convert-haskell-scala

name	convert-haskell-scala
description	Bidirectional conversion between Haskell and Scala. Use when migrating projects between these languages in either direction. Extends meta-convert-dev with Haskell↔Scala specific patterns. Use when migrating Haskell projects to Scala, translating Haskell patterns to idiomatic Scala, or refactoring Haskell codebases. Extends meta-convert-dev with Haskell-to-Scala specific patterns.

Haskell ↔ Scala Conversion

Bidirectional conversion between Haskell and Scala. This skill extends meta-convert-dev with Haskell↔Scala specific type mappings, idiom translations, and tooling.

This Skill Extends

• meta-convert-dev - Foundational conversion patterns (APTV workflow, testing strategies)

For general concepts like the Analyze → Plan → Transform → Validate workflow, testing strategies, and common pitfalls, see the meta-skill first.

This Skill Adds

• Type mappings: Haskell types → Scala types
• Idiom translations: Haskell patterns → idiomatic Scala
• Error handling: Haskell Maybe/Either → Scala Option/Either
• Async patterns: Haskell IO/Async → Scala Future/IO
• Lazy evaluation: Haskell lazy by default → Scala strict with lazy vals
• Type classes: Haskell type classes → Scala implicits/given

This Skill Does NOT Cover

• General conversion methodology - see meta-convert-dev
• Haskell language fundamentals - see lang-haskell-dev
• Scala language fundamentals - see lang-scala-dev

---

Quick Reference

Haskell	Scala	Notes
String	String	Both are immutable strings
Int	Int	32-bit integers
Integer	BigInt	Arbitrary precision
Double	Double	Floating point
Bool	Boolean	Boolean values
[a]	List[A]	Linked list
(a, b)	(A, B)	Tuple
Maybe a	Option[A]	Nullable values
Either a b	Either[A, B]	Sum type for errors
IO a	IO[A] (Cats Effect)	Side effects
data	case class / sealed trait	ADTs
class (type class)	trait + implicit/given	Type classes
-> (function type)	=> (function type)	Function types
Monad	Monad (Cats)	Requires Cats library

When Converting Code

1. Analyze source thoroughly before writing target
2. Map types first - create type equivalence table
3. Preserve semantics over syntax similarity
4. Adopt Scala idioms - don't write "Haskell code in Scala syntax"
5. Handle edge cases - lazy evaluation, null safety, JVM interop
6. Test equivalence - same inputs → same outputs

---

Type System Mapping

Primitive Types

Haskell	Scala	Notes
Int	Int	32-bit signed integer
Integer	BigInt	Arbitrary precision integer
Double	Double	64-bit floating point
Float	Float	32-bit floating point
Bool	Boolean	Boolean type
Char	Char	Single character
String	String	Immutable string
()	Unit	Unit type (void)

Collection Types

Haskell	Scala	Notes
[a]	List[A]	Immutable linked list
(a, b)	(A, B)	Tuple (up to 22 elements in Scala)
(a, b, c)	(A, B, C)	3-tuple
Map k v	Map[K, V]	Immutable map
Set a	Set[A]	Immutable set
Vector a	Vector[A]	Indexed sequence
Seq a	Seq[A]	Generic sequence

Composite Types

Haskell	Scala	Notes
data X = A | B	sealed trait X; case object A extends X; case object B extends X	Sum types
data X = X { field :: Type }	case class X(field: Type)	Product types with named fields
newtype X = X Type	case class X(value: Type) extends AnyVal	Zero-cost wrapper
type X = Y	type X = Y	Type alias
Maybe a	Option[A]	Optional values
Either a b	Either[A, B]	Sum type (Left is error by convention in both)

Function Types

Haskell	Scala	Notes
a -> b	A => B	Function from A to B
a -> b -> c	A => B => C or (A, B) => C	Curried vs uncurried
IO a	IO[A] (Cats Effect)	Effectful computation
Monad m => m a	F[A] (with Monad[F])	Higher-kinded types

---

Idiom Translation

Pattern 1: Maybe/Option Handling

Haskell:

findUser :: String -> Maybe User
findUser userId = lookup userId users

-- Pattern matching
case findUser "123" of
  Just user -> processUser user
  Nothing -> putStrLn "User not found"

-- Maybe functions
fromMaybe defaultUser (findUser "123")
maybe "No user" userName (findUser "123")

Scala:

def findUser(userId: String): Option[User] =
  users.get(userId)

