Day 1 was the typical check that you can actually program. Both parts of the problem were scans to see how fast the depth increased below the submarine.

For the first part, I used a simple loop, starting from 1 and checking the current and previous positions to see if there had been an increase.

``proc part1(input: string): int =  let lines = splitLines(input)  let xs = map(lines, parseInt)  for i in 1 .. xs.len - 1:    let x = xs[i]    let y = xs[i - 1]    if x > y: result += 1``

I tend to implement my solutions as two separate functions which both take the input string (rather than a parsed version of it), even when means doing more work in each part. This makes them easier to test against the example inputs, which I tend to do with inline tests or assertions in every puzzle.

``assert part1("199\n200\n208\n210\n200\n207\n240\n269\n260\n263") == 7assert part2("199\n200\n208\n210\n200\n207\n240\n269\n260\n263") == 5const input = slurp("input.txt")echo "Part 1: ", part1(input)echo "Part 2: ", part2(input)``

The `assert` template is a great first look at Nim's metaprogramming capabilities. It looks like a keyword, but it's actually just a template. If an assertion fails, the template has full access to the syntax tree, and can print the source for the expression that failed, rather than just a generic error message.

Assertions can also be compiled out of the resulting executable with the `--assertions:off` flag to optimise your programs.

Part 2 didn't throw us any curveballs. This time we had to find whether the sum of the current window of three had increased from the previous window.

``proc part2(input: string): int =  let lines = splitLines(input)  let xs = map(lines, parseInt)  for i in 1 .. xs.len - 3:    let x = xs[i + 0] + xs[i + 1] + xs[i + 2]    let y = xs[i - 1] + xs[i + 0] + xs[i + 1]    if x > y: result += 1``

I didn't try to do anything clever here (although I vaguely remember learning about efficient ways to do sliding windows in last year's Day 9). It seemed easiest to just repurpose the code from part one.

Feels good to be learning!

Day 1