paulking dlp

Dynamic Language Practices“Unlearning Java/C#”Dr Paul King - ASERT

Introduction …

• Developer practices– Well understood and documented for

traditional and agile approaches such

as Java, C++ and C# development

– But dynamic languages like Groovy,

Ruby, Python, Boo, JavaScript and

others change the ground rules

– Many of the rules and patterns we

have been taught no longer apply

(c) A

SE

RT

2006-2

009

… Introduction

• Traditional developer practice guidelines– Erich Gamma, Richard Helm, Ralph Johnson, John Vlissides (1995).

Design Patterns: Elements of Reusable Object-Oriented Software.

Addison-Wesley.

– Martin Fowler (1999). Refactoring: Improving the Design of Existing

Code. Addison-Wesley.

– Joshua Bloch (2001). Effective Java Programming Language

Guide. Prentice Hall.

– Robert C Martin (2002), Agile Software Development, Principles,

Patterns, and Practices. Prentice Hall.

• In the dynamic

language world,

are these guidelines

FACT or MYTH !!

(c) A

SE

RT

2006-2

009

Examples to ponder

• What does Immutability mean?– When even constants can be changed

• What does encapsulation

mean?– When I can peek at anything

under the covers

• How can I devise tests

at development time?– When my system can change in

unknown ways at runtime

• How can IDEs help me?– If I no longer spoon feed it static-typing information

or if my language now only allows checks at runtime

(c) A

SE

RT

2006-2

009

What do I mean by Dynamic Language?

• I prefer a loose definition

• One or more of:– Dynamic typing

• Greater polymorphism

– Metaprogramming

• Allow language itself to be dynamically changed

• Allow hooks into object lifecycle and method calls

• Open classes/monkey patching

– Work with code as easily as data

• Closures

• Higher-order programming

– Escape hatches

• Hooks for polyglot programming

(c) A

SE

RT

2006-2

009

Static vs Dynamic Typing …

• The Debate– Static vs dynamic typing

• Static: the type of each variable

(or expression) must be known

at compile time

like wearing a straight-jacket?

• Dynamic: type information is

known only at runtime

like tightrope walking with no net?

– Strong vs weak typing

• Strong: List<Integer> myList

• Weak: Object myList

– Type safety

• How is this provided if at all?

– Type inference

• Is this supported?

(c) A

SE

RT

2006-2

009

…Static vs Dynamic Typing …

• Static Typing Pros– Errors are often detected earlier and with better error

messages

– Code can sometimes be clearer – you don‟t need to infer

the types to understand the code – especially when

revisiting the code later

– Safer because certain kinds of injection hacks don‟t apply

– Code can be more declarative

– Better IDE support: refactoring, editing and other forms

of source processing support is often possible

– Better optimisations are often possible

– Often easier to understand a system from the outside

(“self-documenting” statically-typed APIs and interfaces)

– With generics support you can start to nail down even

complex cases

(c) A

SE

RT

2006-2

009

… Static vs Dynamic Typing …

• Dynamic Typing Pros– Speed development through duck-typing

and less boiler-plate code

– Clearer more concise code is easier to

read and maintain

– Allow more expressiveness through DSLs

– You should have comprehensive tests anyway, why not

cover off types as part of those tests

– Enforced healthy practices:

• Static language developers may get a false sense of

security and not design/test for runtime issues

• Less likely to neglect good documentation and/or good

coding conventions on the grounds that your static

types make everything “inherently” clear

(c) A

SE

RT

2006-2

009


• MYTH or TRUTH?

Static typing is just spoon feeding the

compiler. It represents the old-school way

of thinking and requires extra work while

providing no real value.

(c) A

SE

RT

2006-2

009


• An example

(c) A

SE

RT

2006-2

009

interface Reversible {def reverse()

}

class ReversibleString implements Reversible {def reverse() { /* */ }

}

class ReversibleArray implements Reversible {def reverse() { /* */ }

}

Reversible[] things = [new ReversibleString(), new ReversibleArray()

]

for (i in 0..<things.size()) {things[i].reverse()

}

def things = ["abc", [1, 2 ,3]]def expected = ["cba", [3, 2, 1]]assert things*.reverse() == expected

???

???

… Static vs Dynamic Typing

(c) A

SE

RT

2006-2

009

interface Reversible {def reverse()

}

class ReversibleString implements Reversible {def reverse() { /* */ }

}

class ReversibleArray implements Reversible {def reverse() { /* */ }

}

Reversible[] things = [new ReversibleString(), new ReversibleArray()

]

for (i in 0..<things.size()) {things[i].reverse()

}

With dynamically

typed languages,

there is no need to

explicitly declare the

types of variables or

the “protocols”

observed by our

objects:

Less code

Less declarative

Less IDE support

More testing

Less Robust?

def things = ["abc", [1, 2 ,3]]def expected = ["cba", [3, 2, 1]]assert things*.reverse() == expected


• MYTH or TRUTH?

Static typing is just spoon feeding the

compiler. It represents the old-school way

of thinking and requires extra work while

providing no real value.

(c) A

SE

RT

2006-2

009

...but not a total lie either...

...dynamic languages certainly

assist with removing duplication

and sometimes removing clutter

and boilerplate code...

Typing Approaches…

• Implicit vs Explicit interfaces– Inheritance too restrictive?

– Duck-typing too flexible?

Adapted from Interface-Oriented Design [2]

Shapedraw()

Rectangledraw()

set_sides()

Squaredraw()

set_sides()

Rectangledraw()

set_sides()

Squaredraw()

set_side()

EquilateralTriangledraw()

set_side()

Pistoldraw()

Menuset_sides()

Rectangledraw()

set_sides()

<<interface>>

Shapedraw()

<<interface>>

RegularPolygonset_side()

Squaredraw()

set_side()

EquilateralTriangledraw()

set_side()

(c) A

SE

RT

2006-2

009

I tend to use Explicit types

for major boundaries and

implicit types internally.

…Typing Approaches

• Inheritance hierarchies– Very clear intent but use sparingly

• Interface-oriented design– Use if it adds clarity & your language supports it

– If you do use it, stick to fine-grained interfaces

• Dynamic interface-oriented design– If your language doesn‟t support it natively you

can use a guard: is_a?, kind_of?, instanceof

• Chicken typing– Use a guard: responds_to?, respondsTo

• Duck typing– Use when flexibility is important but have appropriate tests in

place; e.g. you don‟t want to violate the Liskov Substitution

Principal[15] by not considering a refused bequest[13].

• AKA roll your own type safety

(c) A

SE

RT

2006-2

009

Source: Rick DeNatale

© David Friel

http://creativecommons.org/licenses/by/2.0/

http://flickr.com/people/frield

http://flickr.com/people/frield

Typing approaches and IDEs…

• Class A has a bit of duplication

(c) A

SE

RT

2006-2

009

class A {def helperdef make() {

helper.invoke('create')}def get() {

helper.invoke('read')}def change() {

helper.invoke('update')}def remove() {

helper.invoke('delete')}

}

… Typing approaches and IDEs …

• No problems, we can refactor out the dup

class B {def helperdef make() {

invoke('create')}def get() {

invoke('read')}def change() {

invoke('update')}def remove() {

invoke('delete')}private invoke(cmd) {

helper.invoke(cmd)}

}

(c) A

SE

RT

2006-2

009


• But we can do more using a dynamic

language by leveraging metaprogramming

• Which is a whole lot nicer?

• At the expense of IDE completion? …

class C {def helperdef commands = [

make: 'create',get: 'read',change: 'update',remove: 'delete'

]def invokeMethod(String name, ignoredArgs) {

helper.invoke(commands[name])}

}

(c) A

SE

RT

2006-2

009

...


(c) A

SE

RT

2006-2

009

for (x in [A, B, C]) {def o = x.newInstance()o.helper = new Dumper(name: "$x.name's helper")o.make()o.get()o.change()o.remove()

}

class Dumper {def namedef invokeMethod(String methodName, args) {

println "$name: called $methodName with $args"}

}

… Typing approaches and IDEs

• … At the expense of IDE completion?

(c) A

SE

RT

2006-2

009

...

