tests/test_whisper_service.py

from __future__ import annotations

from types import SimpleNamespace

from app.services.whisper_service import WhisperService


class FakeWhisperModel:
    def transcribe(self, path, **kwargs):
        segments = iter(
            [
                SimpleNamespace(id=0, start=0.0, end=1.25, text=" Hello"),
                SimpleNamespace(id=1, start=1.25, end=2.0, text=" world"),
            ]
        )
        info = SimpleNamespace(language="en", language_probability=0.99, duration=2.0)
        return segments, info


async def test_whisper_writes_srt_without_loading_real_model(settings, tmp_path) -> None:
    service = WhisperService(settings)
    service._models["tiny"] = FakeWhisperModel()
    media = tmp_path / "audio.wav"
    media.write_bytes(b"test")
    result = await service.transcribe(
        media, tmp_path / "out", model_name="tiny", output_format="srt"
    )
    assert result.path is not None
    content = result.path.read_text()
    assert "00:00:00,000 --> 00:00:01,250" in content
    assert "Hello" in content
    assert result.metadata["language"] == "en"