SOLID - Principles of Object Oriented Design

SOLID PRINCIPLES
INGEGNERIA DEL SOFTWARE
Università degli Studi di Padova
Dipartimento di Matematica
Corso di Laurea in Informatica, A.A. 2015 – 2016
rcardin@math.unipd.it

Ingegneria del software mod. B
SUMMARY
 Introduction
 Single Responsibility Principle
 Open Closed Principle
 Liskov Substitution Principle
 Interface Segregation Principle
 Dependency Inversion Principle
2Riccardo Cardin

INTRODUCTION
 Structured programming and Object-oriented
programming
 Two of the most important revolution of IT industry
 Everyone uses OO languages, but...
 Today's programmers are unaware of the principles
that are the foundation of Object Orientation
 Dependency management
 The art of making code flexible, robust, and reusable
 It’s too easy to get a bunch of tangled legacy code
 SOLID principles
 A set of class design principles that helps to manage
dependency
3Riccardo Cardin

INTRODUCTION
 SOLID principles
 Single Responsibility Principle
 A class should have one, and only one, reason to change
 Open Closed Principle
 You should be able to extend a classes behavior, without
modifying it
 Liskov Substitution Principle
 Derived classes must be substitutable for their base classes
 Interface Segregation Principle
 Make fine grained interfaces that are client specific
 Dependency Inversion Principle
 Depend on abstractions, not on concretions
4Riccardo Cardin

SINGLE RESPONSIBILITY PRINCIPLE
 Also known as cohesion
 Functional relatedness of the elements of a module
 A module should have only one reason to change
 We call this reason of change responsibility
 Coupled responsibilities
 Changes to one responsibility may impair or inhibit
the class’ ability to meet the others
 Fragile design that break in unexpected ways
 Recompilation, test, deploy, …
5Riccardo Cardin

6Riccardo Cardin
Uses Rectangle to help it with the
mathematics of geometric shapes. It
never draws the rectangle on the
screen
It definitely draws the rectangle
on the screen.
The Rectangle class has two
responsibilities.
• Provides a mathematical
model
• Renders the rectangle on
a GUI

7Riccardo Cardin
Responsibilities are separeted
into two completely different
classes
Renders the rectangle on a
GUI
Provides a mathematical
model

 What is really a responsibility?
 An axis of change is only an axis of change if the
changes actually occur
 The context of the application is also important
 Needless complexity
 Should these two responsibilities be separated?
 That depends upon how the application is changing.
8Riccardo Cardin
public interface Modem {
public void dial(String pno);
public void hangup();
public void send(char c);
public char recv();
}
Connection management
Data communication

 Eventually separate responsibilities avoids rigidity
 They are still coupled in ModemImplementation, but clients
don’t need to worry about interface implementations
9Riccardo Cardin

10Riccardo Cardin

OPEN-CLOSE PRINCIPLE
 There are many heuristics in OOD
 Software entities should be open for extension, but
closed for modification
 You extend behaviour adding new code, not changing the old
 The Open-Close Principle underlines these heuristics
 Abstraction is the key
 Abstract types are the fixed part, derivate classes
points of extension
11Riccardo Cardin
“All member variables should be private”, “Global variables
should be avoided”, “Using run time type identification (RTTI) is
dangerous”

12Riccardo Cardin
Client class
must be changed to
name the new
server class.
If we want Client
objects to use a
different server
class, then a new
derivative of the
AbstractServer
class can be
created.
The Client class
can remain
unchanged.
O
C
P

13Riccardo Cardin
public static void drawAll(Shape[] shapes) {
for (Shape shape : shapes) {
switch (shape.shapeType) {
case Square:
((Square) shape).drawSquare();
break;
case Circle:
((Circle) shape).drawCircle();
break;
}
}
}
Does not conform to the open-
closed principle because it
cannot be closed against new
kinds of shapes. If I wanted to
extend this function, I would
have to modify the function

 Programs conforming to OCP do not experience
«cascade of changes»
 Changes are obtained adding new code
14Riccardo Cardin
public static void drawAll(Shape[] shapes) {
for (Shape shape : shapes) {
shape.draw();
}
}
Solution that conforms to open-
close principle. To extend the
behavior of the drawAll to draw
a new kind of shape, all we need
do is add a new derivative of the
Shape class.

 No program can be 100% closed
 Closure must be strategic
 Closure can be gained through abstraction
 Using interfaces and polimorphim
 The draw abstract method in the Shape class
 ...or can be gained in a «data-driven» fashion
 Sometimes using information configured in external
structure can be the only solution
 What if we want to draw shapes in a specific order that
depends from type!?
15Riccardo Cardin

16Riccardo Cardin

 Conventions and heuristics derived from OCP
 Make all member variables private
 When the member variables of a class change, every function
that depends upon them must be changed
 Encapsulation
 No global variables (ever)
 No module that depends upon a global variable can be closed
against any other module that might write to that variable
 There are very few cases that can disobey (i.e. cin, cout)
 RTTI is dangerous
 The Shape example shows the bad way to use RTTI
 But there are also good cases…
17Riccardo Cardin

LISKOV SUBSTITUTION PRINCIPLE
 Abstraction and polymorphism
 At the basis of OOD and OCP
 What are the characteristics of the best inheritance
hierarchies? What are the traps?
 Liskov Substitution Principle
 It a special case of the real LSP ;)
 Violating this principle means violating OCP
 Function that uses a pointer or reference to a base class, but
must know about all the derivatives of that base class.
18Riccardo Cardin
Functions that use pointers or references to base classes must
be able to use objects of derived classes without knowing it.

19Riccardo Cardin
A Square does not need both
height and width member
variables. Yet it will inherit them
anyway. Clearly this is wasteful.
Square will inherit the setWidth
and setHeight functions. These
functions are utterly inappropriate
for a Square.
But, we could override them...
public void setWidth(double width) {
super.setWidth(width);
super.setHeight(width);
}
public void setHeight(double height) {
this.setWidth(height);
}

 A model, viewed in isolation, can not be
meaningfully validated
 The validity of a model can only be expressed in
terms of its clients
20Riccardo Cardin
public void f(Rectangle r) {
r.setWidth(32);
}
@Test
public void testF() {
Rectangle r = new Square();
r.setHeight(15);
f(r);
// This test will not pass!!!
assertEquals(32, r.getWidth);
}
If we pass a reference to a
Square object into this
function, and the height will
be changed too.
This is a clear violation of
LSP. The f function does not
work for derivatives of its
arguments.

21Riccardo Cardin

 What went wrong?
 What counts is extrinsic public behavior
 Behavior that clients depend upon
 The relation between Square and Rectangle is not a IS-A
relation in OOD
 Design by contract
 Methods of classes declare preconditions and
postconditions (invariants)
22Riccardo Cardin
...when redefining a routine [in a derivative], you may only replace its
precondition by a weaker one, and its postcondition by a stronger one.
// Rectangle.setWidth(double w) postconditions
assert((width == w) && (height == old.height));

 Design by contract
 In a derivate class preconditions must not be stronger
than in the base class
 Using base class interface a client knows only base class
preconditions
 In a derivate class postconditions must be stronger
than in the base class
 Derived class must conform to all base class prostcondition.
The behaviors and outputs must not violate any of the
constraints established for the base class
 Java and JVM base languages have assert primitive.
C++ does not have anything such this
23Riccardo Cardin

INTERFACE SEGREGATION PRINCIPLE
 Reducing coupling means to depend upon
interfaces, not implementations
 The risk is to depend upon a «fat» or «polluted»
interfaces
 Fat interfaces are not cohesive
 Methods can be broken up into groups of functions
 Clients must view only the part they are interested to
 Interface Segregation Principle
24Riccardo Cardin
Clients should not be forced to depend upon interfaces that they do not
use

25Riccardo Cardin
In this system there are Door objects that can
be locked and unlocked, and which know
whether they are open or closed.
Clients used this interface to managed doors.
Now consider that one such
implementation. TimedDoor
needs to sound an alarm when
the door has been left open for too long.
In order to do this the TimedDoor object
communicates with another object called a Timer.
TimeClient method represents the function called
when the timeout expires

 First solution
26Riccardo Cardin
The Door class now depends upon
TimerClient. Not all varieties of Door need
timing. Moreover, the applications that use those
derivatives will have to import the definition of
the TimerClient class, even though it is not
used.
The interface of Door has been polluted with an
interface that it does not require. Each time a
new interface is added to the base class, that
interface must be implemented in derived
classes.
Default implementations violate the
Liskov Substitution Principle (LSP)