But remember:

“clearly express intent”

Language features instead of Patterns …class RoundPeg {def radiusString toString() { "RoundPeg with radius $radius" }

}

class RoundHole {def radiusdef pegFits(peg) { peg.radius <= radius }String toString() { "RoundHole with radius $radius" }

}

def pretty(hole, peg) {if (hole.pegFits(peg)) println "$peg fits in $hole"else println "$peg does not fit in $hole"

}

def hole = new RoundHole(radius:4.0)(3..6).each { w -> pretty(hole, new RoundPeg(radius:w)) }

(c) A

SE

RT

2006-2

009

RoundPeg with radius 3 fits in RoundHole with radius 4.0

RoundPeg with radius 4 fits in RoundHole with radius 4.0

RoundPeg with radius 5 does not fit in RoundHole with radius 4.0

RoundPeg with radius 6 does not fit in RoundHole with radius 4.0

…Language features instead of Patterns…class SquarePeg {def widthString toString() { "SquarePeg with width $width" }

}

class SquarePegAdapter {def pegdef getRadius() { Math.sqrt(((peg.width/2) ** 2)*2) }String toString() {"SquarePegAdapter with width $peg.width (and notional radius $radius)"

}}

def hole = new RoundHole(radius:4.0)

(4..7).each { w ->pretty(hole, new SquarePegAdapter(peg: new SquarePeg(width: w))) }

(c) A

SE

RT

2006-2

009

SquarePegAdapter with width 4 (and notional radius 2.8284271247461903)

fits in RoundHole with radius 4.0


fits in RoundHole with radius 4.0


does not fit in RoundHole with radius 4.0


does not fit in RoundHole with radius 4.0

… Language features instead of Patterns …

SquarePeg.metaClass.getRadius ={ Math.sqrt(((delegate.width/2)**2)*2) }

(4..7).each { w -> pretty(hole, new SquarePeg(width:w)) }

SquarePeg with width 4 fits in RoundHole with radius 4.0

SquarePeg with width 5 fits in RoundHole with radius 4.0

SquarePeg with width 6 does not fit in RoundHole with radius 4.0

SquarePeg with width 7 does not fit in RoundHole with radius 4.0

Adapter Pattern

Do I create a whole

new class or just add

the method I need

on the fly?

(c) A

SE

RT

2006-2

009

Further reading: James Lyndsay, Agile is Groovy, Testing is Square

Adapter Pattern Verdict

• Dynamic languages can make it easier to

apply the adapter pattern to the extent that

its use may not even be apparent:– Express intent more clearly and improves readability

– Aids refactoring

– Can help with test creation

– Avoids class proliferation

– But you still need testing

(c) A

SE

RT

2006-2

009

abstract class Shape {}

class Rectangle extends Shape {def x, y, width, height

Rectangle(x, y, width, height) {this.x = x; this.y = y; this.width = width; this.height = height

}

def union(rect) {if (!rect) return thisdef minx = [rect.x, x].min()def maxx = [rect.x + width, x + width].max()def miny = [rect.y, y].min()def maxy = [rect.y + height, y + height].max()new Rectangle(minx, miny, maxx - minx, maxy - miny)

}

def accept(visitor) {visitor.visit_rectangle(this)

}}

class Line extends Shape {def x1, y1, x2, y2

Line(x1, y1, x2, y2) {this.x1 = x1; this.y1 = y1; this.x2 = x2; this.y2 = y2

}

def accept(visitor) {visitor.visit_line(this)

}}

class Group extends Shape {def shapes = []

def add(shape) { shapes += shape }

def remove(shape) { shapes -= shape }

def accept(visitor) {visitor.visit_group(this)

}}

class BoundingRectangleVisitor {def bounds

def visit_rectangle(rectangle) {if (bounds)bounds = bounds.union(rectangle)

elsebounds = rectangle

}

def visit_line(line) {def line_bounds = new Rectangle(line.x1, line.y1, line.x2 - line.y1, line.x2 - line.y2)if (bounds)bounds = bounds.union(line_bounds)

elsebounds = line_bounds

}

def visit_group(group) {group.shapes.each {shape -> shape.accept(this) }

}}

def group = new Group()group.add(new Rectangle(100, 40, 10, 5))group.add(new Rectangle(100, 70, 10, 5))group.add(new Line(90, 30, 60, 5))def visitor = new BoundingRectangleVisitor()group.accept(visitor)bounding_box = visitor.boundsprintln bounding_box.dump()

abstract class Shape {def accept(Closure yield) { yield(this) }

}

class Rectangle extends Shape {def x, y, w, hdef bounds() { this }def union(rect) {

if (!rect) return thisdef minx = [rect.x, x].min()def maxx = [rect.x + w, x + w].max()def miny = [rect.y, y].min()def maxy = [rect.y + h, y + h].max()new Rectangle(x:minx, y:miny, w:maxx - minx, h:maxy - miny)

}}

class Line extends Shape {def x1, y1, x2, y2def bounds() {

new Rectangle(x:x1, y:y1, w:x2-y1, h:x2-y2)}

}

class Group {def shapes = []def leftShift(shape) { shapes += shape }def accept(Closure yield) { shapes.each{it.accept(yield)} }

}

def group = new Group()group << new Rectangle(x:100, y:40, w:10, h:5)group << new Rectangle(x:100, y:70, w:10, h:5)group << new Line(x1:90, y1:30, x2:60, y2:5)def boundsgroup.accept{ bounds = it.bounds().union(bounds) }println bounds.dump()

… Language features instead of Patterns

(c) A

SE

RT

2006-2

009

See also Ruby Visitor [24]

Visitor Pattern

without closures

with closures

Visitor Pattern Verdict


apply the visitor pattern to the extent that

its use may not even be apparent:– Express intent more clearly and improves readability




(c) A

SE

RT

2006-2

009

Strategy Pattern

(c) A

SE

RT

2006-2

009

Source: http://nealford.com/

interface Calc {def execute(n, m)

}

class CalcByMult implements Calc {def execute(n, m) { n * m }

}

class CalcByManyAdds implements Calc {def execute(n, m) {

def result = 0n.times {result += m

}return result

}}

def sampleData = [[3, 4, 12],[5, -5, -25]

]

Calc[] multiplicationStrategies = [new CalcByMult(),new CalcByManyAdds()

]

sampleData.each {data ->multiplicationStrategies.each {calc ->assert data[2] == calc.execute(data[0], data[1])

}}

def multiplicationStrategies = [{ n, m -> n * m },{ n, m ->

def total = 0; n.times{ total += m }; total },{ n, m -> ([m] * n).sum() }

]

def sampleData = [[3, 4, 12],[5, -5, -25]

]

sampleData.each{ data ->multiplicationStrategies.each{ calc ->

assert data[2] == calc(data[0], data[1])}

}

Language features instead of Patterns…

(c) A

SE

RT

2006-2

009

Strategy Pattern

with interfaces

with closures

Strategy Pattern Verdict


apply the strategy pattern to the extent

that its use may not even be apparent:– Express intent more clearly and improves readability

– Closures open up whole new possibilities for solving

problems





(c) A

SE

RT

2006-2

009

… Language features instead of Patterns …

• Create new builder

• Call pretended methods

(html, head, ...)

• Arguments are Closures

• Builder code looks very

declarative but is ordinary

Groovy program code and

can contain any kind of

logic

• Builder pattern from the GoF at the syntax-level

• Represents easily any nested tree-structured data

NodeBuilder, DomBuilder,

SwingBuilder, AntBuilder, …

(c) A

SE

RT

2006-2

009

import groovy.xml.*def b = new MarkupBuilder()b.html {

head { title 'Hello' }body {

ul {for (count in 1..5) {

li "world $count"} } } }

… Language features instead of Patterns

(c) A

SE

RT

2006-2

009

<html><head>

<title>Hello</title></head><body>

<ul><li>world 1</li><li>world 2</li><li>world 3</li><li>world 4</li><li>world 5</li>

</ul></body>

</html>

import groovy.xml.*def b = new MarkupBuilder()b.html {

head { title 'Hello' }body {

ul {for (count in 1..5) {

li "world $count"} } } }

SwingBuilderimport java.awt.FlowLayoutbuilder = new groovy.swing.SwingBuilder()langs = ["Groovy", "Ruby", "Python", "Pnuts"]

gui = builder.frame(size: [290, 100],title: 'Swinging with Groovy!’) {

panel(layout: new FlowLayout()) {panel(layout: new FlowLayout()) {

for (lang in langs) {checkBox(text: lang)

}}button(text: 'Groovy Button', actionPerformed: {

builder.optionPane(message: 'Indubitably Groovy!').createDialog(null, 'Zen Message').show()

})button(text: 'Groovy Quit',

actionPerformed: {System.exit(0)})}

}gui.show()

Source: http://www.ibm.com/developerworks/java/library/j-pg04125/

(c) A

SE

RT

2006-2

009

JavaFX Script

Frame {title: "Hello World F3"width: 200content: Label {

text: "Hello World"}visible: true

}

(c) A

SE

RT

2006-2

009

Cheri::Swing

# requires JRubyrequire 'rubygems'require 'cheri/swing'include Cheri::Swing

@frame = swing.frame('Hello') {size 500,200flow_layouton_window_closing {|event| @frame.dispose}button('Hit me') {on_click { puts 'button clicked' }

}}@frame.show

(c) A

SE

RT

2006-2

009

AntBuilderdef ant = new AntBuilder()

ant.echo("hello") // let's just call one task

// create a block of Ant using the builder patternant.sequential {

myDir = "target/test/"mkdir(dir: myDir)copy(todir: myDir) {

fileset(dir: "src/test") {include(name: "**/*.groovy")

}}echo("done")

}

// now let's do some normal Groovy againfile = new File("target/test/AntTest.groovy")assert file.exists()

(c) A

SE

RT

2006-2

009

Builder Pattern Verdict

• The builder pattern in combination with

dynamic languages helps me:– Express intent more clearly and improves readability



– Tests are still important

(c) A

SE

RT

2006-2

009

Delegation Pattern ...

• Traditional approach to creating a class that is an

extension of another class is to use inheritance

– Clearest intent & simplest, clearest code for simple cases

class Person {private name, agePerson(name, age) {

this.name = namethis.age = age

}def haveBirthday() { age++ }String toString() { "$name is $age years old" }

}

class StaffMemberUsingInheritance extends Person {private salaryStaffMemberUsingInheritance(name, age, salary) {

super(name, age)this.salary = salary

}String toString() {

super.toString() + " and has a salary of $salary"}

}

(c) A

SE

RT

2006-2

009

… Delegation Pattern ...• Most common alternative is to use delegation

– Intention less clear (can be helped with interfaces)

– Overcomes multiple inheritance issues & inheritance abuse

class StaffMemberUsingDelegation {private delegateprivate salaryStaffMemberUsingDelegation(name, age, salary) {

delegate = new Person(name, age)this.salary = salary

}def haveBirthday() {

delegate.haveBirthday()}String toString() {

delegate.toString() + " and has a salary of $salary"}

}

(c) A

SE

RT

2006-2

009

… Delegation Pattern …• Downside of delegation is maintenance issues

– Refactoring overhead if we change the base class

– Meta-programming allows us to achieve inheritance

like behavior by intercepting missing method calls

(invokeMethod or method_missing)

– You could take this further with Groovy using named

parameters rather than the traditional positional

parameters shown here (future versions of Ruby may

have this too)

(c) A

SE

RT

2006-2

009

… Delegation Pattern …class StaffMemberUsingMOP {

private delegate

private salary

StaffMemberUsingMOP(name, age, salary) {

delegate = new Person(name, age)

this.salary = salary

}

def invokeMethod(String name, args) {

delegate.invokeMethod name, args

}

String toString() {

delegate.toString() + " and has a salary of $salary"

}

}

def p1 = new StaffMemberUsingInheritance("Tom", 20, 1000)

def p2 = new StaffMemberUsingDelegation("Dick", 25, 1100)

def p3 = new StaffMemberUsingMOP("Harry", 30, 1200)

p1.haveBirthday()

println p1

p2.haveBirthday()

println p2

p3.haveBirthday()

println p3

Tom is 21 years old and has a salary of 1000

Dick is 26 years old and has a salary of 1100

Harry is 31 years old and has a salary of 1200

(c) A

SE

RT

2006-2

009

class StaffMemberUsingLibrary {private salaryprivate personStaffMemberUsingLibrary(name, age, salary) {

person = new Person(name, age)this.salary = salarydef delegator = new Delegator(StaffMemberUsingLibrary, person)delegator.delegate haveBirthday

}String toString() {

person.toString() + " and has a salary of $salary"}

}

… Delegation Pattern• Going Further

–The example shown (on the previous slide) codes the

delegate directly but both Groovy and Ruby let you

encapsulate the delegation pattern as a library:

• Groovy: Delegator, Injecto; Ruby: forwardable, delegate

–But only if I don‟t want to add logic as I delegate

• E.g. If I wanted to make haveBirthday() increment salary

(c) A

SE

RT

2006-2

009

Delegation Pattern Verdict

• The delegation pattern can be expressed

more succinctly with dynamic languages:– Express intent more clearly and improves readability


– But don‟t forget the testing implications

(c) A

SE

RT

2006-2

009

Singleton Pattern…• Pattern Intent

– Ensure that only one instance of a class is created

– Provide a global point of access to the object

– Allow multiple instances in the future without affecting a singleton class's clients

• Static language discussion

points

– Need exactly one instance of a class

and a well-known controlled access

point

• Allows for lazy creation of instance

– More flexible than static class

variables and methods alone

• Permits refinement of operations and

representation through subclassing

– Reduces name space clutter

• Compared to using static approach

– Multi-threading implications

– Serializable implications

• need to have readResolve() method to

avoid spurious copies

– Garbage collection implications

• May need "sticky" static self-reference

– Need to be careful subclassing

• Parent may already create instance or be

final or constructor may be hidden

(c) A

SE

RT

2006-2

009

…Singleton Pattern…

• The details quickly get messy …

(c) A

SE

RT

2006-2

009

public final class Singleton {private static final class SingletonHolder {

static final Singleton singleton = new Singleton();}private Singleton() {}public static Singleton getInstance() {

return SingletonHolder.singleton;}

}

public class Singleton implements java.io.Serializable {public static Singleton INSTANCE = new Singleton();protected Singleton() {

// Exists only to thwart instantiation.}private Object readResolve() {

return INSTANCE;}

}


• State of the art approach in Java?– Use an IoC framework, e.g. Spring or Guice

(c) A

SE

RT

2006-2

009

import com.google.inject.*

@ImplementedBy(CalculatorImpl)interface Calculator {

def add(a, b)}

@Singletonclass CalculatorImpl implements Calculator {

private total = 0def add(a, b) { total++; a + b }def getTotalCalculations() { 'Total Calculations: ' + total }String toString() { 'Calc: ' + hashCode()}

}

class Client {@Inject Calculator calc// ...

}

def injector = Guice.createInjector()


• But it is easy using meta-programming– Old style

class Calculator {private total = 0def add(a, b) { total++; a + b }def getTotalCalculations() { 'Total Calculations: ' + total }String toString() { 'Calc: ' + hashCode()}

}

class CalculatorMetaClass extends MetaClassImpl {private final static INSTANCE = new Calculator()CalculatorMetaClass() { super(Calculator) }def invokeConstructor(Object[] arguments) { return INSTANCE }

}

def registry = GroovySystem.metaClassRegistryregistry.setMetaClass(Calculator, new CalculatorMetaClass())

(c) A

SE

RT

2006-2

009


• But it is easy using meta-programming

class Calculator {def total = 0def add(a, b) { total++; a + b }

}

def INSTANCE = new Calculator()Calculator.metaClass.constructor = { -> INSTANCE }

def c1 = new Calculator()def c2 = new Calculator()

assert c1.add(1, 2) == 3assert c2.add(3, 4) == 7

assert c1.is(c2)assert [c1, c2].total == [2, 2]

(c) A

SE

RT

2006-2

009


• And again with Ruby

class Aardvarkprivate_class_method :new@@instance = newdef Aardvark.instance@@instance

endend

(c) A

SE

RT

2006-2

009

class Aardvarkprivate_class_method :newdef Aardvark.instance@@instance = new if not @@instance@@instance

endend

module ThreadSafeSingletondef self.append_features(clazz)require 'thread'clazz.module_eval { private_class_method :new@instance_mutex = Mutex.newdef self.instance@instance_mutex.synchronize {@instance = new if not (@instance)@instance

}end

}end

end Source: http://c2.com/cgi/wiki?RubySingleton

…Singleton Pattern

• Or for Python– Classic class version (pre 2.2)

– Non-classic class version

class Borg:_shared_state = {}def __init__(self):

self.__dict__ = self._shared_state

(c) A

SE

RT

2006-2

009

class Singleton (object):instance = None def __new__(cls, *args, **kargs):

if cls.instance is None:cls.instance = object.__new__(cls, *args, **kargs)

return cls.instance

# UsagemySingleton1 = Singleton()mySingleton2 = Singleton()assert mySingleton1 is mySingleton2

Source: [10] and wikipedia

Singleton Pattern Verdict

• The singleton pattern can be expressed in

better ways with dynamic languages:– Express intent more clearly and improves readability


– But don‟t forgot testing implications

(c) A

SE

RT

2006-2

009

Pattern Summary

• Patterns can be replaced by language

features and libraries

• So patterns aren‟t important any more!

(c) A

SE

RT

2006-2

009

...

Refactoring Refactoring

• Out with the Old– Some refactorings no longer make sense

• In with the New– There are some new refactorings

• Times … they are a changin‟– Some refactorings are done differently

(c) A

SE

RT

2006-2

009

Encapsulate Downcast Refactoring

• Description– Context: A method returns an object that

needs to be downcasted by its callers

– Solution: Move the downcast to within the method

• Is there a point in a dynamic language?– Maybe but not usually

// Before refactoringObject lastReading() {

return readings.lastElement()}

// After refactoringReading lastReading() {

return (Reading) readings.lastElement()}

(c) A

SE

RT

2006-2

009

Introduce Generics Refactoring

• Description– Context: Casting is a runtime hack that allows

JVM to clean up a mess caused by a compiler

that couldn’t infer intent

– Solution: Use Generics to reveal intent to compiler

• Is there a point in a dynamic language?– Maybe but not usually

// Traditional Java styleList myIntList = new LinkedList()myIntList.add(new Integer(0))Integer result = (Integer) myIntList.iterator().next()

// Java generified styleList<Integer> myIntList2 = new LinkedList<Integer>()myIntList2.add(new Integer(0))Integer result2 = myIntList2.iterator().next()

// Groovier styledef myIntList3 = [0]def result3 = myIntList3.iterator().next()

(c) A

SE

RT

2006-2

009

Enabling a functional style …

• Consider the Maximum Segment Sum

(MSS) problem– Take a list of integers; the MSS is the maximum of the sums of

any number of adjacent integers

• Imperative solution:

def numbers = [31,-41,59,26,-53,58,97,-93,-23,84]

def size = numbers.size()def max = null(0..<size).each { from ->(from..<size).each { to ->def sum = numbers[from..to].sum()if (max == null || sum > max) max = sum

