[Executor] Support passing connection provider to script executor (#3004

)

# Description

Currently we use `ConnectionProvider.get_instance` to get connections in
script executor, and it doesn't support passing connections to it
directly, while the flow executor can accept connections, which is
inconsistent. For flow serving, we want the behaviors to be consistent
to refine the conenction resolving logic, so in this PR, we support
directly passing connections to script executor.

# All Promptflow Contribution checklist:
- [x] **The pull request does not introduce [breaking changes].**
- [ ] **CHANGELOG is updated for new features, bug fixes or other
significant changes.**
- [x] **I have read the [contribution guidelines](../CONTRIBUTING.md).**
- [ ] **Create an issue and link to the pull request to get dedicated
review from promptflow team. Learn more: [suggested
workflow](../CONTRIBUTING.md#suggested-workflow).**

## General Guidelines and Best Practices
- [x] Title of the pull request is clear and informative.
- [x] There are a small number of commits, each of which have an
informative message. This means that previously merged commits do not
appear in the history of the PR. For more information on cleaning up the
commits in your PR, [see this
page](https://github.com/Azure/azure-powershell/blob/master/documentation/development-docs/cleaning-up-commits.md).

### Testing Guidelines
- [x] Pull request includes test coverage for the included changes.

Co-authored-by: Lina Tang <[email protected]>

src/promptflow-core/promptflow/executor/_script_executor.py

@@ -25,6 +25,7 @@
 from promptflow.contracts.flow import FlexFlow, Flow
 from promptflow.contracts.tool import ConnectionType
 from promptflow.core import log_metric
+from promptflow.core._connection_provider._dict_connection_provider import DictConnectionProvider
 from promptflow.core._model_configuration import (
     MODEL_CONFIG_NAME_2_CLASS,
     AzureOpenAIModelConfiguration,
@@ -48,7 +49,7 @@ class ScriptExecutor(FlowExecutor):
     def __init__(
         self,
         flow_file: Union[Path, str, Callable],
-        connections: Optional[dict] = None,
+        connections: Optional[Union[dict, ConnectionProvider]] = None,
         working_dir: Optional[Path] = None,
         *,
         storage: Optional[AbstractRunStorage] = None,
@@ -58,6 +59,9 @@ def __init__(
         logger.debug(f"Init params for script executor: {init_kwargs}")
 
         self._flow_file = flow_file
+        if connections and isinstance(connections, dict):
+            connections = DictConnectionProvider(connections)
+        self._connections = connections
         entry = flow_file  # Entry could be both a path or a callable
         self._entry = entry
         self._init_kwargs = init_kwargs or {}
@@ -67,7 +71,6 @@ def __init__(
             self._working_dir = working_dir or Path.cwd()
         self._init_input_sign()
         self._initialize_function()
-        self._connections = connections
         self._storage = storage or DefaultRunStorage()
         self._flow_id = "default_flow_id"
         self._log_interval = 60
@@ -353,9 +356,8 @@ def _resolve_init_kwargs(self, c: type, init_kwargs: dict):
 
         return resolved_init_kwargs
 
-    @classmethod
-    def _resolve_connection_params(cls, connection_params: list, init_kwargs: dict, resolved_init_kwargs: dict):
-        provider = ConnectionProvider.get_instance()
+    def _resolve_connection_params(self, connection_params: list, init_kwargs: dict, resolved_init_kwargs: dict):
+        provider = self._connections or ConnectionProvider.get_instance()
         # parse connection
         logger.debug(f"Resolving connection params: {connection_params}")
         for key in connection_params:

src/promptflow-core/promptflow/executor/flow_executor.py

@@ -205,14 +205,18 @@ def create(
         if hasattr(flow_file, "__call__") or inspect.isfunction(flow_file):
             from ._script_executor import ScriptExecutor
 
-            return ScriptExecutor(flow_file, storage=storage)
+            return ScriptExecutor(flow_file, connections=connections, storage=storage)
         if not isinstance(flow_file, (Path, str)):
             raise NotImplementedError("Only support Path or str for flow_file.")
         if is_flex_flow(flow_path=flow_file, working_dir=working_dir):
             from ._script_executor import ScriptExecutor
 
             return ScriptExecutor(
-                flow_file=Path(flow_file), working_dir=working_dir, storage=storage, init_kwargs=init_kwargs
+                flow_file=Path(flow_file),
+                connections=connections,
+                working_dir=working_dir,
+                storage=storage,
+                init_kwargs=init_kwargs,
             )
         elif is_prompty_flow(file_path=flow_file):
             from ._prompty_executor import PromptyExecutor

src/promptflow-core/tests/conftest.py

@@ -156,6 +156,12 @@ def setup_connection_provider():
         yield
 
 
+@pytest.fixture
+def dev_connections() -> dict:
+    with open(CONNECTION_FILE, "r") as f:
+        return json.load(f)
+
+
 # ==================== serving fixtures ====================
 
 

src/promptflow-core/tests/core/e2etests/test_eager_flow.py

@@ -6,6 +6,7 @@
 from promptflow._core.tool_meta_generator import PythonLoadError
 from promptflow.contracts.run_info import Status
 from promptflow.core import AzureOpenAIModelConfiguration, OpenAIModelConfiguration
+from promptflow.core._connection_provider._dict_connection_provider import DictConnectionProvider
 from promptflow.executor._errors import (
     FlowEntryInitializationError,
     InputNotFound,
@@ -44,7 +45,7 @@ async def func_entry_async(input_str: str) -> str:
 ]
 
 
-@pytest.mark.usefixtures("recording_injection", "setup_connection_provider")
+@pytest.mark.usefixtures("recording_injection", "setup_connection_provider", "dev_connections")
 @pytest.mark.e2etest
 class TestEagerFlow:
     @pytest.mark.parametrize(
@@ -138,6 +139,28 @@ def test_flow_run_with_openai_chat(self):
         token_names = ["prompt_tokens", "completion_tokens", "total_tokens"]
         for token_name in token_names:
             assert token_name in line_result.run_info.api_calls[0]["children"][0]["system_metrics"]
+            assert line_result.run_info.api_calls[0]["children"][0]["system_metrics"][token_name] > 0
+
+    def test_flow_run_with_connection(self, dev_connections):
+        flow_file = get_yaml_file(
+            "dummy_callable_class_with_connection", root=EAGER_FLOW_ROOT, file_name="flow.flex.yaml"
+        )
+
+        # Test submitting eager flow to script executor with connection dictionary
+        executor = ScriptExecutor(
+            flow_file=flow_file, connections=dev_connections, init_kwargs={"connection": "azure_open_ai_connection"}
+        )
+        line_result = executor.exec_line(inputs={}, index=0)
+        assert line_result.run_info.status == Status.Completed, line_result.run_info.error
+
+        # Test submitting eager flow to script executor with connection provider
+        executor = ScriptExecutor(
+            flow_file=flow_file,
+            connections=DictConnectionProvider(dev_connections),
+            init_kwargs={"connection": "azure_open_ai_connection"},
+        )
+        line_result = executor.exec_line(inputs={}, index=0)
+        assert line_result.run_info.status == Status.Completed, line_result.run_info.error
 
     @pytest.mark.parametrize("entry, inputs, expected_output", function_entries)
     def test_flow_run_with_function_entry(self, entry, inputs, expected_output):

src/promptflow/tests/test_configs/eager_flows/dummy_callable_class_with_connection/flow.flex.yaml

@@ -0,0 +1 @@
+entry: simple_callable_with_connection:MyFlow

src/promptflow/tests/test_configs/eager_flows/dummy_callable_class_with_connection/simple_callable_with_connection.py

@@ -0,0 +1,10 @@
+from promptflow.connections import AzureOpenAIConnection
+
+
+class MyFlow:
+    def __init__(self, connection: AzureOpenAIConnection):
+        self._connection = connection
+
+    def __call__(self):
+        assert isinstance(self._connection, AzureOpenAIConnection)
+        return "Dummy output"