bellingcat/auto-archiver
1
0
Fork 0
mirror of https://github.com/bellingcat/auto-archiver.git synced 2026-06-12 05:08:28 +03:00
Code Issues Packages Projects Releases Wiki Activity

improving ignored content from waczs

Browse Source
This commit is contained in:
msramalho
2023-07-28 12:19:14 +01:00
parent 7a5c9c65bd
commit aa71c85a98
3 changed files with 32 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
src/auto_archiver/utils/url.py
View File

@@ -1,7 +1,6 @@
import re
from urllib.parse import urlparse, urlunparse
class UrlUtil:
telegram_private = re.compile(r"https:\/\/t\.me(\/c)\/(.+)\/(\d+)")
is_istagram = re.compile(r"https:\/\/www\.instagram\.com")
@@ -44,6 +43,12 @@ class UrlUtil:
# twitter profile pictures
if "twimg.com/profile_images" in url: return False
if "twimg.com" in url and "/default_profile_images" in url: return False
# instagram profile pictures
if "https://scontent.cdninstagram.com/" in url and "150x150" in url: return False
# instagram recurring images
if "https://static.cdninstagram.com/rsrc.php/" in url: return False
return True
@staticmethod