Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-07 18:24:30 -04:00
Code Issues Packages Projects Releases Wiki Activity

Update Ambito

Browse Source 
Fixes #1499866 [Updated recipe for Ambito](https://bugs.launchpad.net/calibre/+bug/1499866)
This commit is contained in:
Kovid Goyal 2015-09-26 08:54:39 +05:30
parent 323be26c0f
commit dd29778835
1 changed files with 5 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
recipes/ambito.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>'
''' '''
ambito.com ambito.com
''' '''
@ -15,8 +15,9 @@ class Ambito(BasicNewsRecipe):
oldest_article = 2 oldest_article = 2
no_stylesheets = True no_stylesheets = True
encoding = 'cp1252' encoding = 'cp1252'
masthead_url = 'http://www.ambito.com/img/logo_.jpg' masthead_url = 'http://www.ambito.com/img/logo.jpg'
use_embedded_content = False use_embedded_content = False
remove_empty_feeds = True
language = 'es_AR' language = 'es_AR'
publication_type = 'newsportal' publication_type = 'newsportal'
extra_css = """ extra_css = """
@ -33,9 +34,8 @@ class Ambito(BasicNewsRecipe):
, 'language' : language , 'language' : language
} }
keep_only_tags = [dict(name='div', attrs={'align':'justify'})] keep_only_tags = [dict(attrs={'id':['tituloDespliegue','imgDesp','textoDespliegue']})]
remove_tags = [dict(name=['object','link','embed','iframe','meta','link','table','img'])] remove_tags = [dict(name=['object','link','embed','iframe','meta','link'])]
remove_attributes = ['align']
feeds = [ feeds = [
(u'Principales Noticias', u'http://www.ambito.com/rss/noticiasp.asp' ) (u'Principales Noticias', u'http://www.ambito.com/rss/noticiasp.asp' )
@ -49,16 +49,3 @@ class Ambito(BasicNewsRecipe):
,(u'Tecnologia' , u'http://www.ambito.com/rss/noticias.asp?S=Tecnolog%EDa' ) ,(u'Tecnologia' , u'http://www.ambito.com/rss/noticias.asp?S=Tecnolog%EDa' )
,(u'Ambito Nacional' , u'http://www.ambito.com/rss/noticias.asp?S=Ambito%20Nacional' ) ,(u'Ambito Nacional' , u'http://www.ambito.com/rss/noticias.asp?S=Ambito%20Nacional' )
] ]
def print_version(self, url):
return url.replace('/noticia.asp?','/noticias/imprimir.asp?')
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
for item in soup.findAll('a'):
str = item.string
if str is None:
str = self.tag_to_string(item)
item.replaceWith(str)
return soup