Advertisement
Guest User

Untitled

a guest
Sep 19th, 2018
81
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
Python 1.80 KB | None | 0 0
  1. #python2
  2. #-*- coding: utf8 -*-
  3.  
  4. import os
  5.  
  6. class parser:
  7.     def __init__(self):
  8.         self.dico_name = dict()
  9.         self.balise_begin = '<div'
  10.         self.balise_name = 'class="from_name">'
  11.         self.balise_end = '\n       </div>'
  12.  
  13.     def update_dico_name(self, path):
  14.         file = open(path,'r')
  15.         str = file.read()
  16.         file.close()
  17.  
  18.         strtab1= str.split(self.balise_begin)
  19.         strtab2 = [x for x in strtab1 if self.balise_name in x]
  20.         strtab3 = [x.split(self.balise_name)[1] for x in strtab2]
  21.         str_name = [x.split(self.balise_end)[0] for x in strtab3]
  22.  
  23.         for str in str_name:
  24.             if str in self.dico_name.keys():
  25.                 self.dico_name[str] = self.dico_name[str]+1
  26.             else:
  27.                 self.dico_name[str] = 0
  28.  
  29.     def parse_folder_for_name(self):
  30.         for file in os.listdir('.'):
  31.             if ('.html' in file):
  32.                 self.update_dico_name(file)
  33.  
  34.     def output_result(self):
  35.         dico_simple = dict()
  36.         dico_gif = dict()
  37.  
  38.         for k,v in self.dico_name.items():
  39.             if v!=0:
  40.                 if 'via @gif' in k:
  41.                     str_key = k.split('via @gif')[0]
  42.                     dico_gif[str_key] = v
  43.                 elif 'via @bold' not in k :
  44.                     dico_simple[k] = v
  45.  
  46.         file = open('output_result.txt','w')
  47.         file.write('GIF postés :\n')
  48.  
  49.         for k,v in dico_gif.items() :
  50.             file.write( k[1:-1] + ' : ' + str(v) +' Occurences\n')
  51.  
  52.         file.write('\nMessages normaux :\n')
  53.         for k,v in dico_simple.items() :
  54.             file.write( k[1:] + ' : ' + str(v) +' Occurences\n')
  55.  
  56.         file.close()
  57.  
  58.  
  59. if __name__ == "__main__":
  60.     coucou = parser()
  61.     coucou.parse_folder_for_name()
  62.     coucou.output_result()
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement