Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-08 18:54:09 -04:00
Code Issues Packages Projects Releases Wiki Activity

Fix #8487 (NRC Handelsblad feed is empty)

Browse Source
This commit is contained in:
Kovid Goyal 2011-01-21 19:05:33 -07:00
parent 313370ecd9
commit 1a1af35af1
1 changed files with 43 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

58
resources/recipes/nrc.nl.recipe
View File

@ -1,5 +1,5 @@
__license__ = 'GPL v3' __license__ = 'GPL v3'
__copyright__ = '2010, Darko Miletic <darko.miletic at gmail.com>' __copyright__ = '2010-2011, Darko Miletic <darko.miletic at gmail.com>'
''' '''
nrc.nl nrc.nl
''' '''
@ -15,13 +15,18 @@ class Pagina12(BasicNewsRecipe):
oldest_article = 2 oldest_article = 2
max_articles_per_feed = 200 max_articles_per_feed = 200
no_stylesheets = True no_stylesheets = True
encoding = 'cp1252' encoding = 'utf8'
use_embedded_content = False use_embedded_content = False
language = 'nl' language = 'nl'
country = 'NL' country = 'NL'
remove_empty_feeds = True remove_empty_feeds = True
masthead_url = 'http://www.nrc.nl/nrc.nl/images/logo_nrc.png' masthead_url = 'http://www.nrc.nl/nrc.nl/images/logo_nrc.png'
extra_css = ' body{font-family: Verdana,Arial,Helvetica,sans-serif } img{margin-bottom: 0.4em} h1,h2,h3{text-align:left} ' extra_css = """
body{font-family: Georgia,serif }
img{margin-bottom: 0.4em; display: block}
.bijschrift,.sectie{font-size: x-small}
.sectie{color: gray}
"""
conversion_options = { conversion_options = {
'comment' : description 'comment' : description
@ -30,21 +35,42 @@ class Pagina12(BasicNewsRecipe):
, 'language' : language , 'language' : language
} }
keep_only_tags = [dict(name='div',attrs={'class':'article clearfix'})] keep_only_tags = [dict(attrs={'class':'uitstekendekeus'})]
remove_tags = [
dict(name=['meta','base','link','object','embed'])
,dict(attrs={'class':['reclamespace','tags-and-sharing']})
]
remove_attributes=['lang']
feeds = [ feeds = [
(u'Voorpagina' , u'http://feeds.feedburner.com/NRCHandelsbladVoorpagina' ) (u'Voor nieuws', u'http://www.nrc.nl/nieuws/categorie/nieuws/rss.php' )
,(u'Binnenland' , u'http://feeds.feedburner.com/NRCHandelsbladBinnenland' ) ,(u'Binnenland' , u'http://www.nrc.nl/nieuws/categorie/binnenland/rss.php' )
,(u'Buitenland' , u'http://feeds.feedburner.com/NRCHandelsbladBuitenland' ) ,(u'Buitenland' , u'http://www.nrc.nl/nieuws/categorie/buitenland/rss.php' )
,(u'Economie' , u'http://feeds.feedburner.com/NRCHandelsbladEconomie' ) ,(u'Economie' , u'http://www.nrc.nl/nieuws/categorie/economie/rss.php' )
,(u'Kunst & Film' , u'http://feeds.feedburner.com/nrc/NRCHandelsbladKunstEnFilm') ,(u'Cultuur' , u'http://www.nrc.nl/nieuws/categorie/cultuur/rss.php' )
,(u'Sport' , u'http://feeds.feedburner.com/NRCHandelsbladSport' ) ,(u'Sport' , u'http://www.nrc.nl/nieuws/categorie/sport/rss.php' )
,(u'Wetenschap ' , u'http://www.nrc.nl/rss/wetenschap' ) ,(u'Wetenschap ', u'http://www.nrc.nl/nieuws/categorie/wetenschap-nieuws/rss.php')
] ]
def print_version(self, url):
return url + '?service=Print'
def preprocess_html(self, soup): def preprocess_html(self, soup):
return self.adeify_images(soup) for item in soup.findAll(style=True):
del item['style']
for item in soup.findAll('a'):
limg = item.find('img')
if item.string is not None:
str = item.string
item.replaceWith(str)
else:
if limg:
item.name = 'div'
atritems =['href','target','rel']
for atit in atritems:
if item.has_key(atit):
del item[atit]
else:
str = self.tag_to_string(item)
item.replaceWith(str)
for item in soup.findAll('img'):
if not item.has_key('alt'):
item['alt'] = 'image'
return soup