from calibre.web.feeds.news import BasicNewsRecipe class AntywebRecipe(BasicNewsRecipe): encoding = 'utf-8' __license__ = 'GPL v3' __author__ = u'Artur Stachecki ' language = 'pl' version = 1 title = u'Antyweb' category = u'News' description = u'Blog o internecie i nowych technologiach' cover_url='' remove_empty_feeds= True auto_cleanup = False no_stylesheets=True use_embedded_content = False oldest_article = 1 max_articles_per_feed = 100 remove_javascript = True simultaneous_downloads = 3 keep_only_tags =[] keep_only_tags.append(dict(name = 'h1', attrs = { 'class' : 'entry-title '})) keep_only_tags.append(dict(name = 'div', attrs = {'class' : 'news-content'})) extra_css = '''body {font-family: verdana, arial, helvetica, geneva, sans-serif ;}''' feeds = [ (u'Artykuly', u'feed://feeds.feedburner.com/Antyweb?format=xml'), ] def preprocess_html(self, soup): for alink in soup.findAll('a'): if alink.string is not None: tstr = alink.string alink.replaceWith(tstr) return soup