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

Il Giornale by Marini Gabriele

Browse Source
This commit is contained in:
Kovid Goyal 2010-05-01 08:16:10 -06:00
parent b33008a40f
commit 73ba818428
1 changed files with 60 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

60
resources/recipes/il_giornale.recipe Normal file
View File

@ -0,0 +1,60 @@
#!/usr/bin/env python
__license__ = 'GPL v3'
__author__ = 'Gabriele Marini, based on Darko Miletic'
__copyright__ = '2009-2010, Darko Miletic <darko.miletic at gmail.com>'
description = 'Italian daily newspaper - 19-04-2010'
'''
http://www.ilgiornale.it/
'''
from calibre.ebooks.BeautifulSoup import BeautifulSoup
from calibre.web.feeds.news import BasicNewsRecipe
class IlGiornale(BasicNewsRecipe):
__author__ = 'Marini Gabriele'
description = 'Italian daily newspaper'
cover_url = 'http://www.ilgiornale.it/img_v1/logo.gif'
title = u'Il Giornale'
publisher = 'Il Giornale ON-LINE S.r.l.'
category = 'News, politics, culture, economy, general interest'
language = 'it'
timefmt = '[%a, %d %b, %Y]'
oldest_article = 7
max_articles_per_feed = 50
use_embedded_content = False
recursion = 100
no_stylesheets = True
conversion_options = {'linearize_tables':True}
remove_javascript = True
def get_article_url(self, article):
return article.get('guid', article.get('id', None))
def print_version(self, url):
raw = self.browser.open(url).read()
soup = BeautifulSoup(raw.decode('utf8', 'replace'))
all_print_tags = soup.find('div', {'style':'float:left; width:35%;'})
print_link = all_print_tags.contents[1]
if all_print_tags is None:
return url
return print_link['href']
feeds = [
(u'Ultime Notizie',u'http://www.ilgiornale.it/?RSS=S'),
(u'All\'Interno', u'http://www.ilgiornale.it/la_s.pic1?SID=8&RSS=S'),
(u'Esteri', u'http://www.ilgiornale.it/la_s.pic1?SID=6&RSS=S'),
(u'Economia', u'http://www.ilgiornale.it/la_s.pic1?SID=5&RSS=S'),
(u'Cultura', u'http://www.ilgiornale.it/la_s.pic1?SID=4&RSS=S'),
(u'Spettacoli', u'http://www.ilgiornale.it/la_s.pic1?SID=14&RSS=S'),
(u'Sport', u'http://www.ilgiornale.it/la_s.pic1?SID=15&RSS=S'),
(u'Tech&Web', u'http://www.ilgiornale.it/la_s.pic1?SID=35&RSS=S'),
(u'Edizione di Roma', u'http://www.ilgiornale.it/roma.pic1?SID=13&RSS=S'),
(u'Edizione di Milano', u'http://www.ilgiornale.it/milano.pic1?SID=9&RSS=S'),
(u'Edizione di Genova', u'http://www.ilgiornale.it/genova.pic1?SID=7&RSS=S')
]