// Pattern matching
findUser("123") match {
  case Some(user) => processUser(user)
  case None => println("User not found")
}

// Option methods
findUser("123").getOrElse(defaultUser)
findUser("123").map(_.name).getOrElse("No user")

Why this translation:

• Maybe and Option are semantically equivalent
• Both use pattern matching for explicit handling
• Scala's .getOrElse is more concise than fromMaybe
• Method chaining is idiomatic in Scala

Pattern 2: Either for Error Handling

Haskell:

data AppError = NotFound | ValidationError String

parseAge :: String -> Either AppError Int
parseAge str =
  case reads str of
    [(n, "")] -> if n >= 0
                 then Right n
                 else Left (ValidationError "Age must be positive")
    _ -> Left (ValidationError "Not a valid number")

-- Chaining with do-notation
validateUser :: String -> String -> Either AppError User
validateUser ageStr emailStr = do
  age <- parseAge ageStr
  email <- validateEmail emailStr
  return $ User email age

Scala:

sealed trait AppError
case object NotFound extends AppError
case class ValidationError(message: String) extends AppError

def parseAge(str: String): Either[AppError, Int] = {
  str.toIntOption match {
    case Some(n) if n >= 0 => Right(n)
    case Some(_) => Left(ValidationError("Age must be positive"))
    case None => Left(ValidationError("Not a valid number"))
  }
}

// Chaining with for-comprehension
def validateUser(ageStr: String, emailStr: String): Either[AppError, User] = {
  for {
    age <- parseAge(ageStr)
    email <- validateEmail(emailStr)
  } yield User(email, age)
}

Why this translation:

• Both use Either with Left for errors, Right for success
• Haskell's do-notation maps to Scala's for-comprehension
• ADT error types work similarly in both languages
• Scala requires explicit yield at the end of for-comprehension

Pattern 3: List Comprehensions

Haskell:

-- List comprehension
squares = [x^2 | x <- [1..10], even x]

-- With multiple generators
pairs = [(x, y) | x <- [1..3], y <- [1..3], x < y]

-- Nested comprehensions
matrix = [[1..n] | n <- [1..5]]

Scala:

// For-comprehension
val squares = for {
  x <- 1 to 10
  if x % 2 == 0
} yield x * x

// With multiple generators
val pairs = for {
  x <- 1 to 3
  y <- 1 to 3
  if x < y
} yield (x, y)

// Using map/filter (more idiomatic for simple cases)
val squares = (1 to 10).filter(_ % 2 == 0).map(x => x * x)

// Nested
val matrix = (1 to 5).map(n => (1 to n).toList)

Why this translation:

• Both desugar to map/flatMap/filter
• Scala's for-comprehension requires yield keyword
• Guards (filters) use if in both
• Scala often prefers method chaining for simple transformations

Pattern 4: Pattern Matching on ADTs

Haskell:

data Shape = Circle Double
           | Rectangle Double Double
           | Triangle Double Double Double

area :: Shape -> Double
area (Circle r) = pi * r^2
area (Rectangle w h) = w * h
area (Triangle b h _) = 0.5 * b * h

-- Guards
classify :: Int -> String
classify n
  | n < 0 = "negative"
  | n == 0 = "zero"
  | otherwise = "positive"

Scala:

sealed trait Shape
case class Circle(radius: Double) extends Shape
case class Rectangle(width: Double, height: Double) extends Shape
case class Triangle(base: Double, height: Double, side: Double) extends Shape

def area(shape: Shape): Double = shape match {
  case Circle(r) => math.Pi * r * r
  case Rectangle(w, h) => w * h
  case Triangle(b, h, _) => 0.5 * b * h
}

// Guards
def classify(n: Int): String = n match {
  case n if n < 0 => "negative"
  case 0 => "zero"
  case _ => "positive"
}

Why this translation:

• Haskell uses data constructors, Scala uses case classes
• Pattern matching syntax is similar
• Guards use if in Scala match, | in Haskell function definitions
• sealed trait ensures exhaustiveness checking like Haskell

Pattern 5: Type Classes to Implicits/Given

Haskell:

class Show a where
  show :: a -> String

instance Show Int where
  show = Prelude.show

instance Show User where
  show (User name age) = name ++ " (" ++ Prelude.show age ++ ")"

printValue :: Show a => a -> IO ()
printValue x = putStrLn (show x)

