xke - programming paradigms & constructs

www.xebia.fr / blog.xebia.fr 1

Programming Paradigms & Constructs

Nicolas Demengel 2011, July 7th

Inspiration

Ruby, Io, Prolog, Scala,Erlang, Clojure, Haskell

Presentation of thoselanguages' main features

Exercises

Good start...but not enough!

Agenda

▶ Definitions and Examples

▶ Laptops allowed! (and encouraged)

Hands On!

▶ Choose one or several exercises to solve

Retrospective

▶ Comparison of solutions to some exercises

▶ Feelings about this or that feature/language

Material

Image for VirtualBox

Under your home:

▶ A folder for each language

▶ Simple examples of the languagefeatures

▶ HOWTO: instructions for compilingor interpreting the examples

▶ Solved exercises

▶ Unresolved problems

Describing A Language

Paradigm

▶ Style, concepts, abstractions, computation

Typing

▶ Constraints applied to values

▶ Operations provided for kinds of values

Constructs

▶ Decision constructs, data structures, advanced features

Syntax

▶ Rules, expressiveness

Ecosystem, …

▶ Out of scope

The Four Main Paradigms (1/4)

Imperative (Procedural)

▶ “First do this, next do that” => recipe

▶ Control structures and procedures

▶ Hard to organize

▶ Sensitive to small changes

Functional

▶ Theory of functions

▶ Given the same input, a function will produce the same output

▶ Immutable values, no side effect

▶ Higher-order functions

▶ Computation-oriented, suitable for data transformation

▶ Sometimes allows for proof of correctness

▶ Build a knowledge base with facts and rules

▶ Query the program and let it infer solutions

▶ Adapted to knowledge extraction => IA

Object-Oriented

▶ Grouping of data and behavior

▶ Message passing between objects

▶ Inheritance and polymorphism

▶ Domain-oriented, adapted to large and evolutive programs

A word about Prototype-based Programming

▶ A flavor of OO

▶ No classes, but still allows for inheritance via prototypes

▶ Naively put: a given object may be:

› an “instance” of another object (having it as prototype)

› a “class” of another object (being one of its prototypes)

▶ Very powerful! Example:

› a publication

› a newspaper

› Le Monde

› the July 7th edition of Le Monde

› my copy of the July 7th edition of Le Monde

› ...

A language is not always tied to a given paradigm

▶ You can write OO or Functional programs in C

Source: Pulp Fiction. Dir. Quentin Tarentino. Miramax Films. 1994. Film

▶ You can write OO programs with a Functional language and vice-versa

▶ Some languages mixes paradigms (C++, Scala, ...)

There are many other paradigms

▶ Most of them are flavors of the four main ones

Source: Pulp Fiction. Dir. Quentin Tarentino. Miramax Films. 1994. Film

The Main Languages That We Will See (1/2)

Erlang

▶ 1986, functional, dynamic, concurrent, fault-tolerant

▶ “Ericsson Language”

Haskell

▶ 1990, purely functional, static, lazy, “standard”

▶ 1995, object-oriented and functional, dynamic

The Main Languages That We Will See (2/2)

▶ 2003, object-oriented and functional, static, concurrent, on the JVM

Clojure

▶ 2007, functional, dynamic (code as data), concurrent, on the JVM

▶ Lisp dialect

A Word About Prolog (1/3)

What we are used to: assignment

▶ X = 10 assign 10 to variable X

Prolog: Unification

▶ X = 10 given an unknown X, make 10 and X match (binding)

▶ X = 10 given a yet known X, does it match 10? (matching)

What we are used to: assignment

▶ X = 10 assign 10 to variable X

Prolog: Unification

▶ X = 10 given an unknown X, make 10 and X match (binding)

▶ X = 10 given a yet known X, does it match 10? (matching)

▶ Basics of logic programming:

author(what_mad_universe, frederic_brown).author(the_eyre_affair, jasper_fforde).

genre(what_mad_universe, science_fiction).genre(the_eyre_affair, science_fiction).genre(the_eyre_affair, comic_novel).

writes_genre(Author, Genre) :-author(Book, Author),genre(Book, Genre).

|?- writes_genre(Who, science_fiction).

