tests/test_langmap.py

from langmap.langid_mapping import langid_to_language


def test_langid_mapping_is_nonempty():
    assert len(langid_to_language) > 0


def test_langid_mapping_count():
    assert len(langid_to_language) == 22, f"Expected 22 Tier 1 languages, got {len(langid_to_language)}"


def test_keys_are_bcp47_tokens():
    for key in langid_to_language:
        assert key.startswith("<2"), f"Key {key} does not start with '<2'"
        assert key.endswith(">"), f"Key {key} does not end with '>'"


def test_values_are_nonempty_strings():
    for key, value in langid_to_language.items():
        assert isinstance(value, str) and value.strip(), f"Empty language name for {key}"


def test_no_leading_or_trailing_whitespace():
    for key, value in langid_to_language.items():
        assert value == value.strip(), f"Language name for {key} has leading/trailing whitespace: {value!r}"


def test_no_duplicate_language_names():
    names = list(langid_to_language.values())
    assert len(names) == len(set(names)), "Duplicate language names found"


def test_all_tier1_languages_present():
    """Verify all 22 Tier 1 languages are present."""
    expected = {
        "Afrikaans",
        "Bulgarian",
        "Bosnian",
        "Catalan",
        "Czech",
        "Danish",
        "German",
        "Spanish",
        "Filipino",
        "French",
        "Croatian",
        "Indonesian",
        "Italian",
        "Macedonian",
        "Malay",
        "Maltese",
        "Norwegian Nynorsk",
        "Norwegian",
        "Portuguese",
        "Romanian",
        "Slovak",
        "Swedish",
    }
    values = set(langid_to_language.values())
    assert values == expected, f"Mismatch: missing={expected - values}, extra={values - expected}"


def test_removed_languages_absent():
    """Languages outside Tier 1 must not be in the mapping."""
    values = set(langid_to_language.values())
    for lang in ["Japanese", "Hindi", "Swahili", "Arabic", "Russian", "Polish", "Dutch"]:
        assert lang not in values, f"{lang} should have been removed (not Tier 1)"