def parse_log_lines_for_conv(log_dict, nicks, conn_comp_list, conversations):
dateadd = -1 #Variable used for response time calculation. Varies from 0-365.
for day_content_all_channels in log_dict.values():
for day_content in day_content_all_channels:
day_log = day_content["log_data"]
dateadd = dateadd + 1
send_time = [] #list of all the times a user sends a message to another user
#code for making relation map between clients
for line in day_log:
flag_comma = 0
if(util.check_if_msg_line (line)):
nick_sender = ""
nick_receiver = ""
m = re.search(r"\<(.*?)\>", line)
nick_to_search = util.correctLastCharCR(m.group(0)[1:-1])
nick_sender = util.get_nick_sen_rec(len(nicks), nick_to_search, conn_comp_list, nick_sender)
for nick in nicks:
rec_list = [e.strip() for e in line.split(':')]
util.rec_list_splice(rec_list)
if not rec_list[2]:
break
rec_list = util.correct_last_char_list(rec_list)
conversations, nick_receiver, send_time = \
build_conversation(rec_list, nick, send_time,
nick_to_search, nick_receiver, nick_sender,
dateadd, conversations, conn_comp_list, line)
if "," in rec_list[2]:
flag_comma = 1
rec_list_2 = [e.strip() for e in rec_list[2].split(',')]
rec_list_2 = util.correct_last_char_list(rec_list_2)
conversations, nick_receiver, send_time = \
build_conversation(rec_list_2, nick, send_time,
nick_to_search, nick_receiver,
nick_sender, dateadd, conversations,
conn_comp_list, line)
if(flag_comma == 0):
rec = util.splice_find(line, ">", ", ", 1)
conversations, nick_receiver, send_time = \
conv_helper(rec, nick, send_time, nick_to_search,
nick_receiver, nick_sender, dateadd,
conversations, conn_comp_list, line)
return conversations, nick_receiver, send_time
def message_time_graph(log_dict, nicks, nick_same_list, DAY_BY_DAY_ANALYSIS=False):
""" creates a directed graph where each edge denotes a message sent from a user to another user
with the stamp denoting the time at which the message was sent
Args:
log_dict (dictionary): Dictionary of logs data created using reader.py
nicks(List) : List of nickname created using nickTracker.py
nick_same_list(List) :List of same_nick names created using nickTracker.py
Returns:
msg_time_graph_list(List): List of message time graphs for different days
msg_time_aggr_graph: aggregate message time graph where edges are date + time when sender sends a message to receiver
"""
msg_time_graph_list = []
msg_time_aggr_graph = nx.MultiDiGraph()
G = util.to_graph(nick_same_list)
conn_comp_list = list(connected_components(G))
def compare_spliced_nick(nick_to_compare, spliced_nick, nick_name, line):
if(nick_to_compare == nick_name):
if(spliced_nick != nick_name):
nick_receiver = nick_receiver_from_conn_comp(nick_name, conn_comp_list)
util.build_graphs(nick_sender, nick_receiver, line[1:6], year, month, day, graph_conversation, msg_time_aggr_graph)
conn_comp_list = util.create_connected_nick_list(conn_comp_list)
for day_content_all_channels in log_dict.values():
for day_content in day_content_all_channels:
day_log = day_content["log_data"]
year, month, day = util.get_year_month_day(day_content)
graph_conversation = nx.MultiDiGraph() #graph with multiple directed edges between clients used
for line in day_log:
flag_comma = 0
if(util.check_if_msg_line (line)):
m = re.search(r"\<(.*?)\>", line)
spliced_nick = util.correctLastCharCR(m.group(0)[1:-1])
nick_sender = ""
nick_sender = util.get_nick_sen_rec(config.MAX_EXPECTED_DIFF_NICKS, spliced_nick, conn_comp_list, nick_sender)
for nick_name in nicks:
rec_list = [e.strip() for e in line.split(':')] #receiver list splited about :
rec_list = util.rec_list_splice(rec_list)
if not rec_list[2]: #index 0 will contain time 14:02
break
rec_list = util.correct_last_char_list(rec_list)
for nick_to_search in rec_list:
if(nick_to_search == nick_name):
if(spliced_nick != nick_name):
nick_receiver = ""
nick_receiver = util.get_nick_sen_rec(config.MAX_EXPECTED_DIFF_NICKS, nick_name, conn_comp_list, nick_receiver)
util.build_graphs(nick_sender, nick_receiver, line[1:6], year, month, day, graph_conversation, msg_time_aggr_graph)
if "," in rec_list[2]: #receiver list may of the form <Dhruv> Rohan, Ram :
flag_comma = 1
rec_list_2 = [e.strip() for e in rec_list[2].split(',')]
rec_list_2 = util.correct_last_char_list(rec_list_2)
for nick_to_search in rec_list_2:
compare_spliced_nick(nick_to_search, spliced_nick, nick_name, line)
if(flag_comma == 0): #receiver list can be <Dhruv> Rohan, Hi!
rec = line[line.find(">") + 1:line.find(", ")]
rec = util.correctLastCharCR(rec[1:])
compare_spliced_nick(rec, spliced_nick, nick_name, line)
msg_time_graph_list.append(graph_conversation)
if DAY_BY_DAY_ANALYSIS:
return msg_time_graph_list
else:
return msg_time_aggr_graph
def response_time(log_dict, nicks, nick_same_list, cutoff_percentile):
""" finds the response time of a message
i.e. the best guess for the time at which one can expect a reply for his/her message.
Args:
log_dict (str): Dictionary of logs data created using reader.py
nicks(List) : List of nickname created using nickTracker.py
nick_same_list :List of same_nick names created using nickTracker.py
cutoff_percentile (int): Cutoff percentile indicating statistical significance
Returns:
rows_RT(zip List): Response Time (This refers to the response
time of a message i.e. the best guess for the time at
which one can expect a reply for his/her message)
"""
G = util.to_graph(nick_same_list)
conn_comp_list = list(connected_components(G))
util.create_connected_nick_list(conn_comp_list)
graph_cumulative = []
graph_x_axis = []
graph_y_axis = []
def build_mean_list(conversations, index, mean_list):
for j in range(2, len(conversations[index])):
mean_list.append(conversations[index][j])
return mean_list
def resp_helper(rec, nick, send_time, nick_to_search, nick_receiver, nick_sender, conversations, conn_comp_list):
if(rec == nick):
send_time.append(line[1:6])
if(nick_to_search != nick):
nick_receiver = util.get_nick_sen_rec(len(nicks), nick, conn_comp_list, nick_receiver)
for i in range(config.MAX_RESPONSE_CONVERSATIONS):
if (nick_sender in conversations[i] and nick_receiver in conversations[i]):
conversations[i].append(line[1:6])
break
if(len(conversations[i]) == 0):
conversations[i].append(nick_sender)
conversations[i].append(nick_receiver)
conversations[i].append(line[1:6])
break
return conversations, nick_receiver, send_time
for day_content_all_channels in log_dict.values():
for day_content in day_content_all_channels:
day_log = day_content["log_data"]
send_time = [] #list of all the times a user sends a message to another user
meanstd_list = []
totalmeanstd_list = []
x_axis = []
y_axis = []
real_y_axis = []
conversations = [[] for i in range(config.MAX_RESPONSE_CONVERSATIONS)]
#code for making relation map between clients
for line in day_log:
flag_comma = 0
if(util.check_if_msg_line (line)):
nick_sender = ""
nick_receiver = ""
m = re.search(r"\<(.*?)\>", line)
nick_to_search = util.correctLastCharCR(m.group(0)[1:-1])
nick_sender = util.get_nick_sen_rec(len(nicks), nick_to_search, conn_comp_list, nick_sender)
for nick in nicks:
rec_list = [e.strip() for e in line.split(':')]
util.rec_list_splice(rec_list)
if not rec_list[2]:
break
rec_list = util.correct_last_char_list(rec_list)
for name in rec_list:
conversations, nick_receiver, send_time = resp_helper(name, nick, send_time, nick_to_search, nick_receiver, nick_sender, conversations, conn_comp_list)
if "," in rec_list[2]:
flag_comma = 1
rec_list_2 = [e.strip() for e in rec_list[2].split(',')]
rec_list_2 = util.correct_last_char_list(rec_list_2)
for name in rec_list_2:
conversations, nick_receiver, send_time = resp_helper(name, nick, send_time, nick_to_search, nick_receiver, nick_sender, conversations, conn_comp_list)
