Introduction to Functional Programming with Scala

Introduction to Functional Programming with
Scala

Pramode C.E

http://pramode.net

February 26, 2013

Workshop Plan

Here is what we will do:
Learn a bit of Scala syntax and write some simple Scala
programs
Learn some important concepts like: closures, higher order
functions, purity, lazy vs strict evaluation, currying, tail
calls/TCO, immutability, persistent data structures, type
inference etc!
Understand the essence of the ”Functional Programming”
paradigm through Scala

For discussions, please join: scala-kerala@googlegroups.com
Workshop material (slide/code samples):
http://pramode.net/downloads/scala-workshop.tgz

Pramode C.E Introduction to Functional Programming with Scala

Scala - some background info

Scala = ”Scalable Language” - a language which grows with
the demands of its users!
Designed by Prof.Martin Odersky at EPFL
Combines Object Oriented and Functional Programming with
a a powerful static type system and expressive syntax
Compiles to Java byte code and runs on the JVM
Upsurge in adoption since 2010 - Twitter / Foursquare /
Guardian newspaper / LinkedIn are some high proﬁle users
More info: Wikipedia


Create a text ﬁle say hello.scala with the following line in it:

println(”hello, world”)

Type ”scala hello.scala” at the command prompt to run the code!


The Scala REPL

scala> 2*3
res0: Int = 6

scala> val a = 10
a: Int = 10

scala> val b = a + 1
b: Int = 11

scala> a = a + 1
<console>:8: error: reassignment to val
a = a + 1
^
scala>


Function Definition

def add(a:Int, b:Int):Int = a + b

val m:Int = add(1, 2)

println(m)

Note the use of the type declaration ”Int”. Scala is a ”statically
typed” language. We define ”add” to be a function which accepts
two parameters of type Int and returns a value of type Int.
Similarly, ”m” is defined as a variable of type Int.


Function Deﬁnition

def fun(a: Int):Int = {
a + 1
a - 2
a * 3
}

val p:Int = fun(10)
println(p)

Note!
There is no explicit ”return” statement! The value of the last
expression in the body is automatically returned.


Type Inference

def add(a:Int, b:Int) = a + b

val m = add(1, 2)

println(m)

We have NOT speciﬁed the return type of the function or the type
of the variable ”m”. Scala ”infers” that!


Type Inference

def add(a, b) = a + b

val m = add(1, 2)

println(m)

This does not work! Scala does NOT infer type of function
parameters, unlike languages like Haskell/ML. Scala is said to do
local, ”ﬂow-based” type inference while Haskell/ML do
Hindley-Milner type inference

References
1 http://en.wikipedia.org/wiki/Hindley-Milner

2 http://www.scala-lang.org/node/4654

Expression Oriented Programming

val i = 3
val p = if (i > 0) -1 else -2
val q = if (true) "hello" else "world"

println(p)
println(q)

Unlike languages like C/Java, almost everything in Scala is an
”expression”, ie, something which returns a value! Rather than
programming with ”statements”, we program with ”expressions”


Expression Oriented Programming

def errorMsg(errorCode: Int) = errorCode match {
case 1 => "File not found"
case 2 => "Permission denied"
case 3 => "Invalid operation"
}

println(errorMsg(2))

Case automatically ”returns” the value of the expression
corresponding to the matching pattern.


Recursion

// sum n + (n-1) + (n-2) + ... + 0
def sum(n: Int): Int =
if (n == 0) 0 else n + sum(n - 1)

val m = sum(10)
println(m)

Try calling the function ”sum” with a large number (say 10000) as
parameter! You get a stack overﬂow!


Tail Calls and TCO

def sum(n: Int, acc: Int):Int =
if(n == 0) acc else sum(n - 1, acc + n)

val r = sum(10000, 0)

println(r)

This is a ”tail-recursive” version of the previous function - the
Scala compiler converts the tail call to a loop, thereby avoiding
stack overﬂow!


