fix: Fix non-UTC timestamps (#3461)

Fixes a bug where all `datetime` timestamps in an event payload were serialized as if they were UTC timestamps, even if they were non-UTC timestamps, completely ignoring the timezone. Now, we convert all datetime objects to UTC before formatting them as a UTC timestamp.

Fixes #3453

Author: szokeasaurusrex

sentry_sdk/utils.py

@@ -11,7 +11,7 @@
 import threading
 import time
 from collections import namedtuple
-from datetime import datetime
+from datetime import datetime, timezone
 from decimal import Decimal
 from functools import partial, partialmethod, wraps
 from numbers import Real
@@ -228,7 +228,15 @@ def to_timestamp(value):
 
 def format_timestamp(value):
     # type: (datetime) -> str
-    return value.strftime("%Y-%m-%dT%H:%M:%S.%fZ")
+    """Formats a timestamp in RFC 3339 format.
+
+    Any datetime objects with a non-UTC timezone are converted to UTC, so that all timestamps are formatted in UTC.
+    """
+    utctime = value.astimezone(timezone.utc)
+
+    # We use this custom formatting rather than isoformat for backwards compatibility (we have used this format for
+    # several years now), and isoformat is slightly different.
+    return utctime.strftime("%Y-%m-%dT%H:%M:%S.%fZ")
 
 
 def event_hint_with_exc_info(exc_info=None):

tests/test_utils.py

@@ -1,7 +1,7 @@
 import threading
 import re
 import sys
-from datetime import timedelta
+from datetime import timedelta, datetime, timezone
 from unittest import mock
 
 import pytest
@@ -13,6 +13,7 @@
     Components,
     Dsn,
     env_to_bool,
+    format_timestamp,
     get_current_thread_meta,
     get_default_release,
     get_error_message,
@@ -950,3 +951,39 @@ def target():
     thread.start()
     thread.join()
     assert (main_thread.ident, main_thread.name) == results.get(timeout=1)
+
+
+@pytest.mark.parametrize(
+    ("datetime_object", "expected_output"),
+    (
+        (
+            datetime(2021, 1, 1, tzinfo=timezone.utc),
+            "2021-01-01T00:00:00.000000Z",
+        ),  # UTC time
+        (
+            datetime(2021, 1, 1, tzinfo=timezone(timedelta(hours=2))),
+            "2020-12-31T22:00:00.000000Z",
+        ),  # UTC+2 time
+        (
+            datetime(2021, 1, 1, tzinfo=timezone(timedelta(hours=-7))),
+            "2021-01-01T07:00:00.000000Z",
+        ),  # UTC-7 time
+        (
+            datetime(2021, 2, 3, 4, 56, 7, 890123, tzinfo=timezone.utc),
+            "2021-02-03T04:56:07.890123Z",
+        ),  # UTC time all non-zero fields
+    ),
+)
+def test_format_timestamp(datetime_object, expected_output):
+    formatted = format_timestamp(datetime_object)
+
+    assert formatted == expected_output
+
+
+def test_format_timestamp_naive():
+    datetime_object = datetime(2021, 1, 1)
+    timestamp_regex = r"\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}.\d{6}Z"
+
+    # Ensure that some timestamp is returned, without error. We currently treat these as local time, but this is an
+    # implementation detail which we should not assert here.
+    assert re.fullmatch(timestamp_regex, format_timestamp(datetime_object))