From 8b567ce66e8cb8d1ccc75551c67aea4fedc7f9a9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Tomasz=20D=C5=82ugosz?= Date: Sun, 4 May 2014 14:47:55 +0200 Subject: [PATCH] drop too old articles --- recipes/fronda.recipe | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/recipes/fronda.recipe b/recipes/fronda.recipe index f677e38499..43d566bbc4 100644 --- a/recipes/fronda.recipe +++ b/recipes/fronda.recipe @@ -1,7 +1,7 @@ #!/usr/bin/env python __license__ = 'GPL v3' -__copyright__ = u'2010-2013, Tomasz Dlugosz ' +__copyright__ = u'2010-2014, Tomasz Dlugosz ' ''' fronda.pl ''' @@ -71,6 +71,14 @@ class Fronda(BasicNewsRecipe): feeds.append((genName, articles[genName])) return feeds + def preprocess_html(self, soup): + r = soup.find('small') + timestamp = str(r.contents)[3:].split(',')[0] + parts = timestamp.split('.') + art_date = date(int(parts[2]),int(parts[1]),int(parts[0])) + if self.earliest_date < art_date : + return soup + keep_only_tags = [ dict(name='div', attrs={'class':'content content-70 phone-100'}) ]