Advanced C++ Programming Advanced C++ Programming

Advanced
C++ Programming
AMP Ganesh Prabhu, Bangalore,India
Introduction
C++ supports a variety of programming styles:

procedural (as FORTRAN, pascal, C, )
object-oriented (as Eifel, JAVA, )
generic
Features interact with each other.
C++ is very flexible & efficient
Introduction
The goal of this course is to provide insight into the
advanced features and its internal working
To enable the participants to develop good object
oriented design and use C++ effectively.
Prerequisites
A working knowledge of C++ and some basic
object-oriented concepts.
What is a pointer ?
What is a class ?
What is inheritance ?
Outline
Module 1: Basic structure of C++ program
Module 2: Namespace
Module 3: Functions
Module 4: Object Model
Module 5: Initialization & Clean up
Module 6: Dynamic Memory Management
Module 7: Inheritance & Containment
Module 8: Templates and Generic programming
Module 9: Exception Handling
MODULE 1
Steps involved in the process of building the

C++ program
COMPILERS, ASSEMBLERS and LINKERS
Normally the C/C++ program building process involves
four stages and utilizes different tools such as a
preprocessor, compiler, assembler, and linker.
At the end there should be a single executable file.
The following slide illustrates the stages that happen in order

regardless of the operating system/compiler.

C++ program (contd..)
Source code file
- .CPP
C++ Pre-processor
Preprocessed
code
C++ compiler
Object code
Other object
files/modules
C++ linker
Library files
Executable
code- .EXE

Preprocessing is the first pass of any C++ compilation.
-It processes include-files, conditional compilation instructions
and macros.
Compilation is the second pass.

-It takes the output of the preprocessor, and the source code,
and generates assembler source code.
Assembly is the third stage of compilation.

-It takes the assembly source code and produces an assembly
listing with offsets. The assembler output is stored in an object
file.

Linking is the final stage of compilation.
-It takes one or more object files or libraries as input and
combines them to produce a single (usually executable) file.
-In doing so, it resolves references to external symbols,
assigns final addresses to procedures/functions and variables,
-and revises code and data to reflect new addresses (a
process called relocation).
Memory Layout of a process

OBJECT FILES and EXECUTABLE
After the source code has been assembled, it will produce an
Object files (e.g. .o, .obj) and then linked, producing an
executable files.
An object and executable come in several formats such as
- ELF (Executable and Linking Format)
- COFF (Common Object-File Format) and
- PE (Portable Executable Format).
For example, ELF is used on Linux systems, while COFF is used
on Unix systems & PE format is used on Windows systems.
Memory Layout of a process(contd..)

When we examine the content of these object files there
are areas called sections.
Sections can hold executable code, data, dynamic linking
information, debugging data, symbol tables, relocation information,
comments, string tables, and notes.
Some sections are loaded into the process image and some
provide information needed in the building of a process image while
still others are used only in linking object files.

System
mfp, frame pointer for

main
env
argv
argc
Auto variables for main
Stack pointer (grows

downward if func()
calls another
function)
STACK
Auto variables for func()

Available for stack growth
Library functions if
dynamically linked
(.SO , .DLL)
Shared Memory
Available for heap growth
HEAP
(malloc/realloc arena)
Initialized and
unintialized data
Global /static variables

Compiled code
(our code + static libraries)
DATA Segment
CODE/TEXT Segment

There are several sections that are common to all
executable formats (may be named differently, depending
on the compiler/linker) as listed below:
.text
This section contains the executable instruction codes and is shared
among every process running the same binary.
This section usually has READ and EXECUTE permissions only.
This section is the one most affected by optimization.
.bss
BSS stands for Block Started by Symbol.
It holds un-initialized global and static variables.
Since the BSS only holds variables that don't have any values yet, it
doesn't actually need to store the image of these variables.

The size that BSS will require at runtime is recorded in the object
file, but the BSS (unlike the data section) doesn't take up any
actual space in the object file.
.data
Contains the initialized global and static variables and their values.
It is usually the largest part of the executable. It usually has
READ/WRITE permissions.
.rdata
Also known as .rodata (read-only data) section.
This contains constants and string literals.

.reloc
Stores the information required for relocating the image while
loading.
Symbol table
A symbol is basically a name and an address.
Symbol table holds information needed to locate and relocate a
programs symbolic definitions and references.
A symbol table index is a subscript into this array.
Index 0 designates both the first entry in the table and serves as
the undefined symbol index.
The symbol table contains an array of symbol entries.

Relocation records
Relocation is the process of connecting symbolic references with
symbolic definitions. For example,
when a program calls a function, the associated call instruction
must transfer control to the proper destination address at
execution.
Re-locatable files must have relocation entries which are
necessary because they contain information that describes how to
modify their section contents, thus
allowing executable and shared object files to hold the right
information for a process's program image.
Simply said relocation records are information used by the linker
to adjust section contents.
Static and Dynamic link library
A library is a piece of pre-compiled (binary) code

shipped to an end-user developer,
who can implement or consume the functionalities

of the library without getting to see the actual code.
A library can be shipped generally in 2 different

forms.
1. Static link library &
2. Dynamic link library
Static and Dynamic link library (contd..)

Static link library:
This is the simplest form amongst the two.
This library would get linked to the .OBJ file of the

consumer during link time, and tend to be an
integral part of the executable file.
These files go with the extension

1. .LIB (Microsoft)
2. .a (Linux, also called archive file)
Static and Dynamic link library (contd..)

Dynamic link library:
This library is independent of an executable file.
An executable file consuming a dynamic link library
would be more compact or smaller in size when
compared to an executable consuming a static link
library.
The dynamic link library at any given point of time
shall be single copy in main memory and could be
shared by more than one executable file, thus
making it more memory efficient when compared to
static link libraries.
These files go with the extension
1. .DLL (Microsoft)
2. .SO (Linux, also called Shared object)
Why OO technology?
Dependency Management
What is dependency management?

What bearing does DM have on software?
What is the result of poor DM?
What is the advantage of good DM?
simple idea - as interdependencies increase,

Afeatures
like reusability, flexibility, and maintainability
decrease.
management is controlling
Dependency
interdependencies.
What bearing does DM have on software?
and cohesion are the eternal concerns of

Coupling
software development
can say that OO is just a set of tools and
One
techniques for Dependency Management
What is the penalty for practicing poor DM?

A system with poor dependency structure will
typically exhibit these four negative traits:
It is rigid
It is fragile
It is not reusable
It has high viscosity
It is Rigid
Rigidity is the inability to be changed
The impact of a change cannot be predicted

If not predicted, it cannot be estimated
Time and cost cannot be quantified
Managers become reluctant to authorize change
Official Rigidity for Roach Motel modules
Changes with Rigidity
Officially Rigid Area

Where the change should be made
Where the change must be made now

The System
Are we containing risk, or spreading rot?
It is Fragile
Software changes seem to exhibit non-local effects
single change requires a cascade of

Asubsequent
changes
errors appear in areas that seem

New
unconnected to the changed areas
Quality is unpredictable.
The development team loses credibility
Increasing Risk
Probability of
introducing a bug
Defects v. Cumulative Modifications
1.0
Changes
Systems tend to become increasingly fragile over time. Intentional, planned

partial rewrites may be necessary to sustain growth and maintenance.
It is not reusable
parts of the design are

Desirable
dependent upon undesirable
parts
work and risk of extracting
The
the desirable part may exceed
the cost of redeveloping
from scratch.
It has high viscosity

Viscosity is resistance to fluid motion.
the right changes are much more difficult

When
than hacking, the viscosity of the system is high.
time, it will become increasingly difficult to
Over
continue developing the product.
What is the benefit of good DM?

Interdependencies are managed, with firewalls
separating aspects that need to vary independently.
More Flexible
Less fragile,
the bugs are boxed
in
Easier to reuse
Easier to make the right change
MODULE 2
Namespaces
Namespace pollution
Occurs when building large systems from pieces

Identical globally-visible names clash
How many programs have a print function?
Very difficult to fix
Classes suggest a solution

class A { void f(); };
class B { void f(); };
Two fs are separate
Namespaces (contd..)
Scope for enclosing otherwise global declarations
namespace Mine
{
void print(int);
const float pi = 3.1415925635;
class Shape { };
}
void bar(float y)
{
float x = y + Mine::pi;
Mine::print(5);
}
using directive brings namespaces or objects into scope
namespace Mine
{
const float pi = 3.1415926535;
void print(int);
}
using Mine::print;
void foo() { print(5); } // invoke
Mine::print
using namespace Mine;
float twopi = 2*pi;
// Mine::pi
Namespaces are open: declarations can be added
namespace Mine
{
void f(int);
}
namespace Mine
{
void g(int);
}
// Add Mine::g()to Mine
Declarations and definitions can be separated
namespace Mine
{
void f(int);
}
void Mine::f(int a)
{
/* */
}
Function stack frames
MODULE 3
Function stack
One of the most important mechanisms a programmer
should understand is the function call stack (sometimes
referred to as the program execution stack).
This data structure working "behind the scenes supports the

function call/return mechanism.
It also supports the creation, maintenance and destruction of each

called
function's automatic variables.
time a function calls another function, an entry is pushed onto
theEach
stack.
Function stack
This entry, called a stack frame or an activation record, contains the
return
address that the called function needs to return to the calling
function.
Of course, the amount of memory in a computer is finite, so only a
certain
amount of memory can be used to store activation records on
the function call stack.
If more function calls occur than can have their activation records
stored
on the function call stack, an error known as stack overflow
occurs.
Function stack (Example)

// square function used to demonstrate the function
// call stack and activation records.
#include <iostream>
using std::cin;
using std::cout;
using std::endl;
int square( int ); // prototype for function square
int main()
{
int a = 10; // value to square (local automatic variable in main)
cout << a << " squared: " << square( a ) << endl; // display a squared
return 0; // indicate successful termination
} // end main
// returns the square of an integer
int square( int x ) // x is a local variable
{
return x * x; // calculate square and return result
} // end function square
Function stack (Fig .1)
Function stack (Fig. 2)
Function stack (Fig. 3)
CALLING CONVENTIONS
What Are Calling Conventions?
C++ provides a way of defining a function and a way of calling that
function with arguments that soon become second nature to use.
the function and calls to it are compiled there are quite a
fewWhen
different ways in which the process of getting those
arguments to where the functions code can see them before
executing the code can be performed.
There are pros and cons to each, and some compilers have
extensions
which allow you to choose which is used.
of the time there is no need to do anything other than use
theMost
defaults or, in a few cases where you are required to match a
binary specification, do what you are told.
But you might want to make use of this in some cases, or simply
understand what it is going on with calling conventions used by
code you interface with.
CALLING CONVENTIONS (contd..)

Calling Conventions and Pure C++
The C++ Standard [ISO/IEC 14882] has very little to say about
calling conventions. It has a bit to say about language linkage
and linkage specifications, and it notes that a particular
language linkage may be associated with a particular calling
convention. Everything else related to calling conventions is
implementation specific.
Types of Calling Conventions:
__cdecl
__stdcall
__fastcall
__thiscall

The Conventions
__cdecl
The __cdecl calling convention is used for functions marked with
the __cdecl keyword, or for unmarked functions when the /Gd
microsoft compiler option is set. Since /Gd is the default this is then
the default calling convention.
All arguments are stored on the stack, in right to left order. The
code that called the function will have to remove these arguments
when the function returns.
Because the calling code cleans the stack this convention can be
used with functions with a variable number of arguments such as
printf(const char*, ...). Since C++ must allow such functions, as
advisable as it is to avoid them when you can, it makes sense that
this is the default calling convention.

__stdcall
The __stdcall calling convention is used for functions marked
with the __stdcall keyword, or for unmarked functions when
the /Gz compiler option(microsoft specific) is set.
All arguments are stored on the stack, in right to left order. The
function itself will clean the stack. This results in slightly smaller
executables but cannot be used with functions that take a
variable number of arguments.
Microsoft specific:
COM methods, and the Windows API functions are generally
__stdcall functions. The macros WINAPI and CALLBACK and a
few others used in Windows API declarations expand to
__stdcall.

__fastcall
The __fastcall calling convention is used for functions marked
with the __fastcall keyword, or for unmarked functions when
the /Gr compiler option (microsoft specific) is set (/Gr will not
affect main()).
The first two arguments (left to right) which are 32bit or smaller
are placed into two registers (currently ECX and EDX, though
Microsoft have no commitment to keep using those two with later
compilers). All other arguments are stored on the stack, in right to
left order. The function itself will clean the stack.
If you attempt to take the address of an argument stored in a
register it will be copied to a temporary location.
While reasons this would give an advantage in speed are clear,
with modern machines it tends to be negligible.

