2018-05-04 13:53:29 +02:00
|
|
|
#!/usr/bin/env python3
|
2017-03-28 17:42:44 +02:00
|
|
|
# -*-coding:UTF-8 -*
|
|
|
|
"""
|
|
|
|
This Module is used for term frequency.
|
2017-05-09 11:13:16 +02:00
|
|
|
It processes every paste coming from the global module and test the sets
|
|
|
|
supplied in the term webpage.
|
2017-03-28 17:42:44 +02:00
|
|
|
|
|
|
|
"""
|
|
|
|
import redis
|
|
|
|
import time
|
|
|
|
from pubsublogger import publisher
|
|
|
|
from packages import lib_words
|
|
|
|
from packages import Paste
|
|
|
|
import os
|
|
|
|
import datetime
|
|
|
|
import calendar
|
|
|
|
import re
|
|
|
|
import ast
|
|
|
|
from Helper import Process
|
|
|
|
|
2018-02-27 15:12:02 +01:00
|
|
|
# Email notifications
|
|
|
|
from NotificationHelper import *
|
|
|
|
|
2017-03-28 17:42:44 +02:00
|
|
|
# Config Variables
|
|
|
|
BlackListTermsSet_Name = "BlackListSetTermSet"
|
|
|
|
TrackedTermsSet_Name = "TrackedSetTermSet"
|
|
|
|
TrackedRegexSet_Name = "TrackedRegexSet"
|
|
|
|
TrackedSetSet_Name = "TrackedSetSet"
|
2018-02-27 15:12:02 +01:00
|
|
|
|
2017-03-28 17:42:44 +02:00
|
|
|
top_term_freq_max_set_cardinality = 20 # Max cardinality of the terms frequences set
|
|
|
|
oneDay = 60*60*24
|
|
|
|
top_termFreq_setName_day = ["TopTermFreq_set_day_", 1]
|
|
|
|
top_termFreq_setName_week = ["TopTermFreq_set_week", 7]
|
|
|
|
top_termFreq_setName_month = ["TopTermFreq_set_month", 31]
|
|
|
|
top_termFreq_set_array = [top_termFreq_setName_day,top_termFreq_setName_week, top_termFreq_setName_month]
|
|
|
|
|
2018-11-06 13:38:37 +01:00
|
|
|
TrackedTermsNotificationTagsPrefix_Name = "TrackedNotificationTags_"
|
|
|
|
|
2018-07-16 15:51:37 +02:00
|
|
|
# create direct link in mail
|
2018-07-17 15:11:25 +02:00
|
|
|
full_paste_url = "/showsavedpaste/?paste="
|
2018-07-16 15:51:37 +02:00
|
|
|
|
2017-03-28 17:42:44 +02:00
|
|
|
def add_quote_inside_tab(tab):
|
|
|
|
quoted_tab = "["
|
|
|
|
for elem in tab[1:-1].split(','):
|
|
|
|
elem = elem.lstrip().strip()
|
2017-04-18 15:28:21 +02:00
|
|
|
quoted_tab += "\'{}\', ".format(elem)
|
2017-03-28 17:42:44 +02:00
|
|
|
quoted_tab = quoted_tab[:-2] #remove trailing ,
|
|
|
|
quoted_tab += "]"
|
2017-04-18 15:28:21 +02:00
|
|
|
return str(quoted_tab)
|
2017-03-28 17:42:44 +02:00
|
|
|
|
|
|
|
if __name__ == "__main__":
|
|
|
|
publisher.port = 6380
|
|
|
|
publisher.channel = "Script"
|
|
|
|
|
|
|
|
config_section = 'SetForTermsFrequency'
|
|
|
|
p = Process(config_section)
|
|
|
|
|
|
|
|
# REDIS #
|
|
|
|
server_term = redis.StrictRedis(
|
2018-05-07 14:50:40 +02:00
|
|
|
host=p.config.get("ARDB_TermFreq", "host"),
|
|
|
|
port=p.config.get("ARDB_TermFreq", "port"),
|
|
|
|
db=p.config.get("ARDB_TermFreq", "db"),
|
2018-05-04 13:53:29 +02:00
|
|
|
decode_responses=True)
|
2017-03-28 17:42:44 +02:00
|
|
|
|
|
|
|
# FUNCTIONS #
|
|
|
|
publisher.info("RegexForTermsFrequency script started")
|
|
|
|
|
2018-07-17 15:11:25 +02:00
|
|
|
# create direct link in mail
|
|
|
|
full_paste_url = p.config.get("Notifications", "ail_domain") + full_paste_url
|
|
|
|
|
2017-03-28 17:42:44 +02:00
|
|
|
#get the dico and matching percent
|
|
|
|
dico_percent = {}
|
|
|
|
dico_set_tab = {}
|
2017-04-18 15:28:21 +02:00
|
|
|
dico_setname_to_redis = {}
|
2017-03-28 17:42:44 +02:00
|
|
|
for set_str in server_term.smembers(TrackedSetSet_Name):
|
|
|
|
tab_set = set_str[1:-1]
|
|
|
|
tab_set = add_quote_inside_tab(tab_set)
|
|
|
|
perc_finder = re.compile("\[[0-9]{1,3}\]").search(tab_set)
|
|
|
|
if perc_finder is not None:
|
|
|
|
match_percent = perc_finder.group(0)[1:-1]
|
2017-04-18 15:28:21 +02:00
|
|
|
dico_percent[tab_set] = float(match_percent)
|
|
|
|
dico_set_tab[tab_set] = ast.literal_eval(tab_set)
|
|
|
|
