diff --git a/KnowYourMemesParser.py b/KnowYourMemesParser.py
old mode 100755
new mode 100644
index 0dc19b2..650fa7e
--- a/KnowYourMemesParser.py
+++ b/KnowYourMemesParser.py
@@ -1,234 +1,230 @@
-# -*- coding: utf-8 -*-
-"""
-Created on Tue Sep 6 17:40:12 2016
-@author: dmitrys
-"""
+# coding: utf-8
-###############################################################
-#### Meme parsing just for lulz and science <3 ####
-###############################################################
+# authors: @DmitrySerg, @FUlyankin
+# repository: https://github.com/DmitrySerg/memology
+import requests # отправка запросов
+import numpy as np # матрицы, вектора и линал
+import pandas as pd # таблички и операции с ними
+import time # время
-import re
-from bs4 import BeautifulSoup
-import time
-import pandas as pd
-import numpy as np
-from urllib.request import Request, urlopen
-import getpass
-import sys
-sys.path.append('/Users/dmitrys/anaconda2/lib/python2.7/site-packages')
-username = getpass.getuser()
-
-#### Send me a letter
-import smtplib
-from email.mime.multipart import MIMEMultipart
-from email.mime.text import MIMEText
-#from email.mime.base import MIMEBase
-#from email import encoders
-import requests
-import sys
+from tqdm import tqdm # мониторинг прогресса
+from fake_useragent import UserAgent # генерация правдоподобных юзер-агентов
+from bs4 import BeautifulSoup # очень красивый суп для обработки html
+import argparse # чтение аргументов из коммандной строки
-#### GOING TO TOR
-import socks
+import socks # подключение к тору
import socket
-socks.set_default_proxy(socks.SOCKS5, "localhost", 9150)
-socket.socket = socks.socksocket
-#print(urlopen('http://icanhazip.com').read())
-
-
-from fake_useragent import UserAgent
-
-def generateUserAgent():
- return UserAgent().chrome
-
-def html_stripper(text):
- return re.sub('<[^<]+?>', '', str(text))
-
-
-number_of_pages = 369
-page = 1
-main_url = 'http://knowyourmeme.com/'
-columns = ['name', 'added', 'views', 'comments', 'status', 'year', 'tags', 'about', 'origin', 'spread']
-FINAL = pd.DataFrame(columns=columns)
-START = time.time()
-
-def getMemeUrls(page):
- req = Request('http://knowyourmeme.com/memes/all/page/{}'.format(page), headers={'User-Agent': generateUserAgent()})
- webpage = urlopen(req).read()
- soup = BeautifulSoup(webpage, "lxml")
- meme_urls = soup.findAll('a', attrs={'class':'photo'})
- print('Getting all memes from page {}'.format(page))
- return meme_urls
-
-
+import sys
-def getAllFromPage(meme_urls):
- global FINAL
- count = 0
- start = time.time()
- current_shape = FINAL.shape
- for meme in meme_urls:
- count += 1
+def getPageLinks(page_number):
+ """
+ Возвращает список ссылок на мемы, полученный с текущей страницы
- to_append = {x:np.NaN for x in columns}
- #time.sleep(1)
- try:
- meme_url = re.split('href="|" target="|">