__thiscall
The __thiscall calling convention cannot be explicitly used. It is used
for member functions of classes which are not explicitly set to use a
different convention, If the member functions takes a variable
number of arguments, then a combination of __thiscall with __cdecl
is used. If the member functions takes only fixed no. of parameters
then the programmer has the choice __thiscall with __stdcall.
The convention operates much like __stdcall except that the this
pointer (which is effectively a hidden argument) is stored in the ECX
register. This neatly matches with the C++ rule that one cannot take
the address of the this pointer.
In the case where a variable number of arguments requires __cdecl
to be used the this pointer is considered the left-most argument,
and hence pushed on the stack last.
Calling convention syntax in a C++ code:

Microsoft Compiler:
void fun();
------- void __cdecl fun();
GCC Compiler:
void fun(); ------- void fun() __attribute((cdecl))__ ;
Linkage Specification
This is to tell the compiler that the following function is
not defined in this program, and it is the compiler's job to
look for the definition.
This is also to inform the compiler that the following
function was compiled as C code.
C++ specially encodes method names for type-safe
linkage, but C doesnt.
So when an attempt is made to link C code with C++
code, the method compiled in C will not be directly
recognized.
For a single method:
extern C method prototype
For multiple methods:

extern C
{ method prototypes }
Inline Functions
Function calls
Cause execution-time overhead
Qualifier inline before function return type "advises" a
function to be inlined
Puts copy of function's code in place of function call
Speeds up performance but increases file size

Compiler can ignore the inline qualifier
Ignores all but the smallest functions
inline double cube( const double s )

{ return s * s * s; }
FUNCTION POINTER
&
CALLBACKS
FUNCTION POINTER
Function Pointers provide some extremely interesting,
efficient and elegant programming techniques.
You can use them to replace switch/if-statements, to
realize your own late-binding or to implement callbacks.
Unfortunately -probably due to their complicated syntax they are treated quite step motherly in most computer
books and documentations. If at all, they are addressed
quite briefly and superficially.
They are less error prone than normal pointers because
you will never allocate or deallocate memory with them. All
you've got to do is to understand what they are and to
learn their syntax. But keep in mind: Always ask yourself if
you really need a function pointer.
FUNCTION POINTER (contd..)

It's nice to realize one's own late-binding but to use
the existing structures of C++ may make your code
more readable and clear.
One aspect in the case of late-binding is runtime: If
you call a virtual function, your program has got to
determine which one has got to be called. It does this
using a V-Table containing address of all the possible
functions.
This costs some time for each call and maybe you
can save some time using function pointers instead of
virtual functions.

What is a Function Pointer?
Function Pointers are pointers, i.e. variables, which point
to the address of a function.
You must keep in mind, that a running program gets a
certain space in the main-memory.
Both, the executable compiled program code and the
used variables, are put inside this memory. Thus a function
in the program code is, like e.g. a character field, nothing
else than an address.

Define a Function Pointer:
Regarding their syntax, there are two different types
of function pointers:
- Pointers to ordinary C functions or to static C++
member functions.
- Pointers to non-static C++ member functions.
The basic difference is that all pointers to non-static
member functions need a hidden argument: The thispointer to an instance of the class.
Always keep in mind: These two types of function pointers are
incompatible with each other.

Since a function pointer is nothing else than a variable, it
must be defined as usual. In the following example we define
three function pointers named pt2Function, pt2Member and
pt2ConstMember.
They point to functions, which take one float and two char
and return an int.
In the C++ example it is assumed, that the functions, our pointers point to,
are (non-static) member functions of TMyClass.
// define a function pointer and initialize to NULL

int (*pt2Function)(float, char, char) = NULL; // C int
(TMyClass::*pt2Member)(float, char, char) = NULL; // C++ int
(TMyClass::*pt2ConstMember)(float, char, char) const = NULL; // C++
Important note: A function pointer always points to a function with a specific signature! Thus
all functions, you want to use with the same function pointer, must have the same
parameters and return-type!

Calling Convention & Function pointers:
Normally you don't have to think about a function's
calling convention: The compiler assumes __cdecl as
default if you don't specify another convention.
However if we want to specify our own ... The calling
convention belongs to a function's signature: Thus
functions and function pointers with different
calling convention are incompatible with each
other!
For Borland and Microsoft compilers you specify a
specific calling convention between the return type and
the function's or function pointer's name.
Calling Convention & Function pointers: (Contd..)

For the GNU GCC you use the __attribute__ keyword: Write
the function definition followed by the keyword __attribute__
and then state the calling convention in double parentheses.
// define the calling convention
void __cdecl DoIt(float a, char b, char c); //Borland
& Microsoft compiler
void DoIt(float a, char b, char c) __attribute__((cdecl)); //
GNU GCC

Assign an address to a Function Pointer:
It's quite easy to assign the address of a function to a
function pointer. You simply take the name of a suitable

and known function or member function.
Although it's optional for most compilers you should use
the address operator & in front of the function's name in
order to write portable code.
You may have got to use the complete name of the
member function including class-name and scoperesolution operator (::).

Also you have got to ensure, that you are allowed to
access the function right in scope where your assignment
stands.
// assign an address to the function pointer
// Note: Although you may omit the address operator on most compilers
// you should always use the correct way in order to write portable
code. //In C
int DoIt (float a, char b, char c) { printf("DoIt\n"); return a+b+c; }
int DoMore(float a, char b, char c)const
{
printf("DoMore\n");
return a-b+c;
}
pt2Function = DoIt; // short form
pt2Function = &DoMore; // correct assignment using address operator

// C++
class TMyClass
{
public:
int DoIt(float a, char b, char c)
{ cout << "TMyClass::DoIt"<< endl; return a+b+c;};
int DoMore(float a, char b, char c) const
{ cout << "TMyClass::DoMore" << endl; return a-b+c; };
/* more of TMyClass */
};
pt2ConstMember = &TMyClass::DoMore; // correct assignment using & operator

pt2Member = &TMyClass::DoIt; // note: <pt2Member> may also legally point to &DoMore

Calling a Function using a Function Pointer
In C you call a function using a function pointer by
explicitly de-referencing it using the * operator.
Alternatively you may also just use the function pointer's
instead of the functions name.
In C++ the two operators .* and ->* are used together with
an instance of a class in order to call one of their (nonstatic) member functions.
If the call takes place within another member function you
may use the this-pointer.

// calling a function using a function pointer
int result1 = pt2Function (12, 'a', 'b'); // C short way
int result2 = (*pt2Function) (12, 'a', 'b'); //
TMyClass instance1;
int result3 = (instance1.*pt2Member)(12, 'a', 'b'); // C++
int result4 = (*this.*pt2Member)(12, 'a', 'b');
// C++ if this-pointer can be used
TMyClass* instance2 = new TMyClass;
int result4 = (instance2->*pt2Member)(12, 'a', 'b'); // C+
+, instance2 is a pointer
delete instance2;
Function Overloading
Function overloading:
Functions with same name and different parameters
Overloaded functions should perform similar tasks
Function to square ints and function to square floats
int square( int x) {return x * x;}

float square(float x) { return x * x; }
Program chooses function by signature
Signature determined by function name and parameter types
Type safe linkage - ensures proper overloaded function called
Function Overloading (Contd..)

1 //An example
2 // Using overloaded functions
3 #include <iostream>
4
5 using std::cout;
6 using std::endl;
7
8 int square( int x ) { return x * x; }
9
10 double square( double y ) { return y *
11
12 int main()
13 {
14
cout << "The square of integer 7 is
15
<< "\nThe square of double 7.5
square(
7.5<<) endl;
16
17
18
return 0;
OUTPUT
The square of
19 }
y; }
" << square( 7 )

is " <<
integer 7 is 49
The square of double 7.5 is 56.25
Call by value & reference

When passing a value, in either case, the call will look the
same, its the receiving method which determines how the
pass is made.
By Value:
void aMethod(aParameter p) { }
When passing by value a temporary COPY is made in
memory taking up space and making any changes to the
variable useless outside of the aMethod scope. Once
control returns to the calling procedure any changes to the
variable will not be recognized, and you will still have the
original variable.
Call by value & reference

By Reference:
void aMethod(aParameter& p) { }
NOTE: the & can be place either against the parameter type,
against the parameter name, or there can be a space on either
side.
Eg:
By value:
void fun(int x, int y){ .}
By Reference: void fun(int &x, int &y){}
Function Templates
Function templates
Compact way to make overloaded functions

Keyword template
Keyword class or typename before every formal type
parameter (built in or user defined)
template < typename T > T square( T value1)
{
return value1 * value1;
}
T replaced by type parameter in function call
int x;
int y = square(x);
If int parameter, all T's become ints
Can use float, double, long...
Function Templates (Example)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
// An example on
// Using a function template
#include <iostream>
using std::cout;
using std::cin;
using std::endl;
template < class T >
T maximum( T value1, T value2, T value3 )
{
T max = value1;
if ( value2 > max )
max = value2;
if ( value3 > max )
max = value3;
}
return max;
int main()
{
int int1, int2, int3;
cout << "Input three integer values: ";
cin >> int1 >> int2 >> int3;
cout << "The maximum integer value is: "
<< maximum( int1, int2, int3 );
// int version
Function Templates (Example contd..)

31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48 }
double double1, double2, double3;

cout << "\nInput three double values: ";
cin >> double1 >> double2 >> double3;
cout << "The maximum double value is: "
<< maximum( double1, double2, double3 ); // double version
char char1, char2, char3;
cout << "\nInput three characters: ";
cin >> char1 >> char2 >> char3;
cout << "The maximum character value is: "
<< maximum( char1, char2, char3 )
<< endl;
return 0;
Input three
The maximum
Input three
The maximum
Input three
The maximum
// char version
integer values: 1 2 3
integer value is: 3
double values: 3.3 2.2 1.1
double value is: 3.3
characters: A C B
character value is: C
Overloading Vs. Generic

Function overloading when ?
We go for overloading when the functions business
logic should change with input types.
Function Template when ?
We go for templates when irrespective of the input types
the business logic of the function remains same.
Function template Internals

With a function template we realize the compiler carries
out a process called instantiation.
Instantiation is a process where the compiler generates
type specific code out of the template plan.
This happens during compile-time by the compiler,
subject to the compiler can figure out the needs of the
template consumer.
Effectively this is another form of overloading, a
responsibility taken by the compiler and not the
programmer.
Function Template Specialization

Consider the following Generic function:
template<typename T1, typename T2>
void Add(T1 a, T2 b)
{.}
Complete specialization of the above for char type

template<> //template<> usage is optional.
void Add(char a, char b)
{.}
A partial specialization of the above generic code is

not possible.
MODULE 4
OBJECT MODEL
Simple object model

Our first object model is very simple. It might be used for a
C++ implementation
Designed to minimize the complexity of the compiler at
the expense of space and run-time efficiency.
In this model, an object is a sequence of slots, where
each slot points to a member.
The members are assigned a slot in the order of their
declarations.
The no. of slots would equal the total number of members
Simple object model (contd..)

class CA
{
private:
int a,b;
public:
void fun1(){ }
void fun2(){ }
};
int CA::a
int CA::b
void CA::fun1(){ }
CA obj;
void CA::fun2(){ }
Table-driven object model

For an implementation to maintain a uniform representation
for the objects of all classes, an alternative object model
might factor out all member specific information, placing it
in a data member and member function pair of tables.
class CA
{
private:
int a,b;
public:
void fun1(){ }
void fun2(){ }
};
int CA::a
int CA::b
Member Data table
CA obj;
& void CA::fun1(){ }
Member function table
(holds addresses)
& void CA::fun2(){ }
C++ object model

Stroustrups original (and still prevailing) C++ object
model is derived from the simple object model by
optimizing for space and access time.
Non-static data members are allocated directly within
each class object. Static data members are stored
outside the individual class object.
class CA
{
private:
int a,b;
public:
void fun1(){ }
void fun2(){ }
};
int CA::a
This pointer
int CA::b
CA obj;
void CA::fun1(){ }
void CA::fun2(){ }
Class internals
Every member function of a class has an hidden formal
parameter called the this pointer.
The hidden formal parameter is designed to hold the
address of the object which was responsible for calling
that member function.
The compiler translates the code into a more detailed form as
illustrated in the subsequent slide. Consider the following piece of
code.
class CA
{
private:
int a,b;
public:
CA();
CA(int);
CA(int,int);
~CA();
void print() const;
};
//A const method
Class internals
CA::CA(){ .. }
CA::CA(int x){ .. }
CA::CA(int x, int y){..}
CA::~CA(){ .. }
void CA::print() const {..}
main()
{
CA obj1;
CA obj2(200);
CA obj3(30,40);
//.....
obj1.print();
obj3.CA::~CA();
obj2.CA::~CA();
obj1.CA::~CA();
}
---> CA::CA(CA * const this) { ...}

