Sort arbitrary sized file filled with random integers by N biggest

nums.py

#!/usr/bin/env python
import random
import sys
import threading, Queue
from heapq import nlargest
import os

# job queue
queue = Queue.Queue()
# result queue
result = []
FILE = 'nums.txt'

# From: http://effbot.org/zone/wide-finder.htm
class Worker(threading.Thread):
    def run(self):
        while 1:
            args = queue.get()
            if args is None:
                break
            result.append(process(*args))
            queue.task_done()


# Sort chunk by largest number and return it
def process(file, chunk):
    f = open(file, "rb")
    f.seek(chunk[0])
    chunk_list = f.read(chunk[1]).split('\n')
    largest = max(chunk_list)

    return largest

# Chunk input file into 1M chunks - increase size for bigger files
def getchunks(fname, size=1024*1024):
    f = open(fname)
    while 1:
        start = f.tell()
        f.seek(size, 1)
        s = f.readline()
        yield start, f.tell() - start
        if not s:
            break

#Seed a new nums.txt
def seed(nums):
  with open(FILE, 'w+') as f:
      for i in range(0,nums):
          line = str(random.randint(1, 99999999999999))
          f.write(line+'\n')

def sort():
    try:
        num_biggest = int(sys.argv[1])
    except ValueError:
        print('Biggest number value must be a valid integer: i.e. python nums.py 100')
        sys.exit(1)

    # Create 4 worker threads
    for i in range(4):
        w = Worker()
        w.setDaemon(1)
        w.start()

    # Generate chunks of file and push them to queue for workers to process
    for chunk in getchunks(FILE):
        queue.put((FILE, chunk))

    # Start processing
    queue.join()

    #Sort result list by N biggest numbers
    biggest_nums = nlargest(num_biggest, result)
    for num in biggest_nums:
        print(num)

if __name__ == '__main__':
    #seed(1000000000) #seed 100 million ints
    sort()