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

Update HBR

Browse Source
This commit is contained in:
Kovid Goyal 2022-07-27 18:31:44 +05:30
parent 525988b151
commit 7df0187962
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 23 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
recipes/hbr.recipe
View File

@ -25,11 +25,21 @@ class HBR(BasicNewsRecipe):
article-sidebar{font-family:Georgia,"Times New Roman",Times,serif; border:ridge; text-align:left;}
[close-caption]{ border:ridge; font-size:small; text-align:center;}
article-ideainbrief{font-family:Georgia,"Times New Roman",Times,serif; text-align:left; font-style:italic; }
.article-byline-list{font-size:small;}
.credits--hero-image{font-size:small;}
.credits--inline-image{font-size:small;}
.caption--inline-image{font-size:small;}
.description-text{font-size:small; color:gray;}
.right-rail--container{font-size:small; color:#4c4c4c;}
.link--black{font-size:small;}
.article-callout{color:#4c4c4c; text-align:center;}
.slug-content{color:gray;}
'''
keep_only_tags = [
classes(
'headline-container pub-date hero-image-content article-summary article-body standard-content'
'headline-container hero-image-content article-summary article-body standard-content'
' article-dek-group article-dek slug-container'
),
dict(name='article-sidebar'),
]
@ -87,6 +97,18 @@ class HBR(BasicNewsRecipe):
ans = [(key, val) for key, val in feeds.items()]
return ans
def preprocess_html(self, soup):
for slug in soup.findAll(**classes('slug-content')):
del slug['href']
for dek in soup.findAll(**classes('article-byline')):
for by in dek.findAll('span', attrs={'class':'by-prefix'}):
by.extract()
for li in dek.findAll('li'):
li.name = 'span'
for h2 in soup.findAll(('h2','h3')):
h2.name = 'h5'
return soup
# HBR changes the content it delivers based on cookies, so the
# following ensures that we send no cookies
def get_browser(self, *args, **kwargs):