fix(logger): log non-ascii characters as is when JSON stringifying (#…

…3475) * fix(parameters): make cache aware of single vs multiple calls Signed-off-by: heitorlessa <[email protected]> * chore: cleanup, add test for single and nested Signed-off-by: heitorlessa <[email protected]> * fix(logger): utf-8 encoding json Signed-off-by: heitorlessa <[email protected]> * chore: add all non_ascii compatible with 3.7+ * chore: mention issue for future debuggability --------- Signed-off-by: heitorlessa <[email protected]>
aws-powertools · Dec 11, 2023 · 419e05b · 419e05b
1 parent 8839f9c
commit 419e05b
Show file tree

Hide file tree

Showing 2 changed files with 21 additions and 0 deletions.
diff --git a/aws_lambda_powertools/logging/formatter.py b/aws_lambda_powertools/logging/formatter.py
@@ -133,6 +133,7 @@ def __init__(
             default=self.json_default,
             separators=(",", ":"),
             indent=self.json_indent,
+            ensure_ascii=False,  # see #3474
         )
 
         self.datefmt = datefmt

diff --git a/tests/functional/test_logger.py b/tests/functional/test_logger.py
@@ -1120,3 +1120,23 @@ def filter(self, record):
     log = capture_multiple_logging_statements_output(stdout)
     assert log[0]["api_key"] == "REDACTED"
     assert log[1]["api_key"] != "REDACTED"
+
+
+def test_logger_json_unicode(stdout, service_name):
+    # GIVEN Logger is initialized
+    logger = Logger(service=service_name, stream=stdout)
+
+    # WHEN all non-ascii chars are logged as messages
+    # AND non-ascii is also used as ephemeral fields
+    # latest: https://www.unicode.org/versions/Unicode15.1.0/#Summary
+    non_ascii_chars = [chr(i) for i in range(128, 111_411_1)]
+    japanese_field = "47業レルし化"
+    japanese_string = "スコビルデモ２"
+
+    logger.info(non_ascii_chars, **{japanese_field: japanese_string})
+
+    # THEN JSON logs should not try to escape them
+    log = capture_logging_output(stdout)
+
+    assert log["message"] == non_ascii_chars
+    assert log[japanese_field] == japanese_string