src/suggest/models.py

   1 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
   2 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
   3 #
   4 import re
   5 from datetime import timedelta
   6
   7 from django.db import models
   8 from django.contrib.auth.models import User
   9 from django.utils.translation import gettext_lazy as _
  10
  11
  12 class Suggestion(models.Model):
  13     contact = models.CharField(_('contact'), blank=True, max_length=120)
  14     description = models.TextField(_('description'), blank=True)
  15     created_at = models.DateTimeField(_('creation date'), auto_now=True)
  16     ip = models.GenericIPAddressField(_('IP address'))
  17     user = models.ForeignKey(User, models.SET_NULL, blank=True, null=True)
  18
  19     class Meta:
  20         ordering = ('-created_at',)
  21         verbose_name = _('suggestion')
  22         verbose_name_plural = _('suggestions')
  23
  24     def __str__(self):
  25         return str(self.created_at)
  26
  27
  28 class PublishingSuggestion(models.Model):
  29     contact = models.CharField(_('contact'), blank=True, max_length=120)
  30     books = models.TextField(_('books'), null=True, blank=True)
  31     audiobooks = models.TextField(_('audiobooks'), null=True, blank=True)
  32     created_at = models.DateTimeField(_('creation date'), auto_now_add=True)
  33     ip = models.GenericIPAddressField(_('IP address'))
  34     user = models.ForeignKey(User, models.SET_NULL, blank=True, null=True)
  35
  36     class Meta:
  37         ordering = ('-created_at',)
  38         verbose_name = _('publishing suggestion')
  39         verbose_name_plural = _('publishing suggestions')
  40
  41     def is_spam(self):
  42         suggestion_text = (self.books or self.audiobooks).strip(' \r\n,')
  43         # similar = PublishingSuggestion.objects.filter(
  44         #     books__in=('', suggestion_text), audiobooks__in=('', suggestion_text))
  45         similar = PublishingSuggestion.objects.filter(books=self.books, audiobooks=self.audiobooks).exclude(pk=self.pk)
  46         http = 'http' in suggestion_text
  47         spam = False
  48         if re.search(r'([^\W\d_])\1\1\1', suggestion_text):
  49             # same letter repetition outside URL
  50             spam = True
  51         elif re.search(r'[^\W\d_]\d|\d[^\W\d_]', suggestion_text) and not http:
  52             # string of letters and digits outside URL
  53             spam = True
  54         elif re.search(r'[^\W\d_]{17}', suggestion_text):
  55             # long string of letters (usually gibberish)
  56             spam = True
  57         elif ' ' not in suggestion_text:
  58             # single word - usually spam
  59             spam = True
  60         elif len(suggestion_text) < 11:
  61             # too short
  62             spam = True
  63         elif similar.filter(created_at__range=(self.created_at - timedelta(1), self.created_at)):
  64             # the same suggestion within 24h
  65             spam = True
  66         elif similar.filter(ip=self.ip):
  67             # the same suggestion from the same IP
  68             spam = True
  69         return spam
  70
  71     def __str__(self):
  72         return str(self.created_at)