WGriffing/aoc-python

Forked from https://github.com/marcelblijleven/adventofcode

Advent of Code ⭐️

Collection of my Advent of Code solutions in an overkill project setup 👻🎄.

Features ✨

• Solutions are timed with the help of a decorator using time.perf_counter
• Solution and time are printed to console using the rich package with truecolor
• Solution profiler decorator using Cprofile and pstats
• Automatic listing of completed solutions in the README
• Automatic changelog, using semantic versioning and the conventional commit specification
• A badge that is updated automatically with the amount of stars I've collected
• A svg image that is updated automatically with the amount of stars in the style of Advent of Code
  • Has a dark mode and light mode version which are visible only when their respective mode is used by the user
  • Has an animated star emoji which helps the Elves save Christmas
• Pip installable (pip install -e .) with:
  • A generate-readme script, which updates the readme
  • A generate-benchmarks script, which runs all solutions and saves the duration of each solution to these tables
  • A run-all script, which dynamically calls every solution in every adventofcode.year_*.day_* module
  • An add-day script, which add a solution day file using a template and downloads the input data from the AOC site automatically
• Type checked (mypy) and linted (flake8)
• Tested against multiple python versions on each push to master and pull request

Completed ⭐️

2023

Solutions for 2023

day	part one	part two
01	⭐️	⭐️
02	⭐️	⭐️
03	⭐️	⭐️
04	⭐️	⭐️

Benchmarks 🚀

2023

Benchmarks for 2023

day	part	duration
01	part one	1.05 ms
01	part two	6.91 ms
02	part one	3.31 ms
02	part two	2.54 ms
03	part one	98.82 ms
03	part two	56.39 ms
04	part one	2.58 ms
04	part two	2.60 ms

Decorators

What's Christmas without decorations? 🎄

Solution timer

The solution timer times the solution using time.perf_counter and outputs the answer and the duration to the console

Example:

@register_solution(2015, 9, 1)  # year, day, part
def part_one(input_data: list[str]) -> int:
    ...

Output:

2015 day 09 part 01: 251 in 0.1356 ms

Solution profiler

The solution profiler runs the cProfiler against the solution and outputs the profiler stats using pstats to the console. It takes an optional amount kwarg to set the amount of stats to display, and an optional sort kwarg to set the sorting to either time or cumulative.

Example:

@solution_profiler(2015, 9, 1)  # year, day, part
def part_one(input_data: list[str]) -> int:
    ...

Output:

91416 function calls (90941 primitive calls) in 0.159 seconds

Ordered by: internal time
List reduced from 217 to 3 due to restriction <10>

ncalls  tottime  percall  cumtime  percall filename:lineno(function)
    1    0.133    0.133    0.136    0.136 /Users/marcelblijleven/.../day_09_2015.py:39(_get_route_distances)
    1    0.012    0.012    0.015    0.015 /Users/marcelblijleven/.../day_09_2015.py:30(get_all_routes)
82182    0.006    0.000    0.006    0.000 {method 'append' of 'list' objects}

Scripts

add-day

The add-day script creates a file based on a 'solution day' template into the correct year module. If no input is found for that day, it will automatically download the input and save it in the inputs directory. Note: this only works if the session cookie is stored in .session. To get this value:

1. Go to the AOC site.
2. Make sure you're logged in, every user has unique input data
3. View the cookies and copy the value of the session cookie.
4. Paste the cookie value into the .session file

Example:

(venv) add-day 2015 14

Output:

(venv) [adventofcode] add-day 2015 14                                                                                                                                                                   master  ✗ ✭ ✱
Creating solution day file for year 2015 day 14
Wrote template to /Users/marcelblijleven/code/github.com/marcelblijleven/adventofcode/src/adventofcode/year_2015/day_14_2015.py
Input data already exists for year 2015 day 14, skipping download

generate-readme

The generate-readme script dynamically searches for all solutions and writes them to the README.md file. When a solution file has a function called part_one, it adds a star. When it has a function called part_two, it adds another star. The star counter badge at the top of the README.md file is then updated with the total amount of stars found.

This script is only used in the Github workflow update_readme.yml, but can be run locally to using generate-readme

clean-repo

The clean-repo script is used to delete all solutions and inputs from the project. This can be useful if you want to start over, or if you've just forked this repo. The clean-repo command is run in 'dry run mode' by default, to disable it and actually start deleting directories and files, use:

(venv) clean-repo --dry-run false 

Note: not all years/solutions have been migrated yet from my previous repositories