---> CA::CA(CA *const this, int x){ ... }
---> CA::CA(CA *const this, int x, int y) { ...}
----> CA::~CA(CA *const this){ .. }
----> void CA::print(const CA * const this){..}
// obj1.CA::CA(); ---> CA::CA(&obj1);

// obj2.CA::CA(200); --> CA::CA(&obj2, 200);
// obj3.CA::CA(30,40); --> CA::CA(&obj3, 30,40);
// CA::print(&obj1);
---> CA::~CA(&obj3);
---> CA::~CA(&obj2);
---> CA::~CA(&obj1);
Single level Inheritance
CA
Base
class
class CA
{
//.
Derived
class
CB
};
class CB: public CA
{
//
};
Multi-level Inheritance
class CA
CA
{
//.
};
CB
class CB: public CA

{
//
CC
};
class CC: public CB
{
CD
//
};
class CD: public CC
{
//
};
One-to-many Inheritance
class CA
{
//.
};
CA
class CB: public CA

{
CB
CC
CD
//
};
class CC: public CA
{
//
};
class CD: public CA
{
//
};
Multiple Inheritance
class CA
{
//.
CA
CB
CC
};
class CB
{
//
CD
};
class CC
{
//
};
class CD: public CA,public CB, public CC
{
//
};
Hybrid Inheritance (dreaded diamond)

class CA
{
//.
CA
CB
};
CC
CD
class CB:public CA
{
//
};
class CC:public CA
{
//
};
class CD: public CB, public CC
{
//
};
Object Layout
class CA
//BASE CLASS
CA::a
CA::b
private:
int a,b;
CA OBJECT
LAYOUT
public:
//...
};
CA::a
class CB:public CA
CLASS
{
private:
int a,b;
//DERIVED
CA::b
CB::a
CB::b
public:
//....
};
CB OBJECT
LAYOUT
Sub object
Static Members
static Class Members

static class members
Shared by all objects of a class
Normally, each object gets its own copy of each variable
Efficient when a single copy of data is enough

Only the static variable has to be updated
May seem like global variables, but have class scope

Only accessible to objects of same class
Initialized at file scope

Exist even if no instances (objects) of the class exist
Can be variables or functions
public, private, or protected
static
Class Members (II)
Accessing static members

public static variables: accessible through any
object of the class
Or use class name and (::)
Employee::count
private static variables: a public static

member function must be used.
Prefix with class name and (::)
Employee::getCount()
static member functions cannot access non-static

data or functions
No this pointer, function exists independent of objects
Static Member Example

#include<iostream>
using namespace std;
class Employee
{
private:
static int E_count;
public:
Employee()
{
E_count = E_count + 1;
}
virtual ~Employee() //Destructor must be virtual
{ E_count = E_count - 1; }
static void Emp_Count();
};
int Employee::E_count=0;
void Employee::Emp_Count()
{
cout <<"Total no. of Employee objects " << E_count << endl;
}
//------------------------------------------
Static Member Example (contd..)

class Perm_Emp:public Employee
{
private:
static int P_count;
public:
Perm_Emp() { P_count = P_count + 1; }
~Perm_Emp(){ P_count = P_count - 1; }
static void Perm_Count();
};
int Perm_Emp::P_count=0;
void Perm_Emp::Perm_Count()
{ cout <<"Total no. of Permanent Employee objects " << P_count << endl; }
//-------------------------------------------class Temp_Emp:public Employee
{
private:
static int T_count;
public:
Temp_Emp() { T_count = T_count + 1; }
~Temp_Emp(){ T_count = T_count - 1; }
static void Temp_Count();
};
Static Member Example (contd..)

int Temp_Emp::T_count=0;
void Temp_Emp::Temp_Count()
{
cout <<"Total no. of Temporary Employee objects " << T_count << endl;
}
void main()
{
Employee *p1 = new(nothrow) Perm_Emp;
Employee *p2 = new(nothrow) Perm_Emp;
Employee *p3 = new(nothrow) Temp_Emp;
Employee::Emp_Count();
Perm_Emp::Perm_Count();
Temp_Emp::Temp_Count();
cout <<"-----------------"<<endl;
delete p1;
delete p3;
Perm_Emp::Perm_Count();
Temp_Emp::Temp_Count();
cout <<"-----------------"<<endl;
delete p2;
}
Static Member Example (output)
Total no. of Employee objects 3

Total no. of Permanent Employee objects 2
Total no. of Temporary Employee objects 1
----------------Total no. of Employee objects 1
Total no. of Permanent Employee objects 1
Total no. of Temporary Employee objects 0
----------------Total no. of Employee objects 0
mutable Class Member

C++ provides mutable class member as an alternative
to const_cast.
A mutable data member is always modifiable even in
a const method of a const object.
The difference between const_cast and mutable is:
- for a non-mutable data member in a const
method, every time you modify it, you have to use
const_cast.
- This actually reduces the chance of accidental
modification.
mutable Class Member USAGE

class Screen
{
private:
int N_rows, N_cols;
mutable int R_location, C_Location;
public:
//.
};
const Screen TextMode(25,80,1,1);
Given the above example the R_Location and

C_Location are the only data members modifiable on
the Screen object which is const.
Note: mutable is class level qualifier, not applicable for global or
local variables.
Operator Overloading
Introduction (II)
Operator overloading
Use traditional operators with user-defined objects
Straightforward and natural way to extend C++
Requires great care
When overloading misused, program difficult to understand
Fundamentals of Operator Overloading
Use operator overloading to improve readability

Avoid excessive or inconsistent usage
Format
Write function definition as normal
Function name is keyword operator followed by the symbol
for the operator being overloaded.
operator+ would be used to overload the addition operator
(+)
Fundamentals of Operator Overloading (II)
Assignment operator (=)

may be used with every class without explicit overloading
memberwise assignment
Same is true for the address operator (&)
Restrictions on Operator Overloading
Most of C++s operators can be overloaded
Restrictions on Operator Overloading (II)
Arity (number of operands) cannot be changed

Urnary operators remain urnary, and binary operators
remain binary
Operators &, *, + and - each have unary and binary versions
Unary and binary versions can be overloaded separately
Restrictions on Operator Overloading (III)
No new operators can be created

Use only existing operators
Built-in types
Cannot overload operators
You cannot change how two integers are added
Operator Functions as Class Members vs. as

friend Functions
Operator functions
Can be member or non-member functions
Overloading the assignment operators

i.e:(), [], ->,=
Operator must be a member function

friend Functions (II)
Operator functions as member functions

Leftmost operand must be an object (or reference to an
object) of the class
If left operand of a different type, operator function must be a
non-member function
A non-member operator function must be a friend if private
or protected members of that class are accessed directly

friend Functions (III)
Non-member overloaded operator functions
Enable the operator to be commutative
HugeInteger bigInteger;
int integer;
bigInteger = integer + bigInteger;
or
bigInteger = biginteger + integer;
Overloading Stream-Insertion and StreamExtraction Operators
Overloaded << and >> operators

Must have left operand of types ostream &,
istream & respectively
It must be a non-member function (left operand not an object
of the class)
It must be a friend function if it accesses private data
members
Example
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
//An example on
// Overloading the stream-insertion and
// stream-extraction operators.
#include <iostream>
using
using
using
using
using
std::cout;
std::cin;
std::endl;
std::ostream;
std::istream;
#include <iomanip>
using std::setw;
class PhoneNumber {
friend ostream &operator<<( ostream&, const PhoneNumber & );
friend istream &operator>>( istream&, PhoneNumber & );
private:
char areaCode[ 4 ];
char exchange[ 4 ];
char line[ 5 ];
};
// 3-digit area code and null

// 3-digit exchange and null
// 4-digit line and null
// Overloaded stream-insertion operator (cannot be

// a member function if we would like to invoke it with
// cout << somePhoneNumber;).
ostream &operator<<( ostream &output, const PhoneNumber &num )
{
Example (contd..)
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
output << "(" << num.areaCode << ") "

<< num.exchange << "-" << num.line;
return output;
// enables cout << a << b << c;
istream &operator>>( istream &input, PhoneNumber &num )

{
input.ignore();
// skip (
input >> setw( 4 ) >> num.areaCode; // input area code
input.ignore( 2 );
// skip ) and space
input >> setw( 4 ) >> num.exchange; // input exchange
input.ignore();
// skip dash (-)
input >> setw( 5 ) >> num.line;
// input line
return input;
// enables cin >> a >> b >> c;
}
int main()
{
PhoneNumber phone; // create object phone
cout << "Enter phone number in the form (123) 456-7890:\n";
// cin >> phone invokes operator>> function by
// issuing the call operator>>( cin, phone ).
cin >> phone;
// cout << phone invokes operator<< function by

// issuing the call operator<<( cout, phone ).
cout << "The phone number entered was: " << phone << endl;
return 0;
Program Output
OUTPUT:
Enter phone number in the form
(123) 456-7890:
(800) 555-1212
The phone number entered was:
(800) 555-1212
Overloading Unary Operators

Overloading unary operators
Avoid friend functions and friend classes unless
absolutely necessary.
Use of friends forcefully violates the encapsulation of a

class.
As a member function:
class String
{
public:
bool operator!() const;
...
};
Overloading Binary Operators

Overloaded binary operators
Non-static member function, one argument
Non-member function, two arguments
class String
{
public:
const String &operator+=( const String & );
...
};
y += z;
equivalent to
y.operator+=( z );
Overloading Binary Operators

Example
class String
{
friend const String &operator+=
( String &, const String & );
...
};
y += z;
equivalent to
operator+=(y,z);
Converting between Types

Cast operator
Convert objects into built-in types or other objects
Conversion operator must be a non-static member
function.
Cannot be a friend function
Do not specify return type
For user-defined class A
A::operator char *() const;
// A to char
A::operator int() const;
//A to int
A::operator otherClass() const; //A to otherClass
When compiler sees (char *) s it calls

s.operator char*()
Converting between Types (II)
The compiler can call these functions to create

temporary objects.
If s is not of type char *
Calls A::operator char *() const; for
cout << s;
Overloading Unary Operators

Pre/post-incrementing/decrementing operators
Can be overloaded
How does the compiler distinguish between the two?
Prefix versions overloaded same as any other prefix unary
operator would be. i.e. d1.operator++(); for ++d1;
Postfix versions
When compiler sees postincrementing expression, such as
d1++;
Generates the member-function call
d1.operator++( 0 );
Prototype:
Date::operator++( int );
MODULE 5
Constructor
A constructor is a special method or member function
of a class, incidentally taking the name of the class as
the method name.
It is a method that does not return any value or has
any return type associated with it.
This method may or may not take any formal
parameters. That is it can be overloaded if desired.
This method is called constructor for the reason, It
would get implicitly invoked by an object at the time of
its construction hence the name CONSTRUCTOR.
Constructor (contd..)
This method by itself does not do anything on the

object,
It is the programmers responsibility to state inside the
function body as to what he/she wishes to accomplish
on the object at the time of its construction.
A constructor method that does not take any known
formal parameter is also called default constructor.
Destructor
A destructor is a special method or member function
of a class taking the name of the class as its method
name only prefixed with a ~ tilde symbol.
This method does not return any value or has any
return type associated with it.
This method does take any known formal parameters
either. That is it cannot be overloaded.
This method is called destructor for the reason, it
would get implicitly invoked by an object at the time of
objects destruction hence the name DESTRUCTOR.
Destructor (contd..)
If there is any clean-up job a programmer wishes to
accomplish at the time of the objects destruction, then
such business can be employed inside this method.
Note:
It is not compulsory to have constructor or destructor methods in a class.
These methods are need based.
At the same time it does not mean that if a class has constructor
method(s) we should also have destructor method.
The compiler does not assume a default constructor or destructor
method unless and until we provide one.
There are certain specific scenarios where a compiler would assume
default constructor and destructor methods. These scenarios are
highlighted in the slides to follow.
Default constructor synthesis

Case 1:
Base class has a default constructor but derived class
does not have one. Then the compile would assume a
default constructor for the derived class.
class CA
{
public:
CA(){ }
};
class CB:public CA
{
public:
//no default constructor provided, compiler will assume one
};

Case 2:
A class exhibits containment i.e one or more data
members of the class is an object of another class. This
container class does not have a default constructor,
whereas the contained objects class has a default
constructor, then the compiler would assume a default
constructor for the container class.
class CA
{
public:
CA(){ }
};
class CB
{
CA obj1;
//CB has-a CA
public:
//no default constructor provided, compiler will assume one
};
Case 3:
When a class CD multiple inherits virtually. Default
constructors for class CB,CC & CD would be
automatically assumed by the compiler.
class CA {. };
class CB: virtual public CA {. };
class CC: virtual public CA {};
class CD :public CB, public CC {};

Case 4:
All polymorphic classes that do not have a default
constructor provided by the programmer, then the
compiler would assume a default constructor for such
classes.
class CA
{
public:
//no default constructor provided, but compiler will assume one.
virtual void fun(){ }
};
The above cases are aptly applicable for destructor

methods as well.
Initializer list
One of the primary jobs a programmer would wish to use
the constructor for is to initialize the object at the time of
its construction.
Consider the following class.
class rectangle
{
private:
int base, height;
public:
//constructor placing values in objects data
rectangle(int x=0, int y=0)
{
base=x;
height=y;
}
};
The constructor in the above example is rather doing an assignment.
Initializer list (contd..)

