Merge pull request #115 from mokaddem/regexInTerms

New features for terms frequency

all_modules.txt

@@ -1,26 +0,0 @@
-Global
-Duplicates
-Indexer
-Attributes
-Lines
-DomClassifier
-Tokenize
-Curve
-CurveManageTopSets
-Categ
-CreditCards
-Mail
-Onion
-DumpValidOnion
-Web
-WebStats
-SQLInjectionDetection
-ModuleStats
-Browse_warning_paste
-SentimentAnalysis
-Release
-Credential
-Cve
-Phone
-SourceCode
-Keys

bin/Curve.py

@@ -132,6 +132,10 @@ if __name__ == "__main__":
             #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:
 
             if generate_new_graph:

bin/LAUNCH.sh

@@ -157,6 +157,10 @@ function launching_scripts {
     sleep 0.1
     screen -S "Script" -X screen -t "CurveManageTopSets" bash -c './CurveManageTopSets.py; read x'
     sleep 0.1
+    screen -S "Script" -X screen -t "RegexForTermsFrequency" bash -c './RegexForTermsFrequency.py; read x'
+    sleep 0.1
+    screen -S "Script" -X screen -t "SetForTermsFrequency" bash -c './SetForTermsFrequency.py; read x'
+    sleep 0.1
     screen -S "Script" -X screen -t "Indexer" bash -c './Indexer.py; read x'
     sleep 0.1
     screen -S "Script" -X screen -t "Keys" bash -c './Keys.py; read x'

bin/RegexForTermsFrequency.py

@@ -0,0 +1,107 @@
+#!/usr/bin/env python2
+# -*-coding:UTF-8 -*
+"""
+This Module is used for term frequency.
+
+"""
+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
+
+from Helper import Process
+
+# Config Variables
+DICO_REFRESH_TIME = 60 #s
+
+BlackListTermsSet_Name = "BlackListSetTermSet"
+TrackedTermsSet_Name = "TrackedSetTermSet"
+TrackedRegexSet_Name = "TrackedRegexSet"
+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]
+
+
+def refresh_dicos():
+    dico_regex = {} 
+    dico_regexname_to_redis = {}
+    for regex_str in server_term.smembers(TrackedRegexSet_Name):
+        dico_regex[regex_str[1:-1]] = re.compile(regex_str[1:-1])
+        dico_regexname_to_redis[regex_str[1:-1]] = regex_str
+
+    return dico_regex, dico_regexname_to_redis
+
+if __name__ == "__main__":
+    publisher.port = 6380
+    publisher.channel = "Script"
+
+    config_section = 'RegexForTermsFrequency'
+    p = Process(config_section)
+
+    # REDIS #
+    server_term = redis.StrictRedis(
+        host=p.config.get("Redis_Level_DB_TermFreq", "host"),
+        port=p.config.get("Redis_Level_DB_TermFreq", "port"),
+        db=p.config.get("Redis_Level_DB_TermFreq", "db"))
+
+    # FUNCTIONS #
+    publisher.info("RegexForTermsFrequency script started")
+
+    #compile the regex
+    dico_refresh_cooldown = time.time()
+    dico_regex, dico_regexname_to_redis = refresh_dicos()
+
+    message = p.get_from_set()
+
+    # Regex Frequency
+    while True:
+
+        if message is not None:
+            if time.time() - dico_refresh_cooldown > DICO_REFRESH_TIME:
+                dico_refresh_cooldown = time.time()
+                dico_regex, dico_regexname_to_redis = refresh_dicos()
+                print('dico got refreshed')
+
+            filename = message
+            temp = filename.split('/')
+            timestamp = calendar.timegm((int(temp[-4]), int(temp[-3]), int(temp[-2]), 0, 0, 0))
+
+            curr_set = top_termFreq_setName_day[0] + str(timestamp)
+            content = Paste.Paste(filename).get_p_content()
+
+            #iterate the word with the regex
+            for regex_str, compiled_regex in dico_regex.items():
+                matched = compiled_regex.search(content)
+
+                if matched is not None: #there is a match
+                    print('regex matched {}'.format(regex_str))
+                    matched = matched.group(0)
+                    # Add in Regex track set only if term is not in the blacklist
+                    if matched not in server_term.smembers(BlackListTermsSet_Name):
+                        set_name = 'regex_' + dico_regexname_to_redis[regex_str]
+                        new_to_the_set = server_term.sadd(set_name, filename)
+                        new_to_the_set = True if new_to_the_set == 1 else False
+
+                        #consider the num of occurence of this term
+                        regex_value = int(server_term.hincrby(timestamp, dico_regexname_to_redis[regex_str], int(1)))
+                        #1 term per paste
+                        if new_to_the_set:
+                            regex_value_perPaste = int(server_term.hincrby("per_paste_" + str(timestamp), dico_regexname_to_redis[regex_str], int(1)))
+                            server_term.zincrby("per_paste_" + curr_set, dico_regexname_to_redis[regex_str], float(1))
+                    server_term.zincrby(curr_set, dico_regexname_to_redis[regex_str], float(1))
+                else:
+                    pass
+
+        else:
+            publisher.debug("Script RegexForTermsFrequency is Idling")
+            print "sleeping"
+            time.sleep(5)
+        message = p.get_from_set()

