I’ve recently been reading ploeh’s series on doing the Tennis Kata in F#. I decided to have a go in Erlang, we might not have a fancy type system but we do have pattern matching and partial functions.

You can find the source code to follow along here.

I decided to implement a “game” as a gen_server (when you have a hammer…), I refer you to a previous post for the pre-amble.

Love all

The first test is to check the initial score:

initial_score(Pid) ->
    fun() ->
        ?assertEqual(#{p1=>love,p2=>love}, gen_server:call(Pid, get_score))
    end.

I decided to return a map, rather than a tuple or record, and to use the atom “love” rather than 0. The implementation is pretty trivial:

init([]) ->
    {ok, #{p1=>love,p2=>love}}.

handle_call(get_score, _From, State) ->
    {reply, State, State};

If there was more information in the state map, you might want to return a projection rather than the whole object.

Fifteen love

fifteen_love(Pid) ->
    fun() ->
        ?assertEqual(#{p1=>15,p2=>love}, gen_server:call(Pid, {won_point, p1}))
    end.

This time we need to calculate the new score when a player has won a point:

handle_call({won_point, p1}, _From, State = #{p1:=CurrentScore}) ->
    NewState = State#{p1:=new_score(CurrentScore)},
    {reply, NewState, NewState};

...

new_score(love) -> 15.

Love fifteen

love_fifteen(Pid) ->
    fun() ->
        ?assertEqual(#{p1=>love,p2=>15}, gen_server:call(Pid, {won_point, p2}))
    end.

We can re-use the new_score method from before:

handle_call({won_point, p2}, _From, State = #{p2:=CurrentScore}) ->
    NewState = State#{p2:=new_score(CurrentScore)},
    {reply, NewState, NewState};

Fifteen all

fifteen_all(Pid) ->
    fun() ->
        won_point(Pid, p1),
        ?assertEqual(#{p1=>15,p2=>15}, won_point(Pid, p2))
    end.

won_point(Pid, Player) ->
    gen_server:call(Pid, {won_point, Player}).

No need for any new code, but it seemed like time for a little test refactoring.

Thirty fifteen

thirty_fifteen(Pid) ->
    fun() ->
        won_point(Pid, p1),
        won_point(Pid, p2),
        ?assertEqual(#{p1=>30,p2=>15}, won_point(Pid, p1))
    end.

Now we need to add another function head, to handle the new score:

new_score(love) -> 15;
new_score(15) -> 30.

Thirty all

thirty_all(Pid) ->
    fun() ->
        won_point(Pid, p1),
        won_point(Pid, p2),
        won_point(Pid, p2),
        ?assertEqual(#{p1=>30,p2=>30}, won_point(Pid, p1))
    end.

No new code required.

Forty thirty

forty_thirty(Pid) ->
    fun() ->
        won_point(Pid, p1),
        won_point(Pid, p2),
        won_point(Pid, p1),
        won_point(Pid, p2),
        ?assertEqual(#{p1=>40,p2=>30}, won_point(Pid, p1))
    end.

Again, we need to handle a new score:

...
new_score(30) -> 40.

P1 win

p1_wins(Pid) ->
    fun() ->
        won_point(Pid, p1),
        won_point(Pid, p1),
        won_point(Pid, p1),
        ?assertEqual({game_over, p1}, won_point(Pid, p1))
    end.

This requires a special case, for when the player has 40 points:

handle_call({won_point, p1}, _From, State = #{p1:=40}) ->
    {stop, normal, {game_over, p1}, State};

But now we have a problem, as I skipped over the troublesome “deuce”.

Deuce

deuce(Pid) ->
    fun() ->
        won_point(Pid, p1),
        won_point(Pid, p2),
        won_point(Pid, p1),
        won_point(Pid, p1),
        won_point(Pid, p2),
        ?assertEqual(#{p1=>deuce,p2=>deuce}, won_point(Pid, p2))
    end.

Again, I decided to use an atom, rather than 40-40. But now we need to take both player’s scores into account when determining the new score:

handle_call({won_point, p1}, _From, State = #{p1:=P1Score,p2:=P2Score}) ->
    {NewP1Score, NewP2Score} = new_score(P1Score, P2Score),
    NewState = State#{p1:=NewP1Score, p2:=NewP2Score},
    {reply, NewState, NewState};

...

new_score(love, Score) -> {15, Score};
new_score(15, Score) -> {30, Score};
new_score(30, 40) -> {deuce, deuce};
new_score(30, Score) -> {40, Score}.

