dhkim1920 commented on code in PR #73783:
URL: https://github.com/apache/airflow/pull/73783#discussion_r4572000180


##########
airflow-core/tests/unit/api_fastapi/execution_api/versions/head/test_task_instances.py:
##########
@@ -742,6 +743,45 @@ def 
test_startup_xcom_cleanup_keys_belong_to_the_starting_try(
         assert response.status_code == 200
         assert response.json()["xcom_keys_to_clear"] == ["own"]
 
+    @pytest.mark.parametrize(
+        "conf_value",
+        [
+            pytest.param(-(2**63), id="min-int64"),
+            pytest.param(2**64 - 1, id="max-uint64"),
+            pytest.param(
+                -(2**63) - 1,
+                id="below-int64",
+                marks=pytest.mark.backend("postgres", "sqlite"),
+            ),
+            pytest.param(
+                2**64,
+                id="above-uint64",
+                marks=pytest.mark.backend("postgres", "sqlite"),
+            ),
+            pytest.param(str(2**64), id="large-number-string"),
+        ],
+    )
+    def test_run_conf_can_be_sent_to_task_runner(self, client, session, 
create_task_instance, conf_value):

Review Comment:
   Renamed the test to test_run_returns_large_integer_conf as suggested, 
reflecting its scope of checking the database round trip and API response.



##########
task-sdk/src/airflow/sdk/execution_time/comms.py:
##########
@@ -177,14 +181,40 @@ def _new_encoder() -> msgspec.msgpack.Encoder:
     return msgspec.msgpack.Encoder(enc_hook=_msgpack_enc_hook)
 
 
+def _encode_large_ints(value: Any) -> Any:
+    if isinstance(value, int) and not -(2**63) <= value < 2**64:
+        return msgspec.msgpack.Ext(
+            MSGPACK_HUGEINT, value.to_bytes((value.bit_length() + 8) // 8, 
"big", signed=True)
+        )
+    if isinstance(value, dict):
+        return {key: _encode_large_ints(item) for key, item in value.items()}
+    if isinstance(value, (list, tuple)):
+        return [_encode_large_ints(item) for item in value]
+    return value
+
+
+def _decode_msgpack_ext(code: int, data: memoryview) -> Any:
+    if code == MSGPACK_HUGEINT:
+        return int.from_bytes(data, "big", signed=True)
+    return msgspec.msgpack.Ext(code, bytes(data))
+
+
 class _FrameMixin:
     _encoder: ClassVar[msgspec.msgpack.Encoder] = _new_encoder()
 
-    def as_bytes(self) -> bytearray:
+    def as_bytes(self, *, allow_large_ints: bool = True) -> bytearray:
         # https://jcristharif.com/msgspec/perf-tips.html#length-prefix-framing 
for inspiration
         buffer = bytearray(256)
 
-        self._encoder.encode_into(self, buffer, 4)  # type: ignore[arg-type]
+        try:
+            self._encoder.encode_into(self, buffer, 4)  # type: 
ignore[arg-type]
+        except OverflowError:
+            if not allow_large_ints:
+                raise
+            # Keep the usual path free of a Python traversal of every message.

Review Comment:
   Moved the comment above the try block so it explains why we attempt encoding 
without a Python traversal first.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to