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 2023-10-23 12:49:09 +05:30
parent b9ac7c7ed7
commit d9059b3899
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
2 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
recipes/wsj.recipe
View File

@ -67,7 +67,7 @@ class WSJ(BasicNewsRecipe):
dict(attrs={'data-type':'inset'}), dict(attrs={'data-type':'inset'}),
dict(attrs={'data-spotim-app':'conversation'}), dict(attrs={'data-spotim-app':'conversation'}),
dict(attrs={'data-spot-im-class':['message-text', 'conversation-root']}), dict(attrs={'data-spot-im-class':['message-text', 'conversation-root']}),
dict(attrs={'id':lambda x: x and x.startswith(('comments_sector', wrapper-INLINE', 'audio-tag-inner-audio-'))}), dict(attrs={'id':lambda x: x and x.startswith(('comments_sector', 'wrapper-INLINE', 'audio-tag-inner-audio-'))}),
] ]
articles_are_obfuscated = True articles_are_obfuscated = True

6
recipes/wsj_free.recipe
View File

@ -54,7 +54,7 @@ class WSJ(BasicNewsRecipe):
#big-top-caption { font-size:small; text-align:center; } #big-top-caption { font-size:small; text-align:center; }
[data-type:"tagline"] { font-style:italic; color:#202020; } [data-type:"tagline"] { font-style:italic; color:#202020; }
''' '''
keep_only_tags = [ keep_only_tags = [
dict(name=['h1', 'h2']), dict(name=['h1', 'h2']),
dict(attrs={'aria-describedby':'big-top-caption'}), dict(attrs={'aria-describedby':'big-top-caption'}),
@ -68,7 +68,7 @@ class WSJ(BasicNewsRecipe):
dict(attrs={'data-type':'inset'}), dict(attrs={'data-type':'inset'}),
dict(attrs={'data-spotim-app':'conversation'}), dict(attrs={'data-spotim-app':'conversation'}),
dict(attrs={'data-spot-im-class':['message-text', 'conversation-root']}), dict(attrs={'data-spot-im-class':['message-text', 'conversation-root']}),
dict(attrs={'id':lambda x: x and x.startswith(('comments_sector', wrapper-INLINE', 'audio-tag-inner-audio-'))}), dict(attrs={'id':lambda x: x and x.startswith(('comments_sector', 'wrapper-INLINE', 'audio-tag-inner-audio-'))}),
] ]
articles_are_obfuscated = True articles_are_obfuscated = True
@ -85,7 +85,7 @@ class WSJ(BasicNewsRecipe):
pt.close() pt.close()
return pt.name return pt.name
def preprocess_html(self, soup): def preprocess_html(self, soup):
for img in soup.findAll('img', attrs={'old-src':True}): for img in soup.findAll('img', attrs={'old-src':True}):
img['src'] = img['old-src'] img['src'] = img['old-src']
for p in soup.findAll('div', attrs={'data-type':['paragraph', 'image']}): for p in soup.findAll('div', attrs={'data-type':['paragraph', 'image']}):