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

Fix Der Tagesspiegel

Browse Source
This commit is contained in:
Kovid Goyal 2012-04-23 13:11:24 +05:30
parent 8a3dedc7eb
commit b13b7f8a50
1 changed files with 9 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
recipes/tagesspiegel.recipe
View File

@ -34,7 +34,7 @@ class TagesspiegelRSS(BasicNewsRecipe):
no_javascript = True no_javascript = True
remove_empty_feeds = True remove_empty_feeds = True
encoding = 'utf-8' encoding = 'utf-8'
remove_tags = [{'class':'hcf-header'}, {'class':'hcf-atlas'}, {'class':'hcf-date hcf-separate'}] remove_tags = [{'class':'hcf-header'}, {'class':'hcf-atlas'}, {'class':'hcf-colon'}, {'class':'hcf-date hcf-separate'}]
def print_version(self, url): def print_version(self, url):
url = url.split('/') url = url.split('/')
@ -51,6 +51,7 @@ class TagesspiegelRSS(BasicNewsRecipe):
return ''.join(div.findAll(text=True, recursive=False)).strip() if div is not None else None return ''.join(div.findAll(text=True, recursive=False)).strip() if div is not None else None
articles = {} articles = {}
links = set()
key = None key = None
ans = [] ans = []
maincol = soup.find('div', attrs={'class':re.compile('hcf-main-col')}) maincol = soup.find('div', attrs={'class':re.compile('hcf-main-col')})
@ -59,7 +60,7 @@ class TagesspiegelRSS(BasicNewsRecipe):
if div['class'] == 'hcf-header': if div['class'] == 'hcf-header':
try: try:
key = string.capwords(feed_title(div.em.a)) key = string.capwords(feed_title(div.em))
articles[key] = [] articles[key] = []
ans.append(key) ans.append(key)
except: except:
@ -70,6 +71,12 @@ class TagesspiegelRSS(BasicNewsRecipe):
if not a: if not a:
continue continue
url = 'http://www.tagesspiegel.de' + a['href'] url = 'http://www.tagesspiegel.de' + a['href']
# check for duplicates
if url in links:
continue
links.add(url)
title = self.tag_to_string(a, use_alt=True).strip() title = self.tag_to_string(a, use_alt=True).strip()
description = '' description = ''
pubdate = strftime('%a, %d %b') pubdate = strftime('%a, %d %b')