bin/SetForTermsFrequency.py

@@ -0,0 +1,122 @@
+#!/usr/bin/env python2
+# -*-coding:UTF-8 -*
+"""
+This Module is used for term frequency.
+
+"""
+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
+
+# Config Variables
+BlackListTermsSet_Name = "BlackListSetTermSet"
+TrackedTermsSet_Name = "TrackedSetTermSet"
+TrackedRegexSet_Name = "TrackedRegexSet"
+TrackedSetSet_Name = "TrackedSetSet"
+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]
+
+def add_quote_inside_tab(tab):
+    quoted_tab = "["
+    for elem in tab[1:-1].split(','):
+        elem = elem.lstrip().strip()
+        quoted_tab += "\'{}\', ".format(elem)
+    quoted_tab = quoted_tab[:-2] #remove trailing ,
+    quoted_tab += "]"
+    return str(quoted_tab)
+
+if __name__ == "__main__":
+    publisher.port = 6380
+    publisher.channel = "Script"
+
+    config_section = 'SetForTermsFrequency'
+    p = Process(config_section)
+
+    # REDIS #
+    server_term = redis.StrictRedis(
+        host=p.config.get("Redis_Level_DB_TermFreq", "host"),
+        port=p.config.get("Redis_Level_DB_TermFreq", "port"),
+        db=p.config.get("Redis_Level_DB_TermFreq", "db"))
+
+    # FUNCTIONS #
+    publisher.info("RegexForTermsFrequency script started")
+
+    #get the dico and matching percent
+    dico_percent = {}
+    dico_set_tab = {}
+    dico_setname_to_redis = {}
+    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]
+            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
+        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 = {}
+            for word in content.split():
+                for cur_set, array_set in dico_set_tab.items():
+                    for w_set in array_set[:-1]: #avoid the percent matching
+                        if word == w_set:
+                            try:
+                                match_dico[str(array_set)] += 1
+                            except KeyError:
+                                match_dico[str(array_set)] = 1
+
+            #compute matching %
+            for the_set, matchingNum in match_dico.items():
+                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]:
+                    print(the_set, "matched in", filename)
+                    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
+                    
+
+                    #consider the num of occurence of this set
+                    set_value = int(server_term.hincrby(timestamp, dico_setname_to_redis[the_set], int(1)))
+
+                    # FIXME - avoid using per paste as a set is checked over the entire paste
+                    #1 term per paste
+                    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))
+
+
+        else:
+            publisher.debug("Script RegexForTermsFrequency is Idling")
+            print "sleeping"
+            time.sleep(5)
+        message = p.get_from_set()

bin/packages/modules.cfg

@@ -34,6 +34,12 @@ publish = Redis_Words
 subscribe = Redis_Words
 publish = Redis_CurveManageTopSets
 
+[RegexForTermsFrequency]
+subscribe = Redis_Global
+
+[SetForTermsFrequency]
+subscribe = Redis_Global
+
 [CurveManageTopSets]
 subscribe = Redis_CurveManageTopSets
 

var/www/Flasks/Flask_terms.py

@@ -9,7 +9,7 @@ import datetime
 import calendar
 import flask
 from flask import Flask, render_template, jsonify, request
-
+import re
 import Paste
 
 # ============ VARIABLES ============
@@ -18,6 +18,22 @@ import Flask_config
 app = Flask_config.app
 cfg = Flask_config.cfg
 r_serv_term = Flask_config.r_serv_term
+
+DEFAULT_MATCH_PERCENT = 50
+
+#tracked
+TrackedTermsSet_Name = "TrackedSetTermSet"
+TrackedTermsDate_Name = "TrackedTermDate"
+#black
+BlackListTermsDate_Name = "BlackListTermDate"
+BlackListTermsSet_Name = "BlackListSetTermSet"
+#regex
+TrackedRegexSet_Name = "TrackedRegexSet"
+TrackedRegexDate_Name = "TrackedRegexDate"
+#set
+TrackedSetSet_Name = "TrackedSetSet"
+TrackedSetDate_Name = "TrackedSetDate"
+
 # ============ FUNCTIONS ============
 
 def Term_getValueOverRange(word, startDate, num_day, per_paste=""):
@@ -47,15 +63,43 @@ def terms_management():
         per_paste_text = ""
         per_paste = 0
 
-    TrackedTermsSet_Name = "TrackedSetTermSet"
-    BlackListTermsSet_Name = "BlackListSetTermSet"
-    TrackedTermsDate_Name = "TrackedTermDate"
-    BlackListTermsDate_Name = "BlackListTermDate"
-    
     today = datetime.datetime.now()
     today = today.replace(hour=0, minute=0, second=0, microsecond=0)
     today_timestamp = calendar.timegm(today.timetuple())
 
+    #Regex
+    trackReg_list = []
+    trackReg_list_values = []
+    trackReg_list_num_of_paste = []
+    for tracked_regex in r_serv_term.smembers(TrackedRegexSet_Name):
+        trackReg_list.append(tracked_regex)
+        value_range = Term_getValueOverRange(tracked_regex, today_timestamp, [1, 7, 31], per_paste=per_paste_text)
+
+        term_date = r_serv_term.hget(TrackedRegexDate_Name, tracked_regex)
+
+        set_paste_name = "regex_" + tracked_regex
+        trackReg_list_num_of_paste.append(r_serv_term.scard(set_paste_name))
+        term_date = datetime.datetime.utcfromtimestamp(int(term_date)) if term_date is not None else "No date recorded"
+        value_range.append(term_date)
+        trackReg_list_values.append(value_range)
+
+    #Set
+    trackSet_list = []
+    trackSet_list_values = []
+    trackSet_list_num_of_paste = []
+    for tracked_set in r_serv_term.smembers(TrackedSetSet_Name):
+        trackSet_list.append(tracked_set)
+        value_range = Term_getValueOverRange(tracked_set, today_timestamp, [1, 7, 31], per_paste=per_paste_text)
+
+        term_date = r_serv_term.hget(TrackedSetDate_Name, tracked_set)
+
+        set_paste_name = "set_" + tracked_set
+        trackSet_list_num_of_paste.append(r_serv_term.scard(set_paste_name))
+        term_date = datetime.datetime.utcfromtimestamp(int(term_date)) if term_date is not None else "No date recorded"
+        value_range.append(term_date)
+        trackSet_list_values.append(value_range)
+
+    #Tracked terms
     track_list = []
     track_list_values = []
     track_list_num_of_paste = []
@@ -72,23 +116,36 @@ def terms_management():
         track_list_values.append(value_range)
 
 
+    #blacklist terms
     black_list = []
     for blacked_term in r_serv_term.smembers(BlackListTermsSet_Name):
         term_date = r_serv_term.hget(BlackListTermsDate_Name, blacked_term)
         term_date = datetime.datetime.utcfromtimestamp(int(term_date)) if term_date is not None else "No date recorded"
         black_list.append([blacked_term, term_date])
 
-    return render_template("terms_management.html", black_list=black_list, track_list=track_list, track_list_values=track_list_values,  track_list_num_of_paste=track_list_num_of_paste, per_paste=per_paste)
+    return render_template("terms_management.html", 
+            black_list=black_list, track_list=track_list, trackReg_list=trackReg_list, trackSet_list=trackSet_list,
+            track_list_values=track_list_values, track_list_num_of_paste=track_list_num_of_paste, 
+            trackReg_list_values=trackReg_list_values, trackReg_list_num_of_paste=trackReg_list_num_of_paste,
+            trackSet_list_values=trackSet_list_values, trackSet_list_num_of_paste=trackSet_list_num_of_paste,
+            per_paste=per_paste)
 
 
 @app.route("/terms_management_query_paste/")
 def terms_management_query_paste():
     term =  request.args.get('term')
-    TrackedTermsSet_Name = "TrackedSetTermSet"
     paste_info = []
 
