#!/usr/bin/env python2 # -*- coding: utf-8 -*- __license__ = 'GPL v3' __copyright__ = u'2011, Silviu Cotoar\u0103' ''' gsp.ro ''' from calibre.web.feeds.news import BasicNewsRecipe class GSP(BasicNewsRecipe): title = u'Gazeta Sporturilor' language = 'ro' __author__ = u'Silviu Cotoar\u0103' description = u'Gazeta Sporturilor' publisher = u'Gazeta Sporturilor' category = 'Ziare,Sport,Stiri,Romania' oldest_article = 5 max_articles_per_feed = 100 no_stylesheets = True use_embedded_content = False encoding = 'utf-8' remove_javascript = True cover_url = 'http://www.gsp.ro/images/logo.jpg' conversion_options = { 'comments': description, 'tags': category, 'language': language, 'publisher': publisher } keep_only_tags = [dict(name='h1', attrs={'class': 'serif title_2'}), dict(name='div', attrs={'id': 'only_text'}), dict(name='span', attrs={'class': 'block poza_principala'}) # noqa ] feeds = [(u'\u0218tiri', u'http://www.gsp.ro/rss.xml')] def preprocess_html(self, soup): return self.adeify_images(soup)