Scala 2 (implicits):

trait Show[A] {
  def show(a: A): String
}

object Show {
  implicit val intShow: Show[Int] = new Show[Int] {
    def show(a: Int): String = a.toString
  }

  implicit val userShow: Show[User] = new Show[User] {
    def show(user: User): String = s"${user.name} (${user.age})"
  }
}

def printValue[A](x: A)(implicit s: Show[A]): Unit = {
  println(s.show(x))
}

Scala 3 (given/using):

trait Show[A] {
  def show(a: A): String
}

given Show[Int] with {
  def show(a: Int): String = a.toString
}

given Show[User] with {
  def show(user: User): String = s"${user.name} (${user.age})"
}

def printValue[A](x: A)(using s: Show[A]): Unit = {
  println(s.show(x))
}

Why this translation:

• Type classes map to traits with implicit/given instances
• Constraints become implicit/using parameters
• Scala 3 syntax is closer to Haskell's
• Both support type class derivation (Haskell with deriving, Scala with macros)

Pattern 6: Lazy Evaluation

Haskell:

-- Infinite lists (lazy by default)
naturals = [1..]
fibs = 0 : 1 : zipWith (+) fibs (tail fibs)

-- Take first 10
take 10 naturals
take 10 fibs

-- Lazy evaluation of expressions
expensiveComputation = trace "Computing..." (sum [1..1000000])
result = if condition then expensiveComputation else 0  -- Only computed if condition is True

Scala:

// LazyList (Stream in Scala 2.12)
val naturals = LazyList.from(1)
val fibs: LazyList[Int] = 0 #:: 1 #:: fibs.zip(fibs.tail).map { case (a, b) => a + b }

// Take first 10
naturals.take(10).toList
fibs.take(10).toList

// Lazy val for deferred evaluation
lazy val expensiveComputation = {
  println("Computing...")
  (1 to 1000000).sum
}
val result = if (condition) expensiveComputation else 0  // Only computed if condition is true

// By-name parameters for lazy arguments
def ifThenElse[A](cond: Boolean)(thenBranch: => A)(elseBranch: => A): A = {
  if (cond) thenBranch else elseBranch
}

Why this translation:

• Haskell is lazy by default, Scala is strict by default
• Use LazyList for infinite sequences
• Use lazy val for deferred computation
• By-name parameters (=> A) for lazy function arguments
• Scala 2.13+ renamed Stream to LazyList

Pattern 7: Function Composition and Application

Haskell:

-- Function composition
addThenDouble = (*2) . (+1)
process = filter even . map (*2) . filter (>0)

-- Function application
result = f $ g $ h x  -- Equivalent to f (g (h x))

-- Point-free style
sumOfSquares = sum . map (^2)

Scala:

// Function composition
val addThenDouble = ((x: Int) => x + 1) andThen (_ * 2)
val addThenDouble2 = ((x: Int) => x * 2) compose ((x: Int) => x + 1)

// Method chaining (more idiomatic)
def process(list: List[Int]): List[Int] =
  list.filter(_ > 0).map(_ * 2).filter(_ % 2 == 0)

// Infix notation for single-arg methods
val result = f(g(h(x)))  // No special operator needed

// Function style with compose
val sumOfSquares = ((list: List[Int]) => list.map(x => x * x)).andThen(_.sum)

// More idiomatic Scala
def sumOfSquares(list: List[Int]): Int = list.map(x => x * x).sum

Why this translation:

• Haskell's . maps to compose (right-to-left) or andThen (left-to-right)
• Scala prefers method chaining over function composition
• Haskell's $ isn't needed in Scala (no precedence issues)
• Point-free style less common in Scala

Pattern 8: Monadic Composition

Haskell:

-- Do-notation
computation :: IO ()
computation = do
  putStrLn "What's your name?"
  name <- getLine
  putStrLn $ "Hello, " ++ name

-- Maybe monad
safeDivision :: Maybe Int
safeDivision = do
  a <- Just 10
  b <- Just 2
  result <- Just (div a b)
  return (result * 2)

-- List monad
pairs :: [(Int, Int)]
pairs = do
  x <- [1..3]
  y <- [1..3]
  guard (x < y)
  return (x, y)

Scala:

// For-comprehension with IO (Cats Effect)
import cats.effect.IO

val computation: IO[Unit] = for {
  _ <- IO.println("What's your name?")
  name <- IO.readLine
  _ <- IO.println(s"Hello, $name")
} yield ()

