Imitating Scala Futures with Go channels (in Go and Clojure)

Scala's Future abstraction is one of my favourite feature of the language. It's simple, elegant and powerful. I make the beginner's attempt to mimic them in Go in a simplistic domain,
Let's assume we have two functions which will take some time (or http calls that naturally return Futures) and we need to aggregate their results.

The Scala way

	//assume these take time
	def longComp1(): Int = 5
	def longComp2(): Int = 6
	def aggregate(i: Int, j: Int): Unit = {
	println(s"i is $i, j is $j")
	}
	val longComp1Result = Future { longComp1() }
	val longComp2Result = Future { longComp2() }
	val aggregatedFuture: Future[Unit] = for {
	r1 <- longComp1Result
	r2 <- longComp1Result
	} yield aggregate(r1, r2)

view raw Futures.scala hosted with ❤ by GitHub

The Go way

	//define a type for brevity below
	type LongComp func() int
	longComp1 := func() int {
	return 5
	}
	longComp2 := func() int {
	return 6
	}
	aggregate := func(is []int) {
	fmt.Println(is)
	}
	f := func(fs []LongComp, agg func([]int)) {
	channels := []chan int{}
	//loop through the functions
	for _, fun := range fs {
	//create a channel for each and load it's result in
	c := make(chan int)
	channels = append(channels, c)
	go func(f LongComp) {
	c <- f()
	close(c)
	}(fun)
	}
	//get all the elements off the channel and apply the aggregate function
	go func() {
	results := []int{}
	for _, c := range channels {
	results = append(results, <-c)
	}
	agg(results)
	}()
	}
	f([]LongComp{longComp1, longComp2}, aggregate)

view raw futures.go hosted with ❤ by GitHub

It's rather clunky compared to Scala's solution. But the main problem is that it's not polymorphic on the types. The same lump of code would need to be redefined every time the function signatures change.

Here is a more generic version below that uses explicit type casting.

	longComp1 := func() int {
	return 5
	}
	longComp2 := func() int {
	return 6
	}
	aggregate := func(is []int) {
	fmt.Println(is)
	}
	type LongComp2 func() interface{}
	//the "generic" function that works regardless of the underlying type
	f2 := func(fs []LongComp2, agg func([]interface{})) {
	channels := []chan interface{}{}
	for _, fun := range fs {
	c := make(chan interface{})
	channels = append(channels, c)
	go func(f LongComp2) {
	c <- f()
	close(c)
	}(fun)
	}
	go func() {
	results := []interface{}{}
	for _, c := range channels {
	results = append(results, <-c)
	}
	agg(results)
	}()
	}
	//functions needs to be transformed to accept and return interface{}
	//transform a ()->int function to a ()->interface{} function
	//without changing behaviour
	wrap := func(f func() int) func() interface{} {
	return func() interface{} {
	return f()
	}
	}
	f2([]LongComp2{
	wrap(longComp1),
	wrap(longComp2)},
	//transform the aggregate function to accept []interface{} instead of []int
	func(xs []interface{}) {
	is := []int{}
	for _, x := range xs {
	is = append(is, x.(int))
	}
	aggregate(is)
	})

view raw futures2.go hosted with ❤ by GitHub

Not the nicest code I've ever written. If only Go had generics... Any Golang expert out there who could suggest improvements?

The Clojure Way

	(ns home.async-play
	(:require [clojure.core.async :as async]))
	(defn long-comp1 [] 5)
	(defn long-comp2 [] 6)
	(defn aggregate [xs]
	(println xs))
	;;simple version, working with 2 functions
	(let [c1 (async/go (long-comp1))
	c2 (async/go (long-comp2))]
	(async/go
	(aggregate [(async/<! c1) (async/<! c2)])))
	;;generalized version, working with a sequence of functions
	(defn fs->chans
	"Returns channels holding the result of each function execution"
	[fs]
	(for [f fs] (async/go (f))))
	(let [channels (fs->chans [long-comp1 long-comp2])]
	(async/go
	(aggregate (map async/<!! channels))))

view raw futures.clj hosted with ❤ by GitHub

Clojure borrowed its go-routine and channels idea from Golang. Yet the difference between the Clojure and Go versions in terms of brevity is striking. It's the result of 3 independent factors. One, Clojure is dynamically typed, hence there is no need to hassle with neither function signatures nor generics, or rather the lack of them. Two, it's a functional language. Maps and list comprehensions are way more terse than Go's for-loops. The third factor is async/go returns a channel containing the result of the function executed. Go needs to create a slice of channels first, loop through the slice of functions, create anonymous functions in go blocks where the function is executed and its result is put on the channel, ....lot's of hassle.

With generics many of Go's problems would go away. Rob Pike explicitly said, no generics, but hopefully someone will just force him or do it himself instead.

Solving problems in Scala, Clojure and Go - Template Method Design Pattern

I've planning for ages to write a series of posts about comparing OO and FP solutions for common problems. In a similar vein comparing Scala and Clojure. Usually my enthusiasm doesn't survive until I have time to sit down in front of the computer out of working hours. But recently I've started to read about Go and it rekindled my interest in such mental exercises. It also gives me a nice opportunity to learn Go. Also these 3 languages are quite different, which makes these little problems even more educative.

Without further ado, here is the first one.

