calibre/recipes/zaman.recipe
2012-12-04 09:24:43 +05:30

69 lines
5.3 KiB
Plaintext
Raw Permalink Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

# -*- coding: utf-8 -*-
from calibre.web.feeds.news import BasicNewsRecipe
class Zaman (BasicNewsRecipe):
title = u'ZAMAN Gazetesi'
description = ' Zaman Gazetesi''nin internet sitesinden günlük haberler'
__author__ = u'thomass'
oldest_article = 2
max_articles_per_feed =50
no_stylesheets = True
#delay = 1
use_embedded_content = False
encoding = 'utf-8'
publisher = 'Feza Gazetecilik'
category = 'news, haberler,TR,gazete'
language = 'tr'
publication_type = 'newspaper '
extra_css = 'h1{text-transform: capitalize; font-weight: bold; font-size: 22px;color:#0000FF} p{text-align:justify} ' #.introduction{font-weight: bold} .story-feature{display: block; padding: 0; border: 1px solid; width: 40%; font-size: small} .story-feature h2{text-align: center; text-transform: uppercase} '
conversion_options = {
'tags' : category
,'language' : language
,'publisher' : publisher
,'linearize_tables': True
}
cover_img_url = 'https://fbcdn-profile-a.akamaihd.net/hprofile-ak-snc4/188140_81722291869_2111820_n.jpg'
masthead_url = 'http://medya.zaman.com.tr/extentions/zaman.com.tr/img/section/logo-section.png'
ignore_duplicate_articles = { 'title', 'url' }
auto_cleanup = False
remove_empty_feeds= True
#keep_only_tags = [dict(name='div', attrs={'id':[ 'contentposition19']})]#,dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'news-detail-content']}), dict(name='td', attrs={'class':['columnist-detail','columnist_head']}), ]
remove_tags = [ dict(name='img', attrs={'src':['http://cmsmedya.zaman.com.tr/images/logo/logo.bmp']}),dict(name='hr', attrs={'class':['interactive-hr']})]# remove_tags = [ dict(name='div', attrs={'class':[ 'detayUyari']}),dict(name='div', attrs={'class':[ 'detayYorum']}),dict(name='div', attrs={'class':[ 'addthis_toolbox addthis_default_style ']}),dict(name='div', attrs={'id':[ 'tumYazi']})]#,dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='div', attrs={'id':[ 'xxx']}),dict(name='img', attrs={'src':['http://medya.zaman.com.tr/zamantryeni/pics/zamanonline.gif']}),dict(name='div', attrs={'class':['radioEmbedBg','radyoProgramAdi']}),dict(name='a', attrs={'class':['webkit-html-attribute-value webkit-html-external-link']}),dict(name='table', attrs={'id':['yaziYorumTablosu']}),dict(name='img', attrs={'src':['http://medya.zaman.com.tr/pics/paylas.gif','http://medya.zaman.com.tr/extentions/zaman.com.tr/img/columnist/ma-16.png']}),dict(name='div', attrs={'id':[ 'news-detail-gallery']}),dict(name='div', attrs={'id':[ 'news-detail-title-bottom-part']}),dict(name='div', attrs={'id':[ 'news-detail-news-paging-main']})]#
#remove_attributes = ['width','height']
remove_empty_feeds= True
feeds = [
( u'Manşet', u'http://www.zaman.com.tr/manset.rss'),
( u'Yazarlar', u'http://www.zaman.com.tr/yazarlar.rss'),
( u'Politika', u'http://www.zaman.com.tr/politika.rss'),
( u'Ekonomi', u'http://www.zaman.com.tr/ekonomi.rss'),
( u'Dış Haberler', u'http://www.zaman.com.tr/dishaberler.rss'),
( u'Son Dakika', u'http://www.zaman.com.tr/sondakika.rss'),
( u'Gündem', u'http://www.zaman.com.tr/gundem.rss'),
( u'Yorumlar', u'http://www.zaman.com.tr/yorumlar.rss'),
( u'Röportaj', u'http://www.zaman.com.tr/roportaj.rss'),
( u'Dizi Yazı', u'http://www.zaman.com.tr/dizi.rss'),
( u'Bilişim', u'http://www.zaman.com.tr/bilisim.rss'),
( u'Otomotiv', u'http://www.zaman.com.tr/otomobil.rss'),
( u'Spor', u'http://www.zaman.com.tr/spor.rss'),
( u'Kürsü', u'http://www.zaman.com.tr/kursu.rss'),
( u'Eğitim', u'http://www.zaman.com.tr/egitim.rss'),
( u'Kültür Sanat', u'http://www.zaman.com.tr/kultursanat.rss'),
( u'Televizyon', u'http://www.zaman.com.tr/televizyon.rss'),
( u'Aile', u'http://www.zaman.com.tr/aile.rss'),
( u'Cuma Eki', u'http://www.zaman.com.tr/cuma.rss'),
( u'Cumaertesi Eki', u'http://www.zaman.com.tr/cumaertesi.rss'),
( u'Pazar Eki', u'http://www.zaman.com.tr/pazar.rss'),
( u'En çok Okunanlar', u'http://www.zaman.com.tr/max_all.rss'),
( u'Anasayfa', u'http://www.zaman.com.tr/anasayfa.rss'),
]
def print_version(self, url):
return url.replace('http://www.zaman.com.tr/newsDetail_getNewsById.action?newsId=', 'http://www.zaman.com.tr/newsDetail_openPrintPage.action?newsId=')