File handling in Linux

April 04, 2021

Introducing file handling

The exec command is very interesting. Whenever we run any command in a shell, a new subshell or process gets created, and the command runs in this newly created process. When we run any command as an argument to the exec command, exec will replace the current shell with the command to be executed. It does not create or spawn a new process to run the command.

Using exec to assign a file descriptor (fd) to file

In the Bash shell environment, every process has three files opened by default. These are standard input, display, and error. The file descriptors associated with them are 0, 1, and 2 respectively. In the Bash shell, we can assign the file descriptor to any input or output file. These are called file descriptors.

The syntax for declaring output.txt as output is as follows:

exec fd > output.txt

This command will declare the number fd as an output file descriptor.

The syntax for closing the file is as follows:

exec fd<&-

To close fd, which is 5, enter the following:

exec 5<&-

We will try to understand these concepts by writing scripts.

Understanding the opening, writing, and closing of a file

Let’s understand the opening, closing, and writing of a file.

Write a Shell script, shown as follows:

# We will open file for writing purpose 
# We are assigning descriptor number 3 for file sample_out.txt 
exec 3> sample_out.txt 
# We are sending output of command "echo" to sample_out.txt file 
echo "This is a test message for sample_out.txt file" >&3 
# Run command date & store output in file sample_out.txt 
date >&3 
# Closing file with file descriptor 3 
exec 3<&-

Save the file, give the permission to execute, and run the script as follows:

$ chmod u+x
$ ./
$ cat sample_out.txt

This should produce the following output:


This is a test message for sample_out.txt fileTue Sep 29 23:19:22 IST 2015

Understanding reading from a file

Let’s write a script to read from a file:

Write the script, shown as follows:

# We will open file sample_input.txt for reading purpose. 
# We are assigning descriptor 3 to the file. 
exec 3< sample_input.txt 
cat <&3 
# Closing file 
exec 3<&-

Save the file, give the permission to execute, and run the script as follows:

$ chmod u+x

We will create the sample_input.txt file as follows:

$ echo "Hello to All" > sample_input.txt

Run the script and check the result:

$ ./

This should produce the following output:


Hello to All

Understanding reading and writing to a file

In the earlier examples, we opened the file either for reading or writing. Now we will see how to open the file for reading and writing purposes:

exec fd<> fileName

If the file descriptor number is not specified, then 0 will be used in its place. The file will be created if it does not exist. This procedure is useful for updating files.

Let’s understand the following script.

Write the shell script as follows:

# We are assing fd number 3 to file. 
# We will be doing read and write operations on file 
exec 3<> $file_name 
# Writing to file 
echo """ 
 Do not dwell in the past, 
 do not dream of the future, 
 concentrate the mind on the present moment. - Buddha 
""" >&3 
# closing file with fd number 3 
exec 3>&-

Using the read command on a file descriptor (fd)

We can use the read command to get data from a file to store it in variables. The procedure for using the read command to get a text from a file is as follows:

read -u fd variable1 variable2 ... variableN

Reading from one file and writing to another file

Now we will see how to read from one file and write to another. Let’s write the script as follows:

# We are assigning descriptor 3 to in_file.txt 
exec 3< in_file.txt 
# We are assigning descriptor 4 to out_file.txt 
exec 4> out_file.txt 
# We are reading first line of input.txt 
read -u 3 line 
echo $line 
echo "Writing content of in_file.txt to out_file.txt" 
echo  "Line 1 - $line " >&4 
# Closing both the files 
exec 3<&- 
exec 4<&-

Save the file, give the permission to execute, and run the script as follows:

$ chmod u+x
$ echo "Sun is at the centre of Solar System." > in_file.txt
$ cat in_file.txt

It will gives you output similar to below:


Sun is at the centre of Solar System.
$ ./
Sun is at the centre of Solar System.Writing content of in_file.txt to out_file.txt
$ cat out_file.txt
Line 1 - Sun is at the center of Solar System.

In this example, we read the complete line in the variable line and we use the same variable to write it to another file.

Let’s write one more script,, to get the hostname and addresses:

cd /etc/hosts hosts2 
grep -v '^#' hosts2 > hosts3 
exec 3< hosts3    # opening hosts3 as input file 
exec 4> hostsfinal  # opening hostsfinal as output file 
read <& 3 address1 name_1 extra_info 
read <& 3 address2 name_2 extra_info 
echo $name_1 $address1 >& 4 
echo $name_2 $address2 >& 4 
exec 3<&-      # Closing hosts3 
exec 4<&-      # Closing hostsfinal

In this script, we used the variables address1name_1extra_infoaddress2, and name_2 to store useful information.

Displaying the file descriptor information from the /proc folder

We will write the script to display the actual file descriptors associated with the file.

Let’s write the script, shown as follows:

# we are assigning file descriptor 3 to input file test.txt 
exec 3< test.txt 
# we are assigning file descriptor 4 to output.txt 
exec 4> output.txt 
# we are using read command to read line from file 
read -u 3 line 
echo "Process id of current process is $$" 
echo "Currently following files are opened by $0 script :" 
ls -l /proc/$my_pid/fd 
# We are closing both files test.txt and output.txt 
exec 3<&- 
exec 4>&-

File handling – reading line by line

You will learn how to use the while loop and the read command to read a file line by line. You will learn more about the while loop in the upcoming chapters.

Let’s write the script, as follows:

echo "Enter the name of file for reading" 
read file_name 
while read var_line 
  echo $var_line 

For executing the preceding script, we will need to create a file with some text in it. Then, we can pass this filename for reading purposes.

Executing the command and storing the results in a file

The following is the syntax for storing the output of a command in a file:

Command >& fd./script >& fd

The following is the illustrative example script,

exec 4> output.txt 
cat /proc/cpuinfo  >&4 
exec 3<&-

Save the file, give the permission to execute, and run the script as follows:

$ chmod u+x
$ ./

Here’s the output:

In this example, we have executed the command cat /proc/cpuinfo and we have stored the output in the file, output.txt.

Summarizing usage of the exec command

The following is a summary of the exec command for using various file handling-related operations:


What it does

exec command

This command will replace shell and execute it. Therefore, it will not return to its original shell, which started it.

exec > data.txt

This opens data.txt for writing standard output.

exec < data.txt

This opens data.txt for reading standard input.

exec 3< data.txt

This opens data.txt for reading with descriptor 3.

sort <&3

This will sort the data.txt file.

exec 4> data.txt

This opens data.txt for writing with descriptor 4.

ll >&4

The output of ll is redirected to data.txt.

exec 6<&5

This makes fd 6 a copy of fd 5.

exec 4<&-

This closes fd 4.

Related Articles

Calculating and reducing the runtime of a script

In this article, we are going to learn how to calculate and reduce the script’s runtime. A simple time command will help in calculating the execution time.PrerequisitesBesides having a terminal open, make sure you have the necessary scripts present in your...

read more

Lorem ipsum dolor sit amet consectetur


Submit a Comment

Your email address will not be published. Required fields are marked *

11 + eleven =