mirror of
https://github.com/kovidgoyal/calibre.git
synced 2025-09-29 15:31:08 -04:00
395f21054b
Fixes #1190381 [Updated recipe for La Nacion argentina](https://bugs.launchpad.net/calibre/+bug/1190381)
107 lines
5.6 KiB
Plaintext
107 lines
5.6 KiB
Plaintext
__copyright__ = '2008-2013, Darko Miletic <darko.miletic at gmail.com>'
|
|
'''
|
|
lanacion.com.ar
|
|
'''
|
|
|
|
from calibre.web.feeds.news import BasicNewsRecipe
|
|
|
|
class Lanacion(BasicNewsRecipe):
|
|
title = 'La Nacion'
|
|
__author__ = 'Darko Miletic'
|
|
description = "lanacion.com - Informacion actualizada las 24 horas, con noticias de Argentina y del mundo"
|
|
publisher = 'La Nacion S.A.'
|
|
category = 'news, politics, Argentina'
|
|
oldest_article = 1
|
|
max_articles_per_feed = 100
|
|
use_embedded_content = False
|
|
no_stylesheets = True
|
|
language = 'es_AR'
|
|
publication_type = 'newspaper'
|
|
remove_empty_feeds = True
|
|
masthead_url = 'http://www.lanacion.com.ar/_ui/desktop/imgs/layout/logos/ln-home.gif'
|
|
extra_css = """
|
|
h1{font-family: TheSans,Arial,sans-serif}
|
|
body{font-family: Arial,sans-serif}
|
|
img{display: block}
|
|
.firma,.fecha{font-size: small}
|
|
.epigrafe-columna{font-size: x-small}
|
|
"""
|
|
|
|
|
|
conversion_options = {
|
|
'comment' : description
|
|
, 'tags' : category
|
|
, 'publisher': publisher
|
|
, 'language' : language
|
|
}
|
|
|
|
remove_tags = [
|
|
dict(name=['iframe','embed','object','meta','link'])
|
|
,dict(attrs={'id':['herramientas','relacionadas','ampliar']})
|
|
]
|
|
|
|
remove_tags_before = dict(attrs={'id':'encabezado'})
|
|
remove_tags_after = dict(attrs={'id':'relacionadas'})
|
|
|
|
feeds = [
|
|
(u'Politica' , u'http://lanacion.com.ar.feedsportal.com/politica' )
|
|
,(u'Deportes' , u'http://lanacion.com.ar.feedsportal.com/deportes' )
|
|
,(u'Economia' , u'http://lanacion.com.ar.feedsportal.com/economia' )
|
|
,(u'Sociedad' , u'http://lanacion.com.ar.feedsportal.com/sociedad' )
|
|
,(u'Seguridad' , u'http://lanacion.com.ar.feedsportal.com/seguridad' )
|
|
,(u'Buenos Aires' , u'http://lanacion.com.ar.feedsportal.com/buenosaires' )
|
|
,(u'Opinion' , u'http://lanacion.com.ar.feedsportal.com/opinion' )
|
|
,(u'Espectaculos' , u'http://lanacion.com.ar.feedsportal.com/espectaculos' )
|
|
,(u'El Mundo' , u'http://lanacion.com.ar.feedsportal.com/mundo' )
|
|
,(u'Revista' , u'http://lanacion.com.ar.feedsportal.com/revistalanacion' )
|
|
,(u'Enfoques' , u'http://lanacion.com.ar.feedsportal.com/enfoques' )
|
|
,(u'Comercio Exterior' , u'http://lanacion.com.ar.feedsportal.com/comercioexterior' )
|
|
,(u'Tecnologia' , u'http://lanacion.com.ar.feedsportal.com/tecnologia' )
|
|
,(u'Turismo' , u'http://lanacion.com.ar.feedsportal.com/turismo' )
|
|
,(u'Al volante' , u'http://lanacion.com.ar.feedsportal.com/alvolante' )
|
|
,(u'El Campo' , u'http://lanacion.com.ar.feedsportal.com/elcampo' )
|
|
,(u'Moda y Belleza' , u'http://lanacion.com.ar.feedsportal.com/modaybelleza' )
|
|
,(u'Inmuebles Comerciales', u'http://lanacion.com.ar.feedsportal.com/inmueblescomerciales' )
|
|
,(u'Countries' , u'http://lanacion.com.ar.feedsportal.com/countries' )
|
|
,(u'adnCultura' , u'http://lanacion.com.ar.feedsportal.com/adncultura' )
|
|
,(u'The WSJ Americas' , u'http://lanacion.com.ar.feedsportal.com/wallstreetjournalamericas')
|
|
]
|
|
|
|
|
|
def get_article_url(self, article):
|
|
link = article.get('guid', None)
|
|
if link.startswith('http://blogs.lanacion') and not link.endswith('/'):
|
|
return self.browser.open_novisit(link).geturl()
|
|
if link.rfind('galeria=') > 0:
|
|
return None
|
|
return link
|
|
|
|
def get_cover_url(self):
|
|
soup = self.index_to_soup('http://www.lanacion.com.ar/edicion-impresa')
|
|
atap = soup.find(attrs={'class':'tapa'})
|
|
if atap:
|
|
li = atap.find('img')
|
|
if li:
|
|
return li['src']
|
|
return None
|
|
|
|
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
|