CLI: Find largest files in directory

largest_files.py

import os
import argparse
import shutil
import textwrap


def list_files_by_size(directory_path):
    files_with_sizes = []

    for root, dirs, files in os.walk(directory_path):
        for file in files:
            full_path = os.path.join(root, file)
            try:
                size = os.path.getsize(full_path)
                files_with_sizes.append((full_path, size))
            except OSError as e:
                print(f"Could not access file {full_path}: {e}")

    return sorted(files_with_sizes, key=lambda x: x[1])

def human_readable_size(size, decimal_places=2):
    for unit in ['B', 'KB', 'MB', 'GB', 'TB', 'PB']:
        if size < 1024:
            return f"{size:.{decimal_places}f} {unit}"
        size /= 1024
    return f"{size:.{decimal_places}f} PB"

def format_table(data, col_width, size_col_width=12):
    # Headers
    header_path = "File Path".ljust(col_width)
    header_size = "Size".rjust(size_col_width)
    separator = f"+{'-' * col_width}+{'-' * size_col_width}+"

    lines = [separator, f"|{header_path}|{header_size}|", separator]

    # Rows
    for path, size in data:
        wrapped_path = textwrap.wrap(path, width=col_width)
        size_str = human_readable_size(size)

        for i, line in enumerate(wrapped_path):
            if i == 0:
                lines.append(f"|{line.ljust(col_width)}|{size_str.rjust(size_col_width)}|")
            else:
                lines.append(f"|{line.ljust(col_width)}|{'':>{size_col_width}}|")

        lines.append(separator)

    return '\n'.join(lines)


def main():
    parser = argparse.ArgumentParser(description='List files in a directory (recursively), sorted by size.')
    parser.add_argument('directory', type=str, help='Path to the directory')
    parser.add_argument('--reverse', action='store_true', help='Sort by size in descending order')
    parser.add_argument('--limit', type=int, default=None, help='Limit the number of results displayed')
    parser.add_argument('--width', type=int, default=None, help='Max width for file name column')

    args = parser.parse_args()

    files = list_files_by_size(args.directory)
    if args.reverse:
        files.reverse()
    if args.limit is not None:
        files = files[:args.limit]
    default_width = shutil.get_terminal_size(fallback=(80, 20)).columns - 15
    width = args.width or default_width

    print(format_table(files, col_width=width))


if __name__ == "__main__":
    main()