AIL-framework/bin/Curve.py

178 lines
6.0 KiB
Python
Raw Normal View History

2018-05-04 13:53:29 +02:00
#!/usr/bin/env python3
# -*-coding:UTF-8 -*
"""
2014-08-07 14:46:43 +02:00
This module is consuming the Redis-list created by the ZMQ_Sub_Curve_Q Module.
2014-08-20 15:14:57 +02:00
This modules update a .csv file used to draw curves representing selected
words and their occurency per day.
..note:: The channel will have the name of the file created.
..note:: Module ZMQ_Something_Q and ZMQ_Something are closely bound, always put
the same Subscriber name in both of them.
2016-08-22 20:59:56 +02:00
This Module is also used for term frequency.
2016-08-22 20:59:56 +02:00
/!\ Top set management is done in the module Curve_manage_top_set
Requirements
------------
*Need running Redis instances. (Redis)
*Categories files of words in /files/ need to be created
*Need the ZMQ_PubSub_Tokenize_Q Module running to be able to work properly.
"""
2014-08-14 17:55:18 +02:00
import redis
import time
from pubsublogger import publisher
from packages import lib_words
import os
import datetime
import calendar
from Helper import Process
# Email notifications
from NotificationHelper import *
# Config Variables
BlackListTermsSet_Name = "BlackListSetTermSet"
TrackedTermsSet_Name = "TrackedSetTermSet"
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]
2018-07-16 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
def check_if_tracked_term(term, path):
2018-05-04 13:53:29 +02:00
if term in server_term.smembers(TrackedTermsSet_Name):
#add_paste to tracked_word_set
set_name = "tracked_" + term
server_term.sadd(set_name, path)
2018-04-16 14:50:04 +02:00
print(term, 'addded', set_name, '->', path)
p.populate_set_out("New Term added", 'CurveManageTopSets')
# Send a notification only when the member is in the set
if term in server_term.smembers(TrackedTermsNotificationEnabled_Name):
2018-04-16 14:50:04 +02:00
2018-07-16 15:51:37 +02:00
# create mail body
mail_body = ("AIL Framework,\n"
"New occurrence for term: " + term + "\n"
''+full_paste_url + path)
# Send to every associated email adress
for email in server_term.smembers(TrackedTermsNotificationEmailsPrefix_Name + term):
2018-07-16 15:51:37 +02:00
sendEmailNotification(email, 'Term', mail_body)
def getValueOverRange(word, startDate, num_day):
to_return = 0
for timestamp in range(startDate, startDate - num_day*oneDay, -oneDay):
value = server_term.hget(timestamp, word)
to_return += int(value) if value is not None else 0
return to_return
2014-08-20 15:14:57 +02:00
if __name__ == "__main__":
publisher.port = 6380
2014-08-20 15:14:57 +02:00
publisher.channel = "Script"
2014-08-14 17:55:18 +02:00
config_section = 'Curve'
p = Process(config_section)
2014-08-20 15:14:57 +02:00
# REDIS #
r_serv1 = redis.StrictRedis(
2018-05-07 14:50:40 +02:00
host=p.config.get("ARDB_Curve", "host"),
port=p.config.get("ARDB_Curve", "port"),
db=p.config.get("ARDB_Curve", "db"),
2018-05-04 13:53:29 +02:00
decode_responses=True)
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)
# FUNCTIONS #
publisher.info("Script Curve 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
# FILE CURVE SECTION #
csv_path = os.path.join(os.environ['AIL_HOME'],
p.config.get("Directories", "wordtrending_csv"))
wordfile_path = os.path.join(os.environ['AIL_HOME'],
p.config.get("Directories", "wordsfile"))
message = p.get_from_set()
prec_filename = None
2014-09-05 10:42:01 +02:00
generate_new_graph = False
# Term Frequency
top_termFreq_setName_day = ["TopTermFreq_set_day_", 1]
top_termFreq_setName_week = ["TopTermFreq_set_week", 7]
top_termFreq_setName_month = ["TopTermFreq_set_month", 31]
while True:
2014-08-14 17:55:18 +02:00
if message is not None:
generate_new_graph = True
filename, word, score = message.split()
temp = filename.split('/')
date = temp[-4] + temp[-3] + temp[-2]
timestamp = calendar.timegm((int(temp[-4]), int(temp[-3]), int(temp[-2]), 0, 0, 0))
curr_set = top_termFreq_setName_day[0] + str(timestamp)
low_word = word.lower()
2016-08-22 20:59:56 +02:00
#Old curve with words in file
r_serv1.hincrby(low_word, date, int(score))
# Update redis
#consider the num of occurence of this term
curr_word_value = int(server_term.hincrby(timestamp, low_word, int(score)))
#1 term per paste
curr_word_value_perPaste = int(server_term.hincrby("per_paste_" + str(timestamp), low_word, int(1)))
2016-08-22 20:59:56 +02:00
# Add in set only if term is not in the blacklist
2018-05-04 13:53:29 +02:00
if low_word not in server_term.smembers(BlackListTermsSet_Name):
#consider the num of occurence of this term
server_term.zincrby(curr_set, low_word, float(score))
#1 term per paste
server_term.zincrby("per_paste_" + curr_set, low_word, float(1))
2018-04-16 14:50:04 +02:00
#Add more info for tracked terms
check_if_tracked_term(low_word, filename)
#send to RegexForTermsFrequency
to_send = "{} {} {}".format(filename, timestamp, word)
p.populate_set_out(to_send, 'RegexForTermsFrequency')
else:
2016-08-22 20:59:56 +02:00
if generate_new_graph:
generate_new_graph = False
2018-04-16 14:50:04 +02:00
print('Building graph')
today = datetime.date.today()
year = today.year
month = today.month
2018-04-16 14:50:04 +02:00
lib_words.create_curve_with_word_file(r_serv1, csv_path,
wordfile_path, year,
month)
publisher.debug("Script Curve is Idling")
2018-04-16 14:50:04 +02:00
print("sleeping")
time.sleep(10)
message = p.get_from_set()