// Option monad
val safeDivision: Option[Int] = for {
  a <- Some(10)
  b <- Some(2)
  result <- Some(a / b)
} yield result * 2

// List monad
val pairs: List[(Int, Int)] = for {
  x <- (1 to 3).toList
  y <- (1 to 3).toList
  if x < y
} yield (x, y)

Why this translation:

• Haskell's do-notation maps directly to Scala's for-comprehension
• Both desugar to flatMap/map
• Haskell uses return, Scala uses yield
• guard becomes if in for-comprehension
• IO monad requires Cats Effect library in Scala

---

Error Handling

Haskell Maybe/Either → Scala Option/Either

Haskell Pattern	Scala Equivalent	Notes
Nothing	None	Absence of value
Just x	Some(x)	Present value
Left err	Left(err)	Error case
Right val	Right(val)	Success case
fromMaybe default	.getOrElse(default)	Provide default
maybe defaultVal f	.map(f).getOrElse(defaultVal)	Map with default
either errorHandler successHandler	.fold(errorHandler, successHandler)	Fold both cases

Exception Handling

Haskell:

import Control.Exception

readFileSafe :: FilePath -> IO (Either IOException String)
readFileSafe path = try $ readFile path

-- Using try/catch
processFile :: FilePath -> IO ()
processFile path = do
  result <- try (readFile path) :: IO (Either IOException String)
  case result of
    Left ex -> putStrLn $ "Error: " ++ show ex
    Right content -> putStrLn content

Scala:

import scala.util.{Try, Success, Failure}
import java.io.IOException

def readFileSafe(path: String): Either[IOException, String] = {
  Try(scala.io.Source.fromFile(path).mkString).toEither match {
    case Right(content) => Right(content)
    case Left(ex: IOException) => Left(ex)
    case Left(ex) => Left(new IOException(ex))
  }
}

// Using Try
def processFile(path: String): Unit = {
  Try(scala.io.Source.fromFile(path).mkString) match {
    case Success(content) => println(content)
    case Failure(ex) => println(s"Error: ${ex.getMessage}")
  }
}

Why this translation:

• Haskell's try from Control.Exception maps to Scala's Try
• Both can convert to Either for type-safe error handling
• Scala has nullable types from Java, so be careful with interop
• Use Try for exception handling, Either for domain errors

---

Concurrency Model

Haskell IO/Async → Scala Future/IO

Haskell	Scala	Library
IO a	Future[A]	scala.concurrent
IO a	IO[A]	cats-effect
async	Future { ... }	scala.concurrent
forkIO	Future { ... }	scala.concurrent
Async	Async[F]	cats-effect
concurrently	Future.sequence	scala.concurrent
race	Future.firstCompletedOf	scala.concurrent

Basic Async Translation

Haskell:

import Control.Concurrent.Async

fetchData :: IO String
fetchData = do
  threadDelay 1000000
  return "data"

main :: IO ()
main = do
  result <- fetchData
  putStrLn result

-- Concurrent execution
main :: IO ()
main = do
  (data1, data2) <- concurrently fetchData1 fetchData2
  putStrLn $ data1 ++ data2

Scala:

import scala.concurrent.{Future, Await}
import scala.concurrent.duration._
import scala.concurrent.ExecutionContext.Implicits.global

def fetchData: Future[String] = Future {
  Thread.sleep(1000)
  "data"
}

def main(): Unit = {
  val result = Await.result(fetchData, 5.seconds)
  println(result)
}

// Concurrent execution
def main(): Unit = {
  val combined = for {
    data1 <- fetchData1
    data2 <- fetchData2
  } yield data1 + data2

  println(Await.result(combined, 5.seconds))
}

Why this translation:

• Haskell's IO is pure, Scala's Future is eager
• Use Cats Effect IO for pure functional effects in Scala
• concurrently maps to for-comprehension with Futures
• Avoid Await in production; use callbacks or IO

Software Transactional Memory

Haskell:

import Control.Concurrent.STM

type Account = TVar Int

transfer :: Account -> Account -> Int -> STM ()
transfer from to amount = do
  fromBalance <- readTVar from
  when (fromBalance < amount) retry
  modifyTVar from (subtract amount)
  modifyTVar to (+ amount)

main :: IO ()
main = do
  account1 <- newTVarIO 1000
  account2 <- newTVarIO 0
  atomically $ transfer account1 account2 500

