#!/usr/bin/env python __license__ = 'GPL v3' __copyright__ = u'2010-2011, Tomasz Dlugosz ' ''' frazpc.pl ''' from calibre.web.feeds.news import BasicNewsRecipe import re class FrazPC(BasicNewsRecipe): title = u'frazpc.pl' publisher = u'frazpc.pl' description = u'Tw\xf3j Vortal Technologiczny' language = 'pl' __author__ = u'Tomasz D\u0142ugosz' oldest_article = 7 max_articles_per_feed = 100 use_embedded_content = False no_stylesheets = True cover_url='http://www.frazpc.pl/images/logo.png' feeds = [ (u'Aktualno\u015bci', u'http://www.frazpc.pl/feed/aktualnosci'), (u'Artyku\u0142y', u'http://www.frazpc.pl/feed/artykuly') ] keep_only_tags = [dict(name='div', attrs={'class':'article'})] remove_tags = [ dict(name='div', attrs={'class':'title-wrapper'}), dict(name='p', attrs={'class':'tags'}), dict(name='p', attrs={'class':'article-links'}), dict(name='div', attrs={'class':'comments_box'}) ] remove_tags_after=dict(name='div', attrs={'class':'content'}) preprocess_regexps = [(re.compile(r'\| Komentarze \([0-9]*\)'), lambda match: '')] remove_attributes = [ 'width', 'height' ]