Prefer initialization to assignment, which is more efficient.
How? The example revised
class rectangle
{
private:
int base, height;
public:
rectangle(int x=0, int y=0):base(x), height(y)
//initialization
//list
{
// base=x;
// height=y;
//avoid these statements
}
};
NOTE: It is suggested that the order of the data members listed in

the initializer-list shall be in the order of declaration.
Shallow copy and Assignment

All compiler synthesized copy constructors exhibit
shallow copy, also called bit-wise copy or member-tomember copy.
Such approaches may not be a good idea in problem
situations where the class exhibits containment or
aggregation.
Similarly the same is true for assignment functions
synthesized by the compiler, also called shallow
assignment, bit-wise assignment or member-tomember assignment.
Containment (HAS-A relationship)
CB obj1
CB obj2
CA *p
&&
CA *p
&&
S
T
A
C
K
CB obj1(10,20);
CB obj2(30,40);
CA::a 10
CA::b 20
CA::a 30
CA::b 40
CA
CA
H
E
A
P
Shallow Copy
CB obj1
CA *p
&&
CB obj2
Shallow
copy
CA *p
&&
S
T
A
C
K
CB obj2(obj1);
compiler synthesized
copy constructor
CB(const CB &x):p(x.p) {..}
CA::a 10
CA::b 20
CA
H
E
A
P
Shallow Assignment
CB obj3
CB obj3(40,50), obj4;
Obj4 = obj3;
CB obj4
CA *p
&&
Shallow
assignment
CA *p
&&
S
T
A
C
K
compiler synthesized
Assignment function
void operator =(const CB &x)
{
p = x.p;
}
CA::a 40
CA::b 50
CA
LEAKED
MEMORY
CA::a
CA::b
CA
0
0
H
E
A
P
Deep copy and Assignment
When under situations the compiler synthesized copy

constructor is not a good idea we need to go in for
custom copy constructor to tackle the issues.
Note: If any reason the programmer feels the need to have a custom
copy constructor then also make it a point to introduce the
assignment function as well.
Deep Copy
CB obj2(obj1);
CB obj1
CB obj2
CA *p
&&
CA *p
&&
Custom copy constructor

CB(const CB &x)
{
p = new(nothrow) CA(*(x.p));
}
CA::a 10
CA::b 20
CA
deep
copy
CA::a 10
CA::b 20
CA
S
T
A
C
K
H
E
A
P
Deep Assignment
CB obj3(40,50), obj4;
Obj4 = obj3;
CB obj3
CB obj4
CA *p
&&
CA *p
&&
S
T
A
C
K
Custom Assignment function

void operator =(const CB &x)
{
*(this->p) = *(x.p);
}
CA::a 40
CA::b 50
CA
deep
assignment
CA::a 40
CA::b 50
CA
H
E
A
P
Preventing object destruction

#include<iostream>
//preventing object destruction in the destructor
class CA
{
private:
int a,b;
public:
CA(int x=0, int y=0):a(x),b(y)
{
cout <<"CA - constructor" << endl;
}
~CA()
{
if(b==0)
throw b;
cout <<"CA destructed " << endl;
}
void print()
{
cout <<a<<","<<b<<endl;
}
};
Preventing object destruction

void main()
{
CA *p;
try
{
p = new CA(10,10);
p->print();
p->CA::CA();
delete p;
}
catch(int x) {cout <<"caught int - exception" << x <<endl; }
p->print(); //object continues to exist..
}
EXPLICIT CONSTRUCTOR
When a method is expecting an object which has a
one-argument constructor but you pass the argument,
compiler will implicitly call that one-argument
constructor and convert that passed argument to the
object:
class Base
{
public:
Base(int a) : member(a)
{
cout << "Base constructor called with " << a << endl;
}
int member;
};
void test(Base obj1)
{ cout << "Base object's member = " << obj1.member; }
EXPLICIT CONSTRUCTOR(contd..)
int main()
{
test(333);
}
The output will be:
Base constructor called with 333
Base object's member = 333
In same cases such an automatic implicit conversion

may cause trouble.
You can use keyword explicit in front of the
constructor to suppress the implicit conversion.
If you want to convert, you have to use explicit
conversion such as static_cast, without which compiler
will reject the conversion and prompt error message.
EXPLICIT CONSTRUCTOR -Usage

class Base
{
public:
explicit Base(int a) : member(a)
{
cout << "Base constructor called with " << a << endl;
}
int member;
};
void test(Base obj1)
{
cout << "Base object's member = " << obj1.member;
}
int main()
{
test( static_cast<Base>(333) );
return 0;
}
Patterns & Techniques for construction

#include<iostream>
//Prevent object instance
class CA
{
private:
CA();
~CA();
public:
//
};
void main()
{
CA *p = new(nothrow) CA;
CA obj;
//ERROR
}
//ERROR

(contd..)
#include<iostream>
//Prevent Stack based object
class CA
{
private:
CA() { }
public:
static CA* create_heapinst()
{
return new CA;
}
};
void main()
{
CA *p = CA::create_heapinst();
//CA obj; ERROR
}

(contd..)
#include<iostream>
#include<new>
// preventing heap based object
class CA
{
private:
void *operator new(size_t size) { return 0; }
void operator delete(void *pv) { }
void * operator new[](size_t size) { return 0; }
void operator delete[](void *pv) { }
public:
//.....
};
void main()
{
CA obj1;
}

(contd..)
#include<iostream>
#include<new>
//Identifying object is on heap or stack
bool isonheap;
class CA
{
public:
CA()
{
if(isonheap)
cout <<"object created in heap"<<endl;
else
cout <<"object created on stack " << endl;
isonheap = false;
}
void * operator new(size_t size)
{
isonheap = true;
return ::operator new(size); //call the global new operator
}

(contd..)
//Identifying object is on heap or stack (contd..)
void * operator new[](size_t size)
{
cout <<"void * operator new[](size_t size) - called"<<endl;
isonheap = true;
return malloc(size);
}
};
void main()
{
CA *p = new CA;
CA obj;
cout <<" p = "<< p << endl;
cout <<"& obj = " << &obj <<endl;
}
MODULE 6
new & delete
Dynamic Memory Allocation with Operators

new and delete
new and delete

new - automatically creates object of proper size, calls
constructor, returns pointer of the correct type
delete - destroys object and frees space
Better dynamic memory allocation than Cs malloc and free
Example:
TypeName *typeNamePtr;
Creates pointer to a TypeName object
typeNamePtr = new TypeName;

new creates TypeName object, returns pointer (which
typeNamePtr is set equal to)
delete typeNamePtr;
Calls destructor for TypeName object and frees memory
Dynamic Memory Allocation with Operators

new and delete (II)
Initializing objects
double *thingPtr = new double( 3.14159 );
Initializes object of type double to 3.14159
int *arrayPtr = new int[ 10 ];

Create ten element int array, assign to arrayPtr.
Use
delete [] arrayPtr;
to delete arrays
Pointers and Dynamic Memory Management

Static memory is used to store the values of global variables, as
well as static variables within functions and classes.
Stack-based memory is implicitly managed by function calls and
function returns.
Heap-based memory is used to dynamically manage memory as a
result of a programs request, such as Java's new call.
All temporaries created in a statement (numeric values, objects) are
allocated from the stack and destroyed at the end of statement.
string language = string ("C") + "++";
Static memory is by default set zero at the start of the program; no

other memory is initialized by default (all un-initialized primitive
variables contain unspecified random values).

(contd..)

(contd..)
Dynamic memory allocation means that a block of
memory requested by a program is removed from the
heap, and can be used by this program.
Dynamic memory deallocation means that a program
returns a block of memory to the heap - the deallocated memory can be used the next time a memory
request occurs.
The whole heap memory is (usually) returned to the
operating system (OS) only after the termination of the
program.

(contd.)
In Java, garbage collector decides when inaccessible

objects are deallocated.
Java objects are always allocated on the heap and

never on the run-time stack, and all memory
deallocation is done automatically by the garbage
collector.
Memory leakage occurs when memory that is no longer
needed by the program has not been freed and made
available for reuse.
What's a reference to an object? It is (usually) a

memory address; specifically, the address of a block of
memory allocated for this object on the heap.

(contd.)
In C++:
objects may be allocated both on the stack and in the heap
there is no standard garbage collector (utility libraries exist)
This may lead to various programming errors, mainly:
dangling reference problem - the result of a variable
referencing a memory block whose lifetime has expired
(allocated on the stack or heap and then deallocated,
creating a so-called zombie object)
memory leakage.
The programmer is fully responsible for explicitly
deallocating memory (using idioms and/or appropriate
abstractions).

(contd.)
Possible memory management errors are:

memory leakage
dangling references ("zombies")
duplicate deallocation (already deleted)
reallocation and use of memory still used by a zombie (a
"ghost"?)
use and deallocation of not allocated memory (undefined
pointer)
overflow of object or array boundaries (with pointers or
whatever)
but "delete 0;" is allowed and OK.

(contd.)
How to prevent C++ memory management errors:
the run-time system handles local stack-allocated objects,
temporary objects in expressions, and parameters (provided other
things go well..)
define explicit ownership for dynamic objects (who deletes?)
hide dynamic objects behind abstract data types (e.g., std::string)
use appropriate constructors, destructors, copy operations
use auto_ptr <>, reference counting, smart pointers (explained later)
after deallocation, remember to assign zero to the pointer
Placement new
There are many uses of placement new.
The simplest use is to place an object at a particular
location in memory.
This is done by supplying the place as a pointer
parameter to the newpart of a newexpression:
#include <new> // Must #include this to use "placement new"
class Employee {};
void Fun()
{
char memory[sizeof(Employee)];
void* place = memory;
Employee* f = new(place) Employee();
// The pointers f and place will be equal
...
}
ADVICE: Don't use this "placement new" syntax unless you have to.
Use it only when you really care that an object is placed at a
particular location in memory.
Placement new (contd..)

DANGER:
You are taking sole responsibility that the pointer you
pass to the "placement new" operatorpoints to a
region of memory that is big enough
and is properly aligned for the object type that you're
creating.
You are also solely responsible for destructing the
placed object.
This is done by explicitly calling the destructor:
void Fun()
{
char memory[sizeof(Employee)];
void* p = memory;
Employee* f = new(p) Employee();
...
f->~Employee(); // Explicitly call the destructor for the placed object
}
This is about the only time you ever explicitly call a destructor.
MODULE 7
Private Inheritance
Private inheritance gives is-implemented-in-terms-of
relationship:
Inherit just the implementation of the various operations plus data
structures. For example, you may need to operate on data with
more stringent requirements, and so may need to restrict the
original interface.
Further, these traits privately acquired by this derived class is
totally sealed and cannot be inherited further down the inheritance
hierarchy in a multi-level inheritance.
Protected Inheritance
All public features of the base class become protected
features of the derived class (protected features of the
base class continue to be protected in the derived class
- the same for private features)
Useful if you do not want to finalize the derivations, that
is you want to make it possible to have additional derived
classes:
class PermanentEmp : protected Employee

{
public:
PermanentEmp ();
~PermanentEmp ();
//.
};
Inheritance & Access Specifiers

CA
PRIVATE
PRIVATE
PROTECTED
PUBLIC
PROTECTED
PUBLIC
PROTECTED PUBLIC
CB
PRIVATE
class CB:private CA
CB
{.};
PRIVATE INHERITANCE
PRIVATE
PROTECTED
PUBLIC
class CB:protected CA
CB
{.};
PROTECTED INHERITANCE
class CB:public CA
{.};
PUBLIC INHERITANCE
Containment Vs. Inheritance

Private inheritance is a syntactic variant of composition
(aggregation and/or has-a).
E.g., the car has-a Engine" relationship can be expressed
using simple composition:
class Engine
{
public:
Engine(int numCylinders);
void start();
};
// Starts this Engine
class Car
{
public:
Car() : e_(8) { }
// Initializes this Car with 8 cylinders
void start() { e_.start(); }// Start this Car by starting its Engine
private:
Engine e_;
// Car has-a Engine
};

The "Car has-a Engine" relationship can also be
expressed using private inheritance:
class Car : private Engine
{
// Car has-a Engine
public:
Car() : Engine(8) { }
// Initializes this Car with 8 cylinders
using Engine::start;
// Start this Car by starting its Engine
//Expose the hidden member Publicizer Technique.
};
There are several similarities between these two

variants:
In both cases there is exactly one Engine member object
contained in every Car object
In neither case can users (outsiders) convert a Car* to an
Engine*
In both cases the Car class has a start() method that calls the
start() method on the contained Engine object.

