My own personal overengineered helpers to solve AoC problems in Rust!
The project is organized in one crate per year under the years directory.
Each crate has an inputs folder, which is where the tool will look for inputs.
Of course each year can be run independently. Each solution is organized as its
own binary under the src/bin folder, and uses the #[entry] macro (custom to
this project) to define the entrypoint:
A simple way to run a single solution would be to do this:
cargo run -p y2022 --bin d05
If you want to bench a single solution, add --bench:
cargo run -p y2022 --bin d05 -- --bench
The following are the available arguments:
--bench- Run the solution as a benchmark.--verbose- Verbose output.--warmup- Warmup period for benchmark in milliseconds (default400).--time-limit- Time to run the benchmark in milliseconds (default100).--iter- How many iterations to run for a single sample, this is determined by timing the solution once so that we don't try to take timings in the nanosecond realm which would be unreliable.--json- Output JSON which is used by the "run everything" tool below to collect and aggregate output. You can use it yourself if you find it interesting.
The default tool in this project is provided by the lib crate, and allows for
running every project at once (automatically discovered). It works with your
default cargo run:
$ cargo run -- <args> -- <solution args>
<args> is one of the following:
-q | --quiet- less verbose output. *-V | --verbose- more verbose output.-p <project>- only run the specified sub-project, likey2022for 2022 solutions only.--release- run in release mode.--no-prod- disable the "production mode", which removes a bunch of stuff that is solely used to improve diagnostics during development.
How to run every solution in this repo:
cargo run -- -V
Run all benchmarks in this repo:
cargo run -- --release -- --bench