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

Update Clarin. Fixes #1653113 [Updated recipe for Clarin](https://bugs.launchpad.net/calibre/+bug/1653113)

Browse Source
This commit is contained in:
Kovid Goyal 2016-12-30 08:36:52 +05:30
parent 7c35e7dfa8
commit f4703f602b
2 changed files with 23 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
recipes/clarin.recipe
View File

@ -1,6 +1,10 @@
#!/usr/bin/env python2
# -*- mode: python -*-
# -*- coding: utf-8 -*-
from __future__ import unicode_literals from __future__ import unicode_literals
__license__ = 'GPL v3' __license__ = 'GPL v3'
__copyright__ = '2008-2015, Darko Miletic <darko.miletic at gmail.com>' __copyright__ = '2008-2016, Darko Miletic <darko.miletic at gmail.com>'
''' '''
clarin.com clarin.com
''' '''
@ -27,9 +31,8 @@ class Clarin(BasicNewsRecipe):
needs_subscription = 'optional' needs_subscription = 'optional'
INDEX = 'http://www.clarin.com' INDEX = 'http://www.clarin.com'
LOGIN = 'https://app-pase.clarin.com/pase-registracion/app/pase/ingresarNavegable?execution=e1s1' LOGIN = 'https://app-pase.clarin.com/pase-registracion/app/pase/ingresarNavegable?execution=e1s1'
masthead_url = 'http://www.clarin.com/static/CLAClarinV3/images/logo.png' masthead_url = 'http://www.clarin.com/images/logo_clarin.svg'
cover_url = strftime( cover_url = strftime('http://tapas.clarin.com/tapa/%Y/%m/%d/%Y%m%d_thumb.jpg')
'http://tapas.clarin.com/tapa/%Y/%m/%d/%Y%m%d_thumb.jpg')
extra_css = """ extra_css = """
body{font-family: Arial,Helvetica,sans-serif} body{font-family: Arial,Helvetica,sans-serif}
h2{font-family: Georgia,serif; font-size: xx-large} h2{font-family: Georgia,serif; font-size: xx-large}
@ -41,18 +44,24 @@ class Clarin(BasicNewsRecipe):
'comment': description, 'tags': category, 'publisher': publisher, 'language': language 'comment': description, 'tags': category, 'publisher': publisher, 'language': language
} }
remove_tags_before = dict(attrs={'class': 'int-nota-title'}) keep_only_tags = [
dict(name='p' , attrs={'class' : 'volanta'}),
dict(name='h1' , attrs={'itemprop': 'headline'}),
dict(name='div', attrs={'class' : 'bajada'}),
dict(name='div', attrs={'class' : 'body-nota'})
]
remove_tags = [ remove_tags = [
dict(name=['meta', 'base', 'link', 'iframe', 'embed', 'object']), dict(name=['meta', 'base', 'link', 'iframe', 'embed', 'object']),
dict(attrs={'class': ['tags-bar', 'breadcrumb', 'share-bar']}), dict(attrs={'class': ['tags-bar', 'breadcrumb', 'share-bar', 'share']}),
dict(attrs={'id': ['relacionadas']}) dict(name='div', attrs={'class': lambda x: x and 'r-nota' in x.split()}),
dict(attrs={'id': ['relacionadas']}),
dict(name='a', attrs={'class':'content-new'})
] ]
remove_tags_after = dict(name='div', attrs={'id': 'relacionadas'}) remove_tags_after = dict(name='div', attrs={'id': 'relacionadas'})
remove_attributes = ['lang'] remove_attributes = ['lang']
feeds = [ feeds = [
(u'Lo Ultimo', u'http://www.clarin.com/rss/lo-ultimo/'),
(u'Pagina principal', u'http://www.clarin.com/rss/'),
(u'Politica', u'http://www.clarin.com/rss/politica/'), (u'Politica', u'http://www.clarin.com/rss/politica/'),
(u'Deportes', u'http://www.clarin.com/rss/deportes/'), (u'Deportes', u'http://www.clarin.com/rss/deportes/'),
(u'Mundo', u'http://www.clarin.com/rss/mundo/'), (u'Mundo', u'http://www.clarin.com/rss/mundo/'),
@ -67,10 +76,11 @@ class Clarin(BasicNewsRecipe):
br = BasicNewsRecipe.get_browser(self) br = BasicNewsRecipe.get_browser(self)
br.open(self.INDEX) br.open(self.INDEX)
if self.username is not None and self.password is not None: if self.username is not None and self.password is not None:
data = urllib.urlencode({'ingresar_ingresar_paseForm': 'ingresar_ingresar_paseForm', 'ingresar_ingresar_email_paseInputComponent': self.username, 'ingresar_ingresar_palabraClave_paseInputComponent': self.password, 'ingresar_ingresar_ingresar_paseButton': 'Ingresar', 'javax.faces.ViewState': 'e1s1' # noqa data = urllib.urlencode({'ingresar_ingresar_paseForm': 'ingresar_ingresar_paseForm',
'ingresar_ingresar_email_paseInputComponent': self.username,
'ingresar_ingresar_palabraClave_paseInputComponent': self.password,
'ingresar_ingresar_ingresar_paseButton': 'Ingresar',
'javax.faces.ViewState': 'e1s1' # noqa
}) })
br.open(self.LOGIN, data) br.open(self.LOGIN, data)
return br return br
def get_article_url(self, article):
return article.get('guid', None)

BIN
recipes/icons/clarin.png
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 600 B

After

Width:  |  Height:  |  Size: 684 B