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 #7887 (The news fetch for "Times of India" and "The Economic Times" (ENGLISH_ INDIA) not working.)

Browse Source
This commit is contained in:
Kovid Goyal 2010-12-14 20:29:36 -07:00
parent fd20c4ac7f
commit 4c947784ce
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
resources/recipes/theeconomictimes_india.recipe
View File

@ -4,6 +4,7 @@ __copyright__ = '2008-2010, Darko Miletic <darko.miletic at gmail.com>'
economictimes.indiatimes.com economictimes.indiatimes.com
''' '''
from calibre.web.feeds.news import BasicNewsRecipe from calibre.web.feeds.news import BasicNewsRecipe
class TheEconomicTimes(BasicNewsRecipe): class TheEconomicTimes(BasicNewsRecipe):
@ -32,18 +33,17 @@ class TheEconomicTimes(BasicNewsRecipe):
, 'language' : language , 'language' : language
} }
keep_only_tags = [dict(attrs={'class':'printdiv'})]
remove_tags = [dict(name=['object','link','embed','iframe','base','table','meta'])]
remove_attributes = ['name']
remove_tags_before = dict(name='h1')
feeds = [(u'All articles', u'http://economictimes.indiatimes.com/rssfeedsdefault.cms')] feeds = [(u'All articles', u'http://economictimes.indiatimes.com/rssfeedsdefault.cms')]
def print_version(self, url): def print_version(self, url):
rest, sep, art = url.rpartition('/articleshow/') rest, sep, art = url.rpartition('/articleshow/')
return 'http://m.economictimes.com/PDAET/articleshow/' + art
return 'http://economictimes.indiatimes.com/articleshow/' + art + '?prtpage=1' return 'http://economictimes.indiatimes.com/articleshow/' + art + '?prtpage=1'
def get_article_url(self, article): def get_article_url(self, article):
rurl = article.get('link', None) rurl = article.get('guid', None)
if (rurl.find('/quickieslist/') > 0) or (rurl.find('/quickiearticleshow/') > 0): if (rurl.find('/quickieslist/') > 0) or (rurl.find('/quickiearticleshow/') > 0):
return None return None
return rurl return rurl