Tail Calls and TCO

(sum 4)
(4 + sum 3)
(4 + (3 + sum 2))
(4 + (3 + (2 + sum 1)))
(4 + (3 + (2 + (1 + sum 0))))
(4 + (3 + (2 + (1 + 0))))
(4 + (3 + (2 + 1)))
(4 + (3 + 3))
(4 + 6)
(10)
------------------------------
(sum 4 0)
(sum 3 4)
(sum 2 7)
(sum 1 9)
(sum 0 10)
(10)

Summation once again!

def sqr(x: Int) = x * x
def cube(x: Int) = x * x * x

def sumSimple(a: Int, b: Int): Int =
if (a == b) a else a + sumSimple(a + 1, b)

def sumSquares(a: Int, b: Int): Int =
if (a == b) sqr(a) else sqr(a) + sumSquares(a + 1, b)

def sumCubes(a: Int, b: Int): Int =
if (a == b) cube(a) else cube(a) + sumCubes(a + 1, b)


Higher order functions

def identity(x: Int) = x
def sqr(x: Int) = x * x
def cube(x: Int) = x * x * x
def sum(f: Int=>Int, a: Int, b: Int): Int =
if (a == b) f(a) else f(a) + sum(f, a + 1, b)

println(sum(identity, 1, 10))
println(sum(sqr, 1, 10))
println(sum(cube, 1, 10))

”sum” is now a ”higher order” function! It’s ﬁrst parameter is a
function which maps an Int to an Int

The type ”Int” represents a simple Integer value. The type Int =>
Int represents a function which accepts an Int and returns an Int.

Anonymous functions

def sum(f: Int=>Int, a: Int, b: Int): Int =
if (a == b) f(a) else f(a) + sum(f, a + 1, b)

println(sum(x=>x, 1, 10))
println(sum(x=>x*x, 1, 10))
println(sum(x=>x*x*x, 1, 10))

We can create ”anonymous” functions on-the-ﬂy! x => x*x is a
function which takes an ”x” and returns x*x


Methods on collections: Map/Filter/Reduce

val a = List(1,2,3,4,5,6,7)

val b = a.map(x => x * x)
val c = a.filter(x => x < 5)
val d = a.reduce((x, y)=>x+y)

println(b)
println(c)
println(d)

Map applies a function on all elements of a sequence. Filter selects
a set of values from a sequence based on the boolean value
returned by a function passed as its parameter - both functions
return a new sequence. Reduce combines the elements of a
sequence into a single element.

More methods on collections

def even(x: Int) = (x % 2) == 0

val a = List(1,2,3,4,5,6,7)
val b = List(2, 4, 6, 5, 10, 11, 13, 12)

// are all members even?
println(a.forall(even))

// is there an even element in the sequence?
println(a.exists(even))

//take while the element is even -
//stop at the first odd element
println(b.takeWhile(even))

//partition into two sublists: even and odd
println(a.partition(even))

Block structure / Scope

def fun(x: Int) = {
val y = 1

val r = {
val y = 2
x + y
}
println(r)
println(x + y)
}

fun(10)

The ”y” in the inner scope shadows the ”y” in the outer scope


Nested functions / functions returning functions

// fun returns a function of type Int => Int

def fun():Int => Int = {
def sqr(x: Int):Int = x * x

sqr
}

val f = fun()
println(f(10))

def fun():Int=>Int says ”fun is a function which does not take
any argument and returns a function which maps an Int to an Int.
Note that it possible to have ”nested” function deﬁnitions.


Lexical Closure

def fun1():Int => Int = {
val y = 1
def add(x: Int) = x + y

add
}

def fun2() = {
val y = 2
val f = fun1()

// what does it print? 11 or 12
println(f(10))
}

fun2()

Lexical Closure

The function ”fun1” returns a ”closure”.
A ”closure” is a function which carries with it references to
the environment in which it was deﬁned.
When we call f(10), the ”add” function gets executed with
the environment it had when it was deﬁned - in this
environment, the value of ”y” is 1.


Lexical Closure with anonymous functions

def fun1(y: Int):Int=>Int =
x => x + y

def fun2() = {
val f = fun1(10)
println(f(2))
}

fun2()

”y” is now a parameter to fun1
”fun1” returns an anonymous function - there is absolutely no
diﬀerence between returning a ”named” function and
returning an anonymous function.


