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

Improved Newsweek recipe

Browse Source
This commit is contained in:
Kovid Goyal 2009-10-08 22:31:45 -06:00
parent 0a5c6410a8
commit 1d823ba705
1 changed files with 24 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

37
src/calibre/web/feeds/recipes/recipe_newsweek.py
View File

@ -1,7 +1,6 @@
#!/usr/bin/env python
__license__ = 'GPL v3' __license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>' __copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
import re import re
from calibre import strftime from calibre import strftime
from calibre.ebooks.BeautifulSoup import BeautifulSoup from calibre.ebooks.BeautifulSoup import BeautifulSoup
@ -9,37 +8,44 @@ from calibre.web.feeds.news import BasicNewsRecipe
class Newsweek(BasicNewsRecipe): class Newsweek(BasicNewsRecipe):
title = 'Newsweek' title = 'Newsweek'
__author__ = 'Kovid Goyal and Sujata Raman' __author__ = 'Kovid Goyal and Sujata Raman'
description = 'Weekly news and current affairs in the US' description = 'Weekly news and current affairs in the US'
no_stylesheets = True no_stylesheets = True
extra_css = ''' extra_css = '''
h1{color:#383733;font-family:Arial,Helvetica,sans-serif;font-size:large;} h1{font-family:Arial,Helvetica,sans-serif; font-size:large; color:#383733;}
.deck{font-family:Georgia,"Century Schoolbook","Times New Roman",Times,serif;color:#383733;font-size:small;} .deck{font-family:Georgia,sans-serif; color:#383733;}
.articleInfo{color:#474537;font-family:Arial,Helvetica,sans-serif; font-size:xx-small;} .bylineDate{font-family:georgia ; color:#58544A; font-size:x-small;}
.authorName{color:#B61900;font-family:Arial,Helvetica,sans-serif;font-size:medium;} .authorInfo{font-family:arial,helvetica,sans-serif; color:#0066CC; font-size:x-small;}
.authorInfo{color:#0066CC;font-family:Arial,Helvetica,sans-serif;font-size:xx-small;} .articleUpdated{font-family:arial,helvetica,sans-serif; color:#73726C; font-size:x-small;}
.articleUpdated{ font-size:xx-small; color:#73726C; font-family:Arial,Helvetica,sans-serif;} .issueDate{font-family:arial,helvetica,sans-serif; color:#73726C; font-size:x-small; font-style:italic;}
.issueDate{font-family :Arial,Helvetica,sans-serif;font-size:xx-small;font-style:italic;} h5{font-family:arial,helvetica,sans-serif; color:#73726C; font-size:x-small;}
.story{color:#333333; font-family:Georgia,"Century Schoolbook","Times New Roman",Times,serif;font-size:small;} h6{font-family:arial,helvetica,sans-serif; color:#73726C; font-size:x-small;}
.photoCredit{color:#999999;font-family:Arial,Helvetica,sans-serif;font-size:xx-small;} .story{font-family:georgia,sans-serif ; color:#363636;}
.photoCaption{color:#0A0A09;font-family:Arial,Helvetica,sans-serif;font-size:xx-small;font-weight:bold;}''' .photoCredit{color:#999999; font-family:Arial,Helvetica,sans-serif;font-size:x-small;}
.photoCaption{color:#0A0A09;font-family:Arial,Helvetica,sans-serif;font-size:x-small;}
.fwArticle{font-family:Arial,Helvetica,sans-serif;font-size:x-small;font-weight:bold;}
'''
encoding = 'utf-8' encoding = 'utf-8'
language = 'en' language = 'en'
remove_tags = [ remove_tags = [
{'class':['fwArticle noHr','fwArticle','subinfo','hdlBulletItem','head-content','navbar','link', 'ad', 'sponsorLinksArticle', 'mm-content', {'class':['fwArticle noHr','fwArticle','subinfo','hdlBulletItem','head-content','navbar','link', 'ad', 'sponsorLinksArticle', 'mm-content',
'inline-social-links-wrapper', 'email-article', 'inline-social-links-wrapper', 'email-article','ToolBox',
'inlineComponentRight', 'inlineComponentRight',
'comments-and-social-links-wrapper', 'EmailArticleBlock']}, 'comments-and-social-links-wrapper', 'EmailArticleBlock']},
{'id' : ['footer', 'ticker-data', 'topTenVertical', {'id' : ['footer', 'ticker-data', 'topTenVertical',
'digg-top-five', 'mesothorax', 'nw-comments', 'digg-top-five', 'mesothorax', 'nw-comments',
'ToolBox', 'EmailMain']}, 'ToolBox', 'EmailMain']},
{'class': re.compile('related-cloud')}, {'class': re.compile('related-cloud')},
dict(name='li', attrs={'id':['slug_bigbox']})
] ]
keep_only_tags = [{'class':['article HorizontalHeader', 'articlecontent']}]
keep_only_tags = [{'class':['article HorizontalHeader', 'articlecontent','photoBox']}, ]
recursions = 1 recursions = 1
match_regexps = [r'http://www.newsweek.com/id/\S+/page/\d+'] match_regexps = [r'http://www.newsweek.com/id/\S+/page/\d+']
@ -177,3 +183,4 @@ class Newsweek(BasicNewsRecipe):
if article.description is None : if article.description is None :
article.description = extractDescription(article.href) article.description = extractDescription(article.href)
return return