On to the subject of naming. You might have noticed something about the variables we've seen up to now: the Bash variables PATH and BASH_VERSION are written fully uppercase, but in our examples we used lowercase, with words separated by an underscore (hello_text). Consider the following example:
#!/bin/bash
#####################################
# Author: Sebastiaan Tammer
# Version: v1.0.0
# Date: 2018-09-08
# Description: Showing off different styles of variable naming.
# Usage: ./variable-naming.sh
#####################################
# Assign the variables.
name="Sebastiaan"
home_type="house"
LOCATION="Utrecht"
_partner_name="Sanne"
animalTypes="gecko and hamster"
# Print the story.
echo "${name} lives in a ${home_type} in ${LOCATION}, together with ${_partner_name} and their two pets: a ${animalTypes}."
If we run this, we get a nice little story:
reader@ubuntu:~/scripts/chapter_08$ bash variable-naming.sh
Sebastiaan lives in a house in Utrecht, together with Sanne and their two pets: a gecko and hamster.
So, our variables are working great! Technically, everything we did in this example was fine. However, they look a mess. We used four different naming conventions: lowercase_with_underscores, UPPERCASE, _lowercase, and finally camelCase. While these are technically valid, remember that readability counts: it's best to pick one way of naming your variables, and stick with this.
As you might expect, there are many opinions about this (probably as many as in the tabs versus spaces debate!). Obviously, we also have an opinion, which we would like to share: use lowercase_separated_by_underscores for regular variables, and UPPERCASE for constants. From now on, you'll see this practice in all further scripts.
The previous example would look like this:
reader@ubuntu:~/scripts/chapter_08$ cp variable-naming.sh variable-naming-proper.sh
reader@ubuntu:~/scripts/chapter_08$ vim variable-naming-proper.sh
vim variable-naming-proper.sh
reader@ubuntu:~/scripts/chapter_08$ cat variable-naming-proper.sh
#!/bin/bash
#####################################
# Author: Sebastiaan Tammer
# Version: v1.0.0
# Date: 2018-09-08
# Description: Showing off uniform variable name styling.
# Usage: ./variable-naming-proper.sh
#####################################
NAME="Sebastiaan"
HOME_TYPE="house"
LOCATION="Utrecht"
PARTNER_NAME="Sanne"
ANIMAL_TYPES="gecko and hamster"
# Print the story.
echo "${NAME} lives in a ${HOME_TYPE} in ${LOCATION}, together with ${PARTNER_NAME} and their two pets: a ${ANIMAL_TYPES}."
We hope you agree this looks much better. Later in this chapter, when we introduce user input, we will be working with normal variables as well, as opposed to the constants we've been using so far.
To keep things clean, we generally avoid using UPPERCASE variables, except for constants. The main reason for this is that (almost) all environment variables in Bash are written in uppercase. If you do use uppercase variables in your scripts, there is one important thing to keep in mind: make sure the names you choose do not conflict with pre-existing Bash variables. These include PATH, USER, LANG, SHELL, HOME, and so on. Should you use the same names in your script, you might get some unexpected behavior.
It is a much better idea to avoid these conflicts and choose unique names for your variables. You could, for example, choose the SCRIPT_PATH variable instead of PATH.