tests/test_api.py

"""
Integration tests for the FastAPI endpoints.
Uses httpx.AsyncClient with a mocked transcriber so GPU hardware is not required.
"""
from __future__ import annotations

import io
from unittest.mock import AsyncMock, MagicMock, patch

import pytest
import pytest_asyncio


def _make_mock_app():
    """Create a test FastAPI app with mocked model state."""
    from fastapi import FastAPI
    from fastapi.testclient import TestClient

    from src.api.routes import health, iot, transcribe
    from src.engine.transcriber import TranscriptionResult

    app = FastAPI()
    app.include_router(health.router, prefix="/api/v1")
    app.include_router(transcribe.router, prefix="/api/v1")
    app.include_router(iot.router, prefix="/api/v1")

    # Mock adapter manager
    mock_adapter_manager = MagicMock()
    mock_adapter_manager.get_active.return_value = "bam"
    mock_adapter_manager.list_available.return_value = ["bam", "ful"]
    mock_adapter_manager.list_loaded.return_value = ["bam"]

    # Mock transcriber
    mock_transcriber = MagicMock()
    mock_transcriber.transcribe_file.return_value = TranscriptionResult(
        text="bunding nɔgɔ foro",
        language="bam",
        duration_s=3.0,
        processing_time_ms=250,
    )

    # Mock sensor bridge
    mock_sensor_bridge = MagicMock()
    from src.iot.sensor_bridge import SensorData
    from datetime import datetime
    mock_sensor_bridge.fetch = AsyncMock(
        return_value=SensorData(
            sensor_type="soil",
            values={"moisture_pct": 42.0, "ph": 6.5},
            timestamp=datetime.utcnow().isoformat(),
        )
    )

    app.state.adapter_manager = mock_adapter_manager
    app.state.transcriber = mock_transcriber
    app.state.sensor_bridge = mock_sensor_bridge

    return app, TestClient(app)


class TestHealthEndpoint:
    def setup_method(self):
        self.app, self.client = _make_mock_app()

    def test_health_returns_200(self):
        response = self.client.get("/api/v1/health")
        assert response.status_code == 200

    def test_health_response_structure(self):
        data = self.client.get("/api/v1/health").json()
        assert "status" in data
        assert "model_loaded" in data
        assert "adapters_available" in data

    def test_health_model_loaded_true(self):
        data = self.client.get("/api/v1/health").json()
        assert data["model_loaded"] is True

    def test_health_active_adapter(self):
        data = self.client.get("/api/v1/health").json()
        assert data["active_adapter"] == "bam"


class TestTranscribeEndpoint:
    def setup_method(self):
        self.app, self.client = _make_mock_app()

    def _wav_bytes(self) -> bytes:
        """Minimal valid WAV file bytes for testing."""
        import wave
        import struct
        buf = io.BytesIO()
        with wave.open(buf, "wb") as wf:
            wf.setnchannels(1)
            wf.setsampwidth(2)
            wf.setframerate(16000)
            wf.writeframes(struct.pack("<" + "h" * 160, *([0] * 160)))
        return buf.getvalue()

    def test_transcribe_returns_200(self):
        response = self.client.post(
            "/api/v1/transcribe",
            data={"language": "bam"},
            files={"audio_file": ("test.wav", self._wav_bytes(), "audio/wav")},
        )
        assert response.status_code == 200

    def test_transcribe_response_has_text(self):
        data = self.client.post(
            "/api/v1/transcribe",
            data={"language": "bam"},
            files={"audio_file": ("test.wav", self._wav_bytes(), "audio/wav")},
        ).json()
        assert "text" in data
        assert isinstance(data["text"], str)

    def test_invalid_language_returns_422(self):
        response = self.client.post(
            "/api/v1/transcribe",
            data={"language": "xyz"},
            files={"audio_file": ("test.wav", self._wav_bytes(), "audio/wav")},
        )
        assert response.status_code == 422

    def test_unsupported_file_type_returns_422(self):
        response = self.client.post(
            "/api/v1/transcribe",
            data={"language": "bam"},
            files={"audio_file": ("test.txt", b"not audio", "text/plain")},
        )
        assert response.status_code == 422


class TestIoTQueryEndpoint:
    def setup_method(self):
        self.app, self.client = _make_mock_app()

    def _wav_bytes(self) -> bytes:
        import io
        import struct
        import wave
        buf = io.BytesIO()
        with wave.open(buf, "wb") as wf:
            wf.setnchannels(1)
            wf.setsampwidth(2)
            wf.setframerate(16000)
            wf.writeframes(struct.pack("<" + "h" * 160, *([0] * 160)))
        return buf.getvalue()

    def test_query_returns_200(self):
        response = self.client.post(
            "/api/v1/query",
            data={"language": "bam", "field_id": "field_001"},
            files={"audio_file": ("test.wav", self._wav_bytes(), "audio/wav")},
        )
        assert response.status_code == 200

    def test_query_response_structure(self):
        data = self.client.post(
            "/api/v1/query",
            data={"language": "bam"},
            files={"audio_file": ("test.wav", self._wav_bytes(), "audio/wav")},
        ).json()
        assert "transcription" in data
        assert "intent" in data
        assert "sensor_data" in data
        assert "voice_response" in data

    def test_query_voice_response_is_french(self):
        data = self.client.post(
            "/api/v1/query",
            data={"language": "bam"},
            files={"audio_file": ("test.wav", self._wav_bytes(), "audio/wav")},
        ).json()
        # French response should contain at least one French word
        response_text = data["voice_response"]
        french_indicators = ["du", "de", "le", "la", "les", "et", "Humidité", "sol", "pH"]
        assert any(word in response_text for word in french_indicators)