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

Update Telegraph UK

Browse Source
This commit is contained in:
Kovid Goyal 2013-08-06 21:51:56 +05:30
parent a0fb42865d
commit 727944a4b5
1 changed files with 18 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
recipes/telegraph_uk.recipe
View File

@ -12,7 +12,8 @@ class TelegraphUK(BasicNewsRecipe):
description = 'News from United Kingdom' description = 'News from United Kingdom'
oldest_article = 2 oldest_article = 2
category = 'news, politics, UK' category = 'news, politics, UK'
publisher = 'Telegraph Media Group ltd.' publisher = 'Telegraph Media Group ltd.'
compress_news_images = True
max_articles_per_feed = 100 max_articles_per_feed = 100
no_stylesheets = True no_stylesheets = True
language = 'en_GB' language = 'en_GB'
@ -34,30 +35,30 @@ class TelegraphUK(BasicNewsRecipe):
, 'publisher' : publisher , 'publisher' : publisher
, 'language' : language , 'language' : language
} }
keep_only_tags = [ keep_only_tags = [
dict(name='div', attrs={'class':['storyHead','byline']}) dict(name='div', attrs={'class':['storyHead','byline']})
,dict(name='div', attrs={'id':'mainBodyArea' }) ,dict(name='div', attrs={'id':'mainBodyArea'})
] ]
remove_tags = [dict(name='div', attrs={'class':['related_links_inline',"imgindex","next","prev","gutterUnder",'ssImgHide','imageExtras','ssImg hide','related_links_video']}) remove_tags = [dict(name='div', attrs={
'class':['related_links_inline',"imgindex","next","prev","gutterUnder",'ssImgHide','imageExtras','ssImg hide','related_links_video']})
,dict(name='ul' , attrs={'class':['shareThis shareBottom']}) ,dict(name='ul' , attrs={'class':['shareThis shareBottom']})
,dict(name='span', attrs={'class':['num','placeComment']}) ,dict(name='span', attrs={'class':['num','placeComment']})
] ]
feeds = [ feeds = [
(u'UK News' , u'http://www.telegraph.co.uk/news/uknews/rss' ) (u'UK News' , u'http://www.telegraph.co.uk/news/uknews/rss')
,(u'World News' , u'http://www.telegraph.co.uk/news/worldnews/rss' ) ,(u'World News' , u'http://www.telegraph.co.uk/news/worldnews/rss')
,(u'Politics' , u'http://www.telegraph.co.uk/news/newstopics/politics/rss' ) ,(u'Politics' , u'http://www.telegraph.co.uk/news/newstopics/politics/rss')
,(u'Finance' , u'http://www.telegraph.co.uk/finance/rss' ) ,(u'Finance' , u'http://www.telegraph.co.uk/finance/rss')
,(u'Technology News', u'http://www.telegraph.co.uk/scienceandtechnology/technology/technologynews/rss' ) ,(u'Technology News', u'http://www.telegraph.co.uk/scienceandtechnology/technology/technologynews/rss')
,(u'UK News' , u'http://www.telegraph.co.uk/scienceandtechnology/technology/technologyreviews/rss') ,(u'UK News' , u'http://www.telegraph.co.uk/scienceandtechnology/technology/technologyreviews/rss')
,(u'Science News' , u'http://www.telegraph.co.uk/scienceandtechnology/science/sciencenews/rss' ) ,(u'Science News' , u'http://www.telegraph.co.uk/scienceandtechnology/science/sciencenews/rss')
,(u'Sport' , u'http://www.telegraph.co.uk/sport/rss' ) ,(u'Sport' , u'http://www.telegraph.co.uk/sport/rss')
,(u'Earth News' , u'http://www.telegraph.co.uk/earth/earthnews/rss' ) ,(u'Earth News' , u'http://www.telegraph.co.uk/earth/earthnews/rss')
,(u'Comment' , u'http://www.telegraph.co.uk/comment/rss' ) ,(u'Comment' , u'http://www.telegraph.co.uk/comment/rss')
,(u'Travel' , u'http://www.telegraph.co.uk/travel/rss' ) ,(u'Travel' , u'http://www.telegraph.co.uk/travel/rss')
,(u'How about that?', u'http://www.telegraph.co.uk/news/newstopics/howaboutthat/rss' ) ,(u'How about that?', u'http://www.telegraph.co.uk/news/newstopics/howaboutthat/rss')
] ]
def populate_article_metadata(self, article, soup, first): def populate_article_metadata(self, article, soup, first):
if first and hasattr(self, 'add_toc_thumbnail'): if first and hasattr(self, 'add_toc_thumbnail'):
@ -70,3 +71,4 @@ class TelegraphUK(BasicNewsRecipe):
if 'picture-galleries' in url or 'pictures' in url or 'picturegalleries' in url : if 'picture-galleries' in url or 'pictures' in url or 'picturegalleries' in url :
url = None url = None
return url return url