attachment_size.py

from pathlib import Path


def sizeof_fmt(num, suffix='B'):
    """Readable file size"""
    for unit in ['', 'k', 'M', 'G', 'T', 'P', 'E', 'Z']:
        if abs(num) < 1024.0:
            return "%3.1f %s%s" % (num, unit, suffix)
        num /= 1024.0
    return "%.1f%s%s" % (num, 'Yi', suffix)


def calculate_size(data):
    attachments = []
    for item in data['items']:
        for attachment in item['attachments']:
            if 'path' in attachment:
                attachment['size'] = Path(attachment['path']
                    ).stat().st_size
                attachments.append({
                    'path': attachment['path'],
                    'size': attachment['size'],
                    'uri': attachment['uri'],
                    'item': {
                        'title': item['title'],
                        'creators': item['creators'],
                        'date': item['date'],
                    }
                })
    return attachments


def main():
    import argparse
    import json
    from operator import itemgetter
    argparser = argparse.ArgumentParser()
    argparser.add_argument('-o', '--outfile',
                           type=argparse.FileType('w'))
    argparser.add_argument('-p', '--print-big',
                           action='store_true')
    argparser.add_argument('-m', '--min-size',
                           type=int, default=1024 * 1024)
    argparser.add_argument('-s', '--print-sum',
                           action='store_true')
    argparser.add_argument('infile', type=argparse.FileType())
    args = argparser.parse_args()
    data = json.load(args.infile)
    attachments = calculate_size(data)
    if args.outfile:
        json.dump(data, args.outfile, indent=2)
    if args.print_sum:
        print(sizeof_fmt(sum(
            [att['size'] for att in attachments])))
    if args.print_big:
        attachments.sort(key=itemgetter('size'), reverse=True)
        for attachment in attachments:
            if attachment['size'] < args.min_size:
                break
            print(sizeof_fmt(attachment['size']),
                  attachment['uri'].split('/')[-1])

if __name__ == '__main__':
    main()