CatsConcurrentForTask

Alias for productR.

Alias for productL.

Alias for ap.

Given a type A, create a concrete Semigroup[F[A]].

Example:

scala> import cats.implicits._
scala> val s: Semigroup[List[Int]] = SemigroupK[List].algebra[Int]

Replaces the A value in F[A] with the supplied value.

Example:

scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForList

scala> Functor[List].as(List(1,2,3), "hello")
res0: List[String] = List(hello, hello, hello)

Similar to attempt, but it only handles errors of type EE.

Similar to attempt, but wraps the result in a cats.data.EitherT for convenience.

Reifies the value or error of the source and performs an effect on the result, then recovers the original value or error back into F.

Note that if the effect returned by f fails, the resulting effect will fail too.

Alias for fa.attempt.flatTap(f).rethrow for convenience.

Example:

scala> import cats.implicits._
scala> import scala.util.{Try, Success, Failure}

scala> def checkError(result: Either[Throwable, Int]): Try[String] = result.fold(_ => Failure(new java.lang.Exception), _ => Success("success"))

scala> val a: Try[Int] = Failure(new Throwable("failed"))
scala> a.attemptTap(checkError)
res0: scala.util.Try[Int] = Failure(java.lang.Exception)

scala> val b: Try[Int] = Success(1)
scala> b.attemptTap(checkError)
res1: scala.util.Try[Int] = Success(1)

Returns a resource that will start execution of the effect in the background.

In case the resource is closed while the effect is still running (e.g. due to a failure in use), the background action will be canceled.

A basic example with IO:

 val longProcess = (IO.sleep(5.seconds) *> IO(println("Ping!"))).foreverM

 val srv: Resource[IO, ServerBinding[IO]] = for {
   _ <- longProcess.background
   server <- server.run
 } yield server

 val application = srv.use(binding => IO(println("Bound to " + binding)) *> IO.never)

Here, we are starting a background process as part of the application's startup. Afterwards, we initialize a server. Then, we use that server forever using IO.never. This will ensure we never close the server resource unless somebody cancels the whole application action.

If at some point of using the resource you want to wait for the result of the background action, you can do so by sequencing the value inside the resource (it's equivalent to join on Fiber).

This will start the background process, run another action, and wait for the result of the background process:

 longProcess.background.use(await => anotherProcess *> await)

In case the result of such an action is canceled, both processes will receive cancelation signals. The same result can be achieved by using anotherProcess &> longProcess with the Parallel type class syntax.

Evaluates the specified block, catching exceptions of the specified type. Uncaught exceptions are propagated.

Similar to combineK but uses Eval to allow for laziness in the second argument. This can allow for "short-circuiting" of computations.

NOTE: the default implementation of combineKEval does not short-circuit computations. For data structures that can benefit from laziness, SemigroupK instances should override this method.

In the following example, x.combineK(bomb) would result in an error, but combineKEval "short-circuits" the computation. x is Some and thus the result of bomb doesn't even need to be evaluated in order to determine that the result of combineKEval should be x.

scala> import cats.{Eval, Later}
scala> import cats.implicits._
scala> val bomb: Eval[Option[Int]] = Later(sys.error("boom"))
scala> val x: Option[Int] = Some(42)
scala> x.combineKEval(bomb).value
res0: Option[Int] = Some(42)

Compose an Applicative[F] and an Applicative[G] into an Applicative[λ[α => F[G[α]]]].

Example:

scala> import cats.implicits._

scala> val alo = Applicative[List].compose[Option]

scala> alo.pure(3)
res0: List[Option[Int]] = List(Some(3))

scala> alo.product(List(None, Some(true), Some(false)), List(Some(2), None))
res1: List[Option[(Boolean, Int)]] = List(None, None, Some((true,2)), None, Some((false,2)), None)

Compose Invariant F[_] and G[_] then produce Invariant[F[G[_]]] using their imap.

Example:

scala> import cats.implicits._
scala> import scala.concurrent.duration._

scala> val durSemigroupList: Semigroup[List[FiniteDuration]] =
    | Invariant[Semigroup].compose[List].imap(Semigroup[List[Long]])(Duration.fromNanos)(_.toNanos)
scala> durSemigroupList.combine(List(2.seconds, 3.seconds), List(4.seconds))
res1: List[FiniteDuration] = List(2 seconds, 3 seconds, 4 seconds)

Compose an Apply[F] and an Apply[G] into an Apply[λ[α => F[G[α]]]].

Example:

scala> import cats.implicits._

scala> val alo = Apply[List].compose[Option]

scala> alo.product(List(None, Some(true), Some(false)), List(Some(2), None))
res1: List[Option[(Boolean, Int)]] = List(None, None, Some((true,2)), None, Some((false,2)), None)

The composition of two Bifunctors is itself a Bifunctor

"Compose" with a G[_] type to form a SemigroupK for λ[α => F[G[α]]]. Note that this universally works for any G, because the "inner" structure isn't considered when combining two instances.

Example:

scala> import cats.implicits._
scala> type ListOption[A] = List[Option[A]]
scala> val s: SemigroupK[ListOption] = SemigroupK[List].compose[Option]
scala> s.combineK(List(Some(1), None, Some(2)), List(Some(3), None))
res0: List[Option[Int]] = List(Some(1), None, Some(2), Some(3), None)

Compose an Applicative[F] and a ContravariantMonoidal[G] into a ContravariantMonoidal[λ[α => F[G[α]]]].

Example:

scala> import cats.kernel.Comparison
scala> import cats.implicits._

// compares strings by alphabetical order
scala> val alpha: Order[String] = Order[String]

// compares strings by their length
scala> val strLength: Order[String] = Order.by[String, Int](_.length)

scala> val stringOrders: List[Order[String]] = List(alpha, strLength)

// first comparison is with alpha order, second is with string length
scala> stringOrders.map(o => o.comparison("abc", "de"))
res0: List[Comparison] = List(LessThan, GreaterThan)

scala> val le = Applicative[List].composeContravariantMonoidal[Order]

// create Int orders that convert ints to strings and then use the string orders
scala> val intOrders: List[Order[Int]] = le.contramap(stringOrders)(_.toString)

// first comparison is with alpha order, second is with string length
scala> intOrders.map(o => o.comparison(12, 3))
res1: List[Comparison] = List(LessThan, GreaterThan)

// create the `product` of the string order list and the int order list
// `p` contains a list of the following orders:
// 1. (alpha comparison on strings followed by alpha comparison on ints)
// 2. (alpha comparison on strings followed by length comparison on ints)
// 3. (length comparison on strings followed by alpha comparison on ints)
// 4. (length comparison on strings followed by length comparison on ints)
scala> val p: List[Order[(String, Int)]] = le.product(stringOrders, intOrders)

scala> p.map(o => o.comparison(("abc", 12), ("def", 3)))
res2: List[Comparison] = List(LessThan, LessThan, LessThan, GreaterThan)

Compose Invariant F[_] and Functor G[_] then produce Invariant[F[G[_]]] using F's imap and G's map.

Example:

scala> import cats.implicits._
scala> import scala.concurrent.duration._

scala> val durSemigroupList: Semigroup[List[FiniteDuration]] =
    | Invariant[Semigroup]
    |   .composeFunctor[List]
    |   .imap(Semigroup[List[Long]])(Duration.fromNanos)(_.toNanos)
scala> durSemigroupList.combine(List(2.seconds, 3.seconds), List(4.seconds))
res1: List[FiniteDuration] = List(2 seconds, 3 seconds, 4 seconds)

If no interruption happens during the execution of this method, it behaves like .attempt.flatMap.

Unlike .attempt.flatMap however, in the presence of interruption this method offers the continual guarantee: fa is interruptible, but if it completes execution, the effects of f are guaranteed to execute. This does not hold for attempt.flatMap since interruption can happen in between flatMap steps.

The typical use case for this function arises in the implementation of concurrent abstractions, where you have asynchronous operations waiting on some condition (which have to be interruptible), mixed with operations that modify some shared state if the condition holds true (which need to be guaranteed to happen or the state will be inconsistent).

Note that for the use case above:

• We cannot use:

waitingOp.bracket(..., modifyOp)

because it makes waitingOp uninterruptible.

• We cannot use

waitingOp.guaranteeCase {
  case Success => modifyOp(???)
  ...

if we need to use the result of waitingOp.

• We cannot use

waitingOp.attempt.flatMap(modifyOp)

because it could be interrupted after waitingOp is done, but before modifyOp executes.

To access this implementation as a standalone function, you can use Concurrent.continual in the companion object.

Alias for suspend that suspends the evaluation of an F reference and implements cats.Defer typeclass.

Turns a successful value into an error if it does not satisfy a given predicate.

Turns a successful value into an error specified by the error function if it does not satisfy a given predicate.

Defer instances, like functions, parsers, generators, IO, etc... often are used in recursive settings where this function is useful

fix(fn) == fn(fix(fn))

example:

val parser: P[Int] = Defer[P].fix[Int] { rec => CharsIn("0123456789") | P("(") ~ rec ~ P(")") }

Note, fn may not yield a terminating value in which case both of the above F[A] run forever.

Apply a monadic function and discard the result while keeping the effect.

scala> import cats._, implicits._
scala> Option(1).flatTap(_ => None)
res0: Option[Int] = None
scala> Option(1).flatTap(_ => Some("123"))
res1: Option[Int] = Some(1)
scala> def nCats(n: Int) = List.fill(n)("cat")
nCats: (n: Int)List[String]
scala> List[Int](0).flatTap(nCats)
res2: List[Int] = List()
scala> List[Int](4).flatTap(nCats)
res3: List[Int] = List(4, 4, 4, 4)

Alias for map, since map can't be injected as syntax if the implementing type already had a built-in .map method.

Example:

scala> import cats.implicits._

scala> val m: Map[Int, String] = Map(1 -> "hi", 2 -> "there", 3 -> "you")

scala> m.fmap(_ ++ "!")
res0: Map[Int,String] = Map(1 -> hi!, 2 -> there!, 3 -> you!)

Like an infinite loop of >> calls. This is most useful effect loops that you want to run forever in for instance a server.

This will be an infinite loop, or it will return an F[Nothing].

Be careful using this. For instance, a List of length k will produce a list of length k^n at iteration n. This means if k = 0, we return an empty list, if k = 1, we loop forever allocating single element lists, but if we have a k > 1, we will allocate exponentially increasing memory and very quickly OOM.

Tuple the values in fa with the result of applying a function with the value

Example:

scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForOption

scala> Functor[Option].fproduct(Option(42))(_.toString)
res0: Option[(Int, String)] = Some((42,42))

Pair the result of function application with A.

Example:

scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForOption

scala> Functor[Option].fproductLeft(Option(42))(_.toString)
res0: Option[(String, Int)] = Some((42,42))

Convert from scala.Either

Example:

scala> import cats.ApplicativeError
scala> import cats.instances.option._

scala> ApplicativeError[Option, Unit].fromEither(Right(1))
res0: scala.Option[Int] = Some(1)

scala> ApplicativeError[Option, Unit].fromEither(Left(()))
res1: scala.Option[Nothing] = None

Convert from scala.Option

Example:

scala> import cats.implicits._
scala> import cats.ApplicativeError
scala> val F = ApplicativeError[Either[String, *], String]

scala> F.fromOption(Some(1), "Empty")
res0: scala.Either[String, Int] = Right(1)

scala> F.fromOption(Option.empty[Int], "Empty")
res1: scala.Either[String, Int] = Left(Empty)

Convert from cats.data.Validated

Example:

scala> import cats.implicits._
scala> import cats.ApplicativeError

scala> ApplicativeError[Option, Unit].fromValidated(1.valid[Unit])
res0: scala.Option[Int] = Some(1)

scala> ApplicativeError[Option, Unit].fromValidated(().invalid[Int])
res1: scala.Option[Int] = None

An if-then-else lifted into the F context. This function combines the effects of the fcond condition and of the two branches, in the order in which they are given.

The value of the result is, depending on the value of the condition, the value of the first argument, or the value of the second argument.

Example:

scala> import cats.implicits._

scala> val b1: Option[Boolean] = Some(true)
scala> val asInt1: Option[Int] = Apply[Option].ifA(b1)(Some(1), Some(0))
scala> asInt1.get
res0: Int = 1

scala> val b2: Option[Boolean] = Some(false)
scala> val asInt2: Option[Int] = Apply[Option].ifA(b2)(Some(1), Some(0))
scala> asInt2.get
res1: Int = 0

scala> val b3: Option[Boolean] = Some(true)
scala> val asInt3: Option[Int] = Apply[Option].ifA(b3)(Some(1), None)
asInt2: Option[Int] = None

Simulates an if/else-if/else in the context of an F. It evaluates conditions until one evaluates to true, and returns the associated F[A]. If no condition is true, returns els.

scala> import cats._
scala> Monad[Eval].ifElseM(Eval.later(false) -> Eval.later(1), Eval.later(true) -> Eval.later(2))(Eval.later(5)).value
res0: Int = 2

Based on a gist by Daniel Spiewak with a stack-safe implementation due to P. Oscar Boykin

Lifts if to Functor

Example:

scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForList

scala> Functor[List].ifF(List(true, false, false))(1, 0)
res0: List[Int] = List(1, 0, 0)

if lifted into monad.

iterateForeverM is almost exclusively useful for effect types. For instance, A may be some state, we may take the current state, run some effect to get a new state and repeat.

Execute an action repeatedly until its result satisfies the given predicate and return that result, discarding all others.

Apply a monadic function iteratively until its result satisfies the given predicate and return that result.

Execute an action repeatedly until its result fails to satisfy the given predicate and return that result, discarding all others.

Apply a monadic function iteratively until its result fails to satisfy the given predicate and return that result.

apply a function to the "left" functor

Widens A into a supertype AA. Example:

scala> import cats.implicits._
scala> sealed trait Foo
scala> case object Bar extends Foo
scala> val x1: Either[Bar.type, Int] = Either.left(Bar)
scala> val x2: Either[Foo, Int] = x1.leftWiden

Lift a function f to operate on Functors

Example:

scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForOption

scala> val o = Option(42)
scala> Functor[Option].lift((x: Int) => x + 10)(o)
res0: Option[Int] = Some(52)

Inherited from LiftIO, defines a conversion from IO in terms of the Concurrent type class.

N.B. expressing this conversion in terms of Concurrent and its capabilities means that the resulting F is cancelable in case the source IO is.

To access this implementation as a standalone function, you can use Concurrent.liftIO (on the object companion).

Similar to map2 but uses Eval to allow for laziness in the F[B] argument. This can allow for "short-circuiting" of computations.

NOTE: the default implementation of map2Eval does not short-circuit computations. For data structures that can benefit from laziness, Apply instances should override this method.

In the following example, x.map2(bomb)(_ + _) would result in an error, but map2Eval "short-circuits" the computation. x is None and thus the result of bomb doesn't even need to be evaluated in order to determine that the result of map2Eval should be None.

scala> import cats.{Eval, Later}
scala> import cats.implicits._
scala> val bomb: Eval[Option[Int]] = Later(sys.error("boom"))
scala> val x: Option[Int] = None
scala> x.map2Eval(bomb)(_ + _).value
res0: Option[Int] = None

Pair A with the result of function application.

Example:

scala> import cats.implicits._
scala> List("12", "34", "56").mproduct(_.toList)
res0: List[(String, Char)] = List((12,1), (12,2), (34,3), (34,4), (56,5), (56,6))

Returns a non-terminating F[_], that never completes with a result, being equivalent to async(_ => ())

Executes the given finalizer when the source is canceled.

The typical use case for this function arises in the implementation of concurrent abstractions, which generally consist of operations that perform asynchronous waiting after concurrently modifying some state: in case the user asks for cancelation, we want to interrupt the waiting operation, and restore the state to its previous value.

waitingOp.onCancel(restoreState)

A direct use of bracket is not a good fit for this case as it would make the waiting action uncancelable.

NOTE: This function handles interruption only, you need to take care of the success and error case elsewhere in your code

Execute a callback on certain errors, then rethrow them. Any non matching error is rethrown as well.

In the following example, only one of the errors is logged, but they are both rethrown, to be possibly handled by another layer of the program:

scala> import cats._, data._, implicits._

scala> case class Err(msg: String)

scala> type F[A] = EitherT[State[String, *], Err, A]

scala> val action: PartialFunction[Err, F[Unit]] = {
    |   case Err("one") => EitherT.liftF(State.set("one"))
    | }

scala> val prog1: F[Int] = (Err("one")).raiseError[F, Int]
scala> val prog2: F[Int] = (Err("two")).raiseError[F, Int]

scala> prog1.onError(action).value.run("").value

res0: (String, Either[Err,Int]) = (one,Left(Err(one)))

scala> prog2.onError(action).value.run("").value
res1: (String, Either[Err,Int]) = ("",Left(Err(two)))

point lifts any value into a Monoidal Functor.

Example:

scala> import cats.implicits._

scala> InvariantMonoidal[Option].point(10)
res0: Option[Int] = Some(10)

Sequentially compose two actions, discarding any value produced by the second. This variant of productL also lets you define the evaluation strategy of the second action. For instance you can evaluate it only ''after'' the first action has finished:

scala> import cats.Eval
scala> import cats.implicits._
scala> var count = 0
scala> val fa: Option[Int] = Some(3)
scala> def fb: Option[Unit] = Some(count += 1)
scala> fa.productLEval(Eval.later(fb))
res0: Option[Int] = Some(3)
scala> assert(count == 1)
scala> none[Int].productLEval(Eval.later(fb))
res1: Option[Int] = None
scala> assert(count == 1)

Sequentially compose two actions, discarding any value produced by the first. This variant of productR also lets you define the evaluation strategy of the second action. For instance you can evaluate it only ''after'' the first action has finished:

scala> import cats.Eval
scala> import cats.implicits._
scala> val fa: Option[Int] = Some(3)
scala> def fb: Option[String] = Some("foo")
scala> fa.productREval(Eval.later(fb))
res0: Option[String] = Some(foo)

Returns a new value that transforms the result of the source, given the recover or map functions, which get executed depending on whether the result is successful or if it ends in error.

This is an optimization on usage of attempt and map, this equivalence being available:

 fa.redeem(fe, fs) <-> fa.attempt.map(_.fold(fe, fs))

Usage of redeem subsumes handleError because:

 fa.redeem(fe, id) <-> fa.handleError(fe)

Implementations are free to override it in order to optimize error recovery.

Returns a new value that transforms the result of the source, given the recover or bind functions, which get executed depending on whether the result is successful or if it ends in error.

This is an optimization on usage of attempt and flatMap, this equivalence being available:

 fa.redeemWith(fe, fs) <-> fa.attempt.flatMap(_.fold(fe, fs))

Usage of redeemWith subsumes handleErrorWith because:

 fa.redeemWith(fe, F.pure) <-> fa.handleErrorWith(fe)

Usage of redeemWith also subsumes flatMap because:

 fa.redeemWith(F.raiseError, fs) <-> fa.flatMap(fs)

Implementations are free to override it in order to optimize error recovery.

Given fa and n, apply fa n times to construct an F[List[A]] value.

Example:

scala> import cats.data.State

scala> type Counter[A] = State[Int, A]
scala> val getAndIncrement: Counter[Int] = State { i => (i + 1, i) }
scala> val getAndIncrement5: Counter[List[Int]] =
    | Applicative[Counter].replicateA(5, getAndIncrement)
scala> getAndIncrement5.run(0).value
res0: (Int, List[Int]) = (5,List(0, 1, 2, 3, 4))

Inverse of attempt

Example:

scala> import cats.implicits._
scala> import scala.util.{Try, Success}

scala> val a: Try[Either[Throwable, Int]] = Success(Left(new java.lang.Exception))
scala> a.rethrow
res0: scala.util.Try[Int] = Failure(java.lang.Exception)

scala> val b: Try[Either[Throwable, Int]] = Success(Right(1))
scala> b.rethrow
res1: scala.util.Try[Int] = Success(1)

Combines F[A] and F[B] into a F[Either[A,B]]].

