multi-upload-etag-calculation

#!/usr/bin/env python3
# -*- mode: python -*-

import sys
import hashlib

def calculate_multipart_etag(source_path, chunk_size, expected=None):
    md5s = []

    with open(source_path,'rb') as fp:
        while True:
            data = fp.read(chunk_size)
            if not data:
                break
            md5s.append(hashlib.md5(data))

    if len(md5s) > 1:
        digests = b"".join(m.digest() for m in md5s)
        new_md5 = hashlib.md5(digests)
        new_etag = '"%s-%s"' % (new_md5.hexdigest(),len(md5s))
    elif len(md5s) == 1:
        # file smaller than chunk size
        new_etag = '"%s"' % md5s[0].hexdigest()
    else:
        new_etag = '""'

    if expected:
        if not expected==new_etag:
            raise ValueError('new etag %s does not match expected %s' % (new_etag,expected))

    return new_etag

if __name__ == '__main__':

	source_path = sys.argv[1]
	chunk_size = sys.argv[2]
	chunk_size = int(chunk_size) * 1024 * 1024

	try:
		expected = '"%s"' % (sys.argv[3])
	except Exception as e:
		expected = None

	print(calculate_multipart_etag(source_path,chunk_size,expected))