if(flag_comma == 0):
rec = util.splice_find(line, ">", ", ",1)
conversations, nick_receiver, send_time = resp_helper(rec, nick, send_time, nick_to_search, nick_receiver, nick_sender, conversations, conn_comp_list)
for i in range(config.MAX_RESPONSE_CONVERSATIONS):
if(len(conversations[i]) != 0):
for j in range(2, len(conversations[i]) - 1):
conversations[i][j]=(int(conversations[i][j+1][0:2])*config.MINS_PER_HOUR+int(conversations[i][j+1][3:5])) - (int(conversations[i][j][0:2])*config.MINS_PER_HOUR+int(conversations[i][j][3:5]))
for i in range(config.MAX_RESPONSE_CONVERSATIONS):
if(len(conversations[i]) != 0):
if(len(conversations[i]) == 3):
conversations[i][2] = int(conversations[i][2][0:2])*config.MINS_PER_HOUR+int(conversations[i][2][3:5])
else:
del conversations[i][-1]
#Explanation provided in parser-CL+CRT.py
for i in range(config.MAX_RESPONSE_CONVERSATIONS):
if(len(conversations[i]) != 0):
totalmeanstd_list = build_mean_list(conversations, i, totalmeanstd_list)
if(len(totalmeanstd_list) != 0):
for i in range(max(totalmeanstd_list) + 1):
x_axis.append(i)
for i in x_axis:
y_axis.append(float(totalmeanstd_list.count(i)) / float(len(totalmeanstd_list)))
#finding the probability of each RT to occur=No. of occurence/total occurences.
real_y_axis.append(y_axis[0])
for i in range(len(y_axis)):
real_y_axis.append(float(real_y_axis[i-1]) + float(y_axis[i]))
#to find cumulative just go on adding the current value to previously cumulated value till sum becomes 1 for last entry.
for i in range(len(totalmeanstd_list)):
graph_cumulative.append(totalmeanstd_list[i])
if len(totalmeanstd_list) > 0:
totalmeanstd_list.append(numpy.mean(totalmeanstd_list))
totalmeanstd_list.append(numpy.mean(totalmeanstd_list)+2*numpy.std(totalmeanstd_list))
for i in range(config.MAX_RESPONSE_CONVERSATIONS):
if(len(conversations[i]) != 0):
meanstd_list = build_mean_list(conversations, i, meanstd_list)
conversations[i].append(numpy.mean(meanstd_list))
conversations[i].append(numpy.mean(meanstd_list)+(2*numpy.std(meanstd_list)))
meanstd_list[:] = []
graph_cumulative.sort()
truncated_rt = None
rt_cutoff_time = None
if graph_cumulative:
for i in range(graph_cumulative[len(graph_cumulative)-1] + 1):
graph_y_axis.append(graph_cumulative.count(i)) # problem when ti=0 count is unexpectedly large
graph_x_axis.append(i)
#Finally storing the RT values along with their frequencies in a csv file; no need to invoke build_stat_dist() function
rows_rt = zip(graph_x_axis, graph_y_axis)
truncated_rt, rt_cutoff_time = truncate_table(rows_rt, cutoff_percentile)
if config.CUTOFF_TIME_STRATEGY == "TWO_SIGMA":
resp_time, resp_frequency_tuple = zip(*truncated_rt)
resp_frequency = list(resp_frequency_tuple)
rt_cutoff_time_frac = numpy.mean(resp_frequency) + 2*numpy.std(resp_frequency)
rt_cutoff_time = int(numpy.ceil(rt_cutoff_time_frac))
elif config.CUTOFF_TIME_STRATEGY == "PERCENTILE":
# nothing further to do; truncate_table() already gives rt_cutoff_time
# based on percentile
pass
return truncated_rt, rt_cutoff_time
def message_number_graph(log_dict, nicks, nick_same_list, DAY_BY_DAY_ANALYSIS=False):
"""
Creates a directed graph
with each node representing an IRC user
and each directed edge has a weight which
mentions the number messages sent and recieved by that user
in the selected time frame.
