_job_centre() ->
  %%创建两个任务
  Job1 = fun() -> io:format("Doing job 1~n") end,
  Job2 = fun() -> io:format("Doing job 2~n") end,
  %%将两个任务放入任务队列中
  JobNumber1 = add_job(Job1,self()),
  io:format("Job number 1 is ~p~n", [JobNumber1]),
  JobNumber2 = add_job(Job2,self()),
  io:format("Job number 2 is ~p~n", [JobNumber2]),
  %%工人领取任务队列任务
  {JobNumber1, Job1} = work_wanted(),
  io:format("Doing job ~p~n", [{JobNumber1,Job1}]),
  {JobNumber2, Job2} = work_wanted(),
  io:format("Doing job ~p~n", [{JobNumber2,Job2}]),
  %%表示任务完成
  job_done(JobNumber1),
  job_done(JobNumber2),
  case work_wanted() of
    no ->
      io:format("No more jobs~n");
    {JobNumber, Job} ->
      io:format("Doing job ~p~n", [{JobNumber,Job}])
  end,
  %%模拟一个工人进程崩溃退出
  exit(normal),

  ok.

程序输出结果
不知道为什么任务队列中设置的io没有输出，只输出了调用函数中的io，同时正在执行的进程数量为1