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 20 minutes

Browse Source
This commit is contained in:
Kovid Goyal 2015-07-09 16:15:01 +05:30
parent cd60236542
commit 8ff707afd6
1 changed files with 9 additions and 30 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

39
recipes/20minutes.recipe
View File

@ -9,9 +9,9 @@ from calibre.web.feeds.recipes import BasicNewsRecipe
class Minutes(BasicNewsRecipe): class Minutes(BasicNewsRecipe):
title = '20 minutes' title = '20 minutes'
__author__ = 'calibre' __author__ = u'Aurélien Chabot'
description = 'Actualités' description = 'Actualités'
encoding = 'cp1252' encoding = 'utf-8'
publisher = '20minutes.fr' publisher = '20minutes.fr'
category = 'Actualités, France, Monde' category = 'Actualités, France, Monde'
language = 'fr' language = 'fr'
@ -21,30 +21,15 @@ class Minutes(BasicNewsRecipe):
max_articles_per_feed = 15 max_articles_per_feed = 15
no_stylesheets = True no_stylesheets = True
remove_empty_feeds = True remove_empty_feeds = True
filterDuplicates = True
extra_css = ''' keep_only_tags = [
h1 {font-size:xx-large; font-family:Arial,Helvetica,sans-serif;} dict(name='h1'),
.mna-details {font-size:xx-small; color:#4D4D4D; font-family:Arial,Helvetica,sans-serif;} dict(attrs={'class':lambda x: x and 'lt-content' in x.split()}),
.mna-image {font-size:xx-small; color:#4D4D4D; font-family:Arial,Helvetica,sans-serif;}
.mna-body {font-size:medium; font-family:Arial,Helvetica,sans-serif;}
'''
remove_tags = [
dict(name='iframe'),
dict(name='div', attrs={'class':['mn-section-heading']}),
dict(name='a', attrs={'href':['#commentaires']}),
dict(name='div', attrs={'class':['mn-right']}),
dict(name='div', attrs={'class':['mna-box']}),
dict(name='div', attrs={'class':['mna-comment-call']}),
dict(name='div', attrs={'class':['mna-tools']}),
dict(name='div', attrs={'class':['mn-trilist']})
] ]
remove_tags = [
keep_only_tags = [dict(id='mn-article')] dict(attrs={'class':lambda x:x and 'content-related' in x.split()}),
]
remove_tags_after = dict(name='div', attrs={'class':['mna-body','mna-signature']}) remove_tags_after = dict(id='ob_holder')
feeds = [ feeds = [
('France', 'http://www.20minutes.fr/rss/actu-france.xml'), ('France', 'http://www.20minutes.fr/rss/actu-france.xml'),
@ -62,9 +47,3 @@ class Minutes(BasicNewsRecipe):
('Lyon', 'http://www.20minutes.fr/rss/lyon.xml'), ('Lyon', 'http://www.20minutes.fr/rss/lyon.xml'),
('Toulouse', 'http://www.20minutes.fr/rss/toulouse.xml') ('Toulouse', 'http://www.20minutes.fr/rss/toulouse.xml')
] ]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return soup