gen_statem(3erl)

📦

version 25.3.2.8 · 5 versions

💅 `gen_statem(3erl)` — Generic state machine behavior 💅

okay so bestie, welcome to the manual page for gen_statem(3erl) — the Erlang generic state machine behavior. It’s long, it’s iconic, and it’s literally the upgrade path from gen_fsm. Let’s get into it 💖

🌸 NAME 🌸

gen_statem — Generic state machine behavior.

🦋 DESCRIPTION 🦋

gen_statem provides a generic state machine behaviour that for new code replaces its predecessor gen_fsm since Erlang/OTP 20.0. The gen_fsm behaviour remains in OTP "as is".

Note: If you are new to gen_statem and want an overview of concepts and operation, the section gen_statem Behaviour located in the User's Guide OTP Design Principles is recommended to read before this reference manual, possibly after the Description section you are reading here.

This reference manual contains type descriptions generated from types in the gen_statem source code, so they are correct. However, the generated descriptions also reflect the type hierarchy, which sometimes makes it hard to get a good overview. If so, see the section gen_statem Behaviour in the OTP Design Principles User's Guide.

Note: the lore timeline ✨

This behavior appeared in Erlang/OTP 19.0.
In OTP 19.1 a backwards incompatible change of the return tuple from Module:init/1 was made and the mandatory callback function Module:callback_mode/0 was introduced.
In OTP 20.0 generic time-outs were added.
In OTP 22.1 time-out content update and explicit time-out cancel were added.
In OTP 22.3 the possibility to change the callback module with actions change_callback_module, push_callback_module and pop_callback_module, was added.

gen_statem has got the same features that gen_fsm had and adds some really useful:

Co-located state code
Arbitrary term state
Event postponing
Self-generated events
State time-out
Multiple generic named time-outs
Absolute time-out time
Automatic state enter calls
Reply from other state than the request, sys traceable
Multiple sys traceable replies
Changing the callback module

Two callback modes are supported:

One for finite-state machines (gen_fsm like), which requires the state to be an atom and uses that state as the name of the current callback function.
One that allows the state to be any term and that uses one callback function for all states.

The callback model(s) for gen_statem differs from the one for gen_fsm, but it is still fairly easy to rewrite from gen_fsm to gen_statem.

A generic state machine server process (gen_statem) implemented using this module has a standard set of interface functions and includes functionality for tracing and error reporting. It also fits into an OTP supervision tree. For more information, see OTP Design Principles.

A gen_statem assumes all specific parts to be located in a callback module exporting a predefined set of functions. The relationship between the behavior functions and the callback functions is as follows:

gen_statem module            Callback module
-----------------            ---------------
gen_statem:start
gen_statem:start_monitor
gen_statem:start_link -----> Module:init/1

Server start or code change
                      -----> Module:callback_mode/0

gen_statem:stop       -----> Module:terminate/3

gen_statem:call
gen_statem:cast
gen_statem:send_request
erlang:send
erlang:'!'            -----> Module:StateName/3
                             Module:handle_event/4

-                     -----> Module:terminate/3

-                     -----> Module:code_change/4

Events are of different types, so the callback functions can know the origin of an event and how to respond.

If a callback function fails or returns a bad value, the gen_statem terminates, unless otherwise stated. However, an exception of class throw is not regarded as an error but as a valid return from all callback functions.

The state callback for a specific state in a gen_statem is the callback function that is called for all events in this state. It is selected depending on which callback mode that the callback module defines with the callback function Module:callback_mode/0.

When the callback mode is state_functions, the state must be an atom and is used as the state callback name; see Module:StateName/3. This co-locates all code for a specific state in one function as the gen_statem engine branches depending on state name. Note the fact that the callback function Module:terminate/3 makes the state name terminate unusable in this mode.

When the callback mode is handle_event_function, the state can be any term and the state callback name is Module:handle_event/4. This makes it easy to branch depending on state or event as you desire. Be careful about which events you handle in which states so that you do not accidentally postpone an event forever creating an infinite busy loop.

When gen_statem receives a process message it is converted into an event and the state callback is called with the event as two arguments: type and content. When the state callback has processed the event it returns to gen_statem which does a state transition. If this state transition is to a different state, that is: NextState =/= State, it is a state change.

The state callback may return transition actions for gen_statem to execute during the state transition, for example to reply to a gen_statem:call/2,3.

One of the possible transition actions is to postpone the current event. Then it is not retried in the current state. The gen_statem engine keeps a queue of events divided into the postponed events and the events still to process. After a state change the queue restarts with the postponed events.

The gen_statem event queue model is sufficient to emulate the normal process message queue with selective receive. Postponing an event corresponds to not matching it in a receive statement, and changing states corresponds to entering a new receive statement.

The state callback can insert events using the transition actions next_event and such an event is inserted in the event queue as the next to call the state callback with. That is, as if it is the oldest incoming event. A dedicated event_type() internal can be used for such events making them impossible to mistake for external events.

Inserting an event replaces the trick of calling your own state handling functions that you often would have to resort to in, for example, gen_fsm to force processing an inserted event before others.

The gen_statem engine can automatically make a specialized call to the state callback whenever a new state is entered; see state_enter(). This is for writing code common to all state entries. Another way to do it is to explicitly insert an event at the state transition, and/or to use a dedicated state transition function, but that is something you will have to remember at every state transition to the state(s) that need it.

Note: If you in gen_statem, for example, postpone an event in one state and then call another state callback of yours, you have not done a state change and hence the postponed event is not retried, which is logical but can be confusing. 🫠

For the details of a state transition, see type transition_option().

A gen_statem handles system messages as described in sys. The sys module can be used for debugging a gen_statem.

Notice that a gen_statem does not trap exit signals automatically, this must be explicitly initiated in the callback module (by calling process_flag(trap_exit, true).

Unless otherwise stated, all functions in this module fail if the specified gen_statem does not exist or if bad arguments are specified.

The gen_statem process can go into hibernation; see proc_lib:hibernate/3. It is done when a state callback or Module:init/1 specifies hibernate in the returned Actions list. This feature can be useful to reclaim process heap memory while the server is expected to be idle for a long time. However, use this feature with care, as hibernation can be too costly to use after every event; see erlang:hibernate/3.

There is also a server start option {hibernate_after, Timeout} for start/3,4, start_monitor/3,4, start_link/3,4 or enter_loop/4,5,6, that may be used to automatically hibernate the server.

If the gen_statem process terminates, e.g. as a result of a function in the callback module returning {stop,Reason}, an exit signal with this Reason is sent to linked processes and ports. See Processes in the Reference Manual for details regarding error handling using exit signals.

✨ EXAMPLE ✨

Okay bestie, this one's a whole vibe — it's a pushbutton model for a toggling pushbutton, written with callback mode state_functions. You push the button, it tells you whether it went on or off. You ask it for a count of how many times it's been pushed to switch on, and it tells you. Iconic. 🎀

Here's the complete callback module file pushbutton.erl:

-module(pushbutton).
-behaviour(gen_statem).

-export([start/0,push/0,get_count/0,stop/0]).
-export([terminate/3,code_change/4,init/1,callback_mode/0]).
-export([on/3,off/3]).

name() -> pushbutton_statem. % The registered server name

%% API.  This example uses a registered name name()
%% and does not link to the caller.
start() ->
    gen_statem:start({local,name()}, ?MODULE, [], []).
push() ->
    gen_statem:call(name(), push).
get_count() ->
    gen_statem:call(name(), get_count).
stop() ->
    gen_statem:stop(name()).

%% Mandatory callback functions
terminate(_Reason, _State, _Data) ->
    void.
code_change(_Vsn, State, Data, _Extra) ->
    {ok,State,Data}.
init([]) ->
    %% Set the initial state + data.  Data is used only as a counter.
    State = off, Data = 0,
    {ok,State,Data}.
callback_mode() -> state_functions.

%%% state callback(s)

off({call,From}, push, Data) ->
    %% Go to 'on', increment count and reply
    %% that the resulting status is 'on'
    {next_state,on,Data+1,[{reply,From,on}]};
off(EventType, EventContent, Data) ->
    handle_event(EventType, EventContent, Data).

on({call,From}, push, Data) ->
    %% Go to 'off' and reply that the resulting status is 'off'
    {next_state,off,Data,[{reply,From,off}]};
on(EventType, EventContent, Data) ->
    handle_event(EventType, EventContent, Data).

%% Handle events common to all states
handle_event({call,From}, get_count, Data) ->
    %% Reply with the current count
    {keep_state,Data,[{reply,From,Data}]};
handle_event(_, _, Data) ->
    %% Ignore all other events
    {keep_state,Data}.

