Add structlog.stdlib.ProcessorFormatter(use_get_message=True)

CHANGELOG.md

@@ -28,6 +28,9 @@ You can find our backwards-compatibility policy [here](https://github.com/hynek/
 
 - `structlog.threadlocal.tmp_bind()` now also works with `BoundLoggerLazyProxy` (in other words: before anything is bound to a bound logger).
 
+- stdlib: `ProcessorFormatter` can now be told to not render the log record message using `getMessage` and just `str(record.msg)` instead.
+  [#550](https://github.com/hynek/structlog/issues/550)
+
 - stdlib: `structlog.stdlib.BoundLogger.exception()`'s handling of`LogRecord.exc_info` is now set consistent with `logging`.
   [#571](https://github.com/hynek/structlog/issues/571)
   [#572](https://github.com/hynek/structlog/issues/572)

src/structlog/stdlib.py

@@ -964,6 +964,10 @@ class ProcessorFormatter(logging.Formatter):
             This parameter exists for historic reasons. Please use *processors*
             instead.
 
+        use_get_message:
+            If True, use ``record.getMessage`` to get a fully rendered log
+            message, otherwise use ``str(record.msg)``. (default: True)
+
     Raises:
 
         TypeError: If both or neither *processor* and *processors* are passed.
@@ -976,6 +980,7 @@ class ProcessorFormatter(logging.Formatter):
     .. deprecated:: 21.3.0
        *processor* (singular) in favor of *processors* (plural). Removal is not
        planned.
+    .. versionadded:: 23.3.0 *use_get_message*
     """
 
     def __init__(
@@ -987,6 +992,7 @@ def __init__(
         keep_stack_info: bool = False,
         logger: logging.Logger | None = None,
         pass_foreign_args: bool = False,
+        use_get_message: bool = True,
         *args: Any,
         **kwargs: Any,
     ) -> None:
@@ -1011,6 +1017,7 @@ def __init__(
         self.keep_stack_info = keep_stack_info
         self.logger = logger
         self.pass_foreign_args = pass_foreign_args
+        self.use_get_message = use_get_message
 
     def format(self, record: logging.LogRecord) -> str:
         """
@@ -1043,7 +1050,9 @@ def format(self, record: logging.LogRecord) -> str:
             logger = self.logger
             meth_name = record.levelname.lower()
             ed = {
-                "event": record.getMessage(),
+                "event": record.getMessage()
+                if self.use_get_message
+                else str(record.msg),
                 "_record": record,
                 "_from_structlog": False,
             }

tests/test_stdlib.py

@@ -1189,6 +1189,34 @@ def emit(self, record):
 
         assert not records
 
+    def test_use_get_message_false(self):
+        """
+        If use_get_message_is False, the event is obtained using
+        str(record.msg) instead of calling record.getMessage. That means
+        positional formatting is not performed.
+        """
+        event_dicts = []
+
+        def capture(_, __, ed):
+            event_dicts.append(ed.copy())
+
+            return str(ed)
+
+        proc = ProcessorFormatter(processors=[capture], use_get_message=False)
+
+        record = logging.LogRecord(
+            "foo",
+            logging.INFO,
+            "path.py",
+            42,
+            "le msg: %s",
+            ("keep separate",),
+            None,
+        )
+
+        assert proc.format(record)
+        assert "le msg: %s" == event_dicts[0]["event"]
+
 
 @pytest_asyncio.fixture(name="abl")
 async def _abl(cl):