Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-08 10:44:09 -04:00
Code Issues Packages Projects Releases Wiki Activity

Update Dvhn

Browse Source
This commit is contained in:
Kovid Goyal 2011-05-18 08:36:49 -06:00
parent faf5ba7d7c
commit 93f8e4f7c5
1 changed files with 23 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
recipes/dvhn.recipe
View File

@ -1,19 +1,21 @@
import re
from calibre.web.feeds.news import BasicNewsRecipe from calibre.web.feeds.news import BasicNewsRecipe
class AdvancedUserRecipe1302341394(BasicNewsRecipe): class AdvancedUserRecipe1302341394(BasicNewsRecipe):
title = u'DvhN' title = u'DvhN'
oldest_article = 1 __author__ = 'Reijndert'
oldest_article = 7
max_articles_per_feed = 200 max_articles_per_feed = 200
__author__ = 'Reijndert'
no_stylesheets = True no_stylesheets = True
cover_url = 'http://www.dvhn.nl/template/Dagblad_v2.0/gfx/logo_DvhN.gif' cover_url = 'http://members.home.nl/apm.de.haas/calibre/DvhN.jpg'
language = 'nl' language = 'nl'
country = 'NL' country = 'NL'
version = 1 version = 1
publisher = u'Dagblad van het Noorden' publisher = u'Dagblad van het Noorden'
category = u'Nieuws' category = u'Nieuws'
description = u'Nieuws uit Noord Nederland' description = u'Nieuws uit Noord Nederland'
timefmt = ' %Y-%m-%d (%a)'
keep_only_tags = [dict(name='div', attrs={'id':'fullPicture'}) keep_only_tags = [dict(name='div', attrs={'id':'fullPicture'})
@ -21,11 +23,26 @@ class AdvancedUserRecipe1302341394(BasicNewsRecipe):
] ]
remove_tags = [ remove_tags = [
dict(name=['object','link','iframe','base']) dict(name='span',attrs={'class':'location'})
,dict(name='span',attrs={'class':'copyright'})
] ]
feeds = [(u'Drenthe', u'http://www.dvhn.nl/nieuws/drenthe/index.jsp?service=rss'), (u'Groningen', u'http://www.dvhn.nl/nieuws/groningen/index.jsp?service=rss'), (u'Nederland', u'http://www.dvhn.nl/nieuws/nederland/index.jsp?service=rss'), (u'Wereld', u'http://www.dvhn.nl/nieuws/wereld/index.jsp?service=rss'), (u'Economie', u'http://www.dvhn.nl/nieuws/economie/index.jsp?service=rss'), (u'Sport', u'http://www.dvhn.nl/nieuws/sport/index.jsp?service=rss'), (u'Cultuur', u'http://www.dvhn.nl/nieuws/kunst/index.jsp?service=rss'), (u'24 Uur', u'http://www.dvhn.nl/nieuws/24uurdvhn/index.jsp?service=rss&selectiontype=last24hours')] preprocess_regexps = [
(re.compile(r'<a.*?>'), lambda h1: '')
,(re.compile(r'</a>'), lambda h2: '')
,(re.compile(r'Word vriend van Dagblad van het Noorden op Facebook'), lambda h3: '')
,(re.compile(r'Volg Dagblad van het Noorden op Twitter'), lambda h3: '')
]
feeds = [(u'Drenthe', u'http://www.dvhn.nl/nieuws/drenthe/index.jsp?service=rss')
, (u'Groningen', u'http://www.dvhn.nl/nieuws/groningen/index.jsp?service=rss')
, (u'Nederland', u'http://www.dvhn.nl/nieuws/nederland/index.jsp?service=rss')
, (u'Wereld', u'http://www.dvhn.nl/nieuws/wereld/index.jsp?service=rss')
, (u'Economie', u'http://www.dvhn.nl/nieuws/economie/index.jsp?service=rss')
, (u'Sport', u'http://www.dvhn.nl/nieuws/sport/index.jsp?service=rss')
, (u'Cultuur', u'http://www.dvhn.nl/nieuws/kunst/index.jsp?service=rss')
, (u'24 Uur', u'http://www.dvhn.nl/nieuws/24uurdvhn/index.jsp?service=rss&selectiontype=last24hours')
]
extra_css = ''' extra_css = '''
body {font-family: verdana, arial, helvetica, geneva, sans-serif;} body {font-family: verdana, arial, helvetica, geneva, sans-serif;}