}}

println "Maximum Segment Sum of $numbers is $max"

(c) A

SE

RT

2006-2

009

… Enabling a functional style …

• A first attempt at a more functional style:

def numbers = [31,-41,59,26,-53,58,97,-93,-23,84]

def size = numbers.size()def max = [0..<size, 0..<size].combinations().collect{numbers[it[0]..it[1]].sum()

}.max()

println "Maximum Segment Sum of $numbers is $max"

(c) A

SE

RT

2006-2

009

… Enabling a functional style …

• An even more functional style– A known solution using functional composition:

mss = max º sum* º (flatten º tails* º inits)

– Where inits and tails are defined as follows:

(c) A

SE

RT

2006-2

009

assert letters.inits() == [['a'],['a', 'b'],['a', 'b', 'c'],['a', 'b', 'c', 'd']

]

letters = ['a', 'b', 'c', 'd']

assert letters.tails() == [['d'],

['c', 'd'],['b', 'c', 'd'],

['a', 'b', 'c', 'd']]

… Enabling a functional style

• An even more functional stylemss = max º sum* º (flatten º tails* º inits)

Notes:

– sum() is one-level flatten in Groovy, flatten() is recursive

– Metaprogramming allowed us to enhance all Lists

List.metaClass {inits{ (0..<delegate.size()).collect{ delegate[0..it] } }tails{ delegate.reverse().inits() }

}

(c) A

SE

RT

2006-2

009

def segs = { it.inits()*.tails().sum() }

def solve = { segs(it)*.sum().max() }

def numbers = [31,-41,59,26,-53,58,97,-93,-23,84]

println "Maximum Segment Sum of $numbers is ${solve numbers}"

Source: http://hamletdarcy.blogspot.com/2008/07/groovy-vs-f-showdown-side-by-side.html

Refactoring recipes with a curry base

• Static: Replace parameter with method– Refactoring [13]: Chapter 10

• Context– An object invokes a method, then passes the result as

a parameter for a method. The receiver can also

invoke this method.

• Solution– Remove the parameter and let the receiver invoke the

method.

• Dynamic solution– Partial Application: Currying

(c) A

SE

RT

2006-2

009

Replace parameter with method …

class Order {private int quantity, itemPriceOrder(q, p) {quantity = q; itemPrice = p}

double getPrice() {int basePrice = quantity * itemPriceint discountLevelif (quantity > 100) discountLevel = 2else discountLevel = 1double finalPrice = discountedPrice(basePrice, discountLevel)return finalPrice

}

private double discountedPrice(int basePrice, int discountLevel) {if (discountLevel == 2) return basePrice * 0.8return basePrice * 0.9

}}

println new Order(120, 5).price // => 480.0

(c) A

SE

RT

2006-2

009

… Replace parameter with method …


double getPrice() {int basePrice = quantity * itemPriceint discountLevelif (quantity > 100) discountLevel = 2else discountLevel = 1double finalPrice = discountedPrice(basePrice, discountLevel)return finalPrice

}

private double discountedPrice(int basePrice, int discountLevel) {if (discountLevel == 2) return basePrice * 0.8return basePrice * 0.9

}}

println new Order(120, 5).price // => 480.0

(c) A

SE

RT

2006-2

009



double getPrice() {int basePrice = quantity * itemPricedouble finalPrice = discountedPrice(basePrice)return finalPrice

}

private double discountedPrice(int basePrice) {if (getDiscountLevel() == 2) return basePrice * 0.8return basePrice * 0.9

}

private int getDiscountLevel() {if (quantity > 100) return 2return 1

}}println new Order(120, 5).price // => 480.0

(c) A

SE

RT

2006-2

009



double getPrice() {return discountedPrice(getBasePrice())

}

private double discountedPrice(int basePrice) {if (getDiscountLevel() == 2) return basePrice * 0.8return basePrice * 0.9

}

private int getBasePrice() {quantity * itemPrice

}



(c) A

SE

RT

2006-2

009



double getPrice() {return discountedPrice()

}

private double discountedPrice() {if (getDiscountLevel() == 2) return getBasePrice() * 0.8return getBasePrice() * 0.9

}

private int getBasePrice() {quantity * itemPrice

}



(c) A

SE

RT

2006-2

009

… Replace parameter with method


double getPrice() {if (getDiscountLevel() == 2) return getBasePrice() * 0.8return getBasePrice() * 0.9

}

private getBasePrice() {quantity * itemPrice

}

private getDiscountLevel() {if (quantity > 100) return 2return 1


(c) A

SE

RT

2006-2

009

Note the now small

parameter lists

Some functional style …


def discountedPrice = { basePrice, discountLevel ->discountLevel == 2 ? basePrice * 0.8 : basePrice * 0.9 }

def price = {int basePrice = quantity * itemPricedef discountLevel = (quantity > 100) ? 2 : 1discountedPrice(basePrice, discountLevel) }

}println new Order(120, 5).price() // => 480.0

(c) A

SE

RT

2006-2

009

… Some functional style …


def basePrice = { quantity * itemPrice }

def discountLevel = { quantity > 100 ? 2 : 1 }

def price = {discountLevel() == 2 ? basePrice() * 0.8 : basePrice() * 0.9 }


(c) A

SE

RT

2006-2

009

… Some functional style …




def discountedPrice = { basePrice, discountLevel ->discountLevel == 2 ? basePrice * 0.8 : basePrice * 0.9 }

def price = {discountedPrice.curry(basePrice()).curry(discountLevel()).call() }


(c) A

SE

RT

2006-2

009

… Some functional style




def discountedPrice(basePrice, discountLevel) {discountLevel == 2 ? basePrice * 0.8 : basePrice * 0.9

}

def price = {this.&discountedPrice.curry(basePrice()).curry(discountLevel()).call()

}


(c) A

SE

RT

2006-2

009

Closure Refactoring …

• Complex code involving closures

// Before refactoringdef phrase = "The quick brown fox jumps over the lazy dog"def result = phrase.toLowerCase().toList().

findAll{ it in "aeiou".toList() }. // like WHERE ...groupBy{ it }. // like GROUP BY ...findAll{ it.value.size() > 1 }. // like HAVING ...sort{ it.key }.reverse(). // like ORDER BY ...collect{ "$it.key:${it.value.size()}" }.join(", ")

println result

(c) A

SE

RT

2006-2

009

… Closure Refactoring …

• Possible Refactoring

// Refactored helper closuresdef lowercaseLetters = phrase.toLowerCase()def vowels = { it in "aeiou".toList() }def occursMoreThanOnce = { it.value.size() > 1 }def byReverseKey = { a, b -> b.key <=> a.key }def self = { it }def entriesAsPrettyString = { "$it.key:${it.value.size()}" }def withCommaDelimiter = ", "

// Refactored main closureprintln lowercaseLetters.

findAll(vowels).groupBy(self).findAll(occursMoreThanOnce).sort(byReverseKey).collect(entriesAsPrettyString).join(withCommaDelimiter)

(c) A

SE

RT

2006-2

009

… Closure Refactoring

# Add group_by to the Array classclass Array

def group_bygroup_hash = {}uniq.each do |e|

group_hash[e] = select { |i| i == e }.sizeendgroup_hash

endend

# Before refactoringphrase = "The quick brown fox jumps over the lazy dog"puts phrase.downcase.

scan(/[aeiou]/). # like WHERE ...group_by. # like GROUP BY ...select { |key, value| value > 1 }. # like HAVING ...sort.reverse. # like ORDER BY ... DESCcollect{ |key, value| "#{key}:#{value}" }.join(', ')

# Refactored versionlowercase_letters = phrase.downcasevowels = /[aeiou]/occurs_more_than_once = lambda { |key,value| value > 1 }entries_as_pretty_string = lambda { |key, value| "#{key}:#{value}" }

puts lowercase_letters.scan(vowels).group_by.select(&occurs_more_than_once).sort.reverse.collect(&entries_as_pretty_string).join(', ')

(c) A

SE

RT

2006-2

009

• This is the end of the talk

(c) ASERT 2006-2009


• NO!

• We haven’t questioned some fundamental principles yet!

(c) ASERT 2006-2009

Open-Closed Principle...

• Fundamental rule to make

your software flexible– Many other OOP principles, methodologies and

conventions revolve around this principle

• Open-Closed Principle (OCP) states:

• Software entities should be open for

extension, but closed for modification

• References– Bertrand Meyer, Object Oriented Software

Construction (88, 97)

– Robert C Martin, The Open-Closed Principle

– Craig Larman, Protected Variation: The Importance of

Being Closed

(c) A

SE

RT

2006-2

009

Picture source: http://www.vitalygorn.com

...Open-Closed Principle...

• Following the Rules– Encapsulation: Make anything that shouldn‟t be seen

private

– Polymorphism: Force things to be handled using

abstract classes or interfaces

• When making class hierarchies:– Make anything that shouldn‟t be open final

– Polymorphism: Always follow weaker pre stronger

post (object substitutability in the static world)

• When making changes that might break

existing clients– Add a new class into the hierarchy

– No compilation of existing code! No breakages!