There are also several distinctions:
The simple-composition variant is needed if you want to contain
several Engines per Car
The private-inheritance variant can introduce unnecessary
multiple inheritance
The private-inheritance variant allows members of Car to convert
a Car* to an Engine*
The private-inheritance variant allows access to the protected
members of the base class
The private-inheritance variant allows Car to override Engine's
virtual functions
The private-inheritance variant makes it slightly simpler (20
characters compared to 28 characters) to give Car a start()
method that simply calls through to the Engine's start() method
Hybrid Inheritance
class CA
//BASE CLASS
CA
{ private: int a;
public:
//...
};
CA::a
class CB:public CA
CB::b
CB
CC
CA::a
CC:c
{
private: int b;
public:
//...
CB
CB OBJECT
OBJECT
LAYOUT
LAYOUT
CD
};
class CC:public CA
{
private: int c;
public:
//...
};
CA::a
CB::b
class CD:public CB, public CC
CA::a
CC:c
private: int d;
public:
};
CD::d
//...
CD OBJECT
LAYOUT
CC
CC OBJECT
OBJECT
LAYOUT
LAYOUT
Virtual Inheritance
class CA
//BASE CLASS
{ private: int a;
CA
ptr
ptr
};
CA::a
CA::a
class CB: virtual public CA
CB::b
public:
//...
CB
CC
CC:c
{
private: int b;
public:
//...
CB
CB OBJECT
OBJECT
LAYOUT
LAYOUT
CD
};
class CC: virtual public CA
{
private: int c;
public:
//...
};
CA::a
ptr
CB::b
ptr
class CD:public CB, public CC
CC:c
CD::d
private: int d;
public:
};
CA::a
//...
CD OBJECT
LAYOUT
CC
CC OBJECT
OBJECT
LAYOUT
LAYOUT
C++ main programming

rules to remember
What is OOP?
Object-Oriented Programming is a philosophy where the
source code of a program is split into reusable objects.
What is an object, then?
An object is made of two parts:
- The interface = catalog of the object features
-The implementation = internal machinery
The interface in C++

Usually defined in a header (.h) file:
class Car
{
public:
// Members can be accessed from any object
protected:
// Can only be accessed by Car and its derived objects
};
private:
// Can only be accessed by Car for its own use.
Aggregation or Composition?
Car
Wheel
Person
Brain
Aggregation is a relationship
in which one object is a part
of another.
Composition is a relationship
in which one object is an
integral part of another
A aggregates B
=
B is part of A, but their
lifetimes may be different
A composes B
=
B is part of A, and their
lifetimes are the same
Ex: cars
and wheels, engine,
c
etc.
Ex: person and brain, lung,

etc.
Classes: Basic Design Rules

Hide all member variables
Hide implementation functions and data
Minimize the number of public member functions
Avoid default constructors
Avoid overloading (can be ambiguous)
Use const members whenever possible / needed
Be aware of compiler generated functions
Inheritance: quick review
A circle is a shape
GeomShape
Circle
class GeomShape {
//
};
class Circle : public GeomShape {
//
};
file.h
Public Inheritance Philosophy

Public inheritance is a
In other words:
What is applies to a base class applies to its
derived classes
Three aspects to consider:
class public interface
class relationship with derived classes
class internal cookware
Polymorphism
Mechanism that allows a derived class to modify the behavior
of a member declared in a base class
class Employee {
Employee
public :
virtual float income(); // 1000
};
Boss
class Boss : public Employee {
public :
virtual float income(); // 10000
};
Polymorphism (contd..)
A pure virtual function just defines the interface, and leaves
the implementation to derived classes
Employee
class Employee {
public :
Boss
virtual float income() = 0;
// not implemented
};
class Boss : public Employee {
public :
virtual float income(); // implemented
};
Public Inheritance Philosophy

Inheritance of
the interface
Inheritance of
the
implementation
Non virtual
function
Mandatory
Mandatory
Virtual
function
Mandatory
By default,
Possible to
redefine
Pure virtual
function
Mandatory
Re-definition is
mandatory
Private Inheritance Philosophy

Private inheritance is implemented in term of
This is an equivalent variant of aggregation:
class Car : private engine
{
//
};
class Car
{
private:
Engine engine_;
};
Inheritance and functions

Can have
Virtual
Function
Pure
Virtual
function
Construct
or is
Destructor
is
Isolated
class
No
No
Public
Public
Base class
Yes (must)
Yes / No
Public
Public
Virtual
Pure
Abstract
class
Yes (must)
Yes / No
Protected
Public
Virtual
Derived
(concrete
class)
Yes (must)
No
Public
Public
Virtual
Consequences
Never call a virtual function in a constructor

Never declare a virtual function inline
Calling a virtual function is more expensive than calling a
non-virtual function
Be aware of the increased size of classes with virtual
functions
Additional guidelines...
Avoid multiple inheritance: use composition
Forbid default parameters in virtual functions
Dont redefine (overload) a non virtual function
Differentiate between layering and inheritance
Advanced Principles of
Class design
What causes Code Rot?
Its been blamed on stupidity, lack of discipline, and phases of the moon, but...
A case study
The Copy Routine
First Version
All designs start well
Copy
ReadKeyboard
WritePrinter
void copy(void)
{
int ch;
while( (ch=ReadKeyboard()) != EOF)
WritePrinter(ch);
}
The program is an overnight success!
How could it be more simple, elegant, and maintainable?
Second Version
Oh, no! Nobody said the requirements might change!
We sometimes want to read from paper tape reader.

We could put a parameter in the call, but we have
hundreds of users already!
No big deal, this is just an exception we can make it
work.
Second Version Design

Copy
ReadKeyboard
WritePrinter
ReadTape
bool GtapeReader = false; // remember to clear

void copy(void)
{
int ch;
while( (ch=GtapeReader ? ReadTape(): ReadKeyboard())
!= EOF)
WritePrinter(ch);
}
Third Version
How unexpected! Requirements changed again!
It seems that sometimes we need to write to a paper

tape punch. Weve had this problem before, and just
added a flag. Looks like it should work again.
bool GtapeReader = false;
bool GtapePunch = false;
// remember to clear
void copy(void)
{
int ch;
while((ch=GtapeReader ? ReadTape():ReadKeyboard())!= EOF)
GtapePunch ? WritePunch(ch) : WritePrinter(ch);
}
Example of a Good Design

First and only version.
void Copy()
{
int c;
while( (c=getchar()) != EOF)
putchar(c);
}
But wait! Arent we supposed to be learning OO

design? This isnt OO is it?
is it?
It is a small program based on abstractions!
FILE is an abstraction
- It represents some kind of byte stream

- It has many variations
It has methods
- Read, Write, getchar, putchar, etc

- The methods are *dynamically* bound
FILE is a class, just implemented differently.
Rephrased in OO
Copy
interface
Reader
KeyboardReader
interface
Writer
PrinterWriter
class Reader
{ virtual char read()=0; };
class Writer
{ virtual void write(char c)=0; };
void copy(Reader &itsReader, Writer &itsWriter)
{
int c;
while( (c==itsReader.read()) != EOF )
itsWriter.write(c);
}
Dependency Management Review

Why do programs tend to rot over time?
What are four qualities of good designs?
Are OO programs always simpler than non-OO
versions?
Why would anyone want to use a paradigm that may
result in more complex designs?
Why are compile and link time important?
Class Design Principles
SRP: The Single Responsibility Principle

OCP: The Open/Closed Principle
LSP:
The Liskov Substitution Principle
ISP:
The Interface Segregation Principle
DIP:
The Dependency Inversion Principle
The Single Responsibility Principle
A class should have one, and only one, reason to

change.
Payroll
Employee
+ CalcPay
+ ReportHours
+ WriteEmployee
The Single Responsibility Principle. (SRP)
Report
Writer
Employee
Payroll
+ CalcPay
Employee
Repository
Open/Closed Principle
Modules should be open for extension, but closed for modification
-Bertrand Meyer
A principle which states that we should add new

functionality by adding new code, not by editing old
code.
Defines a lot of the value of OO programming
Abstraction is the key
Abstraction is Key
Abstraction is the most importantwordinOOD
Client/Server relationships are

open
Changes to servers cause
changes to clients
Abstract servers close clients
to changes in implementation.
Client
Client
Server
Abstract
Server
Concrete
Server
The Shape Example

Procedural (not closed) implementation
OO (closed) implementation
Procedural (open) version

Shape.h
enum ShapeType {circle, square};
struct Shape
{enum ShapeType itsType;};
Circle.h
struct Circle
{
enum ShapeType itsType;
double itsRadius;
Point itsCenter;
};
void DrawCircle(struct Circle*)
Square.h
struct Square
{
enum ShapeType itsType;
double itsSide;
Point itsTopLeft;
};
void DrawSquare(struct Square*)
DrawAllShapes.cpp
#include <Shape.h>
#include <Circle.h>
#include <Square.h>
typedef struct Shape* ShapePtr;
void
DrawAllShapes(ShapePtr list[], int n)
{
int i;
for( i=0; i< n, i++ )
{
ShapePtr s = list[i];
switch ( s->itsType )
{
case square:
DrawSquare((struct Square*)s);
break;
case circle:
DrawCircle((struct Circle*)s);
break;
}
}
}
What is wrong with the code?

It can be demonstrated to work. Isnt that the important thing?
DrawAllShapes is not closed.

-
Switch/case tend to recur in diverse places.

If we add a shape, we add to the switch/case
All switch/case statements must be found and edited.
Switch/Case statements are seldom this tidy
When we add to the enum, we must rebuild everything
The software is both rigid and brittle
A Closed Implementation
Shape.h
class Shape
{
public:
virtual void Draw() const =0;
};
Square.h
class Square: public Shape
{
public:
virtual void Draw() const;
};
Circle.h
class Circle: public Shape
{
public:
virtual void Draw() const;
};
DrawAllShapes.cpp
#include <Shape.h>
void DrawAllShapes(Shape* list[],int n)
{
for(int i=0; i< n; i++)
list[i]->draw();
}
Strategic Closure
No program is 100% closed.
Closure Against What?
Closure is strategic. You have to choose which changes
youll isolate yourself against.
What if we have to draw all circles first? Now DrawAllShapes
must be edited (or we have to hack something)
Opened Where?
Somewhere, someone has to instantiate the individual
shapes.
Its best if we can keep the dependencies confined
Open/Closed Review
What does the open/closed principle say?

What does that mean practically?
How can it be achieved?
What is strategic closure?
- How can this be achieved in design?
- What if you cant close completely?
Liskov Substitution Principle
Derived classes must be usable through the base class

interface, without the need for the user to know the
difference.
All derived classes must be substitutable for

their base classes
This principle guides us in the creation of
abstractions.
Square/Rectangle
A square is-a rectangle, right? So lets
consider Square as a subtype of Rectangle.
We can make it work:

Rectangle
Uh, oh. This

doesnt quite
seem to fit
-height : real
-width : real
+SetHeight()
+SetWidth()
Square
void Square::SetWidth(double w)
{
width = w;
height = w;
}
void Square::SetHeight(double h)
{
width = h;
height = h;
}
Substitution denied!
It is reasonable for users of a rectangle to expect that
height and width may change independently.
These expectations are preconditions and
postconditions
- Bertrand Meyer calls it Design by Contract
- Post condition contract for rectangle is
width = new Width
height = old height.
Square violates Rectangles contract
Liskov Substitution Principle (cont.)
A client of rectangle expects height and width to be

changed independently
- void setAspectRatio( Rectange* r, double ratio );
By deriving Square from Rectangle, we are allowing

someone to set the aspect ratio of a Square !
We can still make it work
- if ( typeid(r) == typeid(Rectangle) )
- Violates Open/Closed Principle !
Liskov Substitution Principle (cont.)

