fix cnn urls broken bc of amp stripping

master
Aevann 2023-08-24 03:57:03 +03:00
parent 6d90aa8408
commit f8f6e5f37f
1 changed files with 1 additions and 0 deletions

View File

@ -764,6 +764,7 @@ def normalize_url(url):
.replace("/giphy.gif", "/giphy.webp") \ .replace("/giphy.gif", "/giphy.webp") \
.replace('https://www.google.com/amp/s/', 'https://') \ .replace('https://www.google.com/amp/s/', 'https://') \
.replace('https://amp.', 'https://') \ .replace('https://amp.', 'https://') \
.replace('https://cnn.com/cnn/', 'https://edition.cnn.com/') \
.replace('/amp/', '/') \ .replace('/amp/', '/') \
if url.endswith('.amp'): if url.endswith('.amp'):