Skip to main content
Version: 2.0.x

Ref

Ref[A] models a mutable reference to a value of type A in which we can store immutable data. The two basic operations are set, which fills the Ref with a new value, and get, which retrieves its current content.

Ref provides us a way to functionally manage in-memory state. All operations on Ref are atomic and thread-safe, giving us a reliable foundation for synchronizing concurrent programs.

Ref:

  • is purely functional and referentially transparent
  • is concurrent-safe and lock-free
  • updates and modifies atomically

Concurrent Stateful Application​

Ref is the foundation for writing concurrent stateful applications. Anytime we need to share information between multiple fibers, and those fibers have to update the same information, they need to communicate through something that provides the guarantee of atomicity. Because Ref is concurrent-safe, we can share the same Ref among many fibers. All of which can update Ref concurrently, removing the worry of race conditions. Even if we had ten thousand fibers all updating the same Ref, as long as they are using atomic update and modify functions, we will have zero race conditions.

Operations​

Though Ref has many operations, here we will introduce the most common and important ones.

make​

Ref is never empty, it always contains something. We can create a Ref by providing the initial value to its make method, a constructor of the Ref data type. We should pass an immutable value of type A to the constructor, and it returns an UIO[Ref[A]] value:

def make[A](a: A): UIO[Ref[A]]

As we can see, the output is wrapped inUIO, which means creating a Ref is effectful. Whenever we make, update, or modify the Ref, we are performing an effectful operation.

Let's create some Refs from immutable values:

val counterRef = Ref.make(0)
// counterRef: UIO[Ref[Int]] = Sync(
//   trace = "repl.MdocSession.MdocApp.counterRef(ref.md:14)",
//   eval = zio.Ref$$$Lambda$18316/0x00007f5aa2d54298@4f8445f9
// )
val stringRef = Ref.make("initial") 
// stringRef: UIO[Ref[String]] = Sync(
//   trace = "repl.MdocSession.MdocApp.stringRef(ref.md:17)",
//   eval = zio.Ref$$$Lambda$18316/0x00007f5aa2d54298@7149321b
// ) 

sealed trait State
case object Active  extends State
case object Changed extends State
case object Closed  extends State

val stateRef = Ref.make(Active) 
// stateRef: UIO[Ref[Active.type]] = Sync(
//   trace = "repl.MdocSession.MdocApp.stateRef(ref.md:32)",
//   eval = zio.Ref$$$Lambda$18316/0x00007f5aa2d54298@589084c5
// )

Warning:

A big mistake when creating a Ref is trying to store mutable data inside it. ARef must be used with immutable data. Otherwise, we lose our atomic guarantees, which can lead to collisions and race conditions.

The following snippet compiles, but it leads to race conditions due to a mutable variable being provided to make:

// Compiles but don't work properly
var init = 0
// init: Int = 0
val counterRef = Ref.make(init)
// counterRef: UIO[Ref[Int]] = Sync(
//   trace = "repl.MdocSession.MdocApp.<local MdocApp>.counterRef(ref.md:42)",
//   eval = zio.Ref$$$Lambda$18316/0x00007f5aa2d54298@7c93f645
// )

To correct this, we should change the init to be immutable:

val init = 0
// init: Int = 0
val counterRef = Ref.make(init)
// counterRef: UIO[Ref[Int]] = Sync(
//   trace = "repl.MdocSession.MdocApp.<local MdocApp>.counterRef(ref.md:52)",
//   eval = zio.Ref$$$Lambda$18316/0x00007f5aa2d54298@1f80a2e9
// )

get​

The get method returns the current value of the reference. Its return type is IO[EB, B] in which B is the value type of the effect and in the failure case, EB is the error type of that effect.

def get: IO[EB, B]

As the make and get methods of Ref are effectful, we can chain them together with flatMap. In the following example, we create a Ref with initial value, and then we acquire the current state with the get method:

Ref.make("initial")
   .flatMap(_.get)
   .flatMap(current => Console.printLine(s"current value of ref: $current"))

We can refactor this to use a for-comprehension rather than a series of flatMaps to increase readability:

for {
  ref   <- Ref.make("initial")
  value <- ref.get
} yield assert(value == "initial")

Note that, there is no way to access the shared state outside the monadic operations.

set​

The set method atomically writes a new value to the Ref.

for {
  ref   <- Ref.make("initial")
  _     <- ref.set("update")
  value <- ref.get
} yield assert(value == "update")

update​

With update, we can atomically update the state of Ref with a given pure function, that is, it needs to be deterministic and free of side effects.

def update(f: A => A): IO[E, Unit]

Assume we have a counter, we can increase its value with the update method:

val counterInitial = 0
for {
  counterRef <- Ref.make(counterInitial)
  _          <- counterRef.update(_ + 1)
  value <- counterRef.get
} yield assert(value == 1)
caution

update is not the composition of get and set. This composition is not concurrent-safe. Whenever we need to update our state, we should use the update operation which modifies its Ref atomically.

For example, the following snippet is not concurrent-safe:

// Unsafe State Management
object UnsafeCountRequests extends ZIOAppDefault {

  def request(counter: Ref[Int]) = for {
    current <- counter.get
    _ <- counter.set(current + 1)
  } yield ()

  private val initial = 0
  private val myApp =
    for {
      ref <- Ref.make(initial)
      _ <- request(ref) zipPar request(ref)
      rn <- ref.get
      _ <- Console.printLine(s"total requests performed: $rn")
    } yield ()

  def run = myApp
}

The above snippet doesn't behave deterministically. This program sometimes prints 2 and sometimes prints 1. We can fix it by using update:

// Safe State Management
object CountRequests extends ZIOAppDefault {

  def request(counter: Ref[Int]): ZIO[Any, Nothing, Unit] = {
    for {
      _ <- counter.update(_ + 1)
      reqNumber <- counter.get
      _ <- Console.printLine(s"request number: $reqNumber").orDie
    } yield ()
  }

  private val initial = 0
  private val myApp =
    for {
      ref <- Ref.make(initial)
      _ <- request(ref) zipPar request(ref)
      rn <- ref.get
      _ <- Console.printLine(s"total requests performed: $rn").orDie
    } yield ()

  def run = myApp
}

Here is another use case of update to write a repeat combinator:

def repeat[E, A](n: Int)(io: IO[E, A]): IO[E, Unit] =
  Ref.make(0).flatMap { iRef =>
    def loop: IO[E, Unit] = iRef.get.flatMap { i =>
      if (i < n)
        io *> iRef.update(_ + 1) *> loop
      else
        ZIO.unit
    }
    loop
  }

modify​

modify is a more powerful version of update. It atomically modifies Ref by the given function, and also computes a return value. The function that we pass to modify needs to be a pure function; it needs to be deterministic and free of side effects.

def modify[B](f: A => (B, A)): IO[E, B]

Remember the CountRequest example. What if we want to log the number of each request inside the request function? Let's see what happens if we write that function with the composition of update and get methods:

// Unsafe in Concurrent Environment
def request(counter: Ref[Int]) = {
  for {
    _  <- counter.update(_ + 1)
    rn <- counter.get
    _  <- Console.printLine(s"request number received: $rn")
  } yield ()
}

What happens if, between running update and get, a second update occurs on another fiber? This would not behave deterministically in concurrent environments. So we need a way to perform a combination of get, set, get atomically. This is where modify comes in. Here we will edit request to use modify:

// Safe in Concurrent Environment
def request(counter: Ref[Int]) = {
  for {
    rn <- counter.modify(c => (c + 1, c + 1))
    _  <- Console.printLine(s"request number received: $rn")
  } yield ()
}

AtomicReference in Java​

For Java programmers, we can think of Ref as an AtomicReference. Java has a java.util.concurrent.atomic package which contains AtomicReference, AtomicLong, AtomicBoolean and so forth. Ref has roughly the same power, guarantees, and limitations as AtomicReference, but is higher-level and ZIO-friendly.

