calibre/resources/recipes/adevarul.recipe
2011-02-15 09:14:28 -07:00

51 lines
1.7 KiB
Plaintext

# -*- coding: utf-8 -*-
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = u'2011, Silviu Cotoar\u0103'
'''
adevarul.ro
'''
from calibre.web.feeds.news import BasicNewsRecipe
class Adevarul(BasicNewsRecipe):
title = u'Adev\u0103rul'
language = 'ro'
__author__ = u'Silviu Cotoar\u0103'
description = u'\u0218tiri din Rom\u00e2nia'
publisher = 'Adevarul'
category = 'Ziare,Stiri,Romania'
oldest_article = 5
max_articles_per_feed = 100
no_stylesheets = True
use_embedded_content = False
encoding = 'utf-8'
remove_javascript = True
cover_url = 'http://upload.wikimedia.org/wikipedia/en/d/d6/Logo_noul_adevarul.png'
conversion_options = {
'comments' : description
,'tags' : category
,'language' : language
,'publisher' : publisher
}
keep_only_tags = [ dict(name='div', attrs={'class':'article_header'})
,dict(name='div', attrs={'class':'bd'})
]
remove_tags = [ dict(name='div', attrs={'class':'bb-wg-article_related_attachements'})
,dict(name='div', attrs={'class':'bb-md bb-md-article_comments'})
,dict(name='form', attrs={'id':'bb-comment-create-form'})
]
remove_tags_after = [ dict(name='form', attrs={'id':'bb-comment-create-form'}) ]
feeds = [ (u'\u0218tiri', u'http://www.adevarul.ro/rss/latest') ]
def preprocess_html(self, soup):
return self.adeify_images(soup)