And here's the shell session when you actually run it, live and in colour 💅:

1> pushbutton:start().
{ok,<0.36.0>}
2> pushbutton:get_count().
0
3> pushbutton:push().
on
4> pushbutton:get_count().
1
5> pushbutton:push().
off
6> pushbutton:get_count().
1
7> pushbutton:stop().
ok
8> pushbutton:push().
** exception exit: {noproc,{gen_statem,call,[pushbutton_statem,push,infinity]}}
     in function  gen:do_for_proc/2 (gen.erl, line 261)
     in call from gen_statem:call/3 (gen_statem.erl, line 386)

Ngl that last one is giving "texting your ex after the breakup" 💀 — the server's gone, so the call just explodes.

Now, to compare styles, here follows the exact same example but using callback mode handle_event_function — or rather, the code to splice in after function init/1 of the pushbutton.erl example file above. Same energy, different skeleton 🦋:

callback_mode() -> handle_event_function.

%%% state callback(s)

handle_event({call,From}, push, off, Data) ->
    %% Go to 'on', increment count and reply
    %% that the resulting status is 'on'
    {next_state,on,Data+1,[{reply,From,on}]};
handle_event({call,From}, push, on, Data) ->
    %% Go to 'off' and reply that the resulting status is 'off'
    {next_state,off,Data,[{reply,From,off}]};
%%
%% Event handling common to all states
handle_event({call,From}, get_count, State, Data) ->
    %% Reply with the current count
    {next_state,State,Data,[{reply,From,Data}]};
handle_event(_, _, State, Data) ->
    %% Ignore all other events
    {next_state,State,Data}.

🌸 DATA TYPES 🌸

This is the cast of characters, bestie. Every type in the show, with its full spec so nothing gets lost 🫶

server_name() =
    {local, atom()} |
    {global, GlobalName :: term()} |
    {via, RegMod :: module(), Name :: term()}

The name spec you use when starting a gen_statem server. Peep start_link/3 and server_ref() below for the receipts.

server_ref() =
    pid() |
    (LocalName :: atom()) |
    {Name :: atom(), Node :: atom()} |
    {global, GlobalName :: term()} |
    {via, RegMod :: module(), ViaName :: term()}

The server spec you use when addressing a gen_statem server. See call/2 and server_name() above.

It can be:

pid() | LocalName: The gen_statem is locally registered.
{Name,Node}: The gen_statem is locally registered on another node.
{global,GlobalName}: The gen_statem is globally registered in global.
{via,RegMod,ViaName}: The gen_statem is registered in an alternative process registry. The registry callback module RegMod is to export functions register_name/2, unregister_name/1, whereis_name/1, and send/2, which are to behave like the corresponding functions in global. Thus, {via,global,GlobalName} is the same as {global,GlobalName}.

start_opt() =
    {timeout, Time :: timeout()} |
    {spawn_opt, [proc_lib:spawn_option()]} |
    enter_loop_opt()

Options you can use when starting a gen_statem server through, for example, start_link/3.

start_ret() = {ok, pid()} | ignore | {error, term()}

Return value from the start() and start_link() functions — for example, start_link/3.

start_mon_ret() =
    {ok, {pid(), reference()}} | ignore | {error, term()}

Return value from the start_monitor() functions. That reference() is your little monitoring receipt, iykyk ⭐

enter_loop_opt() =
    {hibernate_after, HibernateAfterTimeout :: timeout()} |
    {debug, Dbgs :: [sys:debug_option()]}

Options you can use when starting a gen_statem server through enter_loop/4-6. Two of them, and they're both kinda slay:

hibernate_after: HibernateAfterTimeout specifies that the gen_statem process awaits any message for HibernateAfterTimeout milliseconds and if no message is received, the process goes into hibernation automatically (by calling proc_lib:hibernate/3). Nap time 😴
debug: For every entry in Dbgs, the corresponding function in sys is called. Snitch mode: on 🔍

from() = {To :: pid(), Tag :: reply_tag()}

Destination to use when replying through, for example, the action() {reply,From,Reply} to a process that has called the gen_statem server using call/2.

reply_tag()

A handle that associates a reply to the corresponding request. It's literally the "this is what your message was replying to" sticky note 🎀

state() = state_name() | term()

If the callback mode is handle_event_function, the state can be any term. After a state change (NextState =/= State), all postponed events are retried. So the backlog gets a second chance 👑

state_name() = atom()

If the callback mode is state_functions, the state must be an atom. After a state change (NextState =/= State), all postponed events are retried.

Careful bestie: the state terminate is not possible to use, since it would collide with the optional callback function Module:terminate/3. Name clash, no cap 🙈

data() = term()

A term in which the state machine implementation is to store any server data it needs. The difference between this and the state() itself is that a change in this data does not cause postponed events to be retried. Hence, if a change in this data would change the set of events that are handled, then that data item is to be made a part of the state. Basically: if it changes which events you handle, promote it to state. If not, it can chill in data() 🧚

event_type() =
    external_event_type() | timeout_event_type() | internal

There are 3 categories of events: external, timeout, and internal.

internal events can only be generated by the state machine itself through the transition action next_event.

external_event_type() = {call, From :: from()} | cast | info

External events are of 3 types: {call,From}, cast, or info. Type call originates from the API functions call/2 and send_request/2. For calls, the event contains whom to reply to. Type cast originates from the API function cast/2. Type info originates from regular process messages sent to the gen_statem. Three doorways, one state machine 🚪

timeout_event_type() =
    timeout | {timeout, Name :: term()} | state_timeout

There are 3 types of time-out events that the state machine can generate for itself with the corresponding timeout_action()s. Self-care timers, literally ⏰

event_content() = term()

Any event's content can be any term.

See event_type that describes the origins of the different event types, which is also where the event content comes from. Follow the lore 🔥

callback_mode_result() =
    callback_mode() | [callback_mode() | state_enter()]

This is the return type from Module:callback_mode/0 and selects callback mode and whether to do state enter calls, or not. Two decisions in one return value, multi-tasking icon 💅

callback_mode() = state_functions | handle_event_function

The callback mode is selected with the return value from Module:callback_mode/0:

state_functions: The state must be of type state_name() and one callback function per state, that is, Module:StateName/3, is used.
handle_event_function: The state can be any term and the callback function Module:handle_event/4 is used for all states.

The function Module:callback_mode/0 is called when starting the gen_statem, after code change and after changing the callback module with any of the actions change_callback_module, push_callback_module or pop_callback_module. The result is cached for subsequent calls to state callbacks. So it's not re-asked every single time — it gets remembered, bestie 🧠

state_enter() = state_enter

Whether the state machine should use state enter calls or not is selected when starting the gen_statem and after code change using the return value from Module:callback_mode/0.

If Module:callback_mode/0 returns a list containing state_enter, the gen_statem engine will, at every state change, call the state callback with arguments (enter, OldState, Data) or (enter, OldState, State, Data), depending on the callback mode. This may look like an event but is really a call performed after the previous state callback returned and before any event is delivered to the new state callback. See Module:StateName/3 and Module:handle_event/4. Such a call can be repeated by returning a repeat_state or repeat_state_and_data tuple from the state callback. Looop it 🌀

If Module:callback_mode/0 does not return such a list, no state enter calls are done.

If Module:code_change/4 should transform the state, it is regarded as a state rename and not a state change, which will not cause a state enter call. Important distinction, fr fr 🌈

Okay so before we move on, tiny-but-important note: a state enter call will still happen right before entering the initial state, even though that’s literally not a state change. In this case OldState =:= State, which cannot happen for a later state change, but will happen when repeating the state enter call. 💅

✨ `transition_option()` ✨

       transition_option() =
           postpone() |
           hibernate() |
           event_timeout() |
           generic_timeout() |
           state_timeout()

These transition_option()s can be set by actions and modify the state transition. The state transition takes place when the state callback has processed an event and returns. Here’s the sequence of steps for a state transition, bestie:

All returned actions are processed in order of appearance. In this step all replies generated by any reply_action() are sent. Other actions set transition_option()s that come into play in subsequent steps.
If state enter calls are used, and either it is the initial state or one of the callback results repeat_state_and_data or repeat_state_and_data is used the gen_statem engine calls the current state callback with arguments (enter, State, Data) or (enter, State, State, Data) (depending on callback mode) and when it returns starts again from the top of this sequence.

If state enter calls are used, and the state changes the gen_statem engine calls the new state callback with arguments (enter, OldState, Data) or (enter, OldState, State, Data) (depending on callback mode) and when it returns starts again from the top of this sequence.

If postpone() is true, the current event is postponed.
If this is a state change, the queue of incoming events is reset to start with the oldest postponed.
All events stored with action() next_event are inserted to be processed before previously queued events.
Time-out timers event_timeout(), generic_timeout() and state_timeout() are handled. Time-outs with zero time are guaranteed to be delivered to the state machine before any external not yet received event so if there is such a time-out requested, the corresponding time-out zero event is enqueued as the newest received event; that is after already queued events such as inserted and postponed events.

Any event cancels an event_timeout() so a zero time event time-out is only generated if the event queue is empty.

A state change cancels a state_timeout() and any new transition option of this type belongs to the new state, that is; a state_timeout() applies to the state the state machine enters.