Simple closure examples

def sqr(x: Int) = x*x
def cube(x: Int) = x*x*x

def compose(f: Int=>Int, g: Int=>Int): Int=>Int =
x => f(g(x))

val f = compose(sqr, cube)
println(f(2))

val a = List(1,2,3,4)

println(a.map(f))

println(a.map(cube).map(sqr))


Simple closure examples

def removeLowScores(a: List[Int],
threshold: Int): List[Int] =
a.filter(score => score >= threshold)

val a = List(95, 87, 20, 45, 35, 66, 10, 15)

println(removeLowScores(a, 30))

The anonymous function "score => score >= threshold"
is the closure here.
How do you know that it is a closure? Its body uses a variable
”threshold” which is not in its local environment (the local
environment, in this case, is the parameter list consisting of a
single parameter ”score”)

Currying

Here is the deﬁnition from Wikipedia:

In mathematics and computer science, currying is the technique of
transforming a function that takes multiple arguments (or a tuple
of arguments) in such a way that it can be called as a chain of
functions, each with a single argument. It was originated by Moses
Schonﬁnkel and later re-discovered by Haskell Curry.

Let’s try to do this in Scala!


Currying - two argument functions

def addA(x: Int, y: Int): Int =
x + y

def addB(x: Int):Int=>Int =
y => x + y

val a = addA(10, 20)

val b = addB(10)(20)

println(a)
println(b)


Currying - three argument functions

def addA(x: Int, y: Int, z: Int) = x + y + z

def addB(x: Int): Int => (Int => Int) =
y => (z => x + y + z)

val a = addA(1, 2, 3)

val b = addB(1)(2)(3)

println(a)
println(b)

It is now easy to see how the idea can be generalized to N
argument functions!


Some List operations

val a = List(1,2,3)
val b = Nil
val c = List()
val d = 0::a
val e = 0::b

println(b)
println(c)
println(d) // List(0,1,2,3)
println(e) // List(0)

Nil and List() are both ”empty” lists
a::b returns a new list with ”a” as the ﬁrst item (the ”head”)
and remaining part b (called the ”tail”)

Pattern Matching

def fun(a: List[Int]) = a match {
case List(0, p, q) => p + q
case _ => -1
}

println(fun(List(0, 10, 20)))
println(fun(List(0, 1, 2, 3)))
println(fun(List(1, 10, 20)))

Pattern matching helps you ”pull-apart” complex datastructures
and analyse their components in an elegant way. In the above
case, the function will return the sum of the second and third
elements for any three element List of integers starting with a 0 -
the value -1 is returned for any other list


Pattern Matching

def fun(a: List[Int]) = a match {
case List(0, p, q) => p + q
case List() => -1
}

The compiler warns us when it detects that the match is not
”exhaustive”? Why do we get a warning in the above case?


Pattern Matching

def length(a: List[Int]):Int = a match {
case Nil => 0
case h::t => 1 + length(t)
}

println(length(List()))
println(length(List(10,20,30,40)))

A simple recursive routine for computing length of a list. If list is
empty, length is 0, otherwise, it is 1 + length of remaining part of
the list.

If we have a List(1,2,3) and if we match it with a ”case h::t”, ”h”
will be bound to 1 and ”t” will be bound to the ”tail” of the list,
ie, List(2,3)

Building our own data types with case classes

case class Circle(r: Int)

case class Rectangle(h: Int, w: Int)

val a = Circle(10)
println(a.r) // radius of the circle is 10

val b = Rectangle(5, 7)
println(b.h) // height of the rectangle is 5


Deﬁning an algebraic data type
Problem
How do we deﬁne a function ”area” which will compute the area of
a rectangle as well as a circle?

case class Circle(r: Int)

case class Rectangle(h: Int, w: Int)

def area(c: Circle) = 3.14 * c.r * c.r

val a = Circle(10)
println(area(a))

val b = Rectangle(10, 20)
//println(area(b)) -- does not work!



We need some way to tell the Scala type system that ”Circle” and
”Rectangle” are related. We do this using what is called an
”abstract class”

sealed abstract class Shape
case class Circle(r: Int) extends Shape
case class Rectangle(h: Int, w: Int) extends Shape

// a is a variable of type "Shape"
val a:Shape = Circle(10)

// b is a variable of type "Shape"
val b:Shape = Rectangle(10, 20)

// Note how a variable of type "Shape" can now
// hold either a Circle or a Rectangle


sealed abstract class Shape
case class Circle(r: Int) extends Shape
case class Rectangle(h: Int, w: Int) extends Shape

def area(a: Shape) = a match {
case Circle(r) => 3.14 * r * r
case Rectangle(h, w) => h * w
}

println(area(Circle(10)))
println(area(Rectangle(10, 20)))


