recursive compare two folders

dir_compare.py

import os
import hashlib
import argparse

def md5sum(filename, blocksize=65536):
    hash_md5 = hashlib.md5()
    with open(filename, "rb") as f:
        for chunk in iter(lambda: f.read(blocksize), b""):
            hash_md5.update(chunk)
    return hash_md5.hexdigest()

def get_all_files(root):
    files_set = set()
    for dirpath, _, filenames in os.walk(root):
        for f in filenames:
            rel_path = os.path.relpath(os.path.join(dirpath, f), root)
            files_set.add(rel_path)
    return files_set

def main(dir1, dir2):
    files1 = get_all_files(dir1)
    files2 = get_all_files(dir2)

    only_in_1 = files1 - files2
    only_in_2 = files2 - files1
    in_both = files1 & files2

    print("Files only in", dir1)
    for f in sorted(only_in_1):
        print("  ", f)
    print("\nFiles only in", dir2)
    for f in sorted(only_in_2):
        print("  ", f)

    print("\nFiles present in both but with different content:")
    for f in sorted(in_both):
        file1 = os.path.join(dir1, f)
        file2 = os.path.join(dir2, f)
        if md5sum(file1) != md5sum(file2):
            print("  ", f)

if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="Compare two directories recursively.")
    parser.add_argument("dir1", help="First directory")
    parser.add_argument("dir2", help="Second directory")
    args = parser.parse_args()
    main(args.dir1, args.dir2)