This year I completed AoC using Emacs Lisp. I would say that this year was a more legitimate completion than last, since I used less hints and reference solutions. Here is a link to my solutions. Sorry that they are uncommented and probably incomprehensible. Also a link to my AoC helper package which allowed me to read problems, fetch inputs and submit answers, all within Emacs.

Using Elisp was much more comfortable for me than using Haskell and OCaml (my previously used languages for AoC). It was a real pleasure to have an instant feedback loop using Emacs itself as a REPL, instead of Haskell and OCaml's clunky REPL experience. The REPL is important for these types of problems if you have a small brain like me and can't use your imagination to render data manipulation pipelines.

The libraries I used were fantastic; dash.el, s.el and ht.el should be built-in to Elisp. I do regret not exploring seq.el (built-in) enough, because it does seem to be very useful since it works generically over any type of sequence (lists, vectors, etc).

My only issues were that Emacs completely locks up when running a long computation. A few problems I left running for minutes, and one overnight. Maybe this is actually charming, since I woke up to a correct answer as if it was Christmas morning. I also was a little dissatisfied with iteration in some cases. I spent a lot of time bikeshedding how I would do a simple loop - cl-loop, named-let, dotimes, while, or just a recursive function.

Even though I am still not a Lisp expert, I reached my goal of being able to whip up a script fairly quickly, and now will use Elisp as my "scratch" programming language.

I really recommend Emacs users to try AoC in Elisp for the cozy experience and low barrier for entry - just open up a buffer and start writing then C-x C-e.