import pytest import httpx import wave import io @pytest.mark.asyncio async def test_root_endpoint(http_client): """ Tests if the root endpoint is alive and returns the correct status message. """ print("\n--- Running test_root_endpoint ---") response = await http_client.get("/") assert response.status_code == 200 assert response.json() == {"status": "AI Model Hub is running!"} print("✅ Root endpoint test passed.") @pytest.mark.asyncio async def test_create_speech_stream(http_client): """ Tests the /speech endpoint for a successful audio stream response. """ print("\n--- Running test_create_speech_stream ---") url = "/speech" payload = {"text": "Hello, world!"} # The `stream=True` parameter tells httpx to not read the entire response body # at once. We'll handle it manually to check for content. async with http_client.stream("POST", url, json=payload) as response: assert response.status_code == 200, f"Speech stream request failed. Response: {response.text}" assert response.headers.get("content-type") == "audio/wav" # Check that the response body is not empty by iterating over chunks. content_length = 0 async for chunk in response.aiter_bytes(): content_length += len(chunk) assert content_length > 0 print("✅ TTS stream test passed.") @pytest.mark.asyncio async def test_stt_transcribe_endpoint(http_client): """ Tests the /stt/transcribe endpoint by uploading a dummy audio file and verifying the transcription response. """ print("\n--- Running test_stt_transcribe_endpoint ---") url = "/stt/transcribe" # --- Use a real audio file from the integration test data --- audio_file_path = "integration_tests/test_data/test-audio.wav" with open(audio_file_path, "rb") as audio_file: files = {'audio_file': ('test-audio.wav', audio_file, 'audio/wav')} # --- Send the POST request to the endpoint --- response = await http_client.post(url, files=files) # --- Assertions --- assert response.status_code == 200, f"STT request failed with status code {response.status_code}. Response: {response.text}" response_json = response.json() assert "transcript" in response_json, "Response JSON is missing the 'transcript' key." assert isinstance(response_json["transcript"], str), "Transcript value is not a string." # Assert that the transcript matches the expected text expected_transcript = "This audio is for integration testing of Cortex Hub, which is a wonderful project." assert response_json["transcript"] == expected_transcript, f"Expected: '{expected_transcript}', Got: '{response_json['transcript']}'" print("✅ STT transcription test passed.")