Advent of Code 2024 (Go): Day 1 + Day 2

(side note: I'm done learning Swift for now. I'll come back to it in the new year, perhaps. I was enjoying my C++ adventures in plugin programming [Juce] more anywho. Also like I have skiing to do?!)

I'm doing Advent of Code! The goal is to do it all this year. A great chance to really brush up my Go - Exercism has been nice, but a little easy. The more the merrier though!

My GitHub repo is here: https://github.com/jmontroy90/aoc-2024-golang

I'll be publishing notes under each day as a README, and that's what I'll be reposting here. The notes help me organize my thoughts and perhaps give other people some visibility into how I approached the solution. Today, two days:

Day 1: Easy!

Overall, an easy start. I spent more time setting up the repo structure and the file loader than anything else. The problem itself was really straight-forward - sort and compare. The second part could be done with sorted lists, but making a map and looking up its counts reads cleaner to me.

Day 2: Problems should be composed.

Well, that's more like it! Part 1 was pretty easy, but part 2 messed with me. Here are some notes:

• I tried to re-program part 2 from scratch by just taking out the element at the index it occurred. But this doesn't work, because you need to try taking out each of the three relevant elements - any one of their removals might pop things into working condition.

• Good, clean code mean finding nice boundaries; here, the part 1 boundary clearly was testing each report. But I modified this boundary to return the trouble index, which is required for part 2 to be clean and useful itself. I tried to think of a way to compose the two requirements (just returning "is report safe" bool + returning where the problem was), but it wasn't obvious. The cheap-y way to do it is to just create a SafeReportResult struct and return all relevant data that way - pay me big bucks please!

• Once I started using IsReportSafe directly, Part 2's problem become what it should be - problem index removal. Compose your problems!!

• This whole thing was also a nice exercise in recalling slice behavior. I totally botched my first attempting at using append because appending to a slice modifies its backing array at the slice point indicated. Total nonsense result. Using slices.Concat was nicer, but does too much allocation. Using copy - a rarity for me! - on a preallocated slice was nice to just pop the elements into where they're needed.

• I looked at other AOC2024 Go solutions to figure out whether I was undercounting or overcounting once my first attempt was wrong. To be precise, I ran my input through their programs and then figured out which reports we disagreed on. This was really useful to pop into my table tests and start to realize how I'd misconceived the solution.

  
  

Overall, this took me a little while longer than expected, but that's fair - we're getting back into it!