If there are enqueued events the state callback for the possibly new state is called with the oldest enqueued event, and we start again from the top of this sequence.
Otherwise the gen_statem goes into receive or hibernation (if hibernate() is true) to wait for the next message. In hibernation the next non-system event awakens the gen_statem, or rather the next incoming message awakens the gen_statem, but if it is a system event it goes right back into hibernation. When a new message arrives the state callback is called with the corresponding event, and we start again from the top of this sequence.

🌸 `postpone()` 🌸

If true, postpones the current event and retries it after a state change (NextState =/= State).

💤 `hibernate()` 💤

If true, hibernates the gen_statem by calling proc_lib:hibernate/3 before going into receive to wait for a new external event.

Note: If there are enqueued events to process when hibernation is requested, this is optimized by not hibernating but instead calling erlang:garbage_collect/0 to simulate that the gen_statem entered hibernation and immediately got awakened by an enqueued event.

⏰ `event_timeout()` ⏰

event_timeout() = timeout() | integer()

Starts a timer set by enter_action() timeout. When the timer expires an event of event_type() timeout will be generated. See erlang:start_timer/4 for how Time and Options are interpreted. Future erlang:start_timer/4 Options will not necessarily be supported.

Any event that arrives cancels this time-out. Note that a retried or inserted event counts as arrived. So does a state time-out zero event, if it was generated before this time-out is requested.

If Time is infinity, no timer is started, as it never would expire anyway.

If Time is relative and 0 no timer is actually started, instead the time-out event is enqueued to ensure that it gets processed before any not yet received external event, but after already queued events.

Note that it is not possible nor needed to cancel this time-out, as it is cancelled automatically by any other event.

🧚 `generic_timeout()` 🧚

generic_timeout() = timeout() | integer()

Starts a timer set by enter_action() {timeout,Name}. When the timer expires an event of event_type() {timeout,Name} will be generated. See erlang:start_timer/4 for how Time and Options are interpreted. Future erlang:start_timer/4 Options will not necessarily be supported.

If Time is infinity, no timer is started, as it never would expire anyway.

If Time is relative and 0 no timer is actually started, instead the time-out event is enqueued to ensure that it gets processed before any not yet received external event.

Setting a timer with the same Name while it is running will restart it with the new time-out value. Therefore it is possible to cancel a specific time-out by setting it to infinity.

🦋 `state_timeout()` 🦋

state_timeout() = timeout() | integer()

Starts a timer set by enter_action() state_timeout. When the timer expires an event of event_type() state_timeout will be generated. See erlang:start_timer/4 for how Time and Options are interpreted. Future erlang:start_timer/4 Options will not necessarily be supported.

If Time is infinity, no timer is started, as it never would expire anyway.

If Time is relative and 0 no timer is actually started, instead the time-out event is enqueued to ensure that it gets processed before any not yet received external event.

Setting this timer while it is running will restart it with the new time-out value. Therefore it is possible to cancel this time-out by setting it to infinity.

🧊 `timeout_option()` 🧊

timeout_option() = {abs, Abs :: boolean()}

If Abs is true an absolute timer is started, and if it is false a relative, which is the default. See erlang:start_timer/4 for details.

🛠️ `action()` 🛠️

       action() =
           postpone |
           {postpone, Postpone :: postpone()} |
           {next_event,
            EventType :: event_type(),
            EventContent :: event_content()} |
           {change_callback_module, NewModule :: module()} |
           {push_callback_module, NewModule :: module()} |
           pop_callback_module |
           enter_action()

These transition actions can be invoked by returning them from the state callback when it is called with an event, from Module:init/1 or by giving them to enter_loop/5,6.

Actions are executed in the containing list order.

Actions that set transition options override any previous of the same type, so the last in the containing list wins. For example, the last postpone() overrides any previous postpone() in the list.

postpone — Sets the transition_option() postpone() for this state transition. This action is ignored when returned from Module:init/1 or given to enter_loop/5,6, as there is no event to postpone in those cases.
next_event — This action does not set any transition_option() but instead stores the specified EventType and EventContent for insertion after all actions have been executed. The stored events are inserted in the queue as the next to process before any already queued events. The order of these stored events is preserved, so the first next_event in the containing list becomes the first to process. An event of type internal is to be used when you want to reliably distinguish an event inserted this way from any external event.
change_callback_module — Changes the callback module to NewModule which will be used when calling all subsequent state callbacks. The gen_statem engine will find out the callback mode of NewModule by calling NewModule:callback_mode/0 before the next state callback. Changing the callback module does not affect the state transition in any way, it only changes which module that handles the events. Be aware that all relevant callback functions in NewModule such as the state callback, NewModule:code_change/4, NewModule:format_status/1 and NewModule:terminate/3 must be able to handle the state and data from the old module.
push_callback_module — Pushes the current callback module to the top of an internal stack of callback modules and changes the callback module to NewModule. Otherwise like {change_callback_module, NewModule} above.
pop_callback_module — Pops the top module from the internal stack of callback modules and changes the callback module to be the popped module. If the stack is empty the server fails. Otherwise like {change_callback_module, NewModule} above.

🎀 `enter_action()` 🎀

       enter_action() =
           hibernate |
           {hibernate, Hibernate :: hibernate()} |
           timeout_action() |
           reply_action()

These transition actions can be invoked by returning them from the state callback, from Module:init/1 or by giving them to enter_loop/5,6.

Actions are executed in the containing list order.

Think of transition options as ✨trending topics✨ — if two actions in the same list try to set the same one, the later one just overrides the earlier one, no drama. Last one in the containing list wins, periodt. Like, the last event_timeout() in the list fully cancels out any previous event_timeout() there. 💅

hibernate — sets the transition_option() hibernate() for this state transition. Your machine is literally going to take a nap 😴

timeout_action() =
    (Time :: event_timeout()) |
    {timeout,
     Time :: event_timeout(),
     EventContent :: event_content()} |
    {timeout,
     Time :: event_timeout(),
     EventContent :: event_content(),
     Options :: timeout_option() | [timeout_option()]} |
    {{timeout, Name :: term()},
     Time :: generic_timeout(),
     EventContent :: event_content()} |
    {{timeout, Name :: term()},
     Time :: generic_timeout(),
     EventContent :: event_content(),
     Options :: timeout_option() | [timeout_option()]} |
    {state_timeout,
     Time :: state_timeout(),
     EventContent :: event_content()} |
    {state_timeout,
     Time :: state_timeout(),
     EventContent :: event_content(),
     Options :: timeout_option() | [timeout_option()]} |
    timeout_cancel_action() |
    timeout_update_action()

So these transition actions here? You can slide them in by returning them from the state callback, from Module:init/1, or by handing them to enter_loop/5,6. Three entry points, all valid 💖

And heads up: these time-out actions set time-out transition options. That's their whole job.

Time — this is just short for {timeout,Time,Time}, meaning the time-out message is the time-out time. The reason this form even exists is so the state callback return value {next_state,NextState,NewData,Time} is allowed, just like it was in gen_fsm. Legacy vibes, but make it fashion 🎀
timeout — sets the transition_option() event_timeout() to Time with EventContent and time-out options Options.
{timeout,Name} — sets the transition_option() generic_timeout() to Time for Name with EventContent and time-out options Options.
state_timeout — sets the transition_option() state_timeout() to Time with EventContent and time-out options Options.

timeout_cancel_action() =
    {timeout, cancel} |
    {{timeout, Name :: term()}, cancel} |
    {state_timeout, cancel}

This is the shorter, clearer, iconique form of timeout_action() with Time = infinity, and what it does is cancel a time-out. Way cuter than writing out infinity yourself ngl 🦋

timeout_update_action() =
    {timeout, update, EventContent :: event_content()} |
    {{timeout, Name :: term()},
     update,
     EventContent :: event_content()} |
    {state_timeout, update, EventContent :: event_content()}

Updates a time-out with a fresh new EventContent. Peep timeout_action() for how to actually start a time-out, bestie.

And plot twist: if no time-out of the same type is active? Then it instead inserts the time-out event, exactly like when you start a time-out with relative Time = 0. Sneaky but slay ✨

reply_action() = {reply, From :: from(), Reply :: term()}

This transition action can be invoked by returning it from the state callback, from Module:init/1, or by giving it to enter_loop/5,6.

It does not set any transition_option() — instead it replies to a caller that's sitting there waiting for a reply in call/2. And From has to be the term from argument {call,From} in a call to a state callback. Don't freestyle it.

