calibre/recipes/ultimahora.recipe
Kovid Goyal 567040ee1e Perform PEP8 compliance checks on the entire codebase
Some bits of PEP 8 are turned off via setup.cfg
2016-07-29 21:25:17 +05:30

53 lines
2.1 KiB
Plaintext

__license__ = 'GPL v3'
__copyright__ = '2010, Darko Miletic <darko.miletic at gmail.com>'
'''
ultimahora.com
'''
from calibre.web.feeds.news import BasicNewsRecipe
class UltimaHora_py(BasicNewsRecipe):
title = 'Ultima Hora'
__author__ = 'Darko Miletic'
description = 'Noticias de Paraguay y el resto del mundo'
publisher = 'EDITORIAL EL PAIS S.A.'
category = 'news, politics, Paraguay'
oldest_article = 2
max_articles_per_feed = 200
no_stylesheets = True
encoding = 'cp1252'
use_embedded_content = False
language = 'es_PY'
remove_empty_feeds = True
publication_type = 'newspaper'
masthead_url = 'http://www.ultimahora.com/imgs/uh-com.gif'
extra_css = ' body{font-family: Arial,Helvetica,sans-serif } img{margin-bottom: 0.4em} .sub_titulo_mediano,.TituloNota{font-family: Georgia,"Times New Roman",Times,serif} .sub_titulo_mediano{font-weight: bold} ' # noqa
conversion_options = {
'comment': description, 'tags': category, 'publisher': publisher, 'language': language
}
remove_tags = [
dict(name=['form', 'iframe', 'embed', 'object', 'link', 'base', 'table'])]
keep_only_tags = [
dict(attrs={'id': ['nota_titulo', 'nota_copete', 'texto']})]
feeds = [
(u'Arte y Espectaculos', u'http://www.ultimahora.com/adjuntos/rss/UHEspectaculos.xml'),
(u'Ciudad del Este', u'http://www.ultimahora.com/adjuntos/rss/UHCDE.xml'),
(u'Deportes', u'http://www.ultimahora.com/adjuntos/rss/UHDeportes.xml'),
(u'Ultimo momento', u'http://www.ultimahora.com/adjuntos/rss/UltimoMomento.xml'),
(u'Nacionales', u'http://www.ultimahora.com/adjuntos/rss/uh-rss-nacionales.xml'),
(u'Politica', u'http://www.ultimahora.com/adjuntos/rss/uh-rss-politica.xml'),
(u'Sucesos', u'http://www.ultimahora.com/adjuntos/rss/uh-rss-sucesos.xml'),
(u'Economia', u'http://www.ultimahora.com/adjuntos/rss/uh-rss-economia.xml'),
(u'Ciencia y Tecnologia', u'http://www.ultimahora.com/adjuntos/rss/uh-rss-ciencia.xml')
]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return soup