Who = frederic_brown ?Who = jasper_ffordeyes

|?- writes_genre(frederic_brown, What).

What = science_fictionyes

|?- writes_genre(jasper_fforde, comic_novel).

trueyes

|?- writes_genre(frederic_brown, comic_novel).

Pattern matching

smallest([LonelyNumber], LonelyNumber).

smallest([Head|Tail], SmallestSoFar) :- smallest(Tail, TailSmallest), SmallestSoFar is min(Head, TailSmallest).

smallest([3, 1, 7], S). % S = 1

% more matchingsmallest([], -1). % empty list

smallest(_, -1). % anything

smallest([_|[SecondItem|_]], -1) :-SecondItem > 5, % guard% ...

Typing (1/6)

Static

▶ Type-checking at compile-time

▶ C, Java, Scala, Erlang, Haskell

Typing (1/6)

Static

Dynamic

▶ Type-checking at run-time

▶ JavaScript, Ruby, Python, Clojure

Typing (1/6)

Static

Dynamic

▶ Type-checking at run-time

▶ JavaScript, Ruby, Python, Clojure

“Strong” / “Weak”

▶ What happens when types collide?

› 5 + 'a string'

▶ Can you subvert the type system?

› int an_int = (int) some_bytes;

▶ Few languages are totally “strongly” or “weakly” typed

Typing (2/6)

Type Inference

▶ Deduction of the type of an expression at compile time

▶ Less boiler-plate code: eases writing and reading

▶ Haskell:

▶ Scala:

double x = x * 2-- (Num a) => a -> a

double :: Integer -> Integerdouble x = x * 2-- Integer -> Integer

val ints = List(1, 2, 3)// List[Int]

def addNumbers(first: Int, second: Int) = first + second

// equivalent todef addNumbers(first: Int, second: Int): Int = first + second

Typing (2/6)

Type Inference

▶ Deduction of the type of an expression at compile time

▶ Less boiler-plate code: eases writing and reading

▶ Java:

List<Integer> ints = new ArrayList<Integer>();ints.add(1);ints.add(2);ints.add(3);

// Java 7List<Integer> ints = new ArrayList<>();

// java.util.Arrays, unmodifiableList<Integer> ints = asList(1, 2, 3);

// Guava, modifiableList<Integer> ints = newArrayList(1, 2, 3);

Typing (3/6)

Duck Typing

▶ Example (JS):

▶ If it walks like a duck and quacks like a duck, it’s a duck.

var car = { startEngine: function() { /* vroom */ }}

var blender = { startEngine: function() { /* ouch my fingers! */ }}

var thingsWithEngine = [car, blender];for (thing in thingsWithEngine) { thing.startEngine();}

Typing (3/6)

Duck Typing

▶ Example (JS):

▶ If it walks like a duck and quacks like a duck, it’s a duck.

And if it weighs the same as a duck...

var car = { startEngine: function() { /* vroom */ }}

var blender = { startEngine: function() { /* ouch my fingers! */ }}

var thingsWithEngine = [car, blender];for (thing in thingsWithEngine) { thing.startEngine();}

Typing (3/6)

A witch!

Source: Monthy Python and the Holy Grail. Dir. Terry Gillian, Terry Jones. EMI Films. 1974. Film

Typing (4/6)

Structural Typing: Type-safe Duck Typing

▶ Example (Scala):

class Car { def startEngine = println("vroom")}

class Blender { def startEngine = println("ouch my fingers!")}

type T = {def startEngine}

val thingsWithEngine = List[T](new Car(), new Blender())

thingsWithEngine.foreach(_.startEngine)

Typing (5/6)

Mixins

▶ Interfaces with partial implementations

› Ruby:

module Printabledef print_me

print self.my_representationend

class Thingextend OtherThing # only one extensioninclude Printableinclude Xxx # several inclusions

def my_representation# ...

