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 wsj.recipe

Browse Source
This commit is contained in:
unkn0w7n 2024-07-13 12:14:11 +05:30
parent 484e70ab96
commit 6fe43235a6
1 changed files with 8 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
recipes/wsj.recipe
View File

@ -13,7 +13,7 @@ past_edition = None
def media_bucket(x): def media_bucket(x):
if x.get('type', '') == 'image': if x.get('type', '') == 'image':
if x.get('subtype', '') == 'graphic': if x.get('subtype', '') == 'graphic' or 'images.wsj.net' not in x['manifest-url']:
return '<br><img src="{}"><div class="figc">{}</div>\n'.format( return '<br><img src="{}"><div class="figc">{}</div>\n'.format(
x['manifest-url'], x['caption'] + '<i> ' + x['credit'] + '</i>' x['manifest-url'], x['caption'] + '<i> ' + x['credit'] + '</i>'
) )
@ -82,6 +82,7 @@ class WSJ(BasicNewsRecipe):
byl.insert(0, read) byl.insert(0, read)
url = soup.find('p', attrs={'id':'share-link'}) url = soup.find('p', attrs={'id':'share-link'})
if url: if url:
url.name = 'div'
url['title'] = self.tag_to_string(url).strip() url['title'] = self.tag_to_string(url).strip()
url.string = '' url.string = ''
panel = soup.find('panel', attrs={'id':'metadata'}) panel = soup.find('panel', attrs={'id':'metadata'})
@ -97,6 +98,11 @@ class WSJ(BasicNewsRecipe):
x.insert_after(BeautifulSoup(y, 'html.parser')) x.insert_after(BeautifulSoup(y, 'html.parser'))
return soup return soup
def postprocess_html(self, soup, first_fetch):
for pan in soup.findAll('panel'):
pan.name = 'div'
return soup
if not past_edition: if not past_edition:
def _download_cover(self): def _download_cover(self):
import os import os
@ -178,6 +184,6 @@ class WSJ(BasicNewsRecipe):
return BeautifulSoup(raw).prettify() return BeautifulSoup(raw).prettify()
def populate_article_metadata(self, article, soup, first): def populate_article_metadata(self, article, soup, first):
lnk = soup.find('p', attrs={'id':'share-link'}) lnk = soup.find('div', attrs={'id':'share-link'})
if lnk: if lnk:
article.url = lnk['title'] article.url = lnk['title']