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

New recipe for The Toronto Sun by Darko Miletic. Improved recipe for The Moscow Times.

Browse Source
This commit is contained in:
Kovid Goyal 2009-10-06 11:23:50 -06:00
parent fc84ea2d15
commit 55553753d1
2 changed files with 101 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

56
src/calibre/web/feeds/recipes/recipe_moscow_times.py
View File

@ -10,18 +10,64 @@ from calibre.web.feeds.news import BasicNewsRecipe
class Moscowtimes(BasicNewsRecipe):
title = u'The Moscow Times'
__author__ = 'Darko Miletic'
__author__ = 'Darko Miletic and Sujata Raman'
description = 'News from Russia'
language = 'en'
lang = 'en'
oldest_article = 7
max_articles_per_feed = 100
no_stylesheets = True
use_embedded_content = False
#encoding = 'utf-8'
encoding = 'cp1252'
remove_javascript = True
conversion_options = {
'comment' : description
, 'language' : lang
}
extra_css = '''
h1{ color:#0066B3; font-family: Georgia,serif ; font-size: large}
.article_date{ font-family:Arial,Tahoma,Verdana,Helvetica,sans-serif ; color:#000000; font-size: x-small;}
.autors{color:#999999 ; font-weight: bold ; font-family:Arial,Tahoma,Verdana,Helvetica,sans-serif ; font-size: x-small; }
.photoautors{ color:#999999 ; font-family:Arial,Tahoma,Verdana,Helvetica,sans-serif ; font-size: x-small; }
.text{font-family:Arial,Tahoma,Verdana,Helvetica,sans-serif ; font-size:75%; }
'''
feeds = [
(u'The Moscow Times' , u'http://www.themoscowtimes.com/rss.xml' )
(u'The Moscow Times Top Stories' , u'http://www.themoscowtimes.com/rss/top'),
(u'The Moscow Times Current Issue' , u'http://www.themoscowtimes.com/rss/issue'),
(u'The Moscow Times News' , u'http://www.themoscowtimes.com/rss/news'),
(u'The Moscow Times Business' , u'http://www.themoscowtimes.com/rss/business'),
(u'The Moscow Times Art and Ideas' , u'http://www.themoscowtimes.com/rss/art'),
(u'The Moscow Times Opinion' , u'http://www.themoscowtimes.com/rss/opinion')
]
def print_version(self, url):
return url + '&print=Y'
keep_only_tags = [
dict(name='div', attrs={'class':['newstextblock']})
]
remove_tags = [
dict(name='div', attrs={'class':['photo_nav']})
]
def preprocess_html(self, soup):
soup.html['xml:lang'] = self.lang
soup.html['lang'] = self.lang
mtag = '<meta http-equiv="Content-Type" content="text/html; charset=' + self.encoding + '">'
soup.head.insert(0,mtag)
return self.adeify_images(soup)
def get_cover_url(self):
href = 'http://www.themoscowtimes.com/pdf/'
soup = self.index_to_soup(href)
div = soup.find('div',attrs={'class':'left'})
a = div.find('a')
print a
if a :
cover_url = a.img['src']
return cover_url

50
src/calibre/web/feeds/recipes/recipe_toronto_sun.py Normal file
View File

@ -0,0 +1,50 @@
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2009, Darko Miletic <darko.miletic at gmail.com>'
'''
www.torontosun.com
'''
from calibre.web.feeds.news import BasicNewsRecipe
class TorontoSun(BasicNewsRecipe):
title = 'Toronto SUN'
__author__ = 'Darko Miletic'
description = 'News from Canada'
publisher = 'Toronto Sun'
category = 'news, politics, Canada'
oldest_article = 2
max_articles_per_feed = 100
no_stylesheets = True
use_embedded_content = False
encoding = 'cp1252'
language = 'en_CA'
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
}
keep_only_tags =[
dict(name='div', attrs={'class':'articleHead'})
,dict(name='div', attrs={'id':'channelContent'})
]
remove_tags = [
dict(name='div',attrs={'class':['leftBox','bottomBox clear','bottomBox','breadCrumb']})
,dict(name=['link','iframe','object'])
,dict(name='a',attrs={'rel':'swap'})
,dict(name='ul',attrs={'class':'tabs dl contentSwap'})
]
remove_tags_after = dict(name='div',attrs={'class':'bottomBox clear'})
feeds = [
(u'News' , u'http://www.torontosun.com/news/rss.xml' )
,(u'Canada' , u'http://www.torontosun.com/news/canada/rss.xml' )
,(u'Columnists' , u'http://www.torontosun.com/news/columnists/rss.xml')
,(u'World' , u'http://www.torontosun.com/news/world/rss.xml' )
,(u'Money' , u'http://www.torontosun.com/money/rss.xml' )
]