Notes:
A good examples of .filter_map() / .unzip() to parse a stream into two vectors for subsequent independent sorting.
Part2: big_gaps_undamped() shows a good used of Itertools::tuple_window()
I was stuck until experimented with many_till.
fn parse_instr(input: &str) -> IResult<&str, (u32, u32)> {
let (remain, (_junk, instruction)) = many_till(anychar, parse_mul)(input)?;
Ok((remain, instruction))
}Part 1: A beautiful example of a handcrafted Generator ( Well in RUST I implemented this as a iterator! )
Part 2: Operators Add, Mul, Merge
A good use of Itertools::multi_cartesian_product() Future Refactor: A run time of 2sec implies that I need to investigate performance. All the main loop look like they could be simplified by using fold and in particular reduce(). Also my nom parser could be rewritten.
Makes use of Iterrools::cartesian_product. The correct data structure is :-
let map: HashMap<char, HashSet<(usize, usize)>>Want to return to this problem and visualize the walk.
part 1 just copied the list. part 2 is a walker over the linked list with insertions behind. This puzzle is designed to exceed all possible memory requirement.
The answer is to cache.
Need to rework my failing solution using petgraphs and "condensation"
matrix inversion using cramer rule.
part2 was a tricky test of number conversion.
- u64 has to used to parse large numbers.
- Converted in a f64 for processing
- Converted back into u64 for summation.
A good use of rem_euclid() for modulo arithmetic.
part2 is a really tricky one.
I outputted the map frame by frame and it look like snow I reasoned that the easter egg must have structure unlike the snow so I stopped when there was a line of 10 robots in a row (I tried 4, 6 first but I new as I was on the right track as 4 and 6 were rare.)
A interesting exercise.
htps://en.wikipedia.org/wiki/Clique_problem
A enjoyable puzzle.