Design by Contract
- Bertrand Meyer
- Preconditions, postconditions, invariants
Rectangle's postconditions for setWidth()

- width = newWidth
- length = oldLength
Square can require no more of clients, nor promise

any less
- Doesn't maintain invariant of length
- Violates the contract
LSP Guides the Creation of Abstractions
Abstractions do not stand alone

Abstractions dont always conform to real world
expectations
Violating LSP is tantamount to violating the OCP
LSP Review
What does the LSP state?

What is the risk if LSP is violated?
What is Design by Contract?
Dependency Inversion Principle
Details should depend on abstractions.

Abstractions should not depend on details.
Copy
Copy
Vs.
ReadKeyboard
Reader
Writer
KeyboardReader
PrinterWriter
WritePrinter
DIP Implications
Everything should depend upon abstractions

Avoid deriving from concrete classes
Avoid associating to concrete classes
Avoid aggregating concrete classes
Avoid dependencies on concrete components
Dependency Inversion Principle (cont.)
Legitimate reasons to violate

- Creation of objects
new Circle creates a dependency on concrete class
localize these dependencies using factories
- Nonvolatile classes
string, vector, etc.
- providing they are stable
Interface Segregation Principle

Helps deal with fat or inappropriate interfaces
Sometimes class methods have various groupings.

These classes are used for different purposes.
Not all users rely upon all methods.
This lack of cohesion can cause serious dependency
problems
These problems can be refactored away.
Interface Pollution by collection
Distinct clients of our class have distinct interface needs.
Client 1
Client 2
Server
+ C1Function()
+ C2Function()
+ C3Function()
Client 3
A Segregated Example
Client 1
Client 2
Client 3
interface
interface
interface
Client 1
Client 2
Client 3
+ C1Function()
+ C2Function()
Server
+ C1Function()
+ C2Function()
+ C3Function()
+ C3Function()
ATM UI Example
Withdraw
Deposit
Transfer
interface
ATM UI
+ GetWithdrawAmountAndAccount()
+ GetDepositAmountAndAccount()
+ GetTransferAmountAndAccounts()
French ATM UI
English ATM UI
A Segregated ATM UI Example
Deposit
interface
ATM Deposit UI
Withdraw
Transfer
interface
interface
ATM Transfer UI
ATM Withdraw UI
+ GetWithdrawAmountAndAccount
interface
ATM UI
+ GetWithdrawAmountAndAccount()
French ATM UI
English ATM UI
Logger Example
Log Control
Application
Code needing to
log something
interface
LogManager
interface
+ SendLotToFile(name)
EventLog
+ LogToMem(size)
+ Log(s: string*)
+ LogToConsole()
ManagedEventLog
+ Log(s: string*)
+ SendLotToFile(name)
+ LogToMem(size)
+ LogToConsole()
ISP Review
What is the ISP?

What does it affect?
How do these problems arise?
Does it really provide a real solution or just a
restructuring of the problem?
When is it worth the effort?
Four Class Design Principles - Review
OCP Extend function without editing code

LSP Child instances substitute cleanly for base
DIP
Depend on abstractions instead of details
ISP
Split interfaces to manage dependencies
Virtual function Issues

It is strongly advised not to employ any statement that is going to call
a virtual function in a base class constructor method.
As this statement would always end up in calling only the base class
virtual function and not the appropriate derived class method in spite
of the object belonging to a derived class.
#include<iostream> //Calling virtual function from constructor
class CA
{
public:
CA() { this->fun(); }
virtual void fun(){ cout <<"CA fun"<<endl;}
};
class CB:public CA
{
public:
void fun(){ cout <<"over-ridden CB fun"<<endl;}
};
void main()
{
CB obj1;
}

It is strongly advised not to employ any statement that is going to
call a virtual function in a base class destructor method.
As this statement would always end up in calling only the base
class virtual function and not the appropriate derived class method
in spite of the object belonging to a derived class.
#include<iostream> //Calling virtual function from destructor
class CA
{
public:
~CA(){cout <<"CA destructor"<<endl; this->fun(); }
};
class CB:public CA
{
public:
};
void main()
{
CB obj1;
}

Unlike the base class constructor or destructor methods, it is
appropriate to employ call to virtual member functions in a base
class.
Such a call in an non-virtual member function of the base would
always end up in calling the correct derived class method to which
the object belongs.
#include<iostream> //Calling virtual function from non-virtual
//member function
class CA
{
public:
void nonvirtual(){ this->fun();}
};
class CB:public CA
{
public:
};
void main() //Calling virtual function from non-virtual member

{
//function
CA *p = new(nothrow) CB;
p->nonvirtual();
delete p;
Virtual function issues

Whenever we plan to pass derived polymorphic class
objects as function parameters pass the same by
reference if they are stack objects.
If we pass by value it would lead to object slicing, that is
upon the derived class object getting casted to a base kind
we would end up loosing sub-object part, further the vptr
field would end up pointing to the base class VTABLE and
not the derived.
Consider the example in next slide..

This code leads to OBJECT SLICING
#include<iostream>
class CA
{
public:
};
class CB:public CA
{
public:
};
//global function
void compute(CA x)
{
x.fun();
}
void main()
{
CB obj1;
compute(obj1);
}
//by value, leads to slicing

This code AVOIDS OBJECT SLICING
#include<iostream>
class CA
{
public:
};
class CB:public CA
{
public:
};
//global function
void compute(CA &x) //by reference, AVOIDS slicing
{
x.fun();
}
void main()
{
CB obj1;
compute(obj1);
}
VTABLE Layout
Virtual Tables & Late Binding

void circle::draw() { }
void rectangle::draw() { }
circle::~circle() {}
rectangle::~rectangle() {}
& circle::draw()
& circle::~circle()
& rectangle::draw()
& rectangle::~rect()
circle
class
VTABLE
vptr
@@
rectangle class
VTABLE
Circle object
rectangle
object
@@
&&&
&&&
circle *p
S
E
G
M
E
N
T
vptr radius length breadth
radius
##
D
A
T
A
circle *p
##
##
##
H
E
A
P
S
T
A
C
K
Multi-Level Inheritance & VTABLE Layout
Multiple Inheritance & VTABLE Layout
CC obj1
RTTI
RTTI-Runtime Type Identification

RTTI
Because of polymorphism, when you write your program, you
may not be able to know the exact type you are dealing with.
Run-time type information (RTTI) is used to find out the type of
a polymorphic object.
Header file <typeinfo> defines operator typeid. It returns a const
reference of class type_info, which is a class description of the
operand.
Class type_info has a method name, which returns a char *
which is the type name of the operand.
RTTI Table
OBJECT
OBJECT
LOCATOR
LOCATOR
VTABLE
VTABLE
CLASS
CLASS
HIERARCHY
HIERARCHY
-1 &
& RTTI
RTTI OBJECT
OBJECT
COUNT
COUNT
2
(RTTI
(RTTI BASE
BASE CLASS
CLASS ARRAY)
ARRAY)
LOCATOR
LOCATOR
&
& TYPE
TYPE DESCRIPTOR
DESCRIPTOR
&
& RTTI
RTTI BASE
BASE
CLASS
CLASS ARRAY
ARRAY
4 &
& CLASS
CLASS HIERARCHY
HIERARCHY
DESCRIPTOR
DESCRIPTOR
&
& RTTI
RTTI BASE
BASE CLASS
CLASS
DESCPIPTOR.
DESCPIPTOR.
CONST
CONST TYPEINFO
TYPEINFO &
&
2
3
RTTI
RTTI TYPE
TYPE
DESCRIPTOR
DESCRIPTOR
4
5
6
BASE
BASE CLASS
ARRAY
ARRAY
&
& RTTI
RTTI TYPE
TYPE
DESCRIPTOR
DESCRIPTOR
BASE
BASE CLASS
CLASS
DESCRIPTOR
DESCRIPTOR
Example 1
// Getting the run time type information
#include <iostream>
#include <typeinfo>
// polymorphic base class...
class __rtti Test
{
// This makes Test a polymorphic class type.
virtual void func() {};
};
// derived class...
class Derived : public Test {};
int main(void)
{
// Instantiate Derived type object...
Derived DerivedObj;
// Declare a Derived type pointer
Derived *DerivedPtr;
// Initialize the pointer
DerivedPtr = &DerivedObj;
//continued.
Example 1 (contd..)
// do the run time checking...
if(typeid(*DerivedPtr) == typeid(Derived))
// check the type of *DerivedPtr
cout<<"Ptr *DerivedPtr type name is <<typeid(*DerivedPtr).name();
if(typeid(*DerivedPtr) != typeid(Test))
cout<<"\nPointer DerivedPtr is not a Test class type.\n";
return 0;
}
OUTPUT:
c and f are different type.
int before double: 1
double before int: 0
class A before class B: 1
Example 2
#include <iostream>
#include <typeinfo>
class Base
{
public:
virtual void funct(){}
};
class Derived:public Base{};
int main()
{
Derived* Test1 = new Derived;
Base* Test2 = Test1;
cout<<"The type name of Test1 is: ";
cout<<typeid(Test1).name()<<endl;
cout<<"The type name of *Test1 is: ";
cout<<typeid(*Test1).name()<<endl;
cout<<"The type name of Test2 is: ";
cout<<typeid(Test2).name()<<endl;
cout<<"The type name of *Test2 is: ";
cout<<typeid(*Test2).name()<<endl;
delete Test1;
return 0;
}
Example 2 (output)
OUTPUT:
The type name of Test1 is: class Derived *
The type name of *Test1 is: class Derived
The type name of Test2 is: class Base *
The type name of *Test2 is: class Derived
Gang of Four (GOF) Patterns at a glance

Overview
Design patterns are ways to reuse design solutions that
other software developers have created for common and
recurring problems.
Design patterns are not function or class "building block

solutions".
In other words, a design pattern does not provide a
reusable source code component.
The design pattern provides a framework that is
customized for each particular problem need.
The benefits from these patterns include reductions in
design and debugging time (sometimes quite dramatic).
Abstract Factory
Intent
Provide an interface for creating families of related or
dependent objects without specifying their concrete
classes.
Abstract Factory (contd..)

Problem
If an application is to be portable, it needs to encapsulate platform
dependencies. These "platforms" might include: windowing
system, operating system, database, etc. Too often, this
encapsulation is not engineered in advance, and lots of #ifdef
case statements with options for all currently supported platforms
begin to procreate like rabbits throughout the code.
AbstractProduct
AbstractFactory
+createProduct()::AbstractProduct
ConcreteFactory
ConcreteProduct1
ConcreteProduct2
<<creates>>
<<creates>>
Factory Method
Intent
Define an interface for creating an object, but let
subclasses decide which class to instantiate. Factory
Method lets a class defer instantiation to subclasses.
Factory Method (contd..)

Problem :
A framework needs to standardize the architectural

model for a range of applications, but allow for individual
applications to define their own domain objects and
provide for their instantiation.
Product
Creator
Product=factoryMethod();
+factoryMethod()::Product
+operation():void
Concrete Product
0..*
ConcreteCreator
<<creates>>
+factoryMethod()::ConcreteProduct
return new
ConcreteProduct();
Singleton
Intent
Ensure a class has only one instance, and provide a
global point of access to it.
Problem
Application needs one, and only one, instance of an
object. Additionally, lazy initialization and global access
are necessary.
Singleton
-instance:Singleton
+Singleton()
+operation():void
+getinstance():instance
return a copy of self
Bridge
Intent
Decouple an abstraction from its implementation so
that the two can vary independently.
Bridge (contd..)
Problem
"Hardening of the software arteries" has occurred by using
subclassing of an abstract base class to provide alternative
implementations. This locks in compile-time binding between
Interface and implementation. The abstraction and
implementation cannot be independently extended or
composed.
Abstraction
imp.operatorImpl();
Implementor
imp
<<creates>>
+operationImp():void
+operation():void
RefinedAbstraction
ConcreteImplementorA
ConcreteImplementorB
Template Method
Intent
Define the skeleton of an algorithm in an operation,
deferring some steps to client subclasses. Template
Method lets subclasses redefine certain steps of an
algorithm without changing the algorithm's structure.
Template Method (contd..)

Problem
Two different components have significant similarities,
but demonstrate no reuse of common interface or
implementation. If a change common to both
components becomes necessary, duplicate effort must
be expended.
AbstractClass
primitiveOperation1();
+templateMethod():void
+primitiveOperation1():void
ConcreteClass
primitiveOperation2();
MODULE 8
Class Template
Class templates encourage software reusability by
enabling type-specific versions of generic classes to
be instantiated.
Class templates are called parameterized types,
because they require one or more type parameters to
specify how to customize a "generic class" template to
form a class-template specialization.
The programmer who wishes to produce a variety of
class-template specializations writes only one classtemplate definition.
Class Template
Each time an additional class-template specialization is
needed, the programmer uses a concise, simple
notation, and the compiler writes the source code for
the specialization the programmer requires.
//A complete generic class supporting two unique types
template<typename T1, typename T2> class CA
{.....};
//complete specialization of the above generic class

template<>class CA<int, float>
{...}
//partial specialization, where one type is specific and the

other //generic
template<typename T1, typename T2> class CA<int, T2>
{...}
Example
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
// Stack class template.

