mirror of
https://github.com/kovidgoyal/calibre.git
synced 2025-06-23 15:30:45 -04:00
184 lines
7.4 KiB
Python
184 lines
7.4 KiB
Python
#!/usr/bin/env python
|
|
__license__ = 'GPL v3'
|
|
__copyright__ = '08 Januery 2011, desUBIKado'
|
|
__author__ = 'desUBIKado'
|
|
__description__ = 'Daily newspaper from Biscay'
|
|
__version__ = 'v0.14'
|
|
__date__ = '10, September 2017'
|
|
'''
|
|
http://www.elcorreo.com/
|
|
'''
|
|
|
|
import time
|
|
import re
|
|
from calibre.web.feeds.news import BasicNewsRecipe
|
|
|
|
|
|
class elcorreo(BasicNewsRecipe):
|
|
__author__ = 'desUBIKado'
|
|
description = 'Daily newspaper from Biscay'
|
|
title = u'El Correo'
|
|
publisher = 'Vocento'
|
|
category = 'News, politics, culture, economy, general interest'
|
|
oldest_article = 1
|
|
delay = 1
|
|
max_articles_per_feed = 100
|
|
no_stylesheets = True
|
|
use_embedded_content = False
|
|
masthead_url = 'http://www.elcorreo.com/vizcaya/noticias/201002/02/Media/logo-elcorreo-nuevo.png'
|
|
language = 'es'
|
|
timefmt = '[%a, %d %b, %Y]'
|
|
encoding = 'utf-8'
|
|
remove_empty_feeds = True
|
|
remove_javascript = True
|
|
|
|
feeds = [
|
|
(u'Portada', u'http://www.elcorreo.com/rss/atom/portada'),
|
|
(u'Mundo', u'http://www.elcorreo.com/rss/atom/?section=internacional'),
|
|
(u'Bizkaia', u'http://www.elcorreo.com/rss/atom/?section=bizkaia'),
|
|
(u'Guipuzkoa', u'http://www.elcorreo.com/rss/atom/?section=gipuzkoa'),
|
|
(u'Araba', u'http://www.elcorreo.com/rss/atom/?section=araba'),
|
|
(u'La Rioja', u'http://www.elcorreo.com/rss/atom/?section=larioja'),
|
|
(u'Miranda', u'http://www.elcorreo.com/rss/atom/?section=miranda'),
|
|
(u'Economía', u'http://www.elcorreo.com/rss/atom/?section=economia'),
|
|
(u'Culturas', u'http://www.elcorreo.com/rss/atom/?section=culturas'),
|
|
(u'Politica', u'http://www.elcorreo.com/rss/atom/?section=politica'),
|
|
(u'Tecnología', u'http://www.elcorreo.com/rss/atom/?section=tecnologia'),
|
|
(u'Gente - Estilo', u'http://www.elcorreo.com/rss/atom/?section=gente-estilo'),
|
|
(u'Planes', u'http://www.elcorreo.com/rss/atom/?section=planes'),
|
|
(u'Athletic', u'http://www.elcorreo.com/rss/atom/?section=athletic'),
|
|
(u'Alavés', u'http://www.elcorreo.com/rss/atom/?section=alaves'),
|
|
(u'Bilbao Basket', u'http://www.elcorreo.com/rss/atom/?section=bilbaobasket'),
|
|
(u'Baskonia', u'http://www.elcorreo.com/rss/atom/?section=baskonia'),
|
|
(u'Deportes', u'http://www.elcorreo.com/rss/atom/?section=deportes'),
|
|
(u'Jaiak', u'http://www.elcorreo.com/rss/atom/?section=jaiak'),
|
|
(u'La Blanca', u'http://www.elcorreo.com/rss/atom/?section=la-blanca-vitoria'),
|
|
(u'Aste Nagusia', u'http://www.elcorreo.com/rss/atom/?section=aste-nagusia-bilbao'),
|
|
(u'Semana Santa', u'http://www.elcorreo.com/rss/atom/?section=semana-santa'),
|
|
(u'Festivales', u'http://www.elcorreo.com/rss/atom/?section=festivales')
|
|
]
|
|
|
|
keep_only_tags = [
|
|
dict(name='div', attrs={'class': ['col-xs-12 col-sm-12 col-md-8 col-lg-8']})
|
|
]
|
|
|
|
remove_tags = [
|
|
dict(
|
|
name='div',
|
|
attrs={
|
|
'class': [
|
|
'voc-topics voc-detail-grid ', 'voc-newsletter ',
|
|
'voc-author-social'
|
|
]
|
|
}
|
|
),
|
|
dict(name='section', attrs={'class': ['voc-ficha-detail voc-file-sports']})
|
|
]
|
|
|
|
remove_tags_before = dict(
|
|
name='div', attrs={'class': 'col-xs-12 col-sm-12 col-md-8 col-lg-8'}
|
|
)
|
|
remove_tags_after = dict(
|
|
name='div', attrs={'class': 'col-xs-12 col-sm-12 col-md-8 col-lg-8'}
|
|
)
|
|
|
|
_processed_links = []
|
|
|
|
def get_article_url(self, article):
