[python] Introduce merge_map_with_keytime aggregator function

[PyRSA]

Purpose

This PR introduces the Python FieldMergeMapWithKeyTimeAgg implementation and aligns the Python merge_map_with_keytime aggregation behavior with the Java implementation.

Tests

• python -m pytest pypaimon/tests/test_field_aggregators.py::FieldMergeMapWithKeyTimeAggTest -q
• python -m pytest pypaimon/tests/test_aggregation_e2e.py -q
• python -m pytest pypaimon/tests/test_aggregation_merge_function.py -q
• git diff --check

[JingsongLi]

• When fields.<field>.ts-field is not configured, the Java semantics use the last field in the ROW.
• In Python, when calling input_row.get(None) on a PyArrow/dict ROW, the timestamp returned is always empty, causing the new value to be ignored.

[PyRSA]

• When fields.<field>.ts-field is not configured, the Java semantics use the last field in the ROW.
• In Python, when calling input_row.get(None) on a PyArrow/dict ROW, the timestamp returned is always empty, causing the new value to be ignored.

Thanks! Fixed. The Python implementation now matches the Java behavior: if fields.<field>.ts-field is not configured, it resolves the last field of the ROW as the timestamp field during initialization instead of looking up None at runtime. I've also added a test to cover this default timestamp field behavior.

[JingsongLi]

+1