Here is a recipe to collect articles from
The Business Spectator, an Australian business news and commentary web site.
Code:
__license__ = 'GPL v3'
__copyright__ = '2010, Dean Cording'
'''
abc.net.au/news
'''
import re
from calibre.web.feeds.recipes import BasicNewsRecipe
class BusinessSpectator(BasicNewsRecipe):
title = 'Business Spectator'
__author__ = 'Dean Cording'
description = 'Australian Business News & commentary delivered the way you want it.'
masthead_url = 'http://www.businessspectator.com.au/bs.nsf/logo-business-spectator.gif'
cover_url = masthead_url
oldest_article = 2
max_articles_per_feed = 100
no_stylesheets = True
#delay = 1
use_embedded_content = False
encoding = 'utf8'
publisher = 'Business Spectator'
category = 'News, Australia, Business'
language = 'en_AU'
publication_type = 'newsportal'
preprocess_regexps = [(re.compile(r'<!--.*?-->', re.DOTALL), lambda m: '')]
conversion_options = {
'comments' : description
,'tags' : category
,'language' : language
,'publisher' : publisher
,'linearize_tables': False
}
keep_only_tags = [dict(id='storyHeader'), dict(id='body-html')]
remove_tags = [dict(attrs={'class':'hql'})]
remove_attributes = ['width','height','style']
feeds = [
('Top Stories', 'http://www.businessspectator.com.au/top-stories.rss'),
('Alan Kohler', 'http://www.businessspectator.com.au/bs.nsf/RSS?readform&type=spectators&cat=Alan%20Kohler'),
('Robert Gottliebsen', 'http://www.businessspectator.com.au/bs.nsf/RSS?readform&type=spectators&cat=Robert%20Gottliebsen'),
('Stephen Bartholomeusz', 'http://www.businessspectator.com.au/bs.nsf/RSS?readform&type=spectators&cat=Stephen%20Bartholomeusz'),
('Daily Dossier', 'http://www.businessspectator.com.au/bs.nsf/RSS?readform&type=kgb&cat=dossier'),
('Australia', 'http://www.businessspectator.com.au/bs.nsf/RSS?readform&type=region&cat=australia'),
]