Today's was rough in terms of performance and bugs.
I had a system that built an entire map and could print and calculate the test data fine (obviously), but melted on the real data. I did ultimately take that approach and narrow it down to ONLY the target row and it allowed me to solve Part 1.
Part 2 was a whole new process of sliced scan ranges.
https://github.com/ajohns6/AdventOfCode2022/blob/master/AdventOfCode2022/Day15.cs