Hacker News Reader

Comment by okso | Hacker News Reader

Naïve Python3 is not as fast as Numpy, but pretty elegant:

  def main(filename):
      d = open(filename, 'rb').read()
      result = sum(d) % 256
      print("The answer is: ", result)

View on HN

Note that sum(d) will generate a huge number, possibly using lots of memory and processing power. A better option would be:

    def main(filename):
      d = open(filename, 'rb').read()
      result = reduce(lambda i, j: (i + j) % 256, d)
      print("The answer is: ", result)

Note how this is similar to the squaring algorithm used in cryptography: http://en.wikipedia.org/wiki/Exponentiation_by_squaring

wouldn't you need at least 2^55 byte (36 petabytes) size file to go over 64bit integer 'sum' variable? I think such limitation is far preferable to an extra modulo operation for each and every byte. Not sure though how python does its integers, so the threshold for using bignum might be lower.

That approach would be very well adapted to a low-level language such as C with the risk of overflow.

Python handles long integers transparently, and the overhead of managing the lambda function and additional variables in Python is probably much more time consuming than handling a long integer.

Reply on news.ycombinator.com