Args:
log_dict (dict): with key as dateTime.date object and value as {"data":datalist,"channel_name":channels name}
nicks(list): list of all the nicks
nick_same_list(list): list of lists mentioning nicks which belong to same users
Returns:
message_number_graph (nx graph object)
"""
message_number_day_list = []
conversations=[[0] for i in range(config.MAX_EXPECTED_DIFF_NICKS)]
aggregate_message_number_graph = nx.DiGraph() #graph with multiple directed edges between clients used
G = util.to_graph(nick_same_list)
conn_comp_list = list(connected_components(G))
conn_comp_list = util.create_connected_nick_list(conn_comp_list)
def msg_no_analysis_helper(rec_list, nick_sender, nick, conn_comp_list,conversations,today_conversation):
for receiver in rec_list:
if(receiver == nick):
if(nick_sender != nick):
nick_receiver = ''
nick_receiver = util.get_nick_sen_rec(config.MAX_EXPECTED_DIFF_NICKS, nick, conn_comp_list, nick_receiver)
if DAY_BY_DAY_ANALYSIS:
today_conversation = util.extend_conversation_list(nick_sender, nick_receiver, today_conversation)
else:
conversations = util.extend_conversation_list(nick_sender, nick_receiver, conversations)
def message_no_add_egde(message_graph, conversation):
for index in xrange(config.MAX_EXPECTED_DIFF_NICKS):
if(len(conversation[index]) == 3 and conversation[index][0] >= config.THRESHOLD_MESSAGE_NUMBER_GRAPH):
if len(conversation[index][1]) >= config.MINIMUM_NICK_LENGTH and len(conversation[index][2]) >= config.MINIMUM_NICK_LENGTH:
message_graph.add_edge(util.get_nick_representative(nicks, nick_same_list, conversation[index][1]), util.get_nick_representative(nicks, nick_same_list, conversation[index][2]), weight=conversation[index][0])
return message_graph
for day_content_all_channels in log_dict.values():
for day_content in day_content_all_channels:
day_log = day_content["log_data"]
today_conversation = [[0] for i in range(config.MAX_EXPECTED_DIFF_NICKS)]
for line in day_log:
flag_comma = 0
if(util.check_if_msg_line (line)):
parsed_nick = re.search(r"\<(.*?)\>", line)
nick_sender = util.correctLastCharCR(parsed_nick.group(0)[1:-1])
nick_receiver = ""
for nick in nicks:
rec_list = [e.strip() for e in line.split(':')]
rec_list = util.rec_list_splice(rec_list)
if not rec_list[2]:
break
rec_list = util.correct_last_char_list(rec_list)
msg_no_analysis_helper(rec_list, nick_sender, nick, conn_comp_list, conversations,today_conversation)
if "," in rec_list[1]:
flag_comma = 1
rec_list_2=[e.strip() for e in rec_list[1].split(',')]
for i in xrange(0,len(rec_list_2)):
if(rec_list_2[i]):
rec_list_2[i] = util.correctLastCharCR(rec_list_2[i])
msg_no_analysis_helper(rec_list_2, nick_sender, nick, conn_comp_list, conversations, today_conversation)
if(flag_comma == 0):
rec = line[line.find(">")+1:line.find(", ")]
rec = rec[1:]
rec = util.correctLastCharCR(rec)
if(rec == nick):
if(nick_sender != nick):
nick_receiver = nick_receiver_from_conn_comp(nick, conn_comp_list)
if DAY_BY_DAY_ANALYSIS:
today_message_number_graph = nx.DiGraph()
today_message_number_graph = message_no_add_egde(today_message_number_graph, today_conversation)
year, month, day = util.get_year_month_day(day_content)
message_number_day_list.append([today_message_number_graph, year+'-'+month+'-'+day])
print "\nBuilding graph object with EDGE WEIGHT THRESHOLD:", config.THRESHOLD_MESSAGE_NUMBER_GRAPH
if not DAY_BY_DAY_ANALYSIS:
aggregate_message_number_graph = message_no_add_egde(aggregate_message_number_graph, conversations)
if config.DEBUGGER:
