#!/usr/bin/env python __license__ = 'GPL v3' __copyright__ = '2011, Adrian Gunawan ' __author__ = 'Adrian Gunawan' __version__ = 'v1.0' __date__ = '02 February 2011' ''' http://www.thejakartapost.com/ ''' from calibre.web.feeds.news import BasicNewsRecipe class JakartaPost(BasicNewsRecipe): title = u'Jakarta Post' masthead_url = 'http://www.thejakartapost.com/images/jakartapost_logo.jpg' cover_url = 'http://www.thejakartapost.com/images/jakartapost_logo.jpg' __author__ = u'Adrian Gunawan' description = u'Indonesian Newspaper in English from Jakarta Post Online Edition' category = 'breaking news, national, business, international, Indonesia' language = 'en_ID' oldest_article = 2 max_articles_per_feed = 100 no_stylesheets = True use_embedded_content = False no_javascript = True remove_empty_feeds = True timefmt = ' [%A, %d %B, %Y]' encoding = 'utf-8' keep_only_tags = [dict(name='div', attrs ={'id':'news-main'})] extra_css = ''' h1{font-family:Georgia,"Times New Roman",Times,serif; font-weight:bold; font-size:large;} .cT-storyDetails{font-family:Arial,Helvetica,sans-serif; color:#666666;font-size:x-small;} .articleBody{font-family:Arial,Helvetica,sans-serif; color:black;font-size:small;} .cT-imageLandscape{font-family:Arial,Helvetica,sans-serif; color:#333333 ;font-size:x-small;} .source{font-family:Arial,Helvetica,sans-serif; color:#333333 ;font-size:xx-small;} #content{font-family:Arial,Helvetica,sans-serif;font-size:x-small;} .pageprint{font-family:Arial,Helvetica,sans-serif;font-size:small;} #bylineDetails{font-family:Arial,Helvetica,sans-serif; color:#666666;font-size:x-small;} .featurePic-wide{font-family:Arial,Helvetica,sans-serif;font-size:x-small;} #idfeaturepic{font-family:Arial,Helvetica,sans-serif;font-size:x-small;} h3{font-family:Georgia,"Times New Roman",Times,serif; font-size:small;} h2{font-family:Georgia,"Times New Roman",Times,serif; font-size:small;} h4{font-family:Georgia,"Times New Roman",Times,serif; font-size:small;} h5{font-family:Georgia,"Times New Roman",Times,serif; font-size:small;} body{font-family:Arial,Helvetica,sans-serif; font-size:x-small;} ''' remove_tags = [ dict(name='div', attrs ={'class':['text-size']}), ] feeds = [ (u'Breaking News', u'http://www.thejakartapost.com/breaking/feed'), (u'National', u'http://www.thejakartapost.com/channel/national/feed'), (u'Archipelago', u'http://www.thejakartapost.com/channel/archipelago/feed'), (u'Business', u'http://www.thejakartapost.com/channel/business/feed'), (u'Jakarta', u'http://www.thejakartapost.com/channel/jakarta/feed'), (u'World', u'http://www.thejakartapost.com/channel/world/feed'), (u'Sports', u'http://www.thejakartapost.com/channel/sports/feed'), ]