Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-11-05 12:03:03 -05:00
Code Issues Packages Projects Releases Wiki Activity

Update nytimes.py

Browse Source 
fix JSON decode error
This commit is contained in:
unkn0w7n 2025-09-13 11:38:44 +05:30
parent 3b036b4853
commit 882b2dd517
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
src/calibre/web/site_parsers/nytimes.py
View File

@ -9,7 +9,7 @@ from xml.sax.saxutils import escape, quoteattr
from calibre.utils.iso8601 import parse_iso8601 from calibre.utils.iso8601 import parse_iso8601
module_version = 14 # needed for live updates module_version = 15 # needed for live updates
pprint pprint
@ -198,8 +198,8 @@ def article_parse(data):
def clean_js_json(text): def clean_js_json(text):
text = re.sub(r'\bundefined\b', 'null', text) text = re.sub(r'\bundefined\b', 'null', text)
text = re.sub( text = re.sub(
r',?\s*"[^"]+"\s*:\s*function\s*\([^)]*\)\s*\{.*?\}', r'{\"checkGate\":.*',
'', 'null}}',
text, text,
flags=re.DOTALL flags=re.DOTALL
) )
@ -207,7 +207,7 @@ def clean_js_json(text):
def json_to_html(raw): def json_to_html(raw):
data = json.JSONDecoder(strict=False).raw_decode(raw)[0] data = json.loads(clean_js_json(raw))
# open('/t/raw.json', 'w').write(json.dumps(data, indent=2)) # open('/t/raw.json', 'w').write(json.dumps(data, indent=2))
try: try:
data = data['initialData']['data'] data = data['initialData']['data']