From a689e82b56b04ed7e5245fe9d46d8a2f3c79587e Mon Sep 17 00:00:00 2001 From: Kovid Goyal Date: Sat, 7 Apr 2012 17:26:19 +0530 Subject: [PATCH] Update Sun UK --- recipes/the_sun.recipe | 51 ++++++++++++++++++++++-------------------- 1 file changed, 27 insertions(+), 24 deletions(-) diff --git a/recipes/the_sun.recipe b/recipes/the_sun.recipe index 80b37f329a..9285c0b2c2 100644 --- a/recipes/the_sun.recipe +++ b/recipes/the_sun.recipe @@ -1,57 +1,60 @@ import re from calibre.web.feeds.recipes import BasicNewsRecipe - class AdvancedUserRecipe1325006965(BasicNewsRecipe): title = u'The Sun UK' cover_url = 'http://www.thesun.co.uk/img/global/new-masthead-logo.png' - description = 'A Recipe for The Sun tabloid UK - uses feed43' + description = 'A Recipe for The Sun tabloid UK' __author__ = 'Dave Asbury' - # last updated 20/2/12 + # last updated 7/4/12 language = 'en_GB' oldest_article = 1 max_articles_per_feed = 15 remove_empty_feeds = True no_stylesheets = True + #auto_cleanup = True + #articles_are_obfuscated = True masthead_url = 'http://www.thesun.co.uk/sol/img/global/Sun-logo.gif' - encoding = 'cp1251' + encoding = 'UTF-8' - encoding = 'cp1252' remove_empty_feeds = True remove_javascript = True no_stylesheets = True extra_css = ''' - body{ text-align: justify; font-family:Arial,Helvetica,sans-serif; font-size:11px; font-size-adjust:none; font-stretch:normal; font-style:normal; font-variant:normal; font-weight:normal;} - ''' + body{ text-align: justify; font-family:Arial,Helvetica,sans-serif; font-size:11px; font-size-adjust:none; font-stretch:normal; font-style:normal; font-variant:normal; font-weight:normal;} + ''' preprocess_regexps = [ - (re.compile(r'
', re.IGNORECASE | re.DOTALL), lambda match: '')] + (re.compile(r'
', re.IGNORECASE | re.DOTALL), lambda match: '')] + + keep_only_tags = [ dict(name='h1'),dict(name='h2',attrs={'class' : 'medium centered'}), - dict(name='div',attrs={'class' : 'text-center'}), - dict(name='div',attrs={'id' : 'bodyText'}) - # dict(name='p') - ] - + dict(name='div',attrs={'class' : 'text-center'}), + dict(name='div',attrs={'id' : 'bodyText'}) + # dict(name='p') + ] remove_tags=[ - #dict(name='head'), - dict(attrs={'class' : ['mystery-meat-link','ltbx-container','ltbx-var ltbx-hbxpn','ltbx-var ltbx-nav-loop','ltbx-var ltbx-url']}), + #dict(name='head'), + dict(attrs={'class' : ['mystery-meat-link','ltbx-container','ltbx-var ltbx-hbxpn','ltbx-var ltbx-nav-loop','ltbx-var ltbx-url']}), dict(name='div',attrs={'class' : 'cf'}), - dict(attrs={'title' : 'download flash'}), + dict(attrs={'title' : 'download flash'}), dict(attrs={'style' : 'padding: 5px'}) - ] + ] + feeds = [ - (u'News','http://feed43.com/2517447382644748.xml'), - (u'Sport', u'http://feed43.com/4283846255668687.xml'), - (u'Bizarre', u'http://feed43.com/0233840304242011.xml'), - (u'Film',u'http://feed43.com/1307545221226200.xml'), - (u'Music',u'http://feed43.com/1701513435064132.xml'), - (u'Sun Woman',u'http://feed43.com/0022626854226453.xml'), - ] + #(u'News', u'http://www.thesun.co.uk/sol/homepage/news/rss'), + (u'News','http://feed43.com/2517447382644748.xml'), + (u'Sport', u'http://feed43.com/4283846255668687.xml'), + (u'Bizarre', u'http://feed43.com/0233840304242011.xml'), + (u'Film',u'http://feed43.com/1307545221226200.xml'), + (u'Music',u'http://feed43.com/1701513435064132.xml'), + (u'Sun Woman',u'http://feed43.com/0022626854226453.xml'), +]