Scala:

import scala.concurrent.stm._

type Account = Ref[Int]

def transfer(from: Account, to: Account, amount: Int): Unit = {
  atomic { implicit txn =>
    val fromBalance = from()
    if (fromBalance < amount) retry
    from() = fromBalance - amount
    to() = to() + amount
  }
}

def main(): Unit = {
  val account1 = Ref(1000)
  val account2 = Ref(0)
  transfer(account1, account2, 500)
}

Why this translation:

• Both support STM with similar semantics
• Scala STM requires scala-stm library
• atomic block replaces atomically
• retry works the same way

---

Memory Model

Haskell Lazy Evaluation → Scala Strict Evaluation

Concept	Haskell	Scala	Notes
Default evaluation	Lazy	Strict	Major difference
Force evaluation	seq, deepseq	N/A (already strict)	-
Defer evaluation	Default	lazy val, LazyList	Explicit in Scala
Infinite structures	[1..]	LazyList.from(1)	Requires LazyList
Thunks	Automatic	=> A (by-name)	Explicit in Scala

Space Leaks and Strictness

Haskell:

-- Potential space leak (lazy accumulation)
badSum :: [Int] -> Int
badSum = foldl (+) 0  -- Builds up thunks

-- Strict version
goodSum :: [Int] -> Int
goodSum = foldl' (+) 0  -- Forces evaluation

-- Bang patterns
data Point = Point !Int !Int  -- Strict fields

-- Forcing evaluation
result = x `seq` y  -- Evaluate x, then return y

Scala:

// No space leak (strict by default)
def sum(list: List[Int]): Int =
  list.foldLeft(0)(_ + _)  // Always strict

// Lazy evaluation when needed
lazy val expensiveValue = {
  println("Computing...")
  1 + 1
}

// Strict fields by default
case class Point(x: Int, y: Int)  // Already strict

// All evaluation is forced by default
val result = {
  val _ = x  // x is evaluated
  y          // y is returned
}

Why this matters:

• Haskell's laziness can cause space leaks
• Scala doesn't have this problem by default
• Use lazy val sparingly in Scala
• LazyList for infinite structures

---

Common Pitfalls

1. Assuming Lazy Evaluation

Problem: Expecting Scala to be lazy like Haskell.

Example:

// ❌ This will hang in Scala (strict evaluation)
val naturals = (1 to Int.MaxValue).toList  // Tries to build entire list!

// ✓ Use LazyList for infinite sequences
val naturals = LazyList.from(1)

Solution: Use LazyList for potentially infinite sequences, lazy val for deferred computation.

2. Null Pointer Exceptions

Problem: Scala has null from Java interop, Haskell doesn't.

Example:

// ❌ Dangerous when calling Java code
val name: String = javaObject.getName  // Could be null!
val length = name.length  // NullPointerException

// ✓ Wrap in Option
val name: Option[String] = Option(javaObject.getName)
val length = name.map(_.length).getOrElse(0)

Solution: Always use Option(...) when calling Java code that might return null.

3. Uncurried Functions

Problem: Haskell functions are curried by default, Scala's are not.

Example:

// Haskell: add :: Int -> Int -> Int
// Scala: Either curried or uncurried

// ❌ Uncurried (not partial-application friendly)
def add(a: Int, b: Int): Int = a + b
// Can't do: val add5 = add(5, _)  // Requires placeholder

// ✓ Curried (Haskell-style)
def add(a: Int)(b: Int): Int = a + b
val add5 = add(5) _  // Partial application

Solution: Use curried functions def f(a: A)(b: B) when partial application is needed.

4. Missing Type Classes

Problem: Haskell has many built-in type classes; Scala requires libraries.

Example:

// ❌ No built-in Functor, Monad, etc.
def map[F[_], A, B](fa: F[A])(f: A => B): F[B] = ???  // Can't implement generically

// ✓ Use Cats library
import cats.Functor
import cats.implicits._

def map[F[_]: Functor, A, B](fa: F[A])(f: A => B): F[B] =
  fa.map(f)

Solution: Use Cats library for functional abstractions (Functor, Monad, Applicative, etc.).

5. Pattern Matching Exhaustiveness

Problem: Scala allows non-exhaustive matches without warnings by default.

