Fix remaining mock paths and dynamic aggregation test in util-genai-evals

- Fix 4 instances of handler._load_completion_callbacks -> utils.load_completion_callbacks in test_evaluators.py
- Fix test_evaluation_dynamic_aggregation.py to set _aggregate_results to None instead of False to enable dynamic environment variable reading as per actual implementation

(cherry picked from commit a20742f)

Author: shuningc

util/opentelemetry-util-genai-evals/tests/test_evaluation_dynamic_aggregation.py

@@ -32,10 +32,10 @@ def test_dynamic_aggregation_env_toggle(monkeypatch):  # type: ignore[no-untyped
         [EvaluationResult(metric_name="bias", score=0.1)],
         [EvaluationResult(metric_name="toxicity", score=0.3)],
     ]
-    # Disable internal aggregate flag
-    manager._aggregate_results = False
+    # Set internal aggregate flag to None to enable dynamic env var reading
+    manager._aggregate_results = None
     manager._publish_results(invocation, buckets)
-    assert len(handler.calls) == 2  # two separate batches
+    assert len(handler.calls) == 2  # two separate batches (env var not set)
 
     # Now enable aggregation via env and emit again -> should aggregate
     monkeypatch.setenv(  # type: ignore[attr-defined]

util/opentelemetry-util-genai-evals/tests/test_evaluators.py

@@ -237,7 +237,7 @@ def _build_invocation(self) -> LLMInvocation:
     )
     def test_handler_registers_manager(self) -> None:
         with patch(
-            "opentelemetry.util.genai.handler._load_completion_callbacks",
+            "opentelemetry.util.genai.utils.load_completion_callbacks",
             side_effect=_mock_load_callbacks,
         ):
             handler = get_telemetry_handler()
@@ -263,7 +263,7 @@ def test_handler_registers_manager(self) -> None:
     )
     def test_handler_evaluate_llm_returns_results(self) -> None:
         with patch(
-            "opentelemetry.util.genai.handler._load_completion_callbacks",
+            "opentelemetry.util.genai.utils.load_completion_callbacks",
             side_effect=_mock_load_callbacks,
         ):
             handler = get_telemetry_handler()
@@ -287,7 +287,7 @@ def test_handler_auto_enables_when_env_missing(self) -> None:
                 return_value={"LLMInvocation": ("static_metric",)},
             ),
             patch(
-                "opentelemetry.util.genai.handler._load_completion_callbacks",
+                "opentelemetry.util.genai.utils.load_completion_callbacks",
                 side_effect=_mock_load_callbacks,
             ),
         ):
@@ -305,7 +305,7 @@ def test_handler_auto_enables_when_env_missing(self) -> None:
     )
     def test_handler_disables_when_none(self) -> None:
         with patch(
-            "opentelemetry.util.genai.handler._load_completion_callbacks",
+            "opentelemetry.util.genai.utils.load_completion_callbacks",
             side_effect=_mock_load_callbacks,
         ):
             handler = get_telemetry_handler()