Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

Fix Austin Statesman

Browse Source
This commit is contained in:
Kovid Goyal 2011-03-01 17:45:30 -07:00
parent cb7814dfd5
commit ce2f9396b8
1 changed files with 13 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
resources/recipes/statesman.recipe
View File

@ -10,12 +10,14 @@ class AdvancedUserRecipe1278049615(BasicNewsRecipe):
max_articles_per_feed = 100
feeds = [(u'News', u'http://www.statesman.com/section-rss.do?source=news&includeSubSections=true'),
(u'Business', u'http://www.statesman.com/section-rss.do?source=business&includeSubSections=true'),
(u'Life', u'http://www.statesman.com/section-rss.do?source=life&includesubsection=true'),
(u'Editorial', u'http://www.statesman.com/section-rss.do?source=opinion&includesubsections=true'),
(u'Sports', u'http://www.statesman.com/section-rss.do?source=sports&includeSubSections=true')
]
feeds = [(u'News',
u'http://www.statesman.com/section-rss.do?source=news&includeSubSections=true'),
(u'Local', u'http://www.statesman.com/section-rss.do?source=local&includeSubSections=true'),
(u'Business', u'http://www.statesman.com/section-rss.do?source=business&includeSubSections=true'),
(u'Life', u'http://www.statesman.com/section-rss.do?source=life&includesubsection=true'),
(u'Editorial', u'http://www.statesman.com/section-rss.do?source=opinion&includesubsections=true'),
(u'Sports', u'http://www.statesman.com/section-rss.do?source=sports&includeSubSections=true')
]
masthead_url = "http://www.statesman.com/images/cmg-logo.gif"
#temp_files = []
#articles_are_obfuscated = True
@ -28,8 +30,11 @@ class AdvancedUserRecipe1278049615(BasicNewsRecipe):
conversion_options = {'linearize_tables':True}
remove_tags = [
dict(name='div', attrs={'id':'cxArticleOptions'}),
{'class':['perma', 'comments', 'trail', 'share-buttons',
'toggle_show_on']},
]
keep_only_tags = [
dict(name='div', attrs={'class':'cxArticleHeader'}),
dict(name='div', attrs={'id':'cxArticleBodyText'}),
dict(name='div', attrs={'class':'cxArticleHeader'}),
dict(name='div', attrs={'id':['cxArticleBodyText',
'content']}),
]