Advantage

advantage_p1(Pid) ->
    fun() ->
        won_point(Pid, p1),
        won_point(Pid, p2),
        won_point(Pid, p1),
        won_point(Pid, p1),
        won_point(Pid, p2),
        won_point(Pid, p2),
        ?assertEqual(#{p1=>advantage,p2=>40}, won_point(Pid, p1))
    end.

Another case to handle:

...
new_score(deuce, deuce) -> {advantage, 40}.

Back to Deuce

advantage_p1_back_to_deuce(Pid) ->
    fun() ->
        won_point(Pid, p1),
        won_point(Pid, p2),
        won_point(Pid, p1),
        won_point(Pid, p1),
        won_point(Pid, p2),
        won_point(Pid, p2),
        won_point(Pid, p1),
        ?assertEqual(#{p1=>deuce,p2=>deuce}, won_point(Pid, p2))
    end.

Now we need to add a guard to the win case:

handle_call({won_point, p2}, _From, State = #{p1:=P1Score, p2:=40}) when P1Score =/= advantage ->
    {stop, normal, {game_over, p2}, State};

...

new_score(40, advantage) -> {deuce, deuce}.

Advantage, then win

advantage_p1_then_win(Pid) ->
    fun() ->
        won_point(Pid, p1),
        won_point(Pid, p2),
        won_point(Pid, p1),
        won_point(Pid, p1),
        won_point(Pid, p2),
        won_point(Pid, p2),
        won_point(Pid, p1),
        ?assertEqual({game_over, p1}, won_point(Pid, p1))
    end.

Finally, we need to allow a win from Advantage:

handle_call({won_point, p1}, _From, State = #{p1:=advantage}) ->
    {stop, normal, {game_over, p1}, State};

I think that’s everything covered. As Mark says, “it’s easy enough that it’s fun to play with, but difficult enough that it’s fun to play with”. I personally feel that the use of pattern matching made the intent of the code clearer, but of course YMMV.

Golang has the handy concept of a buffered channel (a similar concept to .NET’s blocking collections). It allows you to throttle your code, by blocking when adding an item to a collection that is already full.

Erlang doesn’t offer anything similar, out of the box (mailboxes will happily increase in size until the VM topples over), but it’s pretty easy to implement using a gen_server. You only need to handle 2 messages, adding an item:

init([Limit]) ->
    State = #{
        limit => Limit,
        count => 0,
        add_q => queue:new(),
        get_q => queue:new()
    },
    {ok, State}.

handle_call(add, From, State = #{count:=Count, limit:=Limit, add_q:=AddQ, get_q:=GetQ}) ->
    UpdatedCount = Count + 1,
    case UpdatedCount > Limit of
        false ->
            case queue:is_empty(GetQ) of
                true ->
                    {reply, ok, State#{count => UpdatedCount}};
                _ ->
                    {{value, Client}, UpdatedQ} = queue:out(GetQ),
                    gen_server:reply(Client, ok),
                    {reply, ok, State#{get_q => UpdatedQ}}
            end;
        _ ->
            {noreply, State#{add_q => queue:in(From, AddQ)}}
    end;

The process is initialised with a limit, the number of items that can be added to the queue before it will block. This collection is FIFO, but another data structure, e.g. a stack, could be used.

When an add message is received, the current size of the collection is checked. If adding another item would breach the limit, then the pid of the caller is added to the wait queue. Otherwise, the number of items is incremented (this version only keeps a count of items, rather than an actual list, but adding that would be trivial).

If any process is blocked waiting to get an item from the collection, then it is unblocked and an item removed immediately. We also need to handle a get message:

handle_call(get, From, State = #{count:=Count, name:=_Name, add_q:=AddQ, get_q:=GetQ}) ->
    case Count > 0 of
        true ->
            case queue:is_empty(AddQ) of
                true ->
                    UpdatedCount = Count - 1,
                    {reply, ok, State#{count => UpdatedCount}};
                _ ->
                    {Client, UpdatedQ} = queue:out(AddQ),
                    gen_server:reply(Client, ok),
                    {reply, ok, State#{add_q => UpdatedQ}}
            end;
        _ ->
            {noreply, State#{get_q => queue:in(From, GetQ)}}
    end;

which is pretty much the opposite, check if the collection is empty. Block if it is, or return an element to the caller.

Once your collection is up & running (and supervised!), using it is as simple as calling any other gen_server:

ok = gen_server:call(queue_1, get, infinity),
do_something(),
ok = gen_server:call(queue_2, add, infinity),

I’ve set the timeout to infinity, but that might not be a good idea for production code.

It would also be pretty simple to back the collection with disk storage, e.g. using a DETS table.