DeepSeek token pattern improvement. (#821)

Author: babenek

.github/workflows/check.yml

@@ -93,7 +93,7 @@ jobs:
         run: |
           banner="$(python -m credsweeper --banner | grep CredSweeper | head -1)"
           echo "banner = '${banner}'"
-          if [ "CredSweeper 1.15.0 crc32:439081df" != "${banner}" ]; then
+          if [ "CredSweeper 1.15.1 crc32:58b9cbea" != "${banner}" ]; then
             echo "Update the check for '${banner}'"
             exit 1
           fi

credsweeper/__init__.py

@@ -24,4 +24,4 @@
     "__version__"
 ]
 
-__version__ = "1.15.0"
+__version__ = "1.15.1"

credsweeper/rules/config.yaml

@@ -1363,7 +1363,7 @@
   confidence: moderate
   type: pattern
   values:
-    - (?P<value>sk-[0-9a-f]{32,64})(?![0-9A-Za-z_-])
+    - (?:^|/|[^\\0-9A-Za-z+_-]|\\[0abfnrtv]|(?:%|\\x)[0-9A-Fa-f]{2}|\\[0-7]{3}|\\[Uu][0-9A-Fa-f]{4}|\x1B\[[0-9;]{0,80}m)(?P<value>sk-[0-9a-f]{32,64})(?![0-9A-Za-z_-])
   min_line_len: 35
   filter_type:
     - ValuePatternCheck

tests/samples/deepseek

@@ -1,2 +1,3 @@
 >>>sk-615ef152bf865a131f88b23571dc4ee6<<<
 <<<sk-2f8d9514cce7a8a3f1f34b64acedfake>>>
+scan-disk-615ef152bf865a131f88b23571dc4ee6

tests/test_app.py

@@ -503,7 +503,7 @@ def test_depth_p(self) -> None:
                     cvs_checksum = hashlib.md5(f.read()).digest()
                 checksum = bytes(a ^ b for a, b in zip(checksum, cvs_checksum))
         # update the checksum manually and keep line endings in the samples as is (git config core.autocrlf false)
-        self.assertEqual("021c37fc3b5d65d32667062cc2ae50e7", binascii.hexlify(checksum).decode())
+        self.assertEqual("b5f001b2a8ce5af6271bbcec2e986fa3", binascii.hexlify(checksum).decode())
         normal_report = []
         sorted_report = []
         with tempfile.TemporaryDirectory() as tmp_dir:
@@ -614,7 +614,7 @@ def test_rules_ml_p(self) -> None:
             rules_text = yaml.dump_all(rules, sort_keys=True)
             checksum = hashlib.md5(rules_text.encode()).hexdigest()
             # update the expected value manually if some changes
-            self.assertEqual("64441b2969e9cf41142c1d0fb45450a8", checksum)
+            self.assertEqual("6363b53925305e2f52fa64798c8e6985", checksum)
             rules_set = set([i["name"] for i in rules if "code" in i["target"]])
             self.assertSetEqual(rules_set, report_set)
             self.assertEqual(SAMPLES_POST_CRED_COUNT, len(report))