Okay but real talk: using this action from Module:init/1 or enter_loop/5,6 would be weird, on the border of witchcraft, since there has been no earlier call to a state callback in this server. Like who are you even replying to 💀

init_result(StateType) = init_result(StateType, term())

init_result(StateType, DataType) =
    {ok, State :: StateType, Data :: DataType} |
    {ok,
     State :: StateType,
     Data :: DataType,
     Actions :: [action()] | action()} |
    ignore |
    {stop, Reason :: term()}

For a succesful initialization, State is the initial state() and Data the initial server data() of the gen_statem. 🎀

The Actions get executed when entering the first state, just like for a state callback — with one exception: the action postpone is forced to false, because there's no event to postpone. Can't postpone what doesn't exist, bestie 💅

For an unsuccesful initialization, use {stop,Reason} or ignore; see start_link/3,4.

state_enter_result(State) = state_enter_result(State, term())

state_enter_result(State, DataType) =
    {next_state, State, NewData :: DataType} |
    {next_state, State,
     NewData :: DataType,
     Actions :: [enter_action()] | enter_action()} |
    state_callback_result(enter_action())

State is the current state, and it cannot be changed since the state callback was called with a state enter call. It is what it is 🫶

next_state — the gen_statem does a state transition to State, which has to be the current state, sets NewData, and executes all Actions.

event_handler_result(StateType) =
    event_handler_result(StateType, term())

event_handler_result(StateType, DataType) =
    {next_state, NextState :: StateType, NewData :: DataType} |
    {next_state,
     NextState :: StateType,
     NewData :: DataType,
     Actions :: [action()] | action()} |
    state_callback_result(action())

Here StateType is state_name() if callback mode is state_functions, or state() if callback mode is handle_event_function. Pick your fighter 🦋

next_state — the gen_statem does a state transition to NextState (which can totally be the same as the current state), sets NewData, and executes all Actions. Oh and if NextState =/= CurrentState, the state transition is a state change. That's the distinction, iykyk 👑

state_callback_result(ActionType) =
    state_callback_result(ActionType, term())

state_callback_result(ActionType, DataType) =
    {keep_state, NewData :: DataType} |
    {keep_state,
     NewData :: DataType,
     Actions :: [ActionType] | ActionType} |
    keep_state_and_data |
    {keep_state_and_data, Actions :: [ActionType] | ActionType} |
    {repeat_state, NewData :: DataType} |
    {repeat_state,
     NewData :: DataType,
     Actions :: [ActionType] | ActionType} |
    repeat_state_and_data |
    {repeat_state_and_data, Actions :: [ActionType] | ActionType} |
    stop |
    {stop, Reason :: term()} |
    {stop, Reason :: term(), NewData :: DataType} |
    {stop_and_reply,
     Reason :: term(),
     Replies :: [reply_action()] | reply_action()} |
    {stop_and_reply,
     Reason :: term(),
     Replies :: [reply_action()] | reply_action(),
     NewData :: DataType}

ActionType is enter_action() if the state callback was called with a state enter call, and action() if the state callback was called with an event. Depends on the occasion ✨

keep_state — literally the same as {next_state,CurrentState,NewData,Actions}.
keep_state_and_data — literally the same as {keep_state,CurrentData,Actions}.
repeat_state — if the gen_statem runs with state enter calls, the state enter call is repeated, see type transition_option(). Other than that, repeat_state is the same as keep_state. Encore performance 🎤
repeat_state_and_data — literally the same as {repeat_state,CurrentData,Actions}.
stop — terminates the gen_statem by calling Module:terminate/3 with Reason and NewData, if specified. An exit signal with this reason is sent to linked processes and ports. The default Reason is normal. Peace out ✌️
stop_and_reply — sends all Replies, then terminates the gen_statem by calling Module:terminate/3 with Reason and NewData, if specified. An exit signal with this reason is sent to linked processes and ports.

Key detail that's gonna age well: all these terms are tuples or atoms, and this property will hold in any future version of gen_statem. Stability, we love to see it 😭💖

request_id()

An opaque request identifier. See send_request/2 for details. Opaque means don't peek under the hood, bestie 🙈

request_id_collection()

An opaque collection of request identifiers (request_id()) where each request identifier can be associated with a label chosen by you, the user. For more information see reqids_new/0. Cute lil labelled squad 🌸

response_timeout() = timeout() | {abs, integer()}

Used to set a time limit on how long to wait for a response, using either receive_response/2, receive_response/3, wait_response/2, or wait_response/3. The time unit used is millisecond. Currently valid values:

0..4294967295 — timeout relative to current time, in milliseconds.
infinity — infinite timeout. That is, the operation will never time out. Living forever, no cap 💫
{abs, Timeout} — an absolute Erlang monotonic time timeout in milliseconds. That is, the operation will time out when erlang:monotonic_time(millisecond) returns a value larger than or equal to Timeout. Timeout is not allowed to identify a time further into the future than 4294967295 milliseconds. Identifying the timeout using an absolute timeout value is especially handy when you have a deadline for responses corresponding to a complete collection of requests (request_id_collection()), since you do not have to recalculate the relative time until the deadline over and over again. Math? Avoided. Slay 🔥

format_status() =
    #{state => state(),
      data => data(),
      reason => term(),
      queue => [{event_type(), event_content()}],
      postponed => [{event_type(), event_content()}],
      timeouts => [{timeout_event_type(), event_content()}],
      log => [sys:system_event()]}

A map that describes the gen_statem status. The keys are:

state — the current state of the gen_statem process.
data — the state data of the the gen_statem process. (yes "the the", that's canon, don't @ me 🎀)
reason — the reason that caused the state machine to terminate.
queue — the event queue of the gen_statem process.
postponed — the postponed events queue of the gen_statem process.
timeouts — the active time-outs of the gen_statem process.
log — the sys log of the server.

Heads up: new associations may be added to the status map without prior notice. So don't hardcode your pattern matches like it's forever, mwah 💋

💅 EXPORTS 💅

call(ServerRef :: server_ref(), Request :: term()) ->
        Reply :: term()

🦋 `call/3` — the sync call bestie

call(ServerRef :: server_ref(),
     Request :: term(),
     Timeout ::
         timeout() |
         {clean_timeout, T :: timeout()} |
         {dirty_timeout, T :: timeout()}) ->
        Reply :: term()

Okay so this is the synchronous call — you slide a request over to the gen_statem at ServerRef and then you literally sit there waiting until the reply shows up 💅. Behind the scenes the gen_statem hits up the state callback with event_type() {call,From} and the event content is your Request.

A Reply gets born when a state callback comes back with {reply,From,Reply} as one action(), and that Reply is what this function hands back to you 🎀.

Timeout is an integer > 0 saying how many milliseconds to wait for a reply, or the atom infinity to wait forever — and infinity is the default, no cap. If no reply lands inside the time you gave it, the call fails. Periodt.

Note: For Timeout < infinity, and to dodge the situation where you're catching exceptions in the caller and a late reply would end up chilling in your inbox — this function spawns a proxy process to do the actual call. A late reply gets delivered to that now-dead proxy, so it just gets yeeted into the void 🗑️. Ngl this is less efficient than using Timeout == infinity.

Timeout can also be a tuple {clean_timeout,T} or {dirty_timeout,T}, where T is the time-out time. {clean_timeout,T} behaves exactly like a plain T described in the note above and uses a proxy process, while {dirty_timeout,T} bypasses the proxy process, which is more lightweight ✨.

Note: If you combine catching exceptions from this function with {dirty_timeout,T} to stop the calling process from dying when the call times out, you gotta be prepared to handle a late reply, okay? And note that there's an odd chance to get a late reply even with {dirty_timeout,infinity} or infinity, for example in the event of network problems. So like… why not just let the calling process die by not catching the exception? 💀

The call can also fail, for example if the gen_statem dies before or during this function call.

When this call fails it exits the calling process. The exit term is on the form {Reason, Location} where Location = {gen_statem,call,ArgList}. Go see gen_server:call/3 — it has a description of the relevant values for the Reason in the exit term 👀.

🌈 `cast/2` — fire and forget

cast(ServerRef :: server_ref(), Msg :: term()) -> ok

Sends an asynchronous event to the gen_statem at ServerRef and returns ok immediately, ignoring whether the destination node or the gen_statem even exists. Like, genuinely doesn't care 😌. The gen_statem calls the state callback with event_type() cast and event content Msg.

🍓 `check_response/2` — did my reply arrive or nah

check_response(Msg, ReqId) -> Result

Types:

Msg = term()
ReqId = request_id()
Response =
    {reply, Reply :: term()} |
    {error, {Reason :: term(), server_ref()}}
Result = Response | no_reply

Checks if Msg is a response that corresponds to the request identifier ReqId. That request must have been made by send_request/2, bestie. If Msg is a reply to the handle ReqId, the result of the request comes back to you in Reply. Otherwise it returns no_reply and no cleanup is done, so you shall invoke this function repeatedly until a reply is returned 🔁.

The return value Reply is generated when a state callback returns with {reply,From,Reply} as one action(), and that Reply becomes the return value of this function.

The function returns an error if the gen_statem dies before or during this request 💀.