A simple binary tree

sealed abstract class Tree

case class Node(value: Int,
left: Tree,
right: Tree) extends Tree

case class Leaf(value: Int) extends Tree

val a = Node(20, Leaf(10), Leaf(30))
val b = Node(40, Leaf(37), Leaf(50))
val c = Node(35, a, b)
println(c)


A simple binary tree

sealed abstract class Tree

case class Node(value: Int,
left: Tree,
right: Tree) extends Tree

case class Leaf(value: Int) extends Tree

def sumTree(t: Tree):Int = t match {
case Leaf(value) => value
case Node(value, left, right) => value +
sumTree(left) +
sumTree(right)
}

println(sumTree(Node(10, Leaf(5), Leaf(20))))

Immutability

val a = 1

a = 2 // does not work

a = a + 1 // does not work

var b = 1
b = 2
b = b + 1

”val” is immutable in the sense once you bind a value to a symbol
deﬁned as ”val”, the binding can’t be changed


Immutability

Let’s look at some Python list manipulation code:

a = [20, 30, 40, 50]
a.insert(0, 10)
print a # prints [10, 20, 30, 40, 50]

Python lists are ”mutable”. The ”insert” operation modiﬁes the
original list.


Immutability
Here is a similar program in Scala:

val a = List(20, 30, 40, 50)
val b = 10 :: a
println(a) // still List(20, 30, 40, 50)
println(b) // List(10, 20, 30, 40, 50)

val c = a ++ List(60)
println(c) // List(20, 30, 40, 50, 60)
println(a) // still List(20, 30, 40, 50)

Good Scala style encourages use immutable versions of data
structures like Lists. Mutable versions are also available though
their use is suggested only when the immutable version may not be
suitable because of reasons like eﬃciency.


Immutability: Structure sharing

This is one way to implement 4::b - but it requires creating a full
copy of the old list ”b” (red-colored cells 1, 2, 3 are copies of cells
containing the same values in ”b”) and adding 4 at the head
position. Scala does NOT do it this way because it is very
ineﬃcient.



This is how scala actually does 4::b. Note that the new list shares
all its elements (except 4) with the old list!



This is what happens when you append two lists A and B (using
++ in Scala). Note that the appending operation needs to copy
the whole of list A.



The ”structure sharing” idea can be used eﬃciently with tree
data structures.
For more info: http:
//en.wikipedia.org/wiki/Persistent_data_structure
Persistent vectors in Scala:
http://www.codecommit.com/blog/scala/
implementing-persistent-vectors-in-scala


Non-strict evaluation

def myIf(cond: Boolean, thenPart: Int, elsePart: Int) =
if (cond) thenPart else elsePart

println(myIf((1 < 2), 10, 20))

We are trying to write a function which behaves similar to the
built-in ”if” control structure in Scala ... does it really work
properly? Let’s try another example!



def fun1() = {
println("fun1")
10
}
def fun2() = {
println("fun2")
20
}
def myIf(cond: Boolean, thenPart: Int, elsePart: Int) =

println(myIf((1 < 2), fun1(), fun2()))



The behaviour of ”if” is ”non-strict”: In the expression ”if
(cond) e1 else e2”, if ”cond” is true e2 is NOT EVALUATED.
Also, if ”cond” is false, e1 is NOT EVALUATED.
By default, the behaviour of function calls in Scala is ”strict”:
In the expression ”fun(e1, e2, ..., en), ALL the expressions e1,
e2 ... en are evaluated before the function is called.
There is a way by which we can make the evaluation of
function parameters non-strict. If we deﬁne a functions as
"def fun(e1: => Int)", the expression passed as a
parameter to ”fun” is evaluated ONLY when its value is
needed in the body of the function. This is the ”call-by-name”
method of parameter passing, which is a ”non-strict” strategy.



def fun1() = {
println("fun1")
10
}
def fun2() = {
println("fun2")
20
}

def myIf(cond: Boolean, thenPart: => Int,
elsePart: => Int) =

println(myIf((1 < 2), fun1(), fun2()))



def hello() = {
println("hello")
10
}

def fun(x: => Int) = {
x + x
}

val t = fun(hello())
println(t)

How many times is the message ”hello” printed? Is there some
way to prevent unnecessary repeated evaluations?


Lazy val’s

def hello() = {
println("hello")
10
}

val a = hello()

The program prints ”hello” once, as expected. The value of the val
”a” will be 10.


