Hi all,
This is my first post and i'd like to contribute my recipe for Australia Daily Telegraph. It's a cleaner than the original version, hopefully you guys enjoy it
Code:
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal kovid@kovidgoyal.net'
__docformat__ = 'restructuredtext en'
'''
http://www.news.com.au/dailytelegraph/
'''
import re
from calibre.web.feeds.news import BasicNewsRecipe
class DailyTelegraph(BasicNewsRecipe):
title = u'Daily Telegraph'
__author__ = u'Adrian G.'
language = 'en_AU'
description = u'Daily Telegraph News'
oldest_article = 5
max_articles_per_feed = 100
no_stylesheets = True
use_embedded_content = False
no_javascript = True
timefmt = ' [%A, %d %B, %Y]'
encoding = 'utf-8'
keep_only_tags = [dict(name='div', attrs ={'id':'story'})]
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 ={'id':['comments','story-related-coverage']}),
dict(name='div', attrs ={'class':['story-header-tools','story-footer','story-extras','story-related']}),
dict(name='div', attrs ={'class':['promo-image','story-extras story-extras-2']}),
dict(name='div', attrs ={'class':['assistive sidebar-jump']})
]
feeds = [
(u'Top Stories', u'http://feeds.news.com.au/public/rss/2.0/dtele_top_stories_253.xml'),
(u'National News', u'http://feeds.news.com.au/public/rss/2.0/dtele_national_news_202.xml'),
(u'World News', u'http://feeds.news.com.au/public/rss/2.0/dtele_world_news_204.xml'),
(u'NSW and ACT', u'http://feeds.news.com.au/public/rss/2.0/dtele_nswact_225.xml'),
(u'Arts', u'http://feeds.news.com.au/public/rss/2.0/dtele_art_444.xml'),
(u'Business News', u'http://feeds.news.com.au/public/rss/2.0/dtele_business_226.xml'),
(u'Entertainment News', u'http://feeds.news.com.au/public/rss/2.0/dtele_entertainment_news_201.xml'),
(u'Lifestyle News', u'http://feeds.news.com.au/public/rss/2.0/dtele_lifestyle_227.xml'),
(u'Music', u'http://feeds.news.com.au/public/rss/2.0/dtele_music_441.xml'),
(u'Property Confidential', u'http://feeds.news.com.au/public/rss/2.0/dtele_property_confidential_463.xml'),
(u'Property - Your Space', u'http://feeds.news.com.au/public/rss/2.0/dtele_property_yourspace_462.xml'),
(u'Confidential News', u'http://feeds.news.com.au/public/rss/2.0/dtele_entertainment_confidential_252.xml'),
(u'Confidential Biographies', u'http://feeds.news.com.au/public/rss/2.0/dtele_confidential_biographies_491.xml'),
(u'Confidential Galleries', u'http://feeds.news.com.au/public/rss/2.0/dtele_confidential_galleries_483.xml'),
]