Pieper 🥧

A tiny, typesafe, asynchronous pipeline for functional programming in TypeScript.

npm | Buy Me a Coffee

What Does It Do?

Pieper allows you to build complex, sequential data flows with a fluent, chainable API. It's designed to be async-native from the ground up, so you can mix synchronous and asynchronous operations seamlessly without cluttering your code with await at every step.

import { Pieper, type SafeResult } from "pieper";

async function slugify(
  title: string | null | undefined
): Promise<SafeResult<string>> {
  const pipe = Pieper.of(title)
    .assert(
      (s): s is string => typeof s === "string" && s.trim().length !== 0,
      "Input title cannot be empty"
    )
    .log("1. Original:")
    .map((s) => s.toLowerCase())
    .map((s) => s.normalize("NFD"))
    .map((s) => s.replace(/[\u0300-\u036f]/g, ""))
    .log("2. Normalized/Lowercased:")
    .map((s) => s.replace(/[^a-z0-9]+/g, "-"))
    .log("3. Hyphenated:")
    .map((s) => s.replace(/^-+|-+$/g, ""))
    .log("4. Trimmed:")
    .ifElse(
      (s) => s.length > 0,
      (s) => s,
      () => "n-a"
    );

  return pipe.runSafe();
}

Features

• Fluent & Chainable: A clean, easy-to-read API (.map().tap().if()...).
• Fully Typesafe: TypeScript tracks the data type as it changes through the pipe.
• Async-Native: The entire pipeline is built on Promises. Any step can be async.
• Robust Error Handling: Use .catch() for internal recovery or .runSafe() for external, throw-free error handling.
• Zero Dependencies: A single, lightweight class.

Installation

npm install pieper
# or
yarn add pieper

Basic Usage

Import Pieper and create a new pipeline using Pieper.of():

import { Pieper } from "pieper";

async function main() {
  const result = await Pieper.of(5)
    .map((x) => x + 2)
    .tap((x) => console.log("Current value:", x))
    .if(
      (x) => x > 5,
      (x) => x * 10
    )
    .log("After .if:")
    .catch((err) => {
      console.error("Oops:", err.message);

      return 0; // Recover with a default value
    })
    .run();

  console.log("Final result:", result); // Final result: 70
}

main();

API Reference

Constructors

Pieper.of<T>(value: T | Promise<T>): Pieper<T>

Creates a new pipeline from a value or a promise.

const p1 = Pieper.of(10);
const p2 = Pieper.of(Promise.resolve(10));

Pieper.from<T>(fn: () => T | Promise<T>): Pieper<T>

Creates a new pipeline by executing a function. This is useful for deferring the creation of the initial value.

const p = Pieper.from(() => heavyComputation());
const pAsync = Pieper.from(() => fs.promises.readFile("..."));

Transformation

.map<R>(fn: (value: T) => R | Promise<R>): Pieper<R>

Transforms the value in the pipe from T to R. The fn can be sync or async.

Pieper.of(5)
  .map((x) => x.toString()) // Pieper<string>
  .map(async (str) => `Value is ${str}`); // Pieper<string>

.if<R>(condition: (value: T) => boolean, thenFn: (value: T) => R | Promise<R>): Pieper<T | R>

Conditionally transforms the value only if the condition is true. If false, the original value is passed through.

Pieper.of(5).if(
  (x) => x < 10,
  (x) => x * 100
); // -> 500

Pieper.of(20).if(
  (x) => x < 10,
  (x) => x * 100
); // -> 20 (passes through)

.ifElse<R1, R2>(condition: (value: T) => boolean, thenFn: (value: T) => R1, elseFn: (value: T) => R2): Pieper<R1 | R2>

Transforms the value into one of two different types based on a condition.

Pieper.of(10).ifElse(
  (x) => x > 5, // condition
  (x) => "is-large", // thenFn
  (x) => "is-small" // elseFn
); // Pieper<string> -> "is-large"

Side Effects

.tap(fn: (value: T) => void | Promise<void>): Pieper<T>

Performs a side effect (like logging) without changing the value.

Pieper.of(10)
  .tap((x) => console.log(x))
  .map((x) => x * 2);

.log(message?: string): Pieper<T>

A convenience helper for .tap() to quickly log the current value.

Pieper.of({ id: 1, name: "Test" }).log("User data:");
// Logs: "User data:" { id: 1, name: "Test" }

Validation

.assert(predicate: (value: T) => boolean, errorMessage: string | Error): Pieper<T>

Enforces a contract. If the predicate is false, the pipe fails (rejects) with the given error.

Pieper.of(email).assert((str) => str.includes("@"), "Invalid email");

Error Handling

.catch<R>(fn: (error: any) => R | Promise<R>): Pieper<T | R>

Recovers from a failed pipeline. The fn receives the error and must return a new "success" value.

Pieper.from(() => {
  throw new Error("Kaboom!");
})
  .map((x) => x * 2)
  .catch((err) => {
    console.error(err.message); // "Kaboom!"

    return 0; // Recover with 0
  }); // Pieper<number>

.finally(fn: () => void | Promise<void>): Pieper<T>

Runs a function when the pipe settles (either on success or failure).

Pieper.of(db.connect())
  .map((conn) => conn.query("..."))
  .finally(() => db.close());

Executors (Getting the Value Out)

The pipeline is lazy and does nothing until you call an executor.

.run(): Promise<T>

Executes the pipeline and returns a promise for the final value. This promise will reject if any step fails.

try {
  const result = await Pieper.of(10).run();

  console.log(result); // 10
} catch (e) {
  console.error("Failed:", e);
}

.runSafe(): Promise<SafeResult<T>>

Executes the pipeline and never throws. It always resolves with a SafeResult object:

• { ok: true, value: T } on success.
• { ok: false, error: any } on failure.

This is the recommended way to handle results in a functional style.

import { Pieper, type SafeResult } from "pieper";

const result = await Pieper.of("not-an-email")
  .assert((x) => x.includes("@"), "Bad email")
  .runSafe();

if (result.ok) {
  // result.value is string
  console.log("Success:", result.value);
} else {
  // result.error is the Error("Bad email")
  console.error("Failure:", result.error.message);
}

.runAndForget(): void

Executes the pipeline and ignores the result. Any unhandled errors will be logged to the console. Use this for "fire-and-forget" operations.

Pieper.of("Data")
  .map((data) => analytics.track(data))
  .runAndForget();