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 tweakers.net

Browse Source
This commit is contained in:
Kovid Goyal 2017-02-19 07:48:28 +05:30
parent 6dcd09d54c
commit 6385a90b1b
1 changed files with 21 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
recipes/tweakers_net.recipe
View File

@ -2,16 +2,15 @@
# vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai # vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai
from __future__ import with_statement from __future__ import with_statement
__license__ = 'GPL v3'
__copyright__ = '2009, Kovid Goyal <kovid@kovidgoyal.net>'
__docformat__ = 'restructuredtext en'
''' Changelog ''' Changelog
2012-04-27 DrMerry: 2012-04-27 DrMerry:
Added cover picture Added cover picture
removed some extra tags removed some extra tags
''' '''
__license__ = 'GPL v3'
__copyright__ = '2009, Kovid Goyal <kovid@kovidgoyal.net>'
__docformat__ = 'restructuredtext en'
import re import re
from calibre.web.feeds.news import BasicNewsRecipe from calibre.web.feeds.news import BasicNewsRecipe
@ -24,12 +23,22 @@ class Tweakers(BasicNewsRecipe):
max_articles_per_feed = 40 max_articles_per_feed = 40
cover_url = 'http://tweakers.net/ext/launch/g/logo.gif' cover_url = 'http://tweakers.net/ext/launch/g/logo.gif'
keep_only_tags = [dict(name='div', attrs={'class': 'columnwrapper news'})] keep_only_tags = [
dict(name='div', attrs={'class': 'columnwrapper news'}),
dict(name='div', attrs={'class': 'article'})
]
remove_tags = [dict(name='div', attrs={'class': 'reacties'}), remove_tags = [
{'id': ['utracker', 'socialButtons', 'b_ac']}, dict(name='div', attrs={'class': 'reacties'}),
{'class': ['sidebar', 'advertorial']}, {
{'class': re.compile('nextPrevious')}, 'id': ['utracker', 'socialButtons', 'b_ac']
},
{
'class': ['sidebar', 'advertorial']
},
{
'class': re.compile('nextPrevious')
},
] ]
no_stylesheets = True no_stylesheets = True
filter_regexps = [r'ads\.doubleclick\.net', r'ad\.doubleclick\.net'] filter_regexps = [r'ads\.doubleclick\.net', r'ad\.doubleclick\.net']