#ifndef STACK_H
#define STACK_H
template< typename T >
class Stack
{
public:
Stack( int = 10 ); // default constructor (Stack size 10)
// destructor
~Stack()
{
delete [] stackPtr; // deallocate internal space for Stack
} // end ~Stack destructor
bool push( const T& ); // push an element onto the Stack
bool pop( T& ); // pop an element off the Stack
// determine whether Stack is empty
bool isEmpty() const
{
return top == -1;
} // end function isEmpty
Example(contd..)
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
// determine whether Stack is full

bool isFull() const
{
return top == size - 1;
} // end function isFull
private:
int size; // # of elements in the Stack
int top; // location of the top element (-1 means empty)
T *stackPtr; // pointer to internal representation of the Stack
}; // end class template Stack
// constructor template
template< typename T >
Stack< T >::Stack( int s )
: size( s > 0 ? s : 10 ), // validate size
top( -1 ), // Stack initially empty
stackPtr( new T[ size ] ) // allocate memory for elements
{
// empty body
} // end Stack constructor template
// push element onto Stack;
// if successful, return true; otherwise, return false
Example(contd..)
51 template< typename T >
52 bool Stack< T >::push( const T &pushValue )
53 {
54
if ( !isFull() )
55
{
56
stackPtr[ ++top ] = pushValue; // place item on Stack
57
return true; // push successful
58
} // end if
59
60
return false; // push unsuccessful
61 } // end function template push
62
63 // pop element off Stack;
64 // if successful, return true; otherwise, return false
65 template< typename T >
66 bool Stack< T >::pop( T &popValue )
67 {
68
if ( !isEmpty() )
69
{
70
popValue = stackPtr[ top-- ]; // remove item from Stack
71
return true; // pop successful
72
} // end if
73
74
return false; // pop unsuccessful
75 } // end function template pop
76 #endif
Code Blow
Whenever we are designing generic classes we got to
look into an very important issue called code-blow.
What is code-blow?
Code-blow is nothing but redundant or unwanted code
instantiated by the compiler without the knowledge of the
generic class consumer.
When does it happen?
If a generic class consists of any member function(s)
dealing only with type-specific data and generic, then
such a situation is possible.
Consider the example in the following slide.
Code Blow (contd..)

#include<iostream>
//An example on code blow
template<typename T> class Blow
{
private:
int a,b;
//type specific members
T c,d;
//generic members
public:
Blow(T x, T y, int i, int j):c(x),d(y),a(i), b(j)
{
}
void PrintAll()
{
cout <<"a = " << a <<", b="<< b << endl;
cout <<"c = " << c <<", d= "<< d << endl;
}
void PrintInt()
{
cout <<"a = " << a <<", b="<< b << endl;
}
};
Code Blow (contd..)

void main()
{
Blow<int> obj1(10,20, 300, 400);
obj1.PrintAll();
cout <<"--------------------"<<endl;
obj1.PrintInt();
cout <<"=============="<<endl;
Blow<float> obj2(12.34f, 54.47f, 500,600);
obj2.PrintAll();
cout <<"--------------------"<<endl;
obj2.PrintInt();
}
Given the code above, we would notice that the compiler would
have instantiated or generated 2 PrintInt() functions where both
functions deal only with int type data no matter the object is
specialized for int or float type.
Code Blow Avoid

#include<iostream>
//An example on avoiding code blow
class AvoidBlow
{
protected:
int a,b;
public:
AvoidBlow(int x, int y):a(x),b(y)
{ }
void PrintInt()
{
cout <<"a = " << a <<", b="<< b << endl;
}
};
template<typename T> class Blow:public AvoidBlow
{
private:
//int a,b;
T c,d;
public:
Blow(T x, T y, int i, int j):c(x),d(y), AvoidBlow(i,j) {
Code Blow Avoid (contd..)

void PrintAll()
{
cout <<"a = " << a <<", b="<< b << endl;
cout <<"c = " << c <<", d= "<< d << endl;
}
/*
void PrintInt()
{ cout <<"a = " << a <<", b="<< b << endl;
*/
};
void main()
{
Blow<int> obj1(10,20, 300, 400);
obj1.PrintAll();
cout <<"--------------------"<<endl;
obj1.PrintInt();
cout <<"=============="<<endl;
Blow<float> obj2(12.34f, 54.47f, 500,600);
obj2.PrintAll();
cout <<"--------------------"<<endl;
obj2.PrintInt();
}
Standard Template Library
Introduction
Industrial revolution - Reuse software, not re-write

Two main paradigms that permit code reuse - Object
oriented design, Generic Programming
OOD facilitates code reuse by data hiding Inheritance and Polymorphism
Generic programming facilitates code reuse by data
independence - Templates and Overloading
STL is a combination of both OOD and Generic
programming - collection of generic algorithms and
containers that communicate through iterators.
Need for STL

Reusability - adaptability and efficiency
Highly adaptive components are usually implemented
using complex inheritance, virtual function and RTTI inefficient to be widely used
Highly efficient components - written in low level,
platform dependent code - non-portable, hard to
maintain
Solution:- Templates - containers and algorithms
Drawback - Programmer implemented (home made),
not portable, less than 100% bug free and no
common interface.
Thus, the lack of platform independent, bug free generic
components gave rise to the creation of a library containing
generic components with generic algorithms and common
interface.
Structure of STL
The Standard Template Library contains the following

five components
Containers
Iterators
Algorithms
Function objects
Adaptors
Containers
Holds other object as its element

The allocation and de-allocation of these objects are
controlled through constructors, destructors, insertion,
and erase operation
There are two types Sequence containers - vector, queue, deque, and list
Associative containers - set, map, and multimap
Sequence Containers
Organizes a finite set of same type objects into a strictly
linear arrangement
There are 3 basic types - vector, list and deque
The containers queue and stack are implemented using
deque
The following are the header files used:
Header
vector
list
deque
queue
stack
Contents
An array of T
A doubly linked list of T
A double ended queue of T
A queue of T
A stack of T
Associative Containers
Facilitates fast retrieval of data based on keys
There are four types of associative containers - set,
multiset, map, multimap
Parameterized on key and an ordering relation Compare
The keys in set and map are unique keys
The keys in multiset and multimap are equal keys
The following header files are used:
Header
map
set
bitset
Contents
An associative array of T
A set of T
A set of Boolean values
Sequence Containers - Example 1

#include<iostream>
#include<vector>
#include<string>
//using namespace std;
int main(){
//An object of vector containing string objects are created
vector <string> vs;
//Make room for 10 strings using the reserve(n) function.
//The reserve(n) function just allocates memory for n elements.
//Thus, this changes the value returned by the capacity() function.
vs.reserve(10);
//The member function push_back(T) appends a single element
//to the end of the container. If the container capacity
//is reached, it reallocates additional storage and appends
//the element.
vs.push_back(string());
//insert an element
//The size() function returns the number of elements that
Sequence Containers - Example 1(contd..)

//are currently stored in the container
cout << "Size: " << vs.size() << endl;
//The capacity() function returns the number of elements that
//the container can hold before reallocation
cout << "Capacity: " << vs.capacity() << endl;
//The difference in capacity() and size() gives the remaining
//space in the container.
cout << "There's room for " << vs.capacity() - vs.size()
<< " elements before reallocation" << endl;
//Allocate 10 more elements using resize(n); The resize(n)
//function,in addition to allocating memory for n elements,
//initializes them to default value. A different value can be
//provided as the second argument if needed. Thus, this changes the
//value returned by both capacity() and size()
vs.resize(20);
return 0;
}
Sequence Containers - Example 1 (output)
Size: 1
Capacity: 10
There's room for 9 elements before reallocation
Size: 20
Capacity: 20

This example illustrates the use of Front and Back operations.
#include<iostream>
#include<vector>
#include<string>
int main(){
vector <string> vs;
vs.push_back("string 1");
vs.push_back("string 2");
//insert an element
//insert an element

//The member function front() accesses a single
//element at the front of the container
//

cout << "Front: " << vs.front() << endl;
//The member function back() accesses a single
//element at the back of the container
//
cout << "Back: " << vs.back() << endl;
//The member function pop_back() removes the
//last element from the container. This funciton
//does not return the removed object.
vs.pop_back(); //Remove vs[1]
return 0;
}

Size: 2
Capacity: 2
Front: string 1
Back: string 2
Front: string 1
Back: string 1
Size: 1
Capacity: 2
This example illustrates the use of Container assignment.
#include<iostream>
#include<vector>
int main(){
//Create a vector of Short integers.
vector <short> vs;
vs.push_back(1);
vs.push_back(2);
vs.push_back(3);
vs.push_back(4);
//insert
//insert
//insert
//insert
an
an
an
an
element
element
element
element
//Get the size and capacity of the container

//Get the front and back elements of the container
Remove the last element
vs.pop_back(); //Remove vs[3]

cout
cout
cout
cout
<<
<<
<<
<<
"Front: " << vs.front() << endl;

"Back: " << vs.back() << endl;
"Size: " << vs.size() << endl;
"Capacity: " << vs.capacity() << endl;
//Create a new vector object

vector <short> new_vector;
//Assign the existing object to the new one
new_vector = vs;
//Print the size and capacity of the new vector
cout << "Size: " << new_vector.size() << endl;
cout << "Capacity: " << new_vector.capacity() << endl;
//Access the elements of the new vector using the index [] operator
cout << "New vector[0]: " << new_vector[0] << endl;
cout << "New vector[1]: " << new_vector[1] << endl;
cout << "Front: " << new_vector.front() << endl;
cout << "Back: " << new_vector.back() << endl;
return 0;
}

Size: 4
Capacity: 4
Front: 1
Back: 4
Front: 1
Back: 3
Size: 3
Capacity: 4
Size: 3
Capacity: 3
New vector[0]: 1
New vector[1]: 2
Front: 1
Back: 3
This example illustrates the use of push(), pop(), Front and Back
operations on a queue object.
#include<iostream>
#include<queue>
int main(){
queue <int> iq; // Create an integer queue object
iq.push(29);
iq.push(239);
iq.push(344);
iq.push(541);
cout << "Size of the queue is: " << iq.size() << endl;
while (!iq.empty()){
cout << "The top most element is: " << iq.front() << endl;
cout << "The bottom most element is: " << iq.back() << endl <<
endl;
iq.pop();
}
return 0;
}

Size of the queue is: 4
The top most element is: 29
The bottom most element is: 541
Algorithms
STL contains a rich collection of algorithms that can be
applied to a variety of containers.
There are approximately 70 standard algorithms
Operate on elements of containers indirectly through
iterators
There are three main types of algorithms.
Non-mutating sequence operations
Mutating sequence operations
Sorting Algorithms
Non-mutating sequence operations

These algorithms do not modify the sequence on which
they operate.
They include searching, checking for equality, and
counting, for example
Example find( ) - Locates an element within a sequence. Takes 3
arguments, the first 2 are iterators that point to begin and end of
sequence respectively. The third element is the value to be
found. Returns an iterator that points to the first element that is
identical to the value to be found.
Non-mutating sequence operations(contd..)
Modifies the sequence on which they operate

Includes operations such as copy, fill, replace, and
transform, for example
Example copy( ) - This is a generic function that can be used to copy a
sequence of objects to specific target. Takes 3 arguments.
The first 2 are the begin and end iterators of the object to be
copied and the third is the target object.
Sorting operations
Algorithms for sorting and merging sequences
Algorithms for set operations for sorted sequences
Includes sort(), partial_sort(), binary_search(), for
example
Example
sort( ) - takes two arguments of type const iterator that
point to the beginning and the end of the sequence
respectively.
MODULE 9
Exception Handling
Purpose of Exception Handling
Deal with errors in an elegant manner

Write clearer, most robust, and fault tolerant program
Error handling code
Minimum --- Casual Systems
Extensive --- Commercial products
Dealing with Errors

Errors are interspersed throughout the software. Errors are
handled where they occur.
Programmer can see the error processing near to the code and
determine the appropriateness of the error processing code.
Code becomes "polluted" with error processing. More difficult for a

programmer concerned with the application itself to read the code
and determine if the code is functioning completely.
User exception handling techniques

Remove error code from the "main code"
better readability and modifiability
Exception Handling remove the error handling code from

the main-line of the programs execution improves
program readability and maintainability catch all kinds of
exceptions, or a subset of them programs become more
robust
Dealing with Errors(contd..)

