tests/test_aztext.py

"""Tests for aztext. Runs under pytest OR as a plain script (python test_aztext.py).

    python -m pytest aztext/tests -q          # if pytest is available
    python aztext/tests/test_aztext.py        # dependency-free fallback
"""

import os
import sys

sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))

import aztext
from aztext import alphabet
import aztext.deasciify            # ensure submodule is imported (name is shadowed by the func export)
deasc = sys.modules["aztext.deasciify"]   # the MODULE, not the re-exported function


# --- Turkic i-family case mapping (the thing Python gets wrong) --------------
def test_az_upper_i_family():
    assert alphabet.az_upper("i") == "İ"
    assert alphabet.az_upper("ı") == "I"
    assert alphabet.az_upper("işıq") == "İŞIQ"


def test_az_lower_i_family():
    assert alphabet.az_lower("İ") == "i"
    assert alphabet.az_lower("I") == "ı"
    assert alphabet.az_lower("İŞIQ") == "işıq"


def test_i_distinction_never_collapses():
    # the four characters must stay pairwise distinct through case round-trips
    assert alphabet.az_lower(alphabet.az_upper("i")) == "i"
    assert alphabet.az_lower(alphabet.az_upper("ı")) == "ı"


# --- number to words --------------------------------------------------------
def test_num_to_words():
    cases = {
        0: "sıfır", 1: "bir", 7: "yeddi", 10: "on", 11: "on bir",
        20: "iyirmi", 25: "iyirmi beş", 100: "yüz", 101: "yüz bir",
        200: "iki yüz", 1000: "min", 2000: "iki min",
        1234: "min iki yüz otuz dörd", 1_000_000: "bir milyon",
    }
    for n, expected in cases.items():
        assert aztext.num_to_words(n) == expected, (n, aztext.num_to_words(n))


def test_num_to_words_negative():
    assert aztext.num_to_words(-5) == "mənfi beş"


# --- language id ------------------------------------------------------------
def test_langid_az_vs_tr():
    assert aztext.is_azerbaijani("Mən Azərbaycan dilini öyrənirəm.")
    # Turkish sentence (no ə, Turkish function words) must NOT be Azerbaijani
    lang, _ = aztext.detect_language("Ben Türkçe öğreniyorum ve kitap okuyorum.")
    assert lang == "tr", lang


def test_langid_rejects_cyrillic():
    lang, _ = aztext.detect_language("Привет как дела")
    assert lang == "other", lang


# --- normalize --------------------------------------------------------------
def test_normalize_idempotent_and_preserves_az():
    s = "Gözəl   şəhər —  “Bakı”."
    once = aztext.normalize(s)
    assert aztext.normalize(once) == once           # idempotent
    for ch in "əöşıü":
        pass
    assert "ə" in once and "ş" in once and "ı" in once   # az letters preserved
    assert '"' in once and "'" not in once               # curly quotes -> ascii


# --- deasciify --------------------------------------------------------------
def test_deasciify_known_words():
    assert deasc.deasciify("ucun") == "üçün"
    assert deasc.deasciify("cixis") == "çıxış"
    assert deasc.deasciify("mekteb") == "məktəb"
    assert deasc.deasciify("usaq") == "uşaq"
    # casing preserved
    assert deasc.deasciify("Mekteb") == "Məktəb"
    # already-diacritic words pass through untouched
    assert deasc.deasciify("gözəl") == "gözəl"
    # unknown words untouched
    assert deasc.deasciify("xyzzy") == "xyzzy"


def test_ascii_fold():
    assert deasc.ascii_fold("gözəl çıxış") == "gozel cixis"
    assert deasc.ascii_fold("işıq") == "isiq"


# --- tokenize / script ------------------------------------------------------
def test_tokenize():
    assert aztext.word_tokenize("Bakı, Azərbaycanın paytaxtıdır.") == \
        ["Bakı", "Azərbaycanın", "paytaxtıdır"]
    assert len(aztext.sent_tokenize("Bir. İki! Üç?")) == 3


def test_is_latin_azerbaijani():
    assert aztext.is_latin_azerbaijani("Azərbaycan dili")
    assert not aztext.is_latin_azerbaijani("Азәрбајҹан дили")


# --- dictionary integrity: no homoglyph / non-Azerbaijani letters -----------
def test_dictionary_is_clean_azerbaijani():
    bad = []
    for w in deasc._CORRECT_WORDS:
        for ch in w:
            if ch.isalpha() and ch not in alphabet.AZ_LETTERS:
                bad.append((w, ch, hex(ord(ch))))
    assert not bad, f"non-Azerbaijani letters in deasciify dict: {bad}"


def _run_all():
    fns = [v for k, v in sorted(globals().items()) if k.startswith("test_") and callable(v)]
    failed = 0
    for fn in fns:
        try:
            fn()
            print(f"PASS {fn.__name__}")
        except AssertionError as e:
            failed += 1
            print(f"FAIL {fn.__name__}: {e}")
        except Exception as e:  # noqa: BLE001
            failed += 1
            print(f"ERROR {fn.__name__}: {type(e).__name__}: {e}")
    print(f"\n{len(fns) - failed}/{len(fns)} passed")
    return failed


if __name__ == "__main__":
    sys.exit(1 if _run_all() else 0)