Skip to content

Commit c183ca5

Browse files
committed
My rounding went away somewhere - I fixed that, I think.
Added hash threshold setting. I think this will be find because of the changes made to my hash regex.
1 parent 109f562 commit c183ca5

File tree

2 files changed

+7
-4
lines changed

2 files changed

+7
-4
lines changed

lib/Paste.py

+5-3
Original file line numberDiff line numberDiff line change
@@ -46,10 +46,12 @@ def match(self):
4646
for regex in regexes['db_keywords']:
4747
if regex.search(self.text):
4848
log(regex.search(self.text).group(1))
49-
self.db_keywords += 1/float(len(regexes['db_keywords']))
49+
self.db_keywords += round(1/float(len(regexes['db_keywords'])), 2)
5050
for regex in regexes['blacklist']:
51-
if regex.search(self.text): self.db_keywords -= 1.25 * (1/float(len(regexes['db_keywords'])))
52-
if (self.num_emails >= settings.EMAIL_THRESHOLD) or (self.db_keywords >= settings.DB_KEYWORDS_THRESHOLD):
51+
if regex.search(self.text):
52+
log(regex.search(self.text).group(1))
53+
self.db_keywords -= round(1.25 * (1/float(len(regexes['db_keywords']))), 2)
54+
if (self.num_emails >= settings.EMAIL_THRESHOLD) or (self.num_hashes >= settings.HASH_THRESHOLD) or (self.db_keywords >= settings.DB_KEYWORDS_THRESHOLD):
5355
self.type = 'db_dump'
5456
if regexes['cisco_hash'].search(self.text) or regexes['cisco_pass'].search(self.text): self.type = 'Cisco'
5557
if regexes['google_api'].search(self.text): self.type = 'google_api'

settings.py

+2-1
Original file line numberDiff line numberDiff line change
@@ -8,7 +8,8 @@
88

99
# Thresholds
1010
EMAIL_THRESHOLD = 20
11-
DB_KEYWORDS_THRESHOLD = .50
11+
HASH_THRESHOLD = 30
12+
DB_KEYWORDS_THRESHOLD = .65
1213

1314
# Time to Sleep for each site
1415
SLEEP_SLEXY = 60

0 commit comments

Comments
 (0)