import time
import random

def randomLine(file):
	""" Selects a line at random from a file.

	Uses only one pass through the file, and still gives a uniform chance of each line
	being picked.  The math is basically:

	Base case:
	If you have 1 line, you return it, giving every line a 1/1 chance.

	After N lines, each line has a 1/N chance of being selected.  Therefore, there is a
	1/N * N/(N+1) chance of it still being selected (a 1 - 1/(N+1) chance of the new
	item not being selected).  The N's reduce, leaving an 1/(N+1) of any previous line being
	selected, and since we took the new line with probability 1/(N+1), all lines have a
	1/(N+1) probability of being selected at this point.

	import boilerplate_induction_verbosity_here
	current = 0
	count = 0 

	random.seed( time.time() ) 

	f = open(file, "r")
	for line in f.readlines():

		if count == 0:
			current = line
			chooseit = random.random() 
			threshold = 1.0 / (count + 1.0)
			if chooseit < threshold: 
				current = line
		count = count + 1

	return current

chose = randomLine("data")
print "Chose line: " + chose