C Programming: Structure and Union

C Programming : User-defined data types
By
Mr.S.Selvaraj
Asst. Professor (SRG) / CSE
Kongu Engineering College
Perundurai, Erode, Tamilnadu, India
Thanks to and Resource from : Sumitabha Das, “Computer Fundamentals and C Programming”, 1st Edition, McGraw Hill, 2018.

20CST11 – Problem Solving and Programming
3/10/2021 5.1 Structure and Union 2
Syllabus

Contents
• There are 4 types of user-defined data types.
1. Structure (struct)
 Structure basics
 Declaring and defining a structure
 Attributes of structures
 Nested structures
 Arrays as structure members
 Arrays of structure
 Passing structures as arguments to functions
 Bit Fields
2. Unions (union)
3. Enumerated type (enum)
4. Type definition (typedef)

Structure Basics
• C offers a basket of data types for handling different
levels of complexity in the organization of data.
• The primary data types (like int and float) are meant
for using simple and small amounts of data.
• The array is offered as a derived data type for handling
a group of data items of the same type.
• Real-life situations often involve the use of
heterogeneous data that need a mix of data types to
be handled as a single unit.
• Structures and unions meet this requirement perfectly.

Structure Basics
• A group of data items are often logically related to one
another.
• For instance,
– the attributes of an employee include the name, address,
date of birth, salary, and so forth.
• A structure is a data type that is designed by the user.
• However, unlike the other data types, a structure
variable is based on a template that must be created
first.
• Each member of the structure is accessed by the
notation structure.member, and can be used in
practically the same way as a variable.

Declaring and Defining a Structure
• a structure variable is created in two steps
which may or may not be combined:
– Declare the structure to create a template which
specifies the components of the structure.
• Declaration simply makes type information available
to the compiler. However, the compiler doesn’t
allocate memory by seeing the declaration.
– Define a variable that conforms to the template.
• The compiler allocates memory for the variable and
creates an instance of the template.

Declaring and Defining a Structure
• Declaration uses the struct keyword.
• The following generalized syntax declares a structure
named struct_name.
• The syntax of the definition also specifies the keyword
struct, the structure tag (struct_name) and the
variable name (struct_var).

Example

Accessing Members of a Structure

Combining Declaration, Definition and Initialization
• For an initialized structure, the variable name is followed by an = and a list of
comma-delimited values enclosed by curly braces.
• These initializers obviously must be provided in the right order.
• In Form 2, “Beethoven” is assigned to title and 3 to qty.
• You can create multiple variables and initialize them at the same time.
• Like with arrays, you can partially initialize a structure. If you initialize only
title, the remaining members, qty and price, are automatically assigned zeroes
• By default, uninitialized structures have junk values unless they are global.

Declaring without Structure Tag
• If declaration and definition are combined, the structure
tag can be omitted.
• This is usually done when no more structure variables of
that type require to be defined later in the program.
• You can’t subsequently create any more variables of this
type.

Memory Layout of Members of music_cd
• Note that sizeof computes the total memory
occupied by a structure.
• This is not necessarily the sum of the sizes of the
members.
• Thus, on this machine having a 4-byte word.
– title occupies 28 bytes (7 words) even though it uses
27 of them.
– qty and price individually occupy an entire word (4
bytes each), but qty uses two of these bytes.
– disk1 thus needs 33 bytes (27 + 2 + 4) but it actually
occupies 36 bytes.

Memory Layout of Members of music_cd
• Alignment issues related to structures lead to
the creation of slack bytes or holes in the
allocated memory segment.

Attributes Of Structures
• (1) There are no name conflicts between
structure templates, their instantiated
variables and members.

• (2) The =, dot, -> and & are the only operators
that can be used on structures.
• (3) A structure can be copied to another
structure provided both objects are based on
the same template.
• This feature is not available with arrays; all
array elements have to be copied individually

• (4) No arithmetic, relational or logical operations can be
performed on structures even when the operation logically makes
sense.
• (5) It is not possible to compare two structures using a single
operator even though such comparison could be logically
meaningful.
– Each member has to be compared individually as shown by the
following code.
– If a structure contains 20 members, you need to use 20 relational
expressions to test for equality.
– Unfortunately, C doesn’t support a better option.

• (6) When a structure is passed by name as an
argument to a function, the entire structure is
copied inside the function.
• (7) A member of a structure can contain a
reference to another structure of the same
type.
– This property of self-referential structures is used
for creating linked lists.

Nested Structures
• A structure member can have any data type—including
another structure (but not of the same type though).
• The inner structure can contain another structure and
so on.
• C supports nested structures and the following outer
structure named student contains an inner structure
named dt_birth as one of its four members.

Example

Initializing a Nested Structure
• When it comes to initialization, nested
structures resemble multi-dimensional arrays.
• For every level of nesting, a set of inner
braces have to be used for members of the
inner structure.

Accessing Members
• A member of an inner structure is connected
by a dot to its immediately enclosing
structure, which is connected by another dot
to the outer structure.

Arrays as structure members

Arrays Of Structures
• We used two structure variables, stud1 and stud2, to
handle data of two students.
• This technique won’t work with 500 students.
• C supports an array of structures, whose definition
may or may not be combined with the declaration of
the structure.
• The following statement defines an array of type struct
student that can hold 50 sets (or records) of student
data.

• Alternatively, you can separate the declaration and
definition.
• You can also use typedef to replace struct student with
STUDENT.

Structures in functions

Structures in functions
• A function using a structure as argument
copies the entire structure.
• If the structure contains an array, it too will be
copied.
• A program may run out of memory when
using structures as arguments.
• A structure containing numerous members
and large arrays can lead to stack overflow
and cause the program to abort

Bit fields
• Have our programs wasted memory?
– Yes, in many cases they have.
• We used 2-byte short integers to store each
component of a date when a few bits would
have sufficed.
• We also wasted 15 bits of the 16 available in
short to set a flag to 0 and 1

Bit fields
• It’s time to know that C supports a memory saving
feature—the bit field.
• It can be used with the right number of bits, rather
than bytes, to store data.
• One or two bits are adequate to represent the sex.
• while six or seven bits are good enough to represent
the age.
• Bit fields can be used only as structure members.
• Properly sized bit fields prevent both data overflow
and memory wastage.

Example
• The first two members, id and age, are bit fields.
• The declaration of a bit field specifies an int (signed or
unsigned), followed by the member name, a colon and the
number of allocated bits.
• Here, id and age occupy 10 and 6 bits, respectively.
• This means that the maximum values of id and age are
restricted to 1023 and 63, respectively.
• sizeof evaluates the size of emp to 32.
• A similar structure using two short variables for id and age
would occupy 34 bytes.

Unions
• When searching a database using a person’s id, you must
have often wanted to search by name as well.
• That is possible if the two attributes are held in a union, a
data type that resembles a structure in form.
• A union comprises two or more members that share the
same memory, with the most recent assignment to a
member being active at any instant.
• You can define a union with many members, but only one
member can contain a value at any given time.
• Unions provide an efficient way of using the same
memory location for multiple purposes.
• Consider the following declaration and definition of a union

Unions
• A union is declared, defined and accessed exactly
like a structure.
• The previous statement declares a two-member
union named uemp and simultaneously defines
the union variable u.
• The members are accessed as u.emp_id and
u.name.
• However, both members share a single memory
region and you can correctly access only one
member at a time—the one that was last
assigned.

Unions

Unique Attributes of Unions
• (1) All members of a union can’t be initialized
simultaneously.
• (2) The size of a union is determined by the size of the
member having the largest footprint in memory. (For
uemp, this would be 30, the size of the name field).
• (3) The compiler doesn’t output an error when you
access a member that was not the last one to be
assigned. A programmer must keep track of the
member that is active at any instant.
• (4) You can use a union to assign a value to a member
and read it back using a different member.

Unions
• A union can easily determine whether a
machine is little-endian or big-endian.
• In a big-endian machine, the most significant
bit is stored first while the reverse is true for
little-endian machines.
• However, unions are mainly used as memory-
saving objects.

Example

Difference between Structure and Union

The enumerated type (enum)
• We have used symbolic constants to represent
integers using the #define feature of the
preprocessor.
• C also supports the user-defined enumeration
data type that assigns names to a set of
integer values called enumerators.
• For instance, you can create a set of named
constants using the enum keyword.

• This statement creates a set of enumerated
integer values which, by default, begin with 0.
• YES evaluates to 0, while NO and CANT_SAY
have the values 1 and 2, respectively.
• If this declaration is made before main, you
can use NO to represent 1 anywhere in the
program.

• You can now create one or more enum
variables of type eday.

Abbreviating a Data Type : The typedef Feature
• The typedef keyword is used to abbreviate
the names of data types.
• Using A Data type LL instead of long long
involves less typing.
• The syntax of typedef is simple; follow
typedef with the existing data type and its
proposed synonym.

Example

Thank you

C Programming: Structure and Union

More Related Content

What's hot

Similar to C Programming: Structure and Union

More from Selvaraj Seerangan

Recently uploaded

In this document

C Programming: Structure and Union