scrapyard VS aoc-2021

Inspired by a blog post of Nicky Meuleman and his solution for this puzzle: https://github.com/NickyMeuleman/scrapyard/blob/main/advent_of_code/2022/src/day_08.rs I refactored my code and it turned out to be ~7-10 times slower than the initial one (700μs vs. 8ms). But hey, the code looks better now and it's more concise.

If you prefer to go straight to the repo: https://github.com/NickyMeuleman/scrapyard/tree/main/advent_of_code/2022/src

rust Per step, this makes a list of all cucumbers that can move first, then mutates those in the overall map.

Solution in Rust

Rust solution, commented both solutions to clearly explain the used logic.

Rust

Rust. The windows() method is really the MVP of the day

Basic C# solution with a straightforward try-everything simulation. This one could've been called 50 Thousand Iterations Under the Sea

C# solution

C# solution, I keep letter pair counts in a square grid and a separate map for individual letter counts

C# solution with basic graph search. I track the step counts to each cave and decide which next caves to take based on them, and prune it after returning from a recursive call to SearchFrom

C# solution with a straightforward stack-based approach. I modeled sum types using abstract record on line 62 to handle the three different return values from the processing function

C# solution, simple flood-fill type algo. I found a use for C# 9.0's record types, explained in the blurb at the top of the file

C# solution with a manual deduction process (one of many possible no doubt) that I figured out in the hours after midnight, then I slept on it and it took all morning to actually code the thing, but I like the result. Again a sprinkling of LINQ makes it bearable to write and read

C# code tries all positions between first and last crab

C# solution same as most others but with a bit of LINQ

C# nothing noteworthy other than LINQ saving the day again