(c) A

SE

RT

2006-2

009

Optional

FactoryOptional

Factory


• Part I: If I violate the Open part of OCP in

static languages

– I can‟t make the future enhancements I need

• Part II: If I violate the Closed part of OCP

– Client applications using my libraries might

break or require recompilation in the future

(c) A

SE

RT

2006-2

009

Extendible

Class A

Class A‟

Class A

User

Class A‟

User

Interface

Class A Class A‟

Class A

User

Class A‟

User

...


• Part I: Consider Java‟s String class– Has methods to convert to upper or

lower case but no swapCase() method?

– Traditionally, consider creating an

EnhancedString class using inheritance?

– I can‟t: String is immutable and final

• In OCP terms, it is not open for extension

• Dynamic language solution: open classes

String.metaClass.swapCase = {delegate.collect{ c ->c in 'A'..'Z' ?c.toLowerCase() :c.toUpperCase()

}.join()}assert "Foo".swapCase() == "fOO"

(c) A

SE

RT

2006-2

009

#lightopen Stringtype System.String with

member x.swapCase =seq { for letter in x.ToCharArray() do

if (System.Char.IsLower(letter))then yield System.Char.ToUpper(letter)else yield System.Char.ToLower(letter)

}printfn "result: %A" "Foo".swapCase

...


• Part II: Violating OCP (see [15])

class Square {def side

}class Circle {

def radius}

class AreaCalculator {double area(shape) {

switch (shape) {case Square:

return shape.side * shape.sidecase Circle:

return Math.PI * shape.radius ** 2}

}}

(c) A

SE

RT

2006-2

009


• What‟s wrong– If we wanted to introduce a Triangle, the

AreaCalculator would need to be recompiled

– If we wanted to change the order the shape

information was displayed, there might be many

changes to make

def shapes = [new Square(side: 3),new Square(side: 2),new Circle(radius: 1.5)

]

def calc = new AreaCalculator()shapes.sort().each {s ->

println "Area of $s.class.name is ${calc.area(s)}"}

(c) A

SE

RT

2006-2

009


• Dynamic shapes– No issue with adding Triangle but sorting is an issue *

class Square {private sidedouble area() { side ** 2 }

}class Circle {

private radiusdouble area() { Math.PI * radius ** 2 }

}

def shapes = [new Square(side:3),new Square(side:2),new Circle(radius:1.5)

]// unsorteddef prettyPrint = { s ->

println "Area of $s.class.name is ${s.area()}" }shapes.each(prettyPrint)

Area of Square is 9.0


Area of Circle is 7.0685834705770345

Note: Duck-type

polymorphism

instead of

inheritance

polymorphism,

i.e. no base Shape

(abstract) class or

interface.

Hmm… what are

the testing

implications when

I add Triangle?

(c) A

SE

RT

2006-2

009

* Our abstractions never designed sorting to be

one of the things open for extension. See [15].

...


• Dynamic sorting using Closures– As long as we are happy having our sort “code”

within a closure we have complete freedom

– Sometimes representing our abstractions within

classes is appropriate; many times closures will do

// sorted by areadef byArea = { s -> s.area() }shapes.sort(byArea).each(prettyPrint)


Area of Circle is 7.0685834705770345


// sorted circles before squares but otherwise by areadef byClassNameThenArea = { sa, sb ->

sa.class.name == sb.class.name ?sa.area() <=> sb.area() :sa.class.name <=> sb.class.name

}shapes.sort(byClassNameThenArea).each(prettyPrint)

Area of Circle is 7.06858...



Note: Make sure your

closures are testable.

(c) A

SE

RT

2006-2

009

...


• Instead of worrying about– Rigidity

– Fragility

– Immobility

(Because they can be easily gotten

around even if you don‟t try to apply OCP)

• We must worry about– Duplication

– Harder refactoring or testing

– Feature interaction

• And of course OCP then leads to ...– Liskov Substitution Principle, Single Responsibility

Principle, Dependency Inversion Principle, ...

(c) A

SE

RT

2006-2

009


• “Clean code” [23] states it this way: – Procedural code (i.e. using data structures) makes it

easy to add new functions without changing existing

data structures but when new data structures are

added, all existing procedures may need to change

– OO code makes it easy to add new classes without

changing existing functions but when new functions

are added, all classes must change

• Recommendation?– Choose procedural or OO approach based on

whether anticipated evolution of system involves

functions or data

– Use Visitor (dual dispatch) Pattern if you think both

functions and data might change

(c) A

SE

RT

2006-2

009


class Square {double side

}

class Rectangle {double height, width

}

class Circle {double radius

}

class Geometry {def area(shape) {switch (shape) {case Square: return shape.side ** 2case Rectangle: return shape.height * shape.widthcase Circle: return PI * shape.radius ** 2

}}

}

(c) A

SE

RT

2006-2

009

Can add perimeter() here without shape classes changing but if we

added a Triangle, area(), perimeter() etc. would need to change.


interface Shape {double area()

}

class Square implements Shape {double sidedouble area() { side ** 2 }

}

class Rectangle implements Shape {double height, widthdouble area() { height * width }

}

class Circle implements Shape {double radiusdouble area() { PI * radius ** 2 }

}

(c) A

SE

RT

2006-2

009

If we add perimeter() here, each

shape class must change but we can

add new shapes with no changes


class Square {double sidedouble area() { side ** 2 }

}

class Rectangle {double height, widthdouble area() { height * width }

}

class Circle {double radiusdouble area() { PI * radius ** 2 }

}

(c) A

SE

RT

2006-2

009

We can easily add perimeter() here

but for any code requiring the perimeter()

method to exist, we should test that code

with all shapes.

...Open-Closed Principle...• “Clean code” [23] recommendation:

– Choose procedural or OO approach or Visitor

• Agile variation:

– Defer moving to complicated solutions, e.g. Visitor

Pattern, but have in place sufficient tests so that you

can confidently refactor to use one later if needed

• Dynamic language variation:

– You won‟t need an explicit visitor (more on this later)

– Duck typing lets you add functions or data without

changing existing classes at the expense of static

type safety

– If you add a function you might need additional tests

for each class associated with that function

– If you add a new class you might need additional

tests for each function associated with that class

(c) A

SE

RT

2006-2

009

...Open-Closed Principle• Sometimes referred to as the Expression

Problem:– Independently Extensible Solutions to the Expression Problem

by Matthias Zenger Martin Odersky

– http://www.scala-lang.org/docu/files/

IC_TECH_REPORT_200433.pdf

(c) A

SE

RT

2006-2

009


(c) ASERT 2006-2009


• NO!

• We haven’t looked at Advanced Topics yet including Aspects, Testability, Feature Interaction and DSLs!

(c) ASERT 2006-2009

What & why of dependency injection?

• Construction by hand

• Factory pattern

• Service locator

• Dependency injection

(c) A

SE

RT

2006-2

009

Loose

Coupling

Flexibility

Testability

Reusability

Lifecycle

Control

Central

Control

Dependency Injection vs Metaprogramming …

• Dependency Injection– Dependencies are explicitly declared and allowed to

be set externally (typically via constructor or setters)

• Transparent injection of dependent service objects into other

service objects by a controlling container hence the name

inversion of control

• Why?– More flexible

• Central configuration of service objects

– Can be less work to do

• Service objects are instantiated by the dependency injection

framework

– Improves testability

– Improves reusability

– Improved lifecycle control

(c) A

SE

RT

2006-2

009

… Dependency Injection vs Metaprogramming …

• Without dependency injection

class PrintService {

private printer = new PhysicalPrinter('Canon i9900')

}

class PrintServiceTest extends GroovyTestCase {

def testPrintService() {

def printService = new PrintService()

printService.print()

// go to the printer and fetch the page

}

}

class PrintService

def initialise

@printer =

PhysicalPrinter.new('Canon i9900')

end

end

class PrintServiceTest < Test::Unit

def test_print_service

print_service = PrinterService.new

print_service.print

# go to the printer and fetch the page

end

end

(c) A

SE

RT

2006-2

009


• Improves testability

class TestablePrintService {

def printer

}

class TestablePrintServiceTest extends GroovyTestCase {

def testPrintService() {

def printService = new TestablePrintService()

printService.printer = new StubPrinter()

printService.print()

//...

}

}class TestablePrintService

attr_accessor :printer

end

class TestablePrintServiceTest < Test::Unit::TestCase


print_service = TestablePrintService.new

print_service.printer = StubPrinter.new

print_service.print

#...

end

end

(c) A

SE

RT

2006-2

009


• Improves reuse

class ReusablePrintService

attr_accessor :printer

# interesting printer logic below

# ...

end

# One customer may use a network PostScript printer

class PostScriptPrinter

# ...

end

# The next customer may use a local Windows printer.

class LocalWindowsPrinter

# ...

end

# No change is required to the ReusablePrintService!

class UnreusablePrintService

def initialise(printer_name)

@printer = PostScriptPrinter.new(ip_address)

end

# printer logic below

# ...

end

# What if I don’t want to use a Physical Printer?

Without DI

(c) A

SE

RT

2006-2

009


• Printer dependency can be redefined in

test code

• Benefits of Dependency Injection only

become apparent in larger projects– where we need lots of reuse and centralized

configuration of services

• In smaller projects, this is just a burden

# Create stub/mock services in a separate module

Printer = Stubs::Printer

class PrintServiceTest < Test::Unit


# The Printer instantiated by this class

# is now a stub!

print_service = UntestablePrintService.new

print_service.do_something_interesting_with_printer

end

end

(c) A

SE

RT

2006-2

009


• Improved lifecycle control– Guice and Spring let you declaratively specify

Singletons

• But see the discussion on Singleton Patterns

(c) A

SE

RT

2006-2

009

… Dependency Injection vs Metaprogramming…

• Downsides.....– Code becomes more complex

– Constructors (and/or interfaces) contain noise

– Potentially need to introduce (many) new interfaces

– There is some magic

• Do I really need this?– In Java, for medium to large projects, the consensus

is yes!

