#!/usr/bin/env python2 __license__ = 'GPL v3' __author__ = '2010, Gustavo Azambuja ' ''' Muy Interesante ''' from calibre.web.feeds.news import BasicNewsRecipe class General(BasicNewsRecipe): title = 'Cosmopolitan' __author__ = 'Gustavo Azambuja' description = 'Revista Cosmopolitan, Edicion Espanola' language = 'es' timefmt = '[%a, %d %b, %Y]' use_embedded_content = False recursion = 1 encoding = 'utf8' remove_javascript = True no_stylesheets = True conversion_options = {'linearize_tables': True} oldest_article = 180 max_articles_per_feed = 100 keep_only_tags = [ dict(id=['contenido']), dict(name='td', attrs={'class': ['contentheading', 'txt_articulo']}) ] remove_tags = [ dict(name='div', attrs={'class': ['breadcrumb', 'bloque1', 'article', 'bajo_title', 'tags_articles', 'otrosenlaces_title', 'otrosenlaces_parent', 'compartir']}), dict(name='div', attrs={'id': 'comment'}), dict(name='table', attrs={'class': 'pagenav'}), dict(name=['object', 'link']) ] remove_attributes = ['width', 'height', 'style', 'font', 'color'] extra_css = ''' h1{font-family:Geneva, Arial, Helvetica, sans-serif;color:#154B7A;} h3{font-size: 14px;color:#999999; font-family:Geneva, Arial, Helvetica, sans-serif;font-weight: bold;} h2{color:#666666; font-family:Geneva, Arial, Helvetica, sans-serif;font-size:small;} img {float:left; clear:both; margin:10px} p {font-family:Arial,Helvetica,sans-serif;} ''' feeds = [ (u'Articulos', u'http://feeds.feedburner.com/cosmohispano/FSSt') ] def preprocess_html(self, soup): attribs = ['style', 'font', 'valign', 'colspan', 'width', 'height', 'rowspan', 'summary', 'align', 'cellspacing', 'cellpadding', 'frames', 'rules', 'border' ] # noqa for item in soup.body.findAll(name=['table', 'td', 'tr', 'th', 'caption', 'thead', 'tfoot', 'tbody', 'colgroup', 'col']): item.name = 'div' for attrib in attribs: item[attrib] = '' del item[attrib] return soup def get_cover_url(self): index = 'http://www.cosmohispano.com/revista' soup = self.index_to_soup(index) link_item = soup.find('img', attrs={'class': 'img_portada'}) if link_item: cover_url = "http://www.cosmohispano.com" + link_item['src'] return cover_url