27Riccardo Cardin

 Clients of Door and TimerClient are different
 The interfaces should remain separate too
 Sometimes it is the client that forces a change to an
interface
 Also the Door interface have to be changed
 Clients that does not need timer doors will also be affected
 The result is a inadvertent coupling between all the
clients
28Riccardo Cardin
public class Timer {
void register(int timeout, int timeOutId, TimerClient client);
}
public interface TimerClient {
// A change to Timer implies a change to TimerClient
void timeOut(int timeOudId);
}

 Separation by delegation
 Object form of the Adapter design pattern
 DoorTimerAdapter translates a Door into a TimerClient
 Clients of Door and TimerClient are not coupled
anymore
29Riccardo Cardin

 Separation through multiple inheritance
 Class form of the Adapter design pattern
 Client can use the same object through different and
separate interfaces
 Possible only when multiple inheritance is supported
 Less types used wrt the solution that uses delegation
30Riccardo Cardin

DEPENDENCY INVERSION PRINCIPLE
 Bad design often derives from degradation due to
new requirement and maintanance
 Rigidity – hard to change because every change affects
to many part of the system
 Fragility – when you make a change, unexpected parts
of the system break
 Immobility – It is hard to reuse in another application
because it cannot be easily disentangled
 TNTWIWHDI – That’s not the way I would have done it
 Interdependence of the modules
31Riccardo Cardin

32Riccardo Cardin
Consider a simple program that
is charged with the task of
copying characters typed on a
keyboard to a printer.
“Read keyboard” and “Write
printer” are quite reusable.
However the “Copy” module is
not reusable in any context that
does not involve keyboard and
printer
public void copy(OutputDevice dev) {
int c;
while ((c = readKeyboard()) != EOF)
if (dev == PRINTER)
writePrinter(c);
else
writeDisk(c);
}
Violates
OCP

 Module containing high level policy should be
independent upon low level details modules
 We have use abstraction to limit dependency
33Riccardo Cardin
We have performed dependency
inversion. the dependencies
have been inverted; the “Copy”
class depends upon abstractions,
and the detailed readers and
writers depend upon the same
abstractions.
Now we can reuse the “Copy”
class, independently of the
“Keyboard Reader” and the
“Printer Writer”.

 Dependency Inversion Principle
 Important policy decisions are in high level modules
 It’s these modules we want to be able to reuse
 Template method design pattern
 In layered application, each layer should expose a
proper level of abstraction (interface)
 A naive implementation can force wrong dependency among
modules
34Riccardo Cardin
High level modules should not depend upon low level modules. Both
should depend upon abstractions.
Abstractions should not depend upon details. Details should depend
upon abstractions.

35Riccardo Cardin
The high level policy class uses a
lower level Mechanism; which in
turn uses a detailed level utility
class. The Policy Layer is sensitive
to changes all the way down in the
Utility Layer.
Each of the lower level layers are
represented by an abstract class.
Each of the higher level classes
uses the next lowest layer through
the abstract interface. Thus, none
of the layers depends upon any of
the other layers.

36Riccardo Cardin

REFERENCES
 The Principles of OOD
http://butunclebob.com/ArticleS.UncleBob.PrinciplesOfOod
 Chap. 8 “The Single-Responsibility Principle (SRP)”, Agile Principles,
Patterns, and Practices in C#, Robert C. Martin, 2006, Prentice Hall
 Chap. 9 “The Open/Closed Principle (OCP)”, Agile Principles,
 Chap. 10 “The Liskov Substitution Principle (LSP)”, Agile Principles,
 Chap. 11 “The Dependency-Inversion Principle (DIP)”, Agile
Principles, Patterns, and Practices in C#, Robert C. Martin, 2006,
Prentice Hall
 Chap. 12 “The Interface Segregation Principle (ISP)”, Agile
Principles, Patterns, and Practices in C#, Robert C. Martin, 2006,
Prentice Hall
37Riccardo Cardin

SOLID - Principles of Object Oriented Design

Recommended

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to SOLID - Principles of Object Oriented Design (20)

More from Riccardo Cardin (9)

Recently uploaded (20)

SOLID - Principles of Object Oriented Design