Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-08-07 09:01:38 -04:00
Code Issues Packages Projects Releases Wiki Activity

...

Browse Source
This commit is contained in:
Kovid Goyal 2013-08-02 16:46:15 +05:30
parent 3c777a8154
commit 211aff74d7
1 changed files with 17 additions and 33 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

50
recipes/the_scotsman.recipe
View File

@ -1,4 +1,3 @@
__license__ = 'GPL v3'
__copyright__ = '2008 - 2011, Darko Miletic <darko.miletic at gmail.com>'
'''
@ -12,54 +11,39 @@ class TheScotsman(BasicNewsRecipe):
__author__ = 'Darko Miletic'
description = 'News from Scotland'
publisher = 'Johnston Publishing Ltd.'
category = 'news, politics, Scotland, UK'
category = 'news, politics, Scotland, UK'
oldest_article = 2
max_articles_per_feed = 100
no_stylesheets = True
use_embedded_content = False
auto_cleanup = True
language = 'en_GB'
encoding = 'utf-8'
publication_type = 'newspaper'
remove_empty_feeds = True
remove_empty_feeds = True
masthead_url = 'http://www.scotsman.com/webimage/swts_thescotsman_image_e_7_25526!image/3142543874.png_gen/derivatives/default/3142543874.png'
extra_css = 'body{font-family: Arial,Helvetica,sans-serif}'
#keep_only_tags = [dict(attrs={'class':'editorialSection'})]
#remove_tags_after = dict(attrs={'class':'socialBookmarkPanel'})
#remove_tags = [
#dict(name=['meta','iframe','object','embed','link']),
#dict(attrs={'class':['secondaryArticlesNav','socialBookmarkPanel']}),
#dict(attrs={'id':'relatedArticles'})
#]
keep_only_tags = [dict(name='div', attrs={'class':'article'})]
remove_attributes = ['lang']
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
}
feeds = [
('Latest News' , 'http://www.scotsman.com/cmlink/1.957140' ),
('UK' , 'http://www.scotsman.com/cmlink/1.957142' ),
('Scotland' , 'http://www.scotsman.com/cmlink/1.957141' ),
('International', 'http://www.scotsman.com/cmlink/1.957143' ),
('Politics' , 'http://www.scotsman.com/cmlink/1.957044' ),
('Latest News' , 'http://www.scotsman.com/cmlink/1.957140'),
('UK' , 'http://www.scotsman.com/cmlink/1.957142'),
('Scotland' , 'http://www.scotsman.com/cmlink/1.957141'),
('International', 'http://www.scotsman.com/cmlink/1.957143'),
('Politics' , 'http://www.scotsman.com/cmlink/1.957044'),
('Arts' , 'http://www.scotsman.com/cmlink/1.1804825'),
('Entertainment', 'http://www.scotsman.com/cmlink/1.957053' ),
('Sports' , 'http://www.scotsman.com/cmlink/1.957151' ),
('Business' , 'http://www.scotsman.com/cmlink/1.957156' ),
('Features' , 'http://www.scotsman.com/cmlink/1.957149' ),
('Opinion' , 'http://www.scotsman.com/cmlink/1.957054' )
('Entertainment', 'http://www.scotsman.com/cmlink/1.957053'),
('Sports' , 'http://www.scotsman.com/cmlink/1.957151'),
('Business' , 'http://www.scotsman.com/cmlink/1.957156'),
('Features' , 'http://www.scotsman.com/cmlink/1.957149'),
('Opinion' , 'http://www.scotsman.com/cmlink/1.957054')
]
#def preprocess_html(self, soup):
#for item in soup.findAll(style=True):
#del item['style']
#for item in soup.findAll('img'):
#if not item.has_key('alt'):
#item['alt'] = 'image'
#return soup