Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-08 02:34:06 -04:00
Code Issues Packages Projects Releases Wiki Activity

Fix #6452 (New recipes for Bolivian newspapers in Spanish)

Browse Source
This commit is contained in:
Kovid Goyal 2010-08-09 14:54:52 -06:00
parent 8f0e4f5367
commit 23875dec1a
4 changed files with 127 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
resources/images/news/la_razon_bo.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 356 B

BIN
resources/images/news/los_tiempos_bo.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 464 B

64
resources/recipes/la_razon_bo.recipe Normal file
View File

@ -0,0 +1,64 @@
__license__ = 'GPL v3'
__copyright__ = '2010, Darko Miletic <darko.miletic at gmail.com>'
'''
www.la-razon.com
'''
from calibre import strftime
from calibre.web.feeds.news import BasicNewsRecipe
class LaRazon_Bol(BasicNewsRecipe):
title = 'La Razón - Bolivia'
__author__ = 'Darko Miletic'
description = 'El diario nacional de Bolivia'
publisher = 'Praxsis S.R.L.'
category = 'news, politics, Bolivia'
oldest_article = 1
max_articles_per_feed = 200
no_stylesheets = True
encoding = 'cp1252'
use_embedded_content = False
language = 'es'
publication_type = 'newspaper'
delay = 1
remove_empty_feeds = True
cover_url = strftime('http://www.la-razon.com/portadas/%Y%m%d_LaRazon.jpg')
masthead_url = 'http://www.la-razon.com/imagenes/logo.jpg'
extra_css = """ body{font-family: Arial,Helvetica,sans-serif }
img{margin-bottom: 0.4em}
.noticia-titulo{font-family: Georgia,"Times New Roman",Times,serif}
.lead{font-weight: bold; font-size: 0.8em}
"""
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
}
keep_only_tags = [dict(name='div', attrs={'class':['noticia-titulo','noticia-desarrollo']})]
remove_tags = [dict(name=['meta','link','form','iframe','embed','object'])]
remove_attributes = ['width','height']
feeds = [
(u'Editorial' , u'http://www.la-razon.com/rss_editorial.php' )
,(u'Opinión' , u'http://www.la-razon.com/rss_opinion.php' )
,(u'Nacional' , u'http://www.la-razon.com/rss_nacional.php' )
,(u'Economia' , u'http://www.la-razon.com/rss_economia.php' )
,(u'Ciudades' , u'http://www.la-razon.com/rss_ciudades.php' )
,(u'Sociedad' , u'http://www.la-razon.com/rss_sociedad.php' )
,(u'Mundo' , u'http://www.la-razon.com/rss_sociedad.php' )
,(u'La Revista' , u'http://www.la-razon.com/rss_larevista.php' )
,(u'Sociales' , u'http://www.la-razon.com/rss_sociales.php' )
,(u'Mia' , u'http://www.la-razon.com/rss_mia.php' )
,(u'Marcas' , u'http://www.la-razon.com/rss_marcas.php' )
,(u'Escape' , u'http://www.la-razon.com/rss_escape.php' )
,(u'El Financiero' , u'http://www.la-razon.com/rss_financiero.php')
,(u'Tendencias' , u'http://www.la-razon.com/rss_tendencias.php')
]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return soup

63
resources/recipes/los_tiempos_bo.recipe Normal file
View File

@ -0,0 +1,63 @@
__license__ = 'GPL v3'
__copyright__ = '2010, Darko Miletic <darko.miletic at gmail.com>'
'''
www.lostiempos.com
'''
from calibre import strftime
from calibre.web.feeds.news import BasicNewsRecipe
class LosTiempos_Bol(BasicNewsRecipe):
title = 'Los Tiempos - Bolivia'
__author__ = 'Darko Miletic'
description = 'El periódico de mayor circulación en la ciudad de Cochabamba, Bolivia'
publisher = 'Los Tiempos'
category = 'news, politics, Bolivia'
oldest_article = 1
max_articles_per_feed = 200
no_stylesheets = True
encoding = 'cp1252'
use_embedded_content = False
language = 'es'
publication_type = 'newspaper'
delay = 1
remove_empty_feeds = True
cover_url = strftime('http://www.lostiempos.com/media_recortes/%Y/%m/%d/portada_md_1.jpg')
masthead_url = 'http://www.lostiempos.com/img_stat/logo_tiempos_sin_beta.jpg'
extra_css = """ body{font-family: Arial,Helvetica,sans-serif }
img{margin-bottom: 0.4em}
h1,.hora,.breadcum,.pie_foto{font-family: Georgia,"Times New Roman",Times,serif}
.hora,.breadcum,.pie_foto{font-size: small}
.en_gris,.pie_foto{color: #666666}
"""
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
}
keep_only_tags = [dict(name='div', attrs={'id':'articulo'})]
remove_tags = [
dict(name=['meta','link','form','iframe','embed','object','hr'])
,dict(attrs={'class':['caja_fonts sin_border_bot','pub']})
]
remove_attributes = ['width','height']
feeds = [
(u'Nacional' , u'http://www.lostiempos.com/rss/lostiempos-nacional.xml' )
,(u'Local' , u'http://www.lostiempos.com/rss/lostiempos-local.xml' )
,(u'Deportes' , u'http://www.lostiempos.com/rss/lostiempos-deportes.xml' )
,(u'Economía' , u'http://www.lostiempos.com/rss/lostiempos-economia.xml' )
,(u'Internacional' , u'http://www.lostiempos.com/rss/lostiempos-internacional.xml' )
,(u'Vida y Futuro' , u'http://www.lostiempos.com/rss/lostiempos-vida-y-futuro.xml' )
,(u'Tragaluz' , u'http://www.lostiempos.com/rss/lostiempos-tragaluz.xml' )
,(u'Opiniones' , u'http://www.lostiempos.com/rss/lostiempos-opiniones.xml' )
]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return soup