Compute total size of git-annexed files with only one local copy

get_size_localcopy_annex.py

# This script computes the total size of git-annex files with only a single local copy.
# It's useful to figure out how much data will be used if all the files were to be archived.
import subprocess
from tqdm import tqdm
import json
import os

def get_files_with_one_copy():
    try:
        result = subprocess.run(['git-annex', 'find', '--copies=1', '--and', '--not', '--copies=2', '--and', '--in=here'], capture_output=True, text=True, check=True)
        files = result.stdout.splitlines()
        return files
    except subprocess.CalledProcessError as e:
        print(f"Error finding files: {e}")
        return []

def get_file_size(file):
    try:
        # result = subprocess.run(['git-annex', 'info', file, '--json', '--bytes', '--fast'], capture_output=True, text=True, check=True)
        # info = json.loads(result.stdout)
        # return int(info['size'])
        result = subprocess.run(['du', '-bL', file], capture_output=True, text=True, check=True)
        return int(result.stdout.split()[0])
    except subprocess.CalledProcessError as e:
        print(f"Error getting info for {file}: {e}")
        return 0
    except (json.JSONDecodeError, KeyError) as e:
        print(f"Error parsing info for {file}: {e}")
        return 0

def main():
    files = get_files_with_one_copy()
    total_size = 0

    for file in tqdm(files, desc="Processing files"):
        size = get_file_size(file)
        total_size += size

    human_readable_size = subprocess.run(['numfmt', '--to=iec-i', '--suffix=B', str(total_size)], capture_output=True, text=True).stdout.strip()
    n_files = len(files)

    print(f"Total size of {n_files} files with only one copy: {human_readable_size}")

if __name__ == "__main__":
    main()