calibre/recipes/el_pais_uy.recipe

81 lines
3.4 KiB
Python

#!/usr/bin/env python
__license__ = 'GPL v3'
__author__ = '2010, Gustavo Azambuja <hola at gazambuja.com>'
'''
http://www.elpais.com.uy/
'''
from calibre.web.feeds.news import BasicNewsRecipe
class General(BasicNewsRecipe):
title = 'El Pais - Uruguay'
__author__ = 'Gustavo Azambuja'
description = 'Noticias de Uruguay y el resto del mundo'
publisher = 'EL PAIS S.A.'
category = 'news, politics, Uruguay'
language = 'es_UY'
timefmt = '[%a, %d %b, %Y]'
use_embedded_content = False
recursion = 2
encoding = 'iso-8859-1'
masthead_url = 'http://www.elpais.com.uy/Images/09/cabezal/logo_PDEP.png'
publication_type = 'newspaper'
remove_javascript = True
no_stylesheets = True
oldest_article = 2
max_articles_per_feed = 200
keep_only_tags = [
dict(name='h1'),
dict(name='div', attrs={'id':'Contenido'})
]
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
}
remove_tags = [
dict(name='div', attrs={'class':['date_text', 'comments', 'form_section', 'share_it']}),
dict(name='div', attrs={'id':['relatedPosts', 'spacer', 'banner_izquierda', 'right_container']}),
dict(name='p', attrs={'class':'FacebookLikeButton'}),
dict(name=['object','form']),
dict(name=['object','table']) ]
extra_css = '''
h1{font-family:Geneva, Arial, Helvetica, sans-serif;color:#154B7A;}
h3{font-size: 14px;color:#999999; font-family:Geneva, Arial, Helvetica, sans-serif;font-weight: bold;}
h2{color:#666666; font-family:Geneva, Arial, Helvetica, sans-serif;font-size:small;}
p {font-family:Arial,Helvetica,sans-serif;}
body{font-family: Verdana,Arial,Helvetica,sans-serif }
img{margin-bottom: 0.4em; display:block;}
'''
feeds = [
(u'Ultimo Momento', u'http://www.elpais.com.uy/formatos/rss/index.asp?seccion=umomento'),
(u'Editorial', u'http://www.elpais.com.uy/formatos/rss/index.asp?seccion=editorial'),
(u'Nacional', u'http://www.elpais.com.uy/formatos/rss/index.asp?seccion=nacional'),
(u'Internacional', u'http://www.elpais.com.uy/formatos/rss/index.asp?seccion=internacional'),
(u'Espectaculos', u'http://www.elpais.com.uy/formatos/rss/index.asp?seccion=espectaculos'),
(u'Deportes', u'http://www.elpais.com.uy/formatos/rss/index.asp?seccion=deportes'),
(u'Ciudades', u'http://www.elpais.com.uy/formatos/rss/index.asp?seccion=ciudades'),
(u'Economia', u'http://www.elpais.com.uy/formatos/rss/index.asp?seccion=economia')
]
def get_cover_url(self):
cover_url = None
index = 'http://www.elpais.com.uy'
soup = self.index_to_soup(index)
link_item = soup.find('div',attrs={'class':'boxmedio box257'})
print link_item
if link_item:
cover_url = 'http://www.elpais.com.uy'+link_item.img['src']
return cover_url
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return soup