|
|
link = article.get('link', None)
|
|
|
|
if link is None:
|
|
return article
|
|
|
|
# modificamos la url de las noticias de los equipos deportivos para que funcionen, por ejemplo:
|
|
# http://athletic.elcorreo.com/noticias/201407/27/muniain-estrella-athletic-para-20140727093046.html
|
|
# http://m.elcorreo.com/noticias/201407/27/muniain-estrella-athletic-para-20140727093046.html?external=deportes/athletic
|
|
|
|
parte = link.split('/')
|
|
|
|
if parte[2] == 'athletic.elcorreo.com':
|
|
link = 'http://www.elcorreo.com/' + parte[3] + '/' + parte[
|
|
4
|
|
] + '/' + parte[5] + '/' + parte[6] + '?external=deportes/athletic'
|
|
else:
|
|
if parte[2] == 'baskonia.elcorreo.com':
|
|
link = 'http://www.elcorreo.com/' + parte[3] + '/' + parte[
|
|
4
|
|
] + '/' + parte[5] + '/' + parte[6] + '?external=deportes/baskonia'
|
|
else:
|
|
if parte[2] == 'bilbaobasket.elcorreo.com':
|
|
link = 'http://www.elcorreo.com/' + parte[3] + '/' + parte[
|
|
4
|
|
] + '/' + parte[5] + '/' + parte[
|
|
6
|
|
] + '?external=deportes/bilbaobasket'
|
|
else:
|
|
if parte[2] == 'alaves.elcorreo.com':
|
|
link = 'http://www.elcorreo.com/' + parte[3] + '/' + parte[
|
|
4
|
|
] + '/' + parte[5] + '/' + parte[
|
|
6
|
|
] + '?external=deportes/alaves'
|
|
|
|
# A veces el mismo articulo aparece en la versión de Alava y en la de Bizkaia. Por ejemplo:
|
|
# http://www.elcorreo.com/alava/deportes/motor/formula-1/201407/27/ecclestone-quiere-briatore-ayude-20140727140820-rc.html
|
|
# http://www.elcorreo.com/bizkaia/deportes/motor/formula-1/201407/27/ecclestone-quiere-briatore-ayude-20140727140820-rc.html
|
|
# para controlar los duplicados, unificamos las url para que sean siempre de bizkaia (excepto para la sección "araba")
|
|
|
|
if ((parte[3] == 'alava') and (parte[4] != 'araba')):
|
|
link = link.replace('elcorreo.com/alava', 'elcorreo.com/bizkaia')
|
|
|
|
# Controlamos si el artículo ha sido incluido en otro feed para eliminarlo
|
|
|
|
if not (link in self._processed_links):
|
|
self._processed_links.append(link)
|
|
else:
|
|
link = None
|
|
|
|
return link
|
|
|
|
# Recuperamos la portada de papel (la imagen format=1 tiene mayor resolucion)
|
|
|
|
def get_cover_url(self):
|
|
cover = None
|
|
st = time.localtime()
|
|
year = str(st.tm_year)
|
|
month = "%.2d" % st.tm_mon
|
|
day = "%.2d" % st.tm_mday
|
|
# http://info.elcorreo.com/pdf/07082013-viz.pdf
|
|
cover = 'http://info.elcorreo.com/pdf/' + day + month + year + '-viz.pdf'
|
|
br = BasicNewsRecipe.get_browser(self)
|
|
try:
|
|
br.open(cover)
|
|
except:
|
|
self.log("\nPortada no disponible")
|
|
cover = 'http://www.elcorreo.com/vizcaya/noticias/201002/02/Media/logo-elcorreo-nuevo.png'
|
|
return cover
|
|
|
|
# Para cambiar el estilo del texto
|
|
|
|
extra_css = '''
|
|
h1 {font-family:Arial,Helvetica,sans-serif; font-weight:bold;font-size:28px;}
|
|
h2 {font-family:georgia,serif; font-style:italic; font-weight:normal;font-size:16px;color:#4D4D4D;}
|
|
h3 {font-family:georgia,serif; font-weight:bold;font-size:18px;}
|
|
'''
|
|
|
|
preprocess_regexps = [
|
|
|
|
# Para presentar la imagen de los video incrustados
|
|
(
|
|
re.compile(r'stillURLVideo: \'', re.DOTALL | re.IGNORECASE),
|
|
lambda match: '</script><img src="'
|
|
),
|
|
(
|
|
re.compile(r'.jpg\',', re.DOTALL | re.IGNORECASE),
|
|
lambda match: '.jpg"><SCRIPT TYPE="text/JavaScript"'
|
|
),
|
|
|
|
# Para quitar el punto de la lista
|
|
(
|
|
re.compile(r'<li class="destacada">', re.DOTALL | re.IGNORECASE),
|
|
lambda match: '<div class="destacada"></div>'
|
|
)
|
|
]
|