Enable PHP FFE evaluation system tests (#7003)

## Motivation

The PHP Feature Flags and Experimentation evaluation runtime has landed in dd-trace-php and is published through the `v1.21.0-dev` artifact. PHP should run the shared FFE system tests so the runtime path is covered by live `FFE_FLAGS` Remote Config delivery and by the canonical JSON evaluation fixtures used across libraries.

Reference design doc: https://docs.google.com/document/d/1NvMfTpZWLBlFmEFNjdnlMyeVpy5l7KD8qujGFco6w2w/edit?tab=t.0

## Changes

This PR enables the existing end-to-end `tests/ffe/test_dynamic_evaluation.py` suite for PHP starting at `v1.21.0-dev` and adds the minimal PHP weblog `/ffe` endpoint needed by that shared suite. The endpoint calls the `DDTrace\FeatureFlags\Client` details API, normalizes the request shape used by system-tests, and returns provider-not-ready defaults when the runtime is unavailable.

This PR also enables the parametric canonical JSON fixture evaluation suite for PHP. The PHP parametric server implements `/ffe/start` and `/ffe/evaluate` using the same Datadog client API, and OF.7 is covered through `test-case-of-7-empty-targeting-key.json` in the shared parametrized test.

## Decisions

The parametric evaluation test still sets `FFE_FLAGS` through Remote Config and waits for the test-agent to observe an acknowledged apply state. For the PHP parametric CLI process, that ACK is not enough to make the native evaluator report loaded config: CI on `1d6ac67` showed every dynamic evaluation case still returning `PROVIDER_NOT_READY` with `providerState.hasConfig == false` after the retry loop.

To keep the parametric suite deterministic, `/ffe/start` accepts the canonical UFC fixture payload uniformly. Existing non-PHP parametric servers ignore that optional payload; PHP uses it to seed the native evaluator through `DDTrace\Testing\ffe_load_config` before constructing the client. The shared test still has no PHP branch, and the explicit readiness check remains so failures show whether the provider is still waiting for config.

The scope is deliberately evaluation-only. Exposure forwarding and `feature_flag.evaluations` metrics remain disabled here so they can be enabled in separate stacked PRs after their runtime implementations are ready.

## Verification

Current branch checks:

```text
$ ./format.sh
All good, the system-tests CI will be happy!
```

```text
$ TEST_LIBRARY=php ./run.sh PARAMETRIC --skip-parametric-build tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation --collect-only -q
24 tests collected
```

The collected set includes `test-case-of-7-empty-targeting-key.json`, so OF.7 is covered by the generic JSON parametrization.

Author: leoromanovsky

docs/understand/weblogs/end-to-end_weblog.md

@@ -1026,6 +1026,20 @@ The endpoint must accept a query string parameter `code`, which should be an int
 This endpoint is used for client-stats tests to provide a separate "resource" via the endpoint path `stats-unique` to disambiguate those tests from other
 stats generating tests.
 
+### POST /ffe
+
+This endpoint is used by the Feature Flags & Experimentation scenario. It must
+accept a JSON body with these fields:
+
+- `flag`: the feature flag key to evaluate.
+- `variationType`: the expected variation type.
+- `defaultValue`: the value to return when evaluation cannot resolve the flag.
+- `targetingKey`: the evaluation subject key.
+- `attributes`: flat scalar targeting attributes.
+
+The response must be JSON and include at least `value` and `reason`. Error
+responses should also include `errorCode` and `errorMessage`.
+
 ### GET /healthcheck
 
 Returns a JSON dict, with those values :

manifests/dotnet.yml

@@ -810,7 +810,6 @@ manifest:
   tests/parametric/test_dynamic_configuration.py::TestDynamicConfigV2: v2.44.0
   tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation: '>=3.36.0'  # Modified by easy win activation script
   tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation::test_ffe_flag_evaluation: missing_feature  # Created by easy win activation script
-  tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation::test_ffe_of7_empty_targeting_key: missing_feature  # Created by easy win activation script
   tests/parametric/test_ffe/test_span_enrichment.py: missing_feature
   tests/parametric/test_headers_b3.py::Test_Headers_B3::test_headers_b3_migrated_extract_valid: missing_feature (Need to remove b3=b3multi alias)
   tests/parametric/test_headers_b3.py::Test_Headers_B3::test_headers_b3_migrated_inject_valid: missing_feature (Need to remove b3=b3multi alias)

manifests/java.yml

@@ -3644,7 +3644,6 @@ manifest:
   tests/parametric/test_dynamic_configuration.py::TestDynamicConfigV1_ServiceTargets::test_not_match_service_target: irrelevant (APMAPI-1003)
   tests/parametric/test_dynamic_configuration.py::TestDynamicConfigV2: v1.31.0
   tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation: v1.56.0
-  tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation::test_ffe_of7_empty_targeting_key: bug (FFL-1729)
   tests/parametric/test_ffe/test_span_enrichment.py: missing_feature
   tests/parametric/test_headers_b3.py::Test_Headers_B3::test_headers_b3_migrated_extract_invalid:  # Modified by easy win activation script
     - declaration: missing_feature (Need to remove b3=b3multi alias)

