Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-06-23 15:30:45 -04:00
Code Issues Packages Projects Releases Wiki Activity

Update Daily Express

Browse Source
This commit is contained in:
Kovid Goyal 2013-11-16 18:37:42 +05:30
parent a5eeba8281
commit d0dbc471d2
1 changed files with 12 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
recipes/daily_express.recipe
View File

@ -1,9 +1,11 @@
import re import re
from calibre.web.feeds.news import BasicNewsRecipe from calibre.web.feeds.news import BasicNewsRecipe
class AdvancedUserRecipe1376229553(BasicNewsRecipe): class AdvancedUserRecipe1376229553(BasicNewsRecipe):
title = u'Daily Express' title = u'Daily Express'
__author__ = 'Dave Asbury' __author__ = 'Dave Asbury'
# 9-9-13 added article author and now use (re.compile(r'>[\w].+? News<' # 9-9-13 added article author and now use (re.compile(r'>[\w].+? News<'
# 16-11-13 cover adjustment
encoding = 'utf-8' encoding = 'utf-8'
remove_empty_feeds = True remove_empty_feeds = True
#remove_javascript = True #remove_javascript = True
@ -68,13 +70,21 @@ class AdvancedUserRecipe1376229553(BasicNewsRecipe):
def get_cover_url(self): def get_cover_url(self):
soup = self.index_to_soup('http://www.express.co.uk/ourpaper/') soup = self.index_to_soup('http://www.express.co.uk/ourpaper/')
cov = soup.find(attrs={'src' : re.compile('http://images.dailyexpress.co.uk/img/covers/')}) cov = soup.find(attrs={'src' : re.compile('http://cdn.images.express.co.uk/img/covers/')})
cov=str(cov) cov=str(cov)
cov2 = re.findall('http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\(\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+', cov) cov2 = re.findall('http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\(\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+', cov)
cov=str(cov2) cov=str(cov2)
cov=cov[2:len(cov)-2] cov=cov[2:len(cov)-2]
cover_url=cov
br = self.get_browser()
br.set_handle_redirect(False)
try:
br.open_novisit(cov)
cover_url = cov
except:
cover_url ='http://cdn.images.express.co.uk/img/static/ourpaper/header-back-issue-papers.jpg'
return cover_url return cover_url
extra_css = ''' extra_css = '''