Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

Fix #5854 (Replacement of broken Publico recipe)

Browse Source
This commit is contained in:
Kovid Goyal 2010-06-16 14:48:05 -06:00
parent 61504a8527
commit b0e140317b
1 changed files with 36 additions and 34 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

70
resources/recipes/publico.recipe
View File

@ -1,41 +1,43 @@
"""
publico.py - v1.0
#!/usr/bin/env python
__author__ = u'Jordi Balcells'
__license__ = 'GPL v3'
description = u'Jornal portugu\xeas - v1.03 (16 June 2010)'
__docformat__ = 'restructuredtext en'
Copyright (c) 2009, David Rodrigues - http://sixhat.net
All rights reserved.
"""
__license__ = 'GPL 3'
'''
publico.pt
'''
from calibre.web.feeds.news import BasicNewsRecipe
import re
class Publico(BasicNewsRecipe):
title = u'P\xfablico'
__author__ = 'David Rodrigues'
oldest_article = 1
max_articles_per_feed = 30
encoding='utf-8'
no_stylesheets = True
language = 'pt'
class PublicoPT(BasicNewsRecipe):
description = u'Jornal portugu\xeas'
cover_url = 'http://static.publico.pt/files/header/img/publico.gif'
title = u'Publico.PT'
category = 'News, politics, culture, economy, general interest'
oldest_article = 2
no_stylesheets = True
encoding = 'utf8'
use_embedded_content = False
language = 'pt'
remove_empty_feeds = True
extra_css = ' body{font-family: Arial,Helvetica,sans-serif } img{margin-bottom: 0.4em} '
preprocess_regexps = [(re.compile(u"\uFFFD", re.DOTALL|re.IGNORECASE), lambda match: ''),]
keep_only_tags = [dict(attrs={'class':['content-noticia-title','artigoHeader','ECOSFERA_MANCHETE','noticia','textoPrincipal','ECOSFERA_texto_01']})]
remove_tags = [dict(attrs={'class':['options','subcoluna']})]
feeds = [
(u'Geral', u'http://feeds.feedburner.com/PublicoUltimaHora'),
(u'Internacional', u'http://www.publico.clix.pt/rss.ashx?idCanal=11'),
(u'Pol\xedtica', u'http://www.publico.clix.pt/rss.ashx?idCanal=12'),
(u'Ci\xcencias', u'http://www.publico.clix.pt/rss.ashx?idCanal=13'),
(u'Desporto', u'http://desporto.publico.pt/rss.ashx'),
(u'Economia', u'http://www.publico.clix.pt/rss.ashx?idCanal=57'),
(u'Educa\xe7\xe3o', u'http://www.publico.clix.pt/rss.ashx?idCanal=58'),
(u'Local', u'http://www.publico.clix.pt/rss.ashx?idCanal=59'),
(u'Media e Tecnologia', u'http://www.publico.clix.pt/rss.ashx?idCanal=61'),
(u'Sociedade', u'http://www.publico.clix.pt/rss.ashx?idCanal=62')
]
remove_tags = [dict(name='script'), dict(id='linhaTitulosHeader')]
keep_only_tags = [dict(name='div')]
feeds = [
(u'Geral', u'http://feeds.feedburner.com/publicoRSS'),
(u'Mundo', u'http://feeds.feedburner.com/PublicoMundo'),
(u'Pol\xedtica', u'http://feeds.feedburner.com/PublicoPolitica'),
(u'Economia', u'http://feeds.feedburner.com/PublicoEconomia'),
(u'Desporto', u'http://feeds.feedburner.com/PublicoDesporto'),
(u'Sociedade', u'http://feeds.feedburner.com/PublicoSociedade'),
(u'Educa\xe7\xe3o', u'http://feeds.feedburner.com/PublicoEducacao'),
(u'Ci\xeancias', u'http://feeds.feedburner.com/PublicoCiencias'),
(u'Ecosfera', u'http://feeds.feedburner.com/PublicoEcosfera'),
(u'Cultura', u'http://feeds.feedburner.com/PublicoCultura'),
(u'Local', u'http://feeds.feedburner.com/PublicoLocal'),
(u'Tecnologia', u'http://feeds.feedburner.com/PublicoTecnologia')
]
def print_version(self,url):
s=re.findall("id=[0-9]+",url);
return "http://ww2.publico.clix.pt/print.aspx?"+s[0]