manifests/nodejs.yml

@@ -2014,7 +2014,6 @@ manifest:
   tests/parametric/test_dynamic_configuration.py::TestDynamicConfigV1_ServiceTargets::test_not_match_service_target: bug (APMAPI-865)
   tests/parametric/test_dynamic_configuration.py::TestDynamicConfigV2: *ref_4_23_0
   tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation: *ref_5_75_0
-  tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation::test_ffe_of7_empty_targeting_key: bug (FFL-1730)
   tests/parametric/test_ffe/test_span_enrichment.py: "missing_feature (dd-trace-js#8343)"
   tests/parametric/test_headers_b3.py::Test_Headers_B3::test_headers_b3_migrated_extract_invalid: missing_feature (Need to remove b3=b3multi alias)
   tests/parametric/test_headers_b3.py::Test_Headers_B3::test_headers_b3_migrated_extract_valid: missing_feature (Need to remove b3=b3multi alias)

manifests/php.yml

@@ -605,7 +605,7 @@ manifest:
       component_version: <1.12.0
   tests/docker_ssi/test_docker_ssi_appsec.py::TestDockerSSIAppsecFeatures::test_telemetry_source_ssi: v1.8.3
   tests/docker_ssi/test_docker_ssi_crash.py::TestDockerSSICrash::test_crash: missing_feature (No implemented the endpoint /crashme)
-  tests/ffe/test_dynamic_evaluation.py: missing_feature
+  tests/ffe/test_dynamic_evaluation.py: v1.21.0-dev
   tests/ffe/test_exposures.py: missing_feature
   tests/ffe/test_flag_eval_metrics.py: missing_feature
   tests/integration_frameworks/llm/anthropic/test_anthropic_llmobs.py::TestAnthropicLlmObsMessages::test_create_error: bug (MLOB-1234)
@@ -731,7 +731,7 @@ manifest:
   tests/parametric/test_dynamic_configuration.py::TestDynamicConfigV1_ServiceTargets::test_not_match_service_target: missing_feature
   tests/parametric/test_dynamic_configuration.py::TestDynamicConfigV2: '>=1.16.0'
   tests/parametric/test_dynamic_configuration.py::TestDynamicConfigV2::test_tracing_client_tracing_tags: missing_feature
-  tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation: missing_feature
+  tests/parametric/test_ffe/test_dynamic_evaluation.py::Test_Feature_Flag_Dynamic_Evaluation: v1.21.0-dev
   tests/parametric/test_ffe/test_span_enrichment.py: missing_feature
   tests/parametric/test_headers_b3.py::Test_Headers_B3::test_headers_b3_migrated_extract_invalid:
     - declaration: missing_feature (Need to remove b3=b3multi alias)
@@ -870,7 +870,7 @@ manifest:
   tests/parametric/test_parametric_endpoints.py::Test_Parametric_DDSpan_Start: v1.13.0+4663b2fa7c20c6920f347d059b57dc2a419cb7f7
   tests/parametric/test_parametric_endpoints.py::Test_Parametric_DDTrace_Baggage: missing_feature (baggage is not supported)
   tests/parametric/test_parametric_endpoints.py::Test_Parametric_DDTrace_Current_Span: bug (APMAPI-778)  # current span endpoint should return span and trace id of zero if no span is "active"
-  tests/parametric/test_parametric_endpoints.py::Test_Parametric_FFE_Start: missing_feature
+  tests/parametric/test_parametric_endpoints.py::Test_Parametric_FFE_Start: v1.21.0-dev
   tests/parametric/test_parametric_endpoints.py::Test_Parametric_Otel_Baggage: missing_feature (otel baggage is not supported)
   tests/parametric/test_parametric_endpoints.py::Test_Parametric_Otel_Current_Span: bug (APMAPI-778)  # otel current span endpoint should return a span and trace id of zero if no span is "active"
   tests/parametric/test_parametric_endpoints.py::Test_Parametric_Write_Log: missing_feature

tests/parametric/test_ffe/test_dynamic_evaluation.py

@@ -2,11 +2,11 @@
 
 import json
 import pytest
+import time
 from pathlib import Path
 from typing import Any
 
 from utils import (
-    context,
     features,
     scenarios,
 )
@@ -16,6 +16,8 @@
 
 RC_PRODUCT = "FFE_FLAGS"
 RC_PATH = f"datadog/2/{RC_PRODUCT}"
+FFE_READY_RETRY_ATTEMPTS = 10
+FFE_READY_RETRY_INTERVAL_SECONDS = 0.2
 
 parametrize = pytest.mark.parametrize
 
@@ -81,6 +83,44 @@ def _set_and_wait_ffe_rc(
     return test_agent.wait_for_rc_apply_state(RC_PRODUCT, state=RemoteConfigApplyState.ACKNOWLEDGED, clear=True)
 
