#!/usr/bin/python
import sys
import os
import humanize

"""
Simple python command line program to estimate the size of a CSV or TSV file.

See also:
https://stackoverflow.com/questions/27642656/estimating-number-of-lines-in-a-file-mismatch-between-file-size-and-size-of-al
https://databio.org/posts/wcle.html
http://nathansheffield.com/wordpress/

Modern update by Abe Usher
"""

def estimate_lines(filename):
    """ Estimate the number of lines in the given file(s) """
    LEARN_SIZE = 65536
    # Get total size of all files                                                                                                                                                                   
    numLines = os.path.getsize(filename)
    with open(filename, 'rb') as file:
         buf = file.read(LEARN_SIZE)
         numLines /= (len(buf) // buf.count(b'\n'))
    return numLines

def main():
    filename = sys.argv[1]
    total_lines = estimate_lines(filename)
    total_lines_word = humanize.intword(total_lines)
    print ('Estimated lines in (filename): %s'%(total_lines_word))

if __name__ == '__main__':
    main()