From 252cd2507ceefb16183b932cf20ae29f5a535372 Mon Sep 17 00:00:00 2001 From: Kovid Goyal Date: Tue, 27 Dec 2011 08:58:52 +0530 Subject: [PATCH] Fix #908912 (Updated recipe for La Razon from Bolivia) --- recipes/la_razon_bo.recipe | 52 ++++++++++++++++++-------------------- 1 file changed, 25 insertions(+), 27 deletions(-) diff --git a/recipes/la_razon_bo.recipe b/recipes/la_razon_bo.recipe index 6af899b760..51ec024350 100644 --- a/recipes/la_razon_bo.recipe +++ b/recipes/la_razon_bo.recipe @@ -1,10 +1,9 @@ __license__ = 'GPL v3' -__copyright__ = '2010, Darko Miletic ' +__copyright__ = '2010-2011, Darko Miletic ' ''' www.la-razon.com ''' -from calibre import strftime from calibre.web.feeds.news import BasicNewsRecipe class LaRazon_Bol(BasicNewsRecipe): @@ -16,19 +15,16 @@ class LaRazon_Bol(BasicNewsRecipe): oldest_article = 1 max_articles_per_feed = 200 no_stylesheets = True - encoding = 'cp1252' + encoding = 'utf8' use_embedded_content = False language = 'es_BO' 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} - """ + masthead_url = 'http://www.la-razon.com/static/LRZRazon/images/lrz-logo.png' + extra_css = """ body{font-family: Georgia,"Times New Roman",Times,serif} + img{margin-bottom: 0.4em; display: block} + .meta{font-size: small; font-family: Arial,Helvetica,sans-serif} + """ conversion_options = { 'comment' : description @@ -37,25 +33,27 @@ class LaRazon_Bol(BasicNewsRecipe): , 'language' : language } - keep_only_tags = [dict(name='div', attrs={'class':['noticia-titulo','noticia-desarrollo']})] - remove_tags = [dict(name=['meta','link','form','iframe','embed','object'])] + keep_only_tags = [dict(name='div', attrs={'class':['pg-hd', 'pg-bd']})] + remove_tags = [ + dict(name=['meta','link','form','iframe','embed','object']) + ,dict(name='div', attrs={'class':'bd'}) + ] 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') + (u'Editorial' , u'http://www.la-razon.com/rss/opinion/editorial/' ) + ,(u'Nacional' , u'http://www.la-razon.com/rss/nacional/' ) + ,(u'Economia' , u'http://www.la-razon.com/rss/economia/' ) + ,(u'Ciudades' , u'http://www.la-razon.com/rss/ciudades/' ) + ,(u'Sociedad' , u'http://www.la-razon.com/rss/sociedad/' ) + ,(u'Mundo' , u'http://www.la-razon.com/rss/mundo/' ) + ,(u'La Revista' , u'http://www.la-razon.com/rss/la_revista/' ) + ,(u'Sociales' , u'http://www.la-razon.com/rss/sociales/' ) + ,(u'Mia' , u'http://www.la-razon.com/rss/suplementos/mia/' ) + ,(u'Marcas' , u'http://www.la-razon.com/rss/marcas/' ) + ,(u'Escape' , u'http://www.la-razon.com/rss/suplementos/escape/' ) + ,(u'El Financiero' , u'http://www.la-razon.com/rss/suplementos/financiero/') + ,(u'Tendencias' , u'http://www.la-razon.com/rss/suplementos/tendencias/') ] def preprocess_html(self, soup):