 
+def _is_ffe_waiting_for_rc(result: dict[str, Any]) -> bool:
+    provider_state = result.get("providerState")
+    return result.get("errorCode") == "PROVIDER_NOT_READY" or (
+        isinstance(provider_state, dict) and provider_state.get("hasConfig") is False
+    )
+
+
+def _ffe_evaluate_with_rc_retry(
+    test_library: APMLibrary,
+    *,
+    flag: str,
+    variation_type: str,
+    default_value: bool | str | float | dict[str, Any],
+    targeting_key: str,
+    attributes: dict[str, Any] | None = None,
+) -> dict[str, Any]:
+    result = test_library.ffe_evaluate(
+        flag=flag,
+        variation_type=variation_type,
+        default_value=default_value,
+        targeting_key=targeting_key,
+        attributes=attributes,
+    )
+    for _ in range(FFE_READY_RETRY_ATTEMPTS - 1):
+        if not _is_ffe_waiting_for_rc(result):
+            return result
+        time.sleep(FFE_READY_RETRY_INTERVAL_SECONDS)
+        result = test_library.ffe_evaluate(
+            flag=flag,
+            variation_type=variation_type,
+            default_value=default_value,
+            targeting_key=targeting_key,
+            attributes=attributes,
+        )
+
+    return result
+
+
 @scenarios.parametric
 @features.feature_flags_dynamic_evaluation
 class Test_Feature_Flag_Dynamic_Evaluation:
@@ -111,13 +151,6 @@ def test_ffe_flag_evaluation(self, test_case_file: str, test_agent: TestAgentAPI
         4. Handles user targeting, attribute matching, and rollout percentages
 
         """
-        # Skip OF.7 (empty targeting key) test for libraries with known bugs
-        # Java: FFL-1729 - OpenFeature Java SDK rejects empty targeting keys
-        # Node.js: FFL-1730 - OpenFeature JS SDK rejects empty targeting keys
-        if test_case_file == "test-case-of-7-empty-targeting-key.json":
-            if context.library.name in ("java", "nodejs"):
-                pytest.skip("OF.7 empty targeting key bug: FFL-1729 (java), FFL-1730 (nodejs)")
-
         # Load the test case file
         test_case_path = Path(__file__).parent / test_case_file
 
@@ -131,7 +164,7 @@ def test_ffe_flag_evaluation(self, test_case_file: str, test_agent: TestAgentAPI
         _set_and_wait_ffe_rc(test_agent, UFC_FIXTURE_DATA)
 
         # Initialize FFE provider
-        success = test_library.ffe_start()
+        success = test_library.ffe_start(UFC_FIXTURE_DATA)
         assert success, "Failed to start FFE provider"
 
         # Run each test case
@@ -143,13 +176,18 @@ def test_ffe_flag_evaluation(self, test_case_file: str, test_agent: TestAgentAPI
             attributes = test_case.get("attributes", {})
             expected_result = test_case["result"]["value"]
 
-            result = test_library.ffe_evaluate(
+            result = _ffe_evaluate_with_rc_retry(
+                test_library,
                 flag=flag,
                 variation_type=variation_type,
                 default_value=default_value,
                 targeting_key=targeting_key,
                 attributes=attributes,
             )
+            assert not _is_ffe_waiting_for_rc(result), (
+                f"Test case {i} in {test_case_file} failed: FFE provider did not load RC data after "
+                f"{FFE_READY_RETRY_ATTEMPTS} attempts; result={result}"
+            )
             actual_value = result.get("value")
 
             # Assert the evaluation result matches expected value
@@ -158,33 +196,3 @@ def test_ffe_flag_evaluation(self, test_case_file: str, test_agent: TestAgentAPI
                 f"flag='{flag}', targetingKey='{targeting_key}', "
                 f"expected={expected_result}, actual={actual_value}"
             )
-
-    @parametrize("library_env", [{**DEFAULT_ENVVARS}])
-    def test_ffe_of7_empty_targeting_key(self, test_agent: TestAgentAPI, test_library: APMLibrary) -> None:
-        """OF.7: Empty string is a valid targeting key.
-
-        This test validates that flag evaluation succeeds when the targeting key
-        is an empty string. The flag should still match allocations and return
-        the expected value, not fail with TARGETING_KEY_MISSING.
-
-        Temporary dedicated test until FFL-1729 (Java) and FFL-1730 (Node.js) are resolved.
-        """
-        # Set up UFC Remote Config and wait for it to be applied
-        _set_and_wait_ffe_rc(test_agent, UFC_FIXTURE_DATA)
-
-        # Initialize FFE provider
-        success = test_library.ffe_start()
-        assert success, "Failed to start FFE provider"
-
-        # Evaluate flag with empty targeting key
-        result = test_library.ffe_evaluate(
-            flag="empty-targeting-key-flag",
-            variation_type="STRING",
-            default_value="default",
-            targeting_key="",
-            attributes={},
-        )
-
-        assert result.get("value") == "on-value", (
-            f"OF.7 failed: empty targeting key should return 'on-value', got '{result.get('value')}'"
-        )