🧚 `check_response/3` — the collection edition

check_response(Msg, ReqIdCollection, Delete) -> Result

Types:

Msg = term()
ReqIdCollection = request_id_collection()
Delete = boolean()
Response =
    {reply, Reply :: term()} |
    {error, {Reason :: term(), server_ref()}}
Result =
    {Response,
     Label :: term(),
     NewReqIdCollection :: request_id_collection()} |
    no_request | no_reply

Checks if Msg is a response corresponding to a request identifier saved in ReqIdCollection 📚. All request identifiers in ReqIdCollection must correspond to requests that have been made using send_request/2 or send_request/4, and all requests must have been made by the process calling this function. No sneaking around from another process, okay?

The Label in the response equals the Label associated with the request identifier that the response corresponds to. A request identifier's Label gets associated when you save the request id in a request identifier collection, or when you send the request using send_request/4.

Compared to check_response/2, the returned result associated with a specific request identifier — or an exception associated with a specific request identifier — will be wrapped in a 3-tuple 🎁. The first element of this tuple equals the value that check_response/2 would have produced, the second element equals the Label associated with that specific request identifier, and the third element NewReqIdCollection is a possibly modified request identifier collection.

If ReqIdCollection is empty, the atom no_request is returned. If Msg doesn't correspond to any of the request identifiers in ReqIdCollection, the atom no_reply is returned.

If Delete equals true, the association with Label will have been deleted from ReqIdCollection in the resulting NewReqIdCollection. If Delete equals false, NewReqIdCollection will equal ReqIdCollection. Note that deleting an association is not for free 🫠 and that a collection containing already handled requests can still be used by subsequent calls to check_response/3, receive_response/3, and wait_response/3. However, without deleting handled associations, those calls won't be able to detect when there are no more outstanding requests to handle — so you'll have to keep track of this some other way than relying on a no_request return. And note: if you pass a collection containing only associations of already handled or abandoned requests to check_response/3, it will always return no_reply. Iconic, honestly.

🌸 `enter_loop/4` — anonymous server, no name needed

enter_loop(Module :: module(),
           Opts :: [enter_loop_opt()],
           State :: state(),
           Data :: data()) ->
              no_return()

The same as enter_loop/6 with Actions = [], except that no server_name() must have been registered. This creates an anonymous server, lowkey mysterious ✨.

👑 `enter_loop/5` — choose your own adventure

enter_loop(Module :: module(),
           Opts :: [enter_loop_opt()],
           State :: state(),
           Data :: data(),
           Server_or_Actions :: server_name() | pid() | [action()]) ->
              no_return()

If Server_or_Actions is a list(), the same as enter_loop/6 except that no server_name() must have been registered and Actions = Server_or_Actions. This creates an anonymous server.

Otherwise, the same as enter_loop/6 with Server = Server_or_Actions and Actions = [].

💜 `enter_loop/6` — become the gen_statem, literally

enter_loop(Module :: module(),
           Opts :: [enter_loop_opt()],
           State :: state(),
           Data :: data(),
           Server :: server_name() | pid(),
           Actions :: [action()] | action()) ->
              no_return()

Makes the calling process become a gen_statem. Main character energy fr fr 👑. It does not return — instead the calling process enters the gen_statem receive loop and becomes a gen_statem server. The process must have been started using one of the start functions in proc_lib. You (yes, you) are responsible for any initialization of the process, including registering a name for it.

This function is useful when you need a more complex initialization procedure than the gen_statem behavior provides.

Module and Opts have the same meaning as when calling start[_link|_monitor]/3,4.

If Server is self() an anonymous server is created just as when using start[_link|_monitor]/3. If Server is a server_name() a named server is created just as when using start[_link|_monitor]/4. However, the server_name() name must have been registered accordingly before this function is called ⚠️.

State, Data, and Actions have the same meanings as in the return value of Module:init/1. Also, the callback module does not need to export a Module:init/1 function 🎉.

The function fails if the calling process wasn't started by a proc_lib start function, or if it isn't registered according to server_name().

⭐ `receive_response/1` — patience of a saint

receive_response(ReqId) -> Result

Types:

ReqId = request_id()
Response =
    {reply, Reply :: term()} |
    {error, {Reason :: term(), server_ref()}}
Result = Response | timeout

The same as calling gen_statem:receive_response(ReqId, infinity). Forever and ever, amen 🙏.

🔥 `receive_response/2` — but with a deadline

receive_response(ReqId, Timeout) -> Result

Types:

ReqId = request_id()
Timeout = response_timeout()
Response =
    {reply, Reply :: term()} |
    {error, {Reason :: term(), server_ref()}}
Result = Response | timeout

Receives a response corresponding to the request identifier ReqId. The request must have been made by send_request/2 to the gen_statem process. This function must be called from the same process from which send_request/2 was made — no outsourcing this 💅.

Timeout specifies how long to wait for a response. If no response is received within the specified time, the function returns timeout. Assuming the server executes on a node supporting aliases (introduced in OTP 24) the request will also be abandoned 🗑️ — that is, no response will be received after a timeout. Otherwise, a stray response might show up at a later time. Awkward.

The return value Reply is generated when a state callback returns with {reply,From,Reply} as one action(), and that Reply becomes the return value of this function.

The function returns an error if the gen_statem dies before or during this function call 💀.

The difference between wait_response/2 and receive_response/2 is that receive_response/2 abandons the request at timeout so that a potential future response is ignored, while wait_response/2 does not. iykyk 🫶

🎀 `receive_response/3`

receive_response(ReqIdCollection, Timeout, Delete) -> Result

Types:

ReqIdCollection = request_id_collection()
Timeout = response_timeout()
Delete = boolean()
Response =
    {reply, Reply :: term()} |
    {error, {Reason :: term(), server_ref()}}
Result =
    {Response,
     Label :: term(),
     NewReqIdCollection :: request_id_collection()} |
    no_request | timeout

okay so bestie, this is the receive_response/3 moment 💅 you're grabbing a response for a request id that’s saved in ReqIdCollection. Every single request id in ReqIdCollection has to correspond to a request made with send_request/2 or send_request/4, and all of those requests must have been made by the process calling this function. no cap, that part matters fr fr.

The Label in the response equals the Label associated with the request id that the response corresponds to. that Label gets associated when you add the request id into a request id collection, or when you send the request with send_request/4 ✨.

Compared to receive_response/2, the returned result for a specific request id gets wrapped in a 3-tuple. the first element equals the value that would’ve come out of receive_response/2, the second element equals the Label linked to that specific request id, and the third element NewReqIdCollection is a possibly modified request id collection 🎀.

If ReqIdCollection is empty, the atom no_request gets returned. simple as that.

Timeout says how long to wait for a response. if no response shows up within the specified time, the function returns timeout. Assuming the server executes on a node supporting aliases (introduced in OTP 24), all requests identified by ReqIdCollection will also be abandoned. that means no responses will be received after a timeout. otherwise, stray responses might be received at a later time 👀.

The difference between receive_response/3 and wait_response/3 is that receive_response/3 abandons the requests at timeout so potential future responses are ignored, while wait_response/3 does not. important vibe check 💛.

If Delete equals true, the association with Label will have been deleted from ReqIdCollection in the resulting NewReqIdCollection. If Delete equals false, NewReqIdCollection will equal ReqIdCollection. note that deleting an association is not for free, and a collection containing already handled requests can still be used by subsequent calls to receive_response/3, check_response/3, and wait_response/3. However, without deleting handled associations, the above calls will not be able to detect when there are no more outstanding requests to handle, so you’ll have to keep track of this some other way than relying on a no_request return. also: if you pass a collection only containing associations of already handled or abandoned requests to receive_response/3, it will always block until a timeout determined by Timeout is triggered 😭.

`reply/1,2`

reply(Replies :: [reply_action()] | reply_action()) -> ok

reply(From :: from(), Reply :: term()) -> ok

This function can be used by a gen_statem to explicitly send a reply to a process that waits in call/2 when the reply cannot be defined in the return value of a state callback. slay ✨

From must be the term from argument {call,From} to the state callback. A reply or multiple replies can also be sent using one or several reply_action()s from a state callback.

Note: a reply sent with this function is not visible in sys debug output. 🙈

`reqids_add/3`

reqids_add(ReqId :: request_id(),
           Label :: term(),
           ReqIdCollection :: request_id_collection()) ->
              NewReqIdCollection :: request_id_collection()

Saves ReqId and associates a Label with the request id by adding this info to ReqIdCollection and returning the resulting request id collection. lowkey useful 🫶

`reqids_new/0`

reqids_new() -> NewReqIdCollection :: request_id_collection()

Returns a new empty request id collection. A request id collection can be utilized in order to handle multiple outstanding requests.

