__license__ = 'GPL v3' __copyright__ = '2010-2011, Darko Miletic ' ''' www.la-razon.com ''' from calibre.web.feeds.news import BasicNewsRecipe class LaRazon_Bol(BasicNewsRecipe): title = u'La Razón - Bolivia' __author__ = 'Darko Miletic' description = 'El diario nacional de Bolivia' publisher = 'Praxsis S.R.L.' category = 'news, politics, Bolivia' oldest_article = 1 max_articles_per_feed = 200 no_stylesheets = True encoding = 'utf8' use_embedded_content = False language = 'es_BO' publication_type = 'newspaper' remove_empty_feeds = True masthead_url = 'http://www.la-razon.com/static/LRZRazon/images/lrz-logo.png' extra_css = """ body{font-family: Georgia,"Times New Roman",Times,serif} img{margin-bottom: 0.4em; display: block} .meta{font-size: small; font-family: Arial,Helvetica,sans-serif} """ INDEX = 'http://www.la-razon.com/' conversion_options = { 'comment': description, 'tags': category, 'publisher': publisher, 'language': language } keep_only_tags = [dict(name='div', attrs={'class': ['pg-hd', 'pg-bd']})] remove_tags = [ dict(name=['meta', 'link', 'form', 'iframe', 'embed', 'object']), dict( name='div', attrs={'class': 'bd'}) ] remove_attributes = ['width', 'height'] feeds = [ (u'Editorial', u'http://www.la-razon.com/rss/opinion/editorial/'), (u'Nacional', u'http://www.la-razon.com/rss/nacional/'), (u'Economia', u'http://www.la-razon.com/rss/economia/'), (u'Ciudades', u'http://www.la-razon.com/rss/ciudades/'), (u'Sociedad', u'http://www.la-razon.com/rss/sociedad/'), (u'Mundo', u'http://www.la-razon.com/rss/mundo/'), (u'La Revista', u'http://www.la-razon.com/rss/la_revista/'), (u'Sociales', u'http://www.la-razon.com/rss/sociales/'), (u'Mia', u'http://www.la-razon.com/rss/suplementos/mia/'), (u'Marcas', u'http://www.la-razon.com/rss/marcas/'), (u'Escape', u'http://www.la-razon.com/rss/suplementos/escape/'), (u'El Financiero', u'http://www.la-razon.com/rss/suplementos/financiero/'), (u'Tendencias', u'http://www.la-razon.com/rss/suplementos/tendencias/') ] def preprocess_html(self, soup): for item in soup.findAll(style=True): del item['style'] return soup def get_cover_url(self): soup = self.index_to_soup(self.INDEX) lightbox = soup.find( 'div', attrs={'class': 'lightbox lightbox-frontpage'}) return lightbox.img['src']