– But Ruby and Groovy are extremely dynamic

– We can rely on duck typing

– We can change classes at runtime

– We can intercept object creation

(c) A

SE

RT

2006-2

009


• But should we abandon using these types

of frameworks altogether?– Not for large systems

– Yes we can step around many of the problems which

make such frameworks especially attractive

– But they typically require us to use a very beneficial

organizational discipline across our dependencies

– A great example: Dependency Injection in Ruby

http://onestepback.org/index.cgi/Tech/Ruby/DependencyInjectionInRuby.rdoc

(c) A

SE

RT

2006-2

009


• Traditional

• Service

Locator

class WebAppdef initialize

@quotes = StockQuotes.new@authenticator = Authenticator.new@database = Database.new@logger = Logger.new@error_handler = ErrorHandler.new

end# ...

end

def create_applicationlocator = {}locator[:logger] = Logger.newlocator[:error_handler] = ErrorHandler.new(locator)locator[:quotes] = StockQuotes.new(locator)locator[:database] = Database.new(locator)locator[:authenticator] = Authenticator.new(locator)locator[:webapp] = WebApp.new(locator)

end

class StockQuotesdef initialize(locator)

@error_handler = locator[:error_handler]@logger = locator[:logger]

end# ...

end

(c) A

SE

RT

2006-2

009

… Dependency Injection vs Metaprogramming

• Using DIdef create_application

container = DI::Container.newcontainer.register(:logfilename) { "logfile.log" }container.register(:db_user) { "jim" }container.register(:db_password) { "secret" }container.register(:dbi_string) { "DBI:Pg:example_data" }

container.register(:app) { |c|app = WebApp.new(c.quotes, c.authenticator, c.database)app.logger = c.loggerapp.set_error_handler c.error_handlerapp

}

container.register(:quotes) { |c|StockQuotes.new(c.error_handler, c.logger)

}

container.register(:authenticator) { |c|Authenticator.new(c.database, c.logger, c.error_handler)

}

container.register(:database) { |c|DBI.connect(c.dbi_string, c.db_user, c.db_password)

}

container.register(:logger) { |c| Logger.new(c.logfilename) }container.register(:error_handler) { |c|

errh = ErrorHandler.newerrh.logger = c.loggererrh

}end

(c) A

SE

RT

2006-2

009

Dependency Injection Framework Verdict

• DI frameworks aren‟t strictly needed when

using dynamic languages

• DI frameworks provide no value for

dynamic languages!

(c) A

SE

RT

2006-2

009

Flexible Mocks …

• Mocking is an important aspect to

interaction-based testing

• But with dynamic languages, you don‟t

always need a complex mocking package:– Stubs

– Closures

– Maps

– Meta-programming

• Same goes for Behavior-Driven Design

packages– But that‟s another story

(c) A

SE

RT

2006-2

009

class ClosureMocking {String methodOne() {

methodTwo()}String methodTwo() {

"television"}

}

def emc = new ExpandoMetaClass(ClosureMocking,true)emc.methodTwo = {-> "radio" }emc.initialize()

assert "radio" == new ClosureMocking().methodOne()

class MyApp {def factorydef loggerdef doBusinessLogic(param) {

def myObj = factory.instancemyObj.doSomething(param)myObj.doSomethingElse(param)logger.log('Something done with: ' + param)

}}

def factory = [instance: businessObj]

def logger = new Expando()logger.log = { msg ->

assert msg == 'Something done with: ' + param }

… Flexible Mocks

<= „mock‟ using meta-programming

Traditionally, we‟d

use mocks for factory

and logger, here we use

Expandos and Maps =>

(c) A

SE

RT

2006-2

009

Aspects vs Meta-programmingclass TimingInterceptor extends TracingInterceptor {

private beforeTimedef beforeInvoke(object, String methodName, Object[] arguments) {

super.beforeInvoke(object, methodName, arguments)beforeTime = System.currentTimeMillis()

}public Object afterInvoke(Object object, String methodName,

Object[] arguments, Object result) {super.afterInvoke(object, methodName, arguments, result)def duration = System.currentTimeMillis() - beforeTimewriter.write("Duration: $duration ms\n")writer.flush()return result

}}

def proxy = ProxyMetaClass.getInstance(util.CalcImpl.class)proxy.interceptor = new TimingInterceptor()proxy.use {

assert 7 == new util.CalcImpl().add(1, 6)}// =>// before util.CalcImpl.ctor()// after util.CalcImpl.ctor()// Duration: 0 ms// before util.CalcImpl.add(java.lang.Integer, java.lang.Integer)// after util.CalcImpl.add(java.lang.Integer, java.lang.Integer)// Duration: 16 ms

(c) A

SE

RT

2006-2

009

Checked vs unchecked exceptions

• From [1]:– “Statically checked exceptions lead to leaking

abstractions. This happens when a component

interacts with another component that declares to

throw unrelated exceptions that stem from a third

component but which can only be handled on a

higher level.”

• Is the debate less relevant now?– Groovy auto converts checked exceptions into

unchecked

– Techniques such as autoboundaries can be used to

overcome issues with existing code that abuses

checked exceptions

See: http://code.google.com/p/agileplugins/

(c) A

SE

RT

2006-2

009

Immutability

• Why?– Simpler classes

– Good for certain kinds of domain objects

– No issues with concurrency

(c) A

SE

RT

2006-2

009

Strict Class Immutability Rules (Bloch #13)

1. No mutator methods (that modify state)

2. Ensure methods can‟t be overridden– Prevents careless or malicious subclasses from

compromising the immutable behavior of the class

3. Make all fields private and final– Clearly expresses your intentions in an enforceable

manner (not strictly required but often desirable)

4. Ensure exclusive access to any mutable

components– Ensure that clients of the class cannot obtain

references to any mutable object fields and make

defensive copies where needed in constructors,

accessors, and readObject methods

(c) A

SE

RT

2006-2

009

Immutability Discussion...

• Java– Strict Immutable

– Read-only accessors

– Read-only super classes

• Ruby– Consider freeze

– But watch out for thaw via dup

• Groovy– Even strict can be broken

– Read-only through meta-programming

(c) A

SE

RT

2006-2

009

...Immutability Discussionstr = 'A simple string. '

str.freeze

begin

str << 'An attempt to modify.'

rescue => err

puts "#{err.class} #{err}"

end

# Output: TypeError can't modify frozen string

str2 = str.dup

puts "#{str.frozen?} #{str2.frozen?}"

# Output: true false

str2 << 'An attempt to modify.'

str2.freeze

puts str2

# Output: A simple string. An attempt to modify.

str = 'Original string - '

str.freeze

str += 'attachment'

puts str

# Output: Original string - attachment

Adapted from

Mutable and Immutable Objects

rubylearning.com

(c) A

SE

RT

2006-2

009

Feature Interaction...• An unintended side-effect of adding features

iteratively [18]– Not helped by emergent behavior

– Nor by dynamic systems which can add behavior in ad-hoc ways

• Common problem in other domains, e.g.

services offered by PABX systems, Intelligent

Networks, Mobile Phones– They have been trying to tackle the problem for some time and

have some success

• From „http://c2.com/cgi/wiki?FeatureInteraction‟:– Good systems don't mind if you try using their features in new

ways; mediocre systems prevent you from using them in new

ways; bad systems break.

(c) A

SE

RT

2006-2

009

...Feature Interaction...• Examples:

– Call Waiting (if a busy subscriber receives a call, the new incoming

call is put on hold) and Call Forwarding (every call to a subscriber is

forwarded to another phone) cannot be activated in conjunction: if

one of them has priority, the other is disabled

– Bob has Call Forwarding, and is forwarding all calls to Carol. Carol

has Do Not Disturb enabled. Alice calls Bob, the call is forwarded to

Carol, and Carol's phone rings, because Do Not Disturb is not

applied to a forwarded call

– A new Mobility service is offered to office workers. When Alice signs

up, her office phone number is forwarded to the Mobility service. On

receiving a call for Alice, the Mobility service forwards it to wherever

Alice's personal data dictates.

However, whenever the data indicates that Alice is in her office, an

incoming call enters a forwarding loop.

(c) A

SE

RT

2006-2

009

...Feature Interaction...• More Examples:

– Alice@host1 and Bob@host2 correspond by email. Because Bob wishes to

remain anonymous in this correspondence, he is known to Alice as

anon@remailer, and the Remailer feature retargets electronic mail for

anon@remailer to Bob@host2.

