calibre/recipes/staradvertiser.recipe

78 lines
3.4 KiB
Plaintext

__license__ = 'GPL v3'
__copyright__ = '2009-2011, Darko Miletic <darko.miletic at gmail.com>'
'''
staradvertiser.com
'''
from calibre.web.feeds.news import BasicNewsRecipe
class Starbulletin(BasicNewsRecipe):
title = 'Honolulu Star Advertiser'
__author__ = 'Darko Miletic'
description = 'Latest national and local Hawaii sports news'
publisher = 'Honolulu Star-Advertiser'
category = 'news, Honolulu, Hawaii'
oldest_article = 2
max_articles_per_feed = 100
language = 'en'
no_stylesheets = True
use_embedded_content = False
encoding = 'utf8'
publication_type = 'newspaper'
masthead_url = 'http://media.staradvertiser.com/designimages/star-advertiser-logo-small.gif'
extra_css = """
body{font-family: Verdana,Arial,Helvetica,sans-serif}
h1,.brown,.postCredit{color: #663300}
.storyDeck{font-size: 1.2em; font-weight: bold}
img{display: block}
"""
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
, 'linearize_tables' : True
}
keep_only_tags = [
dict(attrs={'id':'storyTitle'})
,dict(attrs={'class':['storyDeck','postCredit']})
,dict(name='span',attrs={'class':'brown'})
,dict(name='div',attrs={'class':'storytext'})
]
remove_tags = [
dict(name=['object','link','script','span','meta','base','iframe'])
,dict(attrs={'class':['insideStoryImage','insideStoryAd']})
,dict(attrs={'name':'fb_share'})
]
feeds = [
(u'Headlines' , u'http://www.staradvertiser.com/staradvertiser_headlines.rss' )
,(u'News' , u'http://www.staradvertiser.com/news/index.rss' )
,(u'Sports' , u'http://www.staradvertiser.com/sports/index.rss' )
,(u'Features' , u'http://www.staradvertiser.com/features/index.rss' )
,(u'Editorials', u'http://www.staradvertiser.com/editorials/index.rss' )
,(u'Business' , u'http://www.staradvertiser.com/business/index.rss' )
,(u'Travel' , u'http://www.staradvertiser.com/travel/index.rss' )
]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
for item in soup.findAll('a'):
limg = item.find('img')
if item.string is not None:
str = item.string
item.replaceWith(str)
else:
if limg:
item.name = 'div'
item.attrs = []
else:
str = self.tag_to_string(item)
item.replaceWith(str)
for item in soup.findAll('img'):
if not item.has_key('alt'):
item['alt'] = 'image'
return soup