-    set_paste_name = "tracked_" + term
-    track_list_path = r_serv_term.smembers(set_paste_name)
+    # check if regex or not
+    if term.startswith('/') and term.endswith('/'):
+        set_paste_name = "regex_" + term
+        track_list_path = r_serv_term.smembers(set_paste_name)
+    elif term.startswith('\\') and term.endswith('\\'):
+        set_paste_name = "set_" + term
+        track_list_path = r_serv_term.smembers(set_paste_name)
+    else:
+        set_paste_name = "tracked_" + term
+        track_list_path = r_serv_term.smembers(set_paste_name)
 
     for path in track_list_path:
         paste = Paste.Paste(path)
@@ -131,11 +188,6 @@ def terms_management_query():
 
 @app.route("/terms_management_action/", methods=['GET'])
 def terms_management_action():
-    TrackedTermsSet_Name = "TrackedSetTermSet"
-    TrackedTermsDate_Name = "TrackedTermDate"
-    BlackListTermsDate_Name = "BlackListTermDate"
-    BlackListTermsSet_Name = "BlackListSetTermSet"
-
     today = datetime.datetime.now()
     today = today.replace(microsecond=0)
     today_timestamp = calendar.timegm(today.timetuple())
@@ -149,10 +201,42 @@ def terms_management_action():
     else:
         if section == "followTerm":
             if action == "add":
-                r_serv_term.sadd(TrackedTermsSet_Name, term.lower())
-                r_serv_term.hset(TrackedTermsDate_Name, term, today_timestamp)
+                # check if regex/set or simple term
+                #regex
+                if term.startswith('/') and term.endswith('/'):
+                    r_serv_term.sadd(TrackedRegexSet_Name, term)
+                    r_serv_term.hset(TrackedRegexDate_Name, term, today_timestamp)
+
+                #set
+                elif term.startswith('\\') and term.endswith('\\'):
+                    tab_term = term[1:-1]
+                    perc_finder = re.compile("\[[0-9]{1,3}\]").search(tab_term)
+                    if perc_finder is not None:
+                        match_percent = perc_finder.group(0)[1:-1]
+                        set_to_add = term
+                    else:
+                        match_percent = DEFAULT_MATCH_PERCENT
+                        set_to_add = "\\" + tab_term[:-1] + ", [{}]]\\".format(match_percent)
+                    r_serv_term.sadd(TrackedSetSet_Name, set_to_add)
+                    r_serv_term.hset(TrackedSetDate_Name, set_to_add, today_timestamp)
+
+                #simple term
+                else:
+                    r_serv_term.sadd(TrackedTermsSet_Name, term.lower())
+                    r_serv_term.hset(TrackedTermsDate_Name, term.lower(), today_timestamp)
+            #del action
             else:
-                r_serv_term.srem(TrackedTermsSet_Name, term.lower())
+                if term.startswith('/') and term.endswith('/'):
+                    r_serv_term.srem(TrackedRegexSet_Name, term)
+                    r_serv_term.hdel(TrackedRegexDate_Name, term)
+                elif term.startswith('\\') and term.endswith('\\'):
+                    r_serv_term.srem(TrackedSetSet_Name, term)
+                    print(term)
+                    r_serv_term.hdel(TrackedSetDate_Name, term)
+                else:
+                    r_serv_term.srem(TrackedTermsSet_Name, term.lower())
+                    r_serv_term.hdel(TrackedTermsDate_Name, term.lower())
+
         elif section == "blacklistTerm":
             if action == "add":
                 r_serv_term.sadd(BlackListTermsSet_Name, term.lower())

var/www/templates/terms_management.html

@@ -105,9 +105,16 @@
                         </div>
                         <div class="panel-body">
 
+                            <div style="margin-bottom: 10px;"> 
+                                <table>
+                                    <tr><td><b>Regex</b>: surround the term by '<b>/</b>'. </td> <td><b style="margin-left: 20px;">/([a-z])\w+([a-z])\n/</b></td></tr>
+                                    <tr><td><b>Set of terms</b>: surround the list by '<b>\</b>'. </td> <td><b style="margin-left: 20px;">\[term1, term2, ...]\</b></td></tr>
+                                    <tr><td> - To set a custom matching <b>threshold</b> (defaut=50), append it at the end as a inner list '<b>[thresh]</b>'. </td> <td><b style="margin-left: 20px;">\[term1, term2, ..., [75]]\</b></td></tr>
+                                </table>
+                            </div>
                             <div class="form-group input-group" style="margin-bottom: 30px;">
                                 <span class="input-group-addon"><span class="fa fa-eye"></span></span>
-                                <input id="followTermInput" class="form-control" placeholder="Term to track" type="text" style="max-width: 400px;">
+                                <input id="followTermInput" class="form-control" placeholder="Term to track." type="text" style="max-width: 400px;">
                                 <button id="followTermBtn" class="btn btn-success btn-interaction" style="margin-left: 10px;" data-section="followTerm" data-action="add"> Add term</button>
                             </div>
 
