IBM/fp-go

fp-go: Functional Programming Library for Go

🚧 Work in progress! 🚧 Despite major version 1 (due to semantic-release limitations), we're working to minimize breaking changes.

A comprehensive functional programming library for Go, strongly influenced by the excellent fp-ts library for TypeScript.

📚 Table of Contents

• Getting Started
• Design Goals
• Core Concepts
• Comparison to Idiomatic Go
• Implementation Notes
• Common Operations
• Resources

🚀 Getting Started

Installation

go get github.com/IBM/fp-go

Quick Example

import (
    "errors"
    "github.com/IBM/fp-go/either"
    "github.com/IBM/fp-go/function"
)

// Pure function that can fail
func divide(a, b int) either.Either[error, int] {
    if b == 0 {
        return either.Left[int](errors.New("division by zero"))
    }
    return either.Right[error](a / b)
}

// Compose operations safely
result := function.Pipe2(
    divide(10, 2),
    either.Map(func(x int) int { return x * 2 }),
    either.GetOrElse(func() int { return 0 }),
)
// result = 10

Resources

• 📖 API Documentation
• 💡 Code Samples
• 🆕 V2 Documentation (requires Go 1.24+)

🎯 Design Goals

This library aims to provide a set of data types and functions that make it easy and fun to write maintainable and testable code in Go. It encourages the following patterns:

Core Principles

• Pure Functions: Write many small, testable, and pure functions that produce output only depending on their input and execute no side effects
• Side Effect Isolation: Isolate side effects into lazily executed functions using the IO monad
• Consistent Composition: Expose a consistent set of composition functions across all data types
  • Each data type has a small set of composition functions
  • Functions are named consistently across all data types
  • Semantics of same-named functions are consistent across data types

🧘🏽 Alignment with the Zen of Go

This library respects and aligns with The Zen of Go:

💡 Core Concepts

Data Types

The library provides several key functional data types:

• Option[A]: Represents an optional value (Some or None)
• Either[E, A]: Represents a value that can be one of two types (Left for errors, Right for success)
• IO[A]: Represents a lazy computation that produces a value
• IOEither[E, A]: Represents a lazy computation that can fail
• Reader[R, A]: Represents a computation that depends on an environment
• ReaderIOEither[R, E, A]: Combines Reader, IO, and Either for effectful computations with dependencies
• Task[A]: Represents an asynchronous computation
• State[S, A]: Represents a stateful computation

Monadic Operations

All data types support common monadic operations:

• Map: Transform the value inside a context
• Chain (FlatMap): Transform and flatten nested contexts
• Ap: Apply a function in a context to a value in a context
• Of: Wrap a value in a context
• Fold: Extract a value from a context

Comparison to Idiomatic Go

This section explains how functional APIs differ from idiomatic Go and how to convert between them.

Pure Functions

Pure functions take input parameters and compute output without changing global state or mutating inputs. They always return the same output for the same input.

Without Errors

If your pure function doesn't return an error, the idiomatic signature works as-is:

func add(a, b int) int {
    return a + b
}

With Errors

Idiomatic Go:

func divide(a, b int) (int, error) {
    if b == 0 {
        return 0, errors.New("division by zero")
    }
    return a / b, nil
}

Functional Style:

func divide(a, b int) either.Either[error, int] {
    if b == 0 {
        return either.Left[int](errors.New("division by zero"))
    }
    return either.Right[error](a / b)
}

Conversion:

• Use either.EitherizeXXX to convert from idiomatic to functional style
• Use either.UneitherizeXXX to convert from functional to idiomatic style

Effectful Functions

An effectful function changes data outside its scope or doesn't always produce the same output for the same input.

Without Errors

Functional signature: IO[T]

func getCurrentTime() io.IO[time.Time] {
    return func() time.Time {
        return time.Now()
    }
}

With Errors

Functional signature: IOEither[error, T]

func readFile(path string) ioeither.IOEither[error, []byte] {
    return func() either.Either[error, []byte] {
        data, err := os.ReadFile(path)
        if err != nil {
            return either.Left[[]byte](err)
        }
        return either.Right[error](data)
    }
}

Conversion:

• Use ioeither.EitherizeXXX to convert idiomatic Go functions to functional style

Go Context

Functions that take a context.Context are effectful because they depend on mutable context.

Idiomatic Go:

func fetchData(ctx context.Context, url string) ([]byte, error) {
    // implementation
}

Functional Style:

func fetchData(url string) readerioeither.ReaderIOEither[context.Context, error, []byte] {
    return func(ctx context.Context) ioeither.IOEither[error, []byte] {
        return func() either.Either[error, []byte] {
            // implementation
        }
    }
}

Conversion:

• Use readerioeither.EitherizeXXX to convert idiomatic Go functions with context to functional style

Implementation Notes

Generics

All monadic operations use Go generics for type safety:

• ✅ Pros: Type-safe composition, IDE support, compile-time correctness
• ⚠️ Cons: May result in larger binaries (different versions per type)
• 💡 Tip: For binary size concerns, use type erasure with any type

Ordering of Generic Type Parameters

Go requires all type parameters on the global function definition. Parameters that cannot be auto-detected come first:

// Map: B cannot be auto-detected, so it comes first
func Map[R, E, A, B any](f func(A) B) func(ReaderIOEither[R, E, A]) ReaderIOEither[R, E, B]

// Ap: B cannot be auto-detected from the argument
func Ap[B, R, E, A any](fa ReaderIOEither[R, E, A]) func(ReaderIOEither[R, E, func(A) B]) ReaderIOEither[R, E, B]

This ordering maximizes type inference where possible.

Use of the ~ Operator

Go doesn't support generic type aliases (until Go 1.24), only type definitions. The ~ operator allows generic implementations to work with compatible types:

type ReaderIOEither[R, E, A any] RD.Reader[R, IOE.IOEither[E, A]]

Generic Subpackages:

• Each higher-level type has a generic subpackage with fully generic implementations
• These are for library extensions, not end-users
• Main packages specialize generic implementations for convenience

Higher Kinded Types (HKT)

Go doesn't support HKT natively. This library addresses this by:

• Introducing HKTs as individual types (e.g., HKTA for HKT[A])
• Implementing generic algorithms in the internal package
• Keeping complexity hidden from end-users

Common Operations

Map/Chain/Ap/Flap

Example: Chaining Operations

import (
    "github.com/IBM/fp-go/either"
    "github.com/IBM/fp-go/function"
)

result := function.Pipe3(
    either.Right[error](10),
    either.Map(func(x int) int { return x * 2 }),
    either.Chain(func(x int) either.Either[error, int] {
        if x > 15 {
            return either.Right[error](x)
        }
        return either.Left[int](errors.New("too small"))
    }),
    either.GetOrElse(func() int { return 0 }),
)

📚 Resources

• API Documentation
• Code Samples
• V2 Documentation - New features in Go 1.24+
• fp-ts - Original TypeScript inspiration

🤝 Contributing

Contributions are welcome! Please feel free to submit issues or pull requests.

📄 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.