mirror of
https://github.com/kovidgoyal/calibre.git
synced 2025-09-29 15:31:08 -04:00
47 lines
2.4 KiB
Python
47 lines
2.4 KiB
Python
#!/usr/bin/env python
|
|
# vim:fileencoding=utf-8
|
|
|
|
from calibre.web.feeds.news import BasicNewsRecipe
|
|
|
|
|
|
class Unian(BasicNewsRecipe):
|
|
title = '\u0423\u041D\u0406\u0410\u041D'
|
|
description = (
|
|
'\u0423\u041D\u0406\u0410\u041D (\u0423\u043A\u0440\u0430\u0457\u043D\u0441\u044C\u043A\u0435'
|
|
' \u041D\u0435\u0437\u0430\u043B\u0435\u0436\u043D\u0435 \u0406\u043D\u0444\u043E\u0440\u043C\u0430\u0446\u0456\u0439\u043D\u0435'
|
|
' \u0410\u0433\u0435\u043D\u0442\u0441\u0442\u0432\u043E \u041D\u043E\u0432\u0438\u043D) -'
|
|
' \u043F\u0435\u0440\u0448\u0435 \u0432 \u0423\u043A\u0440\u0430\u0457\u043D\u0456 \u0442\u0430'
|
|
' \u043D\u0430\u0439\u0431\u0456\u043B\u044C\u0448\u0435 \u043D\u0435\u0437\u0430\u043B\u0435\u0436\u043D\u0435'
|
|
' \u0456\u043D\u0444\u043E\u0440\u043C\u0430\u0446\u0456\u0439\u043D\u0435 \u0430\u0433\u0435\u043D\u0442\u0441\u0442\u0432\u043E,'
|
|
' \u0437\u0430\u0441\u043D\u043E\u0432\u0430\u043D\u0435 1993 \u0440\u043E\u043A\u0443, \u043B\u0456\u0434\u0435\u0440'
|
|
' \u0441\u0435\u0440\u0435\u0434 \u043D\u043E\u0432\u0438\u043D\u043D\u0438\u0445 \u043C\u0435\u0434\u0456\u0430'
|
|
' \u043A\u0440\u0430\u0457\u043D\u0438, \u043D\u0430\u0439\u0431\u0456\u043B\u044C\u0448'
|
|
' \u0446\u0438\u0442\u043E\u0432\u0430\u043D\u0435 \u0434\u0436\u0435\u0440\u0435\u043B\u043E'
|
|
' \u043D\u043E\u0432\u0438\u043D \u043F\u0440\u043E \u043F\u043E\u0434\u0456\u0457 \u0432 \u043A\u0440\u0430\u0457\u043D\u0456.')
|
|
__author__ = 'bugmen00t'
|
|
publication_type = 'newspaper'
|
|
oldest_article = 7
|
|
max_articles_per_feed = 100
|
|
language = 'uk'
|
|
cover_url = 'https://www.unian.ua/images/unian-512x512.png'
|
|
auto_cleanup = False
|
|
no_stylesheets = True
|
|
|
|
remove_tags_before = dict(name='h1')
|
|
remove_tags_after = dict(name='div', attrs={'class': 'article-text'})
|
|
remove_tags = [
|
|
dict(name='span', attrs={'class': 'article__info-item comments'}),
|
|
dict(name='span', attrs={'class': 'article__info-item views'}),
|
|
dict(name='div', attrs={'class': 'read-also-slider'}),
|
|
dict(name='div', attrs={'class': 'nts-video-wrapper'})
|
|
]
|
|
|
|
feeds = [
|
|
(u'\u0423\u041D\u0406\u0410\u041D', u'https://rss.unian.net/site/news_ukr.rss')
|
|
]
|
|
|
|
def preprocess_html(self, soup):
|
|
for img in soup.findAll('img', attrs={'data-src': True}):
|
|
img['src'] = img['data-src']
|
|
return soup
|