Lazy val’s

def hello() = {
println("hello")
10
}

lazy val a = hello()

Strange, the program does NOT print ”hello”! Why? The
expression which assigns a value to a ”lazy” val is executed only
when that lazy val is used somewhere in the code!


Lazy val’s

def hello() = {
println("hello")
10
}

lazy val a = hello()

println(a + a)

Unlike a ”call-by-name” parameter, a lazy val is evaluated only
once and the value is stored! This is called ”lazy” or ”call by
need” evaluation.


Referential Transparency

If an expression can be replaced by its value without changing the
behaviour of the program, it is said to be referentially transparent

All occurrences of the expression 1+(2*3) can be replaced by
7 without changing the behaviour of the program.
Say the variable x (in a C program) has initial value 5. It is
NOT possible to replace all occurrences of the statement (x =
x + 1) with the value 6.


Pure Functions

var balance = 1000

def withdraw(amount: Int) = {
balance = balance - amount
balance
}

println(withdraw(100))
println(withdraw(100))

In what way is our ”withdraw” function diﬀerent from a function
like ”sin”?


Pure Functions

A pure function always computes the same value given the
same parameters; for example, sin(0) is always 0. It is
”Referentially Transparent”.
Evaluation of a pure function does not cause any observable
”side eﬀects” or output - like mutation of global variables or
output to I/O devices.


What is Functional Programming?

A style of programming which emphasizes composing your program
out of PURE functions and immutable data.

Theoretical foundation based on Alonzo Church’s Lambda Calculus

In order for this style to be eﬀective in the construction of real
world programs, we make use of most of the ideas seen so far
(higher order functions, lexical closures, currying, immutable and
persistent datastructures, lazy evaluation etc)


What is Functional Programming?

Questions to ask:
Is this the ”silver bullet”?
How practical is a program composed completely out of
”pure” functions?
What are the beneﬁts of writing in a functional style?


Is FP the silver bullet?

Of course, there are NO silver bullets!
(http://en.wikipedia.org/wiki/No_Silver_Bullet)
Writing software is tough - no one methodology or technique
is going to solve all your problems


How practical is a program composed completely out of
pure functions?

Very impractical - unless your aim is to ﬁght the chill by
heating up the CPU (which, by the way, is a ”side eﬀect”)
The idea is to write your program in such a way that it has a
purely functional core, surrounded by a few ”impure”
functions at the outer layers


Functional core + ”impure” outer layers
This example is simple and contrived, but it serves to illustrate the
idea. It is taken from the amazing book ”Functional Programming
in Scala” by Paul Chiusano and Runar Bjarnason.

case class Player(name: String, score: Int)

def declareWinner(p: Player) =
println(p.name + " is the winner!! ")

def winner(p1: Player, p2: Player) =
if(p1.score > p2.score) declareWinner(p1)
else declareWinner(p2)

winner(Player("Ram", 10), Player("John", 20))

Note that ”winner” is an impure function. We will now refactor it
a little!

Functional core + ”impure” outer layers

case class Player(name: String, score: Int)

def declareWinner(p: Player) =
println(p.name + " is the winner!! ")

def maxScore(p1: Player, p2: Player) =
if (p1.score > p2.score) p1 else p2

def winner(p1: Player, p2: Player) =
declareWinner(maxScore(p1, p2))

winner(Player("Ram", 10), Player("John", 20))

Now we have separated the computation from the display logic;
”maxScore” is a pure function and ”winner” is the impure function
at the ”outer” layer!

Beneﬁts of functional style - easy reuse, easy testing

What if we wish to ﬁnd out the winner among a set of N players?
Easy!

val players = List(Player("Ram", 10),
Player("John", 15),
Player("Hari", 20),
Player("Krishna", 17))

println(players.reduceLeft(maxScore))


FP as ”good software engineering”

Pure functions are easy to re-use as they have no ”context”
other than the function parameters. (Think about re-using
the ”winner” function in our ﬁrst version to compute the
winner among N players).
Pure functions are also easy to test. (Think about writing an
automated test for the ”winner” function in the ﬁrst version).
Think of FP as ”Good Software Engineering”!


Pure functions and the benefit of ”local reasoning”

If your function modifies an object which is accessible from
many other functions, the effect of calling the function is
much more complex to analyse because you now have to
analyse how all these other functions get affected by the
mutation.
Similarly, if the value computed by your function depends on
the value of an object which can be modified by many other
functions, it no longer becomes possible to reason about the
working of the function by only looking at the way the
function’s parameters are manipulated.
The evaluation of pure functions can be done by a very simple
process of ”substitution”.