@@ -124,6 +131,37 @@
                                      </tr>
                                 </thead>
                                 <tbody>
+                                <!-- SET -->
+                                {% for set in trackSet_list %}
+                                  <tr style="background-color: #cdffca;">
+                                      <td>{{ set }}</td>
+                                      <td>{{ trackSet_list_values[loop.index0][3] }}</td>
+                                      <td>{{ trackSet_list_values[loop.index0][0] }}</td>
+                                      <td>{{ trackSet_list_values[loop.index0][1] }}</td>
+                                      <td>{{ trackSet_list_values[loop.index0][2] }}</td>
+                                      <td>{{ trackSet_list_num_of_paste[loop.index0] }}</td>
+                                      <td><p style="margin: 0px;">
+                                          <span data-toggle="modal" data-target="#mymodal" data-term="{{ set }}" ><button class="btn-link" data-toggle="tooltip" data-placement="right" title="Show concerned paste(s)"><span class="glyphicon glyphicon-info-sign"></span></button></span>
+                                          <button class="btn-link btn-interaction" data-toggle="tooltip" data-placement="left" title="Remove this term" data-content="{{ set }}" data-section="followTerm" data-action="delete"><span class="glyphicon glyphicon-trash"></span></button>
+                                      </p></td>
+                                  </tr>
+                                {% endfor %}
+                                <!-- REGEX -->
+                                {% for regex in trackReg_list %}
+                                  <tr style="background-color: #fffdca;">
+                                      <td>{{ regex }}</td>
+                                      <td>{{ trackReg_list_values[loop.index0][3] }}</td>
+                                      <td>{{ trackReg_list_values[loop.index0][0] }}</td>
+                                      <td>{{ trackReg_list_values[loop.index0][1] }}</td>
+                                      <td>{{ trackReg_list_values[loop.index0][2] }}</td>
+                                      <td>{{ trackReg_list_num_of_paste[loop.index0] }}</td>
+                                      <td><p style="margin: 0px;">
+                                          <span data-toggle="modal" data-target="#mymodal" data-term="{{ regex }}" ><button class="btn-link" data-toggle="tooltip" data-placement="right" title="Show concerned paste(s)"><span class="glyphicon glyphicon-info-sign"></span></button></span>
+                                          <button class="btn-link btn-interaction" data-toggle="tooltip" data-placement="left" title="Remove this term" data-content="{{ regex }}" data-section="followTerm" data-action="delete"><span class="glyphicon glyphicon-trash"></span></button>
+                                      </p></td>
+                                  </tr>
+                                {% endfor %}
+                                <!-- Normal term -->
                                 {% for term in track_list %}
                                   <tr>
                                       <td>{{ term }}</td>
@@ -251,7 +289,7 @@
             //console.log(data); 
             event.preventDefault();
             var the_modal=$(this);
-            var url = "{{ url_for('terms_management_query_paste')  }}?term=" + $(this).attr('data-term');
+            var url = "{{ url_for('terms_management_query_paste')  }}?term=" + encodeURIComponent($(this).attr('data-term'));
             $.getJSON(url, function (data) {
                 if (data.length != 0) {
                     var html_to_add = "";
@@ -318,9 +356,9 @@ function perform_operation(){
     var curr_section = $(this).attr('data-section');
     var curr_action = $(this).attr('data-action'); 
     if (curr_action == "add") {
-        var curr_term = $('#'+curr_section+'Input').val().toLowerCase();;
+        var curr_term = $('#'+curr_section+'Input').val();
     } else {
-        var curr_term = $(this).attr('data-content').toLowerCase();;
+        var curr_term = $(this).attr('data-content');
     }
     var data_to_send = { section: curr_section, action:curr_action, term: curr_term};
 
@@ -341,6 +379,7 @@ function perform_operation(){
                    } else if (json.action == "delete") {
                        // Find indexes of row which have the term in the first column
                        var index = table_track.rows().eq( 0 ).filter( function (rowIdx) {
+                           console.log(table_track.cell( rowIdx, 0 ).data())
                            return table_track.cell( rowIdx, 0 ).data() === json.term;
                        } );
                        table_track.rows(index).remove().draw( false );