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

New recipe for Publico by David Rodriguez

Browse Source
This commit is contained in:
Kovid Goyal 2009-06-03 09:13:01 -07:00
parent 76dd755dd5
commit ccba32d733
2 changed files with 41 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/calibre/web/feeds/recipes/__init__.py
View File

@ -44,7 +44,7 @@ recipe_modules = ['recipe_' + r for r in (
'stackoverflow', 'telepolis_artikel', 'zaobao', 'usnews',
'straitstimes', 'index_hu', 'pcworld_hu', 'hrt', 'rts',
'h1', 'h2', 'h3', 'phd_comics', 'woz_die', 'elektrolese',
'climate_progress', 'carta', 'slashdot',
'climate_progress', 'carta', 'slashdot', 'publico',
)]
import re, imp, inspect, time, os

40
src/calibre/web/feeds/recipes/recipe_publico.py Normal file
View File

@ -0,0 +1,40 @@
"""
publico.py - v1.0
Copyright (c) 2009, David Rodrigues - http://sixhat.net
All rights reserved.
"""
__license__ = 'GPL 3'
from calibre.web.feeds.news import BasicNewsRecipe
import re
class Publico(BasicNewsRecipe):
title = u'P\xc3\xbablico'
__author__ = 'David Rodrigues'
oldest_article = 1
max_articles_per_feed = 30
encoding='utf-8'
no_stylesheets = True
language = _('Portuguese')
preprocess_regexps = [(re.compile(u"\uFFFD", re.DOTALL|re.IGNORECASE), lambda match: ''),]
feeds = [
(u'Geral', u'http://feeds.feedburner.com/PublicoUltimaHora'),
(u'Internacional', u'http://www.publico.clix.pt/rss.ashx?idCanal=11'),
(u'Política', u'http://www.publico.clix.pt/rss.ashx?idCanal=12'),
(u'Ciências', 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'Educação', 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')]
def print_version(self,url):
s=re.findall("id=[0-9]+",url);
return "http://ww2.publico.clix.pt/print.aspx?"+s[0]