A complex number is also a multivalue object. It consists of two parts: real part and imaginary part. The imaginary part carries symbol i which is equal to √-1 . A complete number is written as sum of the real part and the imaginary part as shown below.

Since structures are multivalue, multitype data objects, they can be considered to form a useful tool in manipulation of quantities such as vectors, complex variables, etc. For this we define functions of the type struct, i.e., the return value of the function is a structure. The function may have arguments of type struct besides other arguments. Let us define a structure with three components of vector as its data members as shown below.

When structures are used to keep record of employees, students, and so on, we would need to sort them according to some criterion. For employees, it may be according to alphabetical order, designation, or pay.

Program provides an illustration of a function with structure as one of its parameter. The function calculates and returns the magnitude of a vector which is an instance of structure defined to hold the three components of a vector as its data members. The structure is declared as below.

Pointers to structures may be declared as we declare pointers to any other data type. Like arrays and functions, the name of structure carries the address of the structure where the values of its various members are stored. Therefore, the pointer may be initialized by the name of structure which is a constant pointer to the structure. Program illustrates the declaration and use of pointer to a structure.

We have used the declaration of nested structures on the parent structure in the same declaration. The structures may also be declared separately and included in the parent structure. The innermost structure should be declared first, then the next enveloping structure and then the next enveloping structure. The code is illustrated in Program.

In this method, the declaration starts with typedef followed by key word struct. The data members of the structure are placed between a pair of curly braces after struct. The type name is placed after the closing right brace (}). This is explained in the following example:

As we already know, a structure is an ordered list of elements of either the same or different types. A union is similar to a structure but with one major difference that the structure stores all its members one after another, whereas the union can store only one member at a time, since all the members in a union are stored beginning from the same memory location. Thus, the total memory required to store a union is the same as that required for the largest member in it. The union is useful when we have to store one of the alternative values of different types and yet conserve space.

The program given below reads the information about a person and prints it on the screen. It uses a structure containing a pointer member name to represent the information about a person. Note that the program uses the dstr_read function.

A program to determine the number of words and average word length is given in Program. It uses the strtok function to separate the words in a given string. The program given below uses the same technique to separate the words in a given string and determine and print the frequency of these words.

The C language permits structures to be nested, in which a structure can contain structure members. A structure may be declared inside another structure. Also, a structure may be a data member of another structure. The important thing to note is how to initialize the data members of inner structures and how to access the data members of inner structures. The use of such nested structures facilitates representation and processing of complex data.

The C language allows an array to be used as a structure member. This enables us to group related structure members and is also useful for declaring character strings as structure members. This section explains how we can effectively use such structures. This includes

Consider the problem of comparison of two valid dates d₁ and d₂. There are three possible outcomes of this comparison: d₁ == d₂ (dates are equal), d₁ > d₂ (date d₁ is greater, i.e., occurs after d₂) and d₁ < d₂(date d₁ is smaller, i.e., occurs before d₂). Let us write a function that accepts two dates as structures dl and d2 of type struct date and returns 0 if the dates are equal, 1 if d₁ is later than d₂ and -1 if date d_l is earlier than d₂.

Let us use structure complex to represent a complex number and write a simple program containing only the main function. Thus, we can define structure complex either in the main function or before it as a global definition.

A union shall determine a single memory location where they can be stored several different variables. The declaration of a union is similar to that of a structure:

The computer has a main memory and CPU registers. The variables (as well as the program as a whole) are stored in memory. The modifier register tells the compiler that the variable in question should be, if possible, used in a CPU register.

Structures and Unions, Giving values to members, Initializing structure, Functions and structures, Passing structure to elements to functions, Passing entire function to functions, Arrays of structure, Structure within a structure and Union.

Union is a collection of data items of different data types. It can hold data of only one member at a time though it has members of different data types. If a union has two members of different data types, they are allocated the same memory.

A structure is a user-defined data type containing a collection of logically related data which may be of different types such as int, double, char, and so on. All of them are encapsulated (packed) in a single unit with a single name. The classes of C++ are in fact generalization of C-structures. A class in C++ with only public members is similar to a C-structure. All members of a structure in C are by default public. However, in a class declaration in C++, if there is no access specifier, the members are private by default. We know that arrays can be used to represent a group of data items that belong to the same type, such as int or float. This restriction is not there in structures.