endend

Typing (5/6)

Mixins

▶ Interfaces with partial implementations

› Scala:

trait Printable { def printMe() = println(myRepresentation) def myRepresentation(): String}

class Thing extends OtherThing with Printable with Xxx { def myRepresentation() = // ...}

Typing (6/6)

Out of scope but still interesting:

▶ Haskell's ad-hoc polymorphism

› type classes

› != subtype polymorphism or parametric polymorhism

› somehow equivalent to Java interfaces

A Word About Dynamic Languages

Concision, productivity

▶ No need to declare value types

Meta-programming

▶ Ruby, Python, Groovy: message interception (method_missing)

▶ Io, Clojure: Deferred argument evaluation

▶ Perfect for DSL

Downside: safety

▶ Test it!

Types we don't have in Java (1/5)

Symbols

▶ Sometimes called atoms or keywords: Prolog, Erlang

▶ Allows for naming things, for conveying meaning

▶ Clearer and more efficient than using strings or ints

› Ruby:

› Prolog:

› Clojure:

› ...

link_to("View Article", :controller => "articles", :action => "show")

friend(laurel, hardy)

(def product {:name "laser", :price 15})

▶ Scala:

val line = """Completed in 100ms (View: 25, DB: 75) | 200 OK [http://example.com?params=here]"""

val LogEntry = """Completed in (\d+)ms \(View: (\d+), DB: (\d+)\) \| (\d+) OK \[http://example.com(.*)\?.*""".r

▶ Scala:

val LogEntry(totalTime, viewTime, dbTime, responseCode, uri) = line

// resulttotalTime: String = 100viewTime: String = 25dbTime: String = 75responseCode: String = 200uri: String = ""

▶ Scala:

val LogEntry(totalTime, viewTime, dbTime, responseCode, uri) = line

// resulttotalTime: String = 100viewTime: String = 25dbTime: String = 75responseCode: String = 200uri: String = ""

Source.fromFile(logFile).getLines.collect { _ match { case LogEntry(tt, vt, dbt, rc, uri) => println(rc) case OtherPattern(var1, var2) => // … case _ => }}

▶ Scala:

val things = <things> <movie genre="action">Pirates of the Caribbean</movie> <movie genre="fairytale">Edward Scissorhands</movie> <book genre={ genre }>{ title }</book></things>

things \ "movie"

(things \\ "@genre")(1).text

things \ "movie" find { node: Node => (node \ "@genre").text == "action" }

▶ Scala:

val things = <things> <movie genre="action">Pirates of the Caribbean</movie> <movie genre="fairytale">Edward Scissorhands</movie> <book genre={ genre }>{ title }</book></things>

things \ "movie"

(things \\ "@genre")(1).text

things \ "movie" find { node: Node => (node \ "@genre").text == "action" }

(things \ "_").foreach { _ match { case <movie>{ title }</movie> => println(title) case <book>{ _ }</book> => println("book") case _ => }}

Ranges

▶ Ruby:

▶ Scala:

▶ Haskell:

▶ Clojure:

myArray[1..3] (1..3).to_a (1..21).each{|n| print n}

0 to 10 for (i <- 0 until 10) (0 until 10 by 5).start //.end .step

(range 1 21) ; actually this produces a sequence, more on that later

[1..5] [1,1.5..5] -- 1, 1.5, 2, 2.5, ...

Ranges

▶ Ruby:

▶ Scala:

▶ Haskell:

▶ Clojure:

Tuples

› (x, y) = (1, 2) {x, y} = {1, 2}

myArray[1..3] (1..3).to_a (1..21).each{|n| print n}

0 to 10 for (i <- 0 until 10) (0 until 10 by 5).start //.end .step

(range 1 21) ; actually this produces a sequence, more on that later

[1..5] [1,1.5..5] -- 1, 1.5, 2, 2.5, ...

Functions!