Request ids of requests made by send_request/2 can be saved in a request id collection using reqids_add/3. Such a collection of request ids can later be used to get one response corresponding to a request in the collection by passing the collection as argument to receive_response/3, wait_response/3, or check_response/3.

reqids_size/1 can be used to determine the amount of request ids in a request id collection. 💖

`reqids_size/1`

reqids_size(ReqIdCollection :: request_id_collection()) ->
               integer() >= 0

Returns the amount of request ids saved in ReqIdCollection. math but make it cute 🎀

`reqids_to_list/1`

reqids_to_list(ReqIdCollection :: request_id_collection()) ->
                  [{ReqId :: request_id(), Label :: term()}]

Returns a list of {ReqId, Label} tuples which corresponds to all request ids with their associated labels present in the ReqIdCollection collection. iykyk ✨

`send_request/2`

send_request(ServerRef :: server_ref(), Request :: term()) ->
                ReqId :: request_id()

Sends an asynchronous call request Request to the gen_statem process identified by ServerRef and returns a request id ReqId. The return value ReqId shall later be used with receive_response/2, wait_response/2, or check_response/2 to fetch the actual result of the request. Besides passing the request id directly to these functions, it can also be saved in a request id collection using reqids_add/3. Such a collection of request ids can later be used to get one response corresponding to a request in the collection by passing the collection as argument to receive_response/3, wait_response/3, or check_response/3. If you’re about to save the request id in a request id collection, you may want to consider using send_request/4 instead 👑

The call gen_statem:wait_response(gen_statem:send_request(ServerRef,Request), Timeout) can be seen as equivalent to gen_statem:call(Server,Request,Timeout), ignoring the error handling.

The gen_statem calls the state callback with event_type() {call,From} and event content Request.

A Reply is generated when a state callback returns with {reply,From,Reply} as one action(), and that Reply becomes the return value of receive_response/1,2, wait_response/1,2, or check_response/2 function. periodt 🌈

`send_request/4`

send_request(ServerRef :: server_ref(),
             Request :: term(),
             Label :: term(),
             ReqIdCollection :: request_id_collection()) ->
                NewReqIdCollection :: request_id_collection()

Sends an asynchronous call request Request to the gen_statem process identified by ServerRef. The Label will be associated with the request id of the operation and added to the returned request id collection NewReqIdCollection. The collection can later be used to get one response corresponding to a request in the collection by passing the collection as argument to receive_response/3, wait_response/3, or check_response/3.

The same as calling gen_statem:reqids_add(statem:send_request(ServerRef, Request), Label, ReqIdCollection), but calling send_request/4 is slightly more efficient. efficiency slay 💅

`start/3,4`

start(Module :: module(), Args :: term(), Opts :: [start_opt()]) ->
         start_ret()

start(ServerName :: server_name(),
      Module :: module(),
      Args :: term(),
      Opts :: [start_opt()]) ->
         start_ret()

Creates a standalone gen_statem process according to OTP design principles (using proc_lib primitives). As it does not get linked to the calling process, this start function cannot be used by a supervisor to start a child.

For a description of arguments and return values, see start_link/3,4.

`start_link/3,4`

start_link(Module :: module(),
           Args :: term(),
           Opts :: [start_opt()]) ->
              start_ret()

start_link(ServerName :: server_name(),
           Module :: module(),
           Args :: term(),
           Opts :: [start_opt()]) ->
              start_ret()

Creates a gen_statem process according to OTP design principles (using proc_lib primitives) that is linked to the calling process. This is essential when the gen_statem must be part of a supervision tree so it gets linked to its supervisor. main character energy 👑

The gen_statem process calls Module:init/1 to initialize the server. To ensure a synchronized startup procedure, start_link/3,4 does not return until Module:init/1 has returned.

ServerName specifies the server_name() to register for the gen_statem. If the gen_statem is started with start_link/3, no ServerName is provided and the gen_statem is not registered.

Module is the name of the callback module.

Args is an arbitrary term that is passed as the argument to Module:init/1.

{timeout,Time} — if this option is present in Opts, the gen_statem is allowed to spend Time milliseconds initializing or it terminates and the start function returns {error,timeout}.
{hibernate_after,HibernateAfterTimeout} — if this option is present, the gen_statem process awaits any message for HibernateAfterTimeout milliseconds and if no message is received, the process goes into hibernation automatically (by calling proc_lib:hibernate/3).
{debug,Dbgs} — if this option is present in Opts, debugging through sys is activated.
{spawn_opt,SpawnOpts} — if this option is present in Opts, SpawnOpts is passed as option list to erlang:spawn_opt/2, which is used to spawn the gen_statem process.

Note: using spawn option monitor is not allowed, it causes this function to fail with reason badarg. 💀

If the gen_statem gets created and initialized all good, this function hands you back {ok,Pid} — and Pid is the pid() of your brand new gen_statem bestie. But if a process with the ServerName you picked is already out there living its life, you get {error,{already_started,Pid}}, where Pid is the pid() of that process. Sorry babe, name's taken. 💅

If Module:init/1 fails with Reason, you get {error,Reason} back. And if Module:init/1 returns {stop,Reason} or ignore, the process gets terminated and the function returns {error,Reason} or ignore respectively. Also — heads up — an exit signal with that same Reason (or normal if Module:init/1 returned ignore) gets sent to linked processes and ports, including the process that called start_link/3,4. So yeah. Everyone's getting the news. 😭

start_monitor(Module :: module(),
              Args :: term(),
              Opts :: [start_opt()]) ->
                 start_mon_ret()

start_monitor(ServerName :: server_name(),
              Module :: module(),
              Args :: term(),
              Opts :: [start_opt()]) ->
                 start_mon_ret()

Creates a standalone gen_statem process according to OTP design principles (using proc_lib primitives) and — this is the cute part — atomically sets up a monitor to the newly created process. Because it does not get linked to the calling process, a supervisor cannot use this start function to start a child. It's a solo act, no strings attached. 🦋

The arguments and return values are described in start_link/3,4, so go peek there. But note that the successful return value is different from start_link/3,4: start_monitor/3,4 returns {ok,{Pid,Mon}}, where Pid is the process identifier of the process and Mon is a reference to the monitor set up to keep an eye on it. 👀 And if the start doesn't work out, the caller gets blocked until the DOWN message has been received and removed from the message queue. So you're waiting, bestie, no way around it.

stop(ServerRef :: server_ref()) -> ok

Literally just stop(ServerRef, normal, infinity). Same energy, fewer args. ✨

stop(ServerRef :: server_ref(),
     Reason :: term(),
     Timeout :: timeout()) ->
        ok

Orders the gen_statem ServerRef to exit with the specified Reason and then waits for it to actually terminate. The gen_statem calls Module:terminate/3 before exiting, so it gets its goodbye moment. 🎀

Returns ok if the server terminates with the expected reason. But any reason other than normal, shutdown, or {shutdown,Term} causes an error report to be issued through logger(3erl). 😬 An exit signal with the same reason also gets sent to linked processes and ports. The default Reason is normal.

Timeout is an integer > 0 — it says how many milliseconds to wait for the server to terminate — or the atom infinity to wait forever and ever. Defaults to infinity. If the server doesn't terminate within the specified time, the call exits the calling process with reason timeout.

And if the process doesn't exist at all? The call exits the calling process with reason noproc. If the connection to the remote Node where the server runs fails, you get {nodedown,Node} instead. Fun. 💀

wait_response(ReqId) -> Result

Types:

ReqId = request_id()
Response =
    {reply, Reply :: term()} |
    {error, {Reason :: term(), server_ref()}}
Result = Response | timeout

Exactly the same as calling gen_statem:receive_response(ReqId, infinity). Short and sweet. 🍓

wait_response(ReqId, WaitTime) -> Result

Types:

ReqId = request_id()
WaitTime = response_timeout()
Response =
    {reply, Reply :: term()} |
    {error, {Reason :: term(), server_ref()}}
Result = Response | timeout

Waits for a response corresponding to the request identifier ReqId. The request must have been made by send_request/2 to the gen_statem process, and — important — this function must be called from the same process that made the send_request/2 call. No outsourcing, bestie. 🙈

WaitTime says how long to wait for a reply. If no reply shows up within the specified time, the function returns timeout and no cleanup is done — so you can just call it again and again until a reply finally returns. Iconic persistence. 🔥

The returned Reply is generated when a state callback returns with {reply,From,Reply} as one action(), and that Reply becomes the return value of this function.

The function returns an error if the gen_statem dies before or during this function call. 😭

And the difference between receive_response/2 and wait_response/2? receive_response/2 abandons the request at timeout, so any potential future response gets ignored. wait_response/2 does not. It stays loyal. 🫶

wait_response(ReqIdCollection, WaitTime, Delete) -> Result

Types:

ReqIdCollection = request_id_collection()
WaitTime = response_timeout()
Delete = boolean()
Response =
    {reply, Reply :: term()} |
    {error, {Reason :: term(), server_ref()}}
Result =
    {Response,
     Label :: term(),
     NewReqIdCollection :: request_id_collection()} |
    no_request | timeout

Waits for a response corresponding to a request identifier saved in ReqIdCollection. All the request identifiers in ReqIdCollection must correspond to requests made using send_request/2 or send_request/4, and all of those requests must have been made by the process calling this function. Same rule as before — your request, your call. ✨

The Label in the response equals the Label associated with the request identifier that the response corresponds to. That Label gets associated when you save the request id in a request identifier collection, or when you send the request using send_request/4.

