Using Shell Functions Casually
Most people know they can add functions to the shell, but often limit themselves to their dotfiles.
Nothing stops you from adding functions to your running session.
Functions in interactive shells
If you’re already in a shell, you can define a function in exactly the same way you would in a startup script, by typing it at the prompt:
# this is bash
> ll() {
ls -l
}
> ll
# runs `ls -l`
# this is fish
> function ll
ls -l
end
> ll
# runs `ls -l`
Why would you do that?
Usually, I use this to handle positional arguments more easily
# this is bash
> g() {
grep -ril "$@" .
}
> g "something"
# grep -ril "something" .
grep invocations look like grep $WHAT $WHERE. Here I’m hardcoding $WHERE (.) and making $WHAT the last argument. While I’m at it, I’m also adding some flags (-ril).
> g "this"
# ...
> g "that"
# ...
> g "the other thing"
# ...
Not having to navigate-and-edit to the second-to-last position every time makes using the command-line delightful.
(although for this very specific case, a better solution would be ripgrep 😄)
When to use this?
If you keep running variations of the same command.
If you keep editing previous commands.
If you have a “one-shot” batch job with a-lot-of-typing but not worth a script.
Tips and Tricks
Remember that your function will disappear when you exit the shell. Consider this a feature rather than a bug: your temporary function garbage collecting itself 🚀
Since this function is going away, don’t feel like you need to give it a great name; the easier to type the better.
Reference: how to define functions and handle arguments
- Passing parameters to a Bash function - arguments as
$1,$2… - Fish : create a function - arguments in
$argvarray
Pro tip: for a better command-line experience, you can edit your command-line in $EDITOR (which is a killer feature, if you didn’t know…)