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 Infobae.com

Browse Source 
Fixes #1499818 [Updated recipe for Infobae](https://bugs.launchpad.net/calibre/+bug/1499818)
This commit is contained in:
Kovid Goyal 2015-09-25 23:33:30 +05:30
parent 4ef778fff8
commit 648b73541a
2 changed files with 15 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
recipes/icons/infobae.png
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 193 B

After

Width:  |  Height:  |  Size: 270 B

48
recipes/infobae.recipe
View File

@ -1,5 +1,5 @@
__license__ = 'GPL v3' __license__ = 'GPL v3'
__copyright__ = '2008-2011, Darko Miletic <darko.miletic at gmail.com>' __copyright__ = '2008-2015, Darko Miletic <darko.miletic at gmail.com>'
''' '''
infobae.com infobae.com
''' '''
@ -16,9 +16,9 @@ class Infobae(BasicNewsRecipe):
max_articles_per_feed = 100 max_articles_per_feed = 100
no_stylesheets = True no_stylesheets = True
use_embedded_content = False use_embedded_content = False
auto_cleanup = False
language = 'es_AR' language = 'es_AR'
encoding = 'utf8' encoding = 'utf8'
masthead_url = 'http://www.infobae.com/media/img/static/logo-infobae.gif'
remove_empty_feeds = True remove_empty_feeds = True
extra_css = ''' extra_css = '''
body{font-family: Arial,Helvetica,sans-serif} body{font-family: Arial,Helvetica,sans-serif}
@ -33,42 +33,20 @@ class Infobae(BasicNewsRecipe):
, 'language' : language , 'language' : language
} }
keep_only_tags = [dict(attrs={'class':['titularnota','nota','post-title','post-entry','entry-title','entry-info','entry-content']})] remove_tags_before = dict(name='h1', attrs={'class':'entry-title'})
remove_tags_after = dict(attrs={'class':['interior-noticia','nota-desc','tags']}) remove_tags_after = dict(name='div', attrs={'class':'narrowcontent'})
remove_tags = [ remove_tags = [
dict(name=['base','meta','link','iframe','object','embed','ins']) dict(name=['base','meta','link','iframe','object','embed','ins'])
,dict(attrs={'class':['barranota','tags']}) ,dict(attrs={'class':['social-hori clearfix','narrowcontent','embed_cont type_freetext']})
,dict(attrs={'id':'captcha-modal'})
] ]
feeds = [ feeds = [
(u'Saludable' , u'http://www.infobae.com/rss/saludable.xml') (u'Ahora' , u'http://cdn02.ib.infobae.com/adjuntos/162/rss/Infobae.xml')
,(u'Economia' , u'http://www.infobae.com/rss/economia.xml' ) ,(u'Economia' , u'http://cdn02.ib.infobae.com/adjuntos/162/rss/economia.xml')
,(u'En Numeros', u'http://www.infobae.com/rss/rating.xml' ) ,(u'Cultura' , u'http://cdn02.ib.infobae.com/adjuntos/162/rss/Cultura.xml')
,(u'Finanzas' , u'http://www.infobae.com/rss/finanzas.xml' ) ,(u'Finanzas' , u'http://cdn02.ib.infobae.com/adjuntos/162/rss/finanzas.xml')
,(u'Mundo' , u'http://www.infobae.com/rss/mundo.xml' ) ,(u'Sociedad' , u'http://cdn02.ib.infobae.com/adjuntos/162/rss/sociedad.xml')
,(u'Sociedad' , u'http://www.infobae.com/rss/sociedad.xml' ) ,(u'Politica' , u'http://cdn02.ib.infobae.com/adjuntos/162/rss/politica.xml')
,(u'Politica' , u'http://www.infobae.com/rss/politica.xml' ) ,(u'Deportes' , u'http://cdn02.ib.infobae.com/adjuntos/162/rss/deportes.xml')
,(u'Deportes' , u'http://www.infobae.com/rss/deportes.xml' )
] ]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
for item in soup.findAll('a'):
limg = item.find('img')
if item.string is not None:
str = item.string
item.replaceWith(str)
else:
if limg:
item.name = 'div'
item.attrs = []
else:
str = self.tag_to_string(item)
item.replaceWith(str)
for item in soup.findAll('img'):
if not item.has_key('alt'):
item['alt'] = 'image'
return soup