weiqingy commented on code in PR #867:
URL: https://github.com/apache/flink-agents/pull/867#discussion_r3524228355
##########
python/flink_agents/plan/tests/compatibility/create_python_option_from_java_option.py:
##########
@@ -15,36 +15,27 @@
# See the License for the specific language governing permissions and
# limitations under the License.
#################################################################################
-from pathlib import Path
-
-from pyflink.util.java_utils import add_jars_to_context_class_loader
-
from flink_agents.api.core_options import (
AgentConfigOptions,
AgentExecutionOptions,
+ EventLogLevel,
ShortTermMemoryTtlUpdate,
ShortTermMemoryTtlVisibility,
)
-# This script is used to verify that Java-defined configuration options
-# (e.g., AgentConfigOptions) are correctly exposed and accessible in the
-# Python environment via the JAR file. It loads a Java JAR into the Python
-# context and performs basic assertions on the configuration keys, types,
-# and default values to ensure compatibility between Java and Python layers.
-#
-# The JAR file path is relative to this script and should be updated if
-# the build structure changes.
+# This script verifies that Python configuration options stay aligned with the
Review Comment:
This comment still says the script verifies the Python options stay aligned
with Java, but after dropping the JAR load it only checks the Python
declarations against hardcoded literals — so it would stay green even if the
Java `AgentConfigOptions` side drifted (a new option, a changed default). One
thing worth noting: this script runs as a standalone process in
`test_java_config_in_python.sh`, not inside the Pemja worker, so the `#866`
gateway concern doesn't apply here — it could keep loading the JAR and compare
each Python `ConfigOption` against the Java key/type/default. Would it be worth
keeping that cross-check (here or in a dedicated parity test) so the two sides
can't silently drift?
--
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]