Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-06-23 15:30:45 -04:00
Code Issues Packages Projects Releases Wiki Activity

Updated La Republica

Browse Source
This commit is contained in:
Kovid Goyal 2010-04-21 06:00:28 -06:00
parent 8057e6cc79
commit 17da75cc50
1 changed files with 28 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
resources/recipes/la_republica.recipe
View File

@ -22,20 +22,35 @@ class LaRepublica(BasicNewsRecipe):
language = 'it'
timefmt = '[%a, %d %b, %Y]'
oldest_article = 1
oldest_article = 5
max_articles_per_feed = 100
use_embedded_content = False
recursion = 10
remove_javascript = True
def get_article_url(self, article):
link = article.get('id', article.get('guid', None))
if link is None:
return article
return link
keep_only_tags = [dict(name='div', attrs={'class':'articolo'}),
dict(name='div', attrs={'class':'body-text'}),
dict(name='div', attrs={'class':'page-content'}),
dict(name='div', attrs={'id':'contA'})
]
keep_only_tags = [dict(name='div', attrs={'class':'articolo'})]
remove_tags = [
dict(name=['object','link']),
dict(name='span',attrs={'class':'linkindice'}),
dict(name='div',attrs={'class':'bottom-mobile'}),
dict(name='div',attrs={'id':['rssdiv','blocco']})
dict(name='div', attrs={'class':'bottom-mobile'}),
dict(name='div', attrs={'id':['rssdiv','blocco']}),
dict(name='div', attrs={'class':'utility'}),
dict(name='div', attrs={'class':'generalbox'})
]
remove_tags_after = [
dict(name='div',attrs={'id':'ugc_linkUpload'})
]
feeds = [
@ -51,5 +66,9 @@ class LaRepublica(BasicNewsRecipe):
(u'Repubblica Cultura', u'http://www.repubblica.it/rss/spettacoli_e_cultura/rss2.0.xml'),
(u'Repubblica Persone', u'http://www.repubblica.it/rss/persone/rss2.0.xml'),
(u'Repubblica Sport', u'http://www.repubblica.it/rss/sport/rss2.0.xml'),
(u'Repubblica Calcio', u'http://www.repubblica.it/rss/sport/calcio/rss2.0.xml')
(u'Repubblica Calcio', u'http://www.repubblica.it/rss/sport/calcio/rss2.0.xml'),
(u'Repubblica Motori', u'http://www.repubblica.it/rss/motori/rss2.0.xml'),
(u'Repubblica Roma', u'http://roma.repubblica.it/rss/rss2.0.xml'),
(u'Repubblica Torino', u'http://torino.repubblica.it/rss/rss2.0.xml')
]