Julia solutions to Advent of Code problems. These are not designed for any particular purpose such as speed, short code, pedagogical value, etc. but are simply whatever solution first occured to me. You may learn something from reading them or you may not. There are no code comments, consider it as part of the challenge.
Ok, who can resist doing a bit of optimization? For some problems
there are alternative versions available, with e.g. a fast suffix in
the filename.
julia aoc.jl YEAR DAY [TEST] [EXTRA]
Examples:
julia aoc.jl 2020 1
Run day 1 of 2020 using the code in 2020/day1.jl and problem input
in 2020/input/day1. See next section about input files.
julia aoc.jl 2020 10 test2
Run day 10 of 2020 using the test input 2020/input/day10test2.
julia aoc.jl 2020 9 test 5
Run day 9 of 2020 using the test input 2020/input/day9test and
passing the value 5 as an extra positional argument to the solution
code.
Input files are named as indicated in the previous section. You need
to enter test files yourself but the real problem inputs will be
downloaded automatically if you extract your session cookie from your
web browser and place it in a file called session_cookie at the top
of your copy of the repository.
Add a --benchmark argument when calling aoc.jl to get the solution
benchmarked. This measurement is done with BenchmarkTools to exclude
compilation time and the input is provided as an IOBuffer,
i.e. eliminating disk read time.
The two parts are benchmarked separately. When comparing with solutions designed to compute both parts in one call, the sum of the two parts will include duplicate input parsing and in some cases additional repeated calculations.
With a --benchmark=SUFFIX argument the code will be loaded from an
alternative file. E.g.
julia aoc.jl 2021 3 --benchmark=fast
will load the code from 2021/day3fast.jl.