chore: Fixes test_fast_api.py (part I for circular deps).

It still fails due to signal used not in main thread. It will be fixed later.

PiperOrigin-RevId: 760050504

src/google/adk/memory/vertex_ai_rag_memory_service.py

@@ -12,23 +12,29 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+
+from __future__ import annotations
+
 from collections import OrderedDict
 import json
 import os
 import tempfile
 from typing import Optional
+from typing import TYPE_CHECKING
 
 from google.genai import types
 from typing_extensions import override
 from vertexai.preview import rag
 
-from ..events.event import Event
-from ..sessions.session import Session
 from . import _utils
 from .base_memory_service import BaseMemoryService
 from .base_memory_service import SearchMemoryResponse
 from .memory_entry import MemoryEntry
 
+if TYPE_CHECKING:
+  from ..events.event import Event
+  from ..sessions.session import Session
+
 
 class VertexAiRagMemoryService(BaseMemoryService):
   """A memory service that uses Vertex AI RAG for storage and retrieval."""
@@ -103,6 +109,8 @@ class VertexAiRagMemoryService(BaseMemoryService):
       self, *, app_name: str, user_id: str, query: str
   ) -> SearchMemoryResponse:
     """Searches for sessions that match the query using rag.retrieval_query."""
+    from ..events.event import Event
+
     response = rag.retrieval_query(
         text=query,
         rag_resources=self._vertex_rag_store.rag_resources,

tests/unittests/fast_api/test_fast_api.py

@@ -12,6 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+from __future__ import annotations
+
 import asyncio
 import json
 import sys
@@ -19,14 +21,14 @@ import threading
 import time
 import types as ptypes
 from typing import AsyncGenerator
+from typing import TYPE_CHECKING
 
-from google.adk.agents import BaseAgent
-from google.adk.agents import LiveRequest
+from google.adk.agents.base_agent import BaseAgent
+from google.adk.agents.live_request_queue import LiveRequest
 from google.adk.agents.run_config import RunConfig
 from google.adk.cli.fast_api import AgentRunRequest
 from google.adk.cli.fast_api import get_fast_api_app
 from google.adk.cli.utils import envs
-from google.adk.events import Event
 from google.adk.runners import Runner
 from google.genai import types
 import httpx
@@ -34,6 +36,9 @@ import pytest
 from uvicorn.main import run as uvicorn_run
 import websockets
 
+if TYPE_CHECKING:
+  from google.adk.events import Event
+
 
 # Here we “fake” the agent module that get_fast_api_app expects.
 # The server code does: `agent_module = importlib.import_module(agent_name)`
@@ -49,33 +54,45 @@ dummy_module.agent = ptypes.SimpleNamespace(
 sys.modules["test_app"] = dummy_module
 envs.load_dotenv_for_agent("test_app", ".")
 
-event1 = Event(
-    author="dummy agent",
-    invocation_id="invocation_id",
-    content=types.Content(
-        role="model", parts=[types.Part(text="LLM reply", inline_data=None)]
-    ),
-)
 
-event2 = Event(
-    author="dummy agent",
-    invocation_id="invocation_id",
-    content=types.Content(
-        role="model",
-        parts=[
-            types.Part(
-                text=None,
-                inline_data=types.Blob(
-                    mime_type="audio/pcm;rate=24000", data=b"\x00\xFF"
-                ),
-            )
-        ],
-    ),
-)
+def _event_1():
+  from google.adk.events import Event
 
-event3 = Event(
-    author="dummy agent", invocation_id="invocation_id", interrupted=True
-)
+  return Event(
+      author="dummy agent",
+      invocation_id="invocation_id",
+      content=types.Content(
+          role="model", parts=[types.Part(text="LLM reply", inline_data=None)]
+      ),
+  )
+
+
+def _event_2():
+  from google.adk.events import Event
+
+  return Event(
+      author="dummy agent",
+      invocation_id="invocation_id",
+      content=types.Content(
+          role="model",
+          parts=[
+              types.Part(
+                  text=None,
+                  inline_data=types.Blob(
+                      mime_type="audio/pcm;rate=24000", data=b"\x00\xFF"
+                  ),
+              )
+          ],
+      ),
+  )
+
+
+def _event_3():
+  from google.adk.events import Event
+
+  return Event(
+      author="dummy agent", invocation_id="invocation_id", interrupted=True
+  )
 
 
 # For simplicity, we patch Runner.run_live to yield dummy events.
@@ -84,13 +101,13 @@ async def dummy_run_live(
     self, session, live_request_queue
 ) -> AsyncGenerator[Event, None]:
   # Immediately yield a dummy event with a text reply.
-  yield event1
+  yield _event_1()
   await asyncio.sleep(0)
 
-  yield event2
+  yield _event_2()
   await asyncio.sleep(0)
 
-  yield event3
+  yield _event_3()
 
   raise Exception()
 
@@ -103,13 +120,13 @@ async def dummy_run_async(
     run_config: RunConfig = RunConfig(),
 ) -> AsyncGenerator[Event, None]:
   # Immediately yield a dummy event with a text reply.
-  yield event1
+  yield _event_1()
   await asyncio.sleep(0)
 
-  yield event2
+  yield _event_2()
   await asyncio.sleep(0)
 
-  yield event3
+  yield _event_3()
 
   return
 
@@ -199,15 +216,15 @@ async def test_sse_endpoint():
           if event_data:
             event_count += 1
             if event_count == 1:
-              assert event_data == event1.model_dump_json(
+              assert event_data == _event_1().model_dump_json(
                   exclude_none=True, by_alias=True
               )
             elif event_count == 2:
-              assert event_data == event2.model_dump_json(
+              assert event_data == _event_2().model_dump_json(
                   exclude_none=True, by_alias=True
               )
             elif event_count == 3:
-              assert event_data == event3.model_dump_json(
+              assert event_data == _event_3().model_dump_json(
                   exclude_none=True, by_alias=True
               )
             else: