New recipes for Ambito and El Argentino by Darko Miletic

This commit is contained in:
Kovid Goyal 2009-01-18 11:28:27 -08:00
parent f16fe3e8e6
commit 8d6bc29107
5 changed files with 100 additions and 1 deletions

Binary file not shown.

After

Width:  |  Height:  |  Size: 508 B

Binary file not shown.

After

Width:  |  Height:  |  Size: 805 B

View File

@ -22,7 +22,7 @@ recipe_modules = ['recipe_' + r for r in (
'time_magazine', 'endgadget', 'fudzilla', 'nspm_int', 'nspm', 'pescanik', 'time_magazine', 'endgadget', 'fudzilla', 'nspm_int', 'nspm', 'pescanik',
'spiegel_int', 'themarketticker', 'tomshardware', 'xkcd', 'ftd', 'zdnet', 'spiegel_int', 'themarketticker', 'tomshardware', 'xkcd', 'ftd', 'zdnet',
'joelonsoftware', 'telepolis', 'common_dreams', 'nin', 'tomshardware_de', 'joelonsoftware', 'telepolis', 'common_dreams', 'nin', 'tomshardware_de',
'pagina12', 'infobae', 'pagina12', 'infobae', 'ambito', 'elargentino',
)] )]
import re, imp, inspect, time, os import re, imp, inspect, time, os

View File

@ -0,0 +1,44 @@
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2008, Darko Miletic <darko.miletic at gmail.com>'
'''
ambito.com
'''
from calibre.web.feeds.news import BasicNewsRecipe
class Ambito(BasicNewsRecipe):
title = 'Ambito.com'
__author__ = 'Darko Miletic'
description = 'Informacion Libre las 24 horas'
oldest_article = 2
max_articles_per_feed = 100
no_stylesheets = True
use_embedded_content = False
encoding = 'iso--8859-1'
cover_url = 'http://www.ambito.com/img/logo_.jpg'
html2lrf_options = [
'--comment' , description
, '--category' , 'news, Argentina'
, '--publisher' , title
]
feeds = [
(u'Principales Noticias', u'http://www.ambito.com/rss/noticiasp.asp' )
,(u'Economia' , u'http://www.ambito.com/rss/noticias.asp?S=Econom%EDa' )
,(u'Politica' , u'http://www.ambito.com/rss/noticias.asp?S=Pol%EDtica' )
,(u'Informacion General' , u'http://www.ambito.com/rss/noticias.asp?S=Informaci%F3n%20General')
,(u'Agro' , u'http://www.ambito.com/rss/noticias.asp?S=Agro' )
,(u'Internacionales' , u'http://www.ambito.com/rss/noticias.asp?S=Internacionales' )
,(u'Deportes' , u'http://www.ambito.com/rss/noticias.asp?S=Deportes' )
,(u'Espectaculos' , u'http://www.ambito.com/rss/noticias.asp?S=Espect%E1culos' )
,(u'Tecnologia' , u'http://www.ambito.com/rss/noticias.asp?S=Tecnologia' )
,(u'Salud' , u'http://www.ambito.com/rss/noticias.asp?S=Salud' )
,(u'Ambito Nacional' , u'http://www.ambito.com/rss/noticias.asp?S=Ambito%20Nacional' )
]
def print_version(self, url):
return url.replace('http://www.ambito.com/noticia.asp?','http://www.ambito.com/noticias/imprimir.asp?')

View File

@ -0,0 +1,55 @@
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2008, Darko Miletic <darko.miletic at gmail.com>'
'''
elargentino.com
'''
from calibre.web.feeds.news import BasicNewsRecipe
class ElArgentino(BasicNewsRecipe):
title = 'ElArgentino.com'
__author__ = 'Darko Miletic'
description = 'Informacion Libre las 24 horas'
oldest_article = 2
max_articles_per_feed = 100
no_stylesheets = True
use_embedded_content = False
encoding = 'utf8'
cover_url = 'http://www.elargentino.com/TemplateWeb/MediosFooter/tapa_elargentino.png'
html2lrf_options = [
'--comment' , description
, '--category' , 'news, Argentina'
, '--publisher' , 'ElArgentino.com'
]
remove_tags = [
dict(name='div', attrs={'id':'noprint' })
,dict(name='div', attrs={'class':'encabezadoImprimir'})
,dict(name='a' , attrs={'target':'_blank' })
]
feeds = [
(u'Portada' , u'http://www.elargentino.com/Highlights.aspx?Content-Type=text/xml&ChannelDesc=Home' )
,(u'Pais' , u'http://www.elargentino.com/Highlights.aspx?ParentType=Section&ParentId=112&Content-Type=text/xml&ChannelDesc=Pa%C3%ADs' )
,(u'Economia' , u'http://www.elargentino.com/Highlights.aspx?ParentType=Section&ParentId=107&Content-Type=text/xml&ChannelDesc=Econom%C3%ADa' )
,(u'Mundo' , u'http://www.elargentino.com/Highlights.aspx?ParentType=Section&ParentId=113&Content-Type=text/xml&ChannelDesc=Mundo' )
,(u'Tecnologia' , u'http://www.elargentino.com/Highlights.aspx?ParentType=Section&ParentId=118&Content-Type=text/xml&ChannelDesc=Tecnolog%C3%ADa' )
,(u'Espectaculos', u'http://www.elargentino.com/Highlights.aspx?ParentType=Section&ParentId=114&Content-Type=text/xml&ChannelDesc=Espect%C3%A1culos')
,(u'Deportes' , u'http://www.elargentino.com/Highlights.aspx?ParentType=Section&ParentId=106&Content-Type=text/xml&ChannelDesc=Deportes' )
,(u'Sociedad' , u'http://www.elargentino.com/Highlights.aspx?ParentType=Section&ParentId=109&Content-Type=text/xml&ChannelDesc=Sociedad' )
,(u'Entrevistas' , u'http://www.elargentino.com/Highlights.aspx?ParentType=Section&ParentId=115&Content-Type=text/xml&ChannelDesc=Entrevistas' )
]
def print_version(self, url):
main, sep, article_part = url.partition('/nota-')
article_id, rsep, rrest = article_part.partition('-')
return u'http://www.elargentino.com/Impresion.aspx?Id=' + article_id
def preprocess_html(self, soup):
mtag = '<meta http-equiv="Content-Type" content="text/html; charset=utf-8">'
soup.head.insert(0,mtag)
soup.prettify()
return soup