advent-of-code VS advent-of-code-swift

My solution in Common Lisp.

Dot file generated by my code and then run through fdp from Graphviz.

Just a simple relief-shaded image like you might find on a map. Principally generated with GDAL. The particular commands used are at the bottom of my solution code for the day. I also made a more flat version, which shows the elevation colors better, but really doesn't feel right in terms of elevation.

The code is written in Common Lisp using Cairo and ffmpeg. https://gitlab.com/asciiphil/advent-of-code/-/blob/master/2022/09.lisp

So I spend more time analyzing the input calculation. I wrote up that analysis and committed it to my repository, too.

This was a bit difficult. My original part one solution worked completely differently to what I had to do for part two.

My solution in Common Lisp, 2335/2041.

No. It's just here, as part of my Advent of Code repository.

The code is at the bottom of my day 9 source file, in Common Lisp. (The AdvMAME3x code was something I originally wrote for my 2020 day 20 visualization.)

As a side note, I took inspiration from a post the other day about adding AoC badges to your repository README file. I use GitLab, which lets you define badges as properties of your repository. So I added badges to my repository's header area, driven by a JSON file in the repository. The JSON file is updated manually by a script. (I might add some automated updates at some point, but manual works okay for now.)

Nice! I struggled with the performance of my Swift solution. I thought the problem was that my huge nextCup array was getting copied every time I updated it due to Swift's copy-on-write. I worked around it by making it an array of integer references which can mutate without triggering the array's copy-on-write process. This helped a lot, but it still takes >30s to complete. 😬