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 The Federalist

Browse Source
This commit is contained in:
Kovid Goyal 2022-03-12 08:43:53 +05:30
parent 5989f3cef7
commit 78333c8683
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 18 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
recipes/the_federalist.recipe
View File

@ -30,7 +30,7 @@ class Federalist(BasicNewsRecipe):
keep_only_tags = [ keep_only_tags = [
classes( classes(
'title-lg post-categories article-excerpt article-author-details' 'title-lg article-thumbnail post-categories article-excerpt article-author-details'
' article-meta-author article-meta-date article-content article-body shortbio entry-header' ' article-meta-author article-meta-date article-content article-body shortbio entry-header'
' byline-month byline-standard alpha-byline article-author-description article-author-details'), ' byline-month byline-standard alpha-byline article-author-description article-author-details'),
] ]
@ -44,9 +44,24 @@ class Federalist(BasicNewsRecipe):
] ]
# def parse_index(self): # def parse_index(self):
# return [('Articles', [{ # return [('Articles', [
# {
# 'title': 'test', # 'title': 'test',
# 'url': 'https://thefederalist.com/2022/03/09/propaganda-press-wield-bidens-russia-blame-game-to-gaslight-americans-about-expensive-gas/'}])] # 'url': 'https://thefederalist.com/2022/03/09/propaganda-press-wield-bidens-russia-blame-game-to-gaslight-americans-about-expensive-gas/'},
# {
# 'title': 'test2',
# 'url': 'https://thefederalist.com/2022/03/10/white-house-will-blame-anyone-but-biden-for-februarys-7-9-inflation-jump/',
# }
# ])]
def preprocess_raw_html_(self, raw_html, url):
soup = self.index_to_soup(raw_html)
# this website puts article-thumbnail images inside article-body in
# some articles and outside it in others, so we have to special case it
for ab in soup.findAll(**classes('article-body')):
for img in ab.findAll(**classes('article-thumbnail')):
del img['class']
return str(soup)
def preprocess_html(self, soup): def preprocess_html(self, soup):
for img in soup.findAll('img', attrs={'data-lazy-src': True}): for img in soup.findAll('img', attrs={'data-lazy-src': True}):