dico_setname_to_redis[tab_set] = set_str
|
2017-03-28 17:42:44 +02:00
|
|
|
else:
|
|
|
|
continue
|
|
|
|
|
|
|
|
message = p.get_from_set()
|
|
|
|
|
|
|
|
while True:
|
|
|
|
|
|
|
|
if message is not None:
|
|
|
|
filename = message
|
|
|
|
temp = filename.split('/')
|
|
|
|
timestamp = calendar.timegm((int(temp[-4]), int(temp[-3]), int(temp[-2]), 0, 0, 0))
|
|
|
|
content = Paste.Paste(filename).get_p_content()
|
|
|
|
|
|
|
|
curr_set = top_termFreq_setName_day[0] + str(timestamp)
|
|
|
|
|
|
|
|
#iterate over the words of the file
|
|
|
|
match_dico = {}
|
2017-04-18 15:28:21 +02:00
|
|
|
for word in content.split():
|
2017-03-28 17:42:44 +02:00
|
|
|
for cur_set, array_set in dico_set_tab.items():
|
2017-04-18 15:28:21 +02:00
|
|
|
for w_set in array_set[:-1]: #avoid the percent matching
|
2017-03-28 17:42:44 +02:00
|
|
|
if word == w_set:
|
|
|
|
try:
|
2017-04-18 15:28:21 +02:00
|
|
|
match_dico[str(array_set)] += 1
|
2017-03-28 17:42:44 +02:00
|
|
|
except KeyError:
|
2017-04-18 15:28:21 +02:00
|
|
|
match_dico[str(array_set)] = 1
|
2017-03-28 17:42:44 +02:00
|
|
|
|
|
|
|
#compute matching %
|
|
|
|
for the_set, matchingNum in match_dico.items():
|
2017-04-18 15:28:21 +02:00
|
|
|
eff_percent = float(matchingNum) / float((len(ast.literal_eval(the_set))-1)) * 100 #-1 bc if the percent matching
|
|
|
|
if eff_percent >= dico_percent[the_set]:
|
2018-02-27 15:12:02 +01:00
|
|
|
# Send a notification only when the member is in the set
|
2018-02-28 11:31:16 +01:00
|
|
|
if dico_setname_to_redis[str(the_set)] in server_term.smembers(TrackedTermsNotificationEnabled_Name):
|
2018-07-16 15:51:37 +02:00
|
|
|
|
|
|
|
# create mail body
|
|
|
|
mail_body = ("AIL Framework,\n"
|
|
|
|
"New occurrence for term: " + dico_setname_to_redis[str(the_set)] + "\n"
|
|
|
|
''+full_paste_url + filename)
|
|
|
|
|
2018-02-27 15:12:02 +01:00
|
|
|
# Send to every associated email adress
|
2018-02-28 11:31:16 +01:00
|
|
|
for email in server_term.smembers(TrackedTermsNotificationEmailsPrefix_Name + dico_setname_to_redis[str(the_set)]):
|
2018-07-16 15:51:37 +02:00
|
|
|
sendEmailNotification(email, 'Term', mail_body)
|
2018-02-28 11:31:16 +01:00
|
|
|
|
2018-11-06 13:38:37 +01:00
|
|
|
# tag paste
|
|
|
|
for tag in server_term.smembers(TrackedTermsNotificationTagsPrefix_Name + dico_setname_to_redis[str(the_set)]):
|
|
|
|
msg = '{};{}'.format(tag, filename)
|
|
|
|
p.populate_set_out(msg, 'Tags')
|
|
|
|
|
2017-03-28 17:42:44 +02:00
|
|
|
print(the_set, "matched in", filename)
|
2017-04-18 15:28:21 +02:00
|
|
|
set_name = 'set_' + dico_setname_to_redis[the_set]
|
|
|
|
new_to_the_set = server_term.sadd(set_name, filename)
|
|
|
|
new_to_the_set = True if new_to_the_set == 1 else False
|
2018-02-28 11:31:16 +01:00
|
|
|
|
2017-03-28 17:42:44 +02:00
|
|
|
#consider the num of occurence of this set
|
2017-04-18 15:28:21 +02:00
|
|
|
set_value = int(server_term.hincrby(timestamp, dico_setname_to_redis[the_set], int(1)))
|
2017-03-28 17:42:44 +02:00
|
|
|
|
|
|
|
# FIXME - avoid using per paste as a set is checked over the entire paste
|
|
|
|
#1 term per paste
|
2017-04-18 15:28:21 +02:00
|
|
|
if new_to_the_set:
|
|
|
|
set_value_perPaste = int(server_term.hincrby("per_paste_" + str(timestamp), dico_setname_to_redis[the_set], int(1)))
|
|
|
|
server_term.zincrby("per_paste_" + curr_set, dico_setname_to_redis[the_set], float(1))
|
|
|
|
server_term.zincrby(curr_set, dico_setname_to_redis[the_set], float(1))
|
2017-03-28 17:42:44 +02:00
|
|
|
|
|
|
|
|
|
|
|
else:
|
|
|
|
publisher.debug("Script RegexForTermsFrequency is Idling")
|
2018-04-16 14:50:04 +02:00
|
|
|
print("sleeping")
|
2017-03-28 17:42:44 +02:00
|
|
|
time.sleep(5)
|
|
|
|
message = p.get_from_set()
|