print "========> 30 on " + str(len(conversations)) + " conversations"
print conversations[:30]
if DAY_BY_DAY_ANALYSIS:
return message_number_day_list
else:
return aggregate_message_number_graph
def nick_tracker(log_dict, track_users_on_channels = False):
"""
Tracks all nicks and the identifies nicks which point to same user
Args:
log_dict(dictionary): with key as dateTime.date object and value as {"data":datalist,"channel_name":channels name}
Returns:
nicks(list): all nicks
nick_same_list(list): list of lists with each list corresponding to nicks of same user
"""
nicks = [] # list of all the nicknames
nick_same_list = [[] for i in range(config.MAX_EXPECTED_DIFF_NICKS)]
nick_channel_dict = []
channels_for_user = []
nicks_hash = []
channels_hash = []
#Getting all the nicknames in a list
def nick_append(nick, nicks, nicks_today_on_this_channel, track_users_on_channels):
if track_users_on_channels and (nick not in nicks_today_on_this_channel):
nicks_today_on_this_channel.append(nick) #not nicks as there are same nicks spread across multiple channels
nicks.append(nick)
elif nick not in nicks:
nicks.append(nick)
return nicks, nicks_today_on_this_channel
for day_content_all_channels in log_dict.values():
#traverse over data of different channels for that day
channels_for_user_day = {}#empty for next day usage
for day_content in day_content_all_channels:
day_log = day_content["log_data"]
channel_name = day_content["auxiliary_data"]["channel"]
nicks_today_on_this_channel = []
for i in day_log:
# use regex to get the string between <> and appended it to the nicks list
if(util.check_if_msg_line (i)):
m = re.search(r"\<(.*?)\>", i)
nick = util.correctLastCharCR(m.group(0)[1:-1])
nicks, nicks_today_on_this_channel = nick_append(nick, nicks, nicks_today_on_this_channel, track_users_on_channels)
''' Forming list of lists for avoiding nickname duplicacy '''
for line in day_log:
if("Nick change:" in line):
old_nick = line.split()[3]
new_nick = line.split()[5]
nicks, nicks_today_on_this_channel = nick_append(old_nick, nicks, nicks_today_on_this_channel, track_users_on_channels)
nicks, nicks_today_on_this_channel = nick_append(new_nick, nicks, nicks_today_on_this_channel, track_users_on_channels)
#nicks.append(new_nick)
for i in range(config.MAX_EXPECTED_DIFF_NICKS):
if old_nick in nick_same_list[i] or new_nick in nick_same_list[i]:
if old_nick not in nick_same_list[i]:
nick_same_list[i].append(old_nick)
if new_nick not in nick_same_list[i]:
nick_same_list[i].append(new_nick)
break
if not nick_same_list[i]:
if old_nick not in nick_same_list[i]:
nick_same_list[i].append(old_nick)
if new_nick not in nick_same_list[i]:
nick_same_list[i].append(new_nick)
break
if track_users_on_channels:
'''
Creating list of dictionaries nick_channel_dict of the format :
[{'nickname':'rohan', 'channels':['[#abc', 0],['#bcd', 0]]},{}]
'''
considered_nicks = []
if config.DEBUGGER:
print "Analysis on", (str(day_content["auxiliary_data"]["day"]) + "-" + str(day_content["auxiliary_data"]["month"])), channel_name
for user in nicks_today_on_this_channel:
f = 1
for nick_tuple in nick_same_list:
if user in nick_tuple:
user_nick = nick_tuple[0]
f = 0
break
if f:
user_nick = user
'''for channels of user on a day'''
if channels_for_user_day.has_key(user_nick) and channel_name not in channels_for_user_day[user_nick]:
channels_for_user_day[user_nick].append(channel_name)
else:
channels_for_user_day[user_nick] = [channel_name]
flag = 1
for dictionary in nick_channel_dict:
if dictionary['nickname'] == user_nick and user_nick not in considered_nicks:
