"""Validation and text-handling tests for the aligner; they never load the model."""
import io
import json
import os
import sys
import unittest
import wave

import numpy as np

sys.path.insert(0, os.path.join(os.path.dirname(__file__), ".."))
from fastapi.testclient import TestClient  # noqa: E402

import app as aligner  # noqa: E402


def wav_bytes(seconds=1.0, sr=16000):
    buf = io.BytesIO()
    with wave.open(buf, "wb") as w:
        w.setnchannels(1); w.setsampwidth(2); w.setframerate(sr)
        w.writeframes((np.zeros(int(seconds * sr))).astype("<i2").tobytes())
    return buf.getvalue()


class NormalizeTests(unittest.TestCase):
    def test_romanizes_german(self):
        self.assertEqual(aligner.normalize_word("Größe,"), "grosse")
        self.assertEqual(aligner.normalize_word("Über"), "uber")
        self.assertEqual(aligner.normalize_word("2024."), "")
        self.assertEqual(aligner.normalize_word("don't"), "don't")

    def test_digits_get_the_gap_between_neighbours(self):
        fake = aligner.align_samples(np.zeros(16000, dtype=np.float32), ["2024", "3"])
        self.assertEqual([(a, ok) for a, _, ok in fake], [(0.0, False), (0.0, False)])
        self.assertAlmostEqual(fake[0][1], 1.0)


class ValidationTests(unittest.TestCase):
    def setUp(self):
        self.c = TestClient(aligner.app)

    def post(self, url="/align", audio=None, **data):
        files = {"audio": ("a.wav", audio if audio is not None else wav_bytes(), "audio/wav")}
        return self.c.post(url, files=files, data=data)

    def test_health_works_without_the_model(self):
        r = self.c.get("/health")
        self.assertEqual(r.status_code, 200)
        self.assertFalse(r.json()["model_loaded"])

    def test_bad_language_text_and_audio_are_rejected(self):
        self.assertEqual(self.post(text="Hallo Welt", language="../x").status_code, 422)
        self.assertEqual(self.post(text="   ", language="de").status_code, 422)
        self.assertEqual(self.post(text="1234 5678", language="de").status_code, 422)
        self.assertEqual(self.post(text="x" * (aligner.MAX_TEXT_CHARS + 1), language="de").status_code, 422)
        self.assertEqual(self.post(audio=b"not audio", text="Hallo", language="de").status_code, 415)

    def test_span_must_lie_inside_the_audio(self):
        self.assertEqual(self.post(text="Hallo", language="de", start="5", end="6").status_code, 422)
        self.assertEqual(self.post(text="Hallo", language="de", start="0.5", end="0.2").status_code, 422)

    def test_oversized_audio_is_rejected(self):
        old = aligner.MAX_AUDIO_BYTES
        aligner.MAX_AUDIO_BYTES = 100
        try:
            self.assertEqual(self.post(text="Hallo", language="de").status_code, 413)
        finally:
            aligner.MAX_AUDIO_BYTES = old

    def test_batch_validates_spans(self):
        self.assertEqual(self.post("/align/batch", spans="nonsense", language="de").status_code, 422)
        self.assertEqual(self.post("/align/batch", spans="[]", language="de").status_code, 422)
        self.assertEqual(self.post("/align/batch", spans=json.dumps([{"start": 0, "end": 1}]), language="de").status_code, 422)
        many = json.dumps([{"start": 0, "end": 1, "text": "a"}] * (aligner.MAX_SPANS + 1))
        self.assertEqual(self.post("/align/batch", spans=many, language="de").status_code, 422)

    def test_letterless_sentence_in_a_batch_needs_no_model(self):
        # no letters: handled by gap interpolation, but the model still loads first; only check validation passes
        r = self.post("/align/batch", spans=json.dumps([{"start": 0, "end": 1, "text": "2024"}]), language="de")
        self.assertNotEqual(r.status_code, 422)


if __name__ == "__main__":
    unittest.main()