Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

ABC and Ultima Hora by Darko Miletic

Browse Source
This commit is contained in:
Kovid Goyal 2010-04-28 15:46:42 -06:00
parent 1913404ddf
commit 67f87e193d
5 changed files with 102 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

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

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 2.6 KiB

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

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 690 B

49
resources/recipes/abc_py.recipe Normal file
View File

@ -0,0 +1,49 @@
__license__ = 'GPL v3'
__copyright__ = '2010, Darko Miletic <darko.miletic at gmail.com>'
'''
abc.com.py
'''
from calibre.web.feeds.news import BasicNewsRecipe
class ABC_py(BasicNewsRecipe):
title = 'ABC digital'
__author__ = 'Darko Miletic'
description = 'Noticias de Paraguay y el resto del mundo'
publisher = 'ABC'
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'
extra_css = ' body{font-family: Arial,Helvetica,sans-serif } img{margin-bottom: 0.4em} '
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
}
remove_tags = [dict(name=['form','iframe','embed','object','link','base','table']),dict(attrs={'class':'toolbox'})]
remove_tags_after = dict(attrs={'class':'date'})
keep_only_tags = [dict(attrs={'class':'zcontent'})]
feeds = [
(u'Ultimo momento' , u'http://www.abc.com.py/ultimo-momento.xml' )
,(u'Nacionales' , u'http://www.abc.com.py/nacionales.xml' )
,(u'Internacionales' , u'http://www.abc.com.py/internacionales.xml' )
,(u'Deportes' , u'http://www.abc.com.py/deportes.xml' )
,(u'Espectaculos' , u'http://www.abc.com.py/espectaculos.xml' )
,(u'Ciencia y Tecnologia', u'http://www.abc.com.py/ciencia-y-tecnologia.xml')
]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return soup

52
resources/recipes/ultimahora.recipe Normal file
View File

@ -0,0 +1,52 @@
__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} '
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

1
src/calibre/utils/localization.py
View File

@ -106,6 +106,7 @@ _extra_lang_codes = {
'en_SG' : _('English (Singapore)'),
'en_YE' : _('English (Yemen)'),
'en_IE' : _('English (Ireland)'),
'es_PY' : _('Spanish (Paraguay)'),
'de_AT' : _('German (AT)'),
'nl' : _('Dutch (NL)'),
'nl_BE' : _('Dutch (BE)'),