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

Fix Daily Mirror

Browse Source
This commit is contained in:
Kovid Goyal 2012-02-11 22:51:41 +05:30
parent 6af42f147e
commit c9a6bd6175
1 changed files with 35 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

47
recipes/daily_mirror.recipe
View File

@ -5,7 +5,7 @@ class AdvancedUserRecipe1306061239(BasicNewsRecipe):
description = 'News as provide by The Daily Mirror -UK' description = 'News as provide by The Daily Mirror -UK'
__author__ = 'Dave Asbury' __author__ = 'Dave Asbury'
# last updated 26/12/11 # last updated 11/2/12
language = 'en_GB' language = 'en_GB'
cover_url = 'http://yookeo.com/screens/m/i/mirror.co.uk.jpg' cover_url = 'http://yookeo.com/screens/m/i/mirror.co.uk.jpg'
@ -14,35 +14,58 @@ class AdvancedUserRecipe1306061239(BasicNewsRecipe):
oldest_article = 1 oldest_article = 1
max_articles_per_feed = 20 max_articles_per_feed = 5
remove_empty_feeds = True remove_empty_feeds = True
remove_javascript = True remove_javascript = True
no_stylesheets = True no_stylesheets = True
auto_cleanup = True auto_cleanup = True
#conversion_options = { 'linearize_tables' : True }
#keep_only_tags = [
# dict(name='h1'),
# dict(name='div',attrs={'id' : 'body-content'}),
#dict(name='div',atts={'class' : 'article-body'}),
#dict(attrs={'class' : ['article-attr','byline append-1','published']}),
#dict(name='p'),
# ]
#remove_tags_after = [dict (name='div',attrs={'class' : 'related'})]
remove_tags = [ remove_tags = [
dict(name='title'), dict(name='title'),
dict(name='div',attrs={'class' : ['inline-ad span-16 last','caption']}), dict(name='div',attrs={'class' : ['inline-ad span-16 last','caption']}),
# dict(name='div',attrs={'id' : ['sidebar','menu','search-box','roffers-top']}),
#dict(name='div',attrs={'class' :['inline-ad span-16 last','article-resize','related','list teasers']}),
#dict(attrs={'class' : ['channellink','article-tags','replace','append-html']}),
] ]
# preprocess_regexps = [
#(re.compile(r'<dl class="q-search">.*?</dl>', re.IGNORECASE | re.DOTALL), lambda match: '')]
preprocess_regexps = [ preprocess_regexps = [
(re.compile(r'- mirror.co.uk', re.IGNORECASE | re.DOTALL), lambda match: '')] (re.compile(r'- mirror.co.uk', re.IGNORECASE | re.DOTALL), lambda match: '')]
preprocess_regexps = [ preprocess_regexps = [
(re.compile(r'Advertisement >>', re.IGNORECASE | re.DOTALL), lambda match: '')] (re.compile(r'Advertisement >>', re.IGNORECASE | re.DOTALL), lambda match: '')]
#preprocess_regexps = [
#(re.compile(r'Sponsored Links', re.IGNORECASE | re.DOTALL), lambda match: '')]
feeds = [ feeds = [
(u'News', u'http://www.mirror.co.uk/news/rss.xml') (u'UK News', u'http://feed43.com/0287771688643868.xml')
,(u'Tech News', u'http://www.mirror.co.uk/news/technology/rss.xml') ,(u'Tech News', u'http://feed43.com/2455520588350501.xml')
,(u'Weird World','http://www.mirror.co.uk/news/weird-world/rss.xml') ,(u'Weird World','http://feed43.com/0863800333634654.xml')
,(u'Film Gossip','http://www.mirror.co.uk/celebs/film/rss.xml') ,(u'Sport','http://feed43.com/7713243036546130.xml')
,(u'Music News','http://www.mirror.co.uk/celebs/music/rss.xml') ,(u'Sport : Boxing ','http://feed43.com/0414732220804255.xml')
,(u'Celebs and Tv Gossip','http://www.mirror.co.uk/celebs/tv/rss.xml') ,(u'Sport : Rugby Union','http://feed43.com/4710138762362383.xml')
,(u'Sport','http://www.mirror.co.uk/sport/rss.xml') ,(u'Sport : Other','http://feed43.com/4501416886323415.xml')
,(u'Life Style','http://www.mirror.co.uk/life-style/rss.xml') ,(u'TV and Film','http://feed43.com/5238302853765104.xml')
,(u'Advice','http://www.mirror.co.uk/advice/rss.xml') ,(u'Celebs','http://feed43.com/8770061048844683.xml')
,(u'Travel','http://www.mirror.co.uk/advice/travel/rss.xml') ,(u'Life Style : Family','http://feed43.com/4356170742410338.xml')
,(u'Travel','http://feed43.com/1436576006476607.xml')
# example of commented out feed not needed ,(u'Travel','http://www.mirror.co.uk/advice/travel/rss.xml') # example of commented out feed not needed ,(u'Travel','http://www.mirror.co.uk/advice/travel/rss.xml')
] ]