Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-07 10:14:46 -04:00
Code Issues Packages Projects Releases Wiki Activity

Fix ABC Australia

Browse Source
This commit is contained in:
Kovid Goyal 2011-11-20 20:56:42 +05:30
parent c82ec4e2fb
commit b21cf64e79
1 changed files with 19 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
recipes/abc_au.recipe
View File

@ -1,5 +1,5 @@
__license__ = 'GPL v3' __license__ = 'GPL v3'
__copyright__ = '2010, Dean Cording' __copyright__ = '2011, Pat Stapleton <pat.stapleton at gmail.com>'
''' '''
abc.net.au/news abc.net.au/news
''' '''
@ -8,7 +8,7 @@ from calibre.web.feeds.recipes import BasicNewsRecipe
class ABCNews(BasicNewsRecipe): class ABCNews(BasicNewsRecipe):
title = 'ABC News' title = 'ABC News'
__author__ = 'Dean Cording' __author__ = 'Pat Stapleton, Dean Cording'
description = 'News from Australia' description = 'News from Australia'
masthead_url = 'http://www.abc.net.au/news/assets/v5/images/common/logo-news.png' masthead_url = 'http://www.abc.net.au/news/assets/v5/images/common/logo-news.png'
cover_url = 'http://www.abc.net.au/news/assets/v5/images/common/logo-news.png' cover_url = 'http://www.abc.net.au/news/assets/v5/images/common/logo-news.png'
@ -23,7 +23,9 @@ class ABCNews(BasicNewsRecipe):
category = 'News, Australia, World' category = 'News, Australia, World'
language = 'en_AU' language = 'en_AU'
publication_type = 'newsportal' publication_type = 'newsportal'
preprocess_regexps = [(re.compile(r'<!--.*?-->', re.DOTALL), lambda m: '')] # preprocess_regexps = [(re.compile(r'<!--.*?-->', re.DOTALL), lambda m: '')]
#Remove annoying map links (inline-caption class is also used for some image captions! hence regex to match maps.google)
preprocess_regexps = [(re.compile(r'<a class="inline-caption" href="http://maps\.google\.com.*?/a>', re.DOTALL), lambda m: '')]
conversion_options = { conversion_options = {
'comments' : description 'comments' : description
,'tags' : category ,'tags' : category
@ -32,23 +34,23 @@ class ABCNews(BasicNewsRecipe):
,'linearize_tables': False ,'linearize_tables': False
} }
keep_only_tags = dict(id='article') keep_only_tags = [dict(attrs={'class':['article section']})]
remove_tags = [dict(attrs={'class':['related', 'tags']}), remove_tags = [dict(attrs={'class':['related', 'tags', 'tools', 'attached-content ready',
dict(id='statepromo') 'inline-content story left', 'inline-content map left contracted', 'published',
] 'story-map', 'statepromo', 'topics', ]})]
remove_attributes = ['width','height'] remove_attributes = ['width','height']
feeds = [ feeds = [
('Top Stories', 'http://www.abc.net.au/news/syndicate/topstoriesrss.xml'), ('Top Stories', 'http://www.abc.net.au/news/feed/45910/rss.xml'),
('Canberra', 'http://www.abc.net.au/news/indexes/idx-act/rss.xml'), ('Canberra', 'http://www.abc.net.au/news/feed/6910/rss.xml'),
('Sydney', 'http://www.abc.net.au/news/indexes/sydney/rss.xml'), ('Sydney', 'http://www.abc.net.au/news/feed/10232/rss.xml'),
('Melbourne', 'http://www.abc.net.au/news/indexes/melbourne/rss.xml'), ('Melbourne', 'http://www.abc.net.au/news/feed/21708/rss.xml'),
('Brisbane', 'http://www.abc.net.au/news/indexes/brisbane/rss.xml'), ('Brisbane', 'http://www.abc.net.au/news/feed/12858/rss.xml'),
('Perth', 'http://www.abc.net.au/news/indexes/perth/rss.xml'), ('Perth', 'feed://www.abc.net.au/news/feed/24886/rss.xml'),
('Australia', 'http://www.abc.net.au/news/indexes/idx-australia/rss.xml'), ('Australia', 'http://www.abc.net.au/news/feed/46182/rss.xml'),
('World', 'http://www.abc.net.au/news/indexes/world/rss.xml'), ('World', 'http://www.abc.net.au/news/feed/52278/rss.xml'),
('Business', 'http://www.abc.net.au/news/indexes/business/rss.xml'), ('Business', 'http://www.abc.net.au/news/feed/51892/rss.xml'),
('Science and Technology', 'http://www.abc.net.au/news/tag/science-and-technology/rss.xml'), ('Science and Technology', 'http://www.abc.net.au/news/feed/2298/rss.xml'),
] ]