The goal was to be able to have a function to randomly pick a movie for me to watch.
I made a text file with one movie title per line that included random flicks top-of-my-head, adding 10. Then, I went and exported some imdb lists, which was so handy because csv files. So, export one list, open it in calc; highlight, and then copy the title column. Simple. I pasted it into my movies list file.
I did this maybe 7 times which gave me a good sized list of movies - then, I needed to remove duplicates, so, I saved the file and ran this on it:
cat films.txt | sort | uniq > films2.txt
I opened the new file and double entries were stacked so I could go through and delete redundant rows, of which, there weren't very many. Now, I have a file with 1892 lines, each line being a movie title. Handy.
Of course, now, I need to randomly be able to choose a title (line) from the file because I can't be bothered to open the thing and read through it every time I need some inspiration, or, guidance. That's where shuf comes in - look:
shuf -n 1 films.txt
Dayum! And the final step of my goal is to make this available, so, bash alias and keybind.
I added the following to my ~/.bash_aliases file:
alias pick='shuf -n 1 films.txt'
I made a bash script with the following command in it and then bound the script to some keys in rc.xml:
#!/bin/bash
bash -c 'notify-send -i /usr/share/icons/gnome/256x256/emotes/face-glasses.png -t 5000 "Watch this one" "$(shuf -n 1 films.txt)" '
the keybind:
<keybind key="C-A-p">
<action name="Execute">
<command>sh ~/bin/pick.sh</command>
</action>
</keybind>
Awesome linux for every day stuff.

Nice. What would you do better the next time, or how could you improve it?
What about using your jellyfin db for the entries and a simple connection so that it randomly chooses one or multiple videos every 3 hours?
Nothing I could think of readily without some sort of infrastructure support. If you look 'lazy' up in a dictionary you will see my smiling face.