CIRCL
/
AIL-framework
mirror of https://github.com/CIRCL/AIL-framework
2
1
Fork 0
Code Issues Releases Wiki Activity

fix: [module_extractor] don't extract onion from crawled items

master
terrtia 2024-12-09 17:25:23 +01:00
parent f4fca05524
commit f77c9268bb
No known key found for this signature in database
GPG Key ID: 1E1B1F50D84613D0
1 changed files with 3 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
bin/modules/Onion.py
View File

@ -56,6 +56,9 @@ class Onion(AbstractModule):
def extract(self, obj, content, tag): def extract(self, obj, content, tag):
extracted = [] extracted = []
if self.obj.type == 'item':
if 'infoleak:submission="crawler"' in obj.get_tags():
return extracted
onions = self.regex_finditer(self.onion_regex, obj.get_global_id(), content) onions = self.regex_finditer(self.onion_regex, obj.get_global_id(), content)
for onion in onions: for onion in onions:
start, end, value = onion start, end, value = onion