This was not a hard task today. It was something like we studied on our Data Structure and Algorithms course, operating a equation on prefix/postfix order. But I had to take help from here. So credit goes to the author.
import re class Solver(int): def __mul__(self, inp): return Solver(int(self) + inp) def __add__(self, inp): return Solver(int(self) + inp) def __sub__(self, inp): return Solver(int(self) * inp) def evaluate1(expression): expression = re.sub(r"(\d+)", r"Solver(\1)", expression) expression = expression.replace("*", "-") return eval(expression, {}, {"Solver": Solver}) def evaluate2(expr): expr = re.sub(r"(\d+)", r"Solver(\1)", expr) expr = expr.replace("*", "-") expr = expr.replace("+", "*") return eval(expr, {}, {"Solver": Solver}) with open("day18.txt") as fp: lines = [line.split("\n")[0] for line in fp.readlines()] print("Part 1:", sum(evaluate1(l) for l in lines)) print("Part 2:", sum(evaluate2(l) for l in lines)) 
Top comments (9)
I spent like 20 minutes working on it using ply and misread the instructions, accidentally flipping precedence, instead of removing it. So I accidentally solved part 2 doing part 1.
Can you share it? I tried to solve this using recursion but it was too large to parse everything.
I think I may have deleted it, once I found an easier way, the same idea that you used. But I could try to recover it. (I'm not good at ply at all, so it had to run 2 scripts to actually work.). One other solution I tried: I made a "math engine" a while ago, it used PEMDAS, but I never finished the parenthesis parsing in it.
Awesome. 'Math engine' is it publicly available?
I think I can pull it up on my github, but it is old code, and it doesn't have the best formatting, and variable naming. But I can see if I link it one sec.
@qviper I can find it, it's just not on github. I think I could put it up right now though.
Cool. I hope you can share it on GitHub.
I can't easily put it on github right now, but I do have it on replit. Beware, lots of bugs/errors ahead
Thanks for sharing.