tests/test_app.py

import pytest
import torch

# --- Fast tests (no model loading required) ---


def test_app_imports_without_model_load():
    """Importing app should not trigger model download."""
    import app

    assert hasattr(app, "translate")
    assert hasattr(app, "_load_tokenizer")
    assert hasattr(app, "_load_model")


# --- Slow tests (require GPU + model download) ---

gpu_available = torch.backends.mps.is_available() or torch.cuda.is_available()


@pytest.fixture(scope="module")
def loaded_app():
    import app

    # Force model/tokenizer loading
    app._load_tokenizer()
    app._load_model()
    return app


@pytest.mark.slow
@pytest.mark.skipif(not gpu_available, reason="Requires MPS or CUDA")
def test_name_to_code_matches_language_names(loaded_app):
    name_to_code, language_names = loaded_app._build_language_mappings()
    assert set(name_to_code.keys()) == set(language_names)


@pytest.mark.slow
@pytest.mark.skipif(not gpu_available, reason="Requires MPS or CUDA")
def test_language_names_sorted(loaded_app):
    _, language_names = loaded_app._build_language_mappings()
    assert language_names == sorted(language_names)


@pytest.mark.slow
@pytest.mark.skipif(not gpu_available, reason="Requires MPS or CUDA")
def test_all_codes_are_bcp47_tokens(loaded_app):
    name_to_code, _ = loaded_app._build_language_mappings()
    for name, code in name_to_code.items():
        assert code.startswith("<2") and code.endswith(">"), f"Invalid code {code} for {name}"


@pytest.mark.slow
@pytest.mark.skipif(not gpu_available, reason="Requires MPS or CUDA")
def test_translate_unsupported_language(loaded_app):
    with pytest.raises(ValueError, match="Unsupported language"):
        loaded_app.translate("hello", "FakeLanguage")


@pytest.mark.slow
@pytest.mark.skipif(not gpu_available, reason="Requires MPS or CUDA")
def test_translate_returns_string(loaded_app):
    result = loaded_app.translate("Hello", "French")
    assert isinstance(result, str)
    assert len(result) > 0