Example:

// ❌ Non-exhaustive match (compiles but warns)
sealed trait Result
case class Success(value: Int) extends Result
case class Failure(error: String) extends Result

def handle(result: Result): String = result match {
  case Success(value) => s"Got $value"
  // Missing Failure case!
}

// ✓ Complete match
def handle(result: Result): String = result match {
  case Success(value) => s"Got $value"
  case Failure(error) => s"Error: $error"
}

Solution: Use sealed trait and enable -Xfatal-warnings compiler option to catch non-exhaustive matches.

6. Do-Notation vs For-Comprehension Differences

Problem: Subtle differences between Haskell do-notation and Scala for-comprehension.

Example:

// Haskell: do { return x }
// Scala: Must use yield

// ❌ Missing yield
val result = for {
  x <- Some(1)
  y <- Some(2)
  x + y  // Does nothing!
}

// ✓ Use yield
val result = for {
  x <- Some(1)
  y <- Some(2)
} yield x + y

Solution: Always use yield at the end of for-comprehensions (equivalent to Haskell's return).

7. Higher-Kinded Types

Problem: Scala 2 requires explicit kind annotation; Scala 3 is better.

Example:

// Haskell: class Functor f where
//           fmap :: (a -> b) -> f a -> f b

// Scala 2: Requires explicit kind
trait Functor[F[_]] {
  def map[A, B](fa: F[A])(f: A => B): F[B]
}

// Usage requires type lambda for partially applied types
// ❌ Can't do: Functor[Either[String, ?]] in Scala 2
// ✓ Scala 2: Need kind-projector plugin
// ✓ Scala 3: Much better support

Solution: Use Scala 3 for better higher-kinded type support, or use kind-projector plugin in Scala 2.

---

Tooling

Category	Haskell	Scala
Build Tool	Cabal, Stack	sbt, Mill
REPL	GHCi	scala, sbt console
Formatter	stylish-haskell, brittany	scalafmt
Linter	hlint	scalafix, wartremover
Testing	HSpec, QuickCheck	ScalaTest, ScalaCheck
Type Checker	GHC	scalac
Package Registry	Hackage	Maven Central
Documentation	Haddock	Scaladoc

Helpful Libraries for Haskell Patterns

Pattern	Haskell	Scala Library
Type classes	Prelude, base	cats, scalaz
Effects	transformers, mtl	cats-effect, zio
Optics	lens	monocle
JSON	aeson	circe, play-json
Parsing	parsec, megaparsec	cats-parse, fastparse
Testing	QuickCheck	scalacheck
STM	stm	scala-stm

---

Examples

Example 1: Simple - List Processing

Before (Haskell):

-- Sum of squares of even numbers
sumOfEvenSquares :: [Int] -> Int
sumOfEvenSquares xs = sum [x^2 | x <- xs, even x]

-- Alternative with functions
sumOfEvenSquares' :: [Int] -> Int
sumOfEvenSquares' = sum . map (^2) . filter even

-- Pattern matching on lists
myLength :: [a] -> Int
myLength [] = 0
myLength (_:xs) = 1 + myLength xs

After (Scala):

// Sum of squares of even numbers
def sumOfEvenSquares(xs: List[Int]): Int = {
  (for {
    x <- xs
    if x % 2 == 0
  } yield x * x).sum
}

// Alternative with method chaining (more idiomatic)
def sumOfEvenSquares(xs: List[Int]): Int =
  xs.filter(_ % 2 == 0).map(x => x * x).sum

// Pattern matching on lists
def myLength[A](xs: List[A]): Int = xs match {
  case Nil => 0
  case _ :: tail => 1 + myLength(tail)
}

Example 2: Medium - Type Classes and ADTs

Before (Haskell):

-- Type class
class Describable a where
  describe :: a -> String

-- ADT
data Shape = Circle Double
           | Rectangle Double Double
           deriving (Show, Eq)

instance Describable Shape where
  describe (Circle r) = "Circle with radius " ++ show r
  describe (Rectangle w h) = "Rectangle " ++ show w ++ "x" ++ show h

-- Using type class
printDescription :: Describable a => a -> IO ()
printDescription x = putStrLn (describe x)

-- Higher-order function with type class
mapDescribe :: Describable a => [a] -> [String]
mapDescribe = map describe

After (Scala):

// Type class
trait Describable[A] {
  def describe(a: A): String
}

object Describable {
  def apply[A](implicit d: Describable[A]): Describable[A] = d

  implicit class DescribableOps[A](val a: A) extends AnyVal {
    def describe(implicit d: Describable[A]): String = d.describe(a)
  }
}

