I was kinda lucky with my input, though. Getting ready for day 4, I'm still a bit behind.

In my solution, I first extracted what I considered “enabled” commands, so, anything between a “do” and “don't”. I then retrieved the last possible command, "don't" until the end of the input and retrieved the enabled chunks.

I'm finally done with part 2!

#aoc

Shout out to Luke Smith's tutorial, I wonder where he is now

www.youtube.com/watch?v=77I4...

Code screenshots look kinda bat, maybe a GitHub gist would be better? Repo coming soon I guess.

Here are my tests. I was doing most of my coding in nvim, but I had to switch to Rider in order to nicely debug each one.

So glad that Rider is now free for noncommercial use, thank you, @jetbrains.com !

Part 2 is done. I'm actually happy with this one. The test input outputted the correct result, so I had to write as many tests as possible to figure out the edge cases.

#aoc

Ok, part 1 was so bad, I couldn't even think about working on part 2.

Here is revised part 1.

I'll gave to learn to comfigure NGINX, I wanted a more hand-off approach with frankenphp. I'll take my time with NGINX

Thankfully, sorting wouldn't have caused trouble in part 2, but I should be careful with mutating the original array.

I know there's mini.files, but give oil.nvim a try.
Do let us know how it went, I'm also interested on mini.

I wish single-spa would migrate to Vite 😢...