However, Bob@host2 also has an Autoresponse feature set to notify his

correspondents that he is on vacation. When electronic mail arrives with

source address Alice@host1 and target address Bob@host2, it immediately

generates a response with source address Bob@host2 and target address

Alice@host1. When Alice receives the response, she learns Bob's identity.

– Bob has Call Forwarding, and is forwarding all calls to Carol. Alice calls

Bob, the feature forwards the call to Carol, and also changes the source

address of the call to Bob's address (on the grounds that Bob is responsible

for this leg of the call).

One scenario: Carol has Call Blocking, and is blocking all calls from Alice.

This call is not blocked, however, because it appears to be from Bob.

Another scenario: Carol misses the call, but later tries to use a Call Return

feature to return it. The feature places a call to Bob's address, which is

forwarded back incorrectly to Carol.

(c) A

SE

RT

2006-2

009

...Feature Interaction...class Phone {

def namedef respond(number) {

println "RING RING ..."println "$name: RECEIVED CALL FROM $number"

}def call(number) {

println "$name: CALLING $number"}

}

class Exchange {def numbersdef makeCall(num1, num2) {

numbers[num1].call(num2)numbers[num2].respond(num1)

}}

def e = new Exchange()e.numbers = [

123 : new Phone(name: 'Alice'),456 : new Phone(name: 'Bob')

]e.makeCall(123, 456)

Alice: CALLING 456RING RING ...Bob: RECEIVED CALL FROM 123

(c) A

SE

RT

2006-2

009

...Feature Interaction...

class ForwardingExchange extends Exchange {def forward = [456: 789]def makeCall(num1, num2) {

numbers[num1].call(forward[num2])numbers[forward[num2]].respond(num1)

}}def fe = new ForwardingExchange()fe.numbers = e.numbersfe.numbers[789] = new Phone(name: 'Carol')fe.makeCall(123, 456)

Alice: CALLING 789RING RING ...Carol: RECEIVED CALL FROM 123

(c) A

SE

RT

2006-2

009

class NoDisturbingPhone extends Phone {def doNotDisturbdef acceptOrDeny(number) {

if (doNotDisturb) println "UNAVAILABLE"else super.respond(number)

}def init(exchange) {

def emc = new ExpandoMetaClass(Exchange, true)exchange.metaClass = emcemc.makeCall = { num1, num2 ->

numbers[num1].call(num2)numbers[num2] instanceof NoDisturbingPhone ?

numbers[num2].acceptOrDeny(num1) :numbers[num2].respond(num1)

}emc.initialize()

}}def ndphone = new NoDisturbingPhone(name: 'Carol',

doNotDisturb: true)ndphone.init(e); e.numbers[789] = ndphonee.makeCall(123, 789)


Alice: CALLING 789UNAVAILABLE

(c) A

SE

RT

2006-2

009


• Now let‟s try to add both features

fe.numbers[789] = ndphonefe.makeCall(123, 456)

Alice: CALLING 456RING RING ...Bob: RECEIVED CALL FROM 123

Alice: CALLING 789RING RING ...Carol: RECEIVED CALL FROM 123

ndphone.init(fe)fe.makeCall(123, 456)

(c) A

SE

RT

2006-2

009


• Strategies to solve?– Design the problem away

• redesign individual features

[Aho et al], [Stephen & Logrippo], [Burns, Mataga &

Sutherland]

• define how combinations of features behave

[Blom, Jonsson & Kempe], [Thistle, Malham, Hoang &

Lafortune]

– Architectural Solutions

• coordinate and constrain the features’ access to shared

variables and resources[Jackson & Zave], [Brooks], [Pinard,

Weiss & Gray], [Utas]

– Detect and resolve it at run-time

• apply corrective action when an interaction actually occurs

[Homayoon & Singh], [Marples & Magill], [Griffeth &

Velthuijsen]

(c) A

SE

RT

2006-2

009

Source: Composing Features and Resolving Interactions [16]


(c) A

SE

RT

2006-2

009

Source: Handling Feature Interactions in LESS [17]

<?xml version="1.0"?><less><incoming><address-switch field="origin"><address is="sip:[email protected]"><accept /></address><otherwise><priority-switch><priority equal="emergency"><accept /></priority><otherwise><reject status="486" /></otherwise></priority-switch></otherwise></address-switch></incoming></less>

{incoming, accept, {{string-switch,organization="ABC Inc."},{address-switch, origin="sip:[email protected]"}, {string-switch,subject="group meeting"}}}

...Feature Interaction

(c) A

SE

RT

2006-2

009

Abstract System State

CW

VM

Feature

Interaction

Manager


CW

VM

Feature

Interaction

Manager


CW

VM

Feature

Interaction

Manager


CW

VM

Feature

Interaction

Manager

1 2

3 4

Source: Composing Features and Resolving Interactions [16]

CW = Call Waiting

VM = Voice Mail

Feature Interaction Verdict

• Dynamic languages allow very flexible

systems to be developed which allow

dynamic behavior to emerge at runtime

• Dynamic languages help avoiding issues

(c) A

SE

RT

2006-2

009

...

DSLs …

• Domain Specific Language– A limited form of computer language designed for a

specific class of problems

• Neal Ford[9]: Layers of abstraction using

language not hierarchies of objects– Object hierarchies still exist underneath a stronger

abstraction layer

– Objects, aspects, generics, et al become the building

blocks for DSLs

– Treat lines of code as sentences

– Allows developers to work at a higher level of

abstraction

• Declarative vs. imperative programming

(c) A

SE

RT

2006-2

009

… DSLs …

• DSL benefits:– DSLs allow solutions to be expressed as closely as

possible in the language of the problem domain

– Consequently, domain experts themselves can

understand, validate, modify, and often even develop

DSL “programs”

– DSL programs are concise, self-documenting to a

large extent, and can be reused for different purposes

– DSLs enhance productivity, reliability, maintainability,

and portability

– DSLs embody domain knowledge, and thus enable

the conservation and reuse of this knowledge

– DSLs allow validation and optimization at the domain

level

– DSLs improve testability

(c) A

SE

RT

2006-2

009

Source: [21]

… DSLs …

• The disadvantages of the use of a DSL

are:– The costs of designing, implementing and

maintaining a DSL

– The costs of education for DSL users

– The limited availability of DSLs

– The difficulty of finding the proper scope for a DSL

– The difficulty of balancing between domain-specificity

and general-purpose programming language

constructs

– The potential loss of efficiency when compared with

hand-coded software

(c) A

SE

RT

2006-2

009

Source: [21]

… DSLs …

• Fluent APIs – Rhino Mocks/Lisp Macros:

using(mocks.Ordered()){Expect.Call(databaseManager.BeginTransaction()).Return(databaseManager);

using(mocks.Unordered()){accountOne.Withdraw(1000);accountTwo.Deposit(1000);

}databaseManager.Dispose();

}mocks.ReplayAll();

Bank bank = new Bank(databaseManager);bank.TransferFunds(accountOne, accountTwo, 1000);

Source: http://www.codeproject.com/useritems/Rhino_Mocks_22.asp

(defun songs-for-album (playlist album)

(select

:from (songs-table playlist)

:where ( matching (songs-table playlist)

:album album)

:order-by :track))

Source: http://blog.secosoft.net/2007/03/13/groovy-dsl-presentation-now-online/

(c) A

SE

RT

2006-2

009

… DSLs …

• Fluent APIs – Custom Java:Event planningMeeting =

Plan.Event("Project planning meeting").RelatedTo(planningTask).WithPriority(1).At("Head office").OrganizedBy("[email protected]", "Jane Doe").StartingAt("12:00").Lasting(45).Minutes.Attendants(

"[email protected]","[email protected]","[email protected]").AreRequired.

Attendant("[email protected]").IsOptional.Resource("Projector").IsRequired.ClassifyAs("Public").CategorizeAs("Businees", "Development").Recurring.Until(2008).EverySingle.Week.On(Day.Thursday).Except.Each.Year.In(Month.July | Month.August);

planningMeeting.SendInvitations();

Source: http://andersnoras.com/blogs/anoras/archive/2007/07/04/i-m-coming-down-with-a-serious-case-of-the-dsls.aspx

(c) A

SE

RT

2006-2

009

… DSLs …

• LINQ:– API form:

– Query expression form:

IEnumerable<string> query = names .Where(s => s.Length == 5) .OrderBy(s => s).Select(s => s.ToUpper());

Source: http://??

