tests/test_mapping.py

import os
import pytest
from unittest.mock import MagicMock, patch


@pytest.fixture
def csv_file(tmp_path):
    p = tmp_path / "index.csv"
    p.write_text(
        "path,token,phrase,word\n"
        "video_dir/hello.mp4,HELLO,,\n"
        "video_dir/thank_you.mp4,THANK-YOU,,\n"
        "video_dir/good_morning.mp4,,GOOD MORNING,\n"
    )
    return str(p)


@pytest.fixture
def dict_dir(tmp_path):
    d = tmp_path / "dictionary"
    d.mkdir()
    (d / "hello.mp4").write_bytes(b"")
    (d / "thank_you.mp4").write_bytes(b"")
    (d / "good_morning.mp4").write_bytes(b"")
    return str(d)


def _mock_vi(result=None):
    vi = MagicMock()
    vi.query_gloss.return_value = result or []
    return vi


def test_ensure_extracted_skips_when_dir_exists(tmp_path):
    """Does nothing if dictionary directory already exists."""
    (tmp_path / "dictionary").mkdir()
    from mapping import ensure_dictionary_extracted
    with patch("mapping.zipfile.ZipFile") as mock_zip:
        ensure_dictionary_extracted(
            zip_path=str(tmp_path / "dict.zip"),
            extract_to=str(tmp_path),
        )
    mock_zip.assert_not_called()


def test_ensure_extracted_unzips_when_missing(tmp_path):
    """Extracts zip when dictionary directory is absent."""
    zip_path = str(tmp_path / "dict.zip")
    from mapping import ensure_dictionary_extracted
    with patch("mapping.zipfile.ZipFile") as mock_zip_cls:
        ensure_dictionary_extracted(zip_path=zip_path, extract_to=str(tmp_path))
    mock_zip_cls.assert_called_once_with(zip_path)
    mock_zip_cls.return_value.__enter__.return_value.extractall.assert_called_once_with(
        str(tmp_path)
    )


def test_exact_single_token_match(csv_file, dict_dir):
    """get_paths returns path on exact single-token match."""
    with patch("mapping.ensure_dictionary_extracted"):
        from mapping import ASLDictionary
        d = ASLDictionary(csv_path=csv_file, dictionary_dir=dict_dir, vector_index=_mock_vi())
    paths = d.get_paths(["hello"])
    assert len(paths) == 1
    assert paths[0].endswith("hello.mp4")


def test_greedy_phrase_match(csv_file, dict_dir):
    """get_paths greedily matches multi-word phrase tokens."""
    with patch("mapping.ensure_dictionary_extracted"):
        from mapping import ASLDictionary
        d = ASLDictionary(csv_path=csv_file, dictionary_dir=dict_dir, vector_index=_mock_vi())
    paths = d.get_paths(["good", "morning"])
    assert len(paths) == 1
    assert paths[0].endswith("good_morning.mp4")


def test_semantic_fallback_on_exact_miss(csv_file, dict_dir):
    """get_paths calls VectorIndex when exact match fails."""
    vi = _mock_vi(result=[("HELLO", 0.85, "hello.mp4")])
    with patch("mapping.ensure_dictionary_extracted"):
        from mapping import ASLDictionary
        d = ASLDictionary(csv_path=csv_file, dictionary_dir=dict_dir, vector_index=vi)
    paths = d.get_paths(["HI"])
    assert len(paths) == 1
    assert paths[0].endswith("hello.mp4")
    vi.query_gloss.assert_called_once_with("HI", top_k=1)


def test_fuzzy_fallback_on_semantic_miss(csv_file, dict_dir):
    """get_paths uses difflib when semantic search returns nothing."""
    vi = _mock_vi(result=[])
    with patch("mapping.ensure_dictionary_extracted"):
        from mapping import ASLDictionary
        d = ASLDictionary(csv_path=csv_file, dictionary_dir=dict_dir, vector_index=vi)
    # "HELO" scores ~0.89 vs "HELLO" with difflib — above default cutoff 0.6
    paths = d.get_paths(["HELO"])
    assert len(paths) == 1
    assert paths[0].endswith("hello.mp4")


def test_no_match_returns_empty(csv_file, dict_dir):
    """get_paths skips token when all three tiers fail."""
    vi = _mock_vi(result=[])
    with patch("mapping.ensure_dictionary_extracted"), \
         patch.dict(os.environ, {"FUZZY_CUTOFF": "0.99"}):
        from mapping import ASLDictionary
        d = ASLDictionary(csv_path=csv_file, dictionary_dir=dict_dir, vector_index=vi)
        paths = d.get_paths(["XYZZY"])
    assert paths == []