Memory allocation, Arithmetic Exceptions, Invalid
Parameters, Array Bounds, etc. synchronous errors.
Used in situations where the system can recover from
the error causing the exception the recovery
procedures is called the exception handler.
Exception handling is typically used when the error will
be dealt with by a different part of the program (i.e. a
different scope) from where the error was detected.
Particularly useful for graceful degradation.
Use only to process exceptional situations.

To process exceptions for program components that are
not geared to handle those exceptions directly.
To process exception from software components, such as
functions, libraries, classes, etc., that are likely to be
widely used and where it does not make sense for those
components to handle their own exceptions.
To handle error processing in a uniform manner across
large-scale projects.
Use assert macro - if an assertion is false, program
terminates - useful at debugging
Ignore exceptions!
Abort program - what if resources were allocated to a
program and it aborted? ("Resource Leak")
Set and test error indicators - need to check them at all
points in the program
Error Handling - Method to Cope with

Uncertainties
A Function which Finds that it Cannot Cope With a
Problem Throws an Exception, Hoping that Caller can
Handle that Problem
A Function that Wants to Handle that Kind of a Problem
can Indicate that it is Willing to Catch that Exception
Method of Transferring Control and Information to an
Associated Exception Handler
A try Block Constitutes the Section of the Program that is
Subject to Exception Checking
A Handler is Invoked with a throw Expression from within
the try Block
The Handler is the catch Function
Error Handling - Method to Cope with

Uncertainties (cont)
Flow of Control
Handler class
try Block
throw Expression
Control
constructor
Control
Catch function
Control
FUNCTION
Invoking of Handler
The Type of the Object Argument of the throw
Expression Determines the Appropriate catch
Invocation
A throw Expression Without an Argument Re-throws
the Exception Being Handled Again. Such a throw
Expression MAY APPEAR ONLY IN A HANDLER or in
a Function Directly Called From the Handler
The catch (...) Handler, if Present, CAN Handle ALL
Exception Classes
Invoking of Handler
If NO Appropriate Handler is Present, the
unexpected() Function is Invoked. It Passes the
Control to terminate() Function and can Cause
Termination
The throw Expression is First Passed to the
Appropriate Constructor for the Class -- A Default
Constructor MAY be Created, if Needed -- No
Constructor is Used for Non-Classes. The Processing
is then Passed to the Handler
catch Handlers
The catch Handler MUST IMMEDIATELY FOLLOW
the Appropriate try Block or Other catch Handlers
The Placement Order is Significant. Control is
Passed to the FIRST HANDLER THAT CAN
HANDLE THE CONDITION
The catch(...) Handler, if Present, SHOULD BE THE
LAST HANDLER
Simple Exception Handler Example 1

#include <iostream>
main()
{
//try Block
try
{
throw(long(1));
}
//Handlers
catch (int x)
{
cout << "Catch Integer " << x
}
catch (long x)
{
cout << "Catch Long " << x <<
}
catch (...)
{
cout << "Catch Rest" << endl;
}
}
OUTPUT WILL BE:

------ ---- -Catch Long 1
<< endl;
endl;
An Exception Handler Class Example 2

#include<stream>
#include<cstdlib> //exit()
class zero
{
public:
zero() //Constructor
{
cout<<"Class Zero Constructor Invoked!!"<<endl;
}
};
//Exception Checker Function
void zero_check(int i)
{
if (i == 0)
throw zero(); //Argument is of zero class type.
}

(contd..)
main()
{
//try block
try
{
for (int i = 2; ; i--)
{
OUTPUT WILL BE:

------ ---- --Reciprocal: 0.5
Reciprocal: 1
Class Zero Constructor Invoked!!
DIVIDE BY ZERO -- ERROR!!
zero_check(i);
cout << "Reciprocal: " << 1.0/i << endl;
}
}
//Handler
catch(zero)
{cout << "DIVIDE BY ZERO -- ERROR!!\n"; exit(-1);}
}
An Exception Handler Class Example -3

// A simple exception handling example.
// Checking for a divide-by-zero exception.
#include <iostream>
using std::cout;
using std::cin;
using std::endl;
// Class DivideByZeroException to be used in exception
// handling for throwing an exception on a division by zero.
class DivideByZeroException {
public:
DivideByZeroException()
: message( "attempted to divide by zero" ) { }
const char *what() const { return message; }
private:
const char *message;
};
// Definition of function quotient. Demonstrates throwing
// an exception when a divide-by-zero exception is encountered.
double quotient( int numerator, int denominator )
{
if ( denominator == 0 )
throw DivideByZeroException();
}
return static_cast< double > ( numerator ) / denominator;

(contd..)
// Driver program
int main()
{
int number1, number2;
double result;
cout << "Enter two integers (end-of-file to end): ";
while ( cin >> number1 >> number2 ) {
// the try block wraps the code that may throw an
// exception and the code that should not execute
// if an exception occurs
try {
result = quotient( number1, number2 );
cout << "The quotient is: " << result << endl;
}
catch ( DivideByZeroException ex ) { // exception handler
cout << "Exception occurred: " << ex.what() << '\n';
}
}
cout << "\nEnter two integers (end-of-file to end): ";
cout << endl;

return 0;
// terminate normally
An Exception Handler Class Example (3)

Output
OUTPUT:
Enter two integers (end-of-file to end): 100 7
The quotient is: 14.2857
Exception occurred: attempted to divide by zero
The quotient is: 3.66667
Enter two integers (end-of-file to end):
Throwing an Exception
throw - indicates an exception has occurred

Usually has one operand (sometimes zero) of any type
If operand an object, called an exception object
Conditional expression can be thrown
Code referenced in a try block can throw an exception

Exception caught by closest exception handler
Control exits current try block and goes to catch handler (if it exists)
Example (inside function definition)
if ( denominator == 0 )
throw DivideByZeroException();
Throws a dividebyzeroexception object
Exception not required to terminate program

However, terminates block where exception occurred
Catching an Exception
Exception handlers are in catch blocks
Format: catch( exceptionType parameterName)
{
exception handling code
}
Caught if argument type matches throw type

If not caught then terminate called which (by default)
calls abort
Example:
catch ( DivideByZeroException ex)
{
cout << "Exception occurred: " << ex.what() <<'\n'
}
Catches exceptions of type DivideByZeroException
Catching an Exception (contd..)
Catch all exceptions

catch(...) - catches all exceptions
- You do not know what type of exception occurred
- There is no parameter name - cannot reference the object
If no handler matches thrown object

Searches next enclosing try block
If none found, terminate called
If found, control resumes after last catch block

If several handlers match thrown object, first one found is
executed
catch parameter matches thrown object when

They are of the same type
Exact match required - no promotions/conversions allowed
The catch parameter is a public base class of the thrown object

The catch parameter is a base-class pointer/ reference type and
the thrown object is a derived-class pointer/ reference type
The catch handler is catch( ... )
Thrown const objects have const in the parameter type

Unreleased resources
Resources may have been allocated when exception
thrown
catch handler should delete space allocated by new and
close any opened files
catch handlers can throw exceptions

Exceptions can only be processed by outer try blocks
Rethrowing an Exception
Rethrowing exceptions
Used when an exception handler cannot process an
exception
Rethrow exception with the statement: throw;
No arguments
If no exception thrown in first place, calls
terminate
Handler can always rethrow exception, even if it

performed some processing
Rethrown exception detected by next enclosing try
block
Rethrowing an Exception -Example

// Demonstration of rethrowing an exception.
#include <iostream>
using std::cout;
using std::endl;
#include <exception>
using std::exception;
void throwException()
{
// Throw an exception and immediately catch it.
try {
cout << "Function throwException\n";
throw exception(); // generate exception
}
catch( exception e )
{
cout << "Exception handled in function throwException\n";
Rethrowing an Exception Example (contd..)

}
}
throw;
// rethrow exception for further processing
cout << "This also should not print\n";
int main()
{
try {
throwException();
cout << "This should not print\n";
}
catch ( exception e )
{
cout << "Exception handled in main\n";
}
cout << "Program control continues after catch in main" << endl;
return 0;
}
OUTPUT:
Function throwException
Exception handled in function throwException
Exception handled in main
Program control continues after catch in main
Exception specifications
Exception specification (throw list)

Lists exceptions that can be thrown by a function
Example:
int g( double h ) throw ( a, b, c )
{
// function body
}
Function can throw listed exceptions or derived types
If other type thrown, function unexpected called
throw() (i.e., no throw list) states that function will not throw any
exceptions
In reality, function can still throw exceptions, but calls unexpected
(more later)
If no throw list specified, function can throw any exception
Processing Unexpected Exceptions
Function unexpected
Calls the function specified with set_unexpected
Default: terminate
Function terminate
Calls function specified with set_terminate
Default: abort
set_terminate and set_unexpected

Prototypes in <exception>
Take pointers to functions (i.E., Function name)
Function must return void and take no arguments
Returns pointer to last function called by terminate or

unexpected
Usages
Arithmetic Exceptions -- Divide by Zero

Range Exceptions -- Overflow, Underflow
Memory Exceptions -- Dynamic Memory Allocations
Example-4
// domain_error and typeid()
#include <iostream>
int main()
{
try
{
throw domain_error("Some error with your domain!");
}
catch (exception &err)
{
cerr<<"Caught: "<<err.what()<<endl;
cerr<<"Type: "<<typeid(err).name()<<endl;
};
}
Output:
Caught: Some error with your domain!
Type: class std::domain_error
Example-5
// bad_typeid
#include <typeinfo>
#include <iostream>
class Test
{
public:
// object for a class needs vtable for the rtti
Test();
virtual ~Test();
};
int main()
{
Test *ptrvar = NULL;
try {
// the error condition
cout<<typeid(*ptrvar).name()<<endl;
}
catch (bad_typeid)
{
cout<<"The object is NULL"<<endl;
}
}
Output:
The object is NULL
Example-6
// set_unexpected()
#include <exception>
#include <iostream>
void myfunction()
{
cout<<"Testing myfunction()."<<endl;
// terminate() handler
terminate();
}
int main( )
{
unexpected_handler oldHandler = set_unexpected(myfunction);
// unexpected() function call
unexpected();
}
Output:
The Debug Error message box should be expected.
Anonymous namespace
References
C++ Primer by Lippman.
Inside Objects by Lippman
C++ By Bjarne Stroustrup
C++ Bible by AL Stevens
Complete Reference by Herbert Schildt.
Advanced C++ James coplin
Effective C++ Scott meyers
More Effective C++ Scott meyers
Effective STL Scott meyers
Design Patterns (GOF)

Advanced C++ Programming Advanced C++ Programming

Uploaded by

Copyright:

Available Formats

Advanced C++ Programming Advanced C++ Programming

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advanced C++ Programming Advanced C++ Programming

Uploaded by

Copyright:

Available Formats

Advanced

AMP Ganesh Prabhu, Bangalore,India

C++ supports a variety of programming styles:

Steps involved in the process of building the

The following slide illustrates the stages that happen in order

Steps involved in the process of building the

Steps involved in the process of building the

Compilation is the second pass.

Assembly is the third stage of compilation.

Steps involved in the process of building the

Memory Layout of a process

Memory Layout of a process(contd..)

Memory Layout of a process(contd..)

mfp, frame pointer for

Stack pointer (grows

Auto variables for func()

Available for heap growth

Global /static variables

Memory Layout of a process(contd..)

Memory Layout of a process(contd..)

Memory Layout of a process(contd..)

Memory Layout of a process(contd..)

Static and Dynamic link library

A library is a piece of pre-compiled (binary) code

who can implement or consume the functionalities

A library can be shipped generally in 2 different

Static and Dynamic link library (contd..)

This library would get linked to the .OBJ file of the

These files go with the extension

Static and Dynamic link library (contd..)

What is dependency management?

What is dependency management?

simple idea - as interdependencies increase,

What bearing does DM have on software?

and cohesion are the eternal concerns of

What is the penalty for practicing poor DM?

The impact of a change cannot be predicted

Changes with Rigidity

Officially Rigid Area

Where the change must be made now

Are we containing risk, or spreading rot?

single change requires a cascade of

errors appear in areas that seem

Defects v. Cumulative Modifications

Systems tend to become increasingly fragile over time. Intentional, planned

parts of the design are

It has high viscosity

the right changes are much more difficult

What is the benefit of good DM?

Occurs when building large systems from pieces

Classes suggest a solution

// Add Mine::g()to Mine

Function stack frames

This data structure working "behind the scenes supports the

It also supports the creation, maintenance and destruction of each

Function stack (Example)

Function stack (Fig .1)

Function stack (Fig. 2)

void DoIt(float a, char b, char c) attribute((cdecl)); //