Ref vs. State Monad​

Basically Ref allows us to have all the power of State Monad inside ZIO. State Monad lacks two important features that we use in real-life application development:

  1. Concurrency Support
  2. Error Handling

Concurrency​

State Monad is an effect system that only includes state. It allows us to do pure stateful computations. We can only get, set, and update (and related computations) state. State Monad updates its state with series of stateful computations sequentially, but it can't be used to do async or concurrent computations. Ref, in contrast, has great support for concurrent and async programming.

Error Handling​

In most real-life,stateful applications, we will involve some database IO and API calls and/or some concurrent and sync operations which can fail in different ways along the path of execution. So besides state management, we need a way to handle errors. The State Monad doesn't have the ability to model error management. We can combine State Monad and Either Monad with StateT monad transformer, but it imposes massive performance overhead. It doesn't buy us anything that we can't do with Ref. So it is an anti-pattern. In the ZIO model, errors are encoded in effects and Ref utilizes that. So, in addition to state management, we have the ability to handle errors without additional work.

State Transformers​

Those who live on the dark side of mutation sometimes have it easy; they can add state everywhere like it's Christmas. Behold:

var idCounter = 0
def freshVar: String = {
  idCounter += 1
  s"var${idCounter}"
}
val v1 = freshVar
val v2 = freshVar
val v3 = freshVar

As functional programmers, we know better and have captured state mutation in the form of functions of type S => (A, S). Ref provides such an encoding, with S being the type of the value, and modify embodying the state mutation function.

Ref.make(0).flatMap { idCounter =>
  def freshVar: UIO[String] =
    idCounter.modify(cpt => (s"var${cpt + 1}", cpt + 1))

  for {
    v1 <- freshVar
    v2 <- freshVar
    v3 <- freshVar
  } yield ()
}

Building more sophisticated concurrency primitives​

Ref is low-level enough that it can serve as the foundation for other concurrency data types.

For example, semaphores are a classic abstract data type for controlling access to shared resources. They are defined as a triplet S = (v, P, V) where v is the number of units of the resource that are currently available, and P and V are operations that decrement and increment v, respectively. P will only complete when v is non-negative and must wait if it isn't.

With Ref, it's easy to implement such a semaphore! The only difficulty is in P, where we must fail and retry when either v is negative, or its value has changed between the moment we read it and the moment we try to update it. A naive implementation could look like:

sealed trait S {
  def P: UIO[Unit]
  def V: UIO[Unit]
}

object S {
  def apply(v: Long): UIO[S] =
    Ref.make(v).map { vref =>
      new S {
        def V = vref.update(_ + 1).unit

        def P = (vref.get.flatMap { v =>
          if (v < 0)
            ZIO.fail(())
          else
            vref.modify(v0 => if (v0 == v) (true, v - 1) else (false, v)).flatMap {
              case false => ZIO.fail(())
              case true  => ZIO.unit
            }
        } <> P).unit
      }
    }
}

Let's rock these crocodile boots we found the other day at the market and test our semaphore at the night club, yee-haw:

import zio.Console._

val party = for {
  dancefloor <- S(10)
  dancers <- ZIO.foreachPar(1 to 100) { i =>
    dancefloor.P *> Random.nextDouble.map(d => Duration.fromNanos((d * 1000000).round)).flatMap { d =>
      printLine(s"${i} checking my boots") *> ZIO.sleep(d) *> printLine(s"${i} dancing like it's 99")
    } *> dancefloor.V
  }
} yield ()

It goes without saying you should take a look at ZIO's own Semaphore, it does all this and more without wasting all those CPU cycles while waiting.