#!/usr/bin/python

# Overview:  Displays the top inputs
# Usage: kallahar_livestats logs/access.log 10
#
# kallahar_livestats:
#
# keeps an array of all the reffers as well as how many hits it's had
# displays the top (20) and their count, in real time
#
# by kallahar@quickwired.com and Aaron Klingaman, February 2006



import curses
import sys
import string
import time 


def main():
	uniquelines= {}

	filename = sys.argv[1]
	column = int(sys.argv[2])

	f = open(sys.argv[1], "rb") 
	f.seek(0, 2) # seek to the end rather than process the entire logfile

	while 1: 
		data = f.read(1000000) 
		if data: 
			raw = str(data)
			lines = raw.split("\n")
			for line in lines:
				line = string.strip(line) # strip extra spaces
				if line != "":
					ra_val = line.split(" ")
					if len(ra_val) < 10:
						print "Error: line too short, try increasing the f.read length; ", line
					else:
						scr.clear()
						val = string.strip(ra_val[column])
						if uniquelines.has_key( val ):
							uniquelines[val] += 1
						else:
							uniquelines[val]= 1

						entries=[(k,v) for v,k in uniquelines.items()]
						total_entries = sum(map(lambda x: x[0], entries))
						entries.sort()
						entries.reverse()
						top_x= entries[:20]

						index= 0
						scr.addstr( 0, 0, "kallahar_livestats: most popular items" )
						for (value, key) in top_x:
							pct = float(float(value) / float(total_entries)) * 100
							scr.addstr( index+1, 0, "%d\t%d %d%%\t%s" % (index+1, value, pct, key) )
							index += 1
						scr.addstr( index+1, 0, "Total Entries: %d" % total_entries )
						scr.refresh()

		else: 
			time.sleep(1)   # wait a second 
			f.seek(0, 1)    # seek to the current position 
	f.close() 

try:
	scr= curses.initscr()
	main()
	curses.endwin()

except KeyboardInterrupt, e:
	curses.endwin()
	print "Usage: kallahar_livestats /var/log/access.log 10"
	print " where 10 is the column number to use"
	print " 1  = ip"
	print " 6  = requested file"
	print " 10 = referer"