// ADT
sealed trait Shape
case class Circle(radius: Double) extends Shape
case class Rectangle(width: Double, height: Double) extends Shape

// Type class instance
object Shape {
  implicit val describableShape: Describable[Shape] = new Describable[Shape] {
    def describe(shape: Shape): String = shape match {
      case Circle(r) => s"Circle with radius $r"
      case Rectangle(w, h) => s"Rectangle ${w}x$h"
    }
  }
}

// Using type class
def printDescription[A: Describable](x: A): Unit = {
  println(Describable[A].describe(x))
}

// Or with extension method
import Describable._
def printDescription[A: Describable](x: A): Unit = {
  println(x.describe)
}

// Higher-order function with type class
def mapDescribe[A: Describable](xs: List[A]): List[String] =
  xs.map(_.describe)

Example 3: Complex - Parser Combinator with Monadic Composition

Before (Haskell):

import Text.Parsec
import Text.Parsec.String (Parser)

-- Simple expression parser
data Expr = Num Int
          | Add Expr Expr
          | Mul Expr Expr
          deriving (Show, Eq)

number :: Parser Expr
number = Num . read <$> many1 digit

expr :: Parser Expr
expr = term `chainl1` addOp

term :: Parser Expr
term = factor `chainl1` mulOp

factor :: Parser Expr
factor = number <|> parens expr

parens :: Parser a -> Parser a
parens p = char '(' *> p <* char ')'

addOp :: Parser (Expr -> Expr -> Expr)
addOp = Add <$ char '+'

mulOp :: Parser (Expr -> Expr -> Expr)
mulOp = Mul <$ char '*'

-- Evaluator
eval :: Expr -> Int
eval (Num n) = n
eval (Add e1 e2) = eval e1 + eval e2
eval (Mul e1 e2) = eval e1 * eval e2

-- Usage
parseAndEval :: String -> Either ParseError Int
parseAndEval input = do
  expr <- parse expr "" input
  return (eval expr)

After (Scala):

import cats.parse.{Parser, Numbers}
import cats.parse.Parser._
import cats.parse.Rfc5234.{char, digit}

// Simple expression parser
sealed trait Expr
case class Num(value: Int) extends Expr
case class Add(left: Expr, right: Expr) extends Expr
case class Mul(left: Expr, right: Expr) extends Expr

object ExprParser {
  val number: Parser[Expr] =
    Numbers.digits.map(s => Num(s.toInt))

  lazy val expr: Parser[Expr] =
    chainl1(term, addOp)

  lazy val term: Parser[Expr] =
    chainl1(factor, mulOp)

  lazy val factor: Parser[Expr] =
    number.orElse(parens(expr))

  def parens[A](p: Parser[A]): Parser[A] =
    (char('(') *> p <* char(')'))

  val addOp: Parser[(Expr, Expr) => Expr] =
    char('+').as((e1: Expr, e2: Expr) => Add(e1, e2))

  val mulOp: Parser[(Expr, Expr) => Expr] =
    char('*').as((e1: Expr, e2: Expr) => Mul(e1, e2))

  // Helper for chainl1 (not built-in)
  def chainl1[A](p: Parser[A], op: Parser[(A, A) => A]): Parser[A] = {
    (p ~ (op ~ p).rep0).map {
      case (initial, ops) =>
        ops.foldLeft(initial) { case (acc, (f, next)) => f(acc, next) }
    }
  }
}

// Evaluator
def eval(expr: Expr): Int = expr match {
  case Num(n) => n
  case Add(e1, e2) => eval(e1) + eval(e2)
  case Mul(e1, e2) => eval(e1) * eval(e2)
}

// Usage
def parseAndEval(input: String): Either[Parser.Error, Int] = {
  ExprParser.expr.parseAll(input).map(eval)
}

---

See Also

For more examples and patterns, see:

• meta-convert-dev - Foundational patterns with cross-language examples
• convert-typescript-rust - Similar functional programming conversion
• lang-haskell-dev - Haskell development patterns
• lang-scala-dev - Scala development patterns

Cross-cutting pattern skills:

• patterns-concurrency-dev - STM, async patterns across languages
• patterns-serialization-dev - JSON, validation patterns
• patterns-metaprogramming-dev - Type classes, macros, generics