test: cover transcript state behavior

Author: Diyoncrz18

tests/test_audio_transcriber.py

@@ -0,0 +1,74 @@
+import time
+import unittest
+import sys
+import types
+from datetime import datetime
+
+sys.modules.setdefault("pyaudiowpatch", types.SimpleNamespace(paInt16=8, PyAudio=object))
+
+from AudioTranscriber import AudioTranscriber, TRANSLATION_SILENCE_DELAY
+
+
+class FakeSource:
+    SAMPLE_RATE = 16000
+    SAMPLE_WIDTH = 2
+    channels = 1
+
+
+class FakeModel:
+    def translate_to_indonesian(self, text):
+        return f"ID: {text}"
+
+
+class AudioTranscriberTests(unittest.TestCase):
+    def make_transcriber(self):
+        return AudioTranscriber(FakeSource(), FakeSource(), FakeModel())
+
+    def wait_for_entry(self, transcriber, predicate, timeout=1.0):
+        deadline = time.monotonic() + timeout
+        while time.monotonic() < deadline:
+            entries = transcriber.get_entries()
+            if entries and predicate(entries[0]):
+                return entries[0]
+            time.sleep(0.01)
+        self.fail("Timed out waiting for transcript entry state")
+
+    def test_merge_transcript_keeps_overlapping_words_once(self):
+        merged = AudioTranscriber._merge_transcript(
+            "hello this is a live transcription",
+            "live transcription test",
+        )
+
+        self.assertEqual(merged, "hello this is a live transcription test")
+
+    def test_finalize_segment_translates_after_silence(self):
+        transcriber = self.make_transcriber()
+        transcriber.update_transcript("Speaker", "hello world", datetime.now())
+
+        with transcriber.lock:
+            segment = transcriber.segment_state["Speaker"]
+            segment["last_activity"] = time.monotonic() - TRANSLATION_SILENCE_DELAY - 0.1
+            generation = segment["timer_generation"]
+
+        transcriber.finalize_segment("Speaker", generation)
+
+        entry = self.wait_for_entry(
+            transcriber,
+            lambda item: not item["translation_pending"],
+        )
+        self.assertEqual(entry["translation"], "ID: hello world")
+        self.assertEqual(entry["translation_state"], "done")
+
+    def test_clear_transcript_resets_public_buffers(self):
+        transcriber = self.make_transcriber()
+        transcriber.update_transcript("You", "testing one two", datetime.now())
+
+        transcriber.clear_transcript_data()
+
+        self.assertEqual(transcriber.get_entries(), [])
+        self.assertEqual(transcriber.mic_buffer, "")
+        self.assertEqual(transcriber.speaker_buffer, "")
+
+
+if __name__ == "__main__":
+    unittest.main()