Why mutability is tricky - an example

class ProtectedResource {
private Resource theResource = ...;
private String [] allowedUsers = ...;
public String[] getAllowedUsers() {
return allowedUsers;
}
public String currentUser() { ... }
public void useTheResource() {
for(int i = 0; i < allowedUsers.length; i++) {
if(currentUser().equals(allowedUsers[i]) {
... // access allowed; use it!
}
}
throw new IllegalAccessException();
}
}


This Java example is taken from Prof.Dan Grossman’s (University
of Washington) excellent coursera.org class on ”Programming
Languages”. Can you identify the problem with the code?


What if client code does this?

p.getAllowedUsers[0] = p.currentUser()
p.useTheResource()

The user can happily use the resource, even if he does not belong
to the group of ”allowed” users! The ﬁx is to return a copy of
”allowedUsers” in the function ”getAllowedUsers”.

What if we had used an immutable Scala list for representing
”allowedUsers”? This problem would never had occurred because
an attempt to modify ”allowedUsers” simply returns a new object
without in any way altering the original!


How it gets even more tricky in the context of concurrency

Multi core CPU’s are becoming commonplace
We need concurrency in our code to make eﬀective use of the
many cores
This throws up a whole bunch of complex problems
A function can no longer assume that nobody else is watching
when it is happily mutating some data



case class Date(var year: Int, var month: String,
var date: Int, var weekDay: String)

val d = Date(2013, "February", 23, "Saturday")
def changeDate(year: Int, month:String,
date: Int, weekDay: String) = {
d.year = year
d.month = month
d.date = date
d.weekDay = weekDay
}
def showDate() = println(d)
changeDate(2013, "February", 24, "Sunday")
showDate()



What happens if the functions changeDate() and showDate() run
as two independent threads on two CPU cores? Will showDate()
always see a consistent date value?

The traditional approach to maintaining correctness in the context
of multithreading is to use locks - but people who do it in practice
will tell you it is extremely tricky business.

It is claimed that one of the reasons for the resurgence of FP is the
emergence of multi-core processors and concurrency - it seems like
FP’s emphasis on pure functions and immutability is a good match
for concurrent programming.


Moving ahead ...

Join Prof.Grossman’s programming languages class on
Coursera: https://www.coursera.org/course/proglang
and learn more about functional programming using SML and
Racket
Join Prof.Odersky’s Functional Programming with Scala
course on Coursera for an introduction to both Scala and FP:
https://www.coursera.org/course/progfun
Watch the classic ”SICP” lectures by Abelson and Sussman:
http://groups.csail.mit.edu/mac/classes/6.001/
abelson-sussman-lectures/
Learn from ”HTDP” if you want something simpler:
http://htdp.org/


Books

”Programming in Scala, 2nd edition” by Martin Odersky, Bill
Venners, Lex Spoon. Perhaps the best introductory book on
the language.
”Functional Programming in Scala” by Paul Chiusano and
Runar Bjarnson - this will soon become a classic!
”Scala in Depth” by Joshua Suereth. This is an advanced
book on Scala
”Learn You a Haskell for Great Good” -
http://learnyouahaskell.com/. Scala programmers can
deﬁnitely beneﬁt from an understanding of Haskell - this is an
amazing book which will get you started with Haskell.
Many others: check out http://blog.typesafe.com/
week-of-scala-with-manning-publications


Other resources

”Learning functional programming without growing a
neckbeard” - http://marakana.com/s/post/1354/
learning_functional_programming_scala_video
”Scala Days 2012” Videos - http:
//skillsmatter.com/event/scala/scala-days-2012
”Out of the tar pit” - paper by Ben Mosely and Peter Marks
(google it)
”Persistent Data Structures and Managed References” - talk
by Rich Hickey (author of the Clojure programming language).
http://www.infoq.com/presentations/
Value-Identity-State-Rich-Hickey
”Can programming be liberated from the von Neumann style”
- by John Backus. http://www.thocp.net/biographies/
papers/backus_turingaward_lecture.pdf


Introduction to Functional Programming with Scala

Recommended

More Related Content

What's hot (20)

Viewers also liked (9)

Similar to Introduction to Functional Programming with Scala (20)

Introduction to Functional Programming with Scala