From 0f84b9b01ad627c4c7c12eedd4520906bc746570 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Wed, 11 Mar 2026 10:52:04 +0000
Subject: [PATCH 1/2] Initial plan


From 5f00b31e02211516f81195205fbd53f2b90b9801 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Wed, 11 Mar 2026 11:18:16 +0000
Subject: [PATCH 2/2] Fix generate_logformat_regex to escape special regex
 characters in literal parts

Co-authored-by: viktorbeck98 <95294548+viktorbeck98@users.noreply.github.com>
---
 .../utils/log_format_utils.py                 |  4 +-
 tests/test_utils/test_log_format_utils.py     | 55 +++++++++++++++++++
 2 files changed, 58 insertions(+), 1 deletion(-)

diff --git a/src/detectmatelibrary/utils/log_format_utils.py b/src/detectmatelibrary/utils/log_format_utils.py
index 5f7c04e..a56478f 100644
--- a/src/detectmatelibrary/utils/log_format_utils.py
+++ b/src/detectmatelibrary/utils/log_format_utils.py
@@ -46,7 +46,9 @@ def generate_logformat_regex(log_format: str) -> Tuple[list[str], re.Pattern[str
     regex_str = ''
     for k in range(len(splitters)):
         if k % 2 == 0:
-            splitter = re.sub(r' +', r'\\s+', splitters[k])
+            # Split on runs of spaces, escape each part, then rejoin with \s+
+            parts = re.split(r' +', splitters[k])
+            splitter = r'\s+'.join(re.escape(p) for p in parts)
             regex_str += splitter
         else:
             header = splitters[k].strip('<').strip('>')
diff --git a/tests/test_utils/test_log_format_utils.py b/tests/test_utils/test_log_format_utils.py
index b79d6b8..c6c1044 100644
--- a/tests/test_utils/test_log_format_utils.py
+++ b/tests/test_utils/test_log_format_utils.py
@@ -1,9 +1,64 @@
 from datetime import datetime, timezone
 
+from detectmatelibrary.utils.log_format_utils import generate_logformat_regex
 from detectmatelibrary.utils.time_format_handler import TimeFormatHandler
 
 tfh = TimeFormatHandler()
 
+
+class TestGenerateLogformatRegex:
+    def test_simple_format(self):
+        headers, regex = generate_logformat_regex("<Time> <Content>")
+        assert headers == ["Time", "Content"]
+        m = regex.match("2024-01-01 hello world")
+        assert m is not None
+        assert m.group("Time") == "2024-01-01"
+        assert m.group("Content") == "hello world"
+
+    def test_special_characters_in_format_literals(self):
+        """Parentheses in literal parts of the format must be treated as literals,
+        not as regex group delimiters.  This is the root cause of the LOGIN bug:
+        the format 'type=<Type> msg=audit(<Time>): <Content>' previously generated
+        a regex where the '(' before <Time> opened an unnamed capture group,
+        causing Time to capture '(1642723741.076:377)' (with parentheses) instead
+        of '1642723741.076:377'."""
+        log_format = "type=<Type> msg=audit(<Time>): <Content>"
+        headers, regex = generate_logformat_regex(log_format)
+
+        audit_lines = [
+            ("type=USER_ACCT msg=audit(1642723741.072:375): pid=10125 uid=0",
+             "USER_ACCT", "1642723741.072:375", "pid=10125 uid=0"),
+            ("type=CRED_ACQ msg=audit(1642723741.072:376): pid=10125 uid=0",
+             "CRED_ACQ", "1642723741.072:376", "pid=10125 uid=0"),
+            ("type=LOGIN msg=audit(1642723741.076:377): pid=10125 uid=0 old-auid=4294967295 auid=0 tty=(none) old-ses=4294967295 ses=65 res=1",
+             "LOGIN", "1642723741.076:377", "pid=10125 uid=0 old-auid=4294967295 auid=0 tty=(none) old-ses=4294967295 ses=65 res=1"),
+        ]
+        for log, expected_type, expected_time, expected_content in audit_lines:
+            m = regex.match(log)
+            assert m is not None, f"No match for log: {log!r}"
+            assert m.group("Type") == expected_type
+            # Time must NOT include the surrounding parentheses
+            assert m.group("Time") == expected_time, (
+                f"Time for '{expected_type}' should be '{expected_time}', "
+                f"got '{m.group('Time')}' (parentheses not properly escaped)"
+            )
+            assert m.group("Content") == expected_content
+
+    def test_timestamp_extraction_from_audit_format(self):
+        """After the fix, the extracted Time value must be parseable as a
+        float-based Unix timestamp (as used in _extract_timestamp)."""
+        log_format = "type=<Type> msg=audit(<Time>): <Content>"
+        _, regex = generate_logformat_regex(log_format)
+
+        log = "type=LOGIN msg=audit(1642723741.076:377): pid=10125 uid=0 res=1"
+        m = regex.match(log)
+        assert m is not None
+        time_val = m.group("Time")
+        # _extract_timestamp does: int(float(time_val.split(":")[0]))
+        time_part = time_val.split(":")[0]
+        parsed = int(float(time_part))
+        assert parsed == 1642723741
+
 def test_iso_z():
     s = "2023-01-02T03:04:05Z"
     expected = int(datetime.fromisoformat("2023-01-02T03:04:05+00:00").timestamp())