Compared to wait_response/2, the returned result associated with a specific request identifier (or an exception associated with a specific request identifier) gets wrapped in a 3-tuple. First element = the value wait_response/2 would have produced. Second element = the Label associated with that specific request identifier. Third element = NewReqIdCollection, a possibly modified request identifier collection. Little glow-up, three elements instead of one. 💅

If ReqIdCollection is empty, no_request is returned. If no response is received before the WaitTime timeout triggers, you get the atom timeout. And it's totally valid to keep waiting for a response as many times as needed — up until a response has been received and completed by check_response(), receive_response(), or wait_response(). ⭐

The difference between receive_response/3 and wait_response/3 is that receive_response/3 abandons requests at timeout so potential future responses get ignored, while wait_response/3 does not. Say it with me: receive_response gives up, wait_response does not. 📣

Okay so this part matters, ngl: deleting an association is not for free. A collection containing already-handled requests can still be used by subsequent calls to wait_response/3, check_response/3, and receive_response/3. However, without deleting handled associations, those calls won't be able to detect when there are no more outstanding requests to handle — so you'll have to keep track of that some other way instead of relying on a no_request return. And one more: if you pass a collection only containing associations of already-handled or abandoned requests to wait_response/3, it will always block until a timeout determined by WaitTime is triggered and then return no_reply. 💀

💫 CALLBACK FUNCTIONS 💫

The following functions are meant to be exported from a gen_statem callback module. Yes, these are the ones your module actually implements. 👑

✨ EXPORTS ✨

`Module:callback_mode() -> CallbackMode`

Types:

CallbackMode = callback_mode() | [ callback_mode() |
       state_enter() ]

This function gets called by a gen_statem when it needs to find out the callback mode of the callback module. The value is cached by gen_statem for efficiency reasons, so this function is only called once after server start, after code change, and after changing the callback module — but before the first state callback in the current callback module's code version is called. More occasions may be added in future versions of gen_statem. So don't count on it being called a specific number of times, k? 🦋

Server start happens either when Module:init/1 returns or when enter_loop/4-6 is called. Code change happens when Module:code_change/4 returns. A change of the callback module happens when a state callback returns any of the actions change_callback_module, push_callback_module, or pop_callback_module.

The CallbackMode is either just callback_mode() or a list containing callback_mode() and possibly the atom state_enter.

Note: If this function's body does not return an inline constant value, the callback module is doing something strange. 👀 Not me judging, just saying.

`Module:code_change(OldVsn, OldState, OldData, Extra) -> Result`

Types:

OldVsn = Vsn | {down,Vsn}
 Vsn = term()
OldState = NewState = term()
Extra = term()
Result = {ok,NewState,NewData} | Reason
 OldState = NewState = state()
OldData = NewData = data()
Reason = term()

Note: This callback is optional, so callback modules need not export it. But if a release upgrade/downgrade with Change = {advanced,Extra} specified in the .appup file is made when code_change/4 is not implemented, the process will crash with exit reason undef. 😭 So like... maybe implement it. Your call bestie, but that's the vibe.

This function is called by a gen_statem when it is to update its internal state during a release upgrade/downgrade — that is, when the instruction {update,Module,Change,...}, where Change = {advanced,Extra}, is specified in the appup file. For more information, see OTP Design Principles. 📚

For an upgrade, OldVsn is Vsn, and for a downgrade, OldVsn is {down,Vsn}. Vsn is defined by the vsn attribute(s) of the old version of the callback module Module. If no such attribute is defined, the version is the checksum of the Beam file. Slay, checksums. 🔥

Okay so picking up right where we left off, bestie: OldState and OldData are the internal state of the gen_statem. 💅

Extra is passed as is from the {advanced,Extra} part of the update instruction.

If successful, the function must return the updated internal state in an {ok,NewState,NewData} tuple.

If the function returns a failure Reason, the ongoing upgrade fails and rolls back to the old release. Note that Reason cannot be an {ok,_,_} tuple since that will be regarded as a {ok,NewState,NewData} tuple, and that a tuple matching {ok,_} is an also invalid failure Reason. It is recommended to use an atom as Reason since it will be wrapped in an {error,Reason} tuple. No cap, that detail matters. ✨

Also note when upgrading a gen_statem, this function and hence the Change = {advanced,Extra} parameter in the appup file is not only needed to update the internal state or to act on the Extra argument. It is also needed if an upgrade or downgrade should change callback mode, or else the callback mode after the code change will not be honoured, most probably causing a server crash. Like, actually — vibe check your callback mode, bestie. 😭

If the server changes callback module using any of the actions change_callback_module, push_callback_module or pop_callback_module, be aware that it is always the current callback module that will get this callback call. That the current callback module handles the current state and data update should be no surprise, but it must be able to handle even parts of the state and data that it is not familiar with, somehow. Somehow is doing a lot of work there, iykyk. 🫶

In the supervisor child specification there is a list of modules which is recommended to contain only the callback module. For a gen_statem with multiple callback modules there is no real need to list all of them, it may not even be possible since the list could change after code upgrade. If this list would contain only the start callback module, as recommended, what is important is to upgrade that module whenever a synchronized code replacement is done. Then the release handler concludes that an upgrade that upgrades that module needs to suspend, code change, and resume any server whose child specification declares that it is using that module. And again; the current callback module will get the Module:code_change/4 call. Periodt. 👑

🌸 `Module:init(Args) -> Result(StateType)`

Types:

                 Args = term()
                  Result(StateType) = init_result(StateType)

Whenever a gen_statem is started using start_link/3,4, start_monitor/3,4, or start/3,4, this function is called by the new process to initialize the implementation state and server data. ✨

Args is the Args argument provided to that start function.

🌸 Note: Note that if the gen_statem is started through proc_lib and enter_loop/4-6, this callback will never be called. Since this callback is not optional it can in that case be implemented as:

-spec init(_) -> no_return().
init(Args) -> erlang:error(not_implemented, [Args]).

💅 `Module:format_status(Status) -> NewStatus`

Types:

                 Status = format_status()
                 NewStatus = format_status()

🌸 Note: This callback is optional, so a callback module does not need to export it. The gen_statem module provides a default implementation of this function that returns {State,Data}.

If this callback is exported but fails, to hide possibly sensitive data, the default function will instead return {State,Info}, where Info says nothing but the fact that format_status/2 has crashed. Yes bestie, that's the note — we keep it exactly as it is. 💀

This function is called by a gen_statem process when any of the following apply: - sys:get_status/1,2 is invoked to get the gen_statem status. - The gen_statem process terminates abnormally and logs an error.

This function is useful for changing the form and appearance of the gen_statem status for these cases. A callback module wishing to change the sys:get_status/1,2 return value and how its status appears in termination error logs exports an instance of format_status/1, which will get a map Status that describes the current states of the gen_statem, and shall return a map NewStatus containing the same keys as the input map, but it may transform some values.

One use case for this function is to return compact alternative state representations to avoid having large state terms printed in log files. Another is to hide sensitive data from being written to the error log. Sensitive data said not today. 🙈

Example:

format_status(Status) ->
  maps:map(
    fun(state,State) ->
            maps:remove(private_key, State);
       (message,{password, _Pass}) ->
            {password, removed};
       (_,Value) ->
            Value
    end, Status).

🦋 `Module:format_status(Opt, [PDict,State,Data]) -> Status`

Types:

                 Opt = normal | terminate
                 PDict = [{Key, Value}]
                  State = state()
                  Data = data()
                 Key = term()
                 Value = term()
                 Status = term()

⚠️ Warning: This callback is deprecated, in new code use format_status/1. If a format_status/1 callback exists, then this function will never be called.

🌸 Note: This callback is optional, so a callback module does not need to export it. The gen_statem module provides a default implementation of this function that returns {State,Data}.

If this callback is exported but fails, to hide possibly sensitive data, the default function will instead return {State,Info}, where Info says nothing but the fact that format_status/2 has crashed.

This function is called by a gen_statem process when any of the following apply: - One of sys:get_status/1,2 is invoked to get the gen_statem status. Opt is set to the atom normal for this case. - The gen_statem terminates abnormally and logs an error. Opt is set to the atom terminate for this case.

This function is useful for changing the form and appearance of the gen_statem status for these cases. A callback module wishing to change the sys:get_status/1,2 return value and how its status appears in termination error logs exports an instance of format_status/2, which returns a term describing the current status of the gen_statem.

PDict is the current value of the process dictionary of the gen_statem.

State is the internal state of the gen_statem.

Data is the internal server data of the gen_statem.

