feat: transcribe Telegram voice messages via APInex

tests/test_voice_stt.py

@@ -0,0 +1,73 @@
+import unittest
+from unittest.mock import patch
+
+import bot
+
+
+class FakeResponse:
+    def __init__(self, status, payload):
+        self.status = status
+        self.payload = payload
+
+    async def __aenter__(self):
+        return self
+
+    async def __aexit__(self, exc_type, exc, tb):
+        return False
+
+    async def json(self, content_type=None):
+        return self.payload
+
+
+class FakeClientSession:
+    def __init__(self, response, **kwargs):
+        self.response = response
+        self.kwargs = kwargs
+        self.request: tuple[str, dict] = ("", {})
+
+    async def __aenter__(self):
+        return self
+
+    async def __aexit__(self, exc_type, exc, tb):
+        return False
+
+    def post(self, url, **kwargs):
+        self.request = (url, kwargs)
+        return self.response
+
+
+class VoiceSttTests(unittest.IsolatedAsyncioTestCase):
+    def test_extract_transcript(self):
+        payload = {"results": {"channels": [{"alternatives": [{"transcript": "  Привет, мир!  "}]}]}}
+        self.assertEqual(bot.extract_transcript(payload), "Привет, мир!")
+
+    def test_empty_response(self):
+        self.assertEqual(bot.extract_transcript({"results": {"channels": []}}), "")
+        self.assertEqual(bot.extract_transcript({}), "")
+
+    def test_split_long_transcript(self):
+        chunks = bot.split_text("слово " * 2000, 4000)
+        self.assertTrue(all(len(chunk) <= 4000 for chunk in chunks))
+        self.assertEqual(" ".join(chunks).split(), ("слово " * 2000).split())
+
+    async def test_transcribe_sends_raw_ogg_and_parses_response(self):
+        response = FakeResponse(200, {"results": {"channels": [{"alternatives": [{"transcript": "Распознано"}]}]}})
+        fake_session = FakeClientSession(response)
+        with patch.object(bot.aiohttp, "ClientSession", return_value=fake_session):
+            result = await bot.transcribe_audio(b"ogg-data", "audio/ogg", "not-logged")
+        self.assertEqual(result, "Распознано")
+        url, kwargs = fake_session.request
+        self.assertEqual(url, bot.APINEX_STT_URL)
+        self.assertEqual(kwargs["data"], b"ogg-data")
+        self.assertEqual(kwargs["headers"]["Content-Type"], "audio/ogg")
+        self.assertEqual(kwargs["headers"]["Authorization"], "Bearer not-logged")
+
+    async def test_transcribe_raises_on_api_error(self):
+        fake_session = FakeClientSession(FakeResponse(401, {}))
+        with patch.object(bot.aiohttp, "ClientSession", return_value=fake_session):
+            with self.assertRaises(RuntimeError):
+                await bot.transcribe_audio(b"ogg-data", "audio/ogg", "not-logged")
+
+
+if __name__ == "__main__":
+    unittest.main()