The Tor Browser: content/media/test/graph

Replace accessor implementation with direct member state manipulation, by
request https://trac.torproject.org/projects/tor/ticket/9701#comment:32

     1 #!/usr/bin/env python

     2 # graph_latency.py - graph media latency

3 #

     4 # This Source Code Form is subject to the terms of the Mozilla Public

     5 # License, v. 2.0. If a copy of the MPL was not distributed with this

     6 # file, You can obtain one at http://mozilla.org/MPL/2.0/.

     8 # needs matplotlib (sudo aptitude install python-matplotlib)

    10 import matplotlib.pyplot as plt

    11 from matplotlib import rc

    12 import sys

    13 from pprint import pprint

    14 import re

    17 # FIX!  needs to be sum of a single mediastreamtrack and any output overhead for it

    18 # So there is one sum per MST

    19 def compute_sum(data):

    20     'Compute the sum for each timestamp. This expects the output of parse_data.'

    21     last_values = {}

    22     out = ([],[])

    24     for i in data:

    25         if i[0] not in last_values.keys():

    26           last_values[i[0]] = 0

    27         last_values[i[0]] = float(i[3])

    28         print last_values

    29         out[0].append(i[2])

    30         out[1].append(sum(last_values.values()))

    31     return out

    34 def clean_data(raw_data):

    35     '''

    36     Remove the PR_LOG cruft at the beginning of each line and returns a list of

    37     tuple.

    38     '''

    39     out = []

    40     for line in raw_data:

    41         match = re.match(r'(.*)#(.*)', line)

    42         if match:

    43 	    continue

    44 	else:

    45             out.append(line.split(": ")[1])

    46     return out

    48 # returns a list of tuples

    49 def parse_data(raw_lines):

    50     '''

    51     Split each line by , and put every bit in a tuple.

    52     '''

    53     out = []

    54     for line in raw_lines:

    55         out.append(line.split(','))

    56     return out

    58 if len(sys.argv) == 3:

    59     name = sys.argv[1]

    60     channels = int(sys.argv[2])

    61 else:

    62     print sys.argv[0] + "latency_log"

    64 try:

    65     f = open(sys.argv[1])

    66 except:

    67     print "cannot open " + name

    69 raw_lines = f.readlines()

    70 lines = clean_data(raw_lines)

    71 data = parse_data(lines)

    73 final_data = {}

    75 for tupl in data:

    76     name = tupl[0]

    77     if tupl[1] != 0:

    78         name = name+tupl[1]

    79     if name not in final_data.keys():

    80         final_data[name] = ([], [])

    81 # sanity-check values

    82     if float(tupl[3]) < 10*1000:

    83         final_data[name][0].append(float(tupl[2]))

    84         final_data[name][1].append(float(tupl[3]))

    86 #overall = compute_sum(data)

    87 #final_data["overall"] = overall

    89 pprint(final_data)

    91 fig = plt.figure()

    92 for i in final_data.keys():

    93     plt.plot(final_data[i][0], final_data[i][1], label=i)

    95 plt.legend()

    96 plt.suptitle("Latency in ms (y-axis) against time in ms (x-axis).")

    98 size = fig.get_size_inches()

    99 # make it gigantic so we can see things. sometimes, if the graph is too big,

   100 # this errors. reduce the factor so it stays under 2**15.

   101 fig.set_size_inches((size[0]*10, size[1]*2))

   102 name = sys.argv[1][:-4] + ".pdf"

   103 fig.savefig(name)

The Tor Browser / file revision