notes.billmill.org / computer_usage / nushell /

nushell

last updated: Oct 20, 2023

https://github.com/nushell/nushell
https://www.nushell.sh/book/quick_tour.html

Got inspired to try this again today because of how annoying it was to head parquet files in bash, while it's built-in to nushell.

commands

ls | sort-by size to... well I think you can guess from reading it

$ echo "https://www.nushell.sh/book/quick_tour.html" | url host
www.nushell.sh

Handy! I've often complained about how difficult this is in bash-land

I'm going to trip up on stuff like this:

For example, in bash, you might use:

> echo "hello" > output.txt

In Nushell, we use the `>` as the greater-than operator. This fits better with the language aspect of Nushell. Instead, you pipe to a command that has the job of saving content:

> echo "hello" | save output.txt

The way Nushell views data is that data flows through the pipeline until it reaches the user or is handled by a final command.

ls **/*.png | get 0.name: list all the PNG files in this directory and its subdirectories, and print the name of the first one

ls **/*.png | get 0.name | each { |f| imgcat $f } I feel like there must be a smoother way to do this, but this takes the result of the previous command and cats it to the terminal window with imgcat

ah ha! the $in variable represents the input to the command, so we can simplify that to:

ls **/*.png | get 0.name | imgcat $in

open foo.db | query db "select * from some_table"

nu natively supports querying sqlite! Neat.

↑ up