test(suite): update all tests for streaming/locking architecture and mock parity

tests/test_headless_verification.py

@@ -89,6 +89,7 @@ async def test_headless_verification_error_and_qa_interceptor(vlogger) -> None:
   mock_resp1.candidates = [MagicMock(content=MagicMock(parts=[mock_part1]), finish_reason=MagicMock(name="STOP"))]
   mock_resp1.usage_metadata.prompt_token_count = 10
   mock_resp1.usage_metadata.candidates_token_count = 5
+  mock_resp1.text = mock_part1.text
   # 2nd round: Final text after tool result
   mock_part2 = MagicMock()
   mock_part2.text = "The command failed but I understand why. Task done."
@@ -97,16 +98,22 @@ async def test_headless_verification_error_and_qa_interceptor(vlogger) -> None:
   mock_resp2.candidates = [MagicMock(content=MagicMock(parts=[mock_part2]), finish_reason=MagicMock(name="STOP"))]
   mock_resp2.usage_metadata.prompt_token_count = 20
   mock_resp2.usage_metadata.candidates_token_count = 10
+  mock_resp2.text = mock_part2.text
   mock_chat.send_message.side_effect = [mock_resp1, mock_resp2]
+  # Handle streaming calls
+  def make_stream_mock(resp):
+   m = MagicMock()
+   m.__iter__.return_value = [resp]
+   m.candidates = resp.candidates
+   m.usage_metadata = resp.usage_metadata
+   return m
+  mock_chat.send_message_stream.side_effect = [make_stream_mock(mock_resp1), make_stream_mock(mock_resp2)]
   # Mock run_powershell behavior: it should call the qa_callback on error
 
   def run_side_effect(script: Any, base_dir: Any, qa_callback: Any) -> Any:
    if qa_callback:
     analysis = qa_callback("Error: file not found")
-    return f"""STDERR: Error: file not found
-
-QA ANALYSIS:
-{analysis}"""
+    return f"STDERR: Error: file not found\n\nQA ANALYSIS:\n{analysis}"
    return "Error: file not found"
   mock_run.side_effect = run_side_effect
   mock_qa.return_value = "FIX: Check if path exists."
@@ -123,8 +130,11 @@ QA ANALYSIS:
   mock_qa.assert_called_once_with("Error: file not found")
   # Verify the 2nd send_message call includes the QA ANALYSIS in its payload (f_resps)
   # The first call is the user message, the second is the tool response.
-  assert mock_chat.send_message.call_count == 2
-  args, kwargs = mock_chat.send_message.call_args_list[1]
+  assert (mock_chat.send_message.call_count + mock_chat.send_message_stream.call_count) == 2
+  
+  # Get the second call's payload (either from send_message or send_message_stream)
+  calls = mock_chat.send_message.call_args_list + mock_chat.send_message_stream.call_args_list
+  args, kwargs = calls[1]
   f_resps = args[0]
   
   found_qa = False