variable / @typed/fx/Fx

mapEffect

Transforms the elements of an Fx using a provided Effectful function.

Package version
2.0.0-beta.13
Category
Transforming values
Since
1.0.0

Follow one execution

Wait one turn for each mapping Effect in this serialized source. Concurrent producers can overlap callback Effects; these rows assume serial delivery.

Fx timelinemapEffect(oneTurnCallback)mapEffect
OperatormapEffect(oneTurnCallback)
Read this diagram

Follow each lane from left to right. Events stacked vertically share a tick; the green cursor marks the current time across every lane.

  • A value
    The text inside the pill is the emitted value.
  • Work starts
    The raised chevron starts an inner run (^ in the source).
  • The run returns
    The vertical bar ends this lane’s run.
  • A cause is delivered
    The exclamation mark belongs to this lane.
  • Work is interrupted
    The cross marks cancellation of this run.
  • Current time
    The line and diamond move together across all lanes.
  • Happening now
    A highlighted event is at the current tick.
  • Still ahead
    Muted, dashed values have not happened yet.
  • Time continues
    The lane’s arrow is not a return marker. An empty stretch can be quiet work that is still running.

Illustrated ticks start at 0. At 1×, one illustrated tick takes one second; captions specify real durations when timing matters. A cause or interruption belongs to its lane, and other work may continue. Scroll horizontally to inspect the rest of a long timeline.

Each callback Effect belongs to the producer delivery that invoked it. Failure is routed to the Sink, services remain required, and interruption is local to that delivery. The combinator adds no queue, semaphore, result retention, or independent fiber.

Source implementation · Learn the surrounding model

Compare all Fx timelines →

Import

import { mapEffect } from "@typed/fx/Fx";

Signatures

export declare const mapEffect: {
    <A, B, E2, R2>(f: (a: A) => Effect.Effect<B, E2, R2>): <E, R>(fx: Fx<A, E | E2, R>) => Fx<B, E | E2, R | R2>;
    <A, E, R, B, E2, R2>(fx: Fx<A, E | E2, R>, f: (a: A) => Effect.Effect<B, E2, R2>): Fx<B, E | E2, R | R2>;
};

Why

mapEffect performs one Effectful transformation for every input and emits once for each successful callback. It does not serialize sink invocations: ordering and concurrency are owned by the producer, so overlapping callbacks may complete and emit out of input order.

Ownership and lifetime

Each callback Effect belongs to the producer delivery that invoked it. Failure is routed to the Sink, services remain required, and interruption is local to that delivery. The combinator adds no queue, semaphore, result retention, or independent fiber.

Examples

import { Effect } from "effect"
import { Fx } from "@typed/fx"

const concurrent = Fx.make<number>((sink) =>
  Effect.all([sink.onSuccess(1), sink.onSuccess(2)], { concurrency: "unbounded", discard: true })
)
const program = Fx.collectAll(
  concurrent.pipe(Fx.mapEffect((n) => Effect.sleep(n === 1 ? "10 millis" : "0 millis").pipe(Effect.as(n))))
) // producer concurrency permits [2, 1]

Other public imports

These import paths expose the same declaration. Each page retains its own public name and signature.

Source