Higher-Order Functions

Produces or consumes other functions

▶ Ruby:

▶ Haskell:

Partial Application (~Currying)

▶ Erlang:

applyToArg f x = f(x)

def do_something(x, y) z = # ... lambda { |w| z + w }end

prod x y = x * ydouble = prod 2triple = prod 3

Functions and Sequences

Sequence: abstraction for “iterable things”▶ Scala:

▶ Clojure:

persons foreach { p => println(p.name) }

persons map { _.name }

persons filter { _.female }

=> (def m1 {:car 2, :doll 5, :table 1})

=> (def m2 {:bike 6, :doll 3})

=> (def m (merge-with + m1 m2)){:bike 6, :car 2, :doll 8, :table 1}

=> (reduce (fn [acc [_ qty]] (+ acc qty)) 0 m)17

Functional Programming Concepts

Immutability

▶ Empty list

▶ Nil.size(), count(nil)

Optional result

▶ Scala: Haskell:

map.get(aKey) match {case Some(x) => println(x)case None => println("Nope")

case aValue ofJust x -> ...Nothing -> ...

List Comprehension

Builds a list from other list(s)

▶ Erlang:

▶ Haskell:

▶ Clojure:

▶ Scala:

[ {X, Y} || X <- lists:seq(1, 4), X < 3, Y <- [5, 6] ].

% [{1,5},{1,6},{2,5},{2,6}]

[ (x, y) | x <- [1..4], x < 3, y <- [5, 6] ]

for (x <- 1 to 2; y <- List(5, 6); if x < 3)yield (x, y)

(for [ x [1,2,3,4], y [5,6], :when (< x 3)] (* x y))

Lazy evaluation

Delays the evaluation of an expression until its value is required

Increases performance

Allows for infinite sequences

▶ Haskell:

▶ Clojure

take 4 (dropWhile (< 34098) [0, 0.5 ..])-- [34098.0, 34098.5, 34099.0, 34099.5]

zip (take 5 (iterate (2*) 10)) (take 5 (iterate (/2) 10))-- [(10, 10.0), (20, 5.0), (40, 2.5), (80, 1.25), (160, 0.625)]

=> (take 5 (interpose :and(cycle [:x :k :e])))

(:x :and :k :and :e)

Lazy evaluation

An alternative to recursion

▶ Fibonacci:

(defn fib-pair [[a b]] [b (+ a b)])

(defn fibonacci[](map first

(iterate fib-pair [1 1])))

(take 10 (fibonacci))

; (1 1 2 3 5 8 13 21 34 55)

Concurrency

The Problem

▶ Shared resources

▶ Interactions between threads/processes

▶ Coordination

▶ Failures

Actors

An independent computational entity having a mailbox

It can send/receive messages (a-)synchronously

▶ Erlang (synchronous example):

loop() -> receive {From, "voiture"} -> From ! "car", loop();

{From, _} -> From ! "unknown word", loop()

Service = spawn(fun loop/0).

Service ! {self(), "voiture"},receive Translation -> Translationend.

Actors

An independent computational entity having a mailbox

It can send/receive messages (a-)synchronously

▶ Scala (asynchronous example):

class Service extends Actor {

def act() = { loop { react { case "voiture" => println("car") case _ => println("unknown word") } } }}

val service = new Service()service.startservice ! "voiture"

Software Transactional Memory

▶ Same approach as for databases (ACID)

▶ Wrap code accessing shared resources in a transaction

▶ The transaction sees values as they were when opening it

› Clojure:

– But also, in a nutshell:

(def i (ref 0))(def j (ref 0))

(dosync (alter i inc) (alter j dec))

Uncoordinated Coordinated

Synchronous Atom Reference (ref)

Asynchronous Agent

Let it Crash!

Erlang philosophy

Cheap processes instead of threads

Monitor your processes

Make them restart on failure

Refer to the “library” example

loop() -> process_flag(trap_exit, true), receive new -> register(revolver, spawn_link(fun russian_roulette_loop/0)), loop();

{'EXIT', From, Reason} -> self() ! new, loop() end.

Questions?

