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 CNET News

Browse Source
This commit is contained in:
Kovid Goyal 2019-08-13 20:12:36 +05:30
parent 6c3fd548a5
commit b9d5842e00
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 11 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
recipes/cnetnews.recipe
View File

@ -18,6 +18,12 @@ You may want to set it to download daily for this reason.
from calibre.web.feeds.news import BasicNewsRecipe from calibre.web.feeds.news import BasicNewsRecipe
def classes(classes):
q = frozenset(classes.split(' '))
return dict(attrs={
'class': lambda x: x and frozenset(x.split()).intersection(q)})
class CnetNews(BasicNewsRecipe): class CnetNews(BasicNewsRecipe):
title = 'CNET News' title = 'CNET News'
__author__ = 'Kovid Goyal' __author__ = 'Kovid Goyal'
@ -44,10 +50,12 @@ class CnetNews(BasicNewsRecipe):
'data-component': 'imageGalleryModal'}), 'data-component': 'imageGalleryModal'}),
dict(attrs={'data-component': 'sharebar'}), dict(attrs={'data-component': 'sharebar'}),
dict(name=['link', 'meta']), dict(name=['link', 'meta']),
classes('playerControls video share-button'),
] ]
keep_only_tags = [ keep_only_tags = [
dict(itemprop='headline'), dict(name='h1'),
dict(section='author'),
dict(id=["article-body", 'cnetReview']), dict(id=["article-body", 'cnetReview']),
dict(attrs={'class': 'deal-content'}), dict(attrs={'class': 'deal-content'}),
] ]
@ -89,4 +97,6 @@ class CnetNews(BasicNewsRecipe):
h1.extract() h1.extract()
if first_fetch: if first_fetch:
soup.find('body').insert(1, h1) soup.find('body').insert(1, h1)
for img in soup.findAll('img'):
img['height'] = img['width'] = ''
return soup return soup