This commit is contained in:
Kovid Goyal 2016-12-30 08:29:28 +05:30
parent 0d032ad273
commit 94d7bce604
3 changed files with 0 additions and 166 deletions

View File

@ -1,56 +0,0 @@
__license__ = 'GPL v3'
__copyright__ = '2008-2011, Darko Miletic <darko.miletic at gmail.com>'
'''
www.cronista.com
'''
from calibre.web.feeds.news import BasicNewsRecipe
class Pagina12(BasicNewsRecipe):
title = 'El Cronista Comercial'
__author__ = 'Darko Miletic'
description = 'El Cronista Comercial es el Diario economico-politico mas valorado. Es la fuente mas confiable de informacion en temas de economia, finanzas y negocios enmarcados politicamente.' # noqa
publisher = 'Cronista.com'
category = 'news, politics, economy, finances, Argentina'
oldest_article = 2
max_articles_per_feed = 200
no_stylesheets = True
encoding = 'utf8'
use_embedded_content = False
language = 'es_AR'
remove_empty_feeds = True
publication_type = 'newspaper'
masthead_url = 'http://www.cronista.com/export/sites/diarioelcronista/arte/header-logo.gif'
extra_css = """
body{font-family: Arial,Helvetica,sans-serif }
h2{font-family: Georgia,"Times New Roman",Times,serif }
img{margin-bottom: 0.4em; display:block}
.nom{font-weight: bold; vertical-align: baseline}
.autor-cfoto{border-bottom: 1px solid #D2D2D2;
border-top: 1px solid #D2D2D2;
display: inline-block;
margin: 0 10px 10px 0;
padding: 10px;
width: 210px}
.under{font-weight: bold}
.time{font-size: small}
"""
conversion_options = {
'comment': description, 'tags': category, 'publisher': publisher, 'language': language
}
remove_tags = [
dict(name=['meta', 'link', 'base', 'iframe', 'object', 'embed']), dict(
attrs={'class': ['user-tools', 'tabsmedia']})
]
remove_attributes = ['lang']
remove_tags_before = dict(attrs={'class': 'top'})
remove_tags_after = dict(attrs={'class': 'content-nota'})
feeds = [(u'Ultimas noticias', u'http://www.cronista.com/rss.html')]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return soup

View File

@ -1,35 +0,0 @@
#!/usr/bin/env python2
__license__ = 'GPL v3'
__copyright__ = '2009, Darko Miletic <darko.miletic at gmail.com>'
'''
www.ieco.clarin.com
'''
from calibre.web.feeds.news import BasicNewsRecipe
class iEco(BasicNewsRecipe):
title = 'iEco Clarin'
__author__ = 'Darko Miletic'
description = 'iEco, el suplemento economico de Clarin. Las ultimas noticias de Empresas y Negocios, Mercados y Finanzas, Economia y Tecnologia de la Argentina y el mundo.' # noqa
oldest_article = 7
max_articles_per_feed = 100
no_stylesheets = True
use_embedded_content = False
encoding = 'utf-8'
publisher = 'Grupo Clarin'
language = 'es_AR'
cover_url = 'http://www.ieco.clarin.com/static2/images/Tapa-PDF.gif'
extra_css = ' #bd{font-family: sans-serif} '
conversion_options = {
'comment': description, 'publisher': publisher, 'author': publisher, 'language': language
}
feeds = [(u'Articulos', u'http://www.ieco.clarin.com/rss/')]
keep_only_tags = [dict(name='div', attrs={'id': 'bd'})]
remove_tags = [dict(name=['object', 'link'])]
def print_version(self, url):
return url + '?print=1'

View File

@ -1,75 +0,0 @@
__license__ = 'GPL v3'
__copyright__ = '2013, Darko Miletic <darko.miletic at gmail.com>'
'''
www.lacapital.com.ar
'''
from calibre.web.feeds.news import BasicNewsRecipe
class LaCapital(BasicNewsRecipe):
title = 'La Capital de Rosario'
__author__ = 'Darko Miletic'
description = 'Noticias, actualidad y toda la informacion de Rosario y la region'
publisher = 'Diario La Capital S. A.'
category = 'news, politics, Rosario, Santa Fe, Argentina'
oldest_article = 2
max_articles_per_feed = 200
no_stylesheets = True
encoding = 'utf8'
use_embedded_content = False
language = 'es_AR'
remove_empty_feeds = True
publication_type = 'newspaper'
masthead_url = 'http://www.lacapital.com.ar/system/modules/com.tfsla.diario.core/resources/images/logoLaCapital_noCom.png'
extra_css = """
body{font-family: Georgia,"Times New Roman",Times,serif }
img{margin-bottom: 0.4em; display:block}
"""
conversion_options = {
'comment': description, 'tags': category, 'publisher': publisher, 'language': language
}
keep_only_tags = [dict(attrs={'class': 'leer'})]
remove_tags_after = dict(attrs={'class': 'notaA'})
remove_tags = [
dict(name=['meta', 'link', 'iframe', 'object']), dict(
name='div', attrs={'class': ['herramientas', 'almargen', 'relacionadas']})
]
feeds = [
(u'Portada', u'http://www.lacapital.com.ar/rss/home.xml'),
(u'La Ciudad', u'http://www.lacapital.com.ar/rss/laciudad.xml'),
(u'Politica', u'http://www.lacapital.com.ar/rss/politica.xml'),
(u'Economia', u'http://www.lacapital.com.ar/rss/economia.xml'),
(u'La Region', u'http://www.lacapital.com.ar/rss/laregion.xml'),
(u'Informacion General', u'http://www.lacapital.com.ar/rss/informaciongral.xml'),
(u'El Mundo', u'http://www.lacapital.com.ar/rss/elmundo.xml'),
(u'Opinion', u'http://www.lacapital.com.ar/rss/opinion.xml'),
(u'Cartas de lectores', u'http://www.lacapital.com.ar/rss/cartasdelectores.xml'),
(u'Escenario', u'http://www.lacapital.com.ar/rss/escenario.xml'),
(u'Policiales', u'http://www.lacapital.com.ar/rss/policiales.xml'),
(u'Ovacion', u'http://www.lacapital.com.ar/rss/ovacion.xml'),
(u'Turismo', u'http://www.lacapital.com.ar/rss/turismo.xml'),
(u'Economia', u'http://www.lacapital.com.ar/rss/economia.xml'),
(u'Señales', u'http://www.lacapital.com.ar/rss/senales.xml'),
(u'Educacion', u'http://www.lacapital.com.ar/rss/educacion.xml'),
(u'Estilo', u'http://www.lacapital.com.ar/rss/estilo.xml'),
(u'Salud', u'http://www.lacapital.com.ar/rss/salud.xml'),
(u'Tecnologia', u'http://www.lacapital.com.ar/rss/tecnologia.xml')
]
def get_cover_url(self):
soup = self.index_to_soup(
'http://www.lacapital.com.ar/impresa/tapa.html')
for image in soup.findAll('img', alt=True):
if image['alt'].startswith('Tapa de papel'):
return image['src']
return None
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return soup