In OO polymorphism is achieved by inheritance. Inheritance also enables sharing data and logic sparing the developer some code duplication. It is a powerful tool, but nowadays generally regarded overused. The Template Method Design Pattern is a nice example and I've implemented it dozens (if not hundreds) of times at work. Let's see how to solve the problems it solves in languages that don't have inheritance.

The idea is to devise a toy problem, solve it in Scala (it would be the same in Java, I just want keep the boilerplate down) to demonstrate the OO-way, then come up with Clojure and Go solutions.

The Problem

Our domain has employees who can fall into 2 broad categories, Office Workers and Field Workers. Their base salary and the calculation of how their years at the company contribute to their salary are different. I let the code speak for itself.

Scala/OO solution

	sealed abstract class Employee(val name: String, years: Int, baseSalary: BigDecimal) {
	//abstract method
	def yearsContribution(): BigDecimal
	//do the same logic with different data and sub-logic in each subclass
	def salary(): BigDecimal = baseSalary + yearsContribution()
	override def toString: String = s"My name is $name and I've been working here for $years years."
	}
	class OfficeWorker(name: String, years: Int)
	extends Employee(name, years, baseSalary = BigDecimal(1000)) {
	override def yearsContribution(): BigDecimal = years * 30
	}
	//extra fields
	class FieldWorker(name: String, years: Int, physicalWork: Boolean)
	extends Employee(name, years, baseSalary = BigDecimal(800)) {
	override def yearsContribution(): BigDecimal =
	years * (if (physicalWork) 40 else 30)
	}
	object EmloyeeTestRunner extends App {
	val o = new OfficeWorker("joe", 12)
	val f = new FieldWorker("jack", 8, true)
	Seq[Employee](o, f).foreach { e ⇒
	println(s"$e. I earn ${e.salary()}")
	}
	}

view raw TemplatePatternExample.scala hosted with ❤ by GitHub

Plain and simple for a OO developer.

Clojure solution

	(ns example
	(:require [clojure.spec :as s]))
	;; Spec section. It's not needed for the solution, but provides
	;; documentation and can generate tests.
	(def employee-types #{:office-worker :field-agent})
	(s/def ::type employee-types)
	(s/def ::years (s/int-in 0 100))
	(s/def ::name string?)
	(s/def ::base-salary (s/int-in 0 1000))
	(s/def ::physical-work? boolean?)
	(s/def ::employee (s/keys :req [::years ::name ::base-salary ::type]
	:opt [::physical-work?]))
	(s/fdef salary
	:args (s/cat :in ::employee)
	:ret number?)
	;; ============================================================
	(defmulti years-contribution ::type)
	(defmethod years-contribution :office-worker [x]
	(* 30 (::years x)))
	(defmethod years-contribution :field-agent [x]
	(let [multiplier (if (::physical-work? x) 40 30)]
	(* multiplier (::years x))))
	(defn to-string [employee]
	(str "My name is " (::name employee) " and I've been working here for " (::years employee) " years."))
	(defn office-worker [name years]
	{::name name ::years years ::base-salary 800 ::type :office-worker})
	(defn field-agent [name years physical-work?]
	{::name name ::years years ::base-salary 1000
	::physical-work? physical-work? ::type :field-agent})
	(defn salary [employee]
	(+ (::base-salary employee) (years-contribution employee)))
	;;=============== demonstrate =======================
	(let [joe (field-agent "joe" 17 false)
	jil (field-agent "jil" 17 true)
	jack (office-worker "jack" 20)]
	(doseq [x [joe jil jack]]
	(println (to-string x))
	(println (salary x))))

view raw template-pattern-example.clj hosted with ❤ by GitHub

No inheritance, obviously and no types either. Therefore the toString and salary functions are not defined on types, but the entity maps are passed as arguments. Another solution could have been using protocols and defrecords, that would have yielded a more OO-like solution. However for a single function it seemed to be an overkill.

Go solution

	package main
	import "fmt"
	type (
	employee interface {
	yearContribution() float32
	baseSalary() int
	toString() string
	}
	baseEmployee struct {
	name string
	years int
	}
	officeWorker struct {
	baseEmployee
	}
	fieldWorker struct {
	baseEmployee
	physicalWork bool
	}
	)
	func (_ officeWorker) baseSalary() int {
	return 500
	}
	func (_ fieldWorker) baseSalary() int {
	return 1000
	}
	func (o officeWorker) yearContribution() float32 {
	return 30 * float32(o.years)
	}
	func (f fieldWorker) yearContribution() float32 {
	if f.physicalWork {
	return 40 * float32(f.years)
	} else {
	return 30 * float32(f.years)
	}
	}
	func salary(e employee) float32 {
	return float32(e.baseSalary()) + e.yearContribution()
	}
	func (e baseEmployee) toString() string {
	return fmt.Sprintf("My name is %s and I've been working here for %d years.", e.name, e.years)
	}
	func main() {
	ow := officeWorker{baseEmployee{name:"joe", years:20}}
	fw := fieldWorker{baseEmployee: baseEmployee{name:"jack", years:20}, physicalWork: true}
	for _, e := range []employee{ow, fw} {
	fmt.Println(e.toString())
	fmt.Println(salary(e))
	}
	}

view raw TemplatePatternExample.go hosted with ❤ by GitHub

It took me some time to come up with a solution. I Go there are no abstract classes or virtual methods. This required me to define to a separate employee interface and a baseEmployee struct. And I couldn't attach the salary method to the interface either, even though it has both the yearContribution() and the baseSalary() methods on it. It is not necessarily a problem, this could be idiomatic in Go.