mirror of
https://github.com/kovidgoyal/calibre.git
synced 2025-06-23 15:30:45 -04:00
47 lines
1.7 KiB
Python
47 lines
1.7 KiB
Python
#!/usr/bin/env python2
|
|
# -*- coding: utf-8 -*-
|
|
|
|
__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': 'bb-tu first-t bb-article-body'})
|
|
]
|
|
|
|
remove_tags = [
|
|
dict(name='li', attrs={'class': 'author'}), dict(name='li', attrs={'class': 'date'}), dict(name='li', attrs={'class': 'comments'}), 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'}), dict(name='div', attrs={'id': 'mediatag'}), dict(name='div', attrs={'id': 'ft'}), dict(name='div', attrs={'id': 'comment_wrapper'}) # noqa
|
|
]
|
|
|
|
remove_tags_after = [
|
|
dict(name='div', attrs={'id': 'comment_wrapper'}),
|
|
]
|
|
|
|
feeds = [(u'\u0218tiri', u'http://www.adevarul.ro/rss/latest')]
|
|
|
|
def preprocess_html(self, soup):
|
|
return self.adeify_images(soup)
|