Example:

scala> import cats.SemigroupK
scala> import cats.data.NonEmptyList
scala> SemigroupK[NonEmptyList].sum(NonEmptyList.one("abc"), NonEmptyList.one(2))
res0: NonEmptyList[Either[String,Int]] = NonEmptyList(Left(abc), Right(2))

Tuples the A value in F[A] with the supplied B value, with the B value on the left.

Example:

scala> import scala.collection.immutable.Queue
scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForQueue

scala> Functor[Queue].tupleLeft(Queue("hello", "world"), 42)
res0: scala.collection.immutable.Queue[(Int, String)] = Queue((42,hello), (42,world))

Tuples the A value in F[A] with the supplied B value, with the B value on the right.

Example:

scala> import scala.collection.immutable.Queue
scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForQueue

scala> Functor[Queue].tupleRight(Queue("hello", "world"), 42)
res0: scala.collection.immutable.Queue[(String, Int)] = Queue((hello,42), (world,42))

Returns the given argument (mapped to Unit) if cond is false, otherwise, unit lifted into F.

Example:

scala> import cats.implicits._

scala> Applicative[List].unlessA(true)(List(1, 2, 3))
res0: List[Unit] = List(())

scala> Applicative[List].unlessA(false)(List(1, 2, 3))
res1: List[Unit] = List((), (), ())

scala> Applicative[List].unlessA(true)(List.empty[Int])
res2: List[Unit] = List(())

scala> Applicative[List].unlessA(false)(List.empty[Int])
res3: List[Unit] = List()

This repeats an F until we get defined values. This can be useful for polling type operations on State (or RNG) Monads, or in effect monads.

Execute an action repeatedly until the Boolean condition returns true. The condition is evaluated after the loop body. Collects results into an arbitrary Alternative value, such as a Vector. This implementation uses append on each evaluation result, so avoid data structures with non-constant append performance, e.g. List.

Execute an action repeatedly until the Boolean condition returns true. The condition is evaluated after the loop body. Discards results.

Un-zips an F[(A, B)] consisting of element pairs or Tuple2 into two separate F's tupled.

NOTE: Check for effect duplication, possibly memoize before

scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForList

scala> Functor[List].unzip(List((1,2), (3, 4)))
res0: (List[Int], List[Int]) = (List(1, 3),List(2, 4))

Empty the fa of the values, preserving the structure

Example:

scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForList

scala> Functor[List].void(List(1,2,3))
res0: List[Unit] = List((), (), ())

Returns the given argument (mapped to Unit) if cond is true, otherwise, unit lifted into F.

Example:

scala> import cats.implicits._

scala> Applicative[List].whenA(true)(List(1, 2, 3))
res0: List[Unit] = List((), (), ())

scala> Applicative[List].whenA(false)(List(1, 2, 3))
res1: List[Unit] = List(())

scala> Applicative[List].whenA(true)(List.empty[Int])
res2: List[Unit] = List()

scala> Applicative[List].whenA(false)(List.empty[Int])
res3: List[Unit] = List(())

Execute an action repeatedly as long as the given Boolean expression returns true. The condition is evaluated before the loop body. Collects the results into an arbitrary Alternative value, such as a Vector. This implementation uses append on each evaluation result, so avoid data structures with non-constant append performance, e.g. List.

Execute an action repeatedly as long as the given Boolean expression returns true. The condition is evaluated before the loop body. Discards results.

Lifts natural subtyping covariance of covariant Functors.

NOTE: In certain (perhaps contrived) situations that rely on universal equality this can result in a ClassCastException, because it is implemented as a type cast. It could be implemented as map(identity), but according to the functor laws, that should be equal to fa, and a type cast is often much more performant. See this example of widen creating a ClassCastException.

Example:

scala> import cats.Functor
scala> import cats.implicits.catsStdInstancesForOption

scala> val s = Some(42)
scala> Functor[Option].widen(s)
res0: Option[Int] = Some(42)