The function is to return Status, a term that contains the appropriate details of the current state and status of the gen_statem. There are no restrictions on the form Status can take, but for the sys:get_status/1,2 case (when Opt is normal), the recommended form for the Status value is [{data, [{"State", Term}]}], where Term provides relevant details of the gen_statem state. Following this recommendation is not required, but it makes the callback module status consistent with the rest of the sys:get_status/1,2 return value. Consistency is cute. ✨

One use for this function is to return compact alternative state representations to avoid having large state terms printed in log files. Another use is to hide sensitive data from being written to the error log.

👑 `Module:StateName/3` and `Module:handle_event/4` — the state callback girlies

Module:StateName(enter, OldState, Data) -> StateEnterResult(StateName)
Module:StateName(EventType, EventContent, Data) -> StateFunctionResult
Module:handle_event(enter, OldState, State, Data) ->
StateEnterResult(State)
Module:handle_event(EventType, EventContent, State, Data) ->
HandleEventResult

Types:

                  EventType = event_type()
                 EventContent = term()
                  State = state()
                  Data = NewData = data()
                  StateEnterResult(StateName) = state_enter_result(StateName)
                  StateFunctionResult = event_handler_result(state_name())
                  StateEnterResult(State) = state_enter_result(State)
                  HandleEventResult = event_handler_result(state())

Whenever a gen_statem receives an event from call/2, cast/2, or as a normal process message, one of these functions is called. If callback mode is state_functions, Module:StateName/3 is called, and if it is handle_event_function, Module:handle_event/4 is called.

If EventType is {call,From}, the caller waits for a reply. The reply can be sent from this or from any other state callback by returning with {reply,From,Reply} in Actions, in Replies, or by calling reply(From, Reply).

If this function returns with a next state that does not match equal (=/=) to the current state, all postponed events are retried in the next state. Retried in the next state, bestie — not a drill. 🔥

The only difference between StateFunctionResult and HandleEventResult is that for StateFunctionResult the next state must be an atom, but for HandleEventResult there is no restriction on the next state.

For options that can be set and actions that can be done by gen_statem after returning from this function, see action(). 🎀

So yeah — when gen_statem is running with state enter calls, those same functions also get called with arguments (enter, OldState, ...) on every state change. Chill, but there are rules 👮‍♀️ — some actions you're just not allowed to return in that mode:

postpone() is a no-go, because a state enter call isn't an event, so there's literally no event to postpone 🙅‍♀️
{next_event,_,_} is also banned, because state enter calls are not supposed to mess with how events get consumed and produced

You also can't change states from this call. Like, if you return {next_state,NextState, ...} with NextState =/= State, the gen_statem crashes 💀. For real.

And here's a cursed-but-technically-legal one: you are actually allowed to use {repeat_state, NewData, ...}, even though it makes almost no sense — because you'll immediately get called again with a fresh state enter call, which makes it just a weird way of looping, and bestie, there are way better ways to loop in Erlang ✨. But heads up: if you don't update NewData and you have some loop termination condition, or if you use {repeat_state_and_data, _} or repeat_state_and_data, congrats, you've built an infinite loop 😭 no cap. You're advised to use {keep_state,...}, {keep_state_and_data,_} or keep_state_and_data instead, since changing states from a state enter call isn't possible anyway. Slay.

Also, note that you can use throw to return the result, and honestly that's kind of useful 🫶. For example, to bail out with throw(keep_state_and_data) from deep inside complex code that can't return {next_state,State,Data} because State or Data is no longer in scope. Iconic escape hatch behavior.

🎀 `Module:terminate(Reason, State, Data) -> Ignored`

Types, bestie:

Reason = normal | shutdown | {shutdown,term()} | term()
State = state()
Data = data()
Ignored = term()

Note: this callback is optional, so callback modules don't have to export it. The gen_statem module ships a default implementation with no cleanup at all.

This function gets called by a gen_statem when it's about to terminate 🌸. Think of it as the mirror opposite of Module:init/1 — it's your moment to do any necessary cleaning up. When it returns, the gen_statem terminates with Reason. The return value is ignored, so don't stress about it.

Reason is a term denoting the stop reason, and State is the internal state of the gen_statem.

Reason depends on why the gen_statem is terminating:

If it's because another callback function has returned a stop tuple {stop,Reason} in Actions, then Reason has the value specified in that tuple.
If it's because of a failure, Reason is the error reason.

If the gen_statem is part of a supervision tree and its supervisor orders it to terminate, this function is called with Reason = shutdown — but only if both of these apply:

The gen_statem has been set to trap exit signals.
The shutdown strategy as defined in the supervisor's child specification is an integer time-out value, not brutal_kill.

Even if the gen_statem isn't part of a supervision tree, this function is called if it receives an 'EXIT' message from its parent. Reason is the same as in the 'EXIT' message.

Otherwise? The gen_statem is immediately terminated 💅.

Heads up: for any other reason than normal, shutdown, or {shutdown,Term}, the gen_statem is assumed to have terminated because of an error, and an error report is issued using logger(3erl).

And when the gen_statem process exits, an exit signal with the same reason is sent to linked processes and ports. Bye bestie 👋

🌈 SEE ALSO 🌈

That's the whole page, bestie, we made it 🎉💖 This gorgeous thing is brought to you by Ericsson AB, sitting pretty at stdlib 4.3.1.3, and it's the gen_statem(3erl) manual page. Go build something iconic with it 👑✨

📦 where this page comes from

ubuntu noble: erlang-manpages 1:25.3.2.8+dfsg-1ubuntu4 · Apache-2.0
ubuntu noble: erlang-manpages 1:25.3.2.8+dfsg-1ubuntu4.8 · Apache-2.0
💖 girlified by deepseek-ai/DeepSeek-V4.1-Flash-1017 on 2026-11-06: a language model's rewrite of the original, under its terms

gen_statem(3erl)

💅 gen_statem(3erl) — Generic state machine behavior 💅

🌸 NAME 🌸

🦋 DESCRIPTION 🦋

✨ EXAMPLE ✨

🌸 DATA TYPES 🌸

✨ transition_option() ✨

🌸 postpone() 🌸

💤 hibernate() 💤

⏰ event_timeout() ⏰

🧚 generic_timeout() 🧚

🦋 state_timeout() 🦋

🧊 timeout_option() 🧊

🛠️ action() 🛠️

🎀 enter_action() 🎀

💅 EXPORTS 💅

🦋 call/3 — the sync call bestie

🌈 cast/2 — fire and forget

🍓 check_response/2 — did my reply arrive or nah

🧚 check_response/3 — the collection edition

🌸 enter_loop/4 — anonymous server, no name needed

👑 enter_loop/5 — choose your own adventure

💜 enter_loop/6 — become the gen_statem, literally

⭐ receive_response/1 — patience of a saint

🔥 receive_response/2 — but with a deadline

🎀 receive_response/3

reply/1,2

reqids_add/3

reqids_new/0

reqids_size/1

reqids_to_list/1

send_request/2

send_request/4

start/3,4

start_link/3,4

💫 CALLBACK FUNCTIONS 💫

✨ EXPORTS ✨

Module:callback_mode() -> CallbackMode

Module:code_change(OldVsn, OldState, OldData, Extra) -> Result

🌸 Module:init(Args) -> Result(StateType)

💅 Module:format_status(Status) -> NewStatus

🦋 Module:format_status(Opt, [PDict,State,Data]) -> Status

👑 Module:StateName/3 and Module:handle_event/4 — the state callback girlies

🎀 Module:terminate(Reason, State, Data) -> Ignored

🌈 SEE ALSO 🌈

💅 `gen_statem(3erl)` — Generic state machine behavior 💅

✨ `transition_option()` ✨

🌸 `postpone()` 🌸

💤 `hibernate()` 💤

⏰ `event_timeout()` ⏰

🧚 `generic_timeout()` 🧚

🦋 `state_timeout()` 🦋

🧊 `timeout_option()` 🧊

🛠️ `action()` 🛠️

🎀 `enter_action()` 🎀

🦋 `call/3` — the sync call bestie

🌈 `cast/2` — fire and forget

🍓 `check_response/2` — did my reply arrive or nah

🧚 `check_response/3` — the collection edition

🌸 `enter_loop/4` — anonymous server, no name needed

👑 `enter_loop/5` — choose your own adventure

💜 `enter_loop/6` — become the gen_statem, literally

⭐ `receive_response/1` — patience of a saint

🔥 `receive_response/2` — but with a deadline

🎀 `receive_response/3`

`reply/1,2`

`reqids_add/3`

`reqids_new/0`

`reqids_size/1`

`reqids_to_list/1`

`send_request/2`

`send_request/4`

`start/3,4`

`start_link/3,4`

`Module:callback_mode() -> CallbackMode`

`Module:code_change(OldVsn, OldState, OldData, Extra) -> Result`

🌸 `Module:init(Args) -> Result(StateType)`

💅 `Module:format_status(Status) -> NewStatus`

🦋 `Module:format_status(Opt, [PDict,State,Data]) -> Status`

👑 `Module:StateName/3` and `Module:handle_event/4` — the state callback girlies

🎀 `Module:terminate(Reason, State, Data) -> Ignored`