#!/usr/bin/env python3

import time, sys
import numpy as np
import statistics as st

time_start  = []
time_end    = []
trans_size  = []
trans_speed = []
stat_size   = []
stat_speed  = []
stat_aver   = []
stat_var    = []

# Read log file


if len(sys.argv) > 1:
    file = open(sys.argv[1], 'r')
else:
    file = open('test.log', 'r')

for line in file:
    if 'ItemStarted' in line:
        words = line.split()
        time_sec = time.strptime(words[1] + ' ' + words[2].split('.')[0], "%Y/%m/%d %H:%M:%S")
        time_start.append(time.mktime(time_sec))

    elif 'LocalIndexUpdated' in line:
        words = line.split()
        time_sec = time.strptime(words[1] + ' ' + words[2].split('.')[0], "%Y/%m/%d %H:%M:%S")
        time_end.append(time.mktime(time_sec))
        for parse in words:
            if 'size:' in parse:
                parse = parse.replace(']', '')
                trans_size.append(float(parse.split(':')[1]))

# Check consistency of table sizes, attempt to correct

if len(time_end) != len(trans_size):
    print('[ERROR] Incoherent data! Exiting...')
    exit(1)

while len(time_start) != len(time_end):
    print('[WARNING] An element has been removed from the time_start list...')
    time_start.pop()

# Calculate transfer speeds in MBps

trans_speed = np.divide(trans_size, np.subtract(time_end, time_start))
n_run = len(time_end)

# Create table for statistics:
#     - stat_size contains the possible file sizes
#     - stat_speed is a 2-D list of the transfer speeds per file size
#
# The idea is to have a table like this:
#
#     file1      file2      file3      file4
#  2.1 MBps   7.3 MBps   0.5 MBps   4.2 MBps
#  2.2 MBps   7.1 MBps   0.3 MBps   4.1 MBps
#  2.0 MBps   7.2 MBps   0.8 MBps   4.7 MBps
#  2.4 MBps   7.5 MBps   0.2 MBps   4.6 MBps


for i in range(n_run):
    if trans_size[i] not in stat_size:
        stat_size.append(trans_size[i])
        stat_speed.append([])


for i in range(n_run):
    idx = stat_size.index(trans_size[i])
    stat_speed[idx].append(trans_speed[i])

# Print results, grouped by file size

n_file = len(stat_size)
for j in range(n_file):
    print()

    print("File size:", round(stat_size[j]/1048576, 2), "MB")

    print ("Run #    Speed (MBps)")

    for jj in range(len(stat_speed[j])):
        print(repr(jj+1).rjust(5), '%15.3f' % float(stat_speed[j][jj]/1048576))

    print("Speed average:   %8.2f" % float(st.mean(stat_speed[j])/1048576), "MBps")
    if len(stat_speed[j]) > 1:
        print("      std. dev.: %8.2f" % float(st.stdev(stat_speed[j])/1048576), "MBps")


exit(0)