New recipe for Glas Javnosti by Darko Miletic

This commit is contained in:
Kovid Goyal 2009-08-19 14:15:49 -06:00
parent 1a35d88952
commit 642506bba0
3 changed files with 81 additions and 1 deletions

Binary file not shown.

After

Width:  |  Height:  |  Size: 731 B

View File

@ -54,7 +54,7 @@ recipe_modules = ['recipe_' + r for r in (
'fastcompany', 'accountancyage', 'laprensa_hn', 'latribuna',
'eltiempo_hn', 'slate', 'tnxm', 'bbcvietnamese', 'vnexpress',
'volksrant', 'theeconomictimes_india', 'ourdailybread',
'monitor', 'republika', 'beta', 'beta_en',
'monitor', 'republika', 'beta', 'beta_en', 'glasjavnosti',
)]

View File

@ -0,0 +1,80 @@
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2009, Darko Miletic <darko.miletic at gmail.com>'
'''
www.glas-javnosti.rs
'''
import re
from calibre.web.feeds.news import BasicNewsRecipe
class GlasJavnosti(BasicNewsRecipe):
title = 'Glas Javnosti'
__author__ = 'Darko Miletic'
description = 'Glas javnosti - Mi ne ulepsavamo stvarnost'
publisher = 'Glas Javnosti'
category = 'news, politics, Serbia'
oldest_article = 2
max_articles_per_feed = 100
no_stylesheets = False
use_embedded_content = False
language = _('Serbian')
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}'
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : lang
, 'pretty_print' : True
}
preprocess_regexps = [(re.compile(u'\u0110'), lambda match: u'\u00D0')]
keep_only_tags = [
dict(name='div', attrs={'id':'above-content'})
,dict(name='div', attrs={'class':'node' })
]
remove_tags = [
dict(name=['object','link'])
,dict(name='div',attrs={'class':['links','meta']})
,dict(name='div',attrs={'id':'block-block-12'})
]
feeds = [
(u'Politika', u'http://www.glas-javnosti.rs/aktuelni-clanci/2')
,(u'Tema', u'http://www.glas-javnosti.rs/aktuelni-clanci/48')
,(u'Drustvo', u'http://www.glas-javnosti.rs/aktuelni-clanci/17')
,(u'Ekonomija', u'http://www.glas-javnosti.rs/aktuelni-clanci/16')
,(u'Dosije', u'http://www.glas-javnosti.rs/aktuelni-clanci/65')
,(u'Svet', u'http://www.glas-javnosti.rs/aktuelni-clanci/18')
,(u'Hronika', u'http://www.glas-javnosti.rs/aktuelni-clanci/19')
,(u'Kultura', u'http://www.glas-javnosti.rs/aktuelni-clanci/6')
,(u'Ljudi i Dogadjaji', u'http://www.glas-javnosti.rs/aktuelni-clanci/37')
,(u'Putovanja', u'http://www.glas-javnosti.rs/aktuelni-clanci/113')
,(u'Feljton', u'http://www.glas-javnosti.rs/aktuelni-clanci/49')
,(u'Sport', u'http://www.glas-javnosti.rs/aktuelni-clanci/1')
,(u'Lov i Ribolov', u'http://www.glas-javnosti.rs/aktuelni-clanci/591')
,(u'Nedelja', u'http://www.glas-javnosti.rs/aktuelni-clanci/1862')
,(u'Glasno', u'http://www.glas-javnosti.rs/aktuelni-clanci/590')
,(u'Tehnologija', u'http://www.glas-javnosti.rs/aktuelni-clanci/609')
,(u'Reflektor', u'http://www.glas-javnosti.rs/aktuelni-clanci/717')
,(u'Saznanja', u'http://www.glas-javnosti.rs/aktuelni-clanci/1694')
,(u'Beograd', u'http://www.glas-javnosti.rs/aktuelni-clanci/40')
,(u'Srbija', u'http://www.glas-javnosti.rs/aktuelni-clanci/114')
,(u'Zapadna Srbija', u'http://www.glas-javnosti.rs/aktuelni-clanci/41')
,(u'Istocna i Juzna Srbija', u'http://www.glas-javnosti.rs/aktuelni-clanci/42')
,(u'Sumadija i Pomoravlje', u'http://www.glas-javnosti.rs/aktuelni-clanci/43')
,(u'Vojvodina', u'http://www.glas-javnosti.rs/aktuelni-clanci/44')
,(u'Republika Srpska', u'http://www.glas-javnosti.rs/aktuelni-clanci/45')
,(u'Slobodno Vreme', u'http://www.glas-javnosti.rs/aktuelni-clanci/61')
,(u'Konjske Snage', u'http://www.glas-javnosti.rs/aktuelni-clanci/46')
]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return self.adeify_images(soup)