Fix #4869 (Updated recipe for Danas newspaper)

This commit is contained in:
Kovid Goyal 2010-02-11 09:39:14 -07:00
parent 81f58b185a
commit 4b670c5f30

View File

@ -1,64 +1,63 @@
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2008-2009, Darko Miletic <darko.miletic at gmail.com>'
__copyright__ = '2008-2010, Darko Miletic <darko.miletic at gmail.com>'
'''
danas.rs
'''
import re
from calibre.web.feeds.news import BasicNewsRecipe
from calibre.ebooks.BeautifulSoup import Tag
class Danas(BasicNewsRecipe):
title = 'Danas'
__author__ = 'Darko Miletic'
description = 'Vesti'
description = 'Dnevne novine sa vestima iz sveta, politike, ekonomije, kulture, sporta, Beograda, Novog Sada i cele Srbije.'
publisher = 'Danas d.o.o.'
category = 'news, politics, Serbia'
oldest_article = 2
max_articles_per_feed = 100
no_stylesheets = False
use_embedded_content = False
encoding = 'utf-8'
masthead_url = 'http://www.danas.rs/images/basic/danas.gif'
language = 'sr'
lang = 'sr-Latn-RS'
direction = 'ltr'
extra_css = '@font-face {font-family: "serif1";src:url(res:///opt/sony/ebook/FONT/tt0011m_.ttf)} @font-face {font-family: "sans1";src:url(res:///opt/sony/ebook/FONT/tt0003m_.ttf)} body{font-family: serif1, serif} .article_description{font-family: sans1, sans-serif}'
extra_css = '@font-face {font-family: "serif1";src:url(res:///opt/sony/ebook/FONT/tt0011m_.ttf)} @font-face {font-family: "sans1";src:url(res:///opt/sony/ebook/FONT/tt0003m_.ttf)} .article_description,body,.lokacija{font-family: Tahoma,Arial,Helvetica,sans1,sans-serif} .nadNaslov,h1,.preamble{font-family: Georgia,"Times New Roman",Times,serif1,serif} .antrfileText{border-left: 2px solid #999999; color:#666666; margin-left: 0.8em; padding-left: 1.2em; margin-bottom: 0; margin-top: 0} h2,.datum,.lokacija,.autor{font-size: small} .antrfileNaslov{border-left: 2px solid #999999; color:#666666; margin-left: 0.8em; padding-left: 1.2em; font-weight:bold; margin-bottom: 0; margin-top: 0} img{margin-bottom: 0.8em} '
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
, 'pretty_print' : True
}
preprocess_regexps = [(re.compile(u'\u0110'), lambda match: u'\u00D0')]
keep_only_tags = [dict(name='div', attrs={'id':'left'})]
remove_tags = [
dict(name='div', attrs={'class':['width_1_4','metaClanka','baner']})
,dict(name='div', attrs={'id':'comments'})
,dict(name=['object','link'])
,dict(name=['object','link','iframe'])
]
feeds = [
(u'Vesti' , u'http://www.danas.rs/rss/rss.asp' )
,(u'Periskop', u'http://www.danas.rs/rss/rss.asp?column_id=4')
feeds = [
(u'Politika' , u'http://www.danas.rs/rss/rss.asp?column_id=27')
,(u'Hronika' , u'http://www.danas.rs/rss/rss.asp?column_id=2' )
,(u'Drustvo' , u'http://www.danas.rs/rss/rss.asp?column_id=24')
,(u'Dijalog' , u'http://www.danas.rs/rss/rss.asp?column_id=1' )
,(u'Ekonomija', u'http://www.danas.rs/rss/rss.asp?column_id=6' )
,(u'Svet' , u'http://www.danas.rs/rss/rss.asp?column_id=25')
,(u'Srbija' , u'http://www.danas.rs/rss/rss.asp?column_id=28')
,(u'Kultura' , u'http://www.danas.rs/rss/rss.asp?column_id=5' )
,(u'Sport' , u'http://www.danas.rs/rss/rss.asp?column_id=13')
,(u'Scena' , u'http://www.danas.rs/rss/rss.asp?column_id=42')
,(u'Feljton' , u'http://www.danas.rs/rss/rss.asp?column_id=19')
,(u'Periskop' , u'http://www.danas.rs/rss/rss.asp?column_id=4' )
]
def preprocess_html(self, soup):
mlang = Tag(soup,'meta',[("http-equiv","Content-Language"),("content",self.lang)])
soup.head.insert(0,mlang)
attribs = [ 'style','font','valign'
,'colspan','width','height'
,'rowspan','summary','align'
,'cellspacing','cellpadding'
,'frames','rules','border'
]
for item in soup.body.findAll(name=['table','td','tr','th','caption','thead','tfoot','tbody','colgroup','col']):
item.name = 'div'
for attrib in attribs:
if item.has_key(attrib):
del item[attrib]
for item in soup.findAll(style=True):
del item['style']
return soup
def print_version(self, url):
return url + '&action=print'