import re
import pytest
from app.service import _normalize_text, _int_to_gujarati, _split_into_chunks
from app.schemas import GenerateRequest
from pydantic import ValidationError


# ── _int_to_gujarati ──────────────────────────────────────────────────────────

def test_zero():          assert _int_to_gujarati(0)       == 'શૂન્ય'
def test_one():           assert _int_to_gujarati(1)       == 'એક'
def test_nine():          assert _int_to_gujarati(9)       == 'નવ'
def test_ten():           assert _int_to_gujarati(10)      == 'દસ'
def test_eleven():        assert _int_to_gujarati(11)      == 'અગિયાર'
def test_nineteen():      assert _int_to_gujarati(19)      == 'ઓગણીસ'
def test_twenty_one():    assert _int_to_gujarati(21)      == 'એકવીસ'
def test_fifty():         assert _int_to_gujarati(50)      == 'પચાસ'
def test_ninety_nine():   assert _int_to_gujarati(99)      == 'નવ્વાણું'
def test_hundred():       assert _int_to_gujarati(100)     == 'સો'
def test_one_fifty():     assert _int_to_gujarati(150)     == 'સો પચાસ'
def test_two_hundred():   assert _int_to_gujarati(200)     == 'બસ્સો'
def test_two_fifty():     assert _int_to_gujarati(250)     == 'બસ્સો પચાસ'
def test_one_thousand():  assert _int_to_gujarati(1000)    == 'એક હજાર'
def test_one_two_fifty(): assert _int_to_gujarati(1250)    == 'એક હજાર બસ્સો પચાસ'
def test_two_thousand():  assert _int_to_gujarati(2000)    == 'બે હજાર'
def test_fifty_thousand():assert _int_to_gujarati(50000)   == 'પચાસ હજાર'
def test_one_lakh():      assert _int_to_gujarati(100000)  == 'એક લાખ'
def test_one_crore():     assert _int_to_gujarati(10000000)== 'એક કરોડ'
def test_year():          assert _int_to_gujarati(2024)    == 'બે હજાર ચોવીસ'


# ── _normalize_text — basic ───────────────────────────────────────────────────

def test_gu_single():     assert _normalize_text('૧')      == 'એક'
def test_gu_comma():      assert _normalize_text('૧,૨૫૦')  == 'એક હજાર બસ્સો પચાસ'
def test_ar_comma():      assert _normalize_text('1,250')   == 'એક હજાર બસ્સો પચાસ'
def test_rupee():         assert _normalize_text('₹500')    == 'રૂપિયા પાંચ સો'
def test_rupee_gu():      assert _normalize_text('₹50,000') == 'રૂપિયા પચાસ હજાર'
def test_percent():       assert _normalize_text('50%')     == 'પચાસ ટકા'
def test_year_gu():       assert _normalize_text('૨૦૨૪')    == 'બે હજાર ચોવીસ'

def test_plain_gujarati_unchanged():
    assert _normalize_text('નમસ્તે') == 'નમસ્તે'

def test_mixed_sentence():
    result = _normalize_text('આ ₹1,250 ની કિંમત છે')
    assert 'એક હજાર બસ્સો પચાસ' in result


# ── _normalize_text — edge cases ─────────────────────────────────────────────

def test_decimal():       assert _normalize_text('3.14')        == 'ત્રણ પોઈન્ટ એક ચાર'
def test_decimal_gu():    assert _normalize_text('૩.૧૪')        == 'ત્રણ પોઈન્ટ એક ચાર'
def test_indian_lakh():   assert _normalize_text('1,00,000')    == 'એક લાખ'
def test_indian_crore():  assert _normalize_text('1,00,00,000') == 'એક કરોડ'
def test_phone():         assert _normalize_text('9876543210')  == 'નવ આઠ સાત છ પાંચ ચાર ત્રણ બે એક શૂન્ય'
def test_ord1():          assert _normalize_text('1st')         == 'એક'
def test_ord2():          assert _normalize_text('2nd')         == 'બે'
def test_ord3():          assert _normalize_text('3rd')         == 'ત્રણ'
def test_ord4():          assert _normalize_text('4th')         == 'ચાર'
def test_ord_upper():     assert _normalize_text('1ST')         == 'એક'


# ── _split_into_chunks ────────────────────────────────────────────────────────

def test_leading_comma_stripped():
    chunks = _split_into_chunks('ગયો?, મેં જાઉં.', 20)
    assert not any(c.startswith(',') for c in chunks)

def test_speakable_chunks_only():
    chunks = _split_into_chunks('Hello world. !!!', 20)
    assert all(re.search(r'[\w઀-૿]', c) for c in chunks)

def test_long_text_chunks_user_input():
    text = 'નમસ્તે! તમારું સ્વાગત છે. આજે તમારો દિવસ કેમ ગયો?'
    normalized = _normalize_text(text)
    chunks = _split_into_chunks(normalized, 20)
    assert len(chunks) >= 1
    for c in chunks:
        assert not c.startswith(',')
        assert re.search(r'[\w઀-૿]', c)


# ── Schema validation ─────────────────────────────────────────────────────────

def test_schema_rejects_whitespace():
    with pytest.raises(ValidationError):
        GenerateRequest(text='   ', description='x', speaker='Yash', language='Gujarati')

def test_schema_rejects_punctuation():
    with pytest.raises(ValidationError):
        GenerateRequest(text='!!!', description='x', speaker='Yash', language='Gujarati')

def test_schema_accepts_gujarati():
    req = GenerateRequest(text='નમસ્તે', description='x', speaker='Yash', language='Gujarati')
    assert req.text == 'નમસ્તે'

def test_schema_accepts_number_in_text():
    req = GenerateRequest(text='test ૧', description='x', speaker='Yash', language='Gujarati')
    assert req.text == 'test ૧'