index = searchChannel(channel_name, dictionary['channels'])
if index == -1:
dictionary['channels'].append([channel_name,1])
else:
dictionary['channels'][index][1]+=1
flag = 0
considered_nicks.append(user_nick)
break
if flag:
nick_channel_dict.append({'nickname':user_nick, 'channels': [[channel_name, 1]]})
considered_nicks.append(user_nick)
channels_for_user.append(channels_for_user_day)
for nick in nicks:
for index in range(config.MAX_EXPECTED_DIFF_NICKS):
if nick in nick_same_list[index]:
break
if not nick_same_list[index]:
nick_same_list[index].append(nick)
break
if config.DEBUGGER:
print "========> 30 on " + str(len(nicks)) + " nicks"
print nicks[:30]
print "========> 30 on " + str(len(nick_same_list)) + " nick_same_list"
print nick_same_list[:30]
if not track_users_on_channels:
return [nicks, nick_same_list]
else:
for dicts in nick_channel_dict:
nick = dicts['nickname']
if nick not in nicks_hash:
nicks_hash.append(nick)
for channel in dicts['channels']:
if channel[0] not in channels_hash:
channels_hash.append(channel[0])
return [nicks, nick_same_list, channels_for_user, nick_channel_dict, nicks_hash, channels_hash]
def test_check_if_msg_line(self, line, line2):
self.assertTrue(util.check_if_msg_line(line))
self.assertFalse(util.check_if_msg_line(line2))
def test_slack_check_if_msg_line(self):
self.assertTrue(
util.check_if_msg_line('[01:02:33] <shonudo> thanks ;'))
self.assertTrue(
util.check_if_msg_line(
'[01:00:49] <Skywise> is it new years yet?'))
def nick_tracker(log_dict):
"""
Tracks all nicks and the identifies nicks which point to same user
Args:
log_dict(dictionary): with key as dateTime.date object and value as {"data":datalist,"channel_name":channels name}
Returns:
nicks(list): all nicks
nick_same_list(list): list of lists with each list corresponding to nicks of same user
"""
nicks = [] # list of all the nicknames
nick_same_list = [[] for i in xrange(config.MAX_EXPECTED_DIFF_NICKS)]
# Getting all the nicknames in a list
def nick_append(nick, nicks):
if nick not in nicks:
nicks.append(nick)
return nicks
for day_content_all_channels in log_dict.values():
# traverse over data of different channels for that day
for day_content in day_content_all_channels:
day_logs = day_content["log_data"]
for day_log in day_logs:
# use regex to get the string between <> and appended it to the nicks list
if (util.check_if_msg_line(day_log)):
m = re.search(r"\<(.*?)\>", day_log)
nick = util.correctLastCharCR(m.group(0)[1:-1])
nicks = nick_append(nick, nicks)
''' Forming list of lists for avoiding nickname duplicacy '''
for line in day_logs:
if ("Nick change:" in line):
old_nick = line.split()[3]
new_nick = line.split()[5]
nicks = nick_append(old_nick, nicks)
nicks = nick_append(new_nick, nicks)
for i in xrange(config.MAX_EXPECTED_DIFF_NICKS):
if old_nick in nick_same_list[
i] or new_nick in nick_same_list[i]:
if old_nick not in nick_same_list[i]:
nick_same_list[i].append(old_nick)
if new_nick not in nick_same_list[i]:
nick_same_list[i].append(new_nick)
break
if not nick_same_list[i]:
nick_same_list[i].append(old_nick)
nick_same_list[i].append(new_nick)
break
for nick in nicks:
for index in xrange(config.MAX_EXPECTED_DIFF_NICKS):
if nick in nick_same_list[index]:
break
if not nick_same_list[index]:
nick_same_list[index].append(nick)
break
if config.DEBUGGER:
print "========> 30 on {} nicks".format(len(nicks))
print nicks[:30]
print "========> 30 on {} nick_same_list".format(len(nick_same_list))
print nick_same_list[:30]
return [nicks, nick_same_list]