IEnumerable<string> query = from s in names where s.Length == 5orderby sselect s.ToUpper();

(c) A

SE

RT

2006-2

009

… DSLs …

• Ruby/Groovy– Malleable languages, conducive to creating natural,

understandable code

then = now + 1.week + 3.days - 2.hours + 17.minutes

InTransaction.do { accOne << 500.euros << accTwo }

Select[:column1].from[:table1].where doequal :column2, 12equal :column3, 13

end

insert into table1 (column1, column2, column3)values (10, 'book', 'start')

(c) A

SE

RT

2006-2

009

… DSLs …

• Rebol Examples

Turn on porch light at 7:30 pm

Search for flight from SFO to JFKprefer "American" or "United"departs 1-June-2000arrives before 10:30PMpriced less than $450

Source: http://www.ddj.com/184404172

Buy 1000 shares of "Acme" symbol ACMEat or below $4.60expires on 6-June-2000account #ACN-456-987confirm with [email protected]

(c) A

SE

RT

2006-2

009

… DSLs …

• Creating a Testing DSL– Ideally, acceptance tests should be written by your

customer

– Examples:

add_book_with :name => "Groovy In Action",

:author => "Paul King",

:cost => 10.dollars

create_new_user_with :email => "[email protected]",

:name => "Jim",

:password => "letmein"

jim = login_as :email => "[email protected]",

:password => "letmein"

jim.add_to_shopping_cart 1.copy,

"Groovy In Action"

jim.checkout

add_book_with name: "Groovy In Action",

author: "Paul King",

cost: 10.dollars

create_new_user_with email: "[email protected]",

name: "Jim",

password: "letmein"

jim = login_as email: "[email protected]",

password: "letmein"

jim.add_to_shopping_cart 1.copy,

"Groovy In Action"

jim.checkout

(c) A

SE

RT

2006-2

009

… DSLs …

(c) A

SE

RT

2006-2

009

Source: http://martinfowler.com/dslwip/Intro.html

events

doorClosed D1CL

drawOpened D2OP

lightOn L1ON

doorOpened D1OP

panelClosed PNCL

end

resetEvents

doorOpened

end

commands

unlockPanel PNUL

lockPanel PNLK

lockDoor D1LK

unlockDoor D1UL

end

state idle

actions {unlockDoor lockPanel}

doorClosed => active

end

state active

drawOpened => waitingForLight

lightOn => waitingForDraw

end

state waitingForLight

lightOn => unlockedPanel

end

state waitingForDraw

drawOpened => unlockedPanel

end

state unlockedPanel

actions {unlockPanel lockDoor}

panelClosed => idle

end

… DSLs …

(c) A

SE

RT

2006-2

009

event ("doorClosed", "D1CL");

event ("drawOpened", "D2OP");

event ("lightOn", "L1ON");

event ("panelClosed", "PNCL");

resetEvent ("doorOpened", "D1OP");

command("unlockPanel", "PNUL");

command("lockPanel", "PNLK");

command("lockDoor", "D1LK");

command("unlockDoor", "D1UL");

state("idle")

.actions("unlockDoor","lockPanel")

.transition("doorClosed").to("active")

;

state("active")

.transition("drawOpened").to("waitingForLight")

.transition("lightOn"). to("waitingForDraw")

;

state("waitingForLight")

.transition("lightOn").to("unlockedPanel");

state("waitingForDraw")

.transition("drawOpened").to("unlockedPanel");

state("unlockedPanel")

.actions("unlockPanel", "lockDoor")

.transition("panelClosed").to("idle");

<stateMachine start = "idle">

<event name="doorClosed" code="D1CL"/>

<event name="drawOpened" code="D2OP"/>

<event name="lightOn" code="L1ON"/>

<event name="doorOpened" code="D1OP"/>

<event name="panelClosed" code="PNCL"/>

<command name="unlockPanel" code="PNUL"/>

<command name="lockPanel" code="PNLK"/>

<command name="lockDoor" code="D1LK"/>

<command name="unlockDoor" code="D1UL"/>

<state name="idle">

<transition event="doorClosed" target="active"/>

<action command="unlockDoor"/>

<action command="lockPanel"/>

</state>

<state name="active">

<transition event="drawOpened" target="waitingForLight"/>

<transition event="lightOn" target="waitingForDraw"/>

</state>

<state name="waitingForLight">

<transition event="lightOn" target="unlockedPanel"/>

</state>

<state name="waitingForDraw">

<transition event="drawOpened" target="unlockedPanel"/>

</state>

<state name="unlockedPanel">

<action command="unlockPanel"/>

<action command="lockDoor"/>

<transition event="panelClosed" target="idle"/>

</state>

<resetEvent name = "doorOpened"/>

</stateMachine>

… DSLs …

• How to write your DSL– Pair with your customer

– Explore the syntax of your DSL using actual inputs

and outputs

– Envision the perfect result

• See the napkin test[9]

– But don‟t over-engineer

– Don‟t go overboard with symbols

– Ensure your „code‟ is syntactically correct

– Then implement it

– Apply different static analysis rules in your IDE and CI

build for DSLs compared to other parts of your code

(c) A

SE

RT

2006-2

009

… DSLs …

• Ruby DSL Toolkit– Open classes enable methods on literals

– Method parenthesis are optional

– Named parameters, using hashes and symbols

– Code blocks to nest statements about something

– Advanced features: Meta-programming

• Ruby Meta-programming– eval, method_missing, define_method

• Groovy Meta-programming– invokeMethod, getProperty, setProperty

– Built-in builder capabilities

• These techniques may seem hackish at first, but they

actually eliminate duplication, and become natural

(c) A

SE

RT

2006-2

009

… DSLs …

(c) A

SE

RT

2006-2

009

… DSLs …

(c) A

SE

RT

2006-2

009

Source: http://martinfowler.com/dslwip/Intro.html

… DSLs …

(c) A

SE

RT

2006-2

009

Source: http://nealford.com/

… DSLs

• Summary– Who will use the DSL?

– When will they use it?

– How will they use it?

– When will changes be required?

– Keep it as cohesive as possible

– Don‟t try to write the entire universe in your DSL

• You may be better off using a bunch of very specific DSLs

– For business DSLs, enable your customer to write

tests as simple phrases, using the terminology of the

business

(c) A

SE

RT

2006-2

009

DSL Verdict

• Dynamic languages help writing DSLs

• Writing DSLs is a trivial and well-

understood area of system design!

(c) A

SE

RT

2006-2

009

...

Further Information…• [1] Dynamic vs. Static Typing — A Pattern-Based Analysis, Pascal Costanza,

University of Bonn, 2004http://p-cos.net/documents/dynatype.pdf

• [2] Interface-Oriented Design, Ken Pugh, Pragmatic Programmers, 2006

• [3] Bruce Eckel, Does Java need Checked Exceptions?www.mindview.net/Etc/Discussions/CheckedExceptions

• [4] Null Object, Kevlin Henney, Proceedings EuroPLoP 2002

• [5] Design Patterns in Dynamic Programming, Peter Norvig, March 1998http://www.norvig.com/design-patterns/

• [6] Advanced Programming Language Features and Software Engineering: Friend or Foe?, Greg Sullivan, April 2002http://people.csail.mit.edu/gregs/proglangsandsofteng.pdf

• [7] JunGL: a Scripting Language for Refactoring, Mathieu Verbaere et al, May 2006http://progtools.comlab.ox.ac.uk/publications/icse06jungl

• [8] Rails for Java Developers, Halloway et al, Pragmatic Bookshelf, 2007, Chapter 3, Ruby Eye for the Java Guy

• [9] Building DSLs in Static & Dynamic languageshttp://www.nealford.com/downloads/conferences/canonical/Neal_Ford-Building_DSLs_in_Static_and_Dynamic_Languages-handouts.pdf

• [10] Five Easy Pieces: Simple Python Non-Patterns, Alex Martellihttp://www.aleax.it/5ep.html

• [11] Emergent Design, Scott L. Bain, 2008

(c) A

SE

RT

2006-2

009

…Further Information• [12] Design Patterns: Elements of Reusable Object-Oriented Software, Erich

Gamma, Richard Helm, Ralph Johnson, John Vlissides, 1995

• [13] Refactoring: Improving the Design of Existing Code, Martin Fowler, 1999

• [14] Effective Java Programming Language Guide, Erich Gamma, Joshua Bloch, 2001

• [15] Agile Software Development, Principles, Patterns, and Practices, Robert C Martin, 2002

• [16] Composing Features and Resolving Interactions, Jonathan Hay and Joanne Atlee, University of Waterloo

• [17] Handling Feature Interactions in the Language for End System Services, Xiaotao Wua and Henning Schulzrinne, January 2007

• [18] FAQ Sheet on Feature Interaction, Pamela Zavehttp://www.research.att.com/~pamela/faq.html

• [19] Liskov Substitution Principle and the Ruby Core Libraries, Dean Wamplerhttp://blog.objectmentor.com/articles/tag/liskovsubstitutionprinciple

• [20] Liskov Substitution in Dynamic Languages, Michael Feathershttp://butunclebob.com/ArticleS.MichaelFeathers.LiskovSubstitutionInDynamicLanguages

• [21] Domain-Specific Languages: An Annotated Bibliography, van Deursen et alhttp://homepages.cwi.nl/~arie/papers/dslbib/

• [22] Agile Principles, Patterns, and Practices in C#, Martin C. Robert et al, 2006

• [23] Clean Code, Robert C. Martin, 2008

• [24] The Craftsman: 51, Ruby Visitor, Robert C Martin, August 2007http://www.objectmentor.com/resources/articles/Craftsman51.pdf

(c) A

SE

RT

2006-2

009


(c) ASERT 2006-2009


• Thank you for listening

• Any Questions?

(c) ASERT 2006-2009

paulking dlp

Technology

count world

li world

title hello

c asert