Exercises 1. Shopping Cart

Represent a catalog in the form of a list of tuples (article, quantity, price)

Write a function that returns the price of an article when present in the catalog or nothing/nil (depending on the language) otherwise

Using recursion, write a function that computes the total value of the catalog

Write a function that feeds a cart with items from the catalog until:

▶ either a maximum amount is reached for the cart value

▶ or the catalog is empty

Note: It might be simpler to represent the cart as a list of tuples (article, price) at first (with the same article occuring serveral times).

Amends your function so that it also returns the catalog minus the items that are in the cart

Using fold/reduce (depending on the language), writes a function that computes the total value of the cart

Using a list comprehension, write a function that takes a cart and returns it with a given tax applied to its items

Interesting candidates are: Erlang, Haskell, Clojure and Scala

Exercises 2. Tic-Tac-Toe

Write a pogram that accepts a tic-tac-toe board and returns whether there is a winner and, if it is the case, which one.

Depending on the language you choose, try to use the most different functional constructs you can think of:

▶ list processing with functions

▶ list comprehensions

▶ pattern matching and data deconstruction

▶ ...

Should you finish the first part early, modify your program so that it tells when there is a tie.

Interesting candidates are: Scala, Clojure, Erlang, Haskell

Alternatively, solve the problem with Prolog

Exercises 3. Erlang: Chat

The aim of this exercise is to write a chat application allowing for discussing between two terminals, based on the "translate_service", "russian_roulette" and "library" examples.

You can find template files for a server, a supervisor and a client.

The server will be in charge to log in clients and dispatch their messages. It will be built with Erlang OTP's generic server feature.

The supervisor will be in charge to restart the server should a crash occur.

The client will spawn a process when the user requests a connection to the server, so that this underlying process can handle messages from the server while the user use the console process to give orders to the client.

Exercises 4. Reverse Dependency Tree

Using Scala, write a program that can scan a local Maven repository for POM files and build a reverse dependency tree for a given module.

Use actors to load and parse files concurrently, and then send pieces of dependency model to a collecting actor.

Use Scala's XML capabilities to read POM files

Exercises 5. The Sleeping Barber

A barber has one barber chair and a waiting room with a number of chairs in it.

When the barber finishes cutting a customer's hair, he dismisses the customer and then goes to the waiting room to see if there are other customers waiting:

▶ if there are, he brings one of them back to the chair and cuts his or her hair,

▶ if there are no other customers waiting, he returns to his chair and sleeps in it.

Each customer, when he arrives, looks to see what the barber is doing:

▶ if the barber is sleeping, then he wakes him up and sits in the chair,

▶ if the barber is cutting hair, then he goes to the waiting room:

› if there is a free chair in the waiting room, he sits in it and waits his turn,

› if there is no free chair, then the customer leaves.

Write a program that print events occurring in the shop: customer arrival and departure (and why), barber state (cutting hair, sleeping), etc...

Use the following durations: a hair cut last 100 to 500 ms, a customer enters the shop every 1 to 450 ms. Stop sending customers after the 20th one.

xke - programming paradigms & constructs

Technology

xke spring boot

wg-s rd t xke...kh u,?\vnd[mie4i]yhq

vuqdzef - moradabad.nic.inmoradabad.nic.in/lohia1/lohia...

constructs, variables, hypotheses

personal constructs the evaluation of personal constructs

bpmn 2.0 analytical constructs - jarrar › courses › bpmn...

m‚å 'kjn dksgyh us % usg: xke hkkjrh ekfur foqofo|ky;] i

paradigms, broadcasting, key issues, analysis. paradigms

the neurodevelopmental constructs

open xke - google compute engine par jean-baptiste...

today's worship constructs

4.3 further programming - weebly · programming paradigms 1...

mta constructs

2011 xke - kanban in action

open xke kanban à grande échelle

open xke - poc d'une architecture distribuée de calculs...

constructs - cdn.filepicker.io

conditional constructs

the new brazilian paradigms the new brazilian paradigms

manager agile open xke 2013