Exit code

Every command you run on the command line of Linux has an exit code. 0 indicates success. Any other number is a failure. The variable $? contains the exit-code of the previous command.

E.g. a successful command:

$ ls
...  (the content of the directory)

$ echo $?
0

A failed command:

$ ls some-place-that-does-not-exist
ls: cannot access 'some-place-that': No such file or directory

$ echo $?
2

Shell script without -e

In this script we have 3 commands. The 2nd fails, but it does not bother the main script. It keeps working and executes the 3rd command.

examples/shell/no-set-e.sh

#!/bin/bash

echo "Start"

ls some-incorrect-path

echo "Still working"

$ ./examples/shell/no-set-e.sh
Start
ls: cannot access 'some-incorrect-path': No such file or directory
Still working


$ echo $?
0

Not only that, but the whole script indicates success (its own exit code is set to be 0) even thought some part of it have failed.

Shell script with -e

We can set the -e either on the sh-bang line or with the set -e command. We can turn it off using the set +e command.

examples/shell/set-e.sh

#!/bin/bash -e

echo "Start"

ls some-incorrect-path

echo "Still working"

$ ./examples/shell/set-e.sh
Start
ls: cannot access 'some-incorrect-path': No such file or directory

$ echo $?
2

Here, as you can see, after the 2nd command fails, the whole script stops and the exit-code of the script is set to a non-zero number indicating failure.