Hashlib's md5 produces different output for same input

I'm trying to write a script that produces the hash value for all filenames and directory names in a specified root directory. This is my script so far:

import hashlib
import os
import sys

class Hasher:
    def __init__(self):
        self.hash_func = hashlib.md5()

    def hash_file(self, file_path):
        with open(file_path, "rb") as file:
            self.hash_func.update(file.read())
        return self.hash_func.digest()

    def hash_dir(self, dir_path):
        for dirpath, dirnames, filenames in os.walk(dir_path):
            self.hash_func.update(dirpath.encode("utf-8"))
            for file_path in filenames:
                self.hash_func.update(file_path.encode("utf-8"))
        return self.hash_func.digest()

hasher = Hasher()
root_dir = "D:/folder/"
hash_1 = str(hasher.hash_dir(root_dir))
hash_2 = str(hasher.hash_dir(root_dir))
print(hash_1)
print(hash_2)

For some reason, it produces two different hash values for the same directory, without any change in the directory whatsoever. How can I make it so that the same hash value is produced if the directory stays the same?

Solution

The problem is that the hashlib.md5 object is reused every time, so you return a hash of the cumulative data, not only over the last/intended data.

You could solve this by creating a new Hasher object every time (so calling Hasher().hash_dir(root_dir) twice in this case). But since your Hasher class does not contain any other data than the md5 object and two methods that could be static, I would recommend to make both class methods static, and create the hashlib.md5 object in the methods itself:

import hashlib
import os


class Hasher:

    @staticmethod  # make it a static method
    def hash_file(file_path):  # no 'self' as first argument
        hash_func = hashlib.md5()  # create the hashlib.md5 object here
        with open(file_path, "rb") as file:
            hash_func.update(file.read())
        return hash_func.digest()

    @staticmethod  # make it a static method
    def hash_dir(dir_path):  # no 'self' as first argument
        hash_func = hashlib.md5()  # create the hashlib.md5 object here
        for dirpath, _, filenames in os.walk(dir_path):
            hash_func.update(dirpath.encode("utf-8"))
            for file_path in filenames:
                hash_func.update(file_path.encode("utf-8"))
        return hash_func.digest()


def main():
    root_dir = "D:/folder/"
    hash_1 = str(Hasher.hash_dir(root_dir))
    hash_2 = str(Hasher.hash_dir(root_dir))
    print(hash_1)
    print(hash_2)


if __name__ == "__main__":
    main()