Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-11-03 19:17:02 -05:00
Code Issues Packages Projects Releases Wiki Activity

Fix USA Today news website recipe

Browse Source
This commit is contained in:
Aurélien Chabot 2011-10-16 11:24:15 +02:00
parent 4bc485f6a4
commit 1bb39ffd77
1 changed files with 32 additions and 24 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

56
recipes/usatoday.recipe
View File

@ -10,27 +10,28 @@ from calibre.web.feeds.news import BasicNewsRecipe
class USAToday(BasicNewsRecipe):
title = 'USA Today'
__author__ = 'Kovid Goyal'
oldest_article = 1
publication_type = 'newspaper'
timefmt = ''
max_articles_per_feed = 20
language = 'en'
no_stylesheets = True
extra_css = '.headline {text-align: left;}\n \
.byline {font-family: monospace; \
text-align: left; \
margin-bottom: 1em;}\n \
.image {text-align: center;}\n \
.caption {text-align: center; \
font-size: smaller; \
font-style: italic}\n \
.credit {text-align: right; \
margin-bottom: 0em; \
font-size: smaller;}\n \
.articleBody {text-align: left;}\n '
#simultaneous_downloads = 1
title = 'USA Today'
__author__ = 'calibre'
description = 'newspaper'
encoding = 'utf-8'
publisher = 'usatoday.com'
category = 'news, usa'
language = 'en'
use_embedded_content = False
timefmt = ' [%d %b %Y]'
max_articles_per_feed = 15
no_stylesheets = True
remove_empty_feeds = True
filterDuplicates = True
extra_css = '''
h1, h2 {font-size:xx-large; font-family:Arial,Helvetica,sans-serif;}
#post-attributes, .info, .clear {font-size:xx-small; color:#4D4D4D; font-family:Arial,Helvetica,sans-serif;}
#post-body, #content {font-size:medium; font-family:Arial,Helvetica,sans-serif;}
'''
feeds = [
('Top Headlines', 'http://rssfeeds.usatoday.com/usatoday-NewsTopStories'),
('Tech Headlines', 'http://rssfeeds.usatoday.com/usatoday-TechTopStories'),
@ -43,15 +44,18 @@ class USAToday(BasicNewsRecipe):
('Sport Headlines', 'http://rssfeeds.usatoday.com/UsatodaycomSports-TopStories'),
('Weather Headlines', 'http://rssfeeds.usatoday.com/usatoday-WeatherTopStories'),
('Most Popular', 'http://rssfeeds.usatoday.com/Usatoday-MostViewedArticles'),
('Offbeat News', 'http://rssfeeds.usatoday.com/UsatodaycomOffbeat-TopStories'),
('Offbeat News', 'http://rssfeeds.usatoday.com/UsatodaycomOffbeat-TopStories')
]
keep_only_tags = [dict(attrs={'class':'story'})]
remove_tags = [
dict(attrs={'class':[
'share',
'reprints',
'inline-h3',
'info-extras',
'info-extras rounded',
'inset',
'ppy-outer',
'ppy-caption',
'comments',
@ -61,9 +65,13 @@ class USAToday(BasicNewsRecipe):
'tags',
'bottom-tools',
'sponsoredlinks',
'corrections'
]}),
dict(name='ul', attrs={'class':'inside-copy'}),
dict(id=['pluck']),
]
dict(id=['updated']),
dict(id=['post-date-updated'])
]
def get_masthead_url(self):