Chapter Seven

File Operations (File I/O)

7.1 Introduction

- The data created by the user and assigned to variables with an assignment
statement is sufficient for some applications. With large volume of data
most real-world applications use a better way of storing that data. For this,
disk files offer the solution.
- When working with disk files, C++ does not have to access much RAM
because C++ reads data from your disk drive and processes the data only
parts at a time.

7.2 Stream
- Stream is a general name given to flow of data. In C++, there are different
types of streams. Each stream is associated with a particular class, which
contains member function and definition for dealing with file. Lets have a
look at the figure:

ios

istream ostream

iostream
ifstream ofstream

fstream

- According to the above hierarchy, the class iostream is derived from the
two classes’ istream and ostream and both istream and ostream are
derived from ios. Similarly the class fstream is derived from iostream.

75
 Generally two main header files are used iostream.h and fstream.h. The
classes used for input and output to the video display and key board are
declared in the header file iostream.h and the classes used for disk file
input output are declared in fstream.h.
- Note that when we include the header file fstream.h in our program then
there is no need to include iostream.h header file. Because all the classes
which are in fstream.h they are derived from classes which are in
iostream.h therefore, we can use all the functions of iostream class.

7.3: Operation With File

- First we will see how files are opened and closed. A file can be defined by
following class ifstream, ofstream, fstream, all these are defined in
fstream.h header file.
 if a file object is declared by ifstream class, then that object can be
used for reading from a file.
 if a file object is declared by ofstream class, then that object can
be used for writing onto a file.
 If a file object is declared by fstream class then, that object can be
used for both reading from and writing to a file

7.4: Types of Disk File Access

- Your program can access files either in sequential manner or random
manner. The access mode of a file determines how one can read, write,
change, add and delete data from a file.
- A sequential file has to be accessed in the same order as the file was
written. This is analogues to cassette tapes: you play music in the same
order as it was recorded.
- Unlike the sequential files, you can have random-access to files in any
order you want. Think of data in a random-access file as being similar to
songs on compact disc (CD): you can go directly to any song you want to
play without having to play or fast-forward through the other songs.

7.4.1: Sequential File Concepts

- You can perform three operations on sequential disk files. You can create
disk files, add to disk files, and read from disk files.

7.4.1.1: Opening and Closing Sequential Files

- When you open a disk file, you only have to inform C++, the file name
and what you want to do with it. C++ and your operating system work
together to make sure that the disk is ready, and they create an entry in
your file directory for the filename (if you are creating a file). When you

76
 close a file, C++ writes any remaining data to the file, releases the file from
the program, and updates the file directory to reflect the file’s new size.
- You can use either of the two methods to open a file in C++:
 using a Constructor or
 using the open function
- The following C++ statement will create an object fout of ofstream class
and this object will be associated with file name “hello.txt”.

Ofstream fout (“hello.txt”);

- This statement uses the constructor method.

- The following C++ statement will create an object fout of ofstream class
and this object will be associated with file name “hello.txt”.

ofstream fout;
fout.open(“hello.txt”);

- If you open a file for writing (out access mode), C++ creates the file. If a
file by that name already exists, C++ overwrite the old file with no
warning. You must be careful when opening files not to overwrite existing
data that you want.
- If an error occurs during opening of a file, C++ does not create a valid file
pointer (file object). Instead, C++ creates a file pointer (object) equal to
zero. For example if you open a file for output, but use an invalid disk
name, C++ can’t open the file and therefore makes the file object equal to
zero.
- You can also determine the file access mode when creating a file in C++. If
you want to use the open function to open a file then the syntax is:
fileobject.open(filename,accessmode);
 File name is a string containing a valid file name for your computer.
 Accessmode is the sought operation to be taken on the file and must be
one of the values in the following table.

Mode Description
app Opens file for appending
ate Seeks to the end of file while opening the file
in Opens the file for reading
out Opens the file for writing
binary Opens the file in binary mode

- You should always check for the successful opening of a file before
starting file manipulation on it. You use the fail() function to do the task:

77
 - Lets have an example here:
ifstream indata;
indata.open(“c:\\myfile.txt”,ios::in);
if(indata.fail())
{
//error discriton here
}
- In this case, the open operation will fail (i.e the fail function will return
true), if there is no file named myfile.txt in the directory C:\
- After you are done with your file manipulations, you should use the close
function to release any resources that were consumed by the file
operation. Here is an example
indata.close();
- The above close() statement will terminate the relation ship b/n the
ifstream object indata and the file name “c:\myfile.txt”, hence releasing
any resource needed by the system.

7.4.1.2: Writing to a sequential File

- The most common file I/O functions are
 get() and put()
 gets() and puts()
- You can also use the output redirection operator (<<) to write to a file.
- The following program creates a file called names.txt in C:\ and saves the
name of five persons in it:
#include<fstream.h>
#include<stdlib.h>
ofstream fp;
void main()
{
fp.open(“c:\\names.txt”,ios::out);
if(fp.fail())
{
cerr<< “\nError opening file”;
getch();
exit(1);
}
fp<< “Abebe Alemu”<<endl;
fp<< “Lemelem Berhanu”<<endl;
fp<< “Tesfaye Mulugeta”<<endl;
fp<< “Mahlet Kebede”<<endl;
fp<< “Assefa Bogale”<<endl;
fp.close();

78
 }//end main
Writing characters to sequential files:
- A character can be written onto a file using the put() function. See the
following code:
#include<fstream.h>
#include<stdlib.h>// for exit() function
…
void main()
{
char c;
ofstream outfile;
outfile.open(“c:\\test.txt”,ios::out);

if(outfile.fail())
{
cerr<< “\nError opening test.txt”;
getch();
exit(1);
}

for(int i=1;i<=15;i++)
{
cout<< “\nEnter a character : ”;
cin>>c;
outfile.put(c);
}
output.close();
}//end main

- The above program reads 15 characters and stores in file test.txt.

- You can easily add data to an existing file, or create new files, by opening
the file in append access mode.
- Files you open for append access mode (using ios::app) do not have to exist.
If the file exists, C++ appends data to the end of the file (as is done when
you open a file for write access).
- The following program adds three more names to the names.txt file
created in the earlier program.
#include<fstream.h>
#include<stdlib.h>
…
void main()

79
 {
ofstream outdata;
outdata.open(“c:\\names.txt”,ios::app);
if(outdata.fail())
{
cerr<< “\nError opening names.txt”;
getch();
exit(1);
}
outdata<< “Berhanu Teka”<<endl;
outdata<< “Zelalem Assefa”<<endl;
outdata<< “Dagim Sheferaw”<<endl;
outdata.close();
}//end main
- If the file names.txt does not exist, C++ creates it and stores the three names
to the file.
- Basically, you have to change only the open() function’s access mode to
turn a file-creation program into a file-appending program.

7.4.1.3: Reading from a File

- Files you open for read access (using ios::in) must exist already, or C++
gives you an error message. You can’t read a file that does not exist.
Open() returns zero if the file does not exist when you open it for read
access.
- Another event happens when you read files. Eventually, you read all the
data. Subsequently reading produces error because there is no more data to
read. C++ provides a solution to the end-of-file occurrence.
- If you attempt to read a file that you have completely read the data from,
C++ returns the value zero. To find the end-of-file condition, be sure to
check for zero when reading information from files.
- The following code asks the user for a file name and displays the content
of the file to the screen.

#include<fstream.h>

#include<stdlib.h>
void main()
{
clrscr();
char name[20],filename[15];
ifstream indata;
cout<<"\nEnter the file name : ";

80
 cin.getline(filename,15);
indata.open(filename,ios::in);
if(indata.fail())
{
cerr<<"\nError opening file : "<<filename;
getch();
exit(1);
}
while(!indata.eof())// checks for the end-of-file
{
indata>>name;
cout<<name<<endl;
}
indata.close();
getch();
}

Reading characters from a sequential file

- You can read a characters from a file using get() function. The following
program asks for a file name and displays each character of the file to the
screen. NB. A space among characters is considered as a character and
hence, the exact replica of the file will be shown in the screen.
#include<fstream.h>

#include<stdlib.h>
void main()
{
char c,filename[15];
ifstream indata;
cout<<"\nEnter the file name : ";
cin.getline(filename,15);
indata.open(filename,ios::in);
if(indata.fail())// check id open succeeded
{
cerr<<"\nError opening file : "<<filename;
getch();
exit(1);
}
while(!indata.eof())// check for eof
{
indata.get(c);
cout<<c;
}

81
 indata.close();
getch();
}

7.4.1.4: File Pointer and their Manipulators

- Each file has two pointers one is called input pointer and second is output
pointer. The input pointer is called get pointer and the output pointer is
called put pointer.
- When input and output operation take places, the appropriate pointer is
automatically set according to mode.
- For example when we open a file in reading mode, file pointer is
automatically set to start of file.
- When we open a file in append mode, the file pointer is automatically set
to the end of file.
- In C++ there are some manipulators by which we can control the
movement of the pointer. The available manipulators are:
1. seekg()
2. seekp()
3. tellg()
4. tellp()
1. seekg(): this moves get pointer i.e input pointer to a specified location.
For eg. infile.seekg(5); move the file pointer to the byte number 5
from starting point.
2. seekp(): this move put pointer (output pointer) to a specified location for
example: outfile.seekp(5);
3. tellg(): this gives the current position of get pointer (input pointer)
4. tellp(): this gives the current position of put pointer (output pointer)
eg. ofstream fileout;
fileout.open(“c:\\test.txt”,ios::app);
int length = fileout.tellp();
- By the above statement in length, the total number byte of files are
assigned. Because the file is opened in append mode that means, the file
pointer is the last part of the file.
- Now lets see the seekg() function in action

#include<fstream.h>

#include<stdlib.h>
void main()
{
clrscr();

82
 fstream fileobj;
char ch; //holds A through Z
//open the file in both output and input mode
fileobj.open("c:\\alph.txt",ios::out|ios::in);
if(fileobj.fail())
{
cerr<<"\nError opening alph.txt";
getch();
exit(1);
}
//now write the characters to the file
for(ch = 'A'; ch <= 'Z'; ch++)
{
fileobj<<ch;
}
fileobj.seekg(8L,ios::beg);//skips eight letters, points to I
fileobj>>ch;
cout<<"\nThe 8th character is : "<<ch;
fileobj.seekg(16L,ios::beg);//skips 16 letters, points to Q
fileobj>>ch;
cout<<"\nThe 16th letter is : "<<ch;
fileobj.close();
getch();
}
- To point to the end of a data file, you can use the seekg() function to
position the file pointer at the last byte. This statement positions the file
pointer to the last byte in the file. Fileobj.seekg(0L,ios::end);
- This seekg() function literally reads “move the file pointer 0 bytes from the
end of the file.” The file pointer now points to the end-of-file marker, but
you can seekg() backwards to find other data in the file.
- The following program is supposed to read “c:\alph.txt” file backwards,
printing each character as it skips back in the file.
- Be sure that the seekg() in the program seeks two bytes backwards from
the current position, not from the beginning or the end as the previous
programs. The for loop towards the end of the program needs to perform
a “skip-two-bytes-back”, read-one-byte-forward” method to skip through
the file backwards.

#include<fstream.h>

#include<stdlib.h>

void main()

83
 {
clrscr();
ifstream indata;
int ctr=0;
char inchar;

indata.open("c:\\alph.txt",ios::in);
if(indata.fail())
{
cerr<<"\nError opening alph.txt";
getch();
exit(1);
}
indata.seekg(-1L,ios::end);//points to the last byte in the file
for(ctr=0;ctr<26;ctr++)
{
indata>>inchar;
indata.seekg(-2L,ios::cur);
cout<<inchar;
}
indata.close();
getch();
}

Text Files And Binary Files (Comparison)

- The default access mode for file access is text mode. A text file is an ASCII
file, compatible with most other programming languages and
applications. Programs that read ASCII files can read data you create as
C++ text files.
- If you specify binary access, C++ creates or reads the file in binary format.
Binary data files are “squeezed”- that is, they take less space than text files.
The disadvantage of using binary files is that other programs can’t always
read the data files. Only C++ programs written to access binary files can
read and write them. The advantage of binary files is that you save disk
space because your data files are more compact.
- The binary format is a system-specific file format. In other words, not all
computers can read a binary file created on another computer.

7.4.2: Random Access File Concepts

84
- Random access enables you to read or write any data in your disk file with
out having to read and write every piece of data that precedes it.
- Generally you read and write file records. A record to a file is analogus to
a C++ structure. A record is a collection of one or more data values (called
fields) that you read and write to disk. Generally you store data in the
structures and write structures to disk.
- When you read a record from disk, you generally read that record into a
structure variable and process it with your program.
- Most random access files are fixed-length records. Each record (a row in
the file) takes the same amount of disk space.
- With fixed length records, your computer can better calculate where on
the disk the desired record is located.

0 100 200 300 400 500

Byte
offsets

100 100 ……………… 100 100

Bytes bytes bytes bytes

7.2.2.1 Opening Random-Access Files

- There is really no difference between sequential files and random files in
C++. The difference between the files is not physical, but lies in the
method that you use to access them and update them.
- The ostream member function write outputs a fixed number of bytes,
beginning at a specific location in memory, to the specified stream.
- When the stream is associated with a file, function write writes the data at
the location in specified by the “put” file position pointer.
- The istream member function read inputs a fixed number of bytes from
the specified stream into an area in memory beginning at the specified
address.
- When the stream is associated with a file, function read inputs bytes at the
location in the filespecified by the “get” file poison pointer.
- Syntax of write: fileobject.write((char*) & name ofobject, sizeof(name of object))
- Function write expects data type const char* as its first argument. The
second argument of write is an integer of type size_t specifying the
number of bytes to be written.

85
Writing randomly to a random access file
- Here is an example that shows how to write a record to a random access
file.
#include<fstream.h>

#include<stdlib.h>
struct stud_info{
int id;
char name[20];
char fname[20];
float CGPA;
}student;

void main()
{
clrscr();
char filename[15];
ofstream outdata;
cout<<"\nenter file name : ";
cin>>filename;
outdata.open(filename,ios::out);
if(outdata.fail())
{
cerr<<"\nError opening "<<filename;
getch();
exit(1);
}
//stud_info student;
//accept data here
cout<<"\nEnter student id : ";
cin>>student.id;
cout<<"\nEnter student name : ";
cin>>student.name;
cout<<"\nEnter student father name : ";
cin>>student.fname;
cout<<"\nEnter student CGPA : ";
cin>>student.CGPA;

//now write to the file

outdata.seekp(((student.id)-1) * sizeof(student));
outdata.write((char*) &student, sizeof(student));
outdata.close();
cout<<"\nData has been saved";

86
 getch();
}

- The above code uses the combination of ostream function seekp and write
to store data at exact locations in the file.
- Function seekp sets the put file-position pointer to a specific position in
the file, then the write outputs the data.
- 1 is subtracted from the student id when calculating the byte location of
the record. Thus, for record 1, the file position pointer is set to the byte 0 of
the file.
- The istream function read inputs a specified number of bytes from the
current position in the specified stream into an object.
- The syntax of read : read((char*)&name of object, sizeof(name of object));
- Function read requires a first argument of type char *. The second
argument of write is an integer of type size_t specifying the number of
bytes to be read.
- Here is a code that shows how to read a random access record from a file.
#include<fstream.h>

#include<stdlib.h>
struct stud_info{
int studid;
char name[20];
char fname[20];
float CGPA;
};
void main()
{
clrscr();
ifstream indata;
char filename[15];
cout<<"\nEnter the file name : ";
cin>>filename;
indata.open(filename,ios::in);
if(indata.fail())
{
cerr<<"\nError opening "<<filename;
getch();
exit(1);
}
stud_info student;
cout<<"\nEnter the id no of the student : ";

87
 int sid;
cin>>sid;
indata.seekg((sid-1) * sizeof(student));
indata.read((char*) &student, sizeof(student));
cout<<"\nhere is the information";
cout<<"\nstudent id : "<<student.studid;
cout<<"\nstudent name : "<<student.name;
cout<<"\nstudent fname : "<<student.fname;
cout<<"\nstudent CGPA : "<<student.CGPA;
indata.close();
getch();
}

7.5: Command Line Argument

- Command line argument means facility by which you can supply
arguments to the main() function. These arguments are supplied to the
program when the main function is called from the command line. Eg.
c:\> file-name arg1 arg2
- Where file-name is the name of file(the program) and arg1, arg2 are
arguments passed to the program.
- If we want to pass arguments to the main function, then main function
should be written as follow.
Return type main(int argc, char * argv[])
- The first argument argc represents the number of arguments in a
command line. The second argument argv is an array of character type
pointers that points to the command line arguments. Argc is known as
argument counter and argv is called argument vector.
- C:\> student A B. the value of argc is 3 (student, A & B) and the argv
would be an array of three pointers to string as follows:
Argv[0] – points to student
Argv[1] – points to A
Argv[2] – points to B
- note that argv[0] always reperesents the command name that invokes the
program.
- Here is a sample command line argument code
#include<fstream.h>

#include<stdlib.h>
void main(int argc, char *argv[])
{
clrscr();
char ch;

88
 if(argc < 3)
{
cerr<<"\ntoo few parameters";
getch();
exit(1);
}
else if(argc > 3)
{
cerr<<"\ntoo many parametes";
getch();
exit(1);
}
else//number of parameters okay
{
ofstream outdata;
ifstream indata;
outdata.open(argv[2],ios::out);
if(outdata.fail())
{
cerr<<"\nlow disk space to create file : "<<argv[2];
getch();
exit(1);
}
indata.open(argv[1],ios::in);
if(indata.fail())
{
cerr<<"\nfile : "<<argv[1]<<" does not exist";
getch();
exit(1);
}
//now start copying the file
while(!indata.eof())
{
indata.get(ch);
outdata.put(ch);
}
indata.close();
outdata.close();
cout<<"\nfinished copying";
}
}

89