Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-08 10:44:09 -04:00
Code Issues Packages Projects Releases Wiki Activity

Update barrons.recipe

Browse Source
This commit is contained in:
unkn0w7n 2025-02-10 09:37:39 +05:30
parent 5da6c49d8f
commit ff8e342269
1 changed files with 25 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
recipes/barrons.recipe
View File

@ -8,6 +8,28 @@ import re
from calibre.web.feeds.news import BasicNewsRecipe
def make_hlinks(zx):
txt = zx['body']['text']
if zx['body'].get('additions'):
for ad in zx['body']['additions']:
if ad.get('type', '') == 'link':
strt = ad['rangeStart']
end = ad['rangeLength']
n_txt = txt[strt : (strt + end)]
lnk = ad['value']
lnk_txt = f'<a href="{lnk}">{n_txt}</a>'
txt = txt[:strt] + lnk_txt + txt[(strt + end) :]
if zx['body'].get('inlineTextStyles'):
for sty in zx['body']['inlineTextStyles']:
if 'Italic' in sty['textStyle']['fontName']:
estrt = sty['rangeStart']
eend = sty['rangeLength']
etxt = txt[estrt : (estrt + eend)]
em_txt = f'<em>{etxt}</em>'
txt = txt[:estrt] + em_txt + txt[(estrt + eend) :]
return txt
def make_html(a):
typ = a.get('type', '')
if typ == 'byline':
@ -20,14 +42,14 @@ def make_html(a):
if typ == 'body':
if 'body-h' in a['body'].get('textStyleID', ''):
return f'<h4>{a["body"]["text"]}</h4>'
return f'<p>{a["body"]["text"]}</p>'
return f'<p>{make_hlinks(a)}</p>'
if typ == 'image':
return f'<img src={a["image"]["url"]}>'
if typ == 'caption':
st = ' style="font-size:small; text-align: center;"'
return f'<div {st}>{a["caption"]["text"]}</div>'
if typ == 'listelement':
return f'<li>{a["body"]["text"]}</li>'
return f'<li>{make_hlinks(a)}</li>'
if typ == 'dynamicinset':
if 'datawrapper-chart-' in a['webview']['value']:
dw = re.search(r'datawrapper-chart-(.{5})', a['webview']['value']).group(1)
@ -115,7 +137,7 @@ class Barrons(BasicNewsRecipe):
body = ''
for x in rdata['screens'][0]['frames']:
body += '\n' + make_html(x)
return '<html><body>' + body + '</body></html>'
return '<html><body>' + body.replace(' _', '') + '</body></html>'
def get_browser(self, *args, **kw):
kw['user_agent'] = 'okhttp/4.12.0'