How to use exit code to read from terminal, from script and with logical operators

Scripting is an important part of automation in Linux and it is one of the key tools every system administrators use to manage day-to-day activities like taking backups, adding/removing users/groups, updating/removing packages, etc. Since bash is the default shell in most of the Linux distributions, bash scripting is very efficient as it can simplify certain operations that are hard to accomplish with GUI.

A bash script is a plain text file that includes single command or series of commands that are read and executed by the bash program. The most crucial part while writing a bash script is to handle errors effectively. When not handled properly, it may leave less of an adverse impact on the system. Exit code is the traditional method of detecting errors in bash script. Thus in this article, we shall discuss various exit codes, and how to read and handle errors using them.

Bash Exit Code

Exit code is returned by bash for every command executed in the terminal. This code can be used to display error messages when there is any problem while running the script. When a command is executed successfully, bash returns 0 as the exit code, and for any error while executing the command non-zero value is returned which varies from 1 to 255 based on the error type. 

Exit code	Description	Examples
0	Successful execution
1	Catch generic errors	“Divide by zero”, “missing operand”, “Permission denied”
2	Improper command usage	“Missing keyword”, “No such file or directory”
	Unable to execute the command	Permission problem
127	The issue in path resolution	Command not found
130	Fatal error	When interrupted by Ctrl+C
255	Out of range	When the exit code exceeds 255

Note : The exit code only informs that the script or command is executed fine, but it doesn’t tell whether the output from the executed command has got any value or not. For example, listing files in the home directory (doesn’t have any files other than hidden files), so simply giving ls will not give any output, but $? gives exit code as 0 which means ls command got executed successfully. 

pwd
ls
echo $?

Read the Exit Code from terminal

Bash uses a special variable ‘$?‘ which holds the exit code of the last command executed either through the terminal or via script. Let us try to execute some commands in the terminal and check their exit code.

Example 1: Exit Code 0

Exit code 0 is returned for commands which ran successfully. For example on running ‘pwd‘ command, $? returns 0. 

$ pwd

Example 2: Exit Code 1

mkdir
echo $?

In Linux, certain commands need to be invoked with options. $? returns 1 when mkdir command is invoked without any options as it results in a missing operand error. 

echo a=10
echo b=0
echo $((a/b))
echp $?

$? returns 1 when any number is divided by zero as it gives the answer “equal to infinity.” No variable can store an infinite amount of data in any programming language. Hence, if any number is divided by zero, we get the arithmetic exception. 

cat /etc/sudoers
echo $?

The sudoers file is primarily for system administrators to manage and by default, only the root user has the privilege to edit this file. When any other user (not in the sudoers list) tries to update this file, it would throw a Permission denied error and $? returns 1.

Example 3: Exit Code 2

ls /root/test.sh
echo $?

No such file or directory error is thrown when any user tries to access the file/directory which is not present in the server. Here /root/test.sh file doesn’t exist, so $? returns 2.

cat /tmp/sleep.t
echo $?

This error occurs when any keyword is missed or misspelled in the script file. In the below example, the done keyword is missed for the while loop, so $? returns 2.

In the below example, the keyword then is misspelled as tthen.

cat /tmp/if.t
echo $?

Example 4: Exit Code 126

ls -lrth /tmp/sleep.t
/tmp/sleep.t
echo $?

Exit code 126 is returned when there is a problem related to permissions. For example, create a script file sleep.t without executing permission set to it. Thus while running the script it would throw permission denied error resulting in a 126 error code.

Example 5: Exit Code 127

hello
echo $?

This error code is returned when the executed command doesn’t exist, or when the command isn’t in the $PATH environment variable. 

This code also appears while attempting to execute a command that is in a different directory. In the below example, the ls command is present in /usr/bin directory. If we run the command from a different directory /usr/sbin, the 127 code is returned.

which ls
/usr/sbin/ls
echo $?

Example 6: Exit Code 130

cat sleep.sh
sh.sleep.sh
echo $?

This error code is returned when any of the processes or command execution is terminated by pressing Ctrl + C. In the below example, while loop runs infinitely and is terminated by Ctrl + C. 

Example 7: Exit Code 255

ip
echo $?

This is a reserved exit status that is easy to reproduce but difficult to interpret. The documentation states that bash receives exit status 255 if an exit code that’s out of the range 0-255 is used. However, it can also be reproduced easily in other ways by running certain commands like ip without any options.

If the code exceeds 256, the exit status returned by bash is the current exit code subtracted by 256. In this example when ‘exit 300’ is given, the returned exit code is 300-256 = 44.

exit 300
echo $?

Read the Exit Code from a script

Exit code can also be read from the bash script using $?. It can either be assigned to a variable or displayed directly in the echo command. Let’s write a sample script list.sh which would list two files and display the exit code based on the file presence.

Script:

#!/usr/bin/sh
ls -a /root/list.sh
echo “The exit code is $?”
ls -a /root/test.sh
echo “The exit code is $?”

Command:

sh list.sh

In this example, /root/list.sh file is present hence 0 exit code is returned. But /root/test.sh file is not present, so 2 exit code is returned.

Use the Exit code to do a task

Let’s create a script named user.sh which will take a username as input. The script will check if the entered user is valid or invalid based on the exit code returned by “cat /etc/passwd | grep $username”. For valid users, exit code 0 is returned and for invalid users, exit code 1 is returned.

Script:

#!/usr/bin/sh
echo “Enter a username”
read username
entry=$(cat /etc/passwd | grep $username)
user_exist=$?
if [ $user_exist -eq 0 ]
then
echo “Valid user”
su $username
else
echo “Invalid user”
fi

Command:

sh user.sh

1. Valid user

For a valid user named test_user, the terminal prompt got changed from root to test_user.

2. Invalid User

If the user is not present, then an “Invalid User” error message is thrown.

Using logical operators with exit code

The ‘&&’ logical operator is used for a successful exit code and ‘||’ logical operator is used for an unsuccessful exit code. Let’s test valid and invalid user scenarios using logical operators. The following command allows test_user to log in as it is a valid user.

cat /etc/passwd | grep test_user && su test_user || echo “invalid user”

The following command prints “invalid user” as the jack is not present in the server.

cat /etc/passwd | grep jack && su jack || echo “invalid user”

Conclusion:

Thus we have learned about various exit codes, how to read them from a terminal or script and how to use them with logical operators.