fix: resolve conflict markers and fix O(n²) edge case in parser_cef

The feature commit accidentally committed unresolved stash conflict markers
into parser_cef.py, making it invalid Python. Resolves to _parse_cef_ext().

Also fixes an O(n²) scan in _parse_cef_ext: when the key scanner reaches
end-of-string with no '=' remaining, the old code rewound to start+1 and
rescanned — degenerating to quadratic on '='-free input. Now breaks out of
the loop immediately.

Co-Authored-By: Claude <noreply@anthropic.com>

Author: sbylica-splunk

package/shared/pylib/parser_cef.py

@@ -35,7 +35,9 @@ class overlapped the trailing whitespace delimiter, so the engine could split
         while p < n and data[p] != "=" and not data[p].isspace():
             p += 1
         if p == start or p >= n or data[p] != "=":
-            p = start + 1  # not a valid "key=" here; advance and retry
+            if p >= n:
+                break  # no '=' anywhere ahead; nothing more to find
+            p = p + 1  # skip the space or bare '='
             continue
         key = data[start:p]
         p += 1  # skip '='
@@ -81,11 +83,7 @@ def parse(self, log_message):
         try:
             data = log_message.get_as_str(".metadata.cef.ext", "")
 
-<<<<<<< Updated upstream
-            rpairs = re.findall(r"([^=\s]+)=((?:\\=|[^=])+)(?:\s|$)", data)
-=======
             rpairs = _parse_cef_ext(data)
->>>>>>> Stashed changes
             pairs = {}
             keys = []
             for p in rpairs: