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

...

Browse Source
This commit is contained in:
Kovid Goyal 2014-09-11 22:36:44 +05:30
parent 0a90dc23d6
commit 41deb58447
1 changed files with 3 additions and 54 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

57
recipes/dawn.recipe
View File

@ -1,5 +1,4 @@
from calibre.web.feeds.news import BasicNewsRecipe from calibre.web.feeds.news import BasicNewsRecipe
from calibre.ebooks.BeautifulSoup import Tag
class DawnRecipe(BasicNewsRecipe): class DawnRecipe(BasicNewsRecipe):
__license__ = 'GPL v3' __license__ = 'GPL v3'
@ -16,20 +15,17 @@ class DawnRecipe(BasicNewsRecipe):
remove_empty_feeds = True remove_empty_feeds = True
oldest_article = 2 oldest_article = 2
max_articles_per_feed = 100 max_articles_per_feed = 100
#auto_cleanup = True
#auto_cleanup_keep = '//dix[@class="slideshow"]'
no_stylesheets = True no_stylesheets = True
remove_javascript = True remove_javascript = True
encoding = 'utf-8' encoding = 'utf-8'
keep_only_tags = [dict(name='div', attrs={'class':'push-half--sides push--top'}), keep_only_tags = [
dict(name='article', attrs={'class':'story story--single push-half'})] dict(name='div', attrs={'class':'push-half--sides push--top'}),
dict(name='article', attrs={'class':'story story--single push-half'})]
# Feeds from http://www.dawn.com/wps/wcm/connect/dawn-content-library/dawn/services/rss # Feeds from http://www.dawn.com/wps/wcm/connect/dawn-content-library/dawn/services/rss
feeds = [] feeds = []
feeds.append((u'Latest News', u'http://feeds.feedburner.com/dawn-news')) feeds.append((u'Latest News', u'http://feeds.feedburner.com/dawn-news'))
#feeds.append((u'', u''))
conversion_options = {'comments': description, 'tags': category, 'language': 'en', conversion_options = {'comments': description, 'tags': category, 'language': 'en',
'publisher': publisher} 'publisher': publisher}
@ -42,50 +38,3 @@ class DawnRecipe(BasicNewsRecipe):
span.news_byline {font-size: x-small; color: #696969; margin-top: 1em;} span.news_byline {font-size: x-small; color: #696969; margin-top: 1em;}
''' '''
#def print_version(self, url):
#url = url.split('?')[0] + '/print'
#print(url)
#return url
#def preprocess_html(self, soup):
#newBody = Tag(soup, 'body')
#for cl in ['page_title', 'news_headline', 'news_byline']:
#tag = soup.find('span', attrs = {'class': cl})
#if tag:
## They like their <br> tags; I don't: does not work well on small screens.
#if tag['class'] == 'news_byline':
#for br in tag.findAll('br'):
#br.extract()
#newBody.append(tag)
#table = soup.find('table', attrs = {'id': 'body table'})
#if table:
#for td in table.findAll('td', attrs = {'class': 'news_story'}):
#for tag in td.findAll(True):
#if tag.has_key('id') and tag['id'] == 'banner-img_slide':
#tag.extract()
#elif tag.has_key('style'):
#del tag['style']
#elif tag.name == 'script':
#tag.extract()
## They like their <br> tags; I don't: does not work well on small screens.
#center = td.find('center')
#if center:
#for br in center.findNextSiblings('br'):
#br.extract()
#for br in center.findPreviousSiblings('br'):
#br.extract()
#for attr in ['align', 'valign']:
#if td.has_key(attr):
#del td[attr]
#td.name = 'div'
#newBody.append(td)
#soup.body.replaceWith(newBody)
#return soup