Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-07 10:14:46 -04:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'master' of https://github.com/unkn0w7n/calibre

Browse Source
This commit is contained in:
Kovid Goyal 2023-10-19 11:06:11 +05:30
commit d61ad01055
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
2 changed files with 20 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
recipes/wsj.recipe
View File

@ -64,7 +64,8 @@ class WSJ(BasicNewsRecipe):
extra_css = '''
#big-top-caption { font-size:small; text-align:center; }
[data-type:"tagline"] { font-style:italic; color:#202020; }
[data-type:"tagline"], em { font-style:italic; color:#202020; }
.auth { font-size:small; }
'''
keep_only_tags = [
@ -113,6 +114,14 @@ class WSJ(BasicNewsRecipe):
for h2 in soup.findAll('h2'):
if self.tag_to_string(h2).startswith('What to Read Next'):
h2.extract()
for ph in soup.findAll('a', attrs={'data-type':['phrase', 'link']}):
if div := ph.findParent('div'):
div.name = 'span'
for auth in soup.findAll('a', attrs={'aria-label': lambda x: x and x.startswith('Author page')}):
if div := auth.find_previous_sibling('div'):
div.name = 'span'
if parent := auth.findParent('div'):
parent['class'] = 'auth'
return soup
# login {{{

10
recipes/wsj_free.recipe
View File

@ -65,6 +65,8 @@ class WSJ(BasicNewsRecipe):
extra_css = '''
#big-top-caption { font-size:small; text-align:center; }
[data-type:"tagline"] { font-style:italic; color:#202020; }
[data-type:"tagline"], em { font-style:italic; color:#202020; }
.auth { font-size:small; }
'''
keep_only_tags = [
@ -113,6 +115,14 @@ class WSJ(BasicNewsRecipe):
for h2 in soup.findAll('h2'):
if self.tag_to_string(h2).startswith('What to Read Next'):
h2.extract()
for ph in soup.findAll('a', attrs={'data-type':['phrase', 'link']}):
if div := ph.findParent('div'):
div.name = 'span'
for auth in soup.findAll('a', attrs={'aria-label': lambda x: x and x.startswith('Author page')}):
if div := auth.find_previous_sibling('div'):
div.name = 'span'
if parent := auth.findParent('div'):
parent['class'] = 'auth'
return soup
# login {{{