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 2012-08-25 17:29:14 +05:30
parent 11e0d74428
commit 47fa81f0bd
1 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
recipes/bwmagazine2.recipe
View File

@ -26,7 +26,6 @@ class BusinessWeekMagazine(BasicNewsRecipe):
#Find date #Find date
mag=soup.find('h2',text='Magazine') mag=soup.find('h2',text='Magazine')
self.log(mag)
dates=self.tag_to_string(mag.findNext('h3')) dates=self.tag_to_string(mag.findNext('h3'))
self.timefmt = u' [%s]'%dates self.timefmt = u' [%s]'%dates
@ -34,9 +33,10 @@ class BusinessWeekMagazine(BasicNewsRecipe):
div0 = soup.find ('div', attrs={'class':'column left'}) div0 = soup.find ('div', attrs={'class':'column left'})
section_title = '' section_title = ''
feeds = OrderedDict() feeds = OrderedDict()
articles = []
for div in div0.findAll('a'): for div in div0.findAll('a'):
articles = []
section_title = self.tag_to_string(div.findPrevious('h3')).strip() section_title = self.tag_to_string(div.findPrevious('h3')).strip()
self.log('Processing section:', section_title)
title=self.tag_to_string(div).strip() title=self.tag_to_string(div).strip()
url=div['href'] url=div['href']
soup0 = self.index_to_soup(url) soup0 = self.index_to_soup(url)
@ -50,10 +50,11 @@ class BusinessWeekMagazine(BasicNewsRecipe):
feeds[section_title] += articles feeds[section_title] += articles
div1 = soup.find ('div', attrs={'class':'column center'}) div1 = soup.find ('div', attrs={'class':'column center'})
section_title = '' section_title = ''
articles = []
for div in div1.findAll('a'): for div in div1.findAll('a'):
articles = []
desc=self.tag_to_string(div.findNext('p')).strip() desc=self.tag_to_string(div.findNext('p')).strip()
section_title = self.tag_to_string(div.findPrevious('h3')).strip() section_title = self.tag_to_string(div.findPrevious('h3')).strip()
self.log('Processing section:', section_title)
title=self.tag_to_string(div).strip() title=self.tag_to_string(div).strip()
url